United States
tn\iionniental
Office of An Quahn
Planning and Standaids
Research Tnanele Park. NC 2771]
EPA-454/R-95-WW
Februan 1995
An
& EPA
AIR/SUPERFUND NATIONAL
TECHNICAL GUIDANCE
STUDY SERIES
Volume V • Procedures for Air
Dispersion Modeling At Superfund
Sites
-------
EPA-454/R-95-003
AIR/SUPERFUND
NATIONAL TECHNICAL
GUIDANCE STUDY SERIES
Volume V - Procedures for Air
Dispersion Modeling At Superfund
Sites
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Technical Support Division
Research Triangle Park, NC 27711
U.S. Environmental Protection Agency
February 1995 Region 5, Library (PL-12J) _,
77 West Jackson Boulevard, 12tn Hoor
Chicago, IL 60604-3590
-------
Disclaimer
This report has been reviewed by the Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, and
has been approved for publication. Any mention of trade names or
commercial products is not intended to constitute endorsement or
recommendations for use. Copies of this report are available for
a fee from the National Technical Information Service, 5285 Royal
Road, Spingfield, VA 22161.
SL
EPA-454/R-95-003
I
}
1
-------
PREFACE
This manual is the fifth in a five-volume series dealing with air pathway assessments
at hazardous waste sites and was developed for the U.S. Environmental Protection Agency
(EPA) Office of Air Quality Planning and Standards in cooperation with the Office of
Emergency and Remedial Response (Superfund). It is an update of the air dispersion modeling
discussion in the original Volume IV of this series. This manual has been reviewed by a
Technical Advisory Committee consisting of EPA Regional modelers and members of the
Air/Superfund program.
This manual is an interim final document offering technical guidance for use by a
diverse audience including EPA Air and Superfund Regional and Headquarters staff. State Air
and Superfund program sta^f, Federal and State remedial and removal contractors, and
potentially responsible parties in analyzing air pathways at hazardous waste sites. This
manual is written to serve the needs of individuals having different levels of scientific
training and experience in designing, conducting, and reviewing air pathway analyses.
Because assumptions and judgements are required in many parts of the analysis, the
individuals conducting air pathway analyses need a strong technical background in air
emission measurements, modeling, monitoring, and risk assessment. Remedial Project
Managers. On-Scene Coordinators, and the Regional Air program staff, supported by the
technical expertise of their contractors, will use this manual when establishing data
quality objectives and the appropriate scientific approach to air pathway analyses. This
manual provides for flexibility in tailoring the air pathway analysis to the specific
conditions of each site.
Air pathway assessments involve complex procedures requiring the use of professional
judgment. The information set forth in this manual is intended solely for technical
guidance. The procedures set out in this manual are not intended, nor can they be relied
upon, to create rights substantive or procedural, enforceable by any party in litigation .
with the United States.
m
-------
It is envisioned that this manual will be periodically updated to incorporate new data
and information on air pathway analysis procedures. The EPA reserves the right to act at
variance with these procedures and to change them without formal public notice as new
information and technical tools become available on air pathway analyses. The EPA Regional
Air/Superfund coordinator should be consulted on the availability and use of the most recent
procedures
Copies of this manual are available, as supplies permit, through the Library Services
Office (MD-35). U.S Environmental Protection Agency. Research Triangle Park. North Carolina
27711 or from the National Technical Information Services (NTIS), 5285 Port Royal Road.
Springfield, Virginia 22161 (Telephone 703/487-4650).
IV
-------
Acknowledgements
This manual was prepared for the U S Environmental Protection Agency (EPA) by Radian
Corporation under EPA Contract No. 68-D30033, Work Assignment No. 3. The EPA Work
Assignment Manager was Mr. Jawad Touma, of the Source Receptor Analysis Branch. Technical
Support Division of the Office of Air Quality Planning and Standards (OAQPS). Special
thanks go to Mr. Dean Wilson and Mr. John Irwin of OAQPS, Ms. Patricia Flores of EPA Region
III and Mr Bill Ryan of EPA Region X for their input on the initial draft.
-------
TABLE OF CONTENTS
Page
DISCLAIMER . . . n
PREFACE . ... m
ACKNOWLEDGEMENTS . . . . . v
ABBREVIATIONS AND ACROMVMS x
1 INTRODUCTION 1-1
Background . . . . ... . 1-2
Objectives and Scope of this Manual 1-3
Overview of Air Modeling at Superfund Sites 1-4
Phases of the Superfund Process 1-4
Typical APA Activities 1-8
2. DEVELOPMENT OF A MODELING PLAN 2-1
3. POLLUTANT RELEASE CHARACTERIZATION 3-1
Selecting Pollutants to Model 3-5
Source Definition 3-10
Point Source Characterization 3-10
Area Source Characterization 3-12
Volume Source Characterization 3-14
Line Source Characterization 3-15
Container/Acute Releases 3-15
4. MODEL SELECTION 4-1
Screening Analysis 4-2
Refined Analysis 4-4
Dense Gas Release Simulations 4-6
Deposition Modeling 4-7
Model Availability 4-11
VI1
-------
TABLE OF CONTENTS (Continued)
Page
5. COMPONENTS OF A MODELING ANALYSIS 5-1
Meteorological Considerations .... . .... 5-1
Wind Speed and Direction 5-3
Atmospheric Stability . . .... 5-4
Ambient Temperature. Relative Humidity, and
Pressure 5-6
Surface Roughness . . . .... 5-6
Mixing Height . .... . . 5-7
Terrain Considerations . 5-9
Receptor Definition 5-11
Urban/Rural Classification 5-15
Plume Downwash 5-17
Averaging Time Considerations 5-19
Worst-Case Impact Determination 5-21
Background Concentrations 5-22
6. ASSESSMENT OF MODEL RESULTS 6-1
Summarizing Model Input and Output 6-1
Topics for Consideration 6-3
Model Uncertainty . ... 6-4
7. REFERENCES 7-1
APPENDIX A: CASE EXAMPLE A-1
APPENDIX B: CHECKLIST OF MODELING CONSIDERATIONS B-l
APPENDIX C: BASIC REQUIREMENTS OF COMMONLY USED DISPERSION
MODELS FOR SUPERFUND APAs C-l
APPENDIX D: USEFUL CONTACTS AND TELEPHONE NUMBERS D-l
APPENDIX E: BIBLIOGRAPHY OF NTGS DOCUMENTS E-l
APPENDIX F: CONTAINER/ACUTE RELEASES F-l
VI11
-------
LIST OF FIGURES
Page
1-1 Phases of the Superfund Process . 1-6
5-1 Wind Roses Exhibiting Distinct Frequencies of Wind Speed
and Direction . 5-5
5-2 Examples of Cartesian and Polar Receptor Grids 5-14
IX
-------
LIST OF TABLES
Page
1-1 APA Activities During Various Superfund Actions .... ... . 1-10
2-1 An Outline for a Superfund APA Modeling Plan . . . . 2-2
3-1 General Characteristics of Sources Associated with Superfund
Activities . . .... . . . 3-3
4-1 SCRAM BBS Communication Parameters . ... 4-11
5-1 Representative Values of Surface Roughness for a Uniform
Distribution of Selected Types of Ground Cover 5-8
5-2 Classification of Land Use Types 5-18
6-1 Example Summary Table of Maximum Predicted Impacts 6-2
-------
ABBREVIATIONS AND ACRONYMS
AAM
ALOHA
APA
ARAR
CERCLA
CFR
CTDMPLUS
DEGADIS
EPA
ER
FDM
FS
GEP
HSL
ISC2
ISCLT2
Ambient Air Monitoring
Area! Locations of Hazardous Atmospheres
Air Pathway Assessment (or Analysis)
Applicable or Relevant and Appropriate Requirement
Bulletin Board System
Comprehensive Environmental Response, Compensation, and Liability Act
Code of Federal Regulations
Complex Terrain Dispersion Model PLUS Algorithms for Unstable Situations
Dense Gjas Dispersion Model
U.S. Environmental Protection Agency
Emergency Removal
Fugitive Dust Model
Feasibility Study
Good Engineering Practice
Hazardous Substances List
Industrial Source Complex Models
Industrial Source Complex Long-Term Model
XI
-------
ISCST2 Industrial Source Complex Short-Term Model
ME I Maximum Exposed Individual
MPRM Meteorological Processor for Regulatory Models
msl Mean Sea Level
NPL National Priorities List
NSR New Source Review
NTG National Technical Guidance
NTGS National Technical Guidance Study
NTIS National Technical Information Service
NWS National Weather Service
OAQPS Office of Air Quality Planning and Standards
O&M Operation and Maintenance
PA Preliminary Assessment
PCB Polychlorinated Biphenyl
PM/PM10 Particulate Matter/Particulate Matter of less than 10 micrometers in
diameter
PRP Potentially Responsible Party (or Parties)
PSD Prevention of Significant Deterioration
QA/QC Quality Assurance/Quality Control
XII
-------
RA Remedial Action
RD Remedial Design
R3 Remedial Investigation
ROD Record of Decision
RTDM Rough Terrain Dispersion Model
SARA Superfund Amendments and Reauthonzatlon Act
SCRAM Support Center for Regulatory Air Models
SI Site Inspection
STAR Stability Array
TCDD Tetrachlorodibenzo-p-dioxin
TCDF Tetrachlorodibenzofuran
USGS U.S. Geological Survey
UTM Universal Transverse Mercator
VOC Volatile Organic Compound
XTM
-------
SECTION 1
INTRODUCTION
This manual presents guidance for designing, conducting, and evaluating air
dispersion modeling analyses for Superfund sites. Its purpose is to provide a logical and
systematic approach for applying air quality models, which are an integral part of several
regulatory programs. This manual is intended to augment the primary U.S. Environmental
Protection Agency (EPA) guidance document on air quality modeling, the Guideline on Air
Quality Models (Revised)1, as such it elaborates on modeling issues particularly related to
Superfund Sites All exposure pathways - including the air pathway - must be evaluated for
every Superfund site; therefore, each site usually requires some level of air dispersion
modeling.
In many cases the nature and complexity of atmospheric dispersion make exposure via
the air pathway more difficult to predict than exposure via other pathways. The air pathway
Is unique in that any on-site release of emissions can have an almost immediate impact.
Furthermore, the locations of impact can shift relatively quickly, with changes in wind
speed and direction. In contrast, exposure through other pathways often requires extended
time periods to occur, and can be minimized by limiting site access or prohibiting use of
contaminated resources (e.g.. drinking water).
Air dispersion models provide the ability to mathematically simulate atmospheric
conditions and behavior and are used to calculate spatial and temporal fields of
concentrations and particle deposition due to emissions from various sources. The output
from air dispersion models is used to fill the gaps in data generated by air monitoring
programs that cannot provide measured concentrations at all locations. Dispersion models
can provide concentration or deposition estimates over an almost unlimited grid of user-
specified locations, and can be used to evaluate both existing and forecasted emissions
scenarios. In this capacity, air dispersion modeling is a vital tool in assessing the
potential risk associated with existing and proposed emissions sources.
1-1
-------
The remainder of this section provides background information related to the
development of this manual, identifies the manual's objectives and scope, and contains an
overview of air modeling for Superfund sites.
BACKGROUND
The EPA's national Air/Superfund Coordination Program helps EPA Headquarters and the
Regional Superfund Offices evaluate Superfund sites and determine appropriate remedial
actions to mitigate their effects on air quality Each Regional Air Program Office has an
Air/Superfund Coordinator who coordinates activities at the Regional level. The
Air/Superfund Coordinator Program has a number of responsibilities, including preparation of
national technical guidance (NTG) documents. A bibliography of national technical guidance
study (NTGS) documents is contained in Appendix E
Continuing EPA involvement in toxic and hazardous pollutant impact activities at
Superfund sites has created a need for guidance in the appropriate modeling methods for such
releases. In 1989, the EPA published guidance for dispersion modeling and air monitoring
for Superfund Air Pathway Assessment (APA). That document (Volume IV of the four-volume APA
series)2 provided technical guidance for activity-specific and source-specific dispersion
modeling and air monitoring. Since its publication, there have been changes in dispersion
modeling guidance for Superfund sites, and a number of documents dealing with specific
modeling techniques applicable to Superfund sources have been developed.
This manual constitutes a new volume (Volume V) in the APA series. It is an update
of the air dispersion modeling discussion in the original Volume IV.2 This fifth volume
contains much of the modeling information of the old Volume IV. but emphasizes newly
developed guidance and techniques applicable to Superfund sources. The guidance for ambient
air monitoring included in the original volume has been revised to form the sole topic of
the new Volume IV of this series. The documents comprising this multi-volume APA series are
listed below:
• Volume I - Overview of Air Pathway Assessments for Suoerfund Sites (Revised)3:
• Volume II - Estimation of Baseline Air Emissions at Suoerfund Sites
(Revised)4:
1-2
-------
Volume III - Estimation of Air Emissions from Clean UP Activities at Superfund
Volume IV - Guidance for Ambient Air Monitoring at Superfund Sites6: and
Volume V - Guidance for Ambient Air Modeling at Super-fund Sites. (Current
document to be published).
OBJECTIVES AND SCOPE OF THIS MANUAL
The overall objective of this project was to create a new and separate document
dealing only with air dispersion modeling issues The new manual would contain relevant
portions of the previous document, and would emphasize newly developed guidance and
techniques applicable to Superfund sources. In designing this fifth volume, it was not the
intention to provide detailed calculations and specific technical procedures, but rather to
present concepts, definitions, and general procedures, and to reference readily available
documentation for more detailed information
This manual offers technical guidance for use by a diverse audience, including EPA
air and Superfund regional and headquarters staff. State air and Superfund staff. Federal
and State remedial and removal contractors, and potentially responsible parties (PRPs).
Remedial project managers, on-scene coordinators, and regional air program staff, supported
by the technical expertise of their contractors, can use the information in this manual when
developing air dispersion modeling programs. This manual is written to serve the needs of
individuals with varying levels of training and experience in implementing air dispersion
modeling methods in support of air pathway assessments. However, professional judgement is
needed to develop air modeling approaches, so the individuals involved in this activity
would benefit from having a strong technical background in source characterization, air
monitoring, and risk assessment
Developing and implementing an air dispersion modeling program can be approached in a
systematic manner, but cannot be reduced to simple "cookbook" procedures (i.e., procedures
that are necessarily absolute). There is always a potential need for professional judgement
and flexibility when developing modeling programs for specific Superfund sites.
1-3
-------
OVERVIEW OF AIR MODELING AT SUPERFUND SITES
Under the Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA) and the Superfund Amendments and Reauthonzation Act (SARA), the EPA is required to
develop and implement measures to clean up hazardous or uncontrolled waste sites. The
cleanup of a contaminated site under the Superfund program proceeds via a series of actions
designed to remove or stabilize the contaminated material in a controlled way
Among the requirements of the National Contingency Plan (NCP) for CERCLA sites are
that a risk assessment be performed, there is compliance with ARARs, and the remedial action
be protective of human health. The compliance with these requirements, and the
documentation of this compliance, are the primary areas where atmospheric dispersion
modeling is needed for Superfund sites In addition, dispersion modeling may be useful for
other aspects of the Superfund process. The remainder of this section is a discussion of
the various steps of the Superfund process and how dispersion modeling may be incorporated
Phases of the Superfund Process
As outlined in Figure 1-1. Superfund activities can be classified in three phases-
pre-remediation, remediation, and post-remediation.
Pre-Remediation Phase--
The pre-remediation phase consists of the Site Discovery, a Preliminary Assessment
(PA), and a Site Inspection (SI). This phase is concerned with evaluating the potential
risk to public health and the environment posed by the discovered (identified) site. PA is
then conducted to collect as much information as possible about the site, with
emphasis on assessing the pollutants present and their physical state. The PA is meant to
be a relatively quick and inexpensive undertaking, involving the collection of all relevant
documentation about the site. In addition, general descriptions of local land use.
topography, demography, and meteorology may be formulated for use in developing a
preliminary modeling approach. The information gathered in the PA is used by the EPA to
determine whether further investigation or action is warranted.
1-4
-------
If further investigation is warranted, a SI is conducted. This is the first action
that involves some form of sample collection, and it is primarily concerned with determining
the urgency of the health risk posed by the site. Samples from various media are collected
and analyzed, and the results are used to rank the site within the Hazard Ranking System
model This model ranks the relative contamination posed by the site over
three pathways air, groundwater, and surface water. The direct contact and fire/explosion
pathways are evaluated by the model, but they are not currently included in the ranking. If
a site ranks above a predetermined score, it is placed on the National Priorities List
(NPL)
During the pre-remediation phase, the main dispersion modeling objectives include
providing a sufficient database of toxic air pollutant concentrations for performing a
detailed assessment of risk to public health and the environment. This assessment can
pertain to both on-site and off-site receptors, and will typically address baseline
conditions and those associated with various remedial alternatives. Generally, a nsk
assessment is performed that is a comprehensive, qualitative determination of the baseline
nsk associated with the site In some cases, only a risk evaluation may be performed,
wherein calculations are performed to develop boundary estimates of the potential risk. Air
dispersion modeling during the pre-remediation phase may also be used to provide input to
the design of an ambient air monitoring (AAM) network.
After a site is placed on the NPL. the necessity of an Emergency Removal (ER) is
evaluated through a site inspection by personnel from the removal program. This site
inspection may take place during the remedial investigation phase. If the site is believed
to pose an immediate and significant health risk, actions are taken to ameliorate the
problem In some cases, the ER action will temporarily increase site emissions. In such
cases, air modeling prior to the ER action may be needed to determine whether, and to what
extent, the local populace should be evacuated. Additional modeling may be needed during
the actual ER action to update earlier predictions as new information becomes available.
Following the SI and any ER actions, the remediation phase begins.
1-5
-------
Site Discovery |
Preliminary Assessment (PA) |
Site Inspection (SI) |
Hazard Ranking System ]
|
I
|
Remedial Investigation (RI)/
Feasibility Study (FS)
• Record of Decision (ROD)
Remedial Design (RD) |
Remedial Action (RA) |
1
1
Operation and Maintenance (O&M) 1
] National
--[Priorities List
! (NPU
—I
1 Emergency j
i_ Removal (ER) j
PRE-REMEDIATION
REMEDIATION
POST-REMEDIATION
Figure 1-1. Phases of the Superfund Process.
1-6
-------
Post-Remediation Phase--
Once the remediation phase has ended, a brief monitoring period is conducted during
which the effectiveness of the cleanup is determined. This is called the post-remediation.
or Operation and Maintenance (O&M) phase. If the monitoring shows that the site no longer
poses a health or environmental threat, the site may be removed from the NPL.
Remediation Phase--
The remediation phase consists of the Remedial Investigation (RI) and Feasibility
Study (FS). production of a Record of Decision (ROD) and Remedial Design (RD), and
performance of the Remedial Action (RA) This phase takes more time to complete than the
pre-remediation phase, and is designed to transform the site into a clean site in a
controlled manner
The RI and FS are separate steps, but are typically conducted simultaneously and
interactively. During the RI, data are collected to determine more precisely the types of
compounds present at the site and the locations and extent of contamination. The data
gathered during the RI are used as input to dispersion models and ultimately the results are
used to estimate the chrome baseline exposure as part of the risk assessment. The data are
also used to help identify appropriate cleanup procedures and remedial alternatives. The FS
is concerned with identifying the preferred cleanup alternative. In this step, dispersion
modeling may be used to help rank alternatives by identifying the potential air impacts of
each, and to aid in siting long-term AAM stations.
After the FS is completed, the ROD is issued. The ROD serves as the official EPA
decision about the preferred course of subsequent action. The next activities are the
preparation of the RD, a detailed plan for the site remediation, and then the actual RA is
initiated. The RA can take a variety of forms, from short-term activities to long-term
activities that can take several years to complete. During the RD, modeling may be
performed to support the development of an air emissions control strategy.
1-7
-------
During the remediation phase, the primary modeling objectives include providing a
sufficient database of toxic air pollutant concentrations for assessing the effects of the
remedial action evaluated. The modeling objectives also include providing predicted
concentration data for routine and non-routine releases in support of protecting on-site
workers, the off-site populace, and the environment3637. Predicted concentration data are
provided as a component of the ER system employed at the site Air dispersion modeling may
also be conducted to provide input to the design of the AAM network for this phase
During the post-remediation phase, the main dispersion modeling objective is to
provide a database of toxic air pollutant concentrations for the site boundary and for off-
site locations. All the main objectives of the Superfund process are part of assessing the
effectiveness of the RA. and for demonstrating that the off-site populace and environment
are protected. Modeling may be performed as part of the O&M if a post-remediation risk
assessment is required.
Typical APA Activities
Typical APA activities associated with steps in the Superfund process are summarized
in Table 1-1. These activities can be divided into the following four categories:
1) Screening evaluation of site emissions and their impacts on air quality under
baseline or undisturbed conditions;
2) Refined evaluation of site emissions and their effect on air quality under
baseline or undisturbed conditions-.
3) Refined evaluation of emissions and their effect on air quality from pilot-
scale remediation activities, and
4) Refined evaluation of emissions and their effects on air quality from full-
scale remediation activities.
Other APA activities may be appropriate for specific site applications. Screening
evaluation activities are most likely to occur during the SI, early RI, or O&M steps of the
Superfund process. Refined evaluations are most likely to occur during the RI. FS, RA and
O&M steps. In general, screening studies are performed to define the nature and extent of a
problem and are considered conservative, particularly for long-term predictions.
1-8
-------
They are often used to eliminate the need for more detailed modeling of a particular
situation. Refined studies are performed to provide more detailed treatment of atmospheric
processes and source-receptor relationships, and provide, at least theoretically, a more
accurate estimate of source impact. Further discussion on screening and refined analyses
with respect to dispersion modeling is presented in Section 4.
1-9
-------
TABLE 1-1 APA ACTIVITIES DURING VARIOUS SUPERFUNO ACTIONS
Action
Data Needs
Typical Ambient
Concentration Levels
Involved (ppb)
Typical Air Modeling
Strategy/Ob jpctive
Typical Air Monitoring
Strategy /Net work Design
Typical Uses of APA Data
SI
Qualitative
Screening Study.
Aid in siting of AAM stations
IH monitoring,
Limited fencel me
monitoring
Assess type and general magnitude
of site emissions
ER
Quantitative
200
Evaluate evacuation options
IH monitoring.
Fencel me monitoring.
Monitoring at ME!
Estimate risk to on-site workers
and off-site populace
RI
Serai-quantitative
10
Evaluate off-site exposure. Aid
in siting of AAM stations
IH monitoring.
Limited fencelme
monitoring
Estimate risk to on-site workers
and improve knowledge of emission
sources, also estimate risk to
off-site populace
FS
Quantitative
100-200
Evaluate impacts of various
remediation alternatives - input
to RO
IH monitoring,
Emission rate
measurements
Estimate air impacts during full-
scale remediation Air impact
issues include evaluation of
chronic and/or acute risk as well
as compliance with ARARs.
RA
Quantitative
* 200
Provide updated, regular
estimates of downwind
impacts
IH monitoring.
Fencelme monitoring.
Monitoring at MEI
Estimate risk to on-site workers
and off-site populace.
O&M Semi-quantitative
10
Input to any final risk
determination
Limited fencelme
monitoring.
Estimate risk to off-site
populace
Notes
ppb - parts per billion
AAM - Ambient Air Monitoring
APA - Air Pathway Assessment
ARAR - Applicable or Relevant and Appropriate Requirement
ER - Emergency Removal
FS - Feasibility Study
IH - Industrial Hygiene
MEI - Maximum Exposed Individual
O&M - Operations and Maintenance
RA • Remedial Action
RO - Remedial Design
RI « Remedial Investigation
SI ' Site Inspection
-------
SECTION 2
DEVELOPMENT OF A MODELING PLAN
A dispersion modeling plan, or protocol, should be developed for each Superfund Air
Pathway Assessment (APA). The purpose of preparing such a plan is to document the modeling
methodology and inputs proposed for use in the APA In addition, all appropriate portions
of the plan should provide an indication of how the selected procedures compare to existing
guidance, and where there are deviations from guidance, provide the rationale for such
deviation
The modeling plan provides an opportunity for peer review and approval of the
modeling program by the Remedial Program Manager, in coordination with the EPA Regional
Air/Superfund Coordinator, and the EPA Regional Modeler. Approval of the protocol before
modeling begins helps ensure that modeling analyses are properly designed and will meet with
regulatory approval. Furthermore, with a plan in place the modeler will know how to proceed
in the event certain outcomes unfold (e g.. if screening-level modeling indicates that risks
are unacceptable, the procedure for conducting a refined analysis will already be known and
approved). A good protocol will also serve as a checklist, clarifying what is relevant and
particularly significant with respect to modeling the site (e.g.. the plan will state that
predicting impacts in complex terrain does not apply to the site in question).
To assist in preparing a modeling plan, a suggested outline is provided in Table 2-1.
It should be noted that, aside from the benefits already mentioned, the effort involved with
preparing a modeling plan typically reduces the eventual effort required in conducting the
modeling analysis.
Specific aspects of the modeling plan are discussed in the following sections of this
manual. Section 3 discusses the source inputs needed for proper application of air
dispersion models. Section 4 addresses model selection, and the concept of screening and
refined analysis. In Section 5. other primary components of a modeling analysis are
discussed. The final section, Section 6, discusses the assessment of model results.
