EPA530-D-03-001e
July 2003
SAB Review Draft
Multimedia, Multipathway, and
Multireceptor Risk Assessment
(3MRA) Modeling System
Volume V: Technology Design and
User's Guide
-------
EPA530-D-03-001 e
July 2003
SAB Review Draft
Multimedia, Multipathway, and
Multireceptor Risk Assessment
(3MRA) Modeling System
Volume V: Technology Design and
User's Guide
prepared by
U.S. Environmental Protection Agency
Office of Research and Development Office of Solid Waste
National Exposure Research Laboratory Washington, D.C.
Athens, GA
-------
This document is the fifth volume of a five-volume set. Volume I describes the conceptual design,
scientific rationale, and supporting data that are the foundation for the 3MRA modeling system. This
volume describes the data developed and used to run the 3MRA modeling system. Volume III describes
the approach to quality assurance, including verification and validation activities ranging from extensive
peer reviews to multimedia model comparisons. Volume IV describes the methodology used to evaluate
sensitivity of model parameters and characterize different types of uncertainty in the 3MRA modeling
system.
-------
Table of Contents
1.0 Introduction 1-1
2.0 3MRA Methodology : A Summary Description 2-1
2.1 Site-based Modeling Approach 2-1
2.1.1 Conceptual Modeling Approach 2-1
2.1.2 3MRA Data 2-7
2.1.3 3MRA Site Modeling Outputs 2-8
2.2 National Protection Measures 2-14
2.3 3MRA Monte Carlo Scheme to Quantify Uncertainty 2-14
3.0 3MRA Technology 2-21
3.1 System Requirements 2-21
3.1.1 Transcription of 3MRA Methodology into
Technology Requirements 2-21
3.1.2 Programmatic Requirements 3-2
3.1.3 Software Environment and Software Engineering Requirements . 3-2
3.2 System Design 3-4
3.2.1 Science Modules 3-5
3.2.2 System Databases 3-7
3.2.3 System Processors 3-7
3.2.4 System Data Files 3-8
3.2.5 Programming Standards and System Utilities 3-9
3.2.6 Integrated Functionality of 3MRA Modeling System 3-10
3.3 Component Design 3-11
3.3.1 Data Representation Standard for the 3MRA Modeling System .3-11
3.3.2 Input/Output Dynamic Linked Library (IOdll) 3-15
3.3.3 Monte Carlo Sampling Dynamic Linked Library (MCdll) 3-17
3.3.4 Chemical Properties Processor (CPP) 3-18
3.3.5 3MRA Databases and Data Files 3-20
3.3.5.1 3MRA Input Data 3-20
3.3.5.2 3MRA Intermediate Data Files 3-22
3.3.5.3 3MRA Output Data 3-23
3.3.6 The System User Interface (SUI) 3-24
3.3.6.1 User Interaction Features 3-24
3.3.6.2 Execution Management Features 3-27
3.3.7 The Site Definition Processor (SDP) 3-27
3.3.8 Multimedia Multipathway Simulation Processor (MMSP) 3-29
3.3.8.1 MMSP Science Module Design Features 3-30
3.3.8.1.1 Science Module Execution 3-31
3.3.8.1.2 Science Module Output 3-32
3.3.9 The Exit Level Processors (ELP I, ELP II, RVP) 3-32
3.3.9.1 Exit Level Processor I (ELP I) 3-33
3.3.9.2 Exit Level Processor II (ELP I) and Risk Visualization
Processor (RVP) 3-38
-------
4.0 3MRA Installation and User's Guide 4-1
4.1 Installing the 3MRA Modeling System 4-1
4.1.1 Installation Steps 4-1
4.1.2 Directory Structure for 3MRA 4-11
4.2 Executing the 3MRA Modeling System 4-21
4.2.1 System User Interface (SUI) 4-22
4.2.2 Invoking the 3MRA Modeling System 4-25
4.2.2.1 System Configuration 4-26
4.2.3 System Management 4-31
4.2.4 System Status 4-34
4.3 Post Simulation Analysis 4-36
4.3.1 Analysis of National Assessment Results (ELP II/RVP) 4-35
4.3.2 Analysis of Site Assessment Results 4-41
4.3.2.1 Site Visualization Tool (SVT) 4-42
-------
List of Figures
Figure 2.1 .a 3MRA Site Layout Example of Human Population Distribution and Extent of
Modeling Area of Interest 2-4
Figure 2.1 .b 3MRA Site Layout Example of Watersheds and Surface Waters 2-4
Figure 2. l.c 3MRA Site Layout Example of Ecological Habitats and Farms 2-5
Figure 2.l.d Integrated 3MRA Site Layout 2-5
Figure 2.2 Relationship Between Exposure Concentration and Pathway Risk 2-10
Figure 2.3 Nf x Ni Pathway Risk Matix Output 2-17
Figure 2.3.a Probability that percent protection is less than P for a given waste concentration
and target risk level 2-18
Figure 2.3.b Percent of receptors protected for different risk levels and Cw=10"3 for N; Monte-
Carlo iterations 2-19
Figure 2.4 Percent of receptors protected for different waste concentrations
and risk levels 2-20
Figure 3.1 3MRA National Assessment High Level Algorithm 3-4
Figure 3.2 3MRA System Design 3-5
Figure 3.3 3MRA Science Modules and Related Connectivity 3-6
Figure 3.4 3MRADictionary Files 3-12
Figure 3.5 3MRA SUI System Configuration Screen 3-25
Figure 3.6 3MRA SUI Systems Management: Selections Screen 3-25
Figure 3.7 3MRA SUI System Management: Options Screen 3-26
Figure 3.8 3MRA SUI System Status Screen 3-27
Figure 3.9 SDP Processing Steps 3-28
Figure 3.10 Details of the Multimedia Multipathway Simulation Processor 3-29
Figure 3.11 Interactions Within the Multimedia Multipathway Simulation Processor ... 3-31
Figure 3.12 How Legacy Models Connect with the FRAMES-HWIR
Technology Software System 3-32
Figure 3.13 Relationship among ELP I, ELP II, and RVP 3-39
Figure 3.14 Protective Summary Output Figure for Human Risk 3-41
Figure 3.15 Protective Summary Output Figures for Ecological HQ 3-42
Figure 4.1 Opening Screen of 3MRA Installation Program 4-1
Figure 4.2 3MRA Installation Welcome Screen 4-2
Figure 4.3 3MRA Licensing Agreement 4-3
Figure 4.4 Description of 3MRA Installation Options 4-3
Figure 4.5 3MRA Installation Selection Screen : Typical 4-4
Figure 4.6 3MRA Installation Selection Screen : Custom 4-4
Figure 4.7 3MRA Installation Options for Database Connectivity Tools 4-5
Figure 4.8 3MRA Installation Target Directory 4-5
Figure 4.9 3MRA "Installation in Progress" Screen 4-6
Figure 4.10 Java Runtime Environment Setup Screen 4-7
Figure 4.11 Java Runtime Environment Installation Screen 4-8
Figure 4.12 Java Installation Target Directory Screen 4-8
Figure 4.13 Java Runtime Browser Selection Screen 4-9
Figure 4.14 3MRA Final Installation Screen 4-10
Figure 4.15 3 MRA Modify /Repair/Remove Installation Screen 4-10
-------
Figure 4.16 3MRA Directory Structure after Installation 4-11
Figure 4.17 3MRA Directory Structure : CPPData Subdirectory 4-12
Figure 4.18 3MRA Directory Structure : Database Subdirectory 4-13
Figure 4.19 3MRA Directory Structure : ExampleOutputs Subdirectory 4-14
Figure 4.20 3MRA Directory Structure : GRF subdirectory 4-15
Figure 4.21 3MRA Directory Structure : MetData Subdirectory 4-16
Figure 4.22 3MRA Directory Structure : PSOF subdirectory 4-17
Figure 4.23 3MRA Directory Structure : RSOF subdirectory 4-18
Figure 4.24 3MRA Directory Structure : SSF subdirectory 4-19
Figure 4.25 3MRA Directory Structure : SVT subdirectory 4-20
Figure 4.26 User Interaction Features of 3MRA Menu Screens 4-23
Figure 4.27 3MRA Directory Listing Displaying Batch File for System Cleanup 4-24
Figure 4.28 Initial 3MRA User Interface Screen 4-25
Figure 4.29 3MRA Header File Selection Screen 4-25
Figure 4.30 3MRA SUI : System Configuration Screen 4-27
Figure 4.31 3MRA SUI: Directories Configuration Screen 4-28
Figure 4.32 3MRA SUI : Processor Configuration Screen 4-29
Figure 4.33 3MRA SUI: MMSP Configuration Screen 4-31
Figure 4.34 3MRA SUI: System Management Selections Screen 4-32
Figure 4.35 3MRA SUI : System Management Options Screen 4-33
Figure 4.36 3MRA SUI : System Status Screen 4-34
Figure 4.37 3MRA SUI: Selection Screen for National Assessment of Benzene 4-37
Figure 4.38 Initial ELP-II Screen 4-38
Figure 4.39 EPL II Options Screen 4-38
Figure 4.40 ELP II Human Risk Protective Summary Screen 4-39
Figure 4.41 ELP II Ecological HQ Protective Summary Screen 4-40
Figure 4.42 ELP II Tabular Output Selection Screen 4-41
Figure 4.43 3MRA Selection Screen for Site Example 4-42
Figure 4.44 SVT User Interface Screen 4-43
Figure 4.45 Example SVT Output 4-44
Figure 4.46 Sample SVT Soil Concentration Graph 4-45
-------
List of Tables
Table 2.1 Dimensions of 3MRA Site-based Risk Assessment 2-6
Table 2.2 3MRA Data Requirements and Sources 2-7
Table 2.3 Summary of Human Risk Bins 2-12
Table 2.4 Summary of Human Hazard Quotient (HQ) Bins 2-12
Table 2.5 Summary of Human Risk Module Output Dimensions Associated
with Risk Bins 2-13
Table 2.6 Summary of Ecological Hazard Quotient (Eco HQ) bins
Table 3.1 Example 3MRA Data File : Watershed Site Simulation File 3-13
Table 3.2 Example 3MRA DIC File for Watershed Site Simulation File 3-14
Table 3.3 3 MR A Mill Data 3-21
Table 3.4. Summary of Parameter Requirements Associated with Human-Health
Risk/Hazard for the ELP-I 3-34
Table 3.5. Summary of Human Risk Bins and Labels 3-35
Table 3.6. Procedure to Compute a Risk Summary Output File 3-36
Table 3.7 Example ELP I Output Table 3-37
Table 3.8 3MRA Regulatory Scenarios 3-43
Table 3.9 Example Lowest Target Exit level Concentrations 3-44
Table 3.10 Example Target Exit level Concentrations by Scenario 3-44
Table 3.11 Example Relative Target Exit Level Concentrations 3-44
Table 3.12 Example Target Exit Level Concentrations, Based on 50% Probability
of Protection 3-45
Table 3.13 Example Cohort Human Risk/HQ for Landfill 3-45
Table 3.14 Example Receptor Human Risk/HQ for WP 3-45
Table 3.15 Exposure Pathway Human Risk/HQ 3-46
-------
Section 1.0
Introduction
1.0 Introduction
A modeling-based methodology entitled "A Framework for Finite-Source Multi-media,
Multi-pathway, and Multi-receptor Risk Assessment (3MRA)" [1] has been developed with the
goal of facilitating the establishment of national regulatory limits related to "safe" constituent
concentration levels in wastestreams entering land-based solid waste management units. The
3MRA methodology is a conceptual approach to "site-based" regulatory risk assessment
problems. Site-based regulatory problems and assessments, in this context, refer to national
scale regulatory decisions that are based on the results of modeling risk at individual sites. The
3MRA methodology calls for the application of a site-based multimedia risk assessment at a
statistically-sampled number of waste management sites across the nation. The national risks are
expressed as a quantitative relationship between constituent concentrations in waste streams and
the percentage of human and ecological receptors that are "protected". This relationship can be
established as a function of various regulatory decision scenarios. For example, decision
analysts may wish to know what the relationship between waste concentrations and
protectiveness is for a particular sensitive receptor group (e.g., children) or for specific
categories of waste management units (e.g., landfills). The 3MRA methodology provides a
means by which to "roll-up" site-based risk results to provide decision makers multiple views of
the national risk picture.
Because the process of establishing such a relationship between cause and effect includes
numerous uncertainties, the 3MRA methodology was also designed to generate estimates of
uncertainty related to the estimates of national protection. Specifically, the methodology
provides assessment procedures for the characterization and separation of natural variability
(irreducible uncertainty) and uncertainty due to errors principally resulting from a lack of
knowledge (reducible uncertainty - whether it be due to errors in measurement, sampling,
model, site conceptualization, etc.).
In order to inform national regulatory decisions the 3MRA methodology has been
expressed in the form of a facilitating technology. The 3MRA technology is an integrated
environmental modeling system consisting of a "modeling domain", an "assessment domain",
and a facilitating software infrastructure. The modeling domain includes a comprehensive set of
science-based models and databases. The models are designed to simulate all aspects of a site-
based human and ecological risk assessment (i.e., source release, multimedia fate and transport,
aquatic and terrestrial foodweb dynamics, and human and ecological exposure and risk). The
databases include data descriptive of waste disposal unit operations, environmental conditions,
chemical properties, and exposure and risk factors. The assessment domain envelopes the
modeling domain and provides a strategy for applying the models and data components to solve
specific regulatory problems. The software infrastructure supports the development and
application of the tools contained in the assessment and modeling domains and consists of 1) a
user interface, 2) data representation and transfer standards and facilitating software, 3)
execution management software, and 4) several software utilities. These elements of the 3MRA
technology are integrated in such a manner as to facilitate the intended national environmental
risk assessment.
The purpose of this document is to provide an overview of the 3MRA software-based
modeling technology designed to automate the 3MRA national risk assessment methodology.
The 3MRA modeling technology is described in terms of its required functionality, integrated
systems design, and user operation. The "science" of the multi-media modeling and the
1-1
-------
statistical theory driving the national assessment strategy are described only to the extent to
make clear the connection between methodology and technology. The presentation is intended
to provide regulatory analysts (who may apply the technology to risk-based regulatory problems)
and model developers (who may either incorporate their models within the modeling system or
apply the system in the conduct of research) sufficient information to make a determination of
the modeling systems applicability to their work.
A summary description of the 3MRA methodology is presented in Section 2 to provide
the basis for describing the technology. The 3MRA technology design is presented in Section 3
(first in terms of set of system requirements [Section 3.1], then a description of the overall
system design for the technology [Section 3.2], and finally a more detailed description of each of
the major components of the 3MRA technology system [Section 3.3]). Section 4 presents a
User's Guide for installation and operation of the technology.
1-2
-------
A note about 3MRA software technologies
Currently there exist three technologies related to 3MRA.
3MRA Version 1.0 :
This technology represents the modeling system designed to execute national
site-based risk assessments.
This technology is the sole subject of this document and is fully available to the
public.
3MRA Version 1.x:
This technology is an extension of 3MRA Version 1.0 that includes software
specifically designed to facilitate 1) the execution of 3MRA Version 1.0 on a
network of 160 Personal Computers linked together to form a Super computer
capability, and 2) the execution of uncertainty analysis and sensitivity analysis
studies using 3MRA Version 1.0.
This technology is currently operational in a research context. It is being
applied to assess the uncertainties and sensitivities in the application of 3MRA
Version 1.0.
3MRA Version 2.0 :
This technology represents an extension of 3MRA Version 1.0 to facilitate the
execution of site-specific risk assessments and to advance the design of the
underlying software infrastructure. The science-based goal of this technology
is to contain the science models and databases needed to conduct both national
and site-specific assessments within the same modeling system. Achieving this
goal is intended to facilitate the assimilation of new science into modeling
technologies that can be used to conduct regulatory-based risk assessments and
to establish consistency across regulatory programs.
This technology is coming online, in Beta release form, during the summer of
2003.
Finally, the infrastructure for each of the 3MRA technologies is based on the Framework for Risk
Analysis in Multimedia Enviromnental Systems (FRAMES). FRAMES represents a collaborative effort
among four Federal Agencies (EPA, DoE, DoD, and NRC) to establish a common modeling infrastructure
for conducting human and ecological risk assessments. The goal of the FRAMES effort is to facilitate
scientific collaboration among the Agencies and to maximize community access to the collective set of
models, databases, and data analysis tools developed and applied by the Agencies. Thus, the reader may
see references to the following technologies that are synonymous with those listed above.
FRAMES 3MRA Version 1.0
FRAMES 3MRA Version 1.x
FRAMES 3MRA Version 2.0
1-3
-------
Section 2.0
3MRA Methodology
2.0 3MRA Methodology : A Summary
Description
This section provides a description of the essential elements of the 3MRA methodology
as they relate to developing a design for the facilitating software system. Detailed descriptions
of the 3MRA methodology are presented in "A Framework for Finite-Source Multi-media,
Multi-pathway, and Multi-receptor Risk Assessment (3MRA)" [1] .
The 3MRA national assessment methodology is a screening-level risk-based assessment
of potential human and ecological health risks resulting from long-term (chronic) exposure to
chemicals released from land-based waste management units (WMUs). The assessment is
national in scale and site-based, that is, risks are assessed at individual sites across the U.S. and
rolled-up to represent a national distribution of risks. The resulting national distribution of risks
forms the basis for determining wastestream constituent concentrations that satisfy regulatory
criteria that are based on the percentage of nationwide receptors and sites that are "protective".
Protective, in this context, means that receptors (human and ecological) do not experience health
risks or hazards greater than those established by Agency policy (e.g., excess cancer risk of 10"6)
The following sections describe the 3MRA assessment methodology in a manner that
leads directly to a statement of requirements for a technology design. First, a brief description of
the site-based risk assessment, including the conceptual modeling approach, data requirements,
and risk outputs, is presented in Section 2.1. Section 2.2 describes the manner in which
expressions of risk at the site level are accumulated and stored in a database that can be queried
to provide expressions of national protection. Finally, Section 2.3 describes the essential
features of the Monte Carlo-based approach, including the general algorithm, that facilitates the
probabilistic applications of the 3MRA modeling tools and provides a means for quantifying and
separating variability and uncertainty. With this background information the 3MRA national
assessment technology is presented in Section 3.
2.1 Site-based Modeling Approach
At the core of the 3MRA methodology is the assessment of human and ecological risks at
a statistically derived sample of sites across the U.S. These risks are estimated using an
integrated multimedia modeling approach. Described in Sections 2.1.1, 2.1.2, and 2.1.3 are,
respectively, the conceptual modeling approach for conducting site-based human and ecological
risk assessments, the modeling input data, and the modeling outputs.
2.1.1 Conceptual Modeling Approach
Figure 2.1 illustrates the conceptual layout of a typical 3MRA site where exposures and
related health risks are to be modeled. Figure 2.1a illustrates that the geographic center of a site,
for modeling purposes, is the waste management unit (WMU). The geographic extent of the
modeling "area of interest" (AOI) is bounded by a circle whose radius extends from the edge of
the source outward 2 kilometers. This extent is a function of a modeling assumption that states
that the peak, and in a cumulative sense, the most significant portion of the risk resulting from
chemical releases from the WMU, occur within 2 kilometers of the source. This geographic
2-1
-------
Section 2.0
3MRA Methodology
extent is not a limitation of the 3MRA modeling system. Also shown in Figure 2. la is the
conceptual view of how the human population distribution in the AOI is assigned. U.S. Census
data is used to locate "Census Block" centroids within the AOI. Block group populations,
characterized by age cohorts, are assigned a resident location at the centroid of the block group.
Thus, for purposes of exposure and risk all receptors within the block group experience the same
exposure concentrations. Further, the 3MRA assumes the population will be present throughout
the duration of a site simulation (which may be on the order of hundreds or thousands of years).
3MRA employs the concept of generational cohorts which assumes that a each receptor lifetime
is followed by a series of identical receptors until the end of the simulation. Finally, Figure 2.1a
includes two additional concentric rings at 0.5 kilometers and 1.0 kilometers, respectively.
These rings define distances for aggregating exposure and risks across receptors, thus providing
decision analysts a risk vs distance from source perspective.
Figure 2. lb illustrates the conceptualization of watersheds and surface waters for 3MRA.
Within the AOI watersheds are delineated using GIS software. There is no limit to the number
of watershed sub-basins that can be modeled in 3MRA. Each watershed sub-basin is assigned to
specific surface waters for purpose of routing runoff and erosive fluxes. The surface waters
within the AOI may include streams, ponds, lakes, and wetlands. Inter-connected surface waters
form a "waterbody network" and there may be multiple water networks within the AOI. Finally,
Figure 2. lb includes "local watershed sub-basins" that represent the land area between the WMU
and the surface water segment receiving the source runoff. This area is specifically modeled in
3MRA.
Figure 2.1.c illustrates the conceptualization of ecological habitats within the AOI.
Habitats are delineated using GIS-based maps displaying landuse and ecological regions.
Individual specie home ranges are randomly assigned in a manner that is consistent with predator
prey relationships among the habitat species. Related to habitats are foodwebs that involve both
plants and animals and associated diets. 3MRA habitat types include several terrestrial and
aquatic environments. Also shown in Figure 2.1c are farms where crops may be exposed and
result in exposure to humans via the food chain.
Figure 2. Id illustrates an integrated view of the site layout features described above.
This is shown to make the point that in reality these features seamlessly overlap and connect.
That is, for example, habitats overalp watersheds that drain into both the sub-surface and surface
waters. In 3MRA, all such connections are explicitly assigned with appropriate modeling of
intermedia fluxes.
Not shown in Figure 2.1 are the atmospheric and groundwater media included in 3MRA.
Atmospheric fate and transport of chemicals released from the WMU is based on meteorological
data associated with a regional weather station. Subsurface components of the site layout
include a vadose zone directly beneath the WMU and a regional aquifer at a uniform depth and
flow direction.
Table 2.1 lists the "dimensions" of modeling associated with the simulation of the
movement of chemical through each of the media of the 3MRA site layout. The dimensions
reflect the collection of physical/chemical/biological processes that are modeled in an attempt to
characterize the release, fate, transport, exposure, and risk associated with waste disposal. The
general steps in the site-based modeling assessment are as follows:
2-2
-------
Section 2.0
3MRA Methodology
1) Simulate the loading of wastestreams to land-based waste management unit
(WMU) over the lifetime of the WMU (including surface impoundments,
landfills, land application units, waste piles, and aerated tanks).
2) Simulate the release of chemical from the WMU to air (volatilization, particle re-
entrainment), vadose zone (leaching), groundwater(leaching), watersheds and
surface waters (overland runoff/erosion).
3) Simulate the fate and transport of chemical in and between major environmental
media (air, watershed soils, vadose zone, groundwater, surface water, and
sediments).
4) Simulate movement of chemical through the farm foodchain and aquatic and
terrestrial foodwebs.
5) Simulate human and ecological exposure via selected pathways (for human
receptors the pathways include air inhalation, shower air inhalation, groundwater
ingestion, soil ingestion, produce ingestion, beef ingestion, milk ingestion, fish
ingestion, and breast milk ingestion for infants).
6) Estimate human and ecological risk per receptor per pathway.
7) Repeat this sequence for each of a series of waste concentrations (Cw) to
establish a quantitative relationship between Cw and risk/hazard.
To execute this series of steps 3MRA utilizes a collection of seventeen science modules,
each simulating a self bounding component of the integrated system. Figure 2.2 identifies the
modules and illustrates their relative position in the 3MRA-based sequential execution of the
steps listed above.
2-3
-------
Section 2.0
3MRA Methodology
2 Kilometer
1 Kilometer
£3 jO.5^ilometer
13
Source
S20
Block Group 2
Block Gri
Figure 2.1 .a 3MRA Site Layout Example of Human Population
Distribution and Extent of Modeling Area of Interest
\^foterehed!j*to-basiii|$ Watershed
I Surface Wkter
S^WBNReach 3 J SOUIVC
Source
\^foterehed Sub-basin 4
LAKE
Figure 2.1 .b 3MRA Site Layout Example of Watersheds and Surface
Waters
2-4
-------
Section 2.0
3MRA Methodology
Habitat 2
Range 2
ange 1
Figure 2.1 .c 3MRA Site Layout Example of Ecological Habitats and Farms
bitat2
Watershed Sub-llasin
WBN Rch 2
N,
WBNXch 1
Watershed Su
b-basin 1
Source
Watershed Sub-basin
£16
Farm 1
Firm 2
\ _
Watershel
WBN Rch 3
ISub-basin 3
lajjii
Watershed
Surface Water
Habitat
Ring
Farm
J Human Receptor
~M Source
LAKE
inge 3 WBN Rch 4
Figure 2.1 .d Integrated 3MRA Site Layout
2-5
-------
Section 2.0
3MRA Methodology
Table 2.1 Dimensions of 3MRA Site-based Risk Assessment
CONTAMINANTS
INTERMEDIA CONTAMINANT FLUXES
Organics (approx. 200)
Source -> Air (vol, resuspension)
Metals (20)
Source -> Vadose zone (leaching)
Source Surface soil -> Local Watershed Soil (erosion, runoff)
SOURCE TYPES
Air -> Watershed/Farm /Habitat Soil
Landfill
(wet/dry dep)
Land Application Unit
Air -> Surface water (wet/dry dep)
Surface Impoundment
Air -> Vegetation (dep/uptake)
Aerated Tank
Farm/Habitat Soil -> Vegetation (root uptake)
Waste Pile
Watershed Soil -> Surface water (erosion, runoff)
Surface water -> Aquatic organisms (uptake)
SOURCE TERM
Surface water -> Sediment (sedimentation)
CHARACTERISTICS
Vadose zone -> Groundwater (percolation)
Mass Balance
Groundwater -> Surface water
Multimedia Partitioning
Soil -> Vegetation (uptake, dep)
Chemical Decay
Vegetation, Soil, Water -> Beef and dairy (uptake)
SOURCE RELEASE MECHANISMS
FOODCHAIN
Erosion
Human (Farm)
Volatilization
Human (Aquatic)
Runoff
Ecological (Aquatic Habitat)
Leaching
Ecological (Terrestrial Habitat)
Particle Resuspension
RECEPTORS
TRANSPORT MEDIA
Human
Atmosphere
Resident (Adult & Child)
Soil
Beef Farmer (Adult & Child)
Vadose zone
Dairy Farmer (Adult & Child)
Saturated zone
Home Gardener (Adult & Child)
Surface water
Recreational Fisher (Adult & Child)
FATE PROCESSES
Ecological
Chemical/Biological Transformation
Mammals, Birds, Soil Communities, Terrestrial Plants,
(and associated products of
Aquatic Communities, Benthic Communities, Aquatic Plants,
transformation)
Amphibians, and Reptiles.
Linear partitioning (water/air,
water/soil, air/plant,
EXPOSURE ROUTES/PATHWAYS
water/biota)
Ingestion (plant, meat, milk, aquatic food, water, soil)
Nonlinear partitioning (metals in
Inhalation (gases, particulates)
vadose zone)
Direct Contact (soil, water)
Chemical Reaction/Speciation
HUMAN AND ECOLOGICAL RISK ENDPOINTS
AGE GROUPS FOR HUMAN
Human Cancer Risk
RECEPTORS
Human Noncancer Hazard Quotient
Infant < 1 year
Ecological Population and Community Hazard Quotients
Child-a 1- 5 years
Child-b 6-11 years
Child-c 12-19 years
Adult 20+ years
2-6
-------
Section 2.0
3MRA Methodology
2.1.2 3MRA Data
The data requirements for 3MRA modeling are substantial. The primary categories of
input data are listed in Table 2.2 and include site data (layout and environmental), human and
ecological exposure data, chemical data, and meteorological data. The 3MRA methodology calls
for the use of "site-based" data, meaning that, to the extent practicable, data used in the modeling
is to be directly reflective of the 3MRA sampled-sites from across the U.S. Because of several
factors, including the lack of availability of various data at specific sites, resource limitations
associated with collecting the data, and the screening level nature of the modeling approach, not
all data is site-specific. Lacking site-specific data, statistical distributions of data values within
the geographic region containing the site is accessed and sampled. The resulting value is
assigned to the site. Further, when a regional source of data is unavailable, a national scale
statistical distribution of the variable sampled and assigned to the site. In all, several hundred
variables are required to model any given site. Table 2.2 lists the categories of data required for
3MRA and the source of the data, i.e., site-specific, regional, national databases, or a
combination of sources.
Included in the 3MRA database containing site data are 201 individual site locations
involving a total of 419 site/WMU combinations. Each site contains one or more of the WMUs
but no site contains all five unit types.
Table 2.2 3MRA Data Requirements and Sources
Data Type
Data Representation
Site-based
Regional
National
Site Layout Data
Waste Management Unit
•
•
Watershed and waterbody layout
•
Human receptor characteristics and location
•
•
Ecological habitat type, receptors, and
location
•
•
Site Environmental Data
Waste properties
•
Atmosphere
•
Surface water
•
•
Soil/vadose zone
•
•
Aquifer
•
•
•
Farm food chain/terrestrial food web
•
2-7
-------
Section 2.0
3MRA Methodology
Aquatic food web
•
•
Human and Ecological Exposure/Risk Data
Human exposure factors
•
Ecological exposure factors
•
•
Risk and control variables
•
Meteorological Data
•
*
Chemical Data
Physical properties
•
Biouptake/bioaccumulation factors
•
Chemical/Biological Transformation Rates
•
Human health benchmarks
•
Ecological benchmarks
•
* The chemical data is labeled under National to imply that the same data is applied to all
sites.
2.1.3 3MRA Site Modeling Outputs
As stated previously the objective of the 3MRA site-based modeling is to estimate the
annual average risk (and/or hazard quotient) for human and ecological receptors residing within
the area of interest surrounding a waste management unit at a site. To arrive at this endpoint the
3MRA site modeling generates the following outputs for each year of simulation :
1) Source Release Chemical Fluxes
• air (volatilization, particle re-entreinment)
• watershed (erosion, runoff)
• sub-surface (leaching)
2) Inter-media Chemical Fluxes
• air to surface soil
• surface soil to vadose zone
• vadose zone to aquifer
• aquifer to surface water
• surface soil to surface water
3) Media Chemical Concentrations at Exposure Locations
• air
• water
• soil
• biota (crop, plant, prey)
4) Receptor Exposures per Pathway
2-8
-------
Section 2.0
3MRA Methodology
• Human (Inhalation Route)
- ambient air
- shower air
• Human (Ingestion Route)
- soil
- water
- crop
- beef
- milk
- fish
- breast milk (infants)
• Ecological
- Ingestion
~ media (soil)
~ plant
~ prey
5) Receptor Health Effects
• Carcinogenic (Human)
• Hazard Quotient (Human)
• Hazard Quotient (Ecological)
Each of the above outputs are reported on an annual basis for the duration of a
simulation, which can be up to 10,000 years. Figure 2.2 illustrates how the primary 3MRA
outputs , i.e., risks/hazard quotients (HQs), are computed based on exposure concentrations and
exposure durations. Risks/HQs are computed for each exposure period (duration of exposure
associated with either carcinogenic risk or hazard quotient). A time series of risks/HQs is
generated for each receptor type/cohort combination, at each location where receptors reside
(e.g., U.S. Census Block centroid), for each exposure pathway (involving each combination of
exposure route and contact medium). Risks and HQs time series are also identified with the
combination of chemical, waste management unit type, wastestream concentration level, site,
and exposure area (i.e., defined by distance from source). These indices of risk/HQ are
maintained in order to allow the decision analyst to accumulate national risk according to
different regulatory scenarios. A regulatory scenario includes an endpoint (e.g., chemical
concentration in a wastestream), a risk-based set of criteria (e.g., that 95% of nationwide
receptors experience less than 10"6 risk of excess cancer), and assessment factors (i.e., the indices
associated with the modeled risk outputs, e.g., WMUs, receptor of concern, distance from
source).
2-9
-------
Section 2.0
3MRA Methodology
c (cj
ubefgijkt
Cbefgijkt^™' = Annual concentration of
chemical e, In contact medium I, pathway
j, exposure route k, in exposure area g, at
site f, in year t, due to waste concentration
Concentration averaging period (Ae)
PRbefghijkt^"^
t,
PRbefghijkt^Cw^
PRbefghijkt^"^ = Pathway specific risk for
cohort that starts exposure at time t1 to t1
+dfgh, associated with representative
receptor of type h, for pathway j, involving
exposure route k, and contact medium i, in
exposure area g, at site f, for chemical e, with
waste concentration Cw and WMU type b.
h+dfgh Cbefgifdt
tn
t,
t (Year)
Figure 2.2 Relationship Between Exposure Concentration and Pathway Risk
2-10
-------
Section 2.0
3MRA Methodology
Consolidation of Risk Time Series Output Data
Anticipating that the amount of computer memory required to store the full extent of the
risk/HQ time series across all site simulations is prohibitively large it was necessary to condense
the information contained in the time series and store only that data required for subsequent
national regulatory decision analysis. To this end, 3MRA employs two specific steps in the risk
module to reduce the risk/HQ time series data. First, risk/HQ time series representing individual
exposure locations are collapsed into three cumulative frequency histograms, one for each of
three areas defined by circular rings drawn at specific distances around the waste management
unit (0.5 km, 1.0 km, 2.0 km). The histograms are constructed annually and include a series of
risk/HQ intervals (referred to as risk bins) and the number of receptors of a given type that incur
risks/HQs within the interval. Tables 2.3 and 2.4 list the human risk and hazard quotient intervals
(bins) within which population counts are accumulated. Thus, for example, assume that 100
receptor locations are present within the AOI. Further, assume that 17 of the locations lie within
0.5 km of the source, 31 locations lie between 0.5 km and 1.0 km, and 52 locations lie between
1.0 km and 2.0 km. Following the consolidation protocol the 17 time series for receptor
locations within 0.5 km are collapsed into a single time series with each year containing a
histogram showing the number of receptors from across the 17 locations, that experience
risks/hazards within the binned range. Similar consolidations are performed for the 1.0 km and
2.0 km distances. After this first step of consolidation only three sets of risk/HQ time series
remain per risk index. This step may reduce the amount of data to be stored by one or more
orders of magnitude, depending on the total number of receptor locations occurring within the
exposure areas.
The second step of consolidation of risk/HQ time series information eliminates the time
series. In this step, each time series of risk/HQ is scanned to determine the year in which the
maximum risk/HQ occurs. This year is referred to as the critical year (Tcrit). Of the complete
time series of cumulative histograms only those associated with Tcrit years are output and
stored. Specifically, for each distance ring, receptor/cohort combination, and exposure pathway
(for which the entire time series of histograms has been developed), the histogram associated
with the Tcrit year for that pathway is output. In addition, however, the histograms associated
with all other pathways at that same Tcrit year are also output. These other histograms will not
necessarily be the histograms corresponding to their own Tcrit years. However, it is of interest
to examine risk distributions for other pathways during the critical year for a given pathway,
because this presents information about the contribution of these pathways to the total risk/HQ.
Thus, for example, if there are M receptor/ring combinations for each of N pathways, then MxN
sets of histograms are output. Storing histograms for only Tcrit may reduce data storage needs
by more than three orders of magnitude, depending on the total number of years included in the
simulation. Table 2.5 lists the full set of dimensions for which the human risk/HQ bins are
produced.
Similar histograms (for HQs only) are produced and stored for ecological receptors,
however, the breakdown of reporting dimensions is different than for human receptors. Rather
than pathway specific HQs the ecological cumulative frequency histograms are stored for
various combinations of habitat group, habitat type, receptor group, and trophic level.
2-11
-------
Section 2.0
3MRA Methodology
Table 2.3 Summary of Human Risk Bins
Risk Bin Number
Risk Bin Range
1
0.0 <= X < 5 x 109
2
5 x 109 <= X < 7.5 x 108
3
7.5 x 108 <= X < 7.5 x 107
4
7.5 x 107 <= X < 2.5 x 106
5
2.5 x 10"6 <= X < 7.5 x 10"6
6
7.5 x 10"6 <= X < 5 x 10"5
7
5 x105 <= X
Table 2.4 Summary of Human Hazard Quotient (HQ) Bins
Human HQ Bin
Human HQ Bin Range
1
0.0 <= X < 0.05
2
0.05 <= X < 0.5
3
0.5 <= X < 5.0
4
5.0 <= X
2-12
-------
Section 2.0
3MRA Methodology
Table 2.5 Summary of Human Risk Module Output Dimensions Associated with Risk Bins
Number of Distances(a)
Number of Exposure Pathways plus Summation of
Pathways(b)
Number of Receptor Types plus Summation of
Receptor Types(c)
Number of Cohorts plus Summation of Cohorts(d)
Number of Bins to Tally Individual Excess Cancers(e)
Number of Bins to Tally hazard Quotients (Non-Cancer)®
Number of Critical Year Percentiles®
Number of Cws(h)
Number of Chemicals'0
WMU Types"
Number of Sites/WMU-Type Combinations®
Parameters
Dimensions
Human-Risk Module
Outputs
12
16
40
419
(a) The distance rings are: 0 to 0.5 km, 0 km to 1 km, and 0 to 2 km from the edge of
the waste site area.
(b) Inhalation Air, Inhalation through Showering, Summation of all Inhalation
Pathways, Ingestion of Groundwater, Ingestion of Soil, Ingestion of Meat, Ingestion of
Milk, Ingestion of Fish, Ingestion of Breast Milk, Ingestion of Vegetables, Summation of
all Ingestion Pathways, Summation of all Inhalation and Ingestion Pathways.
(c) The risk module analyzes 16 receptor types (8 each with and without drinking
water): Beef Farmer, Dairy Farmer, Beef Farmer Fisher, Dairy Farmer Fisher,
Gardener, Gardener Fisher, Resident, and Resident Fisher. Of these 16 receptor types,
the risk module rolls-up the results and passes only 5 receptor types to the ELP-I:
Beef/Dairy Farmer, Gardener, Fisher, Resident, and Summation of Receptor Types.
(d) The risk module analyzes five cohorts: Infants, 1-6 years old, 7-12 years old, 13-17
years old, and 18 years old and older (adult).
(e) Risk bins include (0.0 - 5.0 x 109, (5.0 x 109 - 7.5 x 10 s), (7.5 x io8 - 7.5 x io7), (7.5 x
10 7 - 2.5 x 10"), (2.5 x IO"6 - 7.5 x IO"6), (7.5 x IO"6 - 5.0 x IO"5), and >5.0 x IO"5.
(f) Hazard bins include (0.0 - 0.05), (0.05 - 0.5), (0.5 - 5.0), and >5.0.
(g) The critical year is defined as the year associated with a risk representing a
percentage of the peak
(h) Five levels of Cw, before disposal are stored (mg/L for waste water [SI and AT],
mg/kg dry weight for solids [WP and LF], and mg/kg wet weight [LAU]). These levels
are chemical specific.
(i) Currently, 43 chemicals are included in the 3MRA chemical database
(j) WP, LAU, SI, AT, and LF.
(k) Each site may contain multiple WMU types, but each WMU type will be assessed
one at a time. The maximum possible number of possible combinations is 419, as some
2-13
-------
Section 2.0
3MRA Methodology
2.2 National Protection Measures
To establish national regulatory limits (e.g., concentration thresholds that define
hazardous versus non-hazardous wastestreams), it is necessary to accumulate the site-based risk
results into expressions of national risk. In the case of 3MRA, site-based risk/HQ that quantify
the number of receptors incurring risks/HQ at various levels are transformed into percentages of
receptor populations that are protective at the various levels of risk. This normalization of the
population counts allows site risks to be accumulated in order to determine the percentage of
nationwide receptors that are protected. It is possible to establish a regulatory limit based on the
percentage of protected receptors. For example, a limit could be established based on criteria
that specifies that 95% of all receptors across all sites, across all pathways, across all waste
management unit types, within 2 km of the WMU, incur an excess cancer risk of 10"6 or less.
Because the risk/HQ data at the site level is stored by indices including receptor type, exposure
pathway, exposure ring distance, and waste management unit, it is possible to construct "views"
of the national scale protectiveness that reflect varying combinations of the indices. For
example, protection measures can be applied, without loss of generality, to individual receptor
types, combinations of receptor types, individual waste management units, etc., as required by
the regulatory analyst.
A second measure of protection is the nationwide distribution of sites that are protected.
A site is protective if the percentage of site-based receptors incur a risk/HQ less than a specified
target value.
These measures of protection are combined in 3MRA to allow a decision analyst to
specify both the percentage of receptors nationwide as well as the percentage of sites that are to
be protected (e.g., 95% of the sites are protective of 99% of the site-based receptors).
2.3 3MRA Monte Carlo Scheme to Quantify Uncertainty
The final element of the 3MRA national assessment methodology is associated with the
need to characterize the uncertainty related to the national estimates of protectiveness. There
are two general categories of uncertainty that are important to the 3MRA methodology,
uncertainties that characterize a lack of knowledge or error and those that reflect the natural
variability of the cause and effect relationships being modeled. In terms of error-based
uncertainty there is error associated with the selection of sites sampled to represent the national
population of waste management facilities/locations. There is error in the data collected to
represent environmental conditions at each of the sites. There is error associated with the
simulation models used to simulate the movement of chemicals from waste management units,
through environmental media, to locations where contact with human and ecological receptors
occur. Finally, there is simulation error, that is, the error associated with the finite number of
Monte Carlo simulations conducted. Natural variability associated with the risk/HQ results
from the fact that the myriad of factors that influence exposure and risk process are different
both within and across sites.
The motivation for separating the various sources of uncertainty is to identify those that
can be reduced as opposed to those that can not. Uncertainties due to error are reducible (e.g.,
sampling error may be reduced by increasing the number of samples) while those that reflect
natural variability are non-reducible. To facilitate the characterization and separation of these
2-14
-------
Section 2.0
3MRA Methodology
two types of uncertainty the 3MRA methodology includes a two-stage Monte Carlo simulation
procedure. The Monte-Carlo procedure is designed to meet the following objectives:
Provide an estimate of the uncertainty in the estimated measures of protection
associated with modeling outputs (i.e., limiting waste concentration (Cw);
Provide a mechanism for accounting separately for variability and uncertainty;
Provide a (value of information) basis for comparing the potential benefit
(reduced prediction uncertainty) versus cost of future efforts to reduce the level of
error in the assessment (e.g., collect more data, develop better models, etc.);
Provide a flexible framework that can accommodate alternate policy formulations
including different definitions of protection criteria.
The first stage of the 3MRA Monte Carlo procedure is designed to account for variability
while the second stage addresses error-based uncertainty. Figure 2.2 illustrates the matrix
oriented organization of information that results from a two-stage Monte Carlo simulation
applied to 3MRA. Within each cell of the matrix resides the risk/HQ results from a single site
simulation. A single iteration of the first stage of the Monte Carlo simulation results in one
column of information in the matrix, which represents the variability of risks/HQs occurring
across individual sites. If no error existed in the data, sampling, or modeling a single execution
of the first stage would yield a certain expression of the natural variability in risk. Because error
does exist, the second stage of the procedure allows the error to be characterized and processed
explicitly. For example, a modeling variable that is naturally varying, such as hydraulic
conductivity, may be characterized by collecting a number of random samples and constructing a
statistical distribution to represent the variability. However, there is uncertainty in the
parameters of the statistical distribution due to both measurement error and sampling error. IF
these errors can be
characterized, then they can be processed as part of the second stage of the Monte Carlo
procedure. Executing this second stage is represented across the columns of the matrix shown in
Figure 2.2.
Each iteration of the first stage results in an estimate of variability "with" uncertainty.
When information in this matrix is queried the regulatory analyst can generate quantitative
statements of uncertainty associated with the national measures of protectiveness. Figure 2.3(a)
presents an example corresponding to a query for a target risk level of 10"6from the N; (columns)
iterations of risk matrices corresponding to a waste concentration of 10"3 mg/kg. The figure
indicates that there is a 5% chance that the level of protection (% of receptors that would be
protected at the target risk level for the given waste concentration) would be less than or equal to
85%. Similarly, there is a 25% chance that less than or equal to 93% of the receptors would be
protected at the target risk level for the given waste concentration.
The result of repeating the query for different target risk levels for the same waste
concentration 10"3 mg/kg is illustrated by Figure 2.3(b), which presents the uncertainty in the
percent of protected receptors for each risk level. From Figure 2.3(b), it can be inferred that
there is a 95% chance that setting the waste concentration regulatory limit to 0.001 mg/kg, would
result in at least 85% of the receptors protected to a 10"6 risk level {or 5% chance that, at the risk
level of 10~6, less than 85% of the receptors will be protected), and at least 90% of the receptors
protected to a 10"5 risk level. Similarly, there would be a 95% chance that at least 95% of the
2-15
-------
Section 2.0
3MRA Methodology
receptors would be protected to the 10"4 risk level, and at least 50% of the receptors would be
protected to the 10"4 risk level.
Querying the output data base for different waste concentrations can produce the set of
graphs such as those shown in Figures 2.3(a), (b), and (c). The figure shows how the percent
protection varies as a function of the target risk, the waste concentration and the confidence
limit; and can be used to select the waste concentration that meets a specified protection
measure. These types of figures could also be produced for subsets of receptors to investigate
the effects of selecting a waste concentration on secondary protection measures.
Note: The full two-stage Monte Carlo scheme is not yet implemented within 3MRA. This is
primarily due to the fact that data characterizing the uncertainty associated with the various sources
of error is not available. It is, however, the case that a limited two-stage Monte Carlo capability has
been implemented. The same matrix of risk information shown in Figure 2.2 is produced except
that the uncertainty iterations (i.e., columns) reflect simulation error for the first stage of the Monte
Carlo only.
2-16
-------
Section 2.0
3MRA Methodology
Figure 2.3 Nf x Ni Pathway Risk Matrix Output.
For fixed:
Chemical Type (e)
UNCERTAINTY
Waste concentration (Cw)
WMU Type (b)
ITERATION
i
2
3
H
1
PRb,e,l(Cw, 1)
PRbiM(Cw, 2)
PRb,e,i(Cw, Nj)
HH
H
2
PRb,e,2(Cw, 1)
PRb,e,2(Cw, 2)
PRb,e,i(Cw, Nj)
hJ
HH
3
HH
PQ
HH
PRb,e,f(Cw, IT)
<
u
HH
c
Pi
Ph
<
>
Nf
PR-b,e,Nf(Cw> 1)
PRb,e,Nf(Cw5 2)
PRb,e,Nf (CW3 N;)
Note:
Each element of the above matrix can be any risk matrix, e.g., PRbef(Cw, IT), or MRb e f(Cw, IT), where PRbef(Cw, IT) is the
pathway risk matrix for WMU type b, chemical e, and site for waste concentration Cw and iteration IT, and MRb e f(Cw, IT) is
the contact medium risk matrix for WMU type b, chemical e, and site for waste concentration Cw and iteration IT.
-------
Section 2.0
3MRA Methodology
100-r
(98%, 95%)
(97%, 75%)
Q.
VI
c
o
o
CD
-t—»
o
1
Q_
50--
(93%, 25%)
CL
0 -
80%
85%
90%
95%
100%
P (% Receptor Protected, Cw = 103, TR = 10 6)
Figure 2.3.a Probability that percent protection is less than P for a given waste concentration and
target risk level.
2-18
-------
Section 2.0
3MRA Methodology
97%)
100%
(-4, 95%)
Probability that
protection level
will be less than
P%-^
(-6, 85%)
75%
o
CD
-t—•
O
I—
0_
50%
(/}
L—
O
4—•
Q.
CD
O
CD
tr
xO
0s
95%
25%
75%
50%
25%
5%'
0%
¦9
8
6
7
¦5
¦4
Log (Risk | Cw = 10^mg/kg)
Figure 2.3.b Percent of receptors protected for different risk levels and Cw=10"3 for N Monte-Carlo
iterations.
2-19
-------
100%
3 log(1/Cw | Risk = 10"6)
100% T
(b)
(c)
"O
CD
"5
CD
o
L_
Q_
co
b.
O
Q.
CD
O
CD
cc
Figure 2.4
50% --
0%
Risk = 10E
100% T
"O
CD
¦4—'
o
CD
o
Q_
in
o
¦4—'
Q.
CD
O
CD
CC
50%
0%
log(l/Cw | Risk = 10")
, 95%)
Risk = 10'
r
T
T
1 2 3 log(1/Cw | Risk = 10^)
Percent of receptors protected for different waste concentrations and risk levels.
2-20
-------
3.0 3MRA Technology
This section describes the integrated 3MRA technology design. The design is presented
in three segments. Section 3.1 describes the specific list of system requirements based on the
3MRA methodology. Section 3.2 describes the overall system design and how it reflects the
3MRA methodology and the associated requirements. Section 3.3 presents a more detailed
description of the individual components of the modeling system.
3.1 System Requirements
The 3MRA technology design is "requirements-based", that is, statements of specific
functionality along with statements of hardware and software specifications combine to define
the requirements that the technology must satisfy. Requirements specify "what" needs to be
done but do not dictate "how" the requirement is to be implemented. The collective statement of
requirements is transcribed into a software system design and subsequently implemented into a
fully functional modeling system.
In laying out the systems level design for the 3MRA technology three categories of
requirements are specified. First and foremost, requirements related to the transcription of the
3MRA national assessment methodology into a computer-based technology is presented. These
requirements essentially map the functional needs related to the elements of the methodology
(national assessment strategy, models, data) to functions required of the technology. Second,
there are programmatic requirements that must be met. These requirements reflect the larger
EPA regulatory and science research context within which the 3MRA modeling system will be
applied, maintained, and enhanced. Finally, there are requirements related to the computer
software and hardware environment within which the 3MRA technology is intended to operate
and several computer science oriented requirements that target modern software engineering
practice.
Section 3.1.1 describes the requirements for transcribing the 3MRA national assessment
methodology. Section 3.1.2 describes Agency programmatic requirements and Section 3.1.3
describes the software and hardware requirements .
3.1.1 Transcription of 3MRA Methodology into Technology Requirements
The specific system requirements for transcribing the full 3MRA methodology into a
computer software technology are as follows:
SITE-BASED MODELING
• Provide an integrated site scale environmental risk assessment model. The model
must simulate physical/chemical/biological processes related to :
1) five waste management unit types,
2) full multi-media fate and transport (i.e., air, watershed, vadose
zone, aquifer, and surface water),
3) terrestrial and aquatic foodweb,
4) a farm foodchain,
2-21
-------
Section 3.0
3MRA Technology
5) human/ecological exposure, and
6) human/ecological risk
• Must accommodate legacy models that have long history of use and acceptability
(e.g., EXAMS, ISC).
• Provide a software management system capable of executing a science-based
model for the purpose of simulating site scale human and ecological risk.
DATA
• Store site-based data for 201 sites that are geographically distributed across the
U.S.
• Provide data storage for environmental variables characterized at regional and
national scale.
• Accommodate statistical representation for all data contained in site-based,
regional, and national databases.
• Provide for the storage of all data required to execute a 3MRA risk assessment at
201 facilities distributed across the U.S.
• Allow statistical sampling of data values for any variable contained in the
databases.
• Facilitate statistical sampling from the following distributions : Uniform, Integer
Uniform, Triangular, Normal, Lognormal, Transformed Lognormal, Weibull,
Empirical, Gamma, SU, SB, Johnson SB, Exponential.
• Provide database of chemical-specific properties (physical/chemical/biological
constants/rate s/factor s).
• Store and access meteorological data on hourly, daily, monthly, and annual
average basis for MET stations across the U.S.
• Accommodate "legacy" databases that may be "hardwired" to legacy models.
• Provide for the centralization for all data that may be shared/used by multiple
science modules.
NATIONAL ASSESSMENT STRATEGY
• Provide capability to execute nested loops of 3MRA assessments factors,
including chemicals, sites, waste management units, wastestream concentration
levels, and Monte Carlo iterations.
• Facilitate a two-stage Monte Carlo simulation, the first stage of which is executed
for a specified number of sites containing waste management units.
• Establish tiered or prioritized access to databases containing assessment data
(Site-based, Regional, and National databases).