2-1
-------
TABLE 2-1. AN OUTLINE FOR A SUPERFUND APA MODELING PLAN
I. INTRODUCTION:
General characteristics of the site;
• General site activities (through all planned phases of the Superfund process);
and
• Characteristics of the surrounding environment:
Topography.
Climatology.
Demography.
Presence of water bodies, and
Vegetation types.
II DATA QUALITY OBJECTIVES
• Modeling objectives (consistent with the Superfund activity involved and the
overall project objective);
• Model application to each Superfund activity APA; and
• Overall rationale for the proposed modeling approach.
III. POLLUTANTS TO BE MODELED:
• Physical, chemical and toxicological properties of pollutants to be modeled.
and
• Averaging times associated with pollutants to be modeled, as prescribed by
state and federal applicable or relevant and appropriate requirements (ARARs).
IV. SOURCE CHARACTERISTICS AND EMISSIONS:
• Identification of all point and fugitive sources to be modeled;
Characterization of point sources (e.g., applicability of building downwash);
Characterization of fugitive sources (i.e.. line. area, or volume);
• Methods for estimating emission rates (as a starting point see Volume I of the
Air/Superfund Guidance Study Series3 and Models for Estimating Air Emission
Rates from Suoerfund Remedial Actions)7;
Determination of maximum short-term, and annual average emission rates;
• Characterization of the duration and frequency of emissions from each source
(e.g.. continuous or intermittent): and
Particle size distribution and chemical composition for modeling metals and
fugitive dust.
2-2
-------
TABLE 2. (Continued)
V. MODELING METHODOLOGY:
. Determination of rural/urban classification;
. Treatment of building downwash effects for point sources;
« Treatment of surrounding terrain;
Treatment of particle deposition (if appropriate):
• Model(s) selected and rationale,
« Model options proposed:
Meteorological data
Source of data.
Length of data record.
Quality and completeness of the data, and
Representativeness of the data (for off-site monitoring locations);
• Receptor grid:
Spatial extent.
Resolution around site boundary,
Especially sensitive locations (to address specific public health and
environmental concerns), and
Plan for refined grid to isolate maximum concentrations.
• Background concentrations.
Treatment of nearby sources of pollutant emissions, and
Use of existing ambient air monitoring (AAM) data.
VI. MODEL RESULTS:
Presentation of model results;
« Assessment of model results;
Input to risk assessment:
• Discussion of methods used to determine target compounds and risk assessment
threshold values; and
• Modeling uncertainties and their implications to the APA.
VII. REFERENCES.
(State reference for procedures cited in the modeling plan)
2-3
-------
SECTION 3
POLLUTANT RELEASE CHARACTERIZATION
This section describes the source inputs needed for proper application of air
dispersion models Atmospheric dispersion modeling for Superfund sites includes a
composition of sources that, in general, are different in configuration and characteristics
from the sources traditionally modeled for regulatory air permitting of elevated buoyant
point sources (e g , boiler or process stacks). Superfund sources primarily consist of
fugitive area and volume sources and. to a smaller extent, point sources
Air emissions from Superfund activities can be continuous, intermittent, or a one-
time release of a defined duration, and may have large temporal and spatial variability.
Releases can be anticipated (occurring from routine operations and known sources) or can be
unforeseen (resulting from accidental or nonroutine events). Both gaseous and particulate
emissions to the atmosphere must be considered. Volatile organic compounds are emitted as
gasses. Semi volatile organic compounds are emitted as gasses and particulate depending on
vapor pressure and ambient conditions. Metals and other inorganic substances with the
exception of metallic mercury are emitted as particles. Particles are treated separately
because of their different dynamics, such as settling velocity.
Point sources involve the release of emissions from a well-defined stack or vent, at
a well-defined temperature and flow rate. Air strippers, incinerators, thermal desorption
units, and in situ venting operations constitute the common point sources at Superfund
sites.
Fugitive sources, generally characterized as area, volume, and line sources in
dispersion modeling, involve the release of emissions from a defined surface or depth of
space. The amount of emissions released' from a fugitive source is more directly related to
environmental conditions (e.g., ambient temperature and wind speed above the surface). Area
sources at Superfund sites generally include landfills, lagoons, contaminated soil surfaces.
materials handling and transfer operations, and solidification and stabilization operations.
Volume sources include structures within processing facilities, and may include individual
tanks or tank farms, and chemical storage containers. Line sources include paved and
unpaved roads.
3-1
-------
For particular guidance on determining source emission rates during pre-remediation,
remediation, and post-remediation activities, the reader Is referred to the Air/Superfund
National Technical Guidance Study (NTGS) series documents that exist on this subject (see
Appendix E) A good overview document for those interested in estimating source emissions
would be Models for Estimating Air Emission Rates from Suoerfund Remedial Actions'.
For dispersion modeling, an important consideration in emissions estimation is the
averaging period. Depending on the design of the air pathway assessment (APA), it may be
necessary to estimate both the long-term (i.e.. annual) and short-term (24-hour or less)
emissions potential from the source. Although some of the emissions from Superfund sources
include reactive constituents, the phenomenon of chemical reactivity is not addressed in the
models described in this document.
General characteristics of sources associated with the primary phases of the
Superfund process are shown in Table 3-1. The source classification for modeling of the
primary Superfund sources is indicated, as well as the important air emission mechanisms.
Each source may also be described as having a fundamental release classification (for some
sources, multiple classifications may apply). For pre-remediation sources, the document
Guidance for Baseline Emissions Estimation Procedures for Suoerfund Sites39 should be
reviewed.
SELECTING POLLUTANTS TO MODEL
Selecting the specific toxic air pollutant compounds to model is generally less
critical than when selecting target analytes for ambient air monitoring, where the
selections may be significantly limited by technical, budget, and schedule constraints.
With the sophistication of computer modeling techniques, numerous pollutants may be modeled
for any particular site with relative efficiency. For assessing the impacts of multiple
pollutants from a single source, it is possible to model the source only once with a unit
emission rate (i.e.. 1 gram per second [g/s]). and then scale the results by the actual
pollutant emission rates.
3-2
-------
TABLE 3-1. GENERAL CHARACTERISTICS OF SOURCES ASSOCIATED WITH SUPFRRJND ACTIVITIES
Superfund Source
Typical Source
Classification for
Modeling*1
Important
Air Emission Mechanisms
Release Classification
Gas Phase
Particulate Phase
PRE-REMEDIATION:
co
oo
Landfills
VAPOR CLOOD
SOIL
Fugitive (area)
Gas release from solid
Volati1ization
Wind erosion.
mechanical
disturbances
Lagoons
Fugitive (area)
Low volatility release from
liquid
Volatilization
Wind erosion.
mechanical
disturbances
-------
TABLE 3-1. (Continued)
Typical Source
Classification for
Superfund Source Modeling"
• Contaminated soil Fugitive (area)
surfaces
Important
Air Emission Mechanisms
Release Classification Gas Phase Particulate Phase
Fugitive particulate Volatilization Wind erosion,
mechanical
disturbances
VAPOR CLOUD
co
-Pa.
SOIL
Containers
Fugitive (area,
volume)
Gas release from solid
Low volatility release from
liquid
High volatility release from
liquid
Gas release
Volatilization
Mechanical
disturbance
-------
TABLE 3-1. (Continued)
Superfund Source
Typical Source
Classification for
Modeling3
Important
Air Fmission Mechanisms
Release Classification
Gas Phase
Particulate Phase
Storage tanks
HIGH mtaaan
Fugitive (area)
Gas release from solid
Low volatility release from
liquid
High volatility release from
1iquid
Gas release
Volatilization
NA
co
en
-------
TABLE 3-1. (Continued)
Superfund Source
Typical Source
Classification for
Modeling3
Important
Air Emission Mechanisms
Release Classification
Gas Phase
Particulate Phase
REMEDIATION:
Soil handling
Fugitive (area.
volume)
Gas release from solid
Volati1ization
Wind erosion.
mechanical
disturbances
« Air stripper"
Point
Gas release
Volatilization
NA
BOOZMfT
VAPOR
C^fj^^^^jj^^
-------
TABLE 3-1. (Continued)
Superfund Source
Typical Source
Classification for
Modeling'
Important
Air Emission Mechanisms
Release Classification
Gas Phase
Particulate Phase
Incinerator/
Thermal desorption"
Poi nt
Gas release
Combustion
Combustion
BUOYANT
VAPOR
In situ venting
Solidification/
Stabilization
Point
Fugitive (area.
volume)
Gas release
Gas release from solid
Volatilization
Volatilization
NA
Wind erosion.
mechanical
disturbances
-------
TABLE 3-1. (Continued)
Superfund Source
Typical Source
Classification for
Modeling3
Important
Air Fmission Mechanisms
Release Classification
Gas Phase
Particulate Phase
POST-REMEDIATION:
.c
OJ
CD
Landfills
VAPOR CLOOD
SOIL
• Soil surfaces
VAPOR CLOUD
SOIL
Fugitive (area)
Gas release from solid
Volatilization
Fugitive (area)
Fugitive particulate
Volatilization
Wind erosion.
mechanical
disturbances
Wind erosion,
mechanical
disturbances
"Most Superfund sources are ground level or near ground level, non-buoyant releases.
bSmall stacks where plume is frequently influenced by downwash in the wake of nearby structures.
Emissions may still result during post-remediation, but should be at levels consistent with the goals of the remediation
effort. Sources shown are remediated.
Notes: NA = Not Applicable
-------
For multiple pollutants from multiple sources with the relative emission rates of
pollutants different for each source, the unit emission rate is not suggested. In this
typical case the point of maximum concentration of each pollutant at the fenceline may be at
different locations These different locations would be impossible to identify if all
sources were simply modeled at unity If each source is modeled separately and the
predicted concentrations at each receptor location summed, the resulting concentrations
would be unrealistically high. It is suggested to model on a pollutant by pollutant basis
with the largest source assigned a unit emission rate and all other sources assigned an
emission rate equal to its relative emission strength
The Hazardous Substances List (HSL)36 developed by the EPA for the Superfund program
provides an initial, comprehensive list of target compounds for dispersion modeling. Other
target compounds can be found in the Clean Air Act. Title III (Hazardous Air Pollutants).
Section 112. Compounds included in the Applicable or Relevant and Appropriate Requirements
(ARARs) should also be used to identify candidate pollutants for modeling. To the extent
possible, the target compound list should be based on source and ambient air monitoring
(AAM) results. The Remedial Investigation/Feasibility Study (RI/FS) should identify most or
all of the contaminants present at the site.
It frequently is not practical to address every emitted compound in the modeling
analysis, so typically a subset of compounds - referred to as target compounds - is
selected. When conducting dispersion modeling for refined APAs. target compounds should
include, at a minimum, all contaminants with concentrations greater than or equal to 10
percent of the appropriate health-based action level. These contaminants are expected to
represent the greatest contributors to potential health impacts. This approach provides a
practical basis for addressing refined modeling at sites with a large number of potential
emission compounds (e.g.. over 100) of which only a limited subset significantly affects
inhalation exposure estimates. However, it is generally recommended to also evaluate all
appropriate site/source-specific contaminants, as practical, for refined modeling APAs
(especially if the cumulative effect due to exposure to a mixture of constituents is used
for comparison to health criteria). RAGS Part A34 contains a recommendation that all
compounds be addressed that represent 1 percent or more of the total risk from a given
exposure pathway. The pollutants selected for modeling at a particular site must ultimately
meet the approval of the Remedial Program Manager.
3-9
-------
SOURCE DEFINITION
Generating the source inventory for modeling is intertwined with the creation of the
pollutant inventory. Each emissions source and the constituents each source emits must be
specifically identified. For dispersion modeling, each source will need to be classified as
a point, area, volume, or line source. Building the source inventory usually begins with
mapping the locations of point sources and the locations and spatial extent of fugitive
sources on a site plot plan, drawn to scale. This plot plan should also indicate the
location of the site property boundary, and any on-site receptors of interest. Such a
drawing identifies the necessary near-field, source-receptor relationships for modeling
Refined dispersion models can currently accommodate a large number of sources;
therefore, the modeler should not feel unnecessarily constrained to limit the number of
sources involved in the analysis. A large area source such as a landfill, for example, can
be subdivided into multiple smaller area sources. This is a good way to account for any
spatial or temporal variability in emissions over the source as a whole. Unique sources
will be defined by the multiple emission points resulting from activities at the site. For
example, an initial breakout of sources resulting from soil excavation would be the
excavation pit, the area over which the excavated material is transported, and the short-
term storage piles.
Because source inputs vary with the type of source modeled, an important first step
in creating the inventory is to identify each source of emissions as a point, area, volume,
or line source. With the source types established, the appropriate model inputs can be
determined. The following subsections describe the various source types and associated
inputs for modeling.
Point Source Characterization
Point sources involve the release of emissions from a well-defined stack or vent, at
a known temperature and flow rate. Consequently, characterizing point sources for modeling
is fairly straightforward. The basic model inputs for any point source are: stack height
above ground level; inside diameter at stack exit: gas velocity or flow rate at stack exit;
gas temperature at stack exit: building dimensions (for stacks subject to downwash, to be
discussed in Section 5.5): and emission rate. The location of the source will also need to
be defined in terms of the model receptor grid used (see Section 5.3).
3-10
-------
The influence of any air pollution control equipment (e.g.. carbon absorption and
incineration in the case of air strippers and in situ venting; wet absorbers and scrubbers
in the case of incinerators) on a pollutant specific basis should be taken into account when
defining model inputs The presence of air pollution control equipment can alter the gas
exit temperature and flow rate which effect plume bouyancy. In defining point source inputs
for modeling, the focus should always be on the characteristics of the exhaust as it is
released to the atmosphere. APC equipment may also affect the particulate size distribution
of controlled particulate emissions.
In the event that there are multiple point sources at the site (e.g.. multiple air
stripping towers), it may be possible to conduct the modeling by treating all of the
emissions as coming from a single, representative stack (particularly useful when conducting
a screening-level analysis) when using the SCREEN2 or TSCREEN models. Merging stacks is
appropriate if 1) the individual point sources emit the same pollutant(s), 2) have similar
stack parameters. 3) are within about 100 meters (m) of each other, and 4) the maximum
distance between any two stacks is small relative to the distance between any stack and the
closest receptor. For each stack, the following parameter M would be calculated as shown
below
M
where.
T,
Q
Q
merged stack parameter that accounts for the relative influence of
stack height, plume rise, and emission rate on concentrations
stack height (m)
(n/4)ds2vs = stack gas volumetric flow rate (cubic meters per second
[m3/s])
inside stack diameter (m)
stack gas exit velocity (meters per second [m/s])
stack gas exit temperature (Kelvin [K])
pollutant emission rate (g/s)
3-11
-------
The stack with the lowest calculated value of M is designated the representative stack The
sum of the emissions from all stacks is assumed to be emitted from the representative stack
(i.e.. dispersed based on the parameters of that stack). To be conservative, it is
recommended to use the closest location to the receptors of interest as the source location
of the stacks being merged.
Parameters from dissimilar stacks should be merged with caution. For example, if the
stacks are located more than about 100 m apart, or if stack heights, volumetric flow rates.
or stack gas exit temperatures differ by more than about 20 percent, resulting model impacts
due to the merged-stack procedure may be unacceptably high
Area Source Characterization
Various types of toxic waste sources fall into the area source category. For a
Superfund site these sources include landfills, waste lagoons, evaporation and settling
ponds, and regions where long-term exposure to toxic chemicals has contaminated the soil.8
For all of these sources, pollutants are emitted at or near ground level. The sizes of
these sources can range from a few square meters in the case of settling ponds, to a few
square kilometers or larger in the case of contaminated soils.
Emissions from area sources are assumed to be of neutral buoyancy. Therefore, plume
phenomena such as downwash and impaction on elevated terrain features are not considered
relevant for modeling area sources. The emission rate for area sources is unique in that it
is entered in units of mass per unit time per unit area [e.g.. g/(s-meters squared[m2])]. It
is an emission flux rather than an emission rate. As an example, assume the pollutant
emission rate from a small lagoon is 150 g/s The dimensions of the lagoon are 10 m by 20 m
(total area is 200 m2). If this source were modeled as a single, square area source, then
the modeled emission flux would be 0.75 g/s-m2 (150 g/s - 200 m2). If the source were
subdivided into smaller area sources, the individual area source emission rates would be
determined by multiplying the modeled emission rate based on the total area by the relative
fractions of the total area represented by the individual area sources. The emission flux
for each sub-area will be the same as for the total pool. For example, if the source were
modeled as two square area sources, each of dimensions 10 m by 10 m (100 m2), then the
modeled emission rate for each source would be 75 g/s [(150 g/s) x (100 m2 - 200 m2)].
3-12
-------
The emission flux would still be the same as the total pool example. 0.75 g/s-m2 (75 g/s -
100 m2). Summing the modeled rates for the two areas yields the modeled emission rate for
the total area of 150 g/s. It is easier to work with emission fluxes for area sources than
emission rates since the emission flux is the same for all sub-areas of the source. Only
the total area of the source must be checked to ensure correctness.
For dispersion modeling, the important parameters used to characterize area sources
are location, geometry, and relative height. In current models (for example, the current
versions of the Industrial Source Complex [ISC2]9 and TSCREEN10) area sources are defined by
the location of the southwest corner of a square and a side length. As this document is
being written revisions are being made which expand the definition capability refined
modeling The location is expressed by a single east-west (X) and north-south (Y)
coordinate of a corner (normally southwest) of a square or rectangular geometric shape. The
side lengths need to be defined. By default the area source is assumed to be a square. The
area source can also be rotated about the specified corner for areas not aligned north-
south. An area source of irregular shape (i.e., neither a square nor a rectangle) can be
simulated by dividing the area source into multiple squares and/or rectangles that
approximate the geometry of the source. The particular model user's guide will need to be
consulted to verify whether both square and rectangular definitions of the source are
allowed. The model user's guides normally provide examples of input for complex area
sources.
If the area source is not at ground level, a height for the source may be entered
(for example, a non-zero value would typically be entered for the height of a storage pile).
If the release height of the source is greater than approximately 10 m, it should probably
be modeled as a volume source.
EPA is currently in the process of revising the area source algorithm contained in
the ISC2 model. In this new version, there will be no restrictions on the placement of
receptors relative to the area source(s). Receptors will be allowed within an area source
itself and at the edge of an area source. The model will integrate over the portion of the
area that is upwind of the receptor (specifically for portions of the area that are no
closer than 1 m upwind of the receptor).
3-13
-------
In the interim, an alternative technique is recommended for defining the receptor
grid in relationship to the area source location(s). The general recommendation is to
subdivide the area source into smaller area sources if the separation distance between the
area source and a receptor is less than the length of the side of the area source. Hence.
the area source nearest the receptor(s) would be subdivided into smaller squares such that
there is no area source with a distance to a receptor of less than the source's side length
(this may not be practical if the receptors of interest include on-site workers). For
specific guidance, the particular model user's guide should be consulted. The EPA document
Review and Evaluation of Area Source Dispersion Algorithms for Emission Sources at Suoerfund
Sites8 may also provide some perspective on modeling area sources
Volume Source Characterization
There are two basic types of volume sources: surface-based or ground-level sources
that may also be modeled as area sources, and elevated sources. Most of the Superfund
release sources can be regarded as surface-based sources. The effective emission height of
a surface-based volume source, such as a surface rail line, is usually set equal to zero.
An example of an elevated volume source is an elevated conveyor with an effective emission
height set equal to the height of the conveyor. A source may be defined as a volume source
for modeling when its emissions can be considered to occur over a certain area and within a
certain depth of space. At a Superfund site, fugitive exhaust from on-site structures such
as tanks, or a treatment facility may be modeled as a volume source. A roadway over which
contaminated soil is hauled may also be modeled as a series of volume sources. As with area
sources, emissions from volume sources are assumed to be of neutral buoyancy.
The important parameters used to characterize volume sources for dispersion modeling
are location and initial lateral and vertical dimensions. The particular model user's guide
will have instructions on defining the initial lateral and vertical dimensions of the
source. The length of the side of the volume source will need to be known, as will the
vertical height of the source, and whether it is on or adjacent to a structure or building.
Generally, the north-south and east-west dimensions of each volume source must be the same.
For refined modeling, the location is simply expressed by a single east-west (X) and north-
south (Y) coordinate.
3-14
-------
Line Source Characterization
Line sources are typically used to represent roadways. Certain dispersion models
differentiate line sources from area or volume sources In these cases, basic model inputs
are the overall source length, width, and height. Emissions may be entered in units of
grams per meter per second.
Line sources may also simply be modeled as a series of area or volume sources In
the case of a long and narrow line source, it may be impractical to divide the source into N
volume sources, where N is given by the length of the line source divided by its width.
Dividing the length of the line source by its width effectively splits the line source into
a string of squares (for example, if the length of the line source was 100 m, and the width
was 5 m, then the line source could be split into twenty, adjacent square volume sources)
An approximate representation of the line source can be obtained by placing a smaller number
of volume sources at equal intervals along the line source (for example, for the line source
of length 100 m and width 5 m, a total of 10 square volume sources separated from one
another by 5 m could be defined) With this option, the spacing between individual volume
sources should not be greater than twice the width of the line source. A larger spacing can
be used, however, if the ratio of the minimum source-receptor distance and the spacing
between individual volume sources is greater than about 3.
CONTAINER/ACUTE RELEASES
Although not a typical concern for Superfund sites, except in contingency planning.
the highest concentration impacts can occur from a short term release of a gas or liquid in
a container or a solid such as burning tires. If the potential for accidental releases is
known to exist prior to remediation, the site Health and Safety Plan should include a
contingency plan, based in part on dispersion modeling results, for addressing such
situations.
Descriptions of the possible source terms from such releases and subsequent
dispersion modeling are described in Contingency Analysis Modeling For Superfund Sites and
Other Sources.11 and Guidance on the Application of Refined Dispersion Models to
Hazardous/Toxic Air Pollutant Releases." Appendix F outlines how these two references can
be used to determine the source term for such releases.
3-15
-------
In a typical Superfund source the release is not denser than air. Many of the
container releases result in dense gas releases. Most hazardous chemicals have molecular
weights greater than air's molecular weight. Chemicals stored under pressure cool from
expansion during a release. Chemical vapor parcels which are cooler than the surrounding
air (even if the chemical's molecular weight is the same as air's) will be more dense than
the surrounding air
3-16
-------
SECTION 4
MODEL SELECTION
Determining the level of model sophistication and the specific model to use are
integral to a meaningful Superfund Air Pathway Assessment (APA). This section, therefore.
is designed to assist the modeler in determining what approach to take for a particular
analysis, and presents the major considerations. The EPA has approved numerous models for
use in regulatory application. Non-regulatory models also may be used if it can be shown
that they are more suitable for a given scenario. The use of any non-EPA model should be
reviewed and approved by the RPM 01 their designate before the modeling is performed It is
not the intention of this manual to discuss each available technique; rather, the most
generally applicable and commonly used models are mentioned. For any given APA, it is the
responsibility of the modeler to ensure that the most appropriate technique is selected.
Uhere possible, models selected to provide estimates of ambient concentrations should
be consistent with the requirements and guidance specified in the Guideline.on Air Quality
Models (Revised)' Since dispersion models are periodically revised, the model user should
verify that the most updated version of code is being executed for the APA. For models
issued by the EPA Office of Air Quality Planning and Standards (OAQPS). this can be done by
checking the Support Center for Regulatory Air Models (SCRAM) Bulletin Board System (BBS).
as described in Section 4.5.
In general, determining the level of model sophistication and making the appropriate
model selection will depend on the following key factors:
Site-specific goals of the APA;
« Superfund dispersion modeling objectives:
Legal and liability aspects of the Superfund project: and
Pragmatic aspects of the program, including:
Quality and availability of the input data, including the ability of
the emission models to adequately simulate emission rates and their
variability.
Applicability of existing dispersion models to site-specific
characteristics, including source types.
5-1
-------
Ability of existing dispersion models to reasonably simulate transport
and dispersion of the particular air pollutants of interest released
from the site, given the chemical and physical processes involved, and
Ability to accomplish the dispersion modeling objectives with modest
uncertainties.
The overall goal of a Superfund APA is to evaluate the exposure of the off-site
population, and the impact to the environment, depending on the phase and related activities
of the Superfund program As discussed in Volume I of this APA series, air monitoring is
usually performed to determine worker exposure because on-site personnel may work relatively
close to emission sources and they tend to move around over time. Within the scope of this
document, exposure of the off-site populace and impact to the environment are a basic matter
of producing adequate concentration and deposition estimates in the locations of interest
In this section, determination of the proper model to use is discussed in the context
of a general two-step procedure for assessing air quality impacts. This two-step procedure
involves an initial screening-level analysis to obtain conservative estimates of air quality
based on limited data, followed by a refined analysis, as necessary, to provide more
realistic estimates of air quality based on more detailed model inputs. The following
subsections further describe each of these classifications and their applicability.
SCREENING ANALYSIS
Screening-level dispersion modeling involves simplified calculation procedures
designed with sufficient conservatism to determine if a source of pollutants (1) is clearly
not a threat to air quality, or (2) poses a potential threat that should be examined with
more sophisticated estimation techniques or measurements. Therefore, screening-level
dispersion modeling techniques provide conservative estimates of air quality impacts. These
techniques also eliminate the need for further, more refined modeling if the impacts on air
quality are shown to not pose a risk to public health or the environment. Based on
simplified procedures, screening-level modeling is more readily implemented than a refined
modeling approach.