• Manage the execution of site-based risk assessments using 3MRA science models
and databases, each simulation will represent a combination of site, chemical,
waste management unit, and waste constituent concentration.
• Generate and store risk matrices, i.e., risk estimates as a function of site, monte
carlo realization, receptor, receptor cohorts, exposure areas, contact media,
exposure pathways, and exposure routes.
• Generate measures of protection for both graphical and tabular display.
3-1
-------
Section 3.0
3MRA Technology
3.1.2 Programmatic Requirements
Programmatic requirements for the 3MRA technology design reflect specific constraints
and technology extension needs imposed by the EPA Office of Solid Waste and the Office of
Research and Development. These requirements are especially important from the perspective
of developing a system design. The 3MRA technology design reflects significant consideration
of anticipated future applications of the 3RMA methodology and technology to a myriad of
regulatory assessments.
These requirements include the following :
• Implement the 3MRA methodology in a manner that is adaptable to a wide range
of specific assessments that may use the methodology.
• Initial implementation must be operational within short-term regulatory time
frame of OSW.
• Facilitate integrated research by providing a modeling environment within which
modelers could develop, assess, and compare models and modeling approaches to
addressing environmental problems.
• Must be extensible to allow site-specific assessments in addition to national scale
site-based assessments.
• Free modelers from need to develop all peripheral software (e.g., user interfaces,
data/model analysis/connectivity tools, etc.).
• Perform modeling based assessments : provide users a modeling technology that
provides access to a community of models, data bases, and data analysis tools as
well as an ability to "build" assessment strategies.
3.1.3 Software Environment and Software Engineering Requirements
The software related requirements include the specification of the hardware platform and
operating systems upon which and within which the 3MRA technology is to operate. In
addition, the software engineering requirements include features that accommodate access to and
use of the technology by model developers not involved in the original development. The list of
software-based requirements for the 3MRA technology include the following :
SOFTWARE/HARDWARE/UTILITIES
• Implement on IBM-compatible personal computers (PC) and be designed to run
on a Pentium (586)-compatible computer with a 200 MHZ processing speed, 64
megabytes of RAM, and a 6-gigabyte hard drive or greater. Greater capacity is
recommended.
• Operate in and have applications compiled for a MS Windows 98/2000/NT
environment ( 32-bit)
• Be designed with performance criteria emphasizing run-time efficiencies.
• Accommodate a variety of programming languages (i.e., FORTRAN, C++, C,
Visual Basic, Java)
• Provide system user interface that facilitates the setup of 3MRA national risk
assessments
3-2
-------
Section 3.0
3MRA Technology
• Provide for internal data quality assurance with respect to data representation and
interchange
• Provide flexible model execution management that allows all or a subset of the
full set of multi-media models to be included in an assessment
• Monitor and report Central Processing Unit (CPU) usage per major system
component
• Facilitate user access to results and importing them to other applications for
additional data analyses.
• Facilitate plug and play features that allow immediate system access of new
components (e.g., alternative models of watershed fate and transport)
3-3
-------
Section 3.0
3MRA Technology
3.2 System Design
The objective of the 3MRA software technology design is to transcribe the statement of
requirements into an integrated system design that clearly establishes functional components and
the operational relationship among them. From the statement of requirements we can extract a
high level algorithm, or set of steps, that the technology must execute. These steps help define
the major segments or components of the software system design with the relative order and
sequencing of the steps defining the operational relationship among the components. From here,
the design is hierarchical in that each of the principal steps of the algorithm (or components of
the system) are expanded to deeper levels of detail, and as they are an overall architecture for the
system unfolds. The high level algorithm and the corresponding technology system design are
presented in this section. Section 3.3 then presents a more detailed summary of the system
components, their specific function and the relationship among them.
Figure 3.1 illustrates the 3MRA national assessment methodology in the form of an
algorithm illustrating the highest level functionality that must be accommodated in the system
design.
Solicit from the user the list of chemicals, sites, WMUs, Cw's and the number of
Monte Carlo Iterations to be simulated
For each Monte Carlo iteration
For each chemical
For each site
For each WMU
For each Cw
Populate all Model Input Files
Execute Site Assessment
Store Site Risk Results
Next Cw
Next WMU
Next Site
Next Chemical
Next Monte Carlo Iteration
Solicit from user Regulatory Criteria and query risk results to develop national tables
and plots of Cw vs (Receptor/site) Protectiveness
Figure 3.1 3MRA National Assessment High Level Algorithm
3-4
-------
Section 3.0
3MRA Technology
Figure 3.2 presents a graphical view of the 3MRA modeling system design. The
components of the system shown in Figure 3.2 include four types : science modules, databases,
system processors, and system data files. The roles and relationships among the components can
be concluded from their form and organization in the graphic. However, the operational
relationship among the components, that is, the programming standards and system utilities that
establish and facilitate the data and execution management protocols for the system, are not
shown in Figure 3.2, but are critical to appreciating the full design. The following brief
description of the components and programming aspects of the system design presented in
Figure 3.2 provides a big picture or framework view of the system that forms the basis for
understanding progressively more detailed descriptions presented in subsequent sections.
Sections 3.2.1 through 3.2.6 provide descriptions of the major features of the 3MRA modeling
system.
\/
System User Interface (SUI)
Waste Management Facility Loop (201 National Sites)
Waste Management Unit Loop (5 Source Types)
Sampled Input Data Iteration Loop (nr)
Chemical Loop (43 Metals & Organics)
C„,= Waste stream concentration
Cw Loop
Key
I ~l User Interface
j j Data File
O Processor
| | Database
J. Header Info from SUI
Warnings/Errors to SUI
List ot Sites
Exit
Level
Processor
II
Site-Based
Database
Multimedia
Multipathway
Simulation
Processor
Exit
Level
Processor
I
bite
Definition
Processor
Regional
Database
Risk
Visualization
Processor
National
Database
List of Chemicals
Chemical
Properties
Database
Chemical
Properties
Processor
Metal
Isotherms
MET
Database
Site Input Data
"V
Site Definition
"V"
Multimedia Multipathway
Simulation
"V"
Cw Exit Level Processing
Figure 3.2 3MRA System Design
3.2.1 Science Modules
The extensive site scale modeling required for 3MRA was partitioned into logical units
based on "real world" objects (representing, essentially, well bounded environmental media and
processes such as a "watershed" and "aquatic foodweb"). The resulting set of science modules
and their logical connectivity (i.e., data dependency relationships) are illustrated in Figure 3.3.
The individual science modules include five source release modules (land application unit,
3-5
-------
Section 3.0
3MRA Technology
landfill, surface impoundment, aerated tank, and wastepile), five media-specific fate and
transport modules (watershed, atmospheric, vadose zone, aquifer, and surface water), a farm
foodchain module, two foodweb modules (terrestrial and aquatic), and two exposure and two risk
modules (human and ecological).
In the system diagram they are represented by the circle in the center of the diagram
labeled Multi-media, multi-pathway Simulation Processor (MMSP). This collection of science
modules represents a modeling system or domain that is, itself, contained in the larger 3MRA
modeling system. The coordinated execution of the modules is managed by the MMSP.
Ecological
Risk
Terrestrial Food
Web
Surface
Impoundment
Aerated
Tank
Aquatic Food
Web
Ecological
Exposure
Surface
Water
Watershed
Human
Exposure
Waste Pile
Land
Application
Unit
Vadose
Zone
Aquifer
Farm FoodChain
Transport
Exposure/Risk
Figure 3.3 3MRA Science Modules and Related Connectivity
3-6
-------
Section 3.0
3MRA Technology
3.2.2 System Databases
The following system databases are included in the 3MRA modeling system.
Site-based Database: Contains all data associated with specific 3MRA sites. There are 201
unique site locations, each containing one or more waste management units. There are a total of
419 combinations of site and WMU.
Regional Database: Contains data associated with variables that are characterized on a regional
geographic basis. All data contained in the regional database is in the form of a statistical
distribution to facilitate sampling and assignment to specific sites.
National Database: Contains data associated with variables that are characterized on a national
geographic basis. All data contained in the national database is in the form of statistical
distribution to facilitate sampling and assignment to specific sites.
Chemical Properties Database: A series of data files that contain all chemical-specific data (e.g.,
anaerobic biodegradation rate constants, health effects data).
Meteorological Database: A series of data files that contain meteorological data for weather
stations across the U.S. The MET data is provided in various temporal formats including hourly,
daily, monthly, annual, and long term average.
Metal Isotherm Database: Database used by groundwater modules (vadose and aquifer) only.
Contains media partitioning data for metals.
3.2.3 System Processors
The system processors collectively manage the execution of the 3MRA modeling system.
Processors interact with the user, develop science module input data files, manage the execution
of the of individual system components, and process modeling outputs to form national risk
summaries. The 3MRA Processors are as follows :
System User Interface (SUP : this processor represents the user access point to the technology.
Via the SUI the user selects which combinations of sites, waste management units, chemicals,
constituent concentrations in waste streams to be simulated, and the number of Monte Carlo
simulations to be executed per site. The SUI also provides the user with the ability to configure
the computer directory structure where individual components of the system are stored. Finally,
the SUI manages the overall execution of the user defined national assessment.
Site Definition Processor fSDP) : this processor performs all data retrieval from the site,
regional, national, and chemical databases and organizes it into a series of "site simulation files"
that contain the input data for each of the seventeen science models.
Multi-media Simulation Processor fMMSP) : this processor manages the invocation, execution,
and error handling associated with the seventeen individual science models that simulate source
release, multimedia fate and transport, foodweb dynamics, and human/ecological exposure and
risk.
3-7
-------
Section 3.0
3MRA Technology
Chemical Properties Processor fCPP): this processor accesses the chemical properties database
and either transfers or calculates all requested data. The CPP represents a single location within
the modeling system where chemical data is made available.
Exit Level Processor I (ELP I): this processor assimilates the individual site risk results and
builds a risk summary database containing data used to assess national protection criteria.
Exit Level Processor II (ELP II): this processor reads the ELP I derived risk summary database
and generates, based on regulatory criteria, specific national exemption levels.
Risk Visualization Processor (RVP): accessed through the ELP II (and thus effectively a sub-
component of the ELP II) this processor presents national risk summary results in graphical
form.
3.2.4 System Data Files
System data files are those that are created during execution of the 3MRA modeling
system. The following system data files are included in design.
Site Simulation Files TSSFsV Files containing general simulation information and input data for
individual science modules. The SSFs are produced by the Site Definition Processor based on
general simulation instructions from the SUI and the SDPs access to the set of 3MRA system
databases.
Global Results Files fGRFs): Files generated by individual science modules containing module
simulation results. A single GRF is produced by each science module but may be consumed by
multiple science modules that execute after the producing module (e.g., the source GRF contains
inputs, and is then consumed by the individual media fate and transport modules).
Risk Summary Output File fRSOF): Files generated by the ELP I Processor and containing site-
based risk data in a condensed format for use in developing national protectiveness summaries.
Protective Summary Output File (PSOF): Files generated by the ELP II Processor and
containing tables of national regulatory limits per regulatory scenario (i.e., combination of
protectiveness criteria and decision factors, e.g., combination of waste management units, human
and ecological receptors of concern, exposure pathways, etc.).
Miscellaneous Data Files: Within the 3MRA modeling system science modules, particularly
legacy models, may require input data files in the original model format and produce output data
files not used by the 3MRA system. These files are read and written by module-specific pre- and
post-processors that function as translation routines facilitating communication of data from the
3MRA system to/from legacy modules.
3-8
-------
Section 3.0
3MRA Technology
3.2.5 Programming Standards and System Utilities
Within the 3MRA modeling system there are specific design features that represent
"standard protocols" that must be included as part of any components implementation. These
standards ensure the consistent and accurate transfer of information throughout he system. In
addition, 3MRA includes utilities to facilitate specific types of data preparations and transfers.
The following standards and utilities are included in the 3MRA technology design.
Data representation standard : With the combination of the sheer number of variables included
in the modeling system and the need to share the data among numerous systems components the
3MRA design team developed a standard means for defining and referencing data. This is the
most fundamental aspect of the 3MRA modeling technology system. In 3MRA, data is assigned
to a specific data file that is itself associated with a particular science module or system
processor. Further, each variable is defined with the following attributes : Name, Dimensions,
Data Type, Min, Max, Units, Description, Dimensional Indices. These metadata for the
modeling variables are contained within a DICTIONARY file. A dictionary file is assigned to
each modeling file and used by the 10 dll (see below) to retrieve and assign data to files. During
execution, each software component in the system must repeat the attributes of a variable
whenever it attempts to either read or write a value for the variable.
Input/Output Dynamic Linked Library (10 dll) : this utility contains a library of routines that
perform all the functions needed to open/close files, read/write data from databases and inter-
model files, and perform QA checks on each data transfer. This processor acts as an Application
Programming Interface (API) and is designed to be functional in multiple computer languages
(i.e., C, C++, FORTRAN, Visual Basic, and Java). All system components utilize this common
API to effect data transfers.
Monte Carlo Dynamic Linked Library (MC dll) : this utility contains a library of routines to
perform all the functions needed to randomly sample data values, including correlated data, from
defined statistical distributions.
3-9
-------
Section 3.0
3MRA Technology
3.2.6 Integrated Functionality of 3MRA Modeling System
The 3MRA national assessment methodology that applies the science modules and
related data is reflected in Figure 3.2 in the box labeled System User Interface (SUI) and the Exit
Level Processors I and II. Within the SUI the nested loops that reflect the combinations of
simulation factors that are needed to conduct a national assessment are shown. The SUI acts as a
dispatcher, first interacting with the user to determine which combination of site assessment
factors should be executed and then methodically invoking the SDP, MMSP, and ELP I to
populate the SSFs with data from the 3MRA databases, execute the individual site simulations,
and store the necessary risk summary data, respectively.
As described earlier the 3MRA methodology includes a two-stage Monte Carlo
simulation capability. The first stage is represented by the site loop, that is, a single execution of
all the sites. The second stage is reflected in the SUI loop entitled "Sampled Input Data
Iteration". This stage of Monte Carlo simulation executes the first stage repetitively, i.e.,
multiple executions of the set of sites.
When all site simulations are complete the RSOF is populated with all risk summary data
needed to generate national level expressions of risk protectiveness. The ELP II conducts the
national roll up. The ELP II invokes the RVP to request from the decision analyst the regulatory
scenarios of interest. The RVP displays plots of national scale probability of protectiveness and
allows the user to view calculated regulatory limits that result from the specific regulatory
criteria. The ELP II generates tables containing the national regulatory limits.
3-10
-------
Section 3.0
3MRA Technology
3.3 Component Design
The previous discussion provided an overview of how the 3MRA technology
requirements were transcribed into an integrated software system design. While attention was
paid to the individual components of the system the focus was a description of how the
components inter-relate in order to automate the 3MRA national assessment methodology. The
purpose of this section is to provide additional detail concerning the components of the 3MRA
software system.
Included in this discussion are the following design elements and features :
Data Representation Standard
Data Exchange Utility (IOdll)
Monte Carlo Sampling Utility (MCdll)
Chemical Properties Processor (CPP)
3MRA Databases
System User Interface (SUI)
Site Definition Processor (SDP)
Multimedia, Multipathway Simulation Processor (MMSP)
Exit Level I Processor (ELP I)
Exit Level II Processor (ELP II)
[Section 3.3.1]
[Section 3.3.2]
[Section 3.3.3]
[Section 3.3.4]
[Section 3.3.5]
[Section 3.3.6]
[Section 3.3.7]
[Section 3.3.8]
[Section 3.3.9]
[Section 3.3.10]
3.3.1 Data Representation Standard for the 3MRA Modeling System
To perform a comprehensive site-scale human and ecological risk assessment requires a
significant amount of environmental data. The data characterizes the inter-related physical
features of the site (e.g., watersheds, surface waters, aquifers, ecological habitats, population
distributions, land uses, etc.) along with the detailed data describing the physical, chemical, and
biological processes effecting the fate, transport, uptake, exposure, and risk associated with
contaminant releases to the environment. Data required for any given site (of the 201 included
in the 3MRA database) amounts to several hundred variables. These data are extracted from the
central databases (i.e., Site, Regional, National) by the Site Definition Processor (SDP) which, in
turn, organizes specific input files for each of the seventeen science modules contained in the
3MRA modeling system. The science modules themselves produce output information (e.g.,
concentration time series) that is consumed by yet other modules. The point here is that within a
3MRA-based national risk assessment there are terabytes of data produced and consumed.
Obviously, not all the data can be archived, yet it is essential that each and every data transfer be
correct. To this end, the 3MRA technology design includes a formal mechanism for ensuring the
quality of data, its representation, and its transfer among components of the system.
All data that is transferred within the 3MRA modeling system, whether it be from the
central databases through the SDP to the SSFs or from one science module to another via GRFs,
is formally identified with a series of attributes. The attributes reflect the metadata of the data
and include the following : Variable Name (Code), Dimensions, DataType, Min, Max, Units,
Description, Diml, Dim2, Dim3. The metadata that describes the attributes of each individual
data item is stored in system files referred to as Dictionary files. A dictionary file is assigned to
each modeling file. Figure 3.4 illustrates the system context of Dictionary files within the
3MRA software system. One DIC file exists for each data file included in the system. While
3-11
-------
Section 3.0
3MRA Technology
each data file contains the numerical value for the variable, each associated DIC file contains the
meta data (the data attributes).
To illustrate this concept further Table 3.1 contains the numerical data associated with a
Site Simulation File for the Watershed Module. Note that in addition to the numerical value per
variable there are additional entries. These entries refer to the variable metadata. The definition
of the meta data is contained in the associated DIC file shown in Table 3.2. While the metadata
concept and dictionary files provide the information to ensure data quality within the 3MRA
modeling system it must be implemented in a formal manner by all software (i.e., system
processors and science modules). This is achieved in 3MRA with an Input/Output Dynamic
ExpF actor Dictionary
MediaConc Dictionary
SpecConc Dictionary
Time series of
Air, soil, water
concentrations
Ecological
Exposure
Module
ExpMod Input Dictionary
Time series of
Species Tissue
Concentrations
Specie-specific
Exposure factors
User supplied
Module unique
data
Figure 3.4 3MRADictionary Files
3-12
-------
Section 3.0
3MRA Technology
l,
"wsWPMaineHg.ssf"data group",
23,
"a_BF",0,"FLOAT",0,"m/d",
1.08e-005,
"b_BF",0,"FLOAT",0,"unitless",
1. 16,
"bcm",0,"FLOAT",0,"unitless",
1,
"C",1,"FLOAT",0,"unitless",
5,0.00510101,0.005 045328,0.00961849,0.009241012,0.021134361,
"CN",1,"FLOAT",0,"unitless",
5,72.8 98 98 99,73.0815 8996,74.17942708,73.78739458,75.25770925,
"ConVs",0,"FLOAT",0,"m/d",
0.287852011,
"deltDiv",0,"INTEGER",0, "unitless",
1,
"DRZ",1,"FLOAT",0,"cm",
5,197.979798,199.70083 68,191.63385 42,193.2232579,176.1674009,
"Infild",0,"FLOAT",0,"m/d",
0,
"K",1,"FLOAT",0,"kg/m2",
5,0.18,0.18,0.18,0.18,0.18,
"Ksat",1,"FLOAT",0,"cm/h",
5,2.123995151,10.85980699,1.664272109,1.043309383,0.1275850911,
"P",1,"FLOAT",0, "unitless",
5,1,1,0.972005208,0.97314 6915,0.895374 44 9,
"RunID",0,"STRING", 0,
"Version25",
"SMb",1,"FLOAT",0,"unitless",
5,4.9,4.9,4.9,4.9,4.9,
"SMFC",2,"FLOAT",0,"volume %",
5,
4,22.5,23.2,22.9,21.3,
4,22.5,23.2,22.9,21.3,
4,22.5,23.2,22.9,21.3,
4,22.5,23.2,22.9,21.3,
4,22.5,23.2,22.9,21.3,
"SMWP",2,"FLOAT",0,"volume %",
5,
4,10.4,12.1,11.9,11.5,
4,10.4,12.1,11.9,11.5,
4,10.4,12.1,11.9,11.5,
4,10.4,12.1,11.9,11.5,
4,10.4,12.1,11.9,11.5,
"Theta",1,"FLOAT",0,"degrees",
5,2.919734303,2.919734303,3.662160152,2.17 632 6011,2.4 6235273,
"thetawZld",0,"FLOAT",0,"volume fraction",
0,
"WCS",1,"FLOAT",0,"volume fraction",
5,0.41,0.41,0.41,0.41,0.41,
"X", 1, "FLOAT",0,"m",
5,39.2 64 00 663,46.74481561,37.60883794,51.55062411,62.55905 807,
"zava",0,"FLOAT",0,"m",
0,
"zavb",0,"FLOAT",0,"m",
0. 05,
"zZlsa",0,"FLOAT",0,"m",
0. 05,
Table 3.1 Example 3MRA Data File : Watershed Site Simulation File (ws.ssf)
Linked Library (IOdll), as described in the next section.
3-13
-------
Section 3.0
3MRA Technology
24.watershed
Code. Dimensions. DataTypc.Min. Max. Units. Description. Dim 1 .Dim2.Dim3
a_BF,0,float,0,1000,m/d,regression coefficient a for baseflow model...
b_BF,0,float,0,10,unitless,regression coefficient b for baseflow model...
bcm,0,float,0,l,unitless,boundary condition multiplier (lower boundary),,,
C,l,float,0,l,unitless,USLE cover factor,NumWSSub„
CN,l,float,0,100,unitless,SCS curve number,NumWSSub„
ConVs,0,float,0,10,m/d,settling velocity (suspended solids),,,
deltDiv,0,integer,l,10,unitless,time step divider (for debugging),,,
DRZ,l,float,0,1000,cm,depth (root zone),NumWSSub„
Infild,0,float,0,100,m/d,input infiltration rate (for debugging),,,
K,l,float,0,l,kg/m2,USLE erodibility factor,NumWSSub„
Ksat,l,float,0.00000001,1000000,cm/h,saturated hydraulic conductivity (soil),NumWSSub„
P,l,float,0,l,unitless,USLE erosion control factor (watershed j),NumWSSub„
RunID,0,string,,,,run identification label (optional),,,
SMb,l,float,0,12,unitless,soil moisture coefficient b,NumWSSub„
SMFC.2.float.0.100.volume %,soil moisture field capacity,NumWSSub,nlayer,
SMWP,2,float,0,lOO.volume %,soil moisture wilting point,NumWSSub,nlayer,
Theta,l,float,0,75,degrees,slope (watershed),NumWSSub„
thetawZ ld,0,float,0, 1 .volume fractionjnput volumetric water content in till zone (for debugging),,,
WCS,l,float,0,l,volume fraction saturated water content (total porosity),NumWSSub„
X,l,float,0,50000,m,flow length (watershed),NumWSSub„
zava,0,float,0,10,m,averaging depth upper (depth averaged soil concentration),,,
zavb,0,float,0.01,100,m,averaging depth lower (depth averaged soil concentration),,,
zZlsa,0,float,0.01,l,m,depth (modeled soil column),,.
Table 3.2 Example 3MRA DIC File for Watershed Site Simulation File (ws.dic)
3-14
-------
Section 3.0
3MRA Technology
3.3.2 Input/Output Dynamic Linked Library (IOdll)
The purpose of the IOdll is to standardize and facilitate the exchange of data among
system components of the 3MRA technology. The IOdll is a library of utilities that contains all
the functions needed to open/close files, read/write data from databases and inter-module files
(i.e., SSFs/GRFs), and perform QA checks on each data transfer. The IOdll acts as an
Application Programming Interface (API) and is designed to function with multiple computer
languages (i.e., C, C++, FORTRAN, Visual Basic, and Java). All system components utilize this
common API to effect data transfers.
The design of the 3MRA IOdll encompasses several subroutines and functions, including
those for initialization, input, output, and finalization. In summary, the subroutines and
functions include the following:
FUNCTION
DESCRIPTION
Subroutine
OpenGroups
This subroutine is expected to be the first call in a program and to
first retrieve the call arguments and open the warning and error
files. If the warning and error files are opened successfully, all
appropriate data groups are opened with the correct read/write
mode.
Subroutine
CloseGroups
This subroutine is expected to be the last call in a program and to
close all data group, warning, and error files. It then destroys the
error file, signaling to the SUI that no error occurred in the
component.
Function Readlnt
This function returns the integer value for the given data group,
variable name, and set of indexes.
Function ReadReal
This function returns the real value for the given data group,
variable name, and set of indexes.
Function
ReadLogical
This function returns the logical value for the given data group,
variable name, and set of indexes.
Subroutine
ReadString
This subroutine returns for the variable String, the string value for
the given data group, variable name, and set of indexes.
Subroutine
NumArgs
This subroutine returns the number of call arguments handed to a
module.
Function GetArglnt
This function returns the integer value for the given call-argument
index.
Subroutine
GetArgString
This subroutine returns the string value for the call-argument
index.
Subroutine
Writelnt
This subroutine stores the given integer value in the given data
group, variable name, and set of indexes.
Subroutine
WriteReal
This subroutine stores the given real value in the given data group,
variable name, and set of indexes.
3-15
-------
Section 3.0
3MRA Technology
Subroutine
WriteLog
This subroutine stores the given logical value in the given data
group, variable name, and set of indexes.
Subroutine
WriteString
This subroutine stores a string of up to 80 character values in the
given data group, variable name, and set of indexes.
Subroutine Error
This subroutine will add the string value passed by the calling
program to the error file and then terminate the calling program.
Subroutine
Warning
This subroutine will add the string value passed by the calling
program to the warning file and then return control to the calling
program.
In addition, the 3MRA IOdll performs Quality Assurance checks on all data transfers. To
accomplish this it is required that all software within the 3MRA modeling system read and write
data using the IOdll. This requires that the read and write statements include the attributes of the
data item (including the name of the file within which the variable/value is to be located). The
IOdll checks to ensure that the attributes stated by the reading/writing code are consistent with
the attributes contained in the DIC file containing the variable of interest. For example, if a
science module coded in FORTRAN contained the following line of code:
MyInts(i,j,k)=ReadInt3("Source_SSF," "XYZ," "cm/s,"i,j,k)
the IOdll will execute the following QA tests :
1) Does the science module data group "Source SSF" exist? If not, raise an error.
2) Does the variable "XYZ" exist in "Source SSF"? If not, raise an error.
3) Are any of the indexes negative? If so, raise an error.
4) Do the specified units for "XYZ" in "Source SSF" match the units passed by the
calling program? If not, raise an error.
5) Does the specified dimensionality of "XYZ" match the calling program's
assumptions of dimensionality? If not, raise an error.
6) Is any one of the indexes larger than the maximum indexes stored in the file? If
so, raise a warning and return a 0.
7) Is the value stored in the data file outside the specified bounds for that variable?
If so, raise an error.
8) Does the specified type of the variable match the assumed type from the calling
program? If not, raise an error.
3-16
-------
Section 3.0
3MRA Technology
3.3.3 Monte Carlo Sampling Dynamic Linked Library (MCdll)
One of the requirements for the 3MRA modeling system is that all variables contained in
the site, regional, and national databases be optionally characterized as stochastic. The intent of
this requirement is to facilitate Monte Carlo-based simulations where individual variables that
are either naturally variable in nature or their value is uncertain can be assigned values from a
statistical distribution.
To satisfy these requirements and to provide a single source of such statistical sampling
within the modeling system the 3MRA technology design includes a Monte Carlo dynamic
linked library (MCdll) that contains a core set of Monte Carlo functions and subroutines.
The 3MRA MCdll includes the following statistical distribution types; Normal,
LogNormal, JohnsonSB, Transformed LogNormal, Exponential, Triangular, Uniform, Integer
Uniform, SB, SU, Gamma, Weibull, and Empirical. The MCdll allows correlation among the
variables.
A key feature of the MCdll implementation is that the random number seed is set as an
input parameter rather than, for example, being set by an internal routine. This standardization
of the random number sequence allows for explicit reproducibility of results. This is critical
both in the software development phase as well as in the context of regulatory applications.
The 3MRA MCdll is comprised of the following subroutines:
FUNCTION
DESCRIPTION
Subroutine
StatNumDist
This subroutine sets the number of distributions that will have
values sampled from them.
Subroutine
StatNumCor
This subroutine sets the number of correlations between the
given distributions that will have values sampled from them.
Subroutine
StatDebugOn
This subroutine will turn on debug messages in the MCdll.
Subroutine StatSeed
This subroutine sets the seed to be used by the MCdll. This
allows for a reproducible set of samples from the given set of
distributions.
Subroutine StatClear
This subroutine deallocates any memory resources used by
the MCdll.
Subroutine StatCor
This subroutine sets a correlation between two distributions.
Subroutine
StatSample
This subroutine performs the sampling from the given set of
distributions.
Distribution
Definition Functions
In general, the Distribution Definition Functions provide the
MCdll with the parameters for a specific distribution type.
For example, the Normal() function sets the mean, standard
deviation, minimum, and maximum for a normal distribution.
3-17
-------
Section 3.0
3MRA Technology
While the MCdll is accessible by any 3MRA processor or science module it is the case
that the Site Definition Processor (SDP) executes all statistical sampling. This centralization of
the Monte Carlo sampling is designed for efficiency and quality assurance.
3.3.4 Chemical Properties Processor (CPP)
The Chemical Properties Processor (CPP) represents the single source of chemical-
specific data in the 3MRA modeling system. The CPP represents a library of subroutines and
functions that receive requests for specific chemical property data, accesses the relevant
chemical data file(s), performs any required computations, and returns a value for the requested
property.
Within the 3MRA design the CPP receives requests for chemical property data either
from the Site Definition Processor (SDP) or directly from the science modules. The protocol
followed is that the SDP, which is responsible for populating all module input files (i.e., Site
Simulation Files - SSFs), requests chemical data from the CPP, which, in turn, accesses a series
of specific data files, each containing a group of chemical data, and either transfers directly or
derives the requested chemical property value. In the event that a science module simulates
environmental conditions that affect the value of a chemical property (e.g., temperature and pH
effects on physical properties of organic compounds) the science module can access the CPP and
1) provide the environmental conditions, and 2) request, in return, appropriate condition-specific
chemical property values.
The 3MRA modeling system contains the following series of chemical property data files:
• Organic Chemical Property (OCP)
• Metal/Inorganic Chemical Property (MICP)
• Transformation Products (TP)
• Catalyzation (CAT)
• Aerobic Biodegradation (AerBio)
• Activated Biodegradation (ActBio)
• Anaerobic Biodegration (AnaBio)
• Anaerobic Reduction Biodegradation (AnaRedBio)
• S04 Reduction Biodegradation (S04Bio)
• Methanogenic Biodegradation (MethBio)
• Human Health Benchmarks (HHB)
• Ecological Benchmarks (EB)
• Ecological Bioaccumulation Factors (EBF)
• Aquatic Bioaccumulation Factors (ABF)
• Chemical Ecological Flag
These files are designed and maintained as comma delimited ASCII data files. As such
access and retrieval of data if efficient and the files are easily assimilated into standard
spreadsheet software for review and modification.
The CPP is designed as a dynamic link library (DLL) to facilitate consistent
communication between components. The CPP contains a series of subroutines and functions
that retrieve chemical data from the files and perform any necessary calculations. For example,
the following set of subroutines is included for accessing the Organic Chemical Property (OCP)
data:
3-18
-------
Section 3.0
3MRA Technology
CPP Function
for OCP
Description of Function
Function
ChemADiff
This function computes the air diffusion coefficient for organic
chemicals, in cm2/s.
Function
ChemVol
This function computes the volume for organic chemicals, in mL.
Function
ChemDen
This function computes the density for organic chemicals, in
g/mL.
Function
ChemWDiff
This function computes the water diffusion coefficient for organic
chemicals, in cm2/s.
Function
ChemVP
This function computes the vapor pressure for organic chemicals,
in torr.
Function
ChemSol
This function computes the solubility limit for organic chemicals,
in mg/L.
Function
ChemHLC
This function computes the Henry's Law Constant for organic
chemicals in (atm m3 / mol).
Function
ChemKow
This function computes the Kow for organic chemicals
(dimensionless).
Function
ChemKoc
This function computes the Koc for organic chemicals in mL/g.
Function
ChemHyd
This function computes the hydrolysis rate for organic chemicals
in 1/day units.
There are two sets of subroutines included in the CPP that do not return specific chemical
property data. The first set of subroutines initialize the CPP, that is, the initialization subroutines
establish the computer directory location where the chemical data files can be found, the
chemical for which properties are to be determined, and the environmental conditions relevant
for computing the chemical property (i.e., pH, temperature, and environmental media). The
second set of subroutines that do not return specific chemical property data inform the calling
program about 1) the SMILES string for organic chemicals, 2) the number of valid chemicals for
which the CPP has properties (if a chemical is missing part of its required data, it will not be
included in this count), and 3) to create a list of the available chemicals (and their associated
index, name, and CASID) available in the chemical property file being read by the CPP.
Requests for all other chemical properties are processed with individual subroutines or
function calls specific to the property of interest. In the case when chemical transformation rates
are involved it may also be the case that the transformation reaction results in the formation of
chemical products. For all transformation related chemical properties (e.g., biodegradation rate
constant) the calling program can also request information about the number and name of
chemical products and the yield coefficients associated with the transformation reaction.
3-19
-------
Section 3.0
3MRA Technology
3.3.5 3MRA Databases and Data Files
Within the 3MRA modeling system there are three main categories of data; input,
intermediate, and output. Input data includes all data needed to simulate a comprehensive
multimedia risk assessment at each of 201 sites. Intermediate data includes information
produced for or by any of the science modules within the modeling system. The intermediate
files produced by the science modules contain times series predictions of variables associated
with individual science modules (e.g., concentration time series). These intermediate files report
data on an annual basis (some modules internally process time steps shorter than one year but
report results only on an annual basis). 3MRA output files contain protection data, that is, the
end result of all the site-based simulations are data matrices that contain information necessary to
determine percentages of national population (human and ecological) and/or waste management
unit sites that are protected from risk/hazard quotient values greater than a policy prescribed
threshold.
The following three sections provide summary descriptions of the three categories of
data.
3.3.5.1 3MRA Input Data
The 3MRA input data includes site, waste management unit, chemical, and
meteorological data. The site data includes both site layout and site environmental setting
information. The site layout data reflects all the physical features of the site and their
interconnect vity. Features included in the site layout data for 3MRA include :
• Waste Management Unit (WMU)
• Watershed(s)
• Surface Water Network(s)
• Aquifer(s)
• Ecological Habitat(s)
• Ecological Home Ranges per Specie
• Human Population Distribution
Site environmental setting data includes the following data groups :
• Atmosphere
• Surface water
• Watershed soils/cover
• Vadose zone
• Aquifer
• Terrestrial and aquatic foodwebs
• Farm foodchain
• Waste Management Unit characteristics
As described earlier the site environmental data is assigned via a hierarchical protocol for
populating the series of Site Simulation Files (SSFs) that form the input for the individual
science modules. The protocol calls for a series of three data tables, each reflecting a higher
spatial representation of the data. The site data table contains data that has been collected for a
specific site. If data for a site does not exist the 3MRA modeling system (i.e., the SDP) scans
3-20
-------
Section 3.0
3MRA Technology
the regional data set. If the SDP finds the variable in the regional data set a value is extracted
(may be sampled using the Monte Carlo DLL) and placed in the appropriate SSF. Finally, if the
variable is not located in either the site or regional data set the SDP scans the national data file
for appropriate information from which to derive a site value to be used during simulation.
Also included in the 3MRA input data are chemical and meteorological data. Chemical
properties are included in a series of files that are organized by the type of property. The
categories of chemical properties included in 3MRA are :
• Physical Property
• Chemical and biological transformation rates (with products of transformation)
• Biouptake and bioaccumulation factors
• Metal Isotherms
• Human and ecological health benchmark
The chemical data is stored in a series of comma delimited ASCII files that facilitate their
access by system processors and models and their overall maintenance. The data is read from
the data files by the CPP (primarily upon request of the SDP) using the same IOdll used by all
other processors and models in the modeling system. This provides for, as described earlier, a
high degree of quality assurance in the use of the data.
The last category of input data contained in the 3MRA modeling system is the
meteorological data. Because the science modules contained in 3MRA may simulate processes
at time steps less than the reporting time step (i.e., annual) the MET data must be available for
hourly, daily, and monthly time frames as well as annual and long term average. Most
meteorological data were extracted from Solar and Meteorological Surface Observation Network
(SAMSON) hourly data files and converted as necessary to daily time series, monthly time
series, annual time series, and long-term averages.
Table 5.3 lists the specific meteorological data needs of the 3MRA modeling system and the
time frames needed for each of the data items.
Table 3.3 3MRA MET Data
MET Variable
Hour
Day
Month
Annual
Long
Term
Surface roughness length
~
Minimum Monin-Obukhov
length
~
Friction velocity
~
Precipitation code
~
Stability category
~
Cloud cover
~
~
~
Temperature
~
~
~
~
~
Wind direction
~
3-21
-------
Section 3.0
3MRA Technology
Windspeed
~
~
~
~
~
Precipitation
~
~
~
~
~
Evaporation
~
~
~
~
Windspeed at pile height
~
Number hours wind > 5.4 m/s
at pile height
~
% time wind > 5.4 m/s at pile
height
~
Thornthwaite precipitation-
evaporation index (PE)
~
Days with precipitation > 0.01
in
~
~
Fastest mile of wind
~
~
USLE rainfall erosivity factor
~
~
Maximum/minimum
temperature
~
Temperature of soil column
~
Mixing height
~
Each site included in the 3MRA database is associated with one MET station set of data.
This association is established principally as a function of proximity between the site and the
station.
Finally, the meteorological data files are not designed for use within the 3MRA IOdll
construct, nor are they processed by the SDP with results incorporated into the Site Simulations
Files. Rather, the MET files are contained in a series of standard data files for direct access by
the science modules when needed. The primary reasons for this implementation strategy are
simply resources and schedule.
3.3.5.2 3MRA Intermediate Data Files
There are two primary types of intermediate files contained in the 3MRA modeling
system; Site Simulation Files (SSFs) that contain input data for the science modules and
"header" information that describes runtime parameters, and Global Output Files (GRFs) that
contain the results of individual science module simulations. The software design aspects of
these files is described in Sections 3.3.1 and 3.3.2. Presented here is a brief description of the
different types of SSF and GRF files based on their general content.
There are four types of SSF file within the 3MRA modeling system; Header SSF (hd.ssf),
Site Layout SSF (sl.ssf), Chemical Data SSF (cp*.ssf), and Science Module SSF (*.ssf). The
hd.ssf contains information associated with the overall simulation and includes data that
3-22
-------
Section 3.0
3MRA Technology
specifies the hard drive location on the host computer where all relevant files (data, modules,
processors, etc.) reside. The sl.ssf contains data that describes the physical features of the site
(i.e., area of interest surrounding a waste management unit) and their inter-connectivity, and as
such, is accessed by several science modules. The chemical data required by the science
modules is contained in a series of SSF files, one for each major media or science module. In
all, there exist nine chemical SSF files in the 3MRA; source module (cpsr.ssf), watershed
module (cpws.ssf), vadose zone module (cpvz.ssf), aquifer module (cpaq.ssf), farm food chain
(cpff.ssf), human exposure module (cphe.ssf), and three files associated with the surface water
module (for lakes [cpLake.ssf], streams [cpstream.ssf], and wetlands [cpwetland.ssf]).
The science modules each require a single input SSF that contains the values of variables
unique to that module. Thus, there is a *.ssf for each of the seventeen 3MRA science modules in
the system. Similarly, there is a *.grf for each of the science modules that contains the module
output data. These data files may be saved by the user upon request, however, it is the case that
if conducting a full national assessment it would be impossible to save all of the individual
intermediate SSF/GRF files. Thus, saving the SSF/GRF files is generally useful when either
executing a single site simulation, and this, in turn, has been useful in the context of verifying the
proper functionality of individual science modules.
3.3.5.3 3MRA Output Data
There are two data files that are permanently stored as a result of a 3MRA national
assessment simulation. The Risk Summary Output File (RSOF) contains a consolidation of the
human and ecological risks estimated by the two science modules. The RSOF is appended to for
each site simulation, thus, as the human and ecological risk modules complete their simulations
the results are assimilated and stored in condensed form for later regulatory analysis. The
second output file generated by the 3MRA modeling system is the Protection Summary Output
File (PSOF). The PSOF contains final regulatory limits derived from applying regulatory
criteria to risk/hazard data stored in the RSOF.
3-23
-------
Section 3.0
3MRA Technology
3.3.6 The System User Interface (SUI)
The System User Interface (SUI) has three primary roles within the 3MRA modeling
system. First, the SUI is responsible for interaction with the user for the purpose of establishing
the details of the site-based risk assessment to be executed. The SUI also manages the overall
execution of the assessment process, processing the nested loops of Monte Carlo iterations,
chemicals, sites, waste management units, and constituent concentration levels in the
wastestreams. Finally, the SUI acts as the single source of warning and error messages that
occur during execution. All messages, whether they are initiated within a science module or a
system processor, are directed toward the SUI for final processing to the user.
The SUI, and indeed this version of the 3MRA technology, facilitates only assessments
that utilize the site data contained in the 3MRA databases. While it is possible to execute the
3MRA technology for a site outside the 3MRA list there is no user interface to facilitate the
input of site data.
3.3.6.1 User Interaction Features
The SUI represents the user access point to the technology. The SUI is a Windows™
menu-driven interface that collects information from the user and controls the activities
associated with the simulation. The three main screens associated with the SUI are 1) system
configuration, 2) system management, and 3) system status. These user interface screens are the
main method of interaction between the user and the 3MRA modeling system.
The system configuration screen, Figure 3.5, allows the user to specify the location on
the host computer of all databases, data files, system processor executables, and science module
executables for a simulation. There are seven system configuration screens, one of which,
shown in Figure 3.5, accepts user specifications of the directory location of the individual media
and transport modules (e.g., the vadose zone module resides on Drive: D within the 3MRA
directory and is named vadose.exe).
The system management screen consists of selections and options subscreens. The
selections subscreen, Figure 3.6, allows the user to select the sites, chemicals, source types, and waste
levels for simulation. The options subscreen, Figure 3.7, allows the user to determine the storage level of
input/output files for the simulation. The Minimum storage level saves only the Global Results Files
(GRFs) for the simulation, and the Maximum storage level saves all processor and module files for the
simulation. The options
3-24
-------
Section 3.0
3MRA Technology
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration j System Management | System Status
Databases | Directories | Processors [j^MSP Modules
Source Transport I Food web I Exposure/Risk!
Vadose zone module
Aquifer module
Watershed module
Waterbody module
Air module
D: \3M RAWadose. exe
d:\3M RA'^Aquiferl D.exe
d:\3MRA\Watershed.bat
d: \3M R AVE xamsl 0. exe
d:\3MRA\Airl0.exe
JnlisJ
"J
"J
_|
"J
"J
Figure 3.5 3MRA SUI System Configuration Screen
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration System Management | System Status
\ Selections ]| Options ]
^jnjxj
Site
0114001
0130207
0131104
0131207
0131508
0136703
0220102
0221207
0223504
0224002
0231002
02311 OS
0231407
0231G10
0231 m 1
1
Chemical Source
T etrachloroethylene [Perchl *~| E£2
Thiram [Thiuram][Tetrarnethj
Toluene [Methylbenzene]
T richloroethylene
Vinyl chloride [Chloroethylen
Silver
Arsenic
Barium
Beryllium
Cadmium
Chromium III (insoluble salts)
Chromium VI
Divalent Mercury
zl
Methyl Mercury
Elemental Mernuru
Waste eve
Figure 3.6 3MRA SUI Systems Management: Selections Screen
3-25
-------
Section 3.0
3MRA Technology
subscreen also allows the user to specify the number of Monte Carlo realizations for the simulation and
the seed value used for the statistical sampling. This subscreen allows the user to specify the mode of
operation for the simulation. These modes are: a) Debug, b) Stop on Error c) Stop on Warning,
or d) any combination of these three modes. The Debug mode stops the simulation after every change
in realization, chemical, source type, and waste level; the Stop on Error mode stops the simulation
only if an error occurs in a processor or module; and the Stop on Warning mode stops the simulation
only if a warning occurs in a processor or module. The user can select the Debug, Stop on Error,
and Stop on Warning modes in different combinations to meet the needs of the simulation.
The Comment box on this subscreen enables the user to enter descriptive information about the
current simulation.
The SYSTEM STATUS SCREEN, Figure 3.8, informs the user of the status of the current
simulation. Current realization, site, source type, chemical, and waste level are provided by this
screen. Any error or warning message will display in the Message box on the screen and include
Start/Resume
which processor or module the message came from. This screen also has the
button for the simulation. After stopping, the user can resume the simulation at anytime or use
the
Reset
button to begin the simulation over (that is, put all the indexes back to the original
settings and start the simulation from the beginning). This provides the user with full control
over the simulation.
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
Jnjicj
System Configuration [ System Management ;| System Status
Selections Options
Storage level
Number of realizations
Seed value for realizations 111031
Comments
| Maximum
P
V Debug mode
¦ J] Stop on error
V Stop on warning
zl
Figure 3.7 3MRA SUI System Management: Options Screen
3-26
-------
Section 3.0
3MRA Technology
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
^jn]2(|
System Configuration | System Management : System Status
Current realization
Current site iteration
Current source iteration
Current chemical iteration
Current waste level iteration
Messages
F
|MaineHg
WP
(7439-97-6
I*
Finished
Run 1 of 1, 1 successful runs
Reset
Stop Run
"7]
Warning
"Elp Start",
"Elp Success",
"ELP Time(s),Site,Source,ChemicaLCw,Realization",
"23.000000,MaineHg,WP,"Divalent Mercury",4,1
Successful run
Figure 3.8 3MRA SUI System Status Screen
3.3.6.2 Execution Management Features
The SUI also manages the overall execution of the user specified assessment. As
described above the user selects combinations of sites, chemicals, waste management units,
constituent concentration levels in wastestreams, and the number of Monte Carlo realizations to
execute. To implement the system, the SUI takes control over each of the system processors
before the start of analysis and invokes processors in an ordered and logical manner.
Finally, the SUI receives and manages warning and error messages occurring in
processors and modules. If the SUI receives a warning from a processor or module, the SUI
displays the warning and continues (unless the Stop on Warning mode is selected, in which case
the simulation stops) with the simulation in the system status screen (see Figure 3.8). If an error is
reported to the SUI, the SUI displays the error message and current state of the indexes and stops
execution.
3.3.7 The Site Definition Processor (SDP)
The purpose of the SDP is to populate all Site Simulation Files (SSFs) needed to conduct
individual site assessments required in the context of a 3MRA application. Site assessments are
conducted for each combination of site, waste management unit, chemical, and chemical waste
concentration requested by the user via the System User Interface. Thus, for example, if the
user selects 30 sites, 4 chemicals, 2 waste management units, and all 5 waste concentration levels
3-27
-------
Section 3.0
3MRA Technology
there would be a total of 1200 individual site risk assessments executed during the single session
initiated by the user. The SDP, under the control of the SUI, gathers data for each site risk
assessment from the Site Variable Distribution Table, Regional Variable Distribution Table,
National Variable Distribution Table, and Chemical Properties Database and transfers that data
to the appropriate SSF.
The SUI provides the SDP with a list of Site Simulation Files that must be generated.
The SDP refers to the DIC files associated with each SSF to determine the exact list of data
needed. Because not all sites are completely characterized with site-specific data the SDP is
designed to sequentially search the site, regional, and national databases to find a value for the
variable. In most cases the SDP simply reads the list of variables from a DIC file and queries
the site, regional, national, and chemical databases to either directly transcribe or statistically
sample a value for the variable (Figure 3.9). In many cases however, the SDP must process the
data found in the master databases. For example, in populating the vadose zone module SSF the
SDP must read the value of the vadose zone thickness (VadThick) from the site database and
then calculate the value of "dispersivity" based on the following expression :
Dispr = 0.02 + (0.022 x VadThick)
Thus, the SDP represents a sophisticated data extraction and data processing capability within the 3MRA
modeling system.
For each Chemical
For each Site
For each Source Type
For each Waste Level,
— Specify site layout and transfer all information from the Site
Variable Distribution Database to SSF.
— For each variable that has not been defined in the Site Database,
transfer all constant variables and stochastic variables
(distributions with mean, variance, and range) information,
based on the scenario definition (including chemical
information for all available chemicals), from the regional
statistics database to SSF.
— For each variable that has not been defined in the Site and
Regional Database, transfer all constant variables and stochastic
variables (distributions with mean, variance, and range)
information, based on the scenario definition (including
chemical information for all available chemicals), from the
National Database to SSF.
End loop on Waste Level
End loop on Source Type
End loop on Site.
End loop on Chemical
Figure 3.9 SDP Processing Steps
3-28
-------
Section 3.0
3MRA Technology
System User Interface
Module Execution Manager
Source H
Food- \
Chain J-4—
exposure
& Risk.
Multimedia Multipathway
Simulation Processor
Chemical
Properties
Database
Met
Database
Chemical
Properties
Processor
Figure 3.10 Details of the Multimedia Multipathway Simulation Processor
(Dashed lines indicate input from complementary components that are outside the
MMSP)
3.3.8 Multimedia Multipathway Simulation Processor (MMSP)
The MMSP represents a software system embedded within a larger software system. The
MMSP can be viewed as a site-scale multimedia modeling system, or domain, contained within a
national scale regulatory risk assessment methodology, or assessment domain. The primary
purpose of the MMSP is to manage the ordered execution of the 3MRA science modules for
each site-based assessment requested by the user via the System User's Interface. Management
tasks that the MMSP executes include :
Read input from the Header Site Simulation File (hd.ssf), that contain data that
informs the MMSP concerning the site assessment to be conducted and the
location of all relevant files and databases.
Pass necessary simulation data to science modules and invoke their execution in
proper sequential order.
Direct the disposition of intermediate data files produced and consumed within
the sequential execution of the science modules. The disposition of the data files
3-29
-------
Section 3.0
3MRA Technology
depends on storage options selected by the user and passed to the MMSP by the
SUI.
• Terminate processing of science modules and return control to the SUI, reporting
any erros that occurred during simulation.
• Provide information on the time elapsed for the processing of each science
module.
Figure 3.10 illustrates the MMSP design. The key feature of this design is that the MMSP
Execution Manager functions as an interface between the System User Interface (SUI) and the
individual science modules. In this capacity the MMSP is invoked by the SUI and provided all
information necessary to coordinate a single execution of the science modules. Following
structured design techniques, and in particular, the concept of separation of concerns in the
software context, the MMSP is not aware of the larger national assessment context within which
individual site assessments are conducted. The SUI simply hands control over to the MMSP and
waits for the MMSP to return control. The following section describes the MMSP interactions
with the science modules.
3.3.8.1 MMSP Science Module Design Features
A science module simulates, in an integrated fashion, a set of environmental processes
(physical, chemical, and biological) and in particular, within the 3MRA modeling system, how
these processes interact to bring chemical contaminants into contact with human and ecological
receptors. In simulating human and ecological risk the 3MRA takes a comprehensive approach
to representing the environmental media and processes that play a role. In terms of the software
system design for expressing this comprehensive science-based approach the 3MRA applies
modern software engineering design principles. In particular, the 3MRA design for the
environmental modeling capability centers around establishing modeling components that reflect
real world objects. Thus, within the context of site-scale risk assessment such objects include
the waste management unit, the watershed, the ecological habitats and associated specie home
ranges, surface waters (lakes, streams, wetlands), etc. Within the 3MRA, the integrated science
of environmental processes is expressed through a total of seventeen individual modules, each
simulating a well bounded real world component of the complete environmental risk system.
The collection of 3MRA science modules and their relationship as represented in 3MRA is
shown in Figure 3.11. The following sections describe the science modules operational
characteristics within the larger 3MRA modeling system.