4-2
-------
There are two primary distinctions between screening-level techniques and refined
modeling techniques. One distinction is that the screening-level approach incorporates
"generic" meteorological data, thus eliminating the need for site-specific meteorological
data. The maximum concentration for any averaging period is considered to occur in any
direction from the source Screening-level models typically estimate hourly average
impacts, with impacts for other averaging periods (e g.. 3-hour, 8-hour, 24-hour, and
annual) derived through the use of time scaling factors.
The second primary distinction between screening and refined models is that screening
models can only estimate the impacts from a single source with each model execution. When
impacts are required for multiple sources, the sources must be processed separately, and
their individual, maximum impacts summed to produce the total maximum impact The greater
the distance between sources, and the less similarity they have in dispersion
characteristics, the more conservative this summed impact will be. This is because
dissimilar sources tend to have maximum impacts in different places, especially when located
some distance apart.
Various screening-level techniques are identified in the Guideline on Air Quality
Models {Revised) • Two interactive screening models that are commonly used to predict air
quality impacts are SCREEN213 and TSCREEN10. The SCREEN2 model incorporates the methodology
presented in Screening Procedures for Estimating the,_ALr_Qualitv JmpaQt...of- Stat-ipngry
Sources .(Revised).14 and is capable of predicting impacts for point, volume, and area sources.
Area source calculations are performed using a finite line segment approach, consistent with
the Industrial Source Complex (ISC2) models (described in the following subsection) The
volume source calculations are also consistent with ISC2: however, unlike ISC2, the volume
source algorithm in SCREEN2 is for single-volume sources only. TSCREEN10 incorporates the
dispersion algorithms used in SCREEN2 and also provides algorithms for calculating some
source terms and for estimating noncontinuous releases. The TSCREEN model incorporates the
procedures documented in Workbook of Screening Techniques for Assessing Impacts of Toxic Air
Pollutants (Revised).15
4-3
-------
Both models are capable of predicting concentrations in simple terrain (elevation
below stack top) and intermediate/complex terrain (elevation above stack top). They use a
range of stability classes and wind speeds to identify those meteorological conditions
(e.g., combinations of wind speed and stability) resulting in maximum ground-level
concentrations.
Screening models are most appropriate for assessing the air quality impacts of single
sources, and sources with continuous, constant emission rates. Screening-level dispersion
modeling is applicable to the screening step of the Remedial Investigation/Feasibility Study
(RI/FS). and can be used to provide a preliminary indication of the potential impacts of
possible remedial alternatives. The availability of representative meteorological data is
probably the most significant determining factor in applying a screening model--without
representative meteorological data, refined dispersion modeling cannot be conducted.
In the past, screening techniques for estimating ambient concentrations have been
developed that do not involve use of a computer model. These techniques utilize formulas
and predefined charts or tabular values relating concentration with downwind distance from
the source. Examples of these approaches may be found in A Tiered Modeling Approach for
Assessing the Risks Due to Sources of Hazardous Air Pollutants'6 and Guideline for Predictive
Baseline Emissions Estimation Procedures for Superfund Sites17. With the development of
"user-friendly" computerized screening techniques such as TSCREEN and SCREEN2, it is
recommended that these computerized techniques be used to ensure that the results reflect
the latest modeling guidance.
REFINED ANALYSIS
Refined dispersion modeling requires more detailed and precise input data and.
consequently, provides more accurate estimates of source impacts. Refined dispersion models
have been developed for both simple and complex terrain and for rural and urban
applications. Thus, the topography and land use in the area surrounding the facility must
be evaluated to determine the appropriate model The model selected should most accurately
represent atmospheric transport and dispersion in the area under analysis.
4-4
-------
Refined dispersion modeling requires additional detail in meteorological data.
definition of the receptor grid, and definition of the emission sources. Actual coordinates
of the sources need to be specified, and for point sources, the building dimension data to
account for downwash are considerably more extensive. Another distinguishing feature of
refined dispersion models is their ability to vary source emission rates as a function of
time and/or meteorological conditions.
Many refined models require meteorological data in the form of hourly weather
observations and twice-daily mixing heights that are processed into a format suitable for
model execution Models designed to predict only long-term averages commonly use ^ability
Array (STAR) summaries, which are joint frequency distributions of wind speed, wind
direction, and Pasquill-Gifford atmospheric stability class.18 For the receptor grid, a
spatial array of locations where concentration predictions are desired must be defined
relative to some user-specified grid origin. Unlike screening models, receptor impacts from
refined models are no longer simply a function of downwind distance from the source, but are
a joint function of distance and orientation from the source. Meteorological data and
receptor requirements for refined dispersion modeling are further described in more detail
in Sections 5 1 and 5 3, respectively
One of the most commonly used models for conducting a refined dispersion modeling
analysis is the Industrial Source Complex Model, which has a long-term version (ISCLT2) and
a short-term version (ISCST2).9 The short-term version can also produce long-term
concentrations, and is often used in an analysis when both short-term and annual average
concentrations are required. Although there are numerous models appropriate for regulatory
application, the ISC2 models have experienced widespread use because of their versatility.
The ISC2 models have the following attributes:
• accommodate multiple point and/or fugitive sources;
« allow for a sophisticated treatment of building downwash for point sources;
• predict impacts for flat and rolling terrain;
predict impacts for urban and rural land use classifications; and
• allow input of time-varying emission rates (e.g.. emission rates may vary by
season, month, or hour-of-day)
4-5
-------
Another model, recommended for use in urban areas, is the Gaussian-Plume Multiple
Source Air Quality Algorithm (RAM)19. This model is a steady-state Gaussian plume model used
for estimating the impacts of multiple point and area sources in flat terrain and urban
settings.
In summary, refined dispersion modeling is applicable to all phases of the Superfund
process, provided representative meteorological data are available. Relative to screening
models, refined models generally provide more accurate estimates of the impact of Superfund
sources on public health and the environment by relying on fewer assumptions and providing a
consistent means of making multiple, detailed calculations in a single execution. The
output from refined models can provide an extensive amount of information. For example, the
impacts at a large number of receptors for varying averaging times is a basic outcome of
many refined models. For modeling sources with continuous emission rates located in a rural
area, the ISC2 models should be given first consideration. If multiple sources in an urban
area are to be modeled, and terrain is not a significant consideration, the RAM model should
be used For modeling sources with instantaneous emission rates, the TSCREEN10 model or a
refined model such as described in Guidance on the Application of Refined Dispersion Models
to Hazardous/Toxic Air Pollutant Releases12 may be used.
DENSE GAS RELEASE SIMULATIONS
Dense gas modeling is not a typical consideration for Superfund sites except for
contingency modeling. Releases of liquids or gases from containers can lead to dense gas
clouds. Knowing whether a release should be treated as a dense gas (heavier than air)
release is important to selecting the appropriate model. If a release can be considered
neutrally buoyant or lighter-than-air, standard "passive" dispersion modeling or a model
capable of handling neutrally buoyant releases should be applied. If not neutrally buoyant.
then the analysis requires the use of a specialized model to adequately characterize such
phenomena.
To determine whether a release should be considered a dense gas release, a comparison
is made between the Richardson number describing the release and a selected, known value.
(In atmospheric science, the Richardson number is a measure of dynamic stability.)
-------
The formulation of the Richardson number and. thus, the calculation, depends on the release
density, and whether the release is instantaneous or continuous. The equations for
determining the density of a release are given in Section 4.13 of Guidance on the
Application of Refined Dispersion Models to Hazardous/Toxic Air Pollutant Releases.13
Over the last few years, much focus has been placed on the development of models to
address the dispersion of dense gas releases. Guidance on the Application of Refined
Dispersion Models to Hazardous/Toxic Air Pollutant Releases12 discusses ana presents example
applications for five dense gas dispersion models: the Hgnse £&s Qnpersion model (DEGADIS).
the SLAB model, the ADAM model, the Area! Locations of Hazardous Atmospheres (ALOHA) model.
and Heavy gas System (HGSYSTEM). This document should be consulted as a starting point for
understanding the requirements of dense gas dispersion models.
DEPOSITION MODELING
Remedial activities at National Priorities List (NPL) sites very often involve the
handling of contaminated soil. This handling may result in fugitive dust emissions, that
can carry inorganic and organic constituents in. or attached to, the dust particles One
remediation activity that may be a significant source of fugitive dust is
solidification/stabilization.20 Non-remediation sources of fugitive dust emissions include
storage piles and dry impoundments. Particulate matter (PM) impacts should be evaluated
even when the material is not contaminated, because particulate matter less than 10
micrometers (A/m) in diameter (PM10) is a criteria pollutant. If the particulate emissions
are contaminated, the fraction of contaminant will need to be determined to properly
establish the actual emission rate of the contaminant.
In addition to estimating ambient concentrations of the particulate emissions (used
in evaluating inhalation exposure), it may be important to characterize particle deposition
as part of the overall risk assessment The importance of assessing particle deposition for
a given site will depend on the size of the particulate matter involved and the proximity of
receptors. The smaller the particles, the greater the likelihood that deposition will occur
off site For example, particulate emission factors and equations defined by EPA are given
as a function of particle diameter.21
4-7
-------
Large particles with diameters greater than 100 urn are likely to settle within 5 to 10
meters (m) of the emissions source, while those that are 30 to 100 ym in diameter are likely
to settle within 100 m or so of the source, except in cases of high atmospheric turbulence.
Smaller particles, especially those with diameters less than 10 to 15 /vm. are much more
likely to stay airborne
Deposition is actually a complex process that is highly dependent on the specific
location, meteorology, and chemical species Dry deposition is strongly influenced by the
particle size distribution of the particulate species and by meteorological and surface
characteristics The larger the particle, the greater its deposition velocity; however.
there is a point where the deposition velocity of smaller particles actually increases
because of the effect of turbulence Deposition models calculate a gravitational settling
velocity and a deposition velocity for eacn particle size class. As its name suggests, the
gravitational settling velocity accounts for removal of particulate matter due to gravity.
This mechanism is only significant for particles in the larger size ranges (i.e., greater
than 20 to 30 im in diameter) because only the larger particles have sufficient mass to
overcome turbulent eddies. The deposition velocity accounts for PM removal by all methods.
including turbulent motion, which brings the particles into contact with the surface and
allows them to be removed by impact!on or adsorption at the surface. Wet deposition
includes the effect of precipitation scavenging
The "deposition" produced by dispersion models is actually a flux, or the mass of
particulate deposited over a square area over a unit of time (e.g., micrograms per meter
squared per hour [/vg/m2/hr]). The magnitude of the deposition flux will directly affect the
soil concentration of the pollutant and, consequently, the level of human exposure through
direct soil ingestion. plant and animal consumption, and dermal contact with soil. The
deposition flux over water bodies (e.g.. reservoirs, lakes, and streams) may also affect the
level of exposure through human consumption of fish and drinking water.
Deposition modeling requires the definition of particle size categories (typically up
to 20 categories may be defined). For each category, the particle diameter, mass fraction.
and density are specified. Using as many particle size categories as possible helps ensure
that the most representative results are obtained. Generally, definition of these
-------
parameters will need to be done on a site-specific basis. Proper source testing provides
the best information, however, data previously collected from another site with similar
characteristics (e.g . similar soil characteristics) may provide information. In the event
site-specific data are not available, the modeler should consult with the EPA Regional
Modeler should be committed.
Techniques for evaluating particle deposition are currently being developed by the
EPA Although the ISC2 model allows for deposition estimates to be made, its present
algorithm is appropriate for large particles dominated by gravitational settling (i.e.,
particles with diameters larger than approximately 20 ^m) but not for small particles or
gaseous pollutants A draft revision to the deposition algorithm in the ISC2 model
(ISCSTDFT) incarcerates a new method for estimating deposition velocity and will account for
plume depletion by removing mass from the plume as it is deposited on the surface. The net
result will be a new method for obtaining dry deposition flux of particulate matter. The
model will also include an estimate of wet deposition amount.
In addition to the basic inputs (particle diameter, density, and mass fraction), the
improved methodology requires the following site-specific parameters: surface roughness
height, displacement height, noon-time albedo, soil moisture availability parameter,
fraction of net radiation absorbed by the ground, anthropogenic heat flux, and minimum
Momn-Obukhov length In addition to these parameters and the standard set of
meteorological variables, the methodology requires the Momn-Obukhov length and friction
velocity To provide these parameters an additional processor (provided with the model) is
required. This processor, the Dry DEPosition METeorological processor (DDEPMET) needs as
input a RAMMET output file and NWS surface pressure and cloud cover data in the CD144
format. An additional processor, precipitation merge program (PMERGE) merges hourly
precipitation data (amount and type) with the output from DDEPMET to compute wet deposition.
Guidance on determining these values and using DDEPMET will be provided in the user's
instructions associated with the new model. Section 5.1.4 of this document may be referred
to for guidance on determining surface roughness. In the Air/Superfund National Technical
Guidance Study (NTGS) Series document Estimation of Air Impacts for the Excavation of
Contaminated Soil23, a typical value of 2.65 g/cm3 is presented. Density may be less than 1.0
g/cm3 for particles from combustion sources such as incinerators.
4-9
-------
In the interim, an alternative technique that may be considered for Superfund sites
is the gradient-transfer deposition algorithm contained in the Fugitive Dust Model (FDM).22
The ability to treat both turbulent and gravitational removal mechanisms is a key feature of
this model. Sources entered into the model may be point, line, or area sources. Because
this model is not designed to compute the impact of buoyant point sources, it will be overly
conservative for point sources with significant plume rise.
A unique characteristic of fugitive dust is that its emission rate is often a
function of wind speed. In the FDM. emission rates may be defined as varying with wind
speed based on a simple power law formula and a threshold wind speed. The same formula can
be used to calculate emission rates for use in ISC2
Deposition actually occurs through both dry and wet processes. The emphasis in
regulatory model development has been on the dry deposition process. A discussion of wet
deposition (also known as precipitation scavenging) can be found in Chapter 11 of
Atmospheric Science and Power Production 24
Another phenomenon that could affect air quality is particle resuspension.
Resuspension occurs primarily as a result of mechanical disturbances of the soil, such as
vehicular traffic, but is also a function of wind disturbance. Mechanical stresses can
raise particles from the ground into the main airstream so that they are more rapidly
transported downwind than they would be by general wind resuspension. Variables that affect
resuspension include particle, soil, and surface properties, particle-soil interaction.
topography, and weather conditions. It may be important to consider the effects of particle
resuspension if particle inhalation is of concern. As with wet deposition, particle
resuspension has not been emphasized in regulatory model development. Chapter 12 of
Atmospheric Science and Power Production24 provides a general discussion of resuspension
rates and factors.
4-10
-------
MODEL AVAILABILITY
Source and executable code for regulatory dispersion models can be obtained from the
SCRAM BBS This system is part of the OAQPS Technology Transfer Network (TTN), and is
managed by the Air Quality Modeling Group. Emissions. Monitoring and Analysis Division of
OAQPS Model documentation and other support materials can also be obtained from the SCRAM
BBS. Information necessary to connect with the SCRAM BBS is shown in Table 4-1. The TTN
help line can be accessed by dialing (919) 541-5384.
TABLE 4-1 SCRAM BBS COMMUNICATION PARAMETERS
Modem Telephone Number
(919) 541-5742
Baud Rates
1200 - 9600, 14. 4K
Line Settings
8 data bits
no parity
1 stop bit
Model code can also be obtained for a fee through the National Technical Information
Service (NTIS), which can be reached at (800) 533-6847, and from private vendors. Private
vendors frequently supply interactive or menu-driven data entry programs that can simplify
implementation of the more refined models. When purchasing models through NTIS or from
private vendors, model users should verify that they are acquiring the most up-to-date
versions
5-11
-------
SECTION 5
COMPONENTS OF A MODELING ANALYSIS
This section presents various topics that are relevant to any dispersion modeling
analysis Specifically, this section discusses components of a modeling analysis other than
those directly pertaining to source characterization, which were discussed in Section 3
This section is designed to give the Superfund modeler an understanding of the integral
components of a modeling analysis, and an appreciation for how the treatment of each
component directly affects the concentrations that are predicted. The U.S. Environmental
Protection Agency (EPA) document Gin del me on Air Quality Models (Revised)' is the principal
source of guidance on the various components of an air dispersion modeling analysis
METEOROLOGICAL CONSIDERATIONS
Meteorological conditions govern the transport and dispersion of contaminants and, in
the case of some fugitive sources, such as lagoons or landfills, can affect the amount of
contaminant that becomes airborne It is important, therefore, to use meteorological data
that are representative of the site area and vicinity. A minimum of either one year of on-
site data or five years of off-site (NWS) data is required to run refined dispersion models.
If long-term risk is an issue, it is desirable to have five or more years of on-site
meteorological data to support long-term exposure assessments for refined Air Pathway
Assessments (APAs). As stated in the Guideline on Air Quality Models (Revised).1 a five-year
data set should capture the variability in maximum predicted concentration that could occur
over a longer time span.
Certain models, such as dense gas release models, are executed on a single set of
meteorological conditions. In this situation, meteorological conditions producing worst-
case impacts should be determined (see various subsections within this section).
It is recommended that an on-site meteorological monitoring program be initiated
immediately after a site is included on the National Priorities List (NPL) if representative
data are not available. Even at flat terrain sites where nearby NWS data are available, it
is recommended that an on-site meteorological station be installed and operated during the
remedial action (RA) phase.
5-1
-------
The short-term temporal and spatial variability of wind conditions limits the applicability
of off-site meteorological data for real-time decision making (for example, during non-
routine air releases)
In the absence of a year of on-site data, if representative, data applicable for use
in dispersion modeling may be available from National Weather Service (NWS) stations or
stations operated by state meteorological programs. Selected NWS data may be obtained from
the Support Center for Regulatory Air Models (SCRAM) electronic Bulletin Board System (BBS)
(see Section 4.5) Meteorological data for stations throughout the United States may also
be obtained, for a fee. from the National Climatic Data Center in Asheville. North Carolina
(for placing an order, call 704/271-4800) Other sources of meteorological data may include
nearby universities or military stations. Guidance on determining the representativeness of
off-site data can be found in the Guideline on Air Quality Models (Revised)1 and On-Site
Meteorological Program Guidance for Regulatory Modeling Applications.25
In general, judgements regarding the representativeness of meteorological data must
consider both spatial and temporal dependence. The Superfund site and meteorological
observation locations must have similar spatial characteristics with respect to terrain
features, land use, and synoptic flow patterns. Further, the meteorological data set must
include hourly observations for one year, collected over the four seasons. From a practical
standpoint, if NWS data are available and considered representative. NWS data could be used
in most applications, because such data are subject to well-defined quality
assurance/quality control (QA/QC) programs. The quality of data available from other
sources should be evaluated based on EPA guidance.
If representative data are not available, it is recommended that an on-site
meteorological monitoring program be initiated immediately after a site is included on the
NPL. The meteorological monitoring program should continue throughout the remediation and
post-remediation phases. The quality and siting of the meteorological data collected should
meet EPA requirements, as outlined in the following technical references: On-Site
Meteorological Program Guidance for Regulatory Modeling Applications25. Quality Assurance for
Air Pollution Measurement Systems. Volume IV - Meteorological Measurements32, and Ambient Air
Mom ton na Guidelines for Prevention of Significant Deterioration (PSD)33.
5-2
-------
On-Site Meteorologica1 Program Guidance for Regulatory Modeling Applications25 also
contains guidance on processing of meteorological data for dispersion modeling applications.
To produce a meteorological data set in the proper format for model input, the EPA has
developed the Meteorolpgica1 Processor for Regulatory Models (MPRM)26. This computer program
and the associated user's guide are available from the SCRAM BBS.
Wind Speed and Direction
In dispersion modeling, wind speed is used in determining- (1) plume rise, (2) plume
dilution, and (3) mass transfer rate into the atmosphere (used mostly in fugitive dust and
evaporation rate models). Wind direction is used to approximate the direction of transport
of the plume Most wind data are collected near ground level (the standard height for wind
measurement is 10 meters [m])
The wind speed at release height is frequently determined internally by the
dispersion model using a power law equation. As wind speed increases, plume rise decreases.
plume dilution increases, and concentration estimates predicted by air dispersion models
decrease. However, as speed increases the mass transfer rate also increases (see Section
5.7)
For close-in distances where gravity effects are dominant, dense gas models are less
sensitive to increases in wind speed. Under very light wind conditions, dense gas releases
tend to form "pancake-shaped" clouds near the source, and the dense cloud may not be very
deep until further downwind. At higher wind speeds, the rate of air mixing increases, and
the maximum concentrations decrease. For releases from liquid pools, high wind speed
increases the rate of evaporation and, thus, the emission rate of the source. However, high
wind speed also results in more dilution due to increased entrainment of outside air. which
can lead to a lowering of maximum concentrations.
The variability of the direction of transport (i.e., plume meander) over a period of
time is a major factor in estimating ground-level concentrations averaged over that time
period. Take for example the two "wind roses" shown in Figure 5-1. (The term "wind rose"
is commonly used to refer to an illustration depicting the joint frequency of wind speed and
direction at some location )
5-3
-------
As shown, the distribution of winds in Case A is relatively symmetrical, especially when
compared to that of Case B. Considering wind direction alone, long-term averages predicted
with the Case B data set would be greater than those predicted with the Case A data set. due
to the strong persistence of wind direction in Case B (which reflects data collected in a
mountain valley setting). Wind direction should be estimated from on-site measurements for
emergency removal or accidental release analyses For planning analyses, wind direction
should be chosen to maximize potential off-site impacts.
Calm wind conditions pose a special problem in model applications because Gaussian
models assume that concentration is inversely proportional to wind speed. EPA has developed
a procedure to prevent the occurrence of overly conservative concentration estimates during
periods of calm wind. This procedure acknowledges that a Gaussian plume model does not
apply during calm conditions and that our knowledge of plume behavior and wind patterns
during these conditions does not presently permit the development of a better technique1.
Therefore, the procedure disregards calm hours by using a pre-processor which ignores the
calms. The applicable model user's guide should be consulted to ensure that hours of calm
wind observation are properly interpreted from the meteorological data set. If calm wind
periods are the periods of concern, a nonguideline technique may be used in consultation
with the EPA Regional Modeler.
Atmospheric Stability
Dispersion models currently use stability categories as indicators of atmospheric
turbulence. Based on the work of Pasquill and Gifford, six stability categories have been
defined, where Category A represents extremely unstable conditions and Category F represents
moderately stable conditions.18 Methods for estimating atmospheric stability categories from
on-site data are provided in the Guideline on Air Quality Models (Revised)1 and On-Site
Meteorological Program Guidance for Regulatory Modeling Applications.25
-------
IO
c:
-5
fD
QL
73
O
X
;r
cr
0)
-Q
3
Q.
ro
Q-
-5
ro
r>
o
/ *
Vi i '
• .-' ! : '
o-i •-• »-ie n-n i»-»> J>-«t
in .)(» •)(» •>(' •) (o •) (o •)
WIND SPECO SCALE (KNOTS)
NOTE - WWO DIRECTION IS THE
DIRECTION WIND IS BLOWING FHOM
0-> 4-« J-10 ll-li 1/-II Jl-«
(•i «|( *) <<> «> (o «>
WIND SPEED SCAtE (KNOTS)
NOTE - WIND DIRECTION IS THE
OlBECTION WIND IS Bl OWING fROM
B.
-------
The amount of turbulence in the atmosphere has a major impact on the rise of stack
gas plumes, and upon subsequent plume dispersion by diffusion. Turbulence is a result of
many factors, including: windflow over rough terrain, trees, or buildings (mechanical
turbulence); rising warm air (thermal turbulence); and migrating high and low pressure air
masses Any factor enhancing the vertical motion of air (either rising or sinking), will
increase the amount of turbulence. For a given wind speed, stable atmospheric conditions
provide smaller levels of atmospheric turbulence than do unstable conditions, and can lead
to higher model-predicted concentrations.
For near-field impacts, dense gas releases will be only weakly sensitive to stability
class. As the release becomes neutrally buoyant, the plume is more influenced by
atmospheric conditions such as stability class.
Ambient Temperature. Relative Humidity, and Pressure
Ambient temperature is routinely used in dispersion models to calculate the amount of
rise of a buoyant plume and to calculate evaporation rates. Relative humidity affects the
amount of energy available in the atmosphere for plume entrainment. Atmospheric pressure
data are used in calculating gas and liquid release rates from storage and process vessels,
and from pipes. Therefore, the emission rate from a container at a high altitude (in
Denver, for example) could be different from those of a spill at sea level.
Surface Roughness
The intensity of mechanical turbulence at a site is a function of the surface
roughness. Surface roughness, a required input for some models, is characterized by a
roughness length, which in principle is a measure of the roughness of a surface over which a
fluid (i.e.. the air) is flowing. For a homogeneous surface, the value of the surface
roughness length is sometimes approximated as l/10th of the average height of the surface
irregularity. When the landscape contains obstructions (i.e., is nonhomogeneous). an
effective length must be determined. Typical values of surface roughness length are
provided in Table 5-1 In the event of multiple surface roughness surrounding a site, it is
most conservative to use the lowest value for modeling.
5-6
-------
Mixing Height
The mixing height defines the depth through which pollutants released to the
atmosphere are typically mixed by dispersive processes. The mixing height determines the
vertical extent of dispersion for releases occurring below that height, and releases
occurring above that height are assumed to have no ground-level impact (with the exception
of fumigation episodes) Morning and afternoon mixing heights are estimated for selected
NWS stations from vertical temperature profiles (otherwise known as upper air data) and
surface temperature measurements. Hourly mixing heights for input to dispersion models are
derived from the twice-daily values, based on procedures defined by EPA.