The MMSP design and implementation require that each science module obtain input
from specified SSF Data Groups and, possibly, specified GRF Data Groups. Science modules
use the IOdll subroutines to read all SSF and GRF Data Groups. The MMSP Module Execution
Manager provides each science module with the file name and location of the input data it
requires in the SSF and potentially in the GRF Data Groups. Two exceptions to this provision
are the meteorological data and, or some science modules, the chemical properties data. If a
science module needs meteorological data (observations of wind speed, direction, and stability
over a defined time period), that science module is expected to read the Met database using
"legacy" methods designed and implemented in previous efforts by EPA. If the module requires
chemical properties data that are not provided through an SSF or GRF, the module is expected to
use the Chemical Properties Processor to read from the Chemical Properties Database.
3-30
-------
Section 3.0
3MRA Technology
Surface
Water
Aquifer
Vadose
Ecological
Air
Human
Risk
Human
Exposure
Terrestrial Food
Web
Farm FoodChain
Ecological
Exposure
Aquatic Food
Web
Aerated
Tank
Land
Application
Unit
Landfill
Sources Transport Foodchain Exposure/Risk
Figure 3.11 Interactions Within the Multimedia Multipathway Simulation
Processor
3.3.8.1.1 Science Module Execution
Science modules representing each of the components of the risk assessment process
were formulated independently from one another and the system itself. This was possible
because all input/output requirements were specified before model formulation occurred. Thus,
each module developer knew exactly what data would be input and output from the module (and
in what format). Although the MMSP Module Execution Manager guides the sequence of
execution and provides the modules with names and locations of input files, the science modules
operate as black boxes in the MMSP. Thus, these modules do not require the MMSP to provide
additional information on how to execute them.
Some of the modules used for the MMSP incorporate legacy models, that is, models that
were developed for previous applications outside of the 3MRA context. To incorporate a legacy
model the module developer is expected to reformat and reorder data to meet the specific needs
of their science module. The reformatting and reordering of data within a computational module
can be done by one of three methods:
1. Use pre-processors and/or post-processors without modifying the legacy model.
2. Modify the legacy model to directly read specified files using shared subroutines.
3. Use a combination of processors and model modifications.
These options are shown in Figure 3.12. Note if pre- or post-processors are created for a
computational module, a batch file also is required as part of the computational module. This
batch file serves as the single entry point for execution of that computational module.
3-31
-------
Section 3.0
3MRA Technology
Input_
Specification
Module
Modeled Input
Specification
±
1
r
—Pre-Processor
i
Model
r
Post-Processor
1
r
Output
Specification
Figure 3.12 How Legacy Model s Connect with the
FRAMES-HWIR Technology Software System
3.3.8.1.2 Science Module Output
The primary output of the 3MRA science modules are the Global Output Files (GRFs).
The GRFs contain the results of the module simulation in a standard format that can be read by
other 3MRA modules. A single GRF file is produced by each of the 3MRA science modules.
The modules may produce "non-standard" files as well. These files are not recognized by the
3MRA system and generally reflect either data files generated to facilitate the testing of a
module or, in the case of legacy models, data files associated with the original legacy outputs. In
either case, these files are not used by the 3MRA modeling system.
The MMSP is designed to allow for varying degrees of output storage, addressing
varying degrees in quantity (that is, temporary versus permanent files). With a data storage level
of zero, the MMSP Module Execution Manager does not copy anything to permanent space and
will, delete any temporary files the computational module might have created during its
execution. If a computational module is given a data storage level greater than zero, the
computational module is responsible for determining what files in addition to the GRF are saved.
If a computational module leaves a file in temporary space and the data storage level is greater
than zero, the MMSP Module Execution Manager copies that file into permanent space.
3.3.9 The Exit Level Processors (ELP I, ELP II, RVP)
The 3MRA modeling system contains three processors that collectively accumulate,
transform, and present information generated by the individual site-based risk assessments. The
final outputs of the system are in the form of measures of site and population protectiveness at
3-32
-------
Section 3.0
3MRA Technology
the national level. These measures of protectiveness are presented as a function of several
modeling dimensions that represent regulatory decision factors, including receptor type, cohort,
exposure pathway, chemical, chemical concentration in the wastestream, waste management unit
type, and distance from the source.
The Exit Level Processor I (ELP I) reads individual site-based human-health risk/hazard
and ecological hazard results from the human and ecological risk modules Global Results Files
(hr.grf, er.grf), transforms the population counts into percentages of receptor populations
protected, and stores the resulting risk/hazard information in a series of national Risk Summary
Output Files (RSOFs). The ELP II and RVP processors query the RSOF in response to specific
regulatory criteria for protectiveness and produce graphical and tabular views of the trade off
between levels of chemical concentration in wastestreams and levels of protectiveness. With the
combination of the RSOF database and the ELP II/RVP a regulatory analyst can ask the
following type of question "What is the maximum allowable constituent concentration for
wastestreams entering landfills such that at least 90% of all receptors at 95% of the sites
nationwide incur a risk less than 10"6?" The next two sections provide a description of the exit
level processors and the risk visualization processor.
3.3.9.1 Exit Level Processor I (ELP I)
The primary role of the ELP I is to prepare a Risk Summary Output File (RSOF). The
RSOF is similar to the risk module output file in that it contains a series of cumulative frequency
histograms. However, the ELP I converts the contents of histograms from population counts to
percentage of population "protected" at each risk level. In addition, the ELP I accumulates the
percentage of population protected across all sites and simulations. The end result is an RSOF
database that can provide statistics on the percentage of receptors protected nationwide and the
percentage of sites that are protected. The RSOF, however, can not discriminate among specific
sites. Because the statistics are aggregated over all sites it is not possible, looking in the RSOF,
to determine which sites, for example, incur the highest risk.
Table 3.4 lists the dimensions associated with human health risk/hazard before and after
the ELP I. A similar set of dimensions for ecological hazard quotients is included in 3MRA (but
not shown here). For each combination of dimensions an associated histogram (as shown in
Table 3.5). Most significantly, the ELP I eliminates the dimension associated with the number
of sites.
In the process of accumulating risk data across sites the ELP I converts the data from
population counts per risk/hazard bin into percentages of population incurring less than or equal
the risk/hazard associated with the bin. Table 3.6 describes the ELP I processing in matrix form
and explains the computational method used to construct the RSOF. There is one matrix for
each combination of assessment dimensions shown in Table 3.4. Thus, for a chemical, waste
management unit, receptor type, cohort, exposure pathway, Cw, and risk/hazard bin number the
matrix shows whether various percentages of the receptor/cohort population are protected at the
individual site. For example, the first row in Table 3.6 represents the results for Site#l, Monte
Carlo iteration #1, and any single combination of Cw, Risk bin, receptor, cohort, exposure
pathway, and waste management unit. For this combination the data in the first row shows that
at least 50% of the receptor/cohort population experiences risks less than the particular risk level
associated with the bin (it could be any one of the seven risk bins). The next entry in row 1 is a
zero (in the cell marked 75%). This means that at this site the percentage of the receptor/cohort
population that is protected is not as great as 75%, or, said differently, more than 25% of the
3-33
-------
Section 3.0
3MRA Technology
receptor/cohort population is not protected. When a column is summed the total value represents
the number of sites (and therefore the percentage of all sites) for which the column labeled
percentage of receptor/cohort population is protected. This total is contained in the last row of
Table 3.6 and is the only data from the matrix to be stored in the RSOF (all rows associated with
individual sites are released). Table 3.7 shows the unit matrix stored in the RSOF. For each
assessment dimension shown in Table 3.4 there is a Table 3.7 in the RSOF. The table lists the
waste concentration levels (Cw's) versus individual bin levels of risk/hazard. In each cell is an
array of ten values, each reflecting a total count across all site simulations that resulted in the
specified percentage of the population being protected.
Table 3.4. Summary of Parameter Requirements Associated with
Human-Health Risk/Hazard for the ELP-I
Parameters
Dimensions
Supplied
bv Human-
Risk
Module
Stored bv
ELP-I and
Provided to
RVP
Number of Distances(a)
3
3
Number of Exposure Pathways plus Summation of Pathways®
12
12
Number of Receptor Types plus Summation of Receptor
Types(c)
16
5
Number of Cohorts plus Summation of Cohorts'®
5
4
Number of Bins to Tally Individual Excess Cancers(e)
7
7
Number of Bins to Tally hazard Quotients (Non-Cancer)®
4
4
Number of Critical Year Percentiles®
1
1
Number of Cws(h)
5
5
Number of Chemicals®
40
40
WMU Types®
5
5
Number of SitesAVMU-Type Combinations®
419
—
Percentiles of Protected Population®
—
10
(a) "n" distance rings are designed into the ELP-I, but only three distances are stored for HWIR calculations: 0 to
0.5 km, 0 km to 1 km, and 0 to 2 km from the edge of the waste site area.
(b) Inhalation Air, Inhalation through Showering, Summation of all Inhalation Pathways, Ingestion of
Groundwater, Ingestion of Soil, Ingestion of Meat, Ingestion of Milk, Ingestion of Fish, Ingestion of Breast Milk,
Ingestion of Vegetables, Summation of all Ingestion Pathways, Summation of all Inhalation and Ingestion
Pathways.
(c) The risk module analyzes 16 receptor types (8 each with and without drinking water): Beef Farmer, Dairy
Farmer, Beef Farmer Fisher, Dairy Farmer Fisher, Gardener, Gardener Fisher, Resident, and Resident Fisher. Of
these 16 receptor types, the risk module rolls-up the results and passes only 5 receptor types to the ELP-I:
3-34
-------
Section 3.0
3MRA Technology
Beef/Dairy Farmer, Gardener, Fisher, Resident, and Summation of Receptor Types.
(d) The risk module analyzes five cohorts: Infants, 1-6 years old, 7-12 years old, 13-17 years old, and 18 years
old and older (adult). Risk information on the five cohorts is passed to the ELP-I, where the ELP-I rolls-up the
information into four cohorts: Infants, 1-12 years old, 13 years old and older (adult), and Summation of all
Cohorts.
(e) Risk bins include (0.0 - 5.0 x 10"9, (5.0 x 10"9 - 7.5 x 10"8), (7.5 x 10"8 - 7.5 x 10"7), (7.5 x 10"7 - 2.5 x 10"6), (2.5
x 10"6 - 7.5 x 10"6), (7.5 x 10"6 - 5.0 x 10"5), and >5.0 x 10"5. Each risk bin for human health stores by Chemical
and WMU type the Number of Sites that protects at least some percentage of the human population (0%, 5%,
25%, 50%, 75%), 85%o, 90%o, 95%o, 98%o, or 99%o) for each "risk-bin/Cw" pair by distance, pathway, receptor,
cohort, and critical-year method.
(f) Hazard bins include (0.0 - 0.05), (0.05 - 0.5), (0.5 - 5.0), and >5.0. Each hazard bin for human health stores by
Chemical and WMU type the Number of Sites that protects at least some percentage of the human population for
each "hazard-bin/Cw" pair by distance, pathway, receptor, cohort, and critical-year method.
(g) The critical year is defined as the year associated with a risk representing a percentage of the peak
(h) Five levels of Cw before disposal are stored (mg/L for waste water [SI and AT], mg/kg dry weight for solids
[WP and LF], and mg/kg wet weight [LAU]). These levels are chemical specific.
(i) Although the actual number of chemicals assessed per production run varies, the ELP-1 was designed to at
least address 40 chemicals, where 40 chemicals represent a subset of the total number of chemicals to be assessed,
as part of HWIR.
(j) WP, LAU, SI, AT, and LF.
(k) Each site may contain multiple WMU types, but each WMU type will be assessed one at a time. The
maximum possible number of possible combinations is 419, as some sites may not contain a particular WMU
type.
(1) Currently, 10 population protection percentiles have been identified: 99%o, 98%o, 95%o, 90%, 85%o, 75%o, 50%o,
25%o, 5%o, and 0%. These percentiles represent the percentage of the population that at least has been protected.
For 90%o, for example, at least 90% of the population has been protected.
Table 3.5. Summary of Human Risk Bins and Labels
Risk Bin Number
Risk Bin Range
Risk Bin Label
1
0.0 <= X < 5 x 10"9
1 x 10"9
2
5 x 10"9 <= X < 7.5 x 10"8
1 x 10"8
3
7.5 x 10"8 <= X < 7.5 x 10"7
5 x 10"7
4
7.5 x 10"7 <= X < 2.5 x 10"6
1 x 10"6
5
2.5 x 10"6 <= X < 7.5 x 10"6
5 x 10"6
6
7.5 x 10"6 <= X < 5 x 10"5
1 x 10"5
7
5 x 10"5<=X
1 x 10"4
3-35
-------
Section 3.0
3MRA Technology
Table 3.6. Procedure to Compute a Risk Summary Output File
Cwi and Risk/Hazard Biiij(a)
Index on
Site
(1 toN
sites)
Index on
Iteration
(1 toM
iterations)
Equal To or Greater Than Percentage of Population Protected
0%
5%
25%
50%
75%
85%
90%
95%
98%
99%
1
1
1(b)
1
1
1
0
0
0
0
0
0
2
1
1
1
1
1
1
1
1
0
0
0
N
1
1
0
0
0
0
0
0
0
0
0
1
2
1
1
1
0
0
0
0
0
0
0
2
2
etc.
N
2
1
M
2
M
N
M
Sum of Counts by
Column(c)
Column
1
Column
2
Column
3
Column
4
Column
5
Column
6
Column
7
Column
8
Column
9
Column 10
(a) "i" is the index on the Cw from one to five, and "j" is the index on the Risk/Hazard Bin. For human risk, there are seven excess cancer bins; for human hazard,
four hazard quotient bins; for ecological hazard, five ecological-hazard quotient bins. A similar table is associated with each "Cw&risk/hazard bin" pair.
(b) For each site and iteration combination that meets the population-protection criterion, the cell is assigned a value of unity, representing protection. When the cell
does not meet the population-protection criterion, the cell is assigned a value of zero, representing no protection.
(c) The ELP-1 only stores cumulative counts by column in the RSOFs (that is, last row of Table 3.6).
3-36
-------
Section 3.0
3MRA Technology
Table 3.7 Example ELP I Output Table
Cw
Risk/Hazard Bin
1
2
3
M
1
C(1,1X
C(2,l);
C(3,l)j
C(M,1);
2
C(l,2);
3
C(l,3)i
N
C(l,N)i
C(M,N),
M = Number of Risk/Hazard Bins
N = Number of Cws
C = Summation of counts per percentage of population protected
i = index on the percentage of population protected (i.e., 0%, 5%, 24%, 50%, 75%,
85%, 90%, 95%, 98%, 99%)
3-37
-------
Section 3.0
3MRA Technology
3.3.9.2 Exit Level Processor II (ELP I) and Risk Visualization Processor (RVP)
As described in Section 2 the overall objective of the 3MRA national assessment
methodology is to assess the potential human and ecological risks resulting from the disposal of
industrial wastestreams in land-based waste management units. It is of interest to regulatory
decision analysts to know the quantitative relationship between constituent concentration levels
and national risk (or hazard in the case of non-carcinogenic chemicals). National risk is
expressed in 3MRA in terms of two measures of protectiveness, percentage of the national
population experiencing risks less than an assigned regulatory threshold and the percentage of
sites where a regulatory established percentage of the local population experiences risks less than
an assigned regulatory threshold. For example, regulatory decision analysts may wish to know a
concentration level for benzene that if applied to all wastestreams (i.e., any wastestream
exceeding this concentration of benzene would require disposal as a hazardous waste in a
Subtitle C waste management unit and any wastestream with a lesser concentration of benzene
would be allowed disposal in a Subtitle D waste management unit), in all geographic locations,
regardless of the type of waste management unit, would result in at least 95% of the population
at 99% of the sites experience excess cancer risks less than 10"6.
The Risk Summary Output Files (RSOFs) generated by the ELP I contain the necessary
data to allow the ELP II and RVP to calculate specific constituent concentration levels that
satisfy the regulatory decision criteria related to policy-based protection percentages (population
and sites) and risk thresholds.
The following sections provide a description of the means by which the ELP II and the
RVP facilitate the analysis and selection of national exit levels.
The RVP and ELP II actually share the software that performs the protection-based
calculations, as requested by the user. The difference between the RVP and the ELP II is solely
in how they output the results. The RVP is invoked from within the ELP II and provides the
regulatory analyst a menu driven user interface for specifying protection and risk criteria and
viewing graphically the resulting relationship between constituent concentration levels,
protection criteria, and risk thresholds. The ELP II provides the ability to calculate national exit
levels as a function of the regulatory decision criteria. Figure 3.13 illustrates the RVP and ELP
II use of the Risk Summary Output Files (RSOFs) generated by the ELP I to produce the
graphical and tabular views of the relationship between constituent concentration levels and
population/site protectiveness.
3-38
-------
Section 3.0
3MRA Technology
EXIT
LEVEL
PROCESSOR
I
EXIT
LEVEL
PROCESSOR
II
l.le-9
Risk Summary Output File
M = Number of Risk/Hazard Bins
N = Number of Cws
C = Summation of counts per percentage of
Population Protected
/ = index on the percentage of Population
Protected, varying from 1 to 10 (i.e., 0%,
5%, 25%, 50%, 75%, 85%, 90%, 95%,
98%, and 99%)
Cv
R
isk/Haza
rd Bin
1
2
3
M
1
Cf1.1V
c.( 2.n
cn.n
CfM.1V
2
Cf 1.2V
3
Cf 1.3V
N
m NV
CfMNV
ug/g will provide risks below the
target risk level for the selected 95%
receptors for the exposure at 80% of
the sites
Protective Summary
Output File
m
o
Ph
Ph
£
O
H
O
w
H
O
Ph
Ph
Ph
O
80%
RISK
SUALIZATIO
PROCESSOR
l.le-9
Protective Summary
Output Curves
Figure 3.13 Relationship among ELP I, ELP II, and RVP
Both the RVP and the ELP II allow the regulatory analyst to apply the decision criteria to
risk and hazard data as a function of the assessment factors listed in Table 3.4. Thus, the
regulatory analyst can view risk and hazard information for human and ecological receptors from
several perspectives.
Graphical View of Protectiveness vs Cw
The RVP allows the user (e.g., regulatory decision analyst) to query and summarize the
information stored in the RSOFs and to graphically view the results of such queries. The user
provides a set of assessment factors (i.e.,, WMU type, chemical, distance, receptor type, cohort,
etc.) and a level of protection (for example, risk of 1.0 x 10"6 or HQ of 1.0, etc.). The RVP uses
this information to construct plots showing the probability of protecting specified percentages of
human or ecological receptors as a function of Cw.
3-39
-------
Section 3.0
3MRA Technology
The RVP retrieves data from the RSOFs, generates a plot of probability of protection as a
function of Cw, and displays this plot on the screen. Figures 3.14 and 3.15 illustrate the human
risk protective summary and ecological hazard protective summary plots produced by the RVP,
respectively.
The probability of protection (y-axis) represents the probability that any given site is
protective, or, the percentage of sites found to be protective under the various values of Cw (x-
axis) and for specified percentages of the receptor/cohort population (individual curves). Each
plot is generated for a set of user selected assessment factors (shown in the right hand side of the
Figures). The decision analyst can vary the assessment factors and decision criteria and quickly
determine the change in the Cw vs probability of protection relationship.
3-40
-------
Section 3.0
3MRA Technology
a HW1R - Risk Visualization Processor
File Options
Selections j Human Risk jj Ecological HQ |
Cw Interpolater
mu
Population percentile
Probability of protection
Interpolated Cw
1
Set Defaults
Set Individual Defaults
CCD
0
1
0)
o
I—
CL
-a
o
CL
» i i nnnp Fap^
100
90
SB
70
60
50
40
30
20
10
0
Human Risk Protective Smninnrv
0.001
€
1
a
-------
Section 3.0
3MRA Technology
U HW1R - Risk Visualization Processor
File Options
Selections I Human Risk I Ecological HO
ma
Cw Interpolater
Population percentile
Probability of protection
1
1
Interpolated Cw
Set Defaults
Set Individual Defaults
CD
Li* \y • jlij \M
L_
L.
Ecological HQ Protective Summary
100 -
90 -
SO -
C
o
ts
70 -
ai
o
k_
60 -
Q_
4-
o
50 -
&
la
40 -
.s
o
L_
30 |
a.
20 |
10 -
0
—i
5
F
-------
Section 3.0
3MRA Technology
Tabular View of Protectiveness vs Cw
The ELP II consumes and processes RSOF-based data with the same software used by
the RVP. The ELP-II assesses five scenario options and generates seven types of tables and
places them in the Protective Summary Output Files (PSOFs). A scenario table, such as is
shown in Table 3.8, is created to store the scenario and default selections (representing
regulatory criteria). A scenario is defined to include three risk/hazard thresholds (Human Risk,
Human Health Hazard Quotient, and the Ecological Hazard Quotient), probability of protection,
and a population-percentile. Other scenario descriptors for human health include distance,
exposure pathway, receptor type, cohort, critical-year method. Other descriptors for ecological
health include distance, and habitat group, habitat type, receptor group, and trophic level for
ecological risk.
Table 3.8 3MRA Regulatory Scenarios
SCENARIO^
Scenario
1
Scenario
2
Scenario
3
Scenario
4
Scenario
5
Risk Level
10"6
10"6
10"5
10"5
10"7
Human Health-HQ
1
1
1
10
10
Ecological-HQ
1
1
1
10
10
Population-Protection
Percentile
99
99
99
95
98
Probability of Protection
95
85
85
85
90
(a) When comparing scenarios, the remaining scenario parameters (for example, exposure
pathway, receptor type, etc.) do not change from scenario to scenario and, therefore, are not
presented in this table. The parameters shown in the table can be changed by the user on
subsequent executions of the ELP II.
Seven types of tables are created by the ELP-II in the PSOF directory. The following
tables use the exit level scenario settings created in the RVP/ELP-II user interface to produce the
desired results. The tables generated here reference "exit levels". This term is specific to a
particular regulation entitled the Hazardous Waste Identification Rule (HWIR). The HWIR
focused on defining Cw's that represent thresholds between hazardous and non-hazardous waste
classifications. Other regulatory programs may present ELP II output data in a different format
but the basic data processing of the ELP I and ELP II would remain the same.
Lowest Target Exit Level tables are quantitative criteria for allowing a specific class of
industrial waste streams to no longer require disposal as a hazardous waste (that is, to exit
Subtitle C) and to allow disposal in Industrial Subtitle D facilities. Hazardous waste constituents
with values less than these exit-criteria levels would be reclassified as nonhazardous wastes
under the Resource Conservation and Recovery Act. In these tables, the lowest exit level
3-43
-------
Section 3.0
3MRA Technology
concentrations for human health (that is, human risk and hazard) and ecological hazard will be
evaluated and reported.
Table 3.9 Example Lowest Target Exit level Concentrations
Chemical
Casld
Liquid
Semi-Solid
Solid
Landfill
Other WMU...
HH
Eco
Lowest
HH
Eco
Lowest
HH
Eco
Lowest
HH
Eco
Lowest
HH
Eco
Lowest
Benzene
71-43-2
4.00E-
01
3.00E+01
4.00E-01
Lead
7439-92-1
1.00E-
05
3.00E+01
1.00E-05
The Target Exit Level table summarizes the results contained in the Lowest Target Exit Level tables for
each chemical and WMU type by recording the lowest target exit level concentration from all categories
(human health and ecological hazard) and assigning that one value to its respective scenario. Therefore,
each chemical, WMU type, and scenario will be assigned one target exit level concentration (that is, the
lowest associated with that scenario). In addition, this table will note whether the lowest target exit level
concentration was based on the maximum Cw used in the assessment.
Table 3.10 Example Target Exit level Concentrations by Scenario
Chemical Name
Casld
Scenario 1
Max Used
Scenario
2
Max Used
Other Scenarios...
Benzene
71-43-2
4.00E-01
"no"
1.00E+02
"no"
Lead
7439-92-
1
1.00E-05
"no"
1.00E-04
"no"
Relative Target Exit Level tables give the ratio between the first scenario's (that is, Scenario 1 in Table
2.1) target exit level identified and the target exit level of each other scenario for each WMU type.
Table 3.11 Example Relative Target Exit Level Concentrations
Chemical Name
Casld
Scenario 1
Scenario 2
Scenario 3
Scenario 4
Scenario 5
Benzene
71-43-2
1
2.00E+02
3.00E-02
3.00E+00
3.00E+00
Lead
7439-92-1
1
1.00E+01
1.00E+00
1.00E+00
1.00E+00
50% Probability of Protection tables are similar to the Target Exit Level tables; however, the
probability of protection for each scenario is set to 50%.
3-44
-------
Section 3.0
3MRA Technology
Table 3.12 Example Target Exit Level Concentrations, Based on 50% Probability of Protection
Chemical Name
Casld
Scenario 1
Max Used
Scenario
2
Max Used
Other Scenarios...
Benzene
71-43-2
1.00E+02
"no"
1.00E+02
"no"
Lead
7439-92-
1
1.00E-04
"no"
1.00E-04
"no"
Cohort Human Risk/HQ table contains the risk trigger levels found associated with the Protective
Summary Output curves for each cohort category, WMU type, and waste type. In effect, target
risk/hazard level (for example, 10"6 for risk and 1.0 for human HQ) is determined as a function of an exit
level concentration by cohort category. The previous tables have computed the target exit level
concentration as a function of the target risk/hazard level.
Table 3.13 Example Cohort Human Risk/HQ for Landfill
Chemical
Name
Casld
Landfill
Infants
1-12 years old
13 years old and older
Risk
HQ
Risk
HQ
Risk
HQ
Benzene
71-43-2
1E-04
NA
8E-07
NA
0.000005
NA
Lead
7439-92-1
1E-04
NA
0.0001
NA
0.0001
NA
Receptor Human Risk/HQ table contains the risk trigger levels found associated with the Protective
Summary Output curves for each receptor category, WMU type, and waste type. As with the Cohort
Human Risk/HQ tables, the target risk/hazard level is determined as a function of an exit level
concentration by receptor category.
Table 3.14 Example Receptor Human Risk/HQ for WP
Chemical
Name
Casld
W
>
Beef/Dairy Farmer
Gardener
Fisher
Resit
ent
Risk
HQ
Risk
HQ
Risk
HQ
Risk
HQ
Benzene
71-43-2
4E-08
NA
0.000003
NA
0.000006
NA
8E-07
NA
Lead
7439-92-1
0.0001
NA
0.0001
NA
0.0001
NA
0.0001
NA
3-45
-------
Section 3.0
3MRA Technology
Exposure Pathway Human Risk/HQ table contains the risk trigger levels found associated with the
Protective Summary Output curves for each exposure category, WMU type, and waste type. As with the
Cohort Human Risk/HQ tables, the target risk/hazard level is determined as a function of an exit level
concentration by exposure pathway category.
Table 3.15 Exposure Pathway Human Risk/HQ
Chemical
Name
Casld
Air Inha
ation
Soil Ingestion
Water In
sestion
Crop Ingestion
Risk
HQ
Risk
HQ
Risk
HQ
Risk
HQ
Benzene
71-43-2
0.000006
NA
2E-09
NA
1E-04
NA
0.00002
NA
Lead
53-70-3
1E-04
NA
0.0001
NA
1E-04
NA
0.0001
NA
3-46
-------
Section 4.0
Modeling Approach
4,0 3MRA Installation and User's Guide
The 3MRA modeling system is designed to execute on Personal Computers running the
Microsoft Windows 95/98/2000/NT operating systems. This section describes the installation
(Section 4.1), execution (Section 4.2), and simulation output review (Section 4.3) of the 3MRA
modeling system.
4.1 Installing the 3 MR A Modeling System
Installation of the modeling system for the Windows-based operating systems is managed
by a single installation program located on the 3MRA installation disk. The following two
sections present the specific sequence of installation steps for the 3MRA modeling system and a
description of the directory structure (on the PC harddrive) containing the modeling software.
4.1.1 Installation Steps
To initiate the installation the user inserts the installation CD and, from the START/RUN
menu locates and invokes the file setup.exe on the installation disk (e.g., E:\setup.exe). Figures
4.1 and 4.2 illustrate the opening screens of the InstallShield Wizard. The installation program
prepares the system for installation by copying installation files to the users temporary file space
(e.g., C:\Temp). Once prepared for installation, a Welcome screen appears (Figure 4.2). At all
times during the installation the user is provided options to return to a previous screen, continue
to the next screen, or cancel the installation.
3MRA Modeling System - InstallShield Wizard
Preparing Setup
Please wait while the InstallShield Wizard prepares the setup.
HBB
3MRA Modeling System Setup is preparing the InstallShield "Wizard, which will guide you
through the rest of the setup process. Please wait.
Figure 4.1 Opening Screen of 3MRA Installation Program
4-1
-------
Section 4.0
Modeling Approach
Figure 4.3 displays the licensing agreement that must be agreed upon by the user before
installation of the modeling system will proceed. The 3MRA modeling system software is
"open source", that is, all materials related to the modeling system (i.e., program code,
executables, databases, testing packages, and documentation) are available without charge to the
public.
Figure 4.4 displays information concerning two options for installation; "Typical" and
"Custom". The Typical installation installs three items; 1) the 3MRA modeling system files, i.e.,
all executable programs and databases, 2) a set of database connectivity tools for 3MRA
components that have been compiled using the Borland Compiler, and 3) a Java Runtime
3MRA Modeling System Setup
Welcome to the InstallShield Wizard for 3MRA Modeling System
The InstallShield® Wizard will install 3MRA Modeling System on your computer. To continue, click
Next.
liisliiJlMd
Figure 4.2 3MRA Installation Welcome Screen
Environment needed to run the Site Visualization Tool (SVT) that is written in Java. A Custom
installation provides additional database connectivity files that may be needed for older versions
of Windows (e.g., WIN95). Specific information concerning the need for Custom files is
displayed in Figure 4.4. Figures 4.5 and 4.6 display the user installation selection options and
associated information for Typical and Custom installation, respectively. Figure 4.7 displays the
user screen for selecting/deselecting the various custom items.
4-2
-------
Section 4.0
Modeling Approach
3MRA Modeling System Setup
Installation Information
Important installation information - please n
Redistribution and use in source and binary forms, with or without modification,
|| are permitted provided that the following conditions are met:
Redistributions of source code must retain this list of conditions and the
following disclaimer.
Redistributions in binary form must reproduce this list of conditions and
the following disclaimer in the documentation arid/or other materials
provided with the distribution.
Neither the name of the U.S. Environmental Protection Agency nor any other
contributors may be used to endorse or promote products derived from this
software without specific prior written permission.
THIS SOFTWARE IS PROVIDED "AS IS" AND ANY EXPRESS OR IMPLIED
IWARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
[WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE U.S. ENVIRONMENTAL
^1
Do you accept all the terms of the preceding License Agreement? If you select No, the setup will
close. To install 3MRA Modeling System, you must accept this agreement.
Figure 4.3 3MRA Licensing Agreement
3MRA Modeling System Setup
Selecting Components
Database connectivity components - please n
If a user selects "Typical Installation", this program installs 3MRA's executable
and database files only. Components for installing system oriented database
connectivity files are also provided. Choose the "Custom Installation" option to
allow selection and deselection of the following database connectivity components
0DBC3.51, MDAC2.5SP2, and VB5runtime. Read the information below to help you
The ODBC 3.51 installation program and the Visual Basic 5 runtime library
iritallation program are provided for users whos' systems are running older
versions of WindowsNT and Windows98. Most Win98/NT4 system, Windows2000
and XP systems should already have the functionality these files provide
Figure 4.4 Description of 3MRA Installation Options
4-3
-------
Section 4.0
Modeling Approach
3MRA Modeling System Setup
Setup Type
Select the setup type that best suits your needs.
~D
Select TYPICAL for runtime and database files only, CUSTOM for optional database connectivity
¦D escription
Installs 3MRA's executable,
database files and supporting files.
'ustom
Cancel
Figure 4.5 3MRA Installation Selection Screen : Typical
3MRA Modeling System Setup
Setup Type
Select the setup type that best suits your needs.
Select TYPICAL for runtime and database files only, CUSTOM for optional database connectivity
—Description
Installs 3MRA's runtime, database
and supporting files and/or allows
user database connectivity
component selections. Return to
the "Selecting Components"
screen for more detailed
information.
Figure 4.6 3MRA Installation Selection Screen : Custom
4-4
-------
Section 4.0
Modeling Approach
3MRA Modeling System Setup
Select Components
Select the components setup will install
Select the components you want to install., and deselect the components you do not want to install
g Application Files
~VB5 Runtime
~ MDAC 2.5
~ ODBC 3.51
@BDE
'BUSH
Java Runtime Environment -
component required for Site
Visualization Tool (SVT)
1.67 GB of space required on the C drive
0.37 GB of space available on the C drive
Figure 4.7 3MRA Installation Options for Database Connectivity
Tools
Figure 4.8 displays the user screen for selecting the directory within which the 3 MR A modeling
system will be installed. Figure 4.9 displays a screen that communicates progress toward
installing the modeling system.
Setup will install 3MRA Modeling System in the following folder
To install to this folder, click Next. To install to a different folder, click Browse and select another
folder.
Destination Folder
Figure 4.8 3MRA Installation Target Directory
4-5
-------
Section 4.0
Modeling Approach
RA Modeling System Setup
E
Setup Status
3MRA Modeling System Setup is performing the requested operations.
Installing
d: \3M R A\E coE xposure. exe
r—
%
UijiljJ] jUiHllJ
|Cancel|
Figure 4.9 3 MR A "Installation in Progress" Screen
4-6
-------
Section 4.0
Modeling Approach
Figures 4.10 through 4.13 display screens associated with installing the Java Runtime
Environment (JRE). Figure 4.10 informs the user that the JRE installation files are being
extracted and readied for installation. Figure 4.11 displays the Licensing agreement for the JRE.
Figure 4.12 displays the installation screen for selecting the directory location for installation. It
is recommended that the default directory, shown on the screen, be selected for installation.
Figure 4.13 displays the screen for selecting the Browser to be used to display SVT graphical
outputs.
I he contents or this package are being extracted.
Please wait while the Installshield Wizard extracts the files needed to install Java 2
Runtime Environment on your computer. This may take a few moments.
Extracting data"!.cab...
F Java 2 Runtime Environment - InstallShield Wizard
InstallShield
| Next > | Cancel |
Figure 4.10 Java Runtime Environment Setup Screen
4-7
-------
Section 4.0
Modeling Approach
Java(TM) 2 Runtime Environment. Standard Edition, v1.3.1_08 Setup
License Agreement
Please read the following license agreement carefully.
Press the PAGE DOWN key to see the rest of the agreement.
Sun Microsystems, Inc.
Binary Code License Agreement
READ THE TERMS OF THIS AGREEMENT AND ANY PROVIDED
SUPPLEMENTAL LICENSE TERMS (COLLECTIVELY 'AGREEMENT")
CAREFULLY BEFORE OPENING THE SOFTWARE MEDIA PACKAGE. BY
OPENING THE SOFTWARE MEDIA PACKAGE, YOU AGREE TO THE TERMS
OF THIS AGREEMENT. IF YOU ARE ACCESSING THE SOFTWARE
ELECTRONICALLY, INDICATE YOUFl ACCEPTANCE OF THESE TERMS BY
SELECTING THE "ACCEPT" BUTTON AT THE END OF THIS
Do you accept all the terms of the preceding License Agreement? If you choose No, the
setup will close. To install Java Runtime Environment, you must accept this agreement.
InstallShield
< Back
Figure 4.11 Java Runtime Environment Installation Screen
Java(TM) 2 Runtime Environment, Standard Edition, v1.3.1_08 Setup
Choose Destination Location
Select folder where Setup will install files.
Setup will install Java Runtime Environment in the following folder.
To install to this folder, click Next. To install to a different folder, click. Browse and select
another folder.
Destination Folder
C:'¦.Program FilesVI avaS oft'J R E \1,3.1_08
IristallShield
Browse...
< Back
Next >
Cancel
Figure 4.12 Java Installation Target Directory Screen
4-8
-------
Section 4.0
Modeling Approach
Java(TM] 2 Runtime Environment. Standard Edition. v1_3_1_08 Setup
Select Browsers
Java(TM) Plug-in will be the default Java runtime for the following browser(s):
[Microsoft Internet Explorer;
V Netscape 6
You rinay change the default in the Java(TM) Plug-in Control
Panel.
Histall^^ffl
< Back
Next >
Cancel
Figure 4.13 Java Runtime Browser Selection Screen
Figure 4.14 displays the final installation screen that informs the user that files containing
directory path information have been modified to conform to the users directory specifications.
At this point the installation is complete and the user can proceed to invoke the 3MRA
modeling system by clicking the START menu and PROGRAMS/3MRA/SUI.
Subsequent invocations of the 3MRA installation software begin with the screen shown
in Figure 4.15. This screen allows user the following selections : 1) Adding or removing
components (as in a "Custom" setup) with "Modify", 2) Reinstalling the components previously
installed with "Repair", or 3) Removing the installation with "Remove". If the user selects
"Remove" only those 3MRA files originally installed will be removed. Files that have been
generated by the user, e.g., from performing simulations, will remain, along with the associated
directory structure.
4-9
-------
Section 4.0
Modeling Approach
3MRA Modeling System Setup
InstallShield Wizard Complete
HDPROD.SSF file path strings c:\hwir replaced with d:\3MRA HDPRDEX1 .SSF file path strings
c:\hwir replaced with d:S3MRA JRE installation complete. SVT batchfile updated for correct JRE
path.
To run the 3MRA system click on the SUI desktop icon and open the HDPROD.SSF or
HDPRDEX1.SSF header file located in the SSF subdirectory.
I Finish]
Figure 4.14 3MRA Final Installation Screen
3MRA Modeling System Setup
Modify, repair, or remove the program.
Welcome to the 3MRA Modeling System Setup Maintenance program. This program lets you
modify the current installation. Click one of the options below.
9 [Modify
iiM Select new program components to add or select currently installed components to
0 Repair
-ij§| Reinstall all program components installed by the previous setup.
Remove all installed components.
Figure 4.15 3MRA Modify/Repair/Remove Installation Screen
4-10
-------
Section 4.0
Modeling Approach
4.1.2 Directory Structure for 3 MR A
Upon completion of the 3MRA modeling system installation the complete set of files
needed to execute the system are placed onto the user specified harddrive (e.g., C:\). Figure 4.16
illustrates the Directory structure assuming the 3MRA modeling system has been installed on
D:\3MRA. The root directory (i.e., D:\3MRA) contains all the executable components of the
modeling system, including the seventeen science modules and the system processors. Below
the 3MRA directory are a series of subdirectories containing 3MRA databases, example
simulation output files, and software for the Site Visualization Tool (SVT) that facilitates
visualization of single site output GRF files. The following discussion provides a description of
the contents of each of the 3MRA subdirectories.
gj D:\3MRA
_
id
2£J
File Edit View Tools Help
E
a
n-1 Back " -~ ~ Search Cg] Folders yj
tS X ] Ht
Address ~ D:\3MRA
0
Folders X
Name
bize | Type
| Modified
A.
$-13 Data (D:)
j B 3MRA
j j Cj CPPData
; !-ri Database
$•¦£ 1 ExarnpleOutputs
0-IJ Permanent
B-Q WPO11400171 -55-6Cw4R 1
j~0 grf
LQ ssf
Cj psof
LI RSOF
I | ffl-Q grf
rQ MetData
; i -Q Permanent
j j j- Q Psof
j j I •{ 1 Rsof
j j |-Q ssf
I | i-Q svt _rj
~ CPPData
( 1 Database
i ~l ExarnpleOutputs
~ grf
~ MetData
( 1 Permanent
~ Psof
~ Rsof
~ ssf
~svt
MAirIO.exe
ilAauaticFoodWeb....
[""*1 AauiferlD.exe
3 EcoExposure. exe
^ EcoRisk.exe
I fcelpl.exe
jfeELP2.exe
File Folder
File Folder
File Folder
File Folder-
File Folder
File Folder
File Folder
File Folder
File Folder
File Folder
116 KB Application
32 KB Application
436 KB Application
60 KB Application
76 KB Application
490 KB Application
182 KB Application
6/27/2003 2:24 PM
6/27/2003 2:25 PM
6/27/2003 2:28 PM
6/27/2003 2:24 PM
6/27/2003 2:28 PM
6/27/2003 2:24 PM
6/27/2003 2:24 PM
6/27/2003 2:24 PM
7/7/2003 8:07 AM
6/27/2003 2:35 PM
3/11/2003 1:52 PM
2/6/2003 12:27 PM
4/17/2003 11:48 AM
7/25/2002 1:22 PM
1/22/2003 9:35 AM
1/9/2003 4:26 PM
5/7/2003 1:47 PM
d
« i I jJ
Type: Application Size: 312 KB
|312 KB
|jQI. My Computer
Figure 4.16 3MRA Directory Structure after Installation
4-11
-------
Section 4.0
Modeling Approach
Figure 4.17 illustrates the contents of the Chemical Properties Data subdirectory
(D:\3MRA\CPPData). There are 15 individual chemical property files, each containing a
specific type of chemical data. The data files are comma separated ASCII data files and can
easily be viewed using a spreadsheet program (e.g., Microsoft Excel).
E? D:\3MRA\CPPData
-iQl xl
File Edit View lools Help
•^Back " ^ | Cj, Search | Lgj Folders ^
J RS tg x I !*
n
Address |Q D:\3MRA\CPPData
_»] i^Go
Folders
E-O Data (D:)
&-Q 3MRA
¦Q
CD Database
El CH ExampleOutputs
El CD Permanent
B-Q WPO11400171 -55-6C|ai4R 1
l -Q grf
L-D ssf
•Q P50F
Cl RSOF
S O grf
-•••Cl ifo
; CD MetData
rQ Permanent
i -Q Psof
i- Q Rsof
I -Q ssf
&Gl SVT
j-Q GNUPlot
ei-Q gov
& C3 epa
fi-Q hwir
Q util
Name
Size
Type
j Modified
i^J ABF.csv
5 KB
CSV File
5/17/2002 5:53 PM
!j|| ActBio.csv
1 KB
CSV File
4/19/1999 2:31 PM
g| AerBio.csv
2KB
CSV File
5/12/1999 4:57 PM
g[| AnaBio.csv
12 KB
CSV File
5/12/1999 4:57 PM
gj AnaRed.csv
1 KB
CSV File
3/16/1999 3:56 PM
|jf| CAT.csv
37 KB
CSV File
6/22/1999 11:27 AM
g ChemEco.csv
2KB
CSV File
9/30/1999 8:07 AM
|i§] cw.csv
6 KB
CSV File
6/23/1999 2:09 PM
iff] EB.csv
182 KB
CSV File
9/30/1999 8:05 AM
IjEBF.csv
9 KB
CSV File
11/20/2001 10:05 AM
0 HHB.csv
6 KB
CSV File
5/24/1999 2:58 PM
M\ MethBio.csv
15 KB
CSV File
6/17/1999 6:20 PM
0 MICP.csv
16KB
CSV File
4/16/2001 11:07 AM
!l| OCP.csv
7 KB
CSV File
3/16/1999 4:45 PM
§f| 504Bio.csv
15 KB
CSV File
6/17/1999 6:19 PM
307 KB My Computer
15 object(s) (Disk free space: 5.70 GB)
A
Figure 4.17 3MRA Directory Structure : CPPData Subdirectory
4-12
-------
Section 4.0
Modeling Approach
Figure 4.18 illustrates the contents of the D:\3MRA\Database that contains the three main
data files needed to populate the Site Simulation Files (SSFs). The three data files (Site,
Regional, and National) are Microsoft Access data files and can be viewed and edited with this
program.
El D:\3MRfl\Database
File Edit View Tools Help
Back t 4 ' it] fjj Search £gj Folders Q) I ^ X ^
Address |_I] D:\3MRA\Database
H if^Go
Folders
B S) Data (D:)
e-Q 3MRA
Q CPPData
B I I ExampleOutputs
B'Q Permanent
B-D WP011400171-55-6Cw4R1
¦Cl grf
Q ssf
D P50F
C] RSOF
BQ grf
LQifo
| -C3 MetData
Q Permanent
Q Psof
G Rsof
Q ssf
B- O SVT
!¦¦¦¦ Ci GNUPIot
B Q gov
BQ epa
BQ hwir
Q utii
®National(33-28-03,mdb
Ej Regional 11-18-01. mdb
©5ite03-28-03.mdb
; I T>'Pe
1 Mod
6,024 KB Microsoft Acc...
438 KB Microsoft Acc...
544,940... Microsoft Acc,..
3/26
11/1
3/26
J
±1
" |538 MB m My Computer
3 object(s) (Disk free space: 5.70 GB)
Figure 4.18 3MRA Directory Structure : Database Subdirectory
4-13
-------
Section 4.0
Modeling Approach
Figure 4.19 illustrates the contents of the D:\3MRA\ExampleOuputs directory. This
directory contains the results of two 3MRA simulations; a set of results for a single site
simulation and results from a national simulation to establish exemption levels for benzene. The
Permanent subdirectory contains the results for a 3MRA simulation where a single
site/chemical/WMU/Cw was selected. When the results for individual sites are saved the 3MRA
modeling system places the resulting GRF and SSF files in subdirectories under a directory
named for the site simulation (for example, as shown in the figure the subdirectory
"WP011400171-55-5Cw4Rl", this simulation involved a Waste Pile (WP), at site number
0114001, for the chemical whose CAS number is 71-55-6, and for Cw number 4. These
example site simulation results are included in the 3 MR A modeling system installation to allow
the user to execute the Site Visualization Tool (SVT) described in Section 4.3.
Additionally, the ExampleOuputs directoiy contains subdirectories "PSOF" and "RSOF".
The RSOF subdirectory contains the results of a national simulation that includes all sites, all
WMTIs, and all Cw's for the chemical benzene. Thus, the data files contained in the RSOF
subdirectory can be used to execute the ELP II processor (See Section 4.3). Results of running
the ELP II are tables listing national exemption levels. These results are stored in the PSOF
subdirectory.
I D:\3MRA\ExampleOutputs
-Inl *1
File Edit View Tools Help
vJaBack - ¦=> - Q Search l^-. Folders | %jj X ^ |
Address |Q] D:\3MRA\ExarnpleOutputs
(j^Go
Folders
ExampleOutputs
O Data (D:)
B O 3MRA
h"Q CPPData
i ¦ {_] Database
K -_ij
£3 C] Permanent
: a-Q WP011400171-55-6CW4R1
i-Q grf
j L Q ssf
111 PSOF
C] R50F
a-Q grf
LQ ifo
Q MetData
LJ Permanent
Q Psof
CD Rsof
¦Q ssf
[El Ci SVT
j "C3 GNUPlot
B' Q gov
S CJ epa
B-Q hwir
L -Q util
Size | Type
"I Go to "D:\3MRA\ExampleQutpu
Cl] Permanent
Qpsof
~ rsof
1^ MoveFiles.bat
File Folder
File Folder
File Folder
1 KB MS-DOS Batch File
6/27/2003 2:28 PM
6/27/2003 2:28 PM
6/27/2003 2:28 PM
5/29/2003 6:26 PM
J
jJ
|44 bytes My Computer
4 object(s) (Disk free space: 5.70 GB)
Figure 4.19 3MRA Directory Structure : ExampleOutputs Subdirectory
4-14
-------
Section 4.0
Modeling Approach
Figure 4.20 illustrates the contents of the D:\3MRA\grf subdirectory. Included in this
subdirectory are the Dictionary files for each of the GRFiles associated with the 3MRA science
modules. These files are ASCII data files and can be viewed with any TextEditor (e.g.,
WordPad - do not edit these files with a Word Processor). Also contained in the GRF
subdirectory is the "lfo" subdirectory that contains the normalized output files of the atmospheric
module for each site and WML" combination. These files are created the first time the
atmospheric module is executed for a site/WMU combination. Any subsequent simulations that
include this site/WMU combination will by-pass the atmospheric module and process the
normalized data contained in these data files.
El D:\3MRfl\grf
-js xj
File Edit View Tools Help
* Back
?¦£] Q Search £g| Folders yj %« X ^ I H*
Address [~ D:\3MRA\grf
' | ;£VGO
Folders
B O Data (Di)
0Q 3MRA
;¦ "Cl CPPData
: " Q Database
~¦¦Q ExarnpleOutputs
Fl-n Permanent
&~ WP011400171-55-6Cw4Rl
D grf
Q ssf
I -C3 PSOF
¦ Q RSOF
b-€sBI
LQ lfo
Q MetData
{ ] Permanent
~ Psof
Cl Rsof
!-D ssf
&~ SVT
j Q GNUPIot
ErCll gov
B"Q epa
~ Q hwir
Cl util
Name
Size
Type ^
| Modified
|W| ws.dic
2KB
Text Document
3/28/2003 2:01 AM
Vz.dic
1 KB
Text Document
3/28/2003 2:00 AM
P tf .die
7 KB
Text Document
3/28/2003 2:00 AM
f|] 5W.DIC
3 KB
Text Document
3/28/2003 2:12 AM
1] SR.DIC
4KB
Text Document
3/28/2003 2:00 AM
W| SL.dic
1 KB
Text Document
3/28/2003 2:00 AM
1] HR.DIC
5 KB
Text Document
3/28/2003 2:00 AM
If] HE,DIC
12 KB
Text Document
3/28/2003 2:00 AM
W Hd.dic
1 KB
Text Document
3/28/2003 2:00 AM
1 FF.DIC
5 KB
Text Document
3/28/2003 2:12 AM
J] ER.DIC
4KB
Text Document
3/28/2003 2:00 AM
D EE,DIC
1 KB
Text Document
3/28/2003 2:00 AM
J] AR.DIC
3 KB
Text Document
3/28/2003 2:12 AM
0 AQ.DIC
2KB
Text Document
3/28/2003 2:00 AM
[1] af.dic
3 KB
Text Document
3/28/2003 2:00 AM
~ lfo
File Folder
6/27/2003 2:24 PM
J
Jj
|44.2 KB My Computer
16 object(s) (Disk free space: 5.70 GB)
Figure 4.20 3MRA Directory Structure : GRF subdirectory
4-15
-------
Section 4.0
Modeling Approach
Figure 4.21 illustrates the contents of the D:\3MRAVMetData subdirectory. All meteorological
data files are contained here, including hourly, daily, monthly, annual, and long-term average
data. These data files are ASCII data files and can be viewed with a TextEditor.
| D:\3MRA\MetData
^JnjxJ
File Edit View Tools Help
v3 Back
pq Q Search CgjFolders | IJa1 X ^ | 0T
Address |Q D:\3MRfl\MetData
"3
B"S Data (D:)
&Q 3MRA
¦Q CPPData
Q Database
S-Q ExampieOutputs
B'Q Permanent
B CD WP011400171-S5-6QW4R1
I i-Cj grf
LQ ssf
| Q PSOF
Q RSOF
BCD grf
Cj i-c
Q Permanent
; G Psof
¦¦-C] Rsof
C) ssf
a o svt
Q GNUPIot
a-D gov
HI 1 epa
B -Q hwir
L-Q utii
Size
iwj94860m.dat
|»| 94860l.dat
W] 94860h.dat
09486Od.dat
g] 94860a.dat
094847m.dat
2l94847l.dat
094847h.dat
g]94847d.dat
2 94847a.dat
^94846m.dat
S] 94846l.dat
94846h.dat
g] 94846d.dat
*194846a.dat
jS]94823m.dat
094823l.dat
|S] 94823h.dat
94823d.dat
ji] 94823a.dat
94822m.dat
3 94822l.dat
«] 94822h.dat
r?< , .