For refined dispersion modeling, values of hourly mixing height for selected NWS
stations are available on the SCRAM BBS. In general, upper air data from the closest.
representative NWS station with topography similar to the site may be used for refined
modeling. The EPA Regional Modeler should be contacted to ensure that the appropriate data
set is selected.
From a climatological perspective, seasonal and annual average values of morning and
afternoon mixing heights are available for selected cities throughout the United States
(Mixing Heights. Wind Speeds, and Potentia1 for Urban Air Pollution Throughout the
Contiguous United States27). Where models require a singular value of mixing height, these
values can be used for planning purposes.
TABLE 5-1. REPRESENTATIVE VALUES OF SURFACE ROUGHNESS FOR A UNIFORM
DISTRIBUTION OF SELECTED TYPES OF GROUND COVER
Surface
Surface Roughness (meters)
Water
Ice
Snow
Sand
Soils
Short grass
Long grass
Agriculture crops
Deciduous Forest
Coniferous Forest
0.1 to 10 0 x 10'5
0.00001
0.00005 to 0 0001
0.0003
0.001 to 0.01
0.003 to 0.01
0.04 to 0.10
0.04 to 0,20
1.0 to 6.0
1 0 to 6.0
References:
Pielke. R.A., 1984. Mesoscale Meteorological Modeling. Academic Press. Orlando. FL.
Oke. T R., 1978. Boundary Layer Climates. Methuen and Co. New York. Ny.
5-7
-------
TERRAIN CONSIDERATIONS
Incorporating the effect of elevated terrain in the vicinity of the site may be
significant in situations involving point sources of emissions. This is because with point
sources, high impacts can be predicted due to plume impaction on terrain at elevations
greater than or equal to plume centerline. Incorporating terrain is generally not a
consideration when modeling fugitive releases because these releases are typically neutrally
buoyant with no plume rise to consider and. hence, are essentially ground-level releases.
Maximum impacts from fugitive releases are thus expected to occur at the nearest downwind
location.
The remainder of this subsection discusses terrain considerations as they apply to
modeling point sources. As a general note, complex terrain models for regulatory use
acknowledge only "unique" features, and do not address the influence of intervening terrain.
For example, if a terrain feature, such as a hill, existed between the source location and
the receptor location, the model predicted impacts at the receptor location would not have
taken the effect of the hill into account In reality, the plume may have impacted on the
hill and been diluted or had its trajectory changed before proceeding further downwind.
Because the Gaussian approach used in regulatory dispersion models assumes that the
plume has a normal or Gaussian distribution in both the cross-wind and vertical directions.
the maximum concentration at any downwind distance would be predicted to be at plume
centerline. Introducing receptor elevations for elevated releases can. therefore, increase
predicted concentrations by effectively bringing the receptor closer to plume centerline.
Whether the maximum concentration for a given analysis will be due to plume impaction will
depend on the proximity of the terrain to the source location. Although the phenomenon of
plume impaction can produce high concentrations, it may not produce the maximum predicted
concentration for an analysis, since sufficient dilution of the plume may have occurred by
the time the plume impacts the terrain feature (i.e.. the concentration at plume centerline
at a given downwind distance may be less than an off-center!me concentration predicted much
closer to the source).
5-8
-------
The need to incorporate terrain elevations is a common occurrence in point source
modeling analyses conducted for regulatory compliance. Over the last few years, much
attention has been focused on this aspect of air dispersion modeling within EPA. The term
"complex terrain" modeling has evolved, with complex terrain generally defined as terrain
exceeding the height of the stack being modeled. Another common term, "rolling terrain,"
pertains to terrain elevations above stack base elevation, but below stack top elevation.
Receptor terrain elevations should be included in any point source analysis, even if the
terrain elevations do not exceed that of stack top (as mentioned, elevating a receptor
brings it closer to plume centerline. for elevated releases).
As stated in the Guideline on Air Quality Models (Revised)1 there are currently five
complex terrain screening techniques that are acceptable for estimating concentrations due
to plume impaction. (1) the Valley Screening Technique (for 24-hour impacts); (2) CTSCREEN,
(3) COMPLEX I; (4) SHORTZ/LONGZ; and (5) Rough Terrain Dispersion Model (RTDM). The Valley
Screening technique is incorporated in the complex terrain dispersion estimates made by the
SCREEN2 and TSCREEN models
The Valley model, COMPLEX I. SHORTZ/LONGZ. and RTDM should only be used to estimate
concentrations at receptors whose elevations are greater than or equal to plume, rather than
stack height. For receptors whose elevations are at or below stack height (i.e., simple
terrain receptors), a simple terrain model should be used. For receptors whose elevations
are between stack height and plume height (commonly referred to as intermediate terrain
receptors), the estimation of concentrations should be considered on a case-by-case basis
with the EPA Regional Modeler.
One technique that is generally acceptable, but not necessarily preferred for any
specific application, involves applying both a complex terrain model (except for the Valley
model) and a simple terrain model. For each receptor between stack height and plume height.
an hour-by-hour comparison of the concentration estimates from both models is made. The
higher of the two modeled concentrations should be chosen to represent the impact at that
receptor for that hour, and then used to compute the concentration for the appropriate
averaging time(s)
5-9
-------
The CTSCREEN model presents another technique26. CTSCREEN may be used to estimate
concentrations at all receptors located on terrain above the stack-top elevation (i.e.. in
both intermediate and complex terrain, where "intermediate" terrain refers to terrain above
stack top, but below plume centerline). CTSCREEN is the screening version of the refined
complex terrain model referred to as Complex Terrain Dispersion Model Plus Algorithms for
Unstable Situations (CTDMPLUS). No meteorological data are required to execute SCREEN2.
TSCREEN. Valley, or CTSCREEN, as these models assume worst-case meteorological conditions.
RTDM and SHORTZ/LONGZ require site-specific meteorological data. COMPLEX I requires site-
specific meteorological data unless the Valley option is chosen for its simulation See the
Guideline on Air Quality Models (Revised)1.
For refined complex terrain analyses, the CTDMPLUS model is preferred29. CTDMPLUS is
applicable to all receptors on terrain elevations above that of stack-top. A simple terrain
model and a complex terrain model may also be used in concert to produce the required
intermediate terrain concentrations. Meteorological data that are spatially and temporally
representative must be used when conducting a refined complex terrain analysis. This
essentially requires that the data be collected on site. Where site-specific data are used
for either screening or refined complex terrain models, a database of at least one full year
of meteorological data is preferred. If more than one year of data is available, it should
be used to verify that maximum concentrations have been predicted.
RECEPTOR DEFINITION
In dispersion modeling, receptors are locations where impacts are predicted. A
receptor grid or network for a Superfund analysis defines the locations of predicted air
concentrations that are used as a part of the APA to assess the effect of contaminant air
releases on human health and the environment under various Superfund site activities.
The receptor grid for a Superfund APA should be developed on a case-by-case basis in
consultation with the Remedial Program Manager and should be a function of the goals
outlined by the Superfund data quality objectives. Various types of receptor grids can be
used. Input of the receptor grid is facilitated by some refined models through an option to
automatically generate a grid based on some user specifications, such as desired interval
spacing. In general, receptor grids are based on either a polar coordinate or Cartesian
5-10
-------
coordinate system, or a combination of both systems. In the Cartesian system, the X-axis is
positive to the east and the Y-axis is positive to the north of a user-defined origin.
Specified in this manner, the coordinates are relative to the grid origin. The X and Y
coordinates may also be specified in terms of Universal Transverse Mercator (UTM)
Coordinates, which effectively removes the concept of a grid origin and allows for each
receptor to be readily mapped or identified. The polar receptor grid is based on radial
distances measured from the grid origin and an azimuth bearing (angle) measured clockwise
from true north
In the polar coordinate system, receptors are usually spaced at
10-degree intervals on concentric rings. Radial distances from the origin are user-selected
and are generally set equal to the distances to the expected maximum concentrations of the
sources modeled. In the Cartesian system, the X and Y coordinates of the receptors are
specified by the user. The spacing of the grid points is not required to be uniform, so
that the density of grid points can be greatest in the area of expected maximum
concentrations. Examples of a Cartesian and polar receptor grid are shown in Figure 5-2.
To establish the location of maximum concentration, two levels of receptor grids are
commonly used in a refined modeling analysis. A first-level, or "screening-level." grid
generally comprises a moderate number of receptors located uniformly in all directions from
the source. Typically, this screening-level grid is centered on a prominent source or
feature (e.g.. a water tower) located within the site boundary. A second-level, or
"refined," grid comprising receptors more densely located, is then modeled to pinpoint
maximum concentrations based on the results obtained by using the screening-level grid.
This refined grid is typically centered on areas of maximum impact defined by the screening-
level grid.
From a geographical perspective, receptors should be located along the site boundary
and in the surrounding area off site. The minimum distance to off-site receptors is usually
defined by the property boundary or fence line. Receptors should be located at. and within.
a far enough distance from the source to ensure that the maximum concentration is
identified. A receptor network extends an adequate distance from the source if one can
observe, from the model results, that impacts reach a maximum at some distance and then
diminish with further distance from the source. To isolate maximum impacts, the emphasis
should be placed on receptor resolution and location, and not on the total number of
receptors modeled.
5-11
-------
For the purposes of conducting a Superfund APA, specific receptor locations are also
of interest. All "sensitive" receptor locations within a given distance (e.g , 10
kilometers [km]) of the site, and Individual residences and other habitations near the site
(e.g., within 1 km) should be identified. Sensitive receptor locations include schools.
work areas, and hospitals associated with sensitive population segments, as well as
locations where sensitive environmental flora and fauna exist, including parks, monuments.
and forests. Receptors may also be placed at the work areas on the site and at the
locations of air monitoring stations, so that comparison of predicted and monitored
concentrations can be made. For input to a risk assessment, it may also be necessary to
place receptors within areas relating to specific exposure pathways, such as waterbodies,
dairy farms, playgrounds, and so forth.
Receptor placement requires special attention when modeling in complex terrain. In
such cases, highest pollutant concentrations are often predicted to occur under very stable
atmospheric conditions, when the plume is near, or impinges on, the terrain. Under these
conditions, the plume may be quite narrow in the vertical, so that even relatively small
changes in a receptor's elevation may make a substantial change in the predicted air
pollutant concentrations. Terrain heights should be entered for each receptor and, as
described below, each receptor distance should be entered, if the site is located in an area
of rolling or complex terrain.
Certain screening-level models, such as TSCREEN. estimate the maximum impacts
irrespective of direction from the source Therefore, receptors are simply expressed in
terms of distances considered to be downwind of the source. At a minimum, the user must
specify the nearest and farthest receptor distances at which air pollutant concentrations
are to be predicted. The model will then automatically calculate impacts at distances
within that range, and will interpolate to find the maximum value and associated distance.
The farthest distance should be set sufficiently large to ensure that the maximum
concentration is identified.
5-12
-------
A. Cartesian Receptor Grid
B. Polar Receptor Grid
Figure 5-2. Examples of Cartesian and Polar Receptor Grids
5-13
-------
From a practical standpoint, most Superfund sources involve ground-level releases.
Only a few sources are elevated, and even these can be classified as low-level, elevated
sources. Examples include on-site treatment facilities involving incinerators and/or air
strippers. This implies that, for most Superfund releases, the highest pollutant
concentrations will occur at short distances from the source. Depending on the source
configuration and the release height, such concentrations will occur very close to the
source (i.e. 1 to 2 km).
In determining what terrain height to enter into the model for each downwind
distance, the user should be sensitive to the source plume height. (A quick determination
of plume height can be made by executing the SCREEN2 or TSCREEN models with a hypothetical
receptor above stack top. The final stable plume height is then provided in the complex
terrain portion of the model output for SCREEN2. or in a data entry screen for TSCREEN.) At
each downwind distance, the terrain height closest to the plume height should be entered to
estimate worst-case impacts.
Concentration averaging times should be a factor in establishing the receptor grid.
based on APA objectives. For short-term averaging times (up to 24 hours), the selection of
receptors should be based on the objec"~ve of protecting public health and the environment
at all publicly accessible areas around the Superfund site. In this respect, the receptor
grid should include locations of anticipated maximum concentrations off site. For long-term
averaging times (e.g., monthly, seasonally, annually. 70 years), concentrations should be
predicted at actual receptor locations (i,e , in areas surrounding residences and work
places, and at locations with environmentally sensitive species).
URBAN/RURAL CLASSIFICATION
For the purpose of dispersion modeling, sites are classified as being in a
predominantly "urban" or "rural"'area. This determination is typically based on the land
use in the area surrounding the site to be modeled. The Guideline on Air Quality Models
(Revised)1 and Auer30 provide guidance on appropriate land use classification procedures. In
general, the determination of whether the area should be classified as urban or
5-1'
-------
rural begins by estimating the percentages of urban and rural land use types that occur
within 3 km of the site Table 5-2 lists common land use types and their urban or rural
designation. Zoning maps. U S Geological Survey (USGS) topographic maps (1:24.000 scale).
or aerial photographs of the area surrounding the site typically provide the basis for
distinguishing what land use types exist As stated in the Guideline on Air Quality Models
(Revised)', if land use types II. 12. Cl, R2. and R3 account for 50 percent or more of the
total area (within 3 km of the source), then the site is classified as urban for modeling
purposes, otherwise, it is classified as rural Table 5-2 provides classification of land
use types
Delineation of urban and rural land use types can be difficult for the residential -
type areas listed in Table 5-2 The degree of resolution for residential areas can often
not be identified without conducting an inspection of the site area. This process can be
greatly streamlined for many applications, without lessening confidence in the selection of
the appropriate classification. The fundamental simplifying assumption is that many
applications will have a definite urban or rural designation based on review of the relevant
USGS topographic maps, zoning maps, or aerial photographs.
Sources located in an area classified as urban should be modeled using urban
dispersion coefficients, while sources located in an area classified as rural should be
modeled using rural dispersion coefficients The general effect of an urban area is to
create enough additional turbulence, due to the buildings and urban "heat island," to
enhance plume dispersion. Some models, such as SCREEN2, TSCREEN. and the Industrial Source
Complex Models (ISC2). incorporate both urban and rural dispersion coefficients (the model
user simply specifies which applies). Other models, particularly those addressing complex
terrain, generally accommodate one land use classification or the other.
5-15
-------
PLUME DOWNWASH
Air quality modeling of point sources with stack heights that are less than good
engineering practice (GEP) stack height should consider the impacts associated with building
wake effects. Building wake effects are not considered for area or volume sources.
Incorporating building downwash for stacks with heights less than GEP will increase model -
predicted concentrations As defined by Title 40 of the Code of Federal Regulations (CFR)
Section 51.100, GEP height is calculated as:
GEP - H * \.SL
0
where Hb is the building height and I is the lesser of the building height or maximum
projected width. This formula defines the stack height at which building wake effects on
the stack gas exhaust may be considered insignificant.
A building or structure is considered sufficiently close to a stack to cause wake
effects when the minimum distance between the stack and the building is less than or equal
to five times the lesser of the height or projected width of the building (5L). This
distance is commonly referred to as the building's "region of influence." If the source.
for example an air stripper, is located near more than one building, each building/stack
configuration must be assessed separately.
Note that building projected width is required. This means that the apparent width
of building must be determined. The apparent width is the width as seen from the source
looking towards either the wind direction or the direction of interest. For example, ISC
requires the apparent building widths (and heights) for up to every 10 degrees of azimuth
around each source.
5-16
-------
TABLE 5-2 CLASSIFICATION OF LAND USE TYPES.
Type
11
12
Cl
Rl
R2
R3
R4
Al
A2
A3
A4
A:
Description
Heavy Industrial
Light/Moderate Industrial
Commercial
Common Residential
(Normal Easements)
Compact Residential
(Single Fami ly)
Compact Residential
(Multi -Family)
Estate Residential
(Multi-Acre Plots)
Metropolitan Natural
Agricultural
Undeveloped
(Grasses/Weeds)
Undeveloped
(Heavi ly Wooded)
Water Surfaces
Urban or Rural Designation
Urban
Urban
Urban
Rural
Urban
Urban
Rural
Rural
Rural
Rural
Rural
Rural
References:
EPA. Guideline on Air Quality Models (Revised). EPA-450/2-78-027. Office of Air
Quality Planning and Standards. Research Triangle Park. North Carolina, July 1986
Auer. August H. Jr., "Correlation of Use and Cover with Meteorological Anomalies.
Journal of Applied Meteorology, pp. 636-643, 1978.
5-17
-------
To account for downwash, the SCREEN2 and TSCREEN models require input of a building
(structure) height and the respective maximum and minimum horizontal dimensions. Generally,
to evaluate the greatest downwash effects for each source, the building with dimensions that
result in the highest GEP stack height for that source should be modeled.
The ISC2 models also contain algorithms for determining the impact of plume downwash
on ambient concentration, and should be used for determining refined concentration
estimates. Methods and procedures for determining the appropriate inputs to account for
downwash are discussed in the Guidelines for Determination of Good Engineering Practice
Stack Height (Technical Support Document for the Stack Height Regulations) (Revised).31 Due
to the complexity of GEP guidance, the EPA has developed a computer program for calculating
downwash related inputs for the ISC models. This program, called Building Profile Input
Program (BPIP) is available from the SCRAM BBS. BPIP helps clarify GEP and building
downwash guidance and should be applied in consultation with the EPA Regional Modeler.
AVERAGING TIME CONSIDERATIONS
Superfund sites often contain a complex mixture of contaminants. The potential
adverse health effects vary from compound to compound, and the health-based action levels
may vary by orders of magnitude between compounds with relatively similar structures and
physical properties. Therefore, the most significant compounds at the site, from a health
risk standpoint, may not necessarily be those compounds present in the highest
concentrations in the soil or water.
Action levels are based on health or environmental risk values. The averaging-time
periods that the action levels should address depend on a number of factors. Action levels
associated with the work areas are designed to protect the health of on-site workers.
whereas action levels associated with the exclusion area or fence line and beyond are
designed to protect the surrounding populace and environment.
5-18
-------
Several averaging periods may be of interest for any given analysis, including
instantaneous. 15-minute. 1-hour. 24-hour, monthly, and annual. The averaging periods to
evaluate will depend on the time periods of the applicable action levels. The choice of
time periods will also depend on the specific compounds present and their associated health
effects. The compounds addressed by the action levels will typically be a subset of the
contaminants present at the site. Risk assessments for the air pathway usually indicate
that relatively few compounds account for the great majority of the risk. The compounds
requiring action levels are those compounds, present in significant quantities, that have
high toxicity or a high degree of hazard, and that are capable of being released to the
atmosphere.
Several categories of action levels may be necessary, depending on the compounds of
interest, the operating life of the source, the type of emission sources, and the
potentially exposed population. Categories of action levels used most often are long-term
(annual) action levels for carcinogens and non-carcinogens, and short-term action levels for
acute toxins The EPA Regional Toxicologist is the best source of technical guidance on
determining the appropriate averaging periods for the contaminants at a particular site
Many hazardous air pollutant release models are designed to provide concentration
predictions for unit averaging times ranging from 1 second (instantaneous) to 1 hour. This
is because the concern for hazardous air pollutants may be explosions (where an
instantaneous concentration may be sufficient for ignition) or short term exposures which
can lead to acute effects. By contrast, the regulatory models typically used for air
quality analyses have a basic averaging time of 1 hour for concentration estimates.
To derive impacts for other averaging per-iods such as 3-hour, 8-hour, 24-hour and
annual, screening-level models such as TSCREEN and SCREEN2, use time scaling factors. These
time scaling factors account for the variability in meteorological conditions that may occur
over the longer time period. Concentrations for various averaging periods can be
automatically calculated with refined models, given their use of site-specific
meteorological data.
5-19
-------
WORST-CASE IMPACT DETERMINATION
Establishing the characteristics of a release that maximize the predicted
concentration is what is commonly referred to as determining the "worst-case" impact. It is
important to characterize the worst case impact in order to establish the upper bound of
potential exposure Further, since accidental releases, such as the puncture of a buried
drum or gas cylinder, can occur at any time, it is important to have an understanding of
what the worst impact of these releases could be in order to be as prepared as possible in
the event such a release occurs.
It is also important to understand that what constitutes worst-case conditions is
really a function of the impact of concern Conditions producing the maximum concentration
may not necessarily be the same conditions causing the most people to be exposed. This is
because maximum impacts will result from minimal spatial dispersion of the released
pollutant.
To determine a worst-case impact, one should consider situations that provide the
most effective emissions release rate and the meteorological conditions that produce the
worst dispersion. For most releases, the most effective release rate is equal to the
maximum release rate. The maximum release rate should be defined for each source and for
each averaging period modeled. Often the maximum, annualized emission rate for a source
will be less than the maximum short-term rate, reflecting the fact that emissions are not
occurring continuously over the long-term. The more continuous, as opposed to intermittent.
the emissions are. or can be assumed to be. the higher the rate that should be modeled to
predict worst-case impacts. For fugitive sources, emissions are a function of the spatial
extent of the source. Characterizing worst-case impacts for these sources would involve
modeling the maximum area that may be exposed at any given time, especially for determining
all short-term impacts.
/
Meteorological conditions that produce the worst dispersion for ground-level releases
are those associated with very stable atmospheric conditions and low wind speeds (on the
order of 2 meters per second [m/s]); these conditions normally give poor dispersion. For
elevated, buoyant
5-20
-------
eleases. an unstable atmosphere may result in the maximum concentration predictions, because
an unstable atmosphere can mix the plume to the ground at higher concentrations than would a
stable atmosphere. For other releases, multiple stability classes and wind speeds need to
be modeled to determine the meteorological conditions producing the worst dispersion
Specifying the worst-case conditions is not a simple matter because the same variable
can have conflicting influences For example, high wind speeds can lead to high fugitive
dust emission rates and high evaporative rates from surface lagoons. However, high wind
speeds are also associated with enhanced dispersion (for models that take wind speed into
account in determining dispersion, but are executed with a "fixed" emission rate,
concentration impacts can be lower during high winds, and higher during low winds). The
relationship between ambient temperature and worst-case predicted impact is more
straightforward, as higher ambient temperatures tend to result in higher emission rates of
volatiles and an increase in the difference between ambient and release temperatures
(leading to a tendency for the release to behave as a denser-than-air release). A denser-
than-air release can lead to higher ground-level concentrations, especially at near field
receptors. If a higher ground level temperature is given but the model uses it to increase
the vertical temperature gradient, the turbulence and therefore dispersion will increase.
These are only a few examples of counteracting effects for single parameters which must be
taken into consideration when determining worst-case conditions.
For models executed with a complete year of meteorological data, establishing the
worst-case impact is usually a matter of evaluating and executing different source
characterizations, because the presumption is that the meteorological data set will contain
conditions producing the worst dispersion. To ensure that meteorological conditions
producing the worst dispersion are adequately represented, as many years of representative
meteorological data as are available should be modeled (generally a five-year period should
be adequate).
5-21
-------
BACKGROUND CONCENTRATIONS
It is not uncommon for Superfund sites to be located in industrial areas. In such
cases. It is important to assess the cumulative impact of toxic air pollutants that the site
and existing industry have in common. Establishing the "background" pollutant
concentrations due to other existing sources allows for the incremental impact of the
Superfund site activities to be determined. Background concentrations may also be important
when considering ARARs such as PSD or NSR.
Implementing an air momtonng program in the vicinity of a Superfund site could
provide the necessary information on existing background air quality levels if:
1) the air monitoring network was designed and implemented following procedures
similar to the guidelines provided in Volume IV of this series6; and
2) the network monitored the pollutants of interest at the Superfund site.
Background air quality data could be obtained from previous air monitoring programs
conducted in the site vicinity. In areas where there are large sources of toxic air
pollutants close to the Superfund site, a estimation of background concentrations can be
obtained through dispersion modeling of these sources. The Guideline on Air Quality Models
(Revised)1 provides guidance on determining background concentrations.
5-22
-------
SECTION 6
ASSESSMENT OF MODEL RESULTS
Modeling results need to be summarized and evaluated to provide input to the
site-specific Ai<~ Pathway Assessment (APA) and the Superfund decision-making process. The
output of the dispersion modeling should be summarized together with the pertinent source
and meteorological data to serve as a basis for evaluating the results. In addition,
interpretation of dispersion modeling results should account for other factors such as
complex terrain, multiple sources, and noncontinuous releases. This process is invaluable.
because examining the output will help reveal whether the model was executed properly, and
whether the results make sense compared to the inputs used and the model's simulation of
reality. The ideas presented in this section will help the modeler to understand the model
results, and to view the results from a more learned perspective. Risk assessment
philosophy or considerations are not addressed in this manual; the reader should refer to
other Superfund documents that deal specifically with that subject.
SUMMARIZING MODEL INPUT AND OUTPUT
To understand the model results, it is useful to make some initial summaries of the
output. The output from dispersion models is given in tabular form. These data must be
summarized in a format that is useful both for data evaluation and for presenting the
conclusions of the modeling for the specific APA application. Examples of recommended
tabular data summaries include:
• Maximum short-term and long-term average concentrations predicted off site;
• Maximum concentrations at any sensitive receptor locations;
• Source-specific contributions to the maximum predicted values (for sites with
multiple air release sources);
• Maximum deposition, if modeled, to soil and water bodies;
• Summaries of any predicted model versus measured values; and
Chemical-specific applicable or relevant and appropriate requirements (ARARs)
and health effects data.