JL
I Tvpe
6 KB
1 KB
7,535 KB
285 KB
1 KB
7 KB
1 KB
9,793 KB
373 KB
1 KB
10 KB
1 KB
14,314 KB
543 KB
2KB
12 KB
1 KB
17,326 KB
652 KB
2KB
8KB
1 KB
11,301 KB
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
DAT File
Modified
1/12/1999 11:15 AM
1/12/1999 12:32 PM
1/5/1999 9:01 PM
1/12/1999 12:43 PM
1/12/1999 11:55 AM
12/22/1998 9:44 AM
12/22/1998 10:12 M1
12/22/1998 1:04 PM
12/22/1998 11:44 AN
12/22/1998 9:53 AM
1/12/1999 11:15 AM
1/12/1999 12:32 PM
12/23/1998 3:33 PM
1/12/1999 12:43 PM
1/12/1999 11:55 AM
1/12/1999 11:15 AM
1/12/1999 12:31 PM
12/23/1998 7:31 PM
1/12/1999 12:43 PM
1/12/1999 11:54 AM
1/12/1999 11:15 AM
1/12/1999 12:30 PM
12/30/1998 2:03 PM
¦ rm i
jlI
495 object(s) (Disk free space: 5.70 GB)
My Computer
Figure 4.21 3 MR A Directory Structure : MetData Subdirectory
4-16
-------
Section 4.0
Modeling Approach
Figure 4.22 illustrates the contents of the D:\3MRA\PSOF subdirectory. This subdirectory
contains Microsoft Access database files that contain the standard (or user specified) regulatory
"scenarios" (ELP2.mdb) and a database file that contains the table format for output results of
executing the ELP II Processor. Also, contained in this subdirectory (but not shown here) are the
actual tables produced by the ELP II. These files are produced in Microsoft Excel format.
EJ D:\3MRA\Psof
File Edit View lools Help
Back ~
[jtj ^Search Folders *2) | &Sa X ^ | H'
Address |fj D:\3MRA\Psof
~ | f^Go
I Modified~
Folders
EhS Data (D:)
$ Q 3MRA
h C] CPPData
!¦¦¦¦{_] Database
0 D ExampleOutputs
E] i_J Permanent
El Q WP011400171-55-6CW4R1
! -C3 orf
'• D ssf
Cl PSOF
Q RSOF
grf
CJ Ifo
P I MetData
C] Permanent
¦-¦Pi Rsof
'¦¦¦¦ Q ssf
&Q svt
i -C] GNUPIot
B'CH gov
FlP 1 epa
S'O hwir
D utii
®PSOF,mdb
E] ELP2.mdb
Size 1 Type
150 KB Microsoft Access Ap...
92 KB Microsoft Access Ap..,
5/25/1999 11:53 AM
4/19/2001 12:59 PM
J
li
242 KB 10 My Computer
2 object(s) (Disk free space: 5.70 GB)
A
Figure 4.22 3MRA Directory Structure : PSOF subdirectory
4-17
-------
Section 4.0
Modeling Approach
Figure 4.23 illustrates the contents of the D:\3MRA\RSOF subdirectory. This subdirectory
contains a Microsoft Access database file the includes "templates" that describe the Risk
Summary Output File format. Not shown here (but included in the
D:\3MRA\ExampleOutputs\RSOF subdirectory) are Microsoft Access database files containing
the ELP I results from national simulations. These database files contain the input data for
executing the ELP II Processor.
|fi< D:\3MRfl\Rsof
_Jn]x]
File Edit View Tools Help ^ JJj
-J=Back - ¦+ - f*3 Q Search Folders y
» ] tt <6 X « I H*
Address |Q] D:\3MRA\Rsof
Folders
B-lsP Data (D:)
fi-Q 3MRA
Q CPPData
f"l Database
[=] -Q ExampleOutputs
F1 C~l Permanent
BQ WP011400171-55-6Cw4Rl
Qgrf
ssf
¦Q PSOF
Cl RSOF
a-O grf
O ifo
n MetData
Permanent
r-Q Psof
|-D ssF
6 Q svt
Ci GNUPIot
a-Q gov
R-j I epa
&D hwir
lO utii
I
T ernplateRSOF. mdb
Size | Type
Modified
480 KB Microsoft Access Ap... 3/21 /2000 1:23 PM
jJ
1 object(s) (Disk free space: 5.70 GB)
480 KB g My Computer
Figure 4.23 3MRA Directory Structure : RSOF subdirectory
4-18
-------
Section 4.0
Modeling Approach
Figure 4.24 illustrates the contents of the D:\3MRA\ssf subdirectory. This subirectory contains
the Dictionary files for the Site Simulation Files (SSFs). These files are ASCII data files and
can be viewed with a TestEditor. This directory also contains the individual SSFs resulting from
a 3MRA simulation.
-Inl *1
File Edit View Tools Help
ifi |
Back
" > ' B ^Search 1Folders \'J
| <£ X « I
Address
D D:\3MRA\ssf
~ j i£v*Go
Folders
X
Name
Size |
Type V
| Modified
1
! BO Data (D:) ¦»
0 ws.dic
2KB
Text Document
3/28/2003 2:00 AM
1 1 S-Q 3MRA
0 wp.dic
4KB
Text Document
3/28/2003 1:59 AM
! CD CPPData
0 Vz.dic
1 KB
Text Document
3/28/2003 1:59 AM
| Datahacp
0 tf.die
3 KB
Text Document
3/28/2003 2:17 AM
Hi 1 ExampleOutputs
0 SW.DIC
2KB
Text Document
3/28/2003 1:59 AM
| B O Permanent
0 SL.dic
20 KB
Text Document
3/31/2003 12:02 PM
: i-Cj WP011400171-55-6Cw4Rl
0 SI.DIC
2KB
Text Document
3/28/2003 1:59 AM
i ! Q grf
0 If.die
4KB
Text Document
3/28/2003 1:59 AM
! ! D ssf
0 la.die
5 KB
Text Document
3/28/2003 1:59 AM
Cj psof
0 HR.DIC
1 KB
Text Document
3/28/2003 2:14 AM
i :'-Q R50F
!] HE.DIC
9 KB
Text Document
3/28/2003 1:59 AM
fi-Q grf
0 Hd.dic
5 KB
Text Document
3/28/2003 1:59 AM
1 a ifo
0 FF.DIC
4KB
Text Document
3/28/2003 2:10 AM
; C] MetDdta
(WJ er.dic
1 KB
Text Document
3/28/2003 1:59 AM
I ¦ LJ Permanent
[§J ee.dic
2KB
Text Document
3/28/2003 1:59 AM
; CD Psof
@0 CP.die
15 KB
Text Document
3/28/2003 1:59 AM
i-Q Rsof
0 AT.die
2KB
Text Document
3/28/2003 1:59 AM
0 Ar.dic
3 KB
Text Document
3/28/2003 1:59 AM
e-Q svt
0 aq.dic
2KB
Text Document
3/28/2003 1:59 AM
j-™Q GNUPlot
0 af.dic
3 KB
Text Document
3/28/2003 1:59 AM
B"Q gov
!jj] Hdprod.ssf
3 KB
SSF File
6/27/2003 2:28 PM
S 'CJ epa
§§] Hdprdexl.ssf
5 KB
SSF File
6/27/2003 2:28 PM
EI-Q hwir
j-C3 util |j
<1
I
±1
|22 object(s) (Disk free space: 5.70 GB)
88.3 KB
P, My Computer
m§
Figure 4.24 3MRA Directory Structure : SSF subdirectory
4-19
-------
Section 4.0
Modeling Approach
Figure 4.25 illustrates the content of the D:\3MRA\SVT subdirectory. This subdirectory
contains the software related to the Site Visualization Tool. This software is used to produce
time series plots for single site simulation results. This execution of this software is described in
Section 4.3.
D:\3MRA\SVT
JS|xj
File Edit View Tools Help
v3 Back - - gj | Q Search Folders I ^ X ^
Address |Z] D:\3MRA\SVT
T | ^Go
S-13 Data (D:)
$ Q 3MRA
CD CPPData
Q Database
B-Q ExampleOutputs
Q Permanent
I &G2 WPO11400171 -55-6Cw4R 1
! Q grf
D ssf
•Q P50F
Q RSOF
B-Q grf
I ; Q ifo
i - Q MetData
O Permanent
a Psof
J-Q Rsof
i-Q ssf
i CD GNUPlot
S-Q gov
0-Q epa
il -D hwir
j--C3 util
|W| Readme.txt
H3sVT.bat
|»1 wgnuplot.rnnu
e^wgnuplot.hlp
1^1 wgnuplot.GID
H) prefs.ini
wgnuplot.dll
%] javaHWIRIO.dll
%] hwirio.dll
[Qwgnuplot.exe
CDgov
Cl GNUPlot
Size
Type V
1 Modified
1 KB
1 KB
12 KB
392 KB
49 KB
1 KB
117 KB
237 KB
261 KB
728 KB
Text Document
MS-DOS Batch File
MNU File
Help File
GID File
Configuration Settings
Application Extension
Application Extension
Application Extension
Application
File Folder
File Folder
11/26/2002 1:41 PM
6/27/2003 2:33 PM
11/22/2002 4:21 PM
11/22/2002 4:21 PM
11/22/2002 4:21 PM
6/27/2003 2:40 PM
11/22/2002 4:21 PM
11/22/2002 4:21 PM
11/22/2002 4:21 PM
11/22/2002 4:21 PM
6/27/2003 2:28 PM
6/27/2003 2:28 PM
J
11.75 MB IB" My Computer
12 object(s) (Disk free space: 5.70 GB)
Figure 4.25 3MRA Directory Structure : SVT subdirectory
4-20
-------
Section 4.0
Modeling Approach
4.2 Executing the 3MRA Modeling System
This section describes the sequence of steps necessary to execute a 3MRA assessment.
A 3MRA assessment consists of a combination of chemical(s), site(s), waste mangement
unit(s), wastestream concentration(s),
and Monte Carlo iteration(s). The
number of sites available for an
assessment is dependent upon the WMUs
selected. Each site contains one or more
of the WMUs but no site contains all
WMUs, thus, depending on the WMU
selected the number of sites available
will vary. The adjoining table lists the
WMUs and the number of sites that
contain the particular unit type. National
scale assessments always include at least
one WMU, all sites containing the
WMU, all wastestream concentrations
(Cw's), at least one chemical, and at least
one Monte Carlo iteration. Typical
national assessments include all
combinations of sites and WMUs along
with all Cw's and one chemical and one
Monte Carlo iteration. The total number of site simulations in this case is 2095 (i.e., 419
site/WMU combinations times 5 Cw's). The typical national scale assessment takes on the order
of 2 to 5 days for a modern Personal Computer (PC) to complete, depending on the chemical
(less volatile, more hydrophobic chemicals remain in the environment near the source for longer
periods of time, thus requiring more simulation time). The total number of simulations would
scale proportionately for each additional chemical and Monte Carlo iteration (e.g., 100 Monte
Carlo iterations would require a total of 209,500 individual site simulations). Note that each
Monte Carlo iteration implies that the total set of site simulations will be repeated the specified
number of times.
Smaller scale assessments are typically conducted to either perform a test of the modeling
system or, more typically, to review the results of a single site simulation. Any combination of
3MRA sites, chemicals, WMUs, Cws, and Monte Carlo iterations is possible. In the case when a
user wishes to review the results of a single site the user may specify a single combination of a
site, chemical, waste management unit, wastestream concentration, and a single Monte Carlo
iteration. When executing a single site assessment it is possible to store the results of each
science module. This provides the user with the ability to follow a simulation step by step from
source release to risk.
Waste Management
Unit (WMU) Type
Number of Sites
Containing
WMU
Landfill
56
Land Application Unit
28
Surface Impoundment
137
Aerated Tank
137
Waste Pile
61
All WMU Types
419*
* There is a total of 419 combinations of
WMU and site location.
4-21
-------
Section 4.0
Modeling Approach
NOTE: 3MRA Version 1.0 is not designed to facilitate the definition of "new" sites
by the user. That is, there is no user interface that allows the user to input data
associated with a new site. Sites currently contained in the Site Database represent a
specific collection of sites used to conduct a national assessment of risks from land-
based waste management disposal units. While it is possible to input a new site (i.e.,
non-3MRA national assessment site) the process requires intimate knowledge of the
Microsoft Access database structure used for 3MRA. Information for defining new
sites is not presented in this document. The next version of 3MRA will include a
user interface to facilitate definition of new sites.
The combinations required for each assessment are communicated to the modeling
system, by the user, via the System User Interface(SUI). The users view of the 3MRA modeling
system is presented through a series of SUI menu screens that present simulation options to the
user and accept user choices. Section 4.2.1 presents the complete set of SUI menu screens while
sections 4.2.2 and 4.2.3 describe, by way of an example simulation, the output processing
capbilities of the 3MRA modeling system. The first example is of a National assessment to
establish an exemption level for benzene. The processing of National simulation results by the
Exit Level Processor II/Risk Visualization Processor (ELP II/RVP) are described. The second
example is the execution of a single combination of a site, chemical, waste management unit,
and wastestream concentration level. The Site Visualization Tool (SVT), capable of producing
time series plots of all science modules, is described.
4.2.1 System User Interface (SUI)
There are three main functions of the 3MRA System User Interface; system
configuration, system management, and system status. The system configuration function defines
the locations of all files required to run a simulation (database files and processor and module
executables). This function also specifies where the data files that are produced by the system
are to be stored. The system management function defines which sites, chemicals, source types,
and waste levels are to be included in the simulation, as well as the storage level of output, the
number of iterations, and the run mode. The system status function allows the user to start, stop,
reset, and resume the simulation. It also reports any errors or warnings that may have occurred
during the simulations.
The SUI menu screens include a few general features that facilitate information
exchange between the user and the modeling system. These features, illustrated in Figure 4.26,
include "tabs", "buttons", and "text windows".
4-22
-------
Section 4.0
Modeling Approach
Tab
Textbox
^ 3MRA - riuicimedia Multipathway Multireceptor Kisk Assessment
File
System Configurafisih i| System Management | System Status
Databases | Directories^ Processors | MMSP Modules j
Button
-=lQjiU
Site-based
Regional distributions
National distributions
Static regional
Static national
d:\3MRA\Database\fc ite03-28-U3.mdb
d:\3M RA\Databas^\Regional11 -13-01 .mdb J. |
d:\3MRA\Database\National03-28-03.mdb
d:\3MRA\Database\Regional11 -18-01 .mdb ... |
d:\3MRA\Database\N ational03-28-03.mdb
Figure 4.26 User Interaction Features of 3MRA Menu Screens
Tabs represent categories of data/information and when activated display a menu screen
designed to present and gather information relevant to the data category. Tabs may simply
represent options available to the user or they represent a logical hierarchical data structure.
Tabs are activated with a single left click of the mouse.
Text windows accept information/data from the user in one of two ways. First, the user
can simply type the relevant information/data directly into the window. The second type of text
window is populated with information from the modeling system (e.g,. a list of 3MRA sites) and
requires the user to select specific entries.
Buttons invoke specific actions on the part of the SUI. There are only a few specific
buttons included in the SUI.
The
button is used to browse file/directory locations. Upon a
single click of this button the user is presented access to the directory of files on the host
computer. This button is provided to facilitate the users population of specific text windows that
require file name and directory location data. By double clicking on a file name it will appear
(with all relevant directory relationships) in the text window. Other buttons include Start
Reset Resume Stop Run which invoke intuitive actions with respect to the simulation
execution.
4-23
-------
Section 4.0
Modeling Approach
Pre Simulation Preparations
One final point before describing the SUI functionality. Subsequent to the initial
simulation of the 3MRA modeling system but before each subsequent simulation it is
recommended that the user run a clean up procedure to remove files left over from previous
simulations. Double clicking on the 3MRACleanUp.bat file (shown in Figure 4.27) starts the
cleanup routine. The routine deletes residual Site Simulation Files (SSFs) and Global Results
Files (GRFs) as well as residual Risk Summary Output Files (RSOFs). In addition, it deletes
Error and Warning files left from the previous runs. Lastly, the cleanup routine compares the
files in the SMRAVerl .txt manifest file to the files in the application folder. This check is to
ensure that the official version of the 3MRA software system is installed. If, for example, the
user has updated a database or modified a science module, the 3MRA system will no longer
represent that which was installed. The cleanup procedure reports any differences by writing to
the Messages.all file in the 3MRAAGRF directory.
Ell D:\3MRA
File Edit View Tools Help
<4= Back - -~ ~ g] ^Search Folders ^ K ^
Address _J| D:\3MRfi
Folders
&-(Rl My Computer
01^ 31A Floppy (A:)
ED System (C:)
B '=) Data (D:)
B-Q
Q CPPData
I Database
ra _J GRF
; Q MetData
Fl-f I Permanent
Q PSOF
fj RSOF
Q SSF
d
Name 1
Size
Type
O CPPData
File Folder-
1 Database
File Folder
Qgrf
File Folder
LJ MetData
File Folder
1 Permanent
File Folder
Q PSOF
File Folder
Qrsof
File Folder
~ SSF
File Folder
Cj03MRA.exe
311., 122 KB
Application
0 3MR ACleanUp. bat
3 KB
MS-DOS Bat
if] 3MRAVerl.txt
80 KB
Text Docum
""lAirIO.exe
116 KB
Application
4
j
^2) My Computer
Type: MS-DOS Batch File Size: 2.63 KB
2.63 KB
Figure 4.27 3MRA Directory Listing Displaying Batch File for System Cleanup
4-24
-------
Section 4.0
Modeling Approach
4.2.2 Invoking the 3MRA Modeling System
To invoke the 3MRA modeling system the user enters the Windows Start menu and
selects PROGRAMS/3MRA/SUI. The first screen that appears is shown in Figure 4.28. The
user then clicks on File and then Open. This opens the directory listing window for the purpose
of having the user locate and identify the "header file" to be used to establish the default settings
for the SUI (Figure 4.29). The header file contains all the information necessary for the
modeling system to locate files, databases, science modules, and input/output files necessary to
execute the 3MRA. Any information/data contained in the header file can be modified through
the SUI. Under the standard installation procedure described in Section 4.1 the header file is
located in C:\3MRA\SSF. The user locates this file, highlights the file by left clicking on it, and
then clicking on the Open button (or simply double left clicks on this filename). The SUI then
V- 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
Open
Save F2
Save As
Close
Exit
System Management | System Status |
ctories | Processors | MMSP Modules |
Regional distributions
National distributions
Static regional
Static national
jnjxj
"J
"J
B
"j
-j
Figure 4.28 Initial 3MRA User Interface Screen
Open 3MRA Sytem Configuration file
Look in:
Ssf
|a]iHdprod.ssf
Hdprod20.ssf
File name: |
Open
Files of type: | SUI Config files (HD".SSF)
Cancel
Figure 4.29 3MRA Header File Selection Screen
4-25
-------
Section 4.0
Modeling Approach
reads the file and populates all menu screens with appropriate data from the header file. From
this point on the 3MRA menu screens will contain default information/data. If the user modifies
any information, the updated header file can be saved by pressing the File key, and then the Save
as key. It is convenient to save all header files in the same C:\3MRA\SSF file directory.
4.2.2.1 System Configuration
The system configuration function is managed through the SYSTEM
CONFIGURATION SCREEN of the SUI that, in turn, has several additional subscreens linked
with it. The primary purpose of the series of configuration screens is to allow the user to select
and define the directory location of all files necessary to execute the 3MRA modeling system.
When the "header file", described in the previous section, is accessed and read by the SUI there
is a complete set of file locations read and placed as defaults on the system configuration
screens. So, in essence, the user rarely exercises these screens except to make modifications
(e.g., the user wishes to use a module or database located outside the standard installation
directories).
There are four main categories of file directory information required, represented by the
second tier of tabs (subscreens) below the System Configuration tab in Figure 4.30 and
including Databases, Directories, Processors, and MMSP Modules. Each of these categories are
explained below with a graphic of the screen to aid understanding.
Databases SubScreen
The Databases subscreen shown in Figure 4.30 lists the site related databases required for
3MRA. While five site databases are listed on the screen only three are currently implemented
in the 3MRA modeling system. The Site-based, Regional Distribution, and National
Distribution databases are implemented and must be specified. As noted above the
button can be used to facilitate locating and assigning database files in the appropriate textboxes.
The Static Regional and Static National are not currently utilized by the modeling system (they
represent place holders for a future version of the modeling system) but must contain entries
nonetheless. For purposes of convenience the Regional Distribution database and the National
Distribution database are repeated here to allow the system to function.
4-26
-------
Section 4.0
Modeling Approach
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
¦AD\2Sl
S^em Config^ationjl System Management | System Status
Databases | Directories | Processors | MMSP Modules |
Site-based
| D:\3MFlA\Database\Site11 -18-01 .rridb
Regional distributions
| D: \3M RA\D atabase\R egionall 1 -18-01. rndb
National distributions
|D:\3M RA\Database\N ationall 1 -18-01. rndb
Static regional
|D:\3MR A\Database\Regional11-18-01.mdb
Static national
|DA3M RA\D atabase\N ationall 1 -18-01. rndb
Figure 4.30 3MRA SUI: System Configuration Screen
Directories SubScreen
The Directories subscreen (Figure 4.31) of the SYSTEM CONFIGURATION SCREEN
requires the user to input the Directory location specifying where various sets of data files reside.
The system data files that need to be defined are illustrated in Figure 4.31 and are as follows:
• Site Simulation Files (SSF)
• Global Results Files (GRF)
• Risk Summary Output Files (RSOF)
• Protection Summary Output Files (PSOF)
• Permanent Storage
• Chemical Properties Database
• Meteorological Database
The data files contained in the SSF, GRF, RSOF, and PSOF are produced by the system
during a simulation. The SSF directory contains all input data for the science modules that is
extracted from the Site/Regional/National databases and prepared in SSF format for module
4-27
-------
Section 4.0
Modeling Approach
consumption. The GRF directory contains all the module output files, each being a unique GRF
file. Also contained in both the SSF and GRF directories are the Dictionary files that
correspond to each science module input file (SSF) and output file (GRF). The RSOF directory
contains the Risk Summary Output Files generated by the ELP I. The PSOF directory contains
the Protective Summary Output files generated by the ELP II. The Permanent Storage directory
contains an archive of all SSF and GRF files resulting from a simulation. These files are saved
only if the user specifies a maximum storage option (discussed below). It is typically the case
that the storage option is used only in the case of a single site assessment. Storing all SSF and
GRF files for a national scale assessment is not possible with current memory limits on the
Personal Computers (it would require on the order of several gigabytes for the minimum national
assessment and terabytes of storage for a national assessment involving multiple chemicals and
hundreds of Monte Carlo iterations). The Chemical Properties directory points to the location
where all chemical property data is stored. The Meteorological Data directory points to the
location where all meteorological data reside.
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration || System Management | System Status |
Databases Directories | Processors | MMSP Modules]
SiteSimulation File
Global Results File
Risk Summary Output
Protective Summary Output
Permanent Storage
Chemical Properties
Meteorological Data
D:\3MRA\SSF
D:\3MRA\GRF
D:\3MRA\RS0F
D:\3MRA\PS0F
D:\3MRA\Permanent
D:\3MRA\CPPData
D:\3MRA\MetData
_-iS]x]
"J
'J
'J
"J
'J
"J
'J
Figure 4.31 3MRA SUI: Directories Configuration Screen
4-28
-------
Section 4.0
Modeling Approach
Processors SubScreen
Figure 4.32 displays the Processor subscreen of the System Configuration Screen. A
total of six system processors are listed on the screen but only four are implemented at this time.
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
I System Configuration ;| System Management | System Status |
Databases] Directories Processors | MMSP Modules |
D: \3M R A\D ummyProc. bat
D:\3MRA\sdp.eae
Distribution statistics processor
B Use DSP
Site definition processor
Computational optimization processor
r UseCDP
Multimedia site simulation processor D:\3M RA\MMSP.exe
D: \3M R A\D umrnyProc. bat
Ewit level I processor
Exit level II processor
r Use ELP II
D:\3MRA\ELP1.bat
D:\3M RA\ELP2.exe
Figure 4.32 3MRA SUI: Processor Configuration Screen
JnJxJ
'J
'J
'J
'J
'J
'J
4-29
-------
Section 4.0
Modeling Approach
The following table describes the processors and their current implementation status.
PROCESSOR
FUNCTION
STATUS
Distribution
Statistics Processors
(DSP)
Facilitates Second Stage
Monte Carlo Sampling
and Simulation
To be
implemented in a
future version.
Site Definition
Processor (SDP)
Transfers site data from
Databases to Site
Simulation Files
Implemented
Computational
Optimization
Processor (COP)
Facilitates optimization of
simulations.
To be
implemented in a
future version.
Multimedia
Multipathway
Simulation
Processor (MMSP)
Manages the execution of
3MRA science modules.
Implemented
Exit Level Processor
I (ELP I)
Transforms and stores
human and ecological risk
data.
Implemented
Exit Level Processor
II (ELP II)
Develops tabular and
graphical representations
of exemption levels based
on user requests.
Implemented
(However, the
ELP II is not
invoked through
this SUI).
The screen includes three small textboxes to the left of the DSP, COP, and the ELP II.
These boxes are checked IF the processor is active (which they are not at this point in time). For
the DSP and COP the textboxes specifying the file locations are filled with a filename of
"DummyProc.bat". This filename signals to the system that the processor is not active. The
SDP, MMSP, and the ELP I must be specified by the user. The ELP II must be specified on this
screen even though it is not invoked during a simulation (this is simply an item which has not yet
been removed from the screens).
4-30
-------
Section 4.0
Modeling Approach
The MMSP Modules SubScreen
The MMSP Modules subscreen is linked to four lower level subscreens, each subscreen
collecting directory information about a group of science modules. The four groups of modules
specified by the user to run simulations are the source, transport, food chain, and exposure/risk
modules. The user clicks on the MMSP MODULES subscreen to specify these module data.
Figure 4.33 illustrates the source subscreen. Directory locations and file names for the
executable file representing each of the five source types included in the MMSP are specfied
here. Similar screens are included for the transport, food chain, and exposure/risk modules.
4.2.3 System Management
The functions of the SYSTEM MANAGEMENT SCREEN allow the user to define the
sites, chemicals, source types, and waste levels for simulation. It also allows the user to select
the level of storage for MMSP modules input (SSF) and output (GRF) files, the mode of
operation, the number of Monte Carlo iterations, and seed value for the statistical analysis of the
simulation. Each of these functions is accessed by the user through subscreens associated with
the SYSTEM MANAGEMENT SCREEN. These functions are described in the following
subsections.