6-1
-------
An example of a simple summary table is shown In Table 6-1. This same format could be
applied to show Individual source maximum impacts and their corresponding locations. A
column could also be added to show the percent contributions of the individual sources to
the maximum predicted concentrations for the total site.
TABLE 6-1.
EXAMPLE SUMMARY TABLE OF MAXIMUM PREDICTED IMPACTS
Pollutant
Chloroform
1.1.1-
Tnchloroethane
Trichloroethylene
Averaging
Period
1-hour
Annual
1-hour
Annual
1-hour
Annual
Maximum
Impact
(/ig/m3)
1.050
0.040
11,600
25
2,100
0.420
Maximum Impact
Location
UTN-E
(m)
696000
696200
696100
696300
696100
696300
UTH-N
(m)
5068000
5068000
5068100
5068100
5068100
5068100
Action
Level s
(jig/m3)
98
0.043
19.000
1.000
2.690
0.59
Note: All values shown are for illustration only
It is also extremely useful to present the results graphically. This is easily
accomplished by plotting concentration contours (i.e.. isopleths) for the various pollutant
averaging periods modeled. Since each contour represents a user-specified concentration (or
deposition flux), each contour demarcates the spatial extent of that impact. Plots can be
generated using specialized mapping software or an integrated modeling/plotting software
package. In particular, contour plots of annual average concentration should show a
correlation with the annual wind rose of the meteorological data modeled (it would be easy
to match for example, the respective contour plots with the wind roses shown in Section 5).
6-2
-------
Summaries of the meteorological data should also be made. The following present
useful information:
Annual wind rose;
Daytime wind rose (atmospheric stability classes A-D).
Nighttime wind rose (atmospheric stability classes E and F):
• Tabular summaries of means, mimmums, and maximums of the variables modeled;
and
Summaries of percent data capture for each variable.
Statistical summaries for the meteorological data should be presented on a monthly.
seasonal, and annual basis, as well as for the entire period modeled. For sites with
diurnal wind patterns (e.g., mountain valley or coastal areas), the modeling should include
separate wind roses for daytime and nighttime conditions.
Summaries of the source input, together with those of the meteorological data, are
useful in interpreting the magnitude and locations of maximum impact. The following
illustrations and tables are suggested:
Map showing source locations and locations of maximum exposed individuals
(MEIs):
Source roster, categorized by pollutant; and
• Source physical characteristics modeled.
TOPICS FOR CONSIDERATION
In situations where multiple sources are being modeled, it is important to consider
the source-specific contributions to the predicted concentrations. For example, remediation
sources may involve soil-handling activities and an air stripper, with maximum impacts
dominated by the soil-handling operations. This information is important to determining
which emission controls will be most effective in reducing maximum concentrations to
acceptable levels.
6-3
-------
In the event predicted concentrations exceed acceptable levels, it is important to
first ensure that the model inputs were correctly determined, and that the model options
selected were appropriate. Further, any previously made assumptions should be reconsidered,
particularly with respect to source emission rates or release parameters, as well as the
determination of model selection. The merit of having prepared, and received approval for a
modeling plan will manifest itself at this point, as it will narrow the range of potential
modeling aspects to investigate.
Predicted concentrations may also be compared with ambient air monitoring data to
assess the accuracy of the model predictions. This would involve executing the model with
the actual meteorological conditions occurring during the air quality sampling. The short-
term monitored concentrations could then be compared with the short-term model predictions
for the same time period. Statistical measures to use in comparing the monitored and
modeled data sets can be established through consultation with the Regional Modeler.
The results of the modeling analysis should be tailored to the needs and requirements
of the APA. The level of required detail in the model output will certainly be a function
of the pollutants involved, the magnitude of emissions, and other site-specific concerns.
Understanding the model uncertainties will provide a good basis for developing the
appropriate model output.
MODEL UNCERTAINTY
The accuracy of model estimates varies with the model used, the type of application.
and site-specific characteristics. According to the Guideline on Air Quality Models
(Revised)1, studies of model accuracy have confirmed that (1) models are more reliable for
estimating longer time-averaged concentrations than for estimating short-term concentrations
at specific locations: and (2) the models are reasonably reliable in estimating the
magnitude of highest concentrations occurring sometime, somewhere within an area. For
example, errors in highest estimated concentrations of ± 10 to 40 percent are typical (i.e..
well within the factor-of-two accuracy that has long been reported for these models).
However, estimates of concentrations that occur at a specific time and site are poorly
correlated with actually observed concentrations, and are much less reliable.
6-4
-------
Poor correlations between paired concentrations at fixed stations may be due to
"reducible" uncertainties in knowledge of the precise plume location, and to unquantified,
inherent uncertainties. For example, apart from data input errors, maximum ground-level
concentrations at a given hour for a point source in flat terrain could be in error by 50
percent due to these uncertainties. Uncertainty of five to ten degrees in the measured wind
direction, which transports the plume, can result in concentration errors of twenty to
seventy percent for a particular time and location, depending on stability and station
location Such uncertainties do not indicate that an estimated concentration does not
occur, only that the precise time and locations are in doubt.
In light of model uncertainties, the recommended approach is to consider maximum
concentrations predicted off site as controlling concentrations irrespective of whether the
maximum receptor coincides with an inhabited location This is particularly important for
short-term (i.e., 24-hour or less) concentrations. For long-term concentrations, the
location of maximum impact in relation to residences or inhabited areas can be considered on
a case-by-case basis as a factor in evaluating model results. For predicting impacts in
specific areas of interest (e.g., waterways, residential communities) use of multiple
receptors to characterize impact is recommended.
Because the technical information on measures of model uncertainty most relevant in
decision making is incomplete, no specific guidance on the consideration of model
uncertainty in decision making is presently issued. As procedures for considering
uncertainty develop and are implemented, this guidance will be revised. In the meantime, it
is acceptable to consider model results as a "best estimate."
6-5
-------
SECTION 7
REFERENCES
1 U.S Environmental Protection Agency, 1986. Guideline on Air Quality Models.
(Revised) EPA-450/2-78-027R, Office of Air Quality Planning and Standards, Research
Triangle Park, North Carolina
2. U.S Environmental Protection Agency, 1989 Procedures for Conducting Air Pathway
Analyses for Superfund Activities. Interim Final Document: Volume IV - Procedures for
Dispersion Modeling and Air Mom tori no for Superfund Air Pathway Analyses. EPA-
450/1-89-004 (NTIS PB90-113382/AS)
3. Air/Superfund National Technical Guidance Study Series, 1992. Procedures for
Conducting Air Pathway Analyses for Superfund Activities. Interim Final Document:
Volume I - Overview of Air Pathway Assessments for Suoerfund Sites (Revised). EPA-
450/1-89-OOla.
4 Air/Superfund National Technical Guidance Study Series, 1990. Procedures for
Conducting Air Pathway Analyses for Superfund Activities. Interim Final Document:
Volume II - Estimation of Baseline Air Emissions at Superfund Sites (Revised). EPA-
450/l-89-002a (NTIS PB90-270588).
5. Air/Superfund National Technical Guidance Study Series, 1989. Procedures for
Conducting Air Pathway Analyses for Superfund Activities. Interim Final Document:
Volume III - Estimation of Air Emissions from Cleanup Activities at Superfund Sites.
EPA-450/1-89-003 (NTIS PB89-180061/AS).
6. Air/Superfund National Technical Guidance Study Series. 1993. Procedures for
Conducting Air Pathway Analyses for Superfund Activities. Interim Final Document:
Volume IV - Guidance for Ambient Air Horn toring at Suoerfund Sites. EPA-451/R-93-007
(NTIS PB93-199214).
7-1
-------
7 Air/Superfund National Technical Guidance Study Series. 1993. Models for Estimating
Air Emission Rates from Suoerfund Remedial Actions. EPA-451/R-93-001 Office of Air
Quality Planning and Standards, Research Triangle Park. North Carolina.
8. U.S. Environmental Protection Agency, 1989. Review and Evaluation of Area Source
Dispersion Algorithms for Emission Sources at Superfund Sites. EPA-450/4-89-020.
Office of Air Quality Planning and Standards. Research Triangle Park, North Carolina
9. U.S. Environmental Protection Agency, 1992 User's Guide for the Industrial Source
Complex (ISC2) Dispersion Models. Volume I - User Instructions EPA-450/4-92-008a
Office of Air Quality Planning and Standards, Research Triangle Park. North Carolina.
10. U.S Environmental Protection Agency, 1994. User's Guide to TSCREEN. A Model for
Screening Toxic Air Pollutant Concentrations (Revised). EPA-454/3-94-023. Office of
Air Quality Planning and Standards, Research Triangle Park, North Carolina.
11. U.S. Environmental Protection Agency, 1993. Contingency Analysis Modeling for
Suoerfund Sites and Other Sources EPA-454/R-93-001 Office of Air Quality Planning
and Standards, Office of Air and Radiation, Research Triangle Park, North Carolina.
12. U.S. Environmental Protection Agency, 1993. Guidance on the Application of Refined
Dispersion Models to Hazardous/Toxic Air Pollutant Releases. EPA-454/R-93-002.
Office of Air Quality Planning and Standards. Research Triangle Park. North Carolir-
13. U.S. Environmental Protection Agency, 1992. SCREEN2 Model User's Guide. EPA-450/4-
92-006. Office of Air Quality Planning and Standards. Research Triangle Park, North
Carolina.
14. U.S. Environmental Protection Agency. 1992. Screening Procedures for Estimating the
Air Quality Impact of Stationary Sources (Revised). EPA-454/R-92-019. Office of Air
Quality Planning and Standards, Research Triangle Park, North Carolina.
7-2
-------
15 U.S Environmental Protection Agency. 1992 Workbook of Screening Techniques for
Assessing Impacts of Toxic Air Pollutants (Revised). EPA-454/R-92-024. Office of
Air Quality Planning and Standards, Office of Air and Radiation. Research Triangle
Park. North Carolina.
16. U.S Environmental Protection Agency. 1993. A Tiered Modeling Approach For Assessing
the Risks Due to Sources of Hazardous Air Pollutants. EPA-450/4-92-001 Office of
Air Quality Planning and Standards, Research Triangle Park, North Carolina
17 U S. Environmental Protection Agency, 1992. Guideline for Predictive Baseline
Emissions Estimation Procedures for Superfund Sites EPA-450/1-92-002.
18. F. Pasquill and F.B. Smith. Atmospheric Diffusion. Third Edition John Wiley and
Sons, New York, NY., 1983.
19 U.S Environmental Protection Agency. 1978 User's Guide for.RAM. EPA-600/8-78-016,
Volumes A, B. and C. Research Triangle Park, North Carolina.
20. U.S. Environmental Protection Agency. 1993. Air/Superfund National Technical
Guidance Study Series. Estimation of Air Impacts from Area Sources of Participate
Matter Emissions at Superfund Sites. EPA-451/R-93-004. Office of Air Quality
Planning and Standards, Research Triangle Park, North Carolina.
21. U.S. Environmental Protection Agency. Compilation of Air Pollutant Emission Factors.
Volume 1. Stationary Point and Area Sources. Fourth Edition. EPA AP-42. Chapter
11 2 Fugitive Sources. Research Triangle Park. NC
22. Winges. K., et al. User's Guide for the Fugitive Dust Model (FDM) (Revised). EPA-
910/9-88-202R, 1992.
23. U.S. Environmental Protection Agency, 1992. Estimation of Air Impacts for the
Excavation of Contaminated Soil. EPA-450/1-92-004.
7-3
-------
24. U.S. Department of Energy, 1984 Atmospheric Science and Power Production DOE/TIC-
27601. (DE84005177).
25. US. Environmental Protection Agency, 1993. On-Site Meteorological Program Guidance
for Regulatory Mode1ing Applications. EPA-450/4-87-013. Office of Air Quality
Planning and Standards. Research Triangle Park. North Carolina
26 US. Environmental Protection Agency. 1988. Meteorological Processor for Regulator
Models (HPRH 1 2) User's Guide. EPA-600/3-88/043R. Research Triangle Park. NC (NTIS
No PB 89-127526)
27 U.S.Envu nmental Protection Agency. 1972. Mixing Heights. Wind Speeds, and
Potential for Urban Air Pollution Throughout the Contiguous United States Office of
Air Programs. Research Triangle Park, North Carolina.
28. U.S. Environmental Protection Agency, 1990. User's Guide to CTDMPLUS: Volume 2 The
Screening Mode (CTSCREEN). EPA-600/8-90-087. Research Triangle Park. NC. (NTIS No.
PB 91-136564).
29. U.S. Environmental Protection Agency, 1989. User's Guide to the Complex Terrain
Dispersion Model Plus Algorithms for Unstable Situations (CTDHPLUS) Volume 1: Model
Description and User Instructions. EPA-600/8-89-041. Research Triangle Park, NC.
(NTIS No. PB 89-181424).
30. Auer. A H., Jr. 1978. "Correlation of Land Use and Cover with Meteorological
Anomalies." Journal of Applied Meteorology 17:636-643.
31. U.S. Environmental Protection Agency. 1985 Guidelines for Determination of Good
Engineering Practice Stack Height (Technical Support Documentation for the Stack
Height Regulations) (Revised). EPA-450/4-80-023R.
32. U.S. Environmental Protection Agency. 1989. Quality Assurance for Air Pollution
Measurement Systems. Volume IV - Meteorological Measurements. EPA 600/4-82-060R.
Office of Research and Development. Research Triangle Park, NC.
7-4
-------
33. US. Environmental Protection Agency. 1987. Ambient Air Monitoring Guidelines for
Prevention of Significant Deterioration (PSD). EPA 450/4-87-007. Office of Air
Quality Planning and Standards. Research Triangle Park. NC. (NTIS No. PB 90-168030).
34 U S Environmental Protection Agency, 1989. Risk Assessment Guidance for Superfund
Volume I. Human Health Evaluation Manual (Part A) EPA 540/1-89/002 Office of
Emergency and Remdial Response. Washington DC.
35. LJ.S Environmental Protection Agency. 1991. Risk Assessment Guidance for Suoerfund
Volume I. Human HeaHh Evaluation Manual (Part B. Development of Risk-based
Preliminary Goals) - Interim EPA 9285 7-01B Office of Emergency and Remdial
Response. Washington DC
36. U.S. Environmental Protection Agency. 1991. Risk Assessment Guidance for Suoerfund
Volume I. Human Health Evaluation Manual (Part C. Risk Evaluation of Remedial
Alternatives) - Interim. EPA 9285 7-01C Office of Emergency and Remedial Response.
Washington DC
37. US Environmental Protection Agency, 1990. National Oil and Hazardous Substances
Pollution Contingency Plan (NCP). 40 CFR Part 300, Volume 55. No. 46.
38. 40 CFR Part 302, Table 302 4 or U.S. Environmental Protection Agency. 1991. Title
III List Of Lists. EPA 560/4-91-011
39. U.S. Environmental Protection Agency, 1992. Guideline for Predictive Baseline
Emissions Estimation Procedures for Superfund Sites. EPA-450/1-92-002. Office of
Air Quality Planning and Standards. Research Triangle Park. NC.
7-5
-------
APPENDIX A
CASE EXAMPLE
-------
APPENDIX A
CASE EXAMPLE
This case example illustrates how the concepts presented in this manual can be
implemented to fulfill an air dispersion modeling analysis at a Superfund site. For easy
reference, it is organized to parallel the flow of the manual. This example is meant to
emphasize, in a systematic manner, modeling methodology and input development. Any final
determination on the acceptability of predicted impacts is, therefore, beyond the scope of
this discussion. Finally, because the case described is hypothetical, the example is
naturally not applicable to all situations. This case example draws from the case example
published in the Air/Super fund NTGS document Screening Procedures for Estimating the Air
Impacts of Incineration at Superfund Sites. EPA-450/1-92-003, NTIS PB92-171917.
SITE DESCRIPTION
The hypothetical Superfund site in this example is a 15-acre abandoned waste site
once used by local industries The site is located on the outskirts of an urban area, and
is within 3 kilometers (km) to the west of a small recreational lake that is a popular local
fishing spot. Terrain within the site is relatively flat, and becomes more gently rolling
in the surrounding area. Low foothills of a distant mountain range are a significant
feature of the landscape within 10 km to the north of the site.
General site activities planned during the pre-remediation phase of the project
included collection of historical operations data that might shed light on the nature and
spatial distribution of the contamination, followed by soil sampling during the site
inspection to establish what compounds were actually present. To support dispersion
modeling, a meteorological monitoring program was also initiated.
A-l
-------
Air dispersion modeling conducted during the pre-remediation phase was used to
support the siting of an ambient air monitoring (MM) stations. AAM data was collected
prior to the start of remediation activities to provide baseline air quality information for
the site.
During the remediation phase, further modeling was conducted to support the
development of the Record of Decision (ROD) and remedial design (RD). AAM monitoring was
continued through the remedial action (RA) to provide direct exposure assessment, and to
provide data for comparison with the dispersion modeling results.
NATURE OF CONTAMINATION
In the past, the site was used to bury, dump, and store industrial wastes such as
paint sludges, solvents, oils and greases, phenols, and heavy metals. The area requiring
remediation consisted of a dry surface impoundment containing approximately 60.000 tons of
soil contaminated primarily with polychlorlnated biphenyls (PCBs) and lead. Sampling data
collected during the site inspection indicated the presence of the compounds listed in Table
A-l. The concentrations are in ppm by weight. The value of ppm-wt is the same as A/g/g.
The organic compounds were split into three groups for modeling. These groups were PCBs.
Dioxins (TCDD and TCDF). and Total Hydrocarbons (THC) (all other organics). Potential
exposure to each of these groups during the RA was addressed in the modeling analysis. In
addition, the impact of particulate matter emissions was also determined, since particulate
matter with a diameter less than 10 micrometers (PM10) is a criteria pollutant.
A-2
-------
TABLE A-l CONTAMINANTS PRESENT IN SOIL SAMPLE.
Organics
COMPOUND
Acetone
Benzene
Bis(2-ethylhexyl)phthalate
Carbon Oisulfide
Methyl Ethyl Ketone
Methylene Chloride
Polychlonnated Biphenyls (PCBs)
Phenol
Tetrachlorodibenzo-p-dioxin (TCDD)
Tetrachlorodibenzofuran (TCDF)
Tetrachloroethane
Toluene
Tnchloroethene
Total Xylenes
CONCENTRATION (ppm-wt)
37
5
•5
O
47
160
97?
i- ! L.
28
0.06
Or\ n r
(HIS
. \j \j ^j
50
O V
SO
\JVJ
Inorganics
COMPOUND
Arsenic
Barium
Cadmium
Chromium
Lead
Zinc
CONCENTRATION (ppm-wt)
2
591
20
85
778
301
SOURCE DEFINITION
Given the amount of metals that were present in the soil, a dual remedial action was
planned. Because most of the lead was concentrated in one portion of the impoundment, it
was removed and sent to a nearby hazardous waste landfill. (Impacts due to lead
contamination throughout the waste site were addressed in the modeling analysis.)
A-3
-------
The proposed RA was to incinerate the contaminated soil on site. Therefore, as part
of the risk determination conducted during the RO. air dispersion modeling was performed to
assess the potential impact of this remedial activity. A rotary-kiln incinerator was
proposed, with a maximum waste (soil) feed rate of 10 tons per hour (tons/hr). The
conversion factor for tons/hr to kg/hr is 907.2. No free liquids were incinerated. The air
pollution control system consisted of cyclones to remove large particulate matter, a packed-
tower scrubber for primary removal of acid gases, and an ejector scrubber for removal of
fine particulate matter and additional acid gases before release of the gas from the stack
The point source parameters required for modeling the incinerator are shown in Table A-2.
The incinerator was located approximately 30 meters (m) from the nearest site property
boundary.
TABLE A-2. INCINERATOR STACK PARAMETERS
Stack Height
(meters)
20
Stack Inner Diameter
(meters)
1.0
Exit Gas
Velocity
(meters/second)
20
Exit Gas Temperature
(Kelvin)
344
During remediation, the incinerator was operated 24 hours per day, 7 days per week to
avoid start-ups and shut-downs. At this rate, it took 250 days to incinerate the 60,000
tons of contaminated soil. The incinerator was fueled with propane, which did not
contribute significantly to the emissions of compounds present in the waste soil.
Aside from the incinerator stack, air emissions also occurred from soil handling
operations upstream of the incinerator. These emissions included fugitive emissions from
the soil excavation, transportation to the incinerator, and the temporary storage piles
created near the incinerator. Soil excavation from the surface impoundment proceeded such
that the total surface area exposed for the pit at any given time was no more than
1,000 square meters (m2). A paved roadway was constructed and used for transporting the
contaminated soil to the incinerator. The dimensions associated with the fugitive sources
are shown in Table A-3. A general diagram of the site is shown in Figure A-l.
A-4
-------
TABLE A-3 FUGITIVE SOURCE DIMENSIONS
Source
Excavation Pit
Single Storage Pile
Transport
Dimensions
(meters)
25 x 40 x 3
5 x 10 x 3
250a
Area .
(square meters)
1.000
140
--
Release Height
(meters)
0
1.5
1.5
Maximum, single round-trip distance.
For dispersion modeling, these fugitive sources were represented as square area and
volume sources (rectangular areas were not allowed by the model selected). As illustrated
in Figure A-2, the excavation pit was modeled as five square area sources, each storage pile
was modeled as two square areas, and the roadway was modeled as a series of square volume
sources spaced at equal intervals So defined, the spatial extent of each fugitive source
was matched exactly with the source dimensions Depending on the distance to the nearest
receptor, an alternative technique might have been to assign a single square area source to
both the excavation pit and to the storage piles, each such area source having a square area
equivalent to the total area of the actual source
The actual location of the excavation pit varied as the RA progressed. Therefore.
for dispersion modeling, this source was located within the impoundment such that predicted
impacts would be conservative. Specifically, this source was located a minimal distance to
the property boundary, and located near the storage pile area, allowing the location of the
excavation pit impacts off site to more nearly coincide with those of the storage piles.
A-5
-------
Foothill*
10km
t
Surface Impoundment t
Storage Pile Area
Incinerator Stack
Property Boundary
3 km
Gate
Like
10km
N
NWS Station
Figure A-l Example Site Plan.
A-6
-------
A.
•
•••MM
•
B.
C.
Figure A-2 Area and Volume Source Representations of the Fugitive Sources. (A) Area source
breakout of the excavation pit. (B) Volume source delineation of the roadway. (C) Area source
breakout of a single storage pile. Relative dimensions are not drawn to scale.
A-7
-------
EMISSION ESTIMATION
In order to assess applicable or relevant and appropriate requirements (ARARs) and
health effects, it was necessary to model both long-term and short-term impacts. Therefore.
both long-term and maximum short-term emission rates were determined for each source.
Individual pollutant emission rates per source were defined once the averaging periods of
concern had been determined for each pollutant, based on consultation with the U.S.
Environmental Protection Agency (EPA) Regional Toxicologist In estimating emission rates
from the incinerator, the planned air pollution control equipment was taken into account.
Emission rates were determined for the incinerator, soil excavation and handling.
transport, and storage piles according to procedures outlined in the following documents:
• U.S. Environmental Protection Agency, 1989. Alr/Superfund National Technical
Guidance Study Series. Volume III - Estimation of Air Emissions from Cleanup
Activities at Superfund Sites. EPA Publication No. 450/1-89-003. Office of
Air Quality Planning and Standards. Research Triangle Park, NC;
U.S. Environmental Protection Agency, 1993. Air/Superfund National Technical
Guidance Study Series Models for Estimating Air Emission Rates from
Suoerfund Remedial Actions. EPA Publication No. 451/R-93-001. Office of Air
Quality Planning and Standards. Research Triangle Park, NC.
U.S. Environmental Protection Agency, 1992. Air/Superfund National Technical
Guidance Study Series. Screening Procedures for Estimating the Air Impacts of
Incineration at Superfund Sites. EPA Publication No. 450/1-92-003. Office of
Air Quality Planning and Standards. Research Triangle Park, NC; and
U.S. Environmental Protection Agency. 1992. Air/Suoerfund National Technical
Guidance Study Series. Estimation of Air Impacts for the Excavation of
Contaminated Soil. EPA Publication No. 450/1-92-004. Office of Air Quality
Planning and Standards. Research Triangle Park, NC.
A-8
-------
The TSCREEN model has the capability of doing many of the emission estimation
calculations for a Superfund site. For the incinerator emission rate use the Superfund
Release Type (Thermal incineration). For this example the total feed rate was given as 9072
kg/hr (10 tons/hr) The TSCREEN default efficiency of 99 99 percent was assumed. For the
PCB concentration of 272 /ug/g the emission rate is 6.9 x 10'5 g/s.
For the excavation site and pile TSCREEN can calculate the emission rate by using
Superfund Release Type (Soil Excavation) Using a default vapor pressure of 35 mm Hg (4666
Pa) for PCB from Estimation pf Air Impacts for the Excavation of Contaminated Soil. TSCREEN
gives an emission rate of 0 06542 g/s for PCB from both the pit and a pile. Both the pit
and pile are assumed to be made of the same material. A schematic illustration of the site
is given in Figure A-3 The following assumptions were made (these are the defaults in
TSCREEN):
• Soil excavated for 50 mm/hour:
Each scoop contains 2 m3 of soil:
• 75 scoops per hour;
Pit becomes 10m x 15m x 1m after 1 hour;
• Pile becomes 5m x 10m x 3m after 1 hour:
• Total exposed area of the pit is 150 m2;
• Total exposed area of the pile is 140 m2:
• Density of soil is 1.5 g/m3; and
• Soil and air temperature is assumed to be 25 °C.
In ISC the pile should be treated as a volume source since it extends in the vertical
(emission rate to be given in g/s). The pit should be treated as an area source (emission
rate to be given in g/s-m2). The total surface area exposed is 290 m2 (150 m2 + 140 m2).
These assumptions imply that the PCB emission rate for the pile is 0.03158 g/s ((140
m2 / 290 m2) x 0.06542 g/s). Only three piles will be assumed to be uncovered during the
day. At mght the piles will be assumed to be covered with no emission. This can be
handled explicitly in the ISC input by the use of Emission Factors (EMISFACT input in Table
A-4) Each pile is broken down into two volume sources (see Figure A-2(c)) so the emission
rate must be halved for each source. The PCB area emission rate for the pit is 2.256 x 10"*
g/s-m2 (0.06542 g/s / 290 m2).
A-9
-------
The same area emission rate (E) can be used to estimate the transport emission rate
In this example assume:
Exposed soil surface area (A) on truck is 10 m2;
Round trip travel time (T) is 60 seconds; and
• There are 6 trips per hour (N). 8 hours per day
On an hourly basis the emission would be N x E x A x T. In this example the emission for
PCB over an hour is 0.8122 g (6 x 2.256 x 10'" g/s-m2 x 10 m2 x 60 s). This means the average
emission rate is 2.256 x 10"" g/s. This emission rate is for the entire road. The road is
split into 10 volume sources so the emission rate for each road source is one tenth this
rate
MODEL METHODOLOGY AND SELECTION
Both screening-level and refined dispersion models were used for the analysis. The
screening level analysis was used to determine potential complex terrain impacts from the
incinerator. To determine the collective impact from all sources, the decision was made to
conduct a refined modeling analysis, rather than solely a screening-level analysis, because
of the nature and number of sources involved. (Given the differing dispersion
characteristics involved, the location of maximum impacts from the fugitive sources would
not be collocated with those from the incinerator. Further, as mentioned later in Section
A.13. the total number of sources involved in the analysis also reflected sources from a
nearby manufacturing facility.)
Complex Terrain Screening Analysis
Because of the point source (incinerator) and the presence of complex terrain, it was
necessary to address the potential for plume impaction on an elevated terrain feature. To
determine whether complex terrain impacts would be significant in the analysis, the
screening model TSCREEN1 was executed.
A-10
-------
A receptor gnd comprised of discrete downwind distances was used for running the
TSCREEN model. Discrete receptors were defined in order to account for unique terrain
heights at each downwind distance (see Section A.8 for definition of the terrain heights).
The receptor grid included both simple and complex terrain receptors. The model was
executed to see whether the maximum impact predicted was from the simple terrain algorithm
or the complex terrain algorithm. This analysis determined whether additional complex
terrain modeling would be needed to identify the maximum exposed individual (MED. If the
maximum complex terrain impact was found to be "significantly lower" than the maximum simple
terrain impact, then it would be determined that simple terrain impacts were controlling,
and the identification of the MEI could be adequately addressed with a simple terrain model.
(There is no set criterion on what percentage of the simple terrain impact the complex
terrain impact should be to constitute an impact that is "significantly lower." Such
determination can be made on a case-by-case basis in consultation with the EPA Regional
Modeler.)
Refined Analysis
Based on the screening analysis, it was determined that identification of the MEI
could be adequately addressed with a simple terrain model. (The maximum complex terrain
impact from the incinerator was found to be less than 15 percent of the maximum simple
terrain impact.) Therefore, the Industrial Source Complex Short-Term (ISCST2) model2 was
selected for predicting both short-term and long-term impacts for all sources at the site.
The ISCST2 model was selected based on the following requirements:
Model multiple sources;
Account for rolling terrain;
Model various averaging times;
Model time-varying emission rate(s);
Model with rural dispersion coefficients; and
Model both point and fugitive sources.
A-ll
-------
Since the remedial activity was not continuous at the site on an annual basis, the
modeling analysis was designed such that long-term averages would be based on emissions
occurring only during the time period activities were being performed. In this example, the
soil excavation and incineration occurred continuously for 250 days, during the time period
from the first of March through the first week of November. The emissions due to transport
and piles were limited during the day. Therefore, ISCST2 emission factors (EMISFACT in the
ISC input in Table A-5) were assigned to the transport and pile sources for the time of day.
Emissions from these sources only took place between 8 AM to 5PM. The only meteorological
data used was from March 1 through November 5 To extrapolate the period average
concentrations given in the model output, multiply the averages by the factor (250 days /
365 days).
Other Analysis
To determine whether any of the organic emissions could form a dense gas and,
therefore, require use of a specialized model, specific calculations were made for selected
compounds in accordance with procedures documented in Guidance on the Application of Refined
Dispersion Models to Hazardous/Toxic Air Pollutant Releases.3 The same type of calculation
can be done in TSCREEN through the use of the "Initial Form of Release Menu." Under that
menu use the option "Stacks, Vents. Conventional Point Sources." The next screen allows a
gas density check to be done. Entering "Y" for the gas density check brings up a new screen
requesting emission density. The emission density can be calculated for you by entering a
molecular weight and emission temperature. The calculation assumes that the emission is
100* emitted species. To have the calculation done correctly an emission molecular weight
(Me) is needed. The emission molecular weight can be determined from:
M
where wi is the weight fraction of component i and Mi is the molecular weight of component
i. For this case no dense-gas modeling was determined to be necessary.
A-12
-------
Particle emissions from the point and fugitive sources were also investigated to
determine the need to conduct deposition modeling that would provide input to a multipathway
risk assessment for the site. No site-specific particle size distribution data had been
collected, however, representative data were available from another Superfund site at which
the same remedial activity had been undertaken Based on this data, it was determined that
particulate matter from the excavation, transport, and storage activities would principally
be deposited within the property boundary.
Particulate emissions from the incinerator and subsequent transport off site were a
potential concern. Screening modeling indicated that maximum impacts from the incinerator
were likely occur within 1 km Sensitive receptors of interest, such as the nearby
recreational lake, were located further away (greater than 2.5 km from the source). From a
risk standpoint it was determined, through discussion with the EPA Regional lexicologist and
toxicologist for the Potentially Responsible Parties (PRPs), that exposure throughout the
inhalation pathway would be the most significant. Therefore, for this analysis, it was
decided to estimate only particulate matter concentrations.
METEOROLOGICAL DATA
An on-site meteorological monitoring program had been initiated at the site three
months prior to the analysis. The primary meteorological variables recorded on the 10-meter
tower were wind speed, wind direction, and temperature. The data were used to evaluate the
applicability of meteorological data available from the National Weather Service (NWS)
Station at the local airport, located about 10 km south of the site.
For each meteorological variable, correlation coefficients between the two data sets
were computed, as well as the variable means, ranges from minimum to maximum, and standard
deviations. These statistics were calculated for the entire data set and for smaller time
periods within the larger data set (i.e., for weeks and months). In general, if the
correlation coefficients were found to be much less than 0.8 or 0.9. then the
representativeness of the off-site data would be questionable. The evaluation resulted in
the following information:
A-13
-------
• For the same time period, the off-site meteorological data correlated
reasonably well with the on-site data (correlation coefficients were greater
than 0.9). Wind direction data for the airport showed the same pattern as the
data collected on site, with an apparent small shift of about 10 to 15
degrees. The frequency distribution of wind speed and direction by stability
class was within 10 to 20 percent.
• No significant topographic features or water bodies existed between the NWS
station and the site (The recreational lake to the east of the site is too
small to influence the meteorology at the site.)
Based on the evaluation and consultation with the EPA Regional Modeling contact, it
was decided to use the meteorological data from the NWS station. The most recent five-year
data set was obtained from the Support Center for Regulatory Air Models (SCRAM) Bulletin
Board System (BBS) for use in the analysis. This data set included both surface and mixing
height data. Although required for some models, it was not necessary to determine the
site's surface roughness for this analysis. Examples of surface roughness values are given
in Table 5-1.
TERRAIN CONSIDERATIONS
The base elevation of the site is 1,200 feet (ft) mean sea level (msl). The height
of the incinerator stack was 20 m (65.6 ft msl) above base elevation (1,266 ft msl). The
closest terrain above 1.266 ft occurs approximately 8,700 m to the north of the site. The
presence of rolling terrain and nearby terrain above stack top required the inclusion of
terrain heights (and complex terrain modeling, as discussed previously in Section A.5) in
the modeling analysis. The closest rolling terrain reached about 15 m above stack base at
400 m to the west.
A-14
-------
RECEPTOR GRID
Screening--Two types of receptor grids were established for the analysis. The first
receptor grid was developed for screening incinerator impacts using the TSCREEN model. This
grid was comprised of downwind distances from the incinerator stack, starting with the
nearest distance to the property boundary. Terrain heights were included in this receptor
array by selecting the "worst-case" terrain feature for each distance. The "worst-case"
terrain feature was defined as the terrain height closest to plume centerline height located
at each receptor distance, regardless of the terrain orientation relative to the site. The
plume centerline height was obtained from executing TSCREEN in the complex terrain mode and
noting the value from a data input screen
Refined--For the refined modeling, a Cartesian grid was developed. A Cartesian
coordinate system was selected for the refined modeling, as opposed to a polar coordinate
system, because of an interest in having uniform spacing between receptor points. The grid
was centered on the location of the incinerator stack. Receptors were placed every 100 m
out to a distance of 1 km from the property boundary. From 1 km out to a distance of 3 km.
a 500-meter receptor grid spacing was used In addition, discrete receptors were placed at
100-meter intervals along the site property boundary. To address potential impacts from the
incinerator, a coarse array of discrete receptors was also placed over the low foothills
region to the north of the site.
The initial receptor spacing beyond 1 km from the site boundary was coarse because
preliminary screening with TSCREEN indicated that maximum impacts from the incinerator would
likely fall within 1 km Further, because the other sources were ground-level fugitive
sources, their maximum impacts were anticipated to occur at the property boundary. Although
the emphasis of the analysis was on identifying the MEI, there was a concern for
specifically estimating impacts at the nearby recreational lake. Therefore, a set of
sensitive receptors was placed within the recreational lake area.
In the event the maximum impact occurred in the coarse-grid region (including the
contribution from the nearby facility mentioned in Section A.13 to follow), the maximum
impacts would be refined with a 100-meter interval grid centered on the location of maximum
impacts predicted with the coarse grid. Terrain height elevations for all receptors were
determined from 1:24,000 scale U.S. Geological Survey (USGS) topographic maps of the area.
A-15
-------
LAND USE CLASSIFICATION
The selection of the appropriate dispersion coefficients was dependent upon the land
use within 3 km of the site. According to the guidance in Guideline on Air Quality Models
.(Revised)4, the land use typing scheme of Auer was used to determine the proper land use
classification. Specifically, the total area circumscribed by a 3-kilometer radius about
the site was identified on the pertinent USGS 1:24,000 scale topographic maps. If the Auer
land use types of heavy industrial, light-to-moderate industrial, commercial, and compact
residential account for 50 percent or more of the total area. Guideline on Air Quality
Models (Revised)4 recommends use of urban dispersion coefficients; otherwise, the appropriate
rural coefficients are used.
Although on the outskirts of town, visual inspection of the topographic maps
indicated that the area surrounding the site is predominantly agricultural, recreational.
and low-density residential (i.e.. rural designations visually accounted for greater than 70
percent of the area). Therefore, rural dispersion coefficients were selected for use in the
analysis.
PLUME DOWNWASH
A good engineering practice (GEP) stack height evaluation was conducted to determine
whether inclusion of building wake effects would be required in the modeling analysis. The
procedures used in this analysis were in accordance with those described in Guidelines for
Determination of Good Engineering Practice Stack Height (Technical Support Documentation for
the Stack Height Regulations--Revisgd).5
Operation of the proposed incinerator involved the installation of a temporary
trailer located within 10 m. The dimensions of the trailer were 6.1 m x 4.5 m. with a
height of 3 m. GEP formula height is expressed as: GEP = Hb + 1.5L, where Hb is the
building height and L is the lesser of the building height or maximum projected width. A
building or structure is considered sufficiently close to a stack to cause wake effects when
A-16
-------
the mini mum distance between the stack and the building is less than or equal to five times
the lesser of the height or projected width of the building (5L). Because the trailer was
within 5L of the incinerator [5L = (5 x 3) = 15 m], it was necessary to see whether the
potential for wake effects existed. Based on the trailer dimensions, the GEP formula height
for the stack was 7.5 m [3 + (1 5 x 3) = 7.5]. Since the incinerator stack height was
greater than 7.5 m, no building wake effects due to the trailer were anticipated, and
therefore, no building dimension inputs were used in the analysis.
AVERAGING PERIOD CONSIDERATIONS
To address requirements for the health risk assessment, averaging periods of 1 hour,
24 hours, and annually were predicted In addition, to address the National Ambient Air
Quality Standard for lead, calendar quarterly averages were estimated.
WORST-CASE IMPACT DETERMINATION
To determine the highest potential, or "worst-case" impacts that could occur for each
averaging period, special attention was paid to how the source emission rates were defined.
It was considered important to characterize worst case impacts in order to estimate the
upper bound of potential human exposure to the ambient pollutant concentrations resulting
from the remedial activities at the site.
Since emissions from the incinerator were continuous, pollutant emission rates were
based on a maximum feed rate to the unit. For the fugitive sources, a worst-case, short-
term emission rate was determined by assuming an emitting area of three piles and a single
pit. This emission rate was also used for determining worst-case, long-term impacts. Worst
case dispersion conditions were assumed to be reflected in the five-year meteorological data
set modeled.
A-17
-------
BACKGROUND CONCENTRATIONS
Adjacent to the site on the eastern perimeter is a small manufacturing facility. Air
permit information for the facility indicated that it emits two pollutants addressed in the
Superfund site inventory: methyl ethyl ketone and toluene. To account for the cumulative
impacts of these pollutants, and to identify the incremental change due to the Superfund
remedial activities, the relevant sources from this facility were included in the analysis.
All necessary input data for modeling were obtained from the facility's state air permit
application.
MODEL RESULTS
Screemno--An example output from TSCREEN for the PCB emissions from the incinerator
is given in Table A-4. The results of the model runs were used to verify that complex
terrain modeling was not required. In the case in Table A-4 the maximum simple terrain
impact was 3.231 x 10"3 /ug/m3. the maximum complex terrain impact was 3.591 x 10'" /vg/m3. The
complex terrain concentration was only 11 percent of the simple terrain impact.
In this example, the results of all other pollutants from the incinerator could be
determined by scaling their emission rates to that used for the PCB simulation. The
emission rate can be calculated in TSCREEN just as done for PCB. The predicted
concentrations would then be those given for PCB multiplied by a factor. The factor would
be the pollutant emission rate divided by the PCB emission rate.
The emissions from the excavation area needs to be modeled separately from the
incinerator scenario since it is a different form of source type. Each contaminant has a
different vapor pressure and concentration and needs to be modeled as shown in A.4.
Refined--An example ISCST2 output for the PCB sources is given in Table A-5. The
output is from a run for a single year. Similar runs are needed for each of the five years.
A-18
-------
The emission rate for each species is proportional to its concentration in the contaminated
soil The concentration of each species in the soil is constant. This means that the
relative source strengths of the incinerator, pit, pile, and transport sources are the same
for all the species. This, in turn, means that impacts from the other species can be
determined by multiplying the PCB concentrations by the ratio of the concentration of the
species of interest by the concentration of the PCB (272 pptn-wt).
Over the five years of meteorological data modeled, maximum impacts for the complete
emissions inventory were found to occur at the site boundary. Since the maximum impacts
occurred in the portion of the receptor grid that was resolved to a 100-meter spacing, a
refined-grid model run was not necessary Impacts from just the Superfund site were also
reported (separately from those combined with the adjacent manufacturing facility) to
isolate the predicted impact of the RA. The modeling results indicated that the location of
the ME I was not in an inhabited area In all cases, model results were analyzed to ensure
that no errors were made in the input preparation or model execution.
CONFIRMATORY AIR MEASUREMENTS
An AAM program was performed to validate the dispersion modeling predictions and to
make a direct assessment of exposure at receptors of interest. A network of five monitoring
stations was established around the facility. The preliminary dispersion modeling output
was used to site the monitoring stations to ensure that they would be within the emission
plume. At each monitoring location, canisters were collected and analyzed off site by EPA
Method TO-14 to determine volatile organic compounds (VOCs). Also at each location, PM10
samples were collected. The PMi0 loading was determined from air flow measurements and
gravimetric analysis of the filter catch. A portion of the filter was analyzed off site for
the six metals thought to be present at the site. Sampling was conducted over two 12-hour
periods each day, during each day that the incinerator was in operation or soils/waste
handling operations were underway. In addition, two weeks of baseline AAM data were
collected prior to the start of the remediation activities.
A-19
-------
The MM data collected during remediation were compared to short-term model
predictions developed from the on-site meteorological data for that same time period. The
model predictions were found to be slightly conservative, but of the same magnitude as the
actual AAM results. This outcome was more likely and desirable than the alternative (the
model underpredicts the observed concentration). If the air quality monitor were only a few
degrees off from the exact centerline location of the plume, the maximum concentrations
reported by the monitor could be substantially underestimated.
REFERENCES
1. US. Environmental Protection Agency. 1990- User's Guide to TSCREEN. A Model for
Screening Toxic Air Pollutant Concentrations. EPA-450/ 4-90-013. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina.
2. U.S. Environmental Protection Agency. 1992: User's Guide for the Industrial Source
Complex (1SC2) Dispersion Models. Volume I - User Instructions. EPA-450/4-92-008a.
U.S. Environmental Protection Agency. Research Triangle Park, North Carolina.
3. U.S. Environmental Protection Agency, 1993. Guidance on the Application of Refined
Dispersion Models to Hazardous/Toxic Air Pollutant Releases. EPA-454/R-93-002.
Office of Air Quality Planning and Standards. Research Triangle Park, North Carolina.
4. U.S. Environmental Protection Agency, 1986. Guideline on Air Quality Models.
(Revised). EPA-4502-78-027R, Office of Air Quality Planning and Standards, Research
Triangle Park, North Carolina.
5. U.S. Environmental Protection Agency, 1985. Guidelines for Determination of Good
Engineering Practice Stack Height (Technical Support Documentation for the Stack
Height Regulations--Revised) EPA-450/4-80-023R.
A-20
-------
TABLE A-4. EXAMPLE TSCREEN OUTPUT
10/11/94
10:12.22
*** SCREEN2 MODEL RUN ***
*** VERSION DATED 92245 ***
Example PCB release from incinerator
SIMPLE TERRAIN INPUTS
SOURCE TYPE = POINT
EMISSION RATE (G/S) - 690000E-04
STACK HEIGHT (M) = 20.0000
STK INSIDE DIAM (M) = 1 0000
STK EXIT VELOCITY (M/S)= 20 0000
STK GAS EXIT TEMP (K) - 344.0000
AMBIENT AIR TEMP (K) = 293.0000
RECEPTOR HEIGHT (M) = .0000
URBAN/RURAL OPTION = RURAL
BUILDING HEIGHT (M) = .0000
MIN HORIZ BLDG DIM (M) = 0000
MAX HORIZ BLDG DIM (M) = 0000
***************************************
*** SUMMARY OF SCREEN MODEL RESULTS ***
***************************************
CALCULATION
PROCEDURE
SIMPLE TERRAIN
COMPLEX TERRAIN
MAX CONC
(UG/M**3)
.3231E-02
.3591E-03
DIST TO
MAX (M)
400.
8700.
TERRAIN
HT (M)
15.
55.
55. (24-HR CONC)
***************************************************
** REMEMBER TO INCLUDE BACKGROUND CONCENTRATIONS **
***************************************************
BUOY. FLUX - 7 269 M**4/S**3; MOM. FLUX = 85.174 M**4/S**2.
FINAL STABLE PLUME HEIGHT (M) = 55.2
DISTANCE TO FINAL RISE (M) - 151.3
*VALLEY 24-HR CALCS* **SIMPLE TERRAIN 24-HR CALCS**
TERR MAX 24-HR PLUME HT PLUME HT
HT DIST CONC CONC ABOVE STK CONC ABOVE STK U10M USTK
(M) (M) (UG/M**3) (UG/M**3) BASE (M) (UG/M**3) HGT (M) SC (M/S)
A-21
-------
55. 8700. .3591E-03 .3500E-04 55.2 .3591E-03 42.1 6 1.0 1.5
BUOY. FLUX = 7.269 M**4/S**3; MOM. FLUX = 85.174 M**4/S**2.
*** FULL METEOROLOGY ***
**********************************
*** SCREEN AUTOMATED DISTANCES ***
**********************************
*** TERRAIN HEIGHT OF 0. M ABOVE STACK BASE USED FOR FOLLOWING DISTANCES ***
DIST CONC
(M) (UG/M**3)
U10M USTK MIX HT PLUME SIGMA SIGMA
STAB (M/S) (M/S) (M) HT (M) Y (M) Z (M) DWASH
30.
100.
200.
300.
400.
MAXIMUM
414.
.4057E-14
.7292E-04
.1172E-02
.1242E-02
.1276E-02
6
1
1
2
3
1-HR CONCENTRATION
.1278E-02
3
1.0
3.0
3.0
4.0
4.5
AT OR
4.5
1.5
3 1
3.1
4.2
4.8
BEYOND
4.8
10000
960.
960.
1280.
1440.
30
1440
0
0
0
.0
.0
. M:
.0
62 08
50.12
50.12
42.59
39.67
39.67
7.
27.
50.
52,
45,
46,
,86
53
,71
,60
,00
.52
7.79
15.21
30.54
30.83
27.04
27.92
NO
NO
NO
NO
NO
NO
**********************************
*** SCREEN AUTOMATED DISTANCES ***
**********************************
***
TERRAIN HEIGHT OF 15. M ABOVE STACK BASE USED FOR FOLLOWING DISTANCES ***
DIST
(M)
400.
500.
600.
700.
800.
900.
1000.
1100.
1200.
1300.
1400.
1500.
1600.
1700.
1800.
1900
2000.
2100.
2200.
2300.
CONC
(UG/M**3) STAB
.3231E-02
.2902E-02
.2559E-02
.2287E-02
.2073E-02
.1897E-02
.1741E-02
.1618E-02
.1509E-02
.1407E-02
.1326E-02
.1355E-02
.1375E-02
.1412E-02
.1459E-02
.1498E-02
.1531E-02
.1546E-02
.1556E-02
.1562E-02
4
4
4
4
4
4
4
4
4
4
5
5
5
6
6
6
6
6
6
6
U10M
(M/S)
8.0
5.0
4.5
4.0
3.5
3.0
3.0
2.5
2.5
2.5
1.0
1.0
1.0
1.0
1.0
1 0
1.0
1.0
1.0
1.0
USTK
(M/S)
8.9
5.5
5.0
4.4
3.9
3.3
3.3
2.8
2.8
2.8
1.3
1.3
1.3
1.5
1.5
1.5
1.5
1.5
1.5
1.5
MIX HT
(M)
2560.
1600.
1440.
1280.
1120.
960.
960.
800.
800.
800.
10000.
10000
10000,
10000
10000
10000
10000
10000
10000
10000
0
0
0
0
0
0
0
0
0
0
.0
.0
.0
0
.0
.0
.0
.0
.0
.0
PLUME
HT (M)
15.
22.
24.
26.
29.
69
10
00
37
42
33.49
33.
39,
39
39
58
58
58
47
47
47
47
47
47
47
49
,19
.19
.19
.10
.10
.10
.08
.08
.08
.08
.08
.08
.08
SIGMA
Y (M)
29.61
36.47
43.06
49.57
56.01
62.42
68.61
74.95
81.03
87.07
70.86
75.24
79.61
56.24
59.10
61.96
64.80
67.63
70.46
73.27
SIGMA
Z (M) DWASH
15.57
18.94
21.89
24.80
27.68
30.57
33.11
35.50
37.39
39.24
30.75
31.79
32.81
22.92
23.54
24.14
24.74
25.26
25.76
26.25
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
A-22
-------
2400
2500
2600
2700.
2800
2900
3000.
3500.
4000
4500.
5000
5500.
6000
6500.
7000
7500.
8000.
8500
9000.
9500.
10000
15000
20000.
25000.
30000.
40000
50000.
MAXIMUM
400.