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
Jnjxj
System Management | System Status |
Databases | Directories] Processors MMSP Modules j
Source j Transport | Foodweb | Exposure/Risk |
Land application unit module d:\3MRA\LAU.bat
d:\3MFlA\WP.bat
Waste pile module
Surface impoundment module d:\3MRA\Sl.bat
Aerated tank module
Landfill module
d:\3MFlA\AT.bat
d:\3MRA\LF.bat
"J
-J
J
-J
I
Figure 4.33 3MRA SUI: MMSP Configuration Screen
4-31
-------
Section 4.0
Modeling Approach
Selections SubScreen
The SELECTIONS subscreen (Figure 4.34) of the SYSTEM MANAGEMENT SCREEN
allows the user to select the sites, chemicals, source types, and waste levels used for the
simulation. The user can select a single, subset, or all of the sites, chemicals, source types, and
waste levels to be simulated. The sites appear on the subscreen as site identification names that
come directly from the Site-Based Database. The chemicals also appear on the subscreen as the
actual names defined in the Chemical Properties Database.
^ 3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration System Management j System Status |
Selections j Options ]
^Jnjxj
Site
Chemical
Source
0114001
A
0130207
0131104
0131207
0131508
0136703
0220102
0221207
0223504
0224002
0231002
0231106
0231407
0231610
d
0231911
1 ..1,1 -T richloroethane [M eth
2,3,7,8-T etrachlorodibenzo-[
2..4-D [2,4-Dichlorophenoxy
Acetonitrile [Methyl cyanide]
Acrylonitrile [2-Propenenitri _
Aniline
Benzene
Benzo(a)pyrene
Bis-(2-ethylhe>:yl] phthalate [
Carbon disulfide
Chlorobenzene
Chloroform
D ibenz[a,h]anthracene
Ethylene dibromide [1,2-Dib i
Hftxarhlrirn-I 3-hutRrlimR fl T I
Waste level
Figure 4.34 3MRA SUI: System Management Selections Screen
The five source types are land application unit (LAU), waste pile (WP), surface
impoundment (SI), aerated tank (AT), or landfill (LF). The waste level categories define the
waste concentration in wastestreams entering the source. The selection of multiple entities (e.g.,
sites) is facilitated by two options. By holding down the Ctrl key and left clicking on individual
selections, each selection will be highlighted, thus, including it as part of the simulation. If the
user wishes to highlight a consecutive series of entities (e.g., all chemicals from Benzene
through Chloroform) the first entry (Benzene) is left clicked, and while holding down the Shift
key, left click on the last chemical in the list (Chloroform). To select all available entries of a
list the user left clicks on the first entry, scrolls down to the bottom of the list and, while holding
down the Shift key, left clicks on the last entry.
4-32
-------
Section 4.0
Modeling Approach
Options
When the site, chemicals, source types, and waste levels are selected, the OPTIONS
subscreen (Figure 4.35) of the SYSTEM MANAGEMENT SCREEN is used to select the data
storage level, mode of operation, number of iterations, and seed value used as part of the
probabilistic simulation. The storage level selections are currently minimum or maximum. The
minimum storage level keeps only the RSOF files from the simulation while the maximum
storage level stores all SSF, GRF, and RSOF files created during the simulation. The number of
Monte Carlo iterations (or realizations) for the simulation is defined by the user. The seed value
is used by the statistical component of the simulation. This value defines the random sampling
seed required for the simulation. If the user needs to duplicate a simulation, this value must be
the same as the one used in the original simulation.
The user can select one of three different operation modes for the simulation. The Debug
mode stops the simulation after every change in iteration, chemical, source type, and waste
level. The Stop on Error mode stops the simulation only if an error occurs in a processor or
module when production runs are being implemented (if the Stop on Error is not selected the
simulation simply logs the error and executes the next site simulation). The Stop on Warning
mode stops the simulation if a warning appears in a processor or module. The Debug, Stop on
Error, and Stop on Warning modes can be selected simultaneously by the user. These options
are generally utilized only when testing the software of the modeling system. The Comment
box on this subscreen allows the user to provide descriptive information about the current
simulation.
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration System Management j System Status |
Selections Options
^jnjxj
Storage level IBSBEI V Debug mode
Number of realizations [i ' Stop on error
I ^^^QlStop on warning
Seed value for realizations 11031
Comments
Figure 4.35 3MRA SUI: System Management Options Screen
4-33
-------
Section 4.0
Modeling Approach
4.2.4 System Status
The main function of the SYSTEM STATUS SCREEN (Figure 4.36) is to allow the user
to follow the progress of the simulation, as it steps through the iterations, sites, chemicals, source
types, and waste levels. Depending on the mode of operation the user selected for the simulation
(see the OPTIONS subscreen of the SYSTEM MANAGEMENT SCREEN, Section 4.2.2), this
screen behaves differently. If the Debug mode is selected, the simulation stops after each step
whether it is an iteration, site, chemical, source type, or waste level. The user can then select the
Resume button to continue the simulation or evaluate the results of the simulation (e.g.,
switch out of the 3MRA modeling system to some other software). If the Stop on Error or Stop
on Warning mode is selected, the software continuously cycles through the simulation until it
encounters an error or warning from a processor or module. If an error or warning occurs, the
simulation is suspended. The user can evaluate the situation and decide whether to continue the
simulation or stop it. If the simulation is not completed, all files created before the suspension of
the simulation are stored in the specified locations and are available for future evaluation If the
simulation is suspended, the user can attempt to correct the problem, and then
simulation.
Resume the
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration | System Management System Status I
Current realization
Current site iteration
Current source iteration
Current chemical iteration
Current waste level iteration
Messages
Start
Run 0 of 0
Reset
Stop Run
Figure 4.36 3MRA SUI: System Status Screen
4-34
-------
Section 4.0
Modeling Approach
Any error or warning message generated by a processor or module is displayed in the
message box provided at the bottom of the SYSTEM STATUS SCREEN. These messages help
the user determine the status of the simulation and whether to continue or suspend the
simulation. Error and warning files that correspond to these messages are created by the system.
The user can utilize these files to determine the problem and continue the simulation, or stop it
and correct the problem. If the user stops the simulation and corrects the problem, the simulation
can be set up where it stopped and the simulation resumed using the Resume button.
4-35
-------
Section 4.0
Modeling Approach
4.3 Post Simulation Analysis
This section covers the tools used for post simulation analysis. Section 4.3.1 describes
the tools available to review and analyze results of national assessments. These tools include the
Exit Level II Processor (ELP-II) and the Risk Visualization Processor (RVP). Section 4.3.2
describes the Site Visualization Tool (SVT) that facilitates visual review and analysis of single
site assessments.
Example datasets for both a national assessment and site assessment have been provided
in the installation of the 3MRA modeling system. The essential features of the analysis tools
described by applying them to the review of two sets of example results.
4.3.1 Analysis of National Assessment Results (ELP II/RVP)
At the completion of a national assessment simulation there exists the Risk Summary
Output Files (RSOF). The RSOFs are Microsoft Access databases. The RSOFs can be directly
viewed by using Microsoft Access software. The RSOFs are input to the ELP II for the purpose
of exploring the relationship between waste concentration levels (Cw's) and the national
probabilities of risk protectiveness. The ELP II provides a user interface that allows the
graphical exploration of the risk summary data as well as the ability to generate Protective
Summary Output Files (PSOFs) that provide tables of national exemption levels based on data
contained in the RSOFs and regulatory criteria established by the user.
Let's execute an example of a national assessment and use the ELP II and the RVP to
review risk/hazard results and generate national exemption levels. For this example a national
exemption level will be determined for benzene. Figure 4.37 illustrates the Selection Subscreen
of the SUI System Management tab needed to communicate the national assessment to the
3MRA modeling system. Notice that all sites are selected, only the chemical benzene, all waste
management units, and all five wastestream concentration levels.
With the selections made the user proceeds to the Options Subscreen of the SUI System
Management tab and specifies "Minimum" storage, one Monte Carlo realization, leaves the
default random number seed as 11031, and leaves the options related to debug, error, and
warning modes unchecked. The user then enters the SUI System Status Subscreen and presses
the Start button. The execution of this national assessment for benzene takes on the order of
two full days to complete the 2095 individual site simulations (on a P III 1.0GHz PC).
NOTE : The C:\ is used here as an example
of the directory location where the 3MRA
modeling system has been installed. The
drive specification must correspond to the
one specified by the user during installation
4-36
-------
Section 4.0
Modeling Approach
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
File
System Configuration System Management j System Status |
Selections | Options |
^JnjxJ
Site
Chemical
Source
1522504
3
1,1,1 -T richloroethane [M eth
A
1530605
2,3,7,8-T etrachlorodibenzo-[
1530808
2..4-D [2,4-Dichlorophenoxy
1532401
Acetonitrile [Methyl cyanide]
1621808
Acrylonitrile [2-Propenenitri
1630106
Aniline
1630401
Benzene
1631701
Benzo(a)pyrene
1632106
Bis-(2-ethylhexyl) phthalate [
1632703
Carbon disulfide
1633404
Chlorobenzene
1633405
Chloroform
1635404
D ibenz[a,h]anthracene
1721603
Ethylene dibromide [1,2-Dib
I
1
HfiVRnhlnm-l 3-hutflriipnR fl
ZJ
Waste level
Figure 4.37 3MRA SUI: Selection Screen for National Assessment of Benzene
To facilitate the objective of demonstrating the ELP II/RVP the user has been provided
the header file and RSOFs that result from this assessment. Upon installation of the 3MRA
modeling system the example RSOF is located in the subdirectory
C :\3MRA\ExampleOutput\RSOF.
Before this set RSOF databases can be accessed by the ELP II, the contents of the
directory must by copied to the standard RSOF directory (i.e., C:\3MRA\RSOF). This is
facilitated by providing the batch file named "MoveFiles.bat" in the C:\3MRA\ExampleOutput
directory. The user simply double clicks on the "MoveFiles.bat", the batch file executes
transferring the example RSOF databases to the C:\3MRA\RSOF directory, where the ELP II
expects to find them. The header file is located in the C:\3MRA\SSF directory under the name
"hdprodexl.ssf'.
To start the ELP-II, the user selects the Windows Start button and clicks
PROGRAMS/3MRA/ELPII. Figure 4.38 illustrates the initial screen that appears upon the
invocation of the ELP II. The user then goes to the menu and selects File and then Open, which
accesses the Explorer window for the purpose of locating the appropriate "header file" that was
used to generate the national simulation results. The user locates and highlights the appropriate
"header file" as C:\3MRA\SSF\hdprdexl.ssf and clicks on the OPEN button.
The user then left clicks on the ~
button to display the list of chemicals, selecting
benzene for this example (only benzene is available even though other chemicals are listed).
Similarly, the user selects from among the WMU source options (e.g., Landfill). Upon selection
of the chemical and source type the screen shown in Figure 4.39 is displayed. The screen
displays a list of all factors and options associated with 1) establishing the quantitative
4-37
-------
Section 4.0
Modeling Approach
relationship between Cw and Probability of Protection (i.e., the percentage of sites that protect a
specified percentage of receptors at a specified risk/hazard level, e.g., a human carcinogenic risk
of 10"6), and 2) computing national exemption/exit levels.
There are three type of Cw vs Probability of Protection relationships that can be
generated, each a function of a different measure of risk (i.e., human carcinogenic risk, human
hazard, and ecological hazard). Not all chemicals present all three types of risk and also, effects
data may not be available for any given chemical for any given risk measure. Thus, depending
on the selections made the ELP II may simply inform the user of its inability to develop certain
Cw vs Probability of Protection relationships.
3MRA — Risk Visualization Processor
File ELP-II
; Selections ;|
rGeneral—
^jnjxj
Chemical constituent
WMU source type
"3
"3
Figure 4.38 Initial ELP-II Screen
HWIR — Risk Visualization Processor
File ELP-II
Selections |
-General—
^jnjxj
Chemical constituent
WMU source type
Distance
Method used for critical year
Exposure pathway
Receptor type
Cohort type
Risk level
Hazard quotient
-Ecological
Rollup option
Radius ring distance
Habitat group
Hazard quotient
3
3
3
3
Summation of all Ingestion and In T |
All Receptors ^
3
3
3
Population percentiles
0
95
25
» I
99
5
« I
75
85
90
98
By Ring and Habitat Group
<1000m
3
~3
Figure 4.39 EPL II Options Screen
4-38
-------
Section 4.0
Modeling Approach
The ELP II presents the user with default selections for each regulatory factor. The user
can modify the selection for any factor by left clicking the
~ button and making an
appropriate selection. On the right hand portion of the screen the user selects the population
percentiles for which the Cw vs Probability of Protection relationship should be generated. The
options are listed in the left textbox (i.e., 0, 5, 25, 50, 75, 85, 90, 95, 98, 99). The user highlights
the percentages of interest and clicks the » button to transfer the percentage to the right hand
textbox where it becomes part of the ELP II/RVP processing. Removing the percentages from
the right hand textbox is accomplished in a similar fashion using the « button.
Given the selections the user then clicks the Generate CDFs button. Because benzene
poses a carcinogenic risk and not other health hazards a message box with the message 'No table
available for human hazard quotient summary' is displayed. Figure 4.40 shows the Human Risk
screen after the button has been clicked. This screen has three functions. First, in addition to the
Selection tab there is a tab for each risk measure selected by the user, allowing the user to view
the various Cw vs Probability of Protection graph. Secondly, it displays the relationship between
Cw and Probability of Protection. A curve is included for each population percentile selected by
the user. Notice that as Cw increases the Probability of Protection either remains the same or
decreases. In this example, for instance, the graph shows there is an 80 percent probability that
any given site is protective of 95 percent of all receptors located at the site at a risk level of le-6
and a Cw of le3 ug/g. Or, said a bit differently, that 80 percent of the sites are protective of 95
percent of the receptors located at individual sites at the site at a risk level of le-6 and a Cw of
le3 ug/g. The third function of the screen is to facilitate the computation of interpolated values
for exemption levels (shown as interpolated Cw). The user specifies a population percentile and
Probability of protection in the text windows at the top of the screen and the ELP II computes a
value for the corresponding Cw (interpolated between plotted points if necessary).
3MRA -- Risk Visualization Processor
File ELP-II
Selections I Hi..man Hisk l| Ecological HQ |
Cw Interpolater
Population percentile
Probability of protection
3
~3
Interpolated Cw
Set Defaults
J
Set Individual Defaults |
ITIiijEUrE7! [WPP
Human Risk Protective Summai
o
CL
100
90
80
70
60
50
40
30
20
10
0
—£
c
'—^
"H
rl
I I 95% pop.
I~l 99% pop.
1:0.001
2:0.1
0.001
0.1
1 100
CWs in ug/g
1000
3:1
4:100
5:1000
"Description
Chemical constituent:
Wlenzene I
WMU source type:
WP
Distance:
500.000000
Critical year method:
Maximum
Exposure pathway:
Summation of all Ingestion an
Receptor type:
^t^.ll Receptors |
Cohort type:
All Cohorts
Risk level:
1E-G
Figure 4.40 ELP II Human Risk Protective Summary Screen
4-39
-------
Section 4.0
Modeling Approach
These values of Cw represent possible exemption levels that may, for example, define the
threshold between hazardous and non-hazardous waste. The user can return to the selection
screen and modify various regulatory factors and develop new plots. This interaction between
the user and the RSOF database can be fully explored via the ELP II/RVP.
Figure 4.41 Shows the Ecological Hazard Quotient screen. Notice that in this graph all
sites are protective of all terrestrial receptors, even at the highest possible Cw. Thus, from this
graph one would conclude that under no scenario does benzene pose a risk to terrestrial receptors
(keep in mind this is a hypothetical graph for illustrative purposes only).
HWIR — Risk Visualization Processor
File ELP-II
Selections | Human Risk
_Cw Interpolator
^jnjxj
Population percentile
Probability of protection
~ | Interpolated Cw |
Set Defaults
3
Set Individual Defaults I
loffl 1 |£|tu|^|Lllh,l fflHHl ITP \m®n\
Eco
100
= 90
¦2 80
o
2 70
0
CL 60
° 50
£ 40
Is
£ 30
1 20
10
logical HQ Protective Summ
\ I 95% pop.
I~l 99% pop.
1:0.001
2:0.1
3:1
4:100
0.001 0.1 1 100 1000
CWs in ug/g
5:1000
Description—
Chemical constituent:
^Kenzene^^^_^^^^~
WMU source type:
WP
Rollup option:
By Ring and Habitat Group
Radius ring distance
<1000m
Habitat group
Bterrestrial
Hazard quotient:
1
Figure 4.41 ELP II Ecological HQ Protective Summary Screen
The user, via this graphical plotting capability, is free to explore the possibilities
regarding the relationship between Cw and Probability of Protection. Additionally, the user can
specify a population percentile and probability of protection (i.e., percentage of sites protective
at the given population percentile) and have the ELP II compute an interpolated value of the
exemption level (Cw). This functionality appears at the top of Figures 4.40 and 4.41 within the
box labeled Cw Interpolator.
The ELP II also allows the user to generate output tables containing exemption levels
under varying conditions (as described in Section 3.3.9.2). Figure 4.42 illustrates the options
that will appear when the ELP II tab is invoked. The tables are stored in comma separated data
files contained in C:\3MRA\PSOF. The files are easily read into a spreadsheet such as Microsoft
Excel.
4-40
-------
Section 4.0
Modeling Approach
4.3.2 Analysis of Site Assessment Results
While the 3MRA modeling system is designed principally to execute national assessments
it is also capable of singling out a specific combination of site, chemical, WMU, and Cw for
simulation. This capability has been particularly useful during the verification testing of the
3MRA modeling system. It is also important in the context of comparing results of the 3MRA
with other modeling systems designed for similar purposes. To demonstrate the application of the
3MRA to a single site and show how detailed results of a site application can be reviewed
visually we will use an example. Outputs for this example are provided in the installation of
3MRA.
¦i£t- HWIR -- Risk Visualization Processor
JnJxJ
File ELP-II
5ave Protection Group
Ctrl+S
Allow Combined Source Tables
Ctrl+G
Create Exit Levels Tables
Create Cohort Risky HQ Tables
Create Receptor Type Risk/HQ Tables
Create Exposure Pathway RiskJHQ Tables
Create All Tables
"3
"3
Generate CDFs
Distance
Method used for critical year
Exposure pathway
Receptor type
Cohort type
Risk level
Hazard quotient
-Ecological—
Rollup option
Radius ring distance
Habitat group
Hazard quotient
500.000000
Maximum
"31
"31
Air Inhalation
All Receptors
31
31
All Cohorts
1E-6
31
31
31
"Population percentiles-
0
95
25
» 1
99
5
50
« '
75
85
90
98
By Ring and Habitat Group
<1000m
31
31
terrestrial
31
3]
Figure 4.42 ELP II Tabular Output Selection Screen
Figure 4.43 illustrates the Selection Subscreen of the SUI System Management tab
needed to communicate the single site application to the 3MRA modeling system. Notice that
one site is selected (ID# 0114001) one chemical (CAS ID 71-55-6), one waste management units
(WastePile, WP)), and one wastestream concentration level (4). All other SUI input screens
should be populated as described in Section 4.2.
With the selections made, the user proceeds to the Options Subscreen of the SUI System
Management tab and specifies "Maximum" storage, one Monte Carlo realization, leaves the
default random number seed as 11031, and leaves the options related to debug, error, and warning
modes unchecked. The user then enters the SUI System Status Subscreen and presses the Start
button. The execution of this site assessment for benzene takes on the order of a few minutes to
complete (on a P III 1.0GHz PC).
4-41
-------
Section 4.0
Modeling Approach
Selecting Maximum storage causes the modeling system to save all SSF and GRF files
generated during the simulation. Because only one site is executed there will be one set of SSF
and GRF files generated in this example. The SSF and GRF files are saved under the
C:\3MRA\Permanent directory. For this example the SSF and GRF files have been placed under
the C:\3MRA\ExampleOutput directory. To continue to follow this example the user must either
execute the simulation or copy the SSF and GRF files from the ExampleOutput subdirectory into
the C:\3MRA\Permanent\SSF and C:\3MRA\Permanent\GRF directories. If the user elects to
execute this site assessment the resulting SSF and GRF files will be placed in the
C:\3MRA\Permanent directory.
File
System Management System Status
Selections j Options ]
System Configuration
3MRA - Multimedia Multipathway Multireceptor Risk Assessment
Site
Chemical
c
iource
0114001
3
1,1,1-Trichloroethane [MethH
I LA.U
0130207
3
2,3,7,8-T etrachlorodibenzo-[
WP
0131104
0131207
013150S
0136703
2,4-D [2,4-Dichlorophenoxy
Acetonitrile [Methyl cyanide]
Acrylonitrile [2-Propenenitri _
Aniline
SI
AT
LF
0220102
Benzene
0221207
0223504
0224002
0231002
0231106
0231407
0231610
0231 Hi 1
Benzo(a)pyrene
Bis-(2-ethylhexyl) phthalate [
Carbon disulfide
Chlorobenzene
Chloroform
D ibenz[a,h]anthracene
Ethylene dibrornide [1,2-Dib i
Hfiyflnhlnrn-1 3-hi jtflrlimfi N I
Waste level
mmmmm
3
2
1
Figure 4.43 3MRA Selection Screen for Site Example
4.3.2.1 Site Visualization Tool (SVT)
The purpose to the Site Visualization Tool is to allow the user to view a complete set of
science module results from a single site simulation. Each of the 3MRA science modules
produces a key set of results in the context of a multimedia risk assessment. For example, each of
the source modules estimates time series releases of contaminant to air, groundwater, watershed,
and surface water. Each of these time series is input to a medium-specific fate and transport
module with the result being a time series of medium concentrations at human and ecological
exposure locations. From here, the foodweb modules estimate future concentrations of
contaminant within various trophic levels. Finally, the exposure and risk modules generate
estimates of the time series of exposure and risk. The SVT is designed to read each of the Global
Output Files (GRFs) generated during the execution of a site and display all the time series results
4-42
-------
Section 4.0
Modeling Approach
for simultaneous viewing via an HTML plotting format. When viewed in this fashion the user
can identify trends within and across time series.
There are several important uses of this graphical tool. First, with respect to Quality
Assurance the SVT provides an efficient means by which to identify anomalies in the results
generated by the science modules.
The SVT is a Java based application that requires the Java runtime environment to be
installed for the SVT to work.. This runtime environment is installed during the 3MRA
installation as described in Section 4.1. The SVT is currently Beta version software. This means
that the software is fully functional but has not undergone final review, testing, and
documentation.
To launch the SVT, the user selects the Windows Start PROGRAMS/3MRA/SVT.
Figure 4.44 displays the SVT user interface. To execute, the SVT must receive the 3MRA
"header file" name associated with the site simulation, the location of the SSF and GRF files that
resulted from simulation, and the directory location for storing the graphical output files
generated by the SVT. The SSF, GRF directory specifies the parent folder in which the SSF and
GRF folders are located. For this example the SSF, GRF directory is
D:\3MRA\ExampleOutputs\Permanent\WP011400171-55-6Cw4Rl. The header file specifies
which header file the user would like to use for the SVT analysis. For this example, the header
file name and location is D:\3MRA\ExampleOutputs\Permanent\WP011400171-55-
6Cw4Rl\ssf\hdprod.ssf. The output directory specifies where the generated output graphics will
be placed. For this example the results should be
D :\3MRA\ExampleOutputs\Permanent\WPO 11400171 -5 5-6C w4Rl.
The status label on the bottom of the window informs the user of the current SVT status.
Site Visualization Tool
SSF, GRF directory:
Headerfile:
Output directory:
Go
D:\3MRA
D:\3MRA\SSF\Hdprod.ssf
D:\3MRA\plot
Cancel
^JnjxJ
A
_l
J
Status: Ready
Figure 4.44 SVT User Interface Screen
Once the required information has been correctly entered, the user clicks the
Cancel
Go button.
button will end the
display: 'Status:
Alternatively, if the user does not wish to proceed, the clicking the
SVT application. When the SVT has finished processing, the Status bar wil
Done!'. The output of the SVT is a collection of graphics illustrating various time-varying
characteristics of the specified site. To view the SVT output, the user should open the index.html
file located in the folder indicated by the 'Output directory:' box on the initial SVT screen.
Double-clicking on the index.html file will open the file in the default Internet Browser (e.g.
4-43
-------
Section 4.0
Modeling Approach
Internet Explorer, Netscape Navigator). Figure 4.45 illustrates an example index.html file.
Clicking on an individual graph will bring up the full size image in the browser. For example,
clicking on the 'Soil Concentration' graph brings up a full size Soil Concentration image as
shown in Figure 4.46.
D:\3MRA\Permanent\WP01140017 l-55-6Cw4Rl\index.html - Microsoft Internet Explorer provided by U.S. EPA
File Edit View Favorites Tools Help
.Jj9j J£j
¦ Back
•J Si £uS U Search (j] Favorites 'ijjMedia | ^ Q C?
Address D:\3MRA\Permanent\WP011400171-55-6Cw4Rl\index.html
T | ^Go Links
1,1,1-Trichloroetliaiie [Methyl(71-55-6) at 0114001 in a(aii) WP
Source
Release flux
Le achate rate
SW Load Chem/Solid
Runoff
Particle
Soil Concentration
TSS to SW
Air
Vapor Concentration
PM10 Concentration
Vapor Wet Deposition
PM10 Dry Deposition
PM10 Wet Deposition
CT
m
d My Computer
Figure 4.45 Example SVT Output
4-44
-------
Section 4.0
Modeling Approach
File Edit View Favorites Tools Help
Address
Soil Concentration
18000
1000
100
10
0.1
20
year
— CTss(l,l) — CTss(l,2) — CTssd^J — CTda(l,l>
— CTda(l,3)
Done
| My Computer
3 D:\3MRA\plot\Image3.gif - Microsoft Internet Explorer
D:\3MRA\plot\Image3.gif
A
Figure 4.46 Sample SVT Soil Concentration Graph
4-45
------- |