DIST
CONC
STAB
U10M
USTK
MIX HT
PLUME
SIGMA
SIGMA
DWASH
1566E-02 6 1 0 1.5 10000.0 47.08
1566E-02 6 1.0 1.5 10000.0 47.08
1565E-02 6 1.0 15 10000.0 47.08
.1561E-02 6 1.0 1.5 10000.0 47.08
1555E-02 6 1.0 1.5 10000.0 47.08
1547E-02 6 1 0 1.5 10000.0 47.08
1538E-02 6 1.0 1.5 10000 0 47.08
.1459E-02 6 1 0 1.5 10000.0 47.08
.1376E-02 6 1 0 1.5 10000.0 47.08
.1295E-02 6 1 0 1.5 10000 0 47.08
. 1218E-02 6 10 15 10000.0 47.08
.1147E-02 6 1.0 1 5 10000 0 47.08
.1081E-02 6 10 15 10000.0 47 08
.1020E-02 6 1.0 1 5 10000.0 47.08
.9642E-03 6 1.0 1.5 10000 0 47.08
.9124E-03 6 1 0 1.5 10000 0 47.08
.8651E-03 6 1.0 1.5 10000.0 47.08
.8219E-03 6 1.0 1.5 10000.0 47.08
.7824E-03 6 1.0 1.5 10000.0 47.08
.7460E-03 6 1 0 1.5 10000.0 47.08
.7125E-03 6 1 0 1.5 10000 0 47.08
4837E-03 6 1.0 1.5 10000.0 47.08
.3632E-03 6 1.0 1.5 10000.0 47.08
.2891E-03 6 1.0 1.5 10000.0 47.08
.2391E-03 6 1 0 1.5 10000 0 47.08
.1778E-03 6 1 0 1.5 10000.0 47.08
1410E-03 6 1.0 1.5 10000.0 47.08
1-HR CONCENTRATION AT OR BEYOND 400. M:
.3231E-02 4 8.0 8 9 2560.0 15.69
= DISTANCE FROM CENTER OF THE AREA SOURCE
= MAXIMUM GROUND LEVEL CONCENTRATION
= ATMOSPHERIC STABILITY CLASS (1=A. 2-B. 3-C.
= WIND SPEED AT THE 10-M LEVEL
= WIND SPEED AT STACK HEIGHT
- MIXING HEIGHT
HT= PLUME CENTERLINE HEIGHT
Y - LATERAL DISPERSION PARAMETER
Z - VERTICAL DISPERSION PARAMETER
= BUILDING DOWNWASH:
DWASH= MEANS NO CALC MADE (CONC =0.0)
DWASH=NO MEANS NO BUILDING DOWNWASH USED
DWASH=HS MEANS HUBER-SNYDER DOWNWASH USED
DWASH=SS MEANS SCHULMAN-SCIRE DOWNWASH USED
76.07
78.87
81.65
84.43
87.20
89.96
92.71
106.33
119.77
133 05
146.17
159.15
172.00
184.73
197.36
209.88
222.31
234.65
246.90
259.07
271.17
388 61
501.09
609.87
715.69
920 30
1117.49
29.61
4=0. 5-E.
26.74
27.22
27.70
28.17
28.63
29 08
29.53
31.37
33.10
34.72
36.26
37.72
39.13
40.47
41.77
42 88
43.96
44.99
46.00
46.97
47.92
56.19
61.48
65.96
69.88
75.45
80.10
15.57
6-F)
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
NO
DWASH=NA MEANS DOWNWASH NOT APPLICABLE. X<3*LB
A-23
-------
********************************************
* SUMMARY OF TERRAIN HEIGHTS ENTERED FOR *
* SIMPLE ELEVATED TERRAIN PROCEDURE *
********************************************
TERRAIN DISTANCE RANGE (M)
HT (M) MINIMUM MAXIMUM
0. 30 400.
15 400. 50000
**************************************
*** USER SPECIFIED AVERAGING TIMES ***
**************************************
ESTIMATED MAXIMUM CONCENTRATION FOR 24 HR AVERAGING TIME:
0.001292 (+/- 0.000646) UG/M**3
ESTIMATED MAXIMUM CONCENTRATION FOR ANNUAL AVERAGING TIME:
0.000258 (+/- 0.000065) UG/M**3
**********************************
*** END OF SCREEN MODEL OUTPUT ***
**********************************
A-24
-------
TABLE A-5. EXAMPLE ISCST2 OUTPUT
CO STARTING
TITLEONE An Example Superfund Site for the ISCST2 Model
MODELOPT DFAULT RURAL CONC
TERRHGTS ELEV
1 24 PERIOD
PCB
RUN
EVENTEXP.INP
ERRORS.OUT
AVERTIME
POLLUTID
RUNORNOT
EVENTFIL
ERRORFIL
CO FINISHED
SO STARTING
LOCATION INCIN1 POINT 00 00 0.0
** Point Source QS HS TS VS DS
**
rai aniCLCi i
SRCPARAM
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
INCIN1
PIT1
PIT2
PITS
PIT4
PITS
6.9E-5
AREA 0.
AREA 25.
AREA 25.
AREA 35.
AREA 35
** Area Source
** Parameters
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
PIT1
PIT2
PIT3
PIT4
PITS
ROAD1
ROA02
ROADS
ROAD4
ROADS
ROAD6
ROAD7
ROAD8
ROAD9
ROADO
-
2
2
2
2
2
** Volume Source
** Parameters
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
ROAD1
ROAD2
ROADS
ROAD4
ROADS
ROAD6
ROAD7
OS
—
.256E-4
.256E-4
.256E-4
.256E-4
.256E-4
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
OS
2.256E-5
2.256E-5
2.256E-5
2.256E-5
2.256E-5
2.256E-5
2.256E-5
20
.0
0 105
0 115.
0 105.
0 110
0 105.
HS
0
0
0
0
0
27
27
27
27
27
27
27
15
2
-10
1
1
1
1
1
1
1
.0
.0
.0
.0
.0
.5
.5
.5
.5
.5
.5
.5
.0
.5
.0
HS
.5
.5
.5
.5
.5
.5
.5
344
0
0
0
0
0
DX
25.
15.
10.
5.
5.
85.
72.
60.
47.
35.
22
10
10.
10.
10.
0
0.
0.
0
0
0
0
0
0
0
0
5
0
5
0
5
0
0
0
0
20
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
DY
5.
5.
5.
5
5.
5.
5.
81
81
81
81
81
81
81
0
0
0
0
0
0
0
.0 1.0
.0
0
.0 -.
.0
0
.0
.0
.0
.0
.0
DZ
.698
698
698
.698
698
698
698
A-25
-------
SRCPARAM
SRCPARAM
SRCPARAM
EM IS FACT
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
LOCATION
ROADS
ROAD9
ROADO
2.256E-5
2.256E-5
2.256E-5
1.
1.
1.
5
5
5
ROADO-ROAD9 HROFDY
PILE1A
PILE1B
PILE2A
PILE2B
PILE3A
PILE3B
** Volume Source
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
VOLUME
OS
-30
-25
-45
-40
-45
-40
7*0
0
0
0
0
0
0
HS
5.
5.
5.
.0
10
10
25
25
-5
-5
81
81
81
4*
.0
.0
0
.0
.0
.0
0
0
0
1
DY
0
0
0
0
0
0
0
698
698
698
0.0 4*1.0 8*0.0
.0
.0
.0
.0
.0
.0
DZ
rut QIMC L C I j .
SRCPARAM PILE1A
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
SRCPARAM
EMISFACT
SRCGROUP
SO FINISHED
PILE1B
PILE2A
PILE2B
PILE3A
PILE3B
0 01579
0 01579
0.01579
0.01579
0.01579
0.01579
i
1
1.
1.
1.
1.
5
5
5
5
5
5
PILE1A-PILE3B HROFDY
ALL
2.
2.
2.
2.
2.
2.
7*0.0
33
33
33
33
33
33
9*
0
0
0
0
0
0
1.
698
698
698
698
698
698
0 8*0.0
RE STARTING
GRIDCART GRID1 STA
XYINC -1000.0 21 100.0 -1000.0 21 100.0
ELEV 1 315*0.0 126*15.0
GRID1 END
GRIDCART GRID2 STA
XYINC -3000.0 13 500.0 -3000.0 13 500.0
ELEV 1 91*0.0 78*15.0
GRID2 END
** Site boundary points
DISCCART -150. -30. 0.0
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
-50.
50.
50.
50.
50.
50.
-50.
-150.
-150.
-150.
-150.
-30.
-30.
70.
170.
270.
370.
370.
370.
270.
170.
70.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
** High elevation points
OISCCART -2000. 8700. 55.2
DISCCART -1500. 8700. 55.2
A-26
-------
DISCCART -1000 8700. 55.2
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
DISCCART
RE FINISHED
ME STARTING
INPUTFIL
ANEMHGHT
SURFDATA
UAIRDATA
STARTEND
ME FINISHED
-500. 8700. 55.2
0 8700. 55.2
500. 8700 55.2
1000. 8700. 55.2
1500 8700. 55 2
2000. 8700. 55.2
METDATA\METDATA INP
10 METERS
12960 1991
3937 1991
91 03 01 91 11 05
OU STARTING
RECTABLE ALLAVE FIRST-SECOND
MAXTABLE ALLAVE 50
OU FINISHED
***********************************
*** SETUP Finishes Successfully ***
***********************************
A-27
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94 ***
*** 07:59:09
PAGE 1
*** MODELING OPTIONS USED: CONC RURAL ELEV DFAULT
MODEL SETUP OPTIONS SUMMARY
**Model Is Setup For Calculation of Average CONCentration Values.
**Model Use? RURAL Dispersion.
**Model Uses Regulatory DEFAULT Options-
1. Final Plume Rise.
2. Stack-tip Downwash.
3. Buoyancy-induced Dispersion.
4. Use Calms Processing Routine.
5. Not Use Missing Data Processing Routine.
6. Default Wind Profile Exponents.
7. Default Vertical Potential Temperature Gradients.
8. "Upper Bound" Values for Supersquat Buildings.
9. No Exponential Decay for RURAL Mode
**Model Accepts Receptors on ELEV Terrain.
**Model Assumes No FLAGPOLE Receptor Heights.
**Model Calculates 2 Short Term Average(s) of: 1-HR 24-HR
and Calculates PERIOD Averages
**This Run Includes: 22 Source(s): 1 Source Group(s): and 631 Receptor(s)
**The Model Assumes A Pollutant Type of: PCB
**Model Set To Continue RUNmng After the Setup Testing.
**0utput Options Selected:
Model Outputs Tables of PERIOD Averages by Receptor
Model Outputs Tables of Highest Short Term Values by Receptor (RECTABLE Keyword)
Model Outputs Tables of Overall Maximum Short Term Values (MAXTABLE Keyword)
**NOTE: The Following Flags May Appear Following CONC Values: c for Calm Hours
m for Missing Hours
b for Both Calm and Missing
Hours
**Misc. Inputs: Anem. Hgt. (m) = 10.00 : Decay Coef. = 0.0000 : Rot. Angle =
0.0
Emission Units = GRAMS/SEC ; Emission Rate
Unit Factor = 0.10000E+07
A-28
-------
Output Units = MICROGRAMS/M**3
**Input Runstream File, examl inp ; **0utput Print File:
examl.out
**Fi1e Created for Event Model: EVENTEXP.INP
**0etailed Error/Message File: ERRORS.OUT
A-29
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
07:59:09
PAGE 2
*** MODELING OPTIONS USED- CONC RURAL ELEV
DFAULT
POINT SOURCE DATA
NUMBER EMISSION RATE
STACK BUILDING EMISSION RATE
SOURCE PART. (GRAMS/SEC)
DIAMETER EXISTS SCALAR VARY
BASE STACK STACK STACK
ELEV. HEIGHT TEMP. EXIT VEL.
ID
(METERS)
CATS.
(METERS) (METERS) (METERS) (METERS) (DEG.K) (M/SEC)
BY
INCIN1
1,00 NO
0 0.69000E-04
0.0
0.0
0.0 20.1
344.00 20.
A-30
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
*•*•*
***
07:59:09
PAGE 3
*** MODELING OPTIONS USED: CONC RURAL ELEV
DFAULT
NUMBER EMISSION RATE
EMISSION RATE
SOURCE PART. (GRAMS/SEC)
SCALAR VARY
ID CATS
BY
*** VOLUME SOURCE DATA ***
BASE RELEASE INIT. INIT.
X Y ELEV. HEIGHT SY SZ
(METERS) (METERS) (METERS) (METERS) (METERS) (METERS)
ROAD1
HROFDY
ROAD2
HROFDY
ROAD3
HROFDY
ROAD4
HROFDY
ROADS
HROFDY
ROAD6
HROFDY
ROAD7
HROFDY
ROADS
HROFDY
ROAD9
HROFDY
ROADO
HROFDY
PILE1A
HROFDY
PILE1B
HROFDY
PILE2A
HROFDY
PILE2B
HROFDY
PILE3A
HROFDY
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.22560E-04
0.22560E-04
0.22560E-04
0 22560E-04
0.22560E-04
0 22560E-04
0 22560E-04
0.22560E-04
0.22560E-04
0.22560E-04
0.15790E-01
0.15790E-01
0.15790E-01
0.15790E-01
0.15790E-01
27.5
27.5
27 5
27.5
27 5
27.5
27.5
15 0
2.5
-10.0
-30 0
-25.0
-45 0
-40.0
-45.0
85.0
72.5
60.0
47.5
35.0
22.5
10.0
10.0
10.0
10.0
10.0
10.0
25.0
25 0
-5.0
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
0.0 1.50
5.81
5.81
5.81
5.81
5.81
5.81
5.81
5.81
5.81
5.81
2.33
2.33
2.33
2.33
2.33
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
0.70
A-31
-------
PILE3B 0 0.15790E-01 -40.0 -5.0 0.0 1.50 2 33 0.70
HROFDY
A-32
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
*** 07.59:09
PAGE 4
*** MODELING OPTIONS USED. CONC RURAL ELEV DFAULT
*** AREA SOURCE DATA ***
NUMBER EMISSION RATE COORD (SW CORNER) BASE RELEASE WIDTH EMISSION
RATE
SOURCE PART. (GRAMS/SEC X Y ELEV. HEIGHT OF AREA SCALAR
VARY
ID CATS. /METER**2) (METERS) (METERS) (METERS) (METERS) (METERS) BY
PIT1 0 0.22560E-03 0.0 105.0 0.0 0.00 25.00
PIT2 0 0.22560E-03 25.0 115.0 0.0 0.00 15.00
PITS 0 0.22560E-03 25.0 105.0 0.0 0.00 10.00
PIT4 0 0 22560E-03 35.0 110.0 0 0 0.00 5.00
PITS 0 0 22560E-03 35.0 105 0 0.0 0,00 5.00
A-33
-------
*** ISCST2 - VERSION 93109 *** *** An Example Super-fund Site for the ISCST2 Model
*** 10/12/94
***
***
07:59:09
PAGE 5
*** MODELING OPTIONS USED: CONG RURAL ELEV DFAULT
*** SOURCE IDs DEFINING SOURCE GROUPS ***
GROUP ID SOURCE IDs
ALL INCIN1 . PIT! , PIT2 , PIT3 . PIT4 , PIT5 . ROAD1 . ROAD2
ROAD3 , ROAD4 , ROAD5 . ROAD6
ROAD7 , ROADS . ROAD9 . ROADO . PILE1A , PILE1B . PILE2A , PILE2B
PILE3A . PILE3B ,
A-34
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
•*••**
07 59:09
PAGE 6
*** MODELING OPTIONS USED' CONC RURAL ELEV
DFAULT
SOURCE EMISSION RATE SCALARS WHICH VARY FOR EACH HOUR OF THE
DAY
HOUR SCALAR HOUR SCALAR HOUR SCALAR HOUR SCALAR HOUR
SCALAR HOUR SCALAR
SOURCE ID = ROAD1
1 .OOOOOE+00
.OOOOOE+00 6
7 .OOOOOE+00
.10000E+01 12 .
13 .10000E+01
.OOOOOE+00 18 .
19 .OOOOOE+00
.OOOOOE+00 24 .
SOURCE ID = ROAD2
1 .OOOOOE+00
.OOOOOE+00 6 .
7 .OOOOOE+00
.10000E+01 12 .
13 .10000E+01
.OOOOOE+00 18
19 .OOOOOE+00
.OOOOOE+00 24
SOURCE ID = ROAD3
1 .OOOOOE+00
.OOOOOE+00 6 .
7 .OOOOOE+00
.10000E+01 12
13 .10000E+01
.OOOOOE+00 18 .
19 .OOOOOE+00
; SOURCE TYPE =
2 OOOOOE+00
.OOOOOE+00
8 .10000E+01
.OOOOOE+00
14 .10000E+01
.OOOOOE+00
20 .OOOOOE+00
.OOOOOE+00
; SOURCE TYPE =
2 .OOOOOE+00
OOOOOE+00
8 .10000E+01
OOOOOE+00
14 .10000E+01
, OOOOOE+00
20 .OOOOOE+00
.OOOOOE+00
; SOURCE TYPE =
2 .OOOOOE+00
.OOOOOE+00
8 .10000E+01
.OOOOOE+00
14 .10000E+01
.OOOOOE+00
20 .OOOOOE+00
VOLUME :
3
9
15
21
VOLUME :
3
9
15
21
VOLUME :
3
9
15
21
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
4
10
16
22
4
10
16
22
4
10
16
22
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
5
11
17
23
5
11
17
23
5
11
17
23
SOURCE ID = ROAD4 ; SOURCE TYPE = VOJJME .
A-35
-------
1 .OOOOOE+00 2 .OOOOOE+00 3 .OOOOOE+00 4 .OOOOOE+00 5
.OOOOOE+00 6 OOOOOE+00
7 OOOOOE+OQ 8 .10000E+01 9 .10000E+01 10 .10000E+01 11
.10000E+01 12 .OOOOOE+00
13 .10000E+01 14 .10000E+01 15 .10000E+01 16 .10000E+01 17
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00 21 .OOOOOE+00 22 .OOOOOE+00 23
.OOOOQE+00 24 .OOOOOE+00
SOURCE ID = ROADS ; SOURCE TYPE - VOLUME •
1 .OOOOOE+00 2 .OOOOOE+00 3 .OOOOOE+00 4 .OOOOOE+00 5
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01 9 .10000E+01 10 .10000E+01 11
.10000E+01 12 .OOOOOE+00
13 .10000E+01 14 .10000E+01 15 .10000E+01 16 .10000E+01 17
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00 21 .OOOOOE+00 22 .OOOOOE+00 23
.OOOOOE+00 24 .OOOOOE+00
A-36
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
*** 07:59'09
PAGE 7
*** MODELING OPTIONS USED: CONC RURAL ELEV
DFAULT
* SOURCE EMISSION RATE SCALARS WHICH VARY FOR EACH HOUR OF THE
DAY
HOUR SCALAR HOUR SCALAR
SCALAR HOUR SCALAR
SOURCE ID = ROAD6 ; SOURCE TYPE •
1 .OOOOOE+OO 2 .OOOOOE+OO
.OOOOOE+OO 6 .OOOOOE+OO
7 .OOOOOE+OO 8 . 10000E+01
.10000E+01 12 .OOOOOE+OO
13 10000E+01 14 .10000E+01
.OOOOOE+OO 18 .OOOOOE+OO
19 .OOOOOE+OO 20 OOOOOE+OO
.OOOOOE+OO 24 .OOOOOE+OO
SOURCE ID = ROAD7 ; SOURCE TYPE =
1 .OOOOOE+OO 2 .OOOOOE+OO
.OOOOOE+OO 6 .OOOOOE+OO
7 .OOOOOE+OO 8 10000E+01
.10000E+01 12 .OOOOOE+OO
13 10000E+01 14 .10000E+01
.OOOOOE+OO 18 .OOOOOE+OO
19 .OOOOOE+OO 20 .OOOOOE+OO
.OOOOOE+OO 24 .OOOOOE+OO
SOURCE ID - ROAD8 : SOURCE TYPE =
1 OOOOOE+OO 2 .OOOOOE+OO
.OOOOOE+OO 6 .OOOOOE+OO
7 OOOOOE+OO 8 .10000E+01
.10000E+01 12 .OOOOOE+OO
13 .10000E+01 14 .10000E+01
.OOOOOE+OO 18 .OOOOOE+OO
19 OOOOOE+OO 20 .OOOOOE+OC
HOUR
VOLUME :
3
9
15
21
VOLUME :
3
9
15
21
VOLUME :
3
9
15
i 21
SCALAR
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
HOUR
4
10
16
22
4
10
16
22
4
10
16
22
SCALAR
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
.OOOOOE+OO
.10000E+01
.10000E+01
.OOOOOE+OO
HOUR
5
11
17
23
5
11
17
23
5
11
17
23
.OOOOOE+OO 24 .OOOOOE+OO
SOURCE ID - ROAD9
SOURCE TYPE = VOLUME
A-37
-------
1 .OOOOOE+00 2 .OOOOOE+00 3 .OOOOOE+00 4 OOOOOE+00 5
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01 9 .10000E+01 10 .10000E+01 11
.10000E+01 12 .OOOOOE+00
13 .10000E+01 14 .10000E+01 15 .10000E+01 16 .10000E+01 17
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00 21 .OOOOOE+00 22 .OOOOOE+00 23
.OOOOOE+00 24 OOOOOE+00
SOURCE ID = ROADO ; SOURCE TYPE = VOLUME •
1 .OOOOOE+00 2 .OOOOOE+00 3 .OOOOOE+00 4 OOOOOE+00 5
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01 9 .10000E+01 10 .10000E+01 11
.lOOOOE+Oi 12 .OOOOOE+00
13 10000E+01 14 .10000E+01 15 .10000E+01 16 10000E+01 17
.OOOOOE+00 18 .OOOOOE+00
19 OOOOOE+00 20 .OOOOOE+00 21 .OOOOOE+00 22 .OOOOOE+00 23
.OOOOOE+00 24 .OOOOOE+00
A-38
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
07:59:09
PAGE 8
*** MODELING OPTIONS USED. CONC RURAL ELEV
DFAULT
SOURCE EMISSION RATE SCALARS WHICH VARY FOR EACH HOUR OF THE
DAY
HOUR SCALAR HOUR SCALAR HOUR SCALAR HOUR SCALAR HOUR
SCALAR HOUR SCALAR
SOURCE ID = PILE1A . SOURCE TYPE =
1 .OOOOOE+00 2 .OOOOOE+00
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01
.10000E+01 12 .10000E+01
13 .10000E+01 14 .10000E+01
OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00
.OOOOOE+00 24 .OOOOOE+00
SOURCE ID = PILE1B ; SOURCE TYPE =
1 .OOOOOE+00 2 .OOOOOE+00
.OOOOOE+00 6 .OOOOOE+00
7 OOOOOE+00 8 .10000E+01
.10000E+01 12 .10000E+01
13 .10000E+01 14 .10000E+01
.OOOOOE+00 18 .OOOOOE+00
19 , OOOOOE+00 20 .OOOOOE+00
.OOOOOE+00 24 .OOOOOE+00
SOURCE ID - PILE2A . SOURCE TYPE =
1 .OOOOOE+00 2 .OOOOOE+00
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01
.10000E+01 12 .10000E+01
13 .10000E+01 14 .10000E+01
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00
VOLUME :
3
9
15
21
VOLUME :
3
9
15
21
VOLUME :
3
9
15
21
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
4
10
16
22
4
10
16
22
4
10
16
22
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
.OOOOOE+00
.10000E+01
.10000E+01
.OOOOOE+00
5
11
17
23
5
11
17
23
5
11
17
23
SOURCE ID - PILE2B : SOURCE TYPE
VOLUME .
A-39
-------
1 .OOOOOE+C
.OOOOOE+00 6
7 .OOOOOE+C
.10000E+01 12
2 .OOOOOE+00
.OOOOOE+00
8 .10000E+01
10000E+01
13 10000E+01 14 .10000E+01
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00
.OOOOOE+00 24 .OOOOOE+00
SOURCE ID = PILE3A ; SOURCE TYPE = VOLUME :
1 .OOOOOE+00 2 .OOOOOE+00
.OOOOOE+00 6 .OOOOOE+00
7 .OOOOOE+00 8 .10000E+01
.10000E+C1 12 .10000E+01
13 10000E+01 14 .10000E+01
.OOOOOE+00 18 .OOOOOE+00
19 .OOOOOE+00 20 .OOOOOE+00
.OOOOOE+00 24 .OOOOOE+00
3
9
15
21
IE :
3
9
15
21
.OOOOOE+00
.10000E+01
10000E+01
OOOOOE+00
.OOOOOE+00
.10000E+01
.iOOOOE+01
.OOOOOE+00
4
10
16
22
4
10
16
22
.OOOOOE+00
.IOOOOE+01
.IOOOOE+01
OOOOOE+00
.OOOOOE+00
.IOOOOE+01
.IOOOOE+01
.OOOOOE+00
5
11
17
23
5
11
17
23
A-40
-------
£2 00+300000' 22
LI 10+300001' 91
II 10+300001' 01
S 00+300000' fr
00+300000'
10+300001'
10+300001'
00+300000'
12
SI
6
e
: 3WniOA
00+300000' n 00+300000'
00+300000' 02 00+300000' 61
00+300000' 81 00+300000'
10+300001' frl 10+300001' SI
10+300001' 21 10+300001'
10+300001' 8 00+300000' L
00+300000 9 00+300000'
00+300000 2 00+300000' I
= 3dAi 3oanos : ae3iid = QI 33anos
yvivos ynoH avivos
anon yvivDS anon yvivos anon avivos ynoH avivos anon
3Hi do
H3V3
HOIHM swivos 3iva N0issiw3 33anos
AVQ
21S3SI
unvjo
A313
DNOD 'Q3sn
6 39Vd
60'6S:ZO
t?6/2I/OI
60ie6
9Nii3aow
- 21S3SI
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
*** 07:59:09
PAGE 10
*** MODELING OPTIONS USED. CONC RURAL ELEV DFAULT
*** GRIDDED RECEPTOR NETWORK SUMMARY ***
*** NETWORK ID: GRID1 : NETWORK TYPE: GRIDCART ***
*** X-COORDINATES OF GRID ***
(METERS)
-1000 0. -900.0, -800 0. -700.0. -600.0, -500.0. -400.0. -300.0.
-200.0. -100.0.
0.0, 100.0. 200.0, 300.0, 400.0. 500.0, 600.0, 700.0,
800.0, 900.0,
1000.0,
*** Y-COORDINATES OF GRID ***
(METERS)
-1000.0. -900.0. -800 0. -700 0. -600.0, -500.0, -400.0, -300.0,
-200.0, -100.0.
0.0. 100.0, 200.0, 300 0. 400.0. 500.0, 600.0, 700.0,
800.0, 900.0,
1000.0.
A-42
-------
EtrV
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
00 '0
OO'O
OO'O
OO'SI
00 '91
00 '91
OO'SI
OO'SI
OO'SI
00'009-
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'SI
OO'SI
00 '91
OO'SI
00 '91
OO'SI
00'009-
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'SI
OO'SI
OO'SI
00 '91
OO'SI
00 91
OO'OOZ-
OO'O
00 0
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'SI
OO'SI
OO'SI
OO'SI
OO'SI
OO'SI
00 '008-
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
00 0
00 0
OO'O
OO'O
OO'O
00 0
OO'O
00 0
00 '91
OO'SI
00 91
OO'SI
OO'SI
OO'SI
OO'SI
00 '91
OO'SI
OO'SI
OO'SI
OO'SI
OO'OOZ-
00 006-
OO'O
OO'O
OO'O
00 0
OO'O
OO'O
OO'O
OO'O
00 0
OO'O
00 0
OO'O
00 0
OO'O
OO'O
OO'O
OO'O
OO'O
OO'O
00 0
OO'SI
OO'SI
OO'SI
OO'SI
OO'SI
OO'SI
00 91
OO'SI
OO'SI
OO'SI
OO'SI
OO'SI
oo ooc-
OO'OOOI-
00 '0
1 00 '009-
00 '0
1 00'00l7-
OO'O
1 oo'ooe-
OO'O
1 OO'OOZ-
OO'O
1 OO'OOI-
00 '0
1 OO'O
OO'O
1 OO'OOI
00 0
1 OO'OOZ
OO'O
1 oo'ooe
OO'O
1 OO'OOfr
OO'SI
1 00'009
OO'SI
1 00 '009
OO'SI
1 00 00^
00 91
1 00 '008
OO'SI
1 OO'OOG
OO'SI
1 00 '0001
00'0017-
1 (SH313W)
(Sy313W) OH003-X
NI S1H9I3H NOI1VA313 x
:3dAi MRL3N : 101*19 'QI MRL3N xxx
A313 ivyny DNOD 'Q3sn SNOiidO BNIIBOOW
II 39Vd
60:69^0 xxx
XXX
176/31/01 xxx
21SDSI a^ Jo; a^s pun^jadns atdujgyB uv xxx xxx 60IS6 NOISy3A - 21S3SI xxx
-------
-600.00 |
0.00
-700.00 |
0.00
-800.00 |
0.00
-900.00 |
0.00
1000.00 |
0.00
0.00
0 00
0.00
0 00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0 00
0.00
0.00
0 00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 00
Q.OO
0.00
0.00
0.00
0 00
A-44
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
*** 07:59:09
PAGE 12
*** MODELING OPTIONS USED:
Y- COORD |
(METERS) |
500.00
1000.00 |
15.00
900.00 |
15 00
800.00 |
15.00
700.00 |
15.00
600.00 |
15.00
500.00 |
15.00
400.00 |
0.00
300.00 |
0.00
200.00 |
0.00
100.00 |
0.00
0.00 |
0.00
-100.00 |
0.00
-200.00 |
0.00
-300.00 |
0.00
-400.00 |
0.00
-500.00 |
0.00
-100 00
600.00
15.00
15.00
15.00
15.00
15 00
15 00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0 00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CONC RURAL ELEV DFAULT
*** NETWORK
0.00
700.00
15.00
15.00
15 00
15.00
15 00
15 00
15 00
15 00
15,00
15 00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0 00
0.00
0.00
0 00
0 00
0.00
0.00 ,
0.00
0.00
0.00
ID' GRID1
* ELEVATION
100.00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
.0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NETWORK
HEIGHTS IN
X- COORD
200 00
15.00
15.00
15 00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
TYPE- GRIDCART
METERS *
(METERS)
300.00
15.00
15.00
15 00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
***
400 00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
O.OQ
0.00
0.00
0.00
0.00
0.00
0.00
0.00
A-45
-------
-600.00 |
0.00
-700 00 |
0.00
-800.00 |
0.00
-900.00 |
0.00
1000.00 |
0.00
0.00
0.00
0 00
0 00
0.00
0.00
0.00
O.QO
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0.00
A-46
-------
ISCST2
VERSION 93109 ***
*** 10/12/94
***
07:59 09
PAGE 13
*** MODELING OPTIONS USED CONC
An Example Superfund Site for the ISCST2 Model
RURAL ELEV
DFAULT
*** NETWORK ID. GRID1 , NETWORK TYPE GR1DCART ***
* ELEVATION HEIGHTS IN METERS *
Y-COORD |
(METERS) 800
1000
900
800
700
600
500
400
300
200
100
0
-100
-200
-300
-400
-500
-600
-700
-800
-900
-1000
00 | 15
00 | 15
00 | 15
00 | 15
00 | 15
00
00
00
00
00
15
0
0
0
0
00 i 0
00 | 0
00 | 0
00
0
00 | 0
00
00
00
00
00
00
0
0
0
0
0
0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
900
15
15
15
15
15
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
1000
15
15
15
15
15
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
X-COORD (METERS)
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
A-47
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
*** 07:59:09
PAGE 14
*** MODELING OPTIONS USED. CONC RURAL ELEV DFAULT
*** GRIDDED RECEPTOR NETWORK SUMMARY ***
*** NETWORK ID. GRID2 : NETWORK TYPE: GRIDCART ***
*** X-COORDINATES OF GRID ***
(METERS)
-3000.0. -2500.0. -2000 0. -1500 0. -1000.0. -500.0. 0.0. 500.0.
1000.0, 1500.0,
2000.0, 2500.0. 3000.0,
*** Y-COORDINATES OF GRID ***
(METERS)
-3000.0. -2500.0. -2000 0, -1500 0. -1000.0, -500.0. 0 0. 500.0.
1000 0. 1500 0.
2000.0, 2500.0. 3000.0.
A-48
-------
*** ISCST2 - VERSION 93109 *** *** An Example Superfund Site for the ISCST2 Model
*** 10/12/94
***
***
07:59.09
PAGE 15
*** MODELING OPTIONS USED: CONC RURAL ELEV
DFAULT
*** NETWORK ID: GRID2 ; NETWORK TYPE. GRIDCART ***
* ELEVATION HEIGHTS IN METERS *
Y- COORD
(METERS)
0 00
3000.00
15.00
2500.00
15 00
2000.00
15.00
1500.00
15,00
1000.00
15.00
500.00
15.00
0.00
0.00
-500.00
0.00
-1000.00
0.00
-1500.00
0.00
-2000.00
0.00
-2500.00
0.00
-3000.00
0.00
-3000 00
500.00
15 00
15.00
15.00
15.00
15 00
15.00
15.00
15.00
15.00
15.00
15 00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0 00
-2500 00
1000.00
15 00
15.00
15.00
15.00
15 00
15.00
15.00
15.00
15.00
15.00
15 00
15.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0.00
-2000 00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
X-COORD (METERS)
-1500.00 -1000.00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-500.00
15.00
15.00
15.00
15.00
15.00
15.00
0 00
0.00
0.00
0.00
0.00
0.00
0.00
A-49
-------
*** ISCST2 - VERSION 93109 ***
*** 10/12/94
An Example Superfund Site for the ISCST2 Model
***
07:59:09
PAGE 16
*** MODELING OPTIONS USED: CONC
RURAL ELEV
DFAULT
*** NETWORK ID: GRID2 ; NETWORK TYPE. GRIDCART ***
* ELEVATION HEIGHTS IN METERS *
Y-COORD
(METERS)
3000.00
2500.00
2000.00
1500.00
1000.00
500.00
0.00
-500.00
-1000.00
-1500.00
-2000.00
-2500.00
-3000.00
1500.00
15.00
15.00
15 00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2000.00
15.00
15.00
15.00
15 00
15.00
15.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
2500.00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
X-COORD (METERS)
3000.00
15.00
15.00
15.00
15.00
15.00
15.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
A-50
-------
*** ISCST2 - VERSION 93109 ***
*** 10/12/94
An Example Superfund Site for the ISCST2 Model
*** 07 59.09
PAGE 17
*** MODELING OPTIONS USED: CONC RURAL ELEV
DFAULT
*** DISCRETE CARTESIAN RECEPTORS ***
(X-COORD. Y-COORD, 2ELEV. ZFLAG)
(METERS)
( -150.1
0.0);
-30.0,
0 0),
-50 0.
-30.0.
0 0,
A-51
-------
APPENDIX B
CHECKLIST OF MODELING CONSIDERATIONS
-------
APPENDIX B
SUPERFUND AIR PATHWAY ANALYSIS
CHECKLIST OF MODELING INPUT REQUIREMENTS
I PHYSICAL CHARACTERISTICS OF THE SITE
D Has a plot plan of the site been acquired?
D Has the site boundary been identified on a topographic map?
D Have the topographic features and water bodies at the site and in the
surrounding vicinity been identified?
D Have the soil and vegetation characteristics of the site and vicinity been
identified?
II. SOURCE DEFINITION
A. Nature and Extent of Contamination
D Have the locations of the sources within the site boundary been identified?
D Have the contaminants from each source been identified?
D Has the emissions mode (continuous or instantaneous) been identified for each
source?
D Has the physical state (gas or particle) of emissions to the atmosphere been
identified?
D Have fugitive emission sources been identified?
D - Has the spatial extent of the source areas been well-defined?
D - Have sources been subdivided into appropriate line, volume, or area
sources?
D Has the possibility of a dense-gas release been identified?
D Have any emissions control effects on source parameters been taken into
account?
D Has the potential for building wake effects been addressed for point source
emissions?
B. Emissions Estimation
6-1
-------
SUPERFUND AIR PATHWAY ANALYSIS
CHECKLIST OF MODELING INPUT REQUIREMENTS (Continued!
D Have maximum emission rates for short-term impacts prediction been identified?
D Have average long-term emission rates been identified?
D Has the particle size distribution been estimated for deposition modeling?
III. METEOROLOGICAL DATA
D Have representative meteorological data for the site been identified?
D Do the meteorological data meet the U.S. Environmental Protection Agency (EPA)
data quality and completeness requirements?
D Do climatological data for the site exist?
D Has worst-case meteorology been defined for models requiring a single set of
meteorological variables?
IV. RECEPTOR DATA
D Have sensitive population receptors been identified?
D Have sensitive environmental areas (wildlife preserves, wilderness areas,
etc.) been identified?
D If point sources are to be modeled, have elevated terrain receptors been
considered?
D If exposure pathways other than inhalation are to be addressed, have adequate
receptors been identified (e.g.. receptors placed on nearby water bodies to
account for potential exposure through the drinking water and fish consumption
pathways)?
V. LAND USE CLASSIFICATION
D Has the land use within 3 kilometers of the site been classified as urban or
rural, in accordance with EPA procedures?
VI. BACKGROUND CONCENTRATIONS
D Have any off-site contaminant sources been identified?
D Do background data exist from a past or existing air monitoring program in the
vicinity of the site?
-------
SUPERFUNO AIR PATHWAY ANALYSIS
CHECKLIST OF MODELING INPUT REQUIREMENTS (Continued)
VII EXPOSURE ASSESSMENT
D Have all air-related local, state, and federal ambient applicable or relevant
and appropriate requirements (ARARs) been identified?
D Does a population density map of the area exist?
D Have exposure pathways other than inhalation been considered?
D Have the uncertainties associated with the modeling technique been defined and
their implications to the Air Pathway Assessment (APA) discussed?
-------
APPENDIX C
BASIC REQUIREMENTS OF COMMONLY USED
DISPERSION MODELS FOR SUPERFUND APAs
-------
APPENDIX C
BASIC REQUIREMENTS OF COMMONLY USED DISPERSION MODELS
Common
Models
Principal Input
Principal Output
Major Assumptions
Screening
TSCREEN
SCREEN2
CTSCREEN"
Source
Emission rate
Height
Stack inner diameter (point source)
Gas exit velocity (point source)
Gas exit temperature (point source)
Building dimensions (point source)
Length of side of square (square area source)
Initial lateral dimension (volume source)
Initial vertical dimension (volume source)
Receptor
Other
Height (e.g.. ground level or breathing zone)
Downwind distance from source
Terrain elevation (height above source elevation)
Urban/rural classification
Digitized terrain contours (CTSCREENa)
Maximum
concentration and
associated downwind
distance (TSCREEN.
SCREEN?)
Maximum
concentration over
receptor grid
(CTSCREEN")
Worst-case meteorology
(TSCREEN. SCRECN2. CTSCREEN')
Short-term emissions occur
simultaneously
Maximum impacts are co-located
-------
APPENDIX C
BASIC REQUIREMENTS OF COMMONLY USED DISPERSION MODELS
(Continued)
Common
Models
Principal Input
Principal Output
Major Assumptions
Refined
ISCST2
ISCLT2
COMPLEX Ib
Source
Emission rate
Height
• Location
Stack inner diameter (point source)
• Gas exit velocity (point source)
Gas exit temperature (point source)
Building dimensions (point source)
• Length of side of square (square area source)
• Initial lateral dimension (volume source)
• Initial vertical dimension (volume source)
Meteorological Data -
Hourly surface and mixing height data (ISCST2 and
COMPLEX Ic)
Joint frequencies of wind speed and stability class
(STAR data) and average mixing heights and air
temperature (ISCLT2)
Receptor
. Height (e.g.. ground level or breathing zone)
« Cartesian or Polar coordinates
. Terrain elevation (height above source elevation)
Other
Urban/rural classification
Concentrations at
spatially
distributed receptor
points for varying
averaging periods
Meteorological data reflect
transport and dispersion
conditions at the site (if data
are not collected on site)
'Valid for receptors with terrain elevations above stack top.
"A complex terrain screening technique. Valid if used alone for receptors with terrain elevations above that of plume centerline.
Otherwise, used in conjunction with a simple terrain model for receptors with terrain elevations above stack top, but below plume centerline
(i.e., used for conducting an intermediate terrain analysis).
C0nly on-site meteorological data should be used if this model is run outside of the VALLEY screening mode.
Notes:
ISCST2 = Industrial Source Complex Short-Term Model.
Ten ro _
-------
APPENDIX D
USEFUL CONTACTS AND TELEPHONE NUMBERS
-------
APPENDIX D
USEFUL CONTACTS AND TELEPHONE NUMBERS
U S. ENVIRONMENTAL PROTECTION AGENCY (EPA) REGIONAL OFFICES
Each EPA regional office has the following staff positions:
Air/Superfund Coordinator.
Applicable or Relevant and Appropriate Requirements (ARARs) Coordinator; and
Air Toxics Coordinator.
The Air/Superfund coordinator is the best single point of contact for air issues
related to Superfund Sites The individuals in the staff positions listed above can be
reached through the office switchboards at the following numbers:
Region
I
II
III
IV
V
VI
VII
VIII
IX
X
Location
Boston
New York
Philadelphia
Atlanta
Chicago
Dallas
Kansas City
Denver
San Francisco
Seattle
Telephone
(617) 565-3420
(212) 264-2657*
(215) 597-9800
(404) 347-2864
(312) 353-2000
(214) 655-6444
(913) 551-7000
(303) 293-1603
(415) 744-1305
(206) 442-1200
*Air Programs Branch x-2517
AIR/SUPERFUND PROGRAM CONTACT
The primary contact for the Air/Superfund program is Ms. Patricia Flores.
Air/Superfund Coordinator of EPA Region III at (215) 597-9134.
D-l
-------
DOCUMENT ORDERING INFORMATION
Documents can be obtained through the National Technical Information Service (NTIS)
at (703) 487-4650. Information of Air/Superfund reports that are not yet in the NTIS system
can be obtained from Environmental Quality Management at (919) 489-5299.
Other sources of documents include:
EPA's Control Technology Center (CTC) at (919) 541-0800;
EPA's Center for Environmental Research Information (CERI) at (513) 569-7562.
and
U S. Government Printing Office (USGPO) at (202) 783-3238.
OTHER USEFUL CONTACTS
Air and Waste Management Association (412) 232-3444.
D-2
-------
APPENDIX E
BIBLIOGRAPHY OF NTGS DOCUMENTS
-------
APPENDIX E
BIBLIOGRAPHY OF NATIONAL TECHNICAL GUIDANCE STUDY (NTGS) DOCUMENTS
ASF-1 Eklund, B Procedures for Conducting Air Pathway Analyses for Superfund
Activities, Interim Final Document: Volume 1 - Overview of Air Pathway
Assessments for Superfund Sites (Revised) EPA-450/l-89-001a. February 1993.
ASF-2 Schmidt, C , et al Procedures for Conducting Air Pathway Analyses for
Superfund Activities, Interim Final Document. Volume 2 - Estimation of Baseline
Air Emissions at Superfund Sites (Revised). EPA-450/l-89-002a (NTIS PB90-
270588). August 1990.
ASF-3 Eklund. B.. et al. Procedures for Conducting Air Pathway Analyses for
Superfund Activities, Interim Final Document: Volume 3 - Estimation of Air
Emissions From Clean-up Activities at Superfund Sites. EPA-450/1-89-003 (NTIS
PB89-180061/AS). January 1989.
ASF-4 Hendler, A., et al. Procedures for Conducting Air Pathway Analyses for
Superfund Activities, Interim Final Document: Volume 4 - Guidance for Ambient
Air Monitoring at Superfund Sites. EPA-451/R-93-OQ7 (NTIS PB93-199214). May
1993.
ASF-5 U.S. EPA. Procedures for Conducting Air Pathway Assessments for Superfund
Sites, Interim Final Document: Volume 5 - Dispersion Modeling. [Proposed
document]
ASF-6 TRC Environmental Consultants. A Workbook of Screening Techniques For
Assessing Impacts of Toxic Air Pollutants. EPA-450/4-88-009 (NTIS PB89-
134340) September 1988
ASF-7 Salmons, C., F. Smith, and M. Messner. Guidance on Applying the Data Quality
Objectives For Ambient Air Monitoring Around Superfund Sites (Stages I & II).
EPA-450/4-89-015 (NTIS PB90-204603/AS). August 1989.
ASF-8 Pacific Environmental Services. Soil Vapor Extraction VOC Control Technology
Assessment. EPA-450/4-89-017 (NTIS PB90-216995). September 1989.
ASF-9 TRC Environmental Consultants. Review and Evaluation of Area Source Dispersion
Algorithms for Emission Sources at Superfund Sites. EPA-450/4-89-020 (NTIS
PB90-142753). November 1989.
ASF-10 Letkeman, J. Superfund Air Pathway Analysis Review Criteria Checklists. EPA-
450/1-90-001 (NTIS PB90-182544/AS). January 1990.
ASF-11 Smith. F., C. Salmons, M. Messner, and R. Shores. Guidance on Applying the
Data Quality Objectives For Ambient Air Monitoring Around Superfund Sites
(Stage III). EPA-450/4-90-005 (NTIS PB90-204611/AS). March 1990.
ASF-12 Saunders, G. Comparisons of Air Stripper Simulations and Field Performance
Data. EPA-450/1-90-002 (NTIS PB90-207317). March 1990.
ASF-13 Damle, A.S., and T.N. Rogers. Air/Superfund National Technical Guidance Study
Series- Air Stripper Design Manual. EPA-450/1-90-003 (NTIS PB91-125997). May
1990.
E-l
-------
ASF-14 Saunders, G. Development of Example Procedures for Evaluating the Air Impacts
of Soil Excavation Associated with Superfund Remedial Actions EPA-450/4-90-
014 (NTIS PB90-255662/AS). July 1990.
ASF-15 Paul, R. Contingency Plans at Superfund Sites Using Air Monitoring. EPA-
450/1-90-005 (NTIS PB91-102129). September 1990.
ASF-16 Stroupe. K., S Boone, and C. Thames. User's Guide to TSCREEN - A Model For
Screening Toxic Air Pollutant Concentrations. EPA-450/4-90-013 (NTIS PB91-
141820). December 1990.
ASF-17 Winges, K D . User's Guide for the Fugitive Dust Model (FDMXRevised). User's
Instructions. EPA-910/9-88-202R (NTIS PB90-215203. PB90-502410). January
1991.
ASF-18 Thompson. P., A. Ingles, and B. Eklund. Emission Factors For Superfund
Remediation Technologies EPA-450/1-91-001 (NTIS PB91-190-975). March 1991.
ASF-19 Eklund, B., C. Petnnec, D. Ranum. and L. Hewlett. Database of Emission Rate
Measurement Projects -Draft Technical Note. EPA-450/1-91-003 (NTIS PB91-
222059). June 1991.
ASF-20 Eklund, B.. S. Smith, and M. Hunt. Estimation of Air Impacts For Air Stripping
of Contaminated Water. EPA-450/1-91-002 (NTIS PB91-211888). May 1991 (Revised
August 1991).
ASF-21 Mann, C. and J. Carroll. Guideline For Predictive Baseline Emissions
Estimation Procedures For Superfund Sites. EPA-450/1-92-002 (NTIS PB92-
171909). January 1992.
ASF-22 Eklund, B.. S. Smith, P. Thompson, and A. Malik. Estimation of Air Impacts For
Soil Vapor Extraction (SVE) Systems. EPA-450/1-92-001 (NTIS PB92-143676/AS).
January 1992.
ASF-23 Carroll, J. Screening Procedures For Estimating the Air Impacts of
Incineration i ouperfund Sites. EPA-450/1-92-003 (NTIS PB92-171917).
February 1992.
ASF-24 Eklund, B.. S. Smith, and A. Hendler. Estimation of Air Impacts For the
Excavation of Contaminated Soil. EPA-450/1-92-004 (NTIS PB92-171925), March
1992.
ASF-25 Draves, J. and B. Eklund. Applicability of Open Path Monitors for Superfund
Site Cleanup. EPA-451/R-92-001 (NTIS PB93-138154). May 1992.
ASF-26 U S. EPA. Assessing Potential Air Impacts for Superfund Sites. EPA-451/R-92-
002. September 1992.
ASF-27 Hueske, K.. B. Eklund, and J. Barnett. Evaluation of Short-Term Air Action
Levels for Superfund Sites. EPA-451/R-93-009. April 1993.
ASF-28 Ranum, D. and B. Eklund. Compilation of Information on Real-Time Air Monitors
for Use at Superfund Sites. EPA-451/R-93-008. April 1993.
ASF-29 U.S. EPA. Air Emissions From Area Sources: Estimating Soil and Soil-Gas Sample
Number Requirements. EPA-451/R-93-002. March 1993.
ASF-30 Eklund. B. and C. Albert. Models For Estimating Air Emission Rates From
Superfund Remedial Actions. EPA-451/R-93-001. March 1993.
E-2
-------
ASF-31 U.S. EPA. Contingency Analysis Modeling for Superfund Sites and Other Sources.
EPA-454/R-93-001. 1993.
ASF-32 Eklund. B., C. Thompson, and S. Mischler. Estimation of Air Impacts From Area
Sources of Particulate Matter Emissions at Superfund Sites. EPA-451/R-93-004.
April 1993
ASF-33 Dulaney. W . B Eklund. C Thompson, and S. Mischler. Estimation of Air
Impacts For Bioventing Systems Used at Superfund Sites. EPA-451/R-93-003.
April 1993
ASF-34 Eklund. B . C. Thompson, and S Mischler. Estimation of Air Impacts For
Solidification and Stabilization Processes Used at Superfund Sites. EPA-451/R-
93-006. April 1993.
ASF-35 Dulaney. W . B. Eklund. C. Thompson, and S. Mischler. Estimation of Air
Impacts For Thermal Desorption Systems Used at Superfund Sites. EPA-451/R93-
005. April 1993
ASF-36 Options for Developing and Evaluating Mitigation Strategies for Indoor Air
Impacts at CERCLA Sites. EPA-451/R-93-012 April 1993.
Affiliated Reports
Eklund. B . et al. Control of Air Emissions From Superfund Sites. Final
Revised Report. EPA/625/R-92-012. U.S. EPA, Center for Environmental Research
Information. November 1992
E-3
-------
APPENDIX F
CONTAINER/ACUTE RELEASES
-------
APPENDIX F
CONTAINER/ACUTE RELEASES
At some Super-fund sites, buried drums containing liquid or cylinders containing gas
liquified under pressure may be encountered. During remediation activities, these items may
accidentally be punctured, causing a pollutant release. Characterizing the release in such
cases may be complicated, requiring knowledge of the source conditions at the time of
release and the thermodynamic properties of the released material. The source parameters
such as emission rate, exit velocity, and exit temperature are highly dependent on release
characterization The remainder of this section discusses some of the main concepts
involved with characterizing these types of releases.
For an air pathway assessment, it is ultimately the emissions to the atmosphere that
drive the predicted impacts For a punctured drum or cylinder, the pollutant may be
initially released as a liquid that consequently evaporates and enters the atmosphere as a
gas. Therefore, it is important to understand how the gas emission rate for this source may
be determined.
Each source of emissions can be defined by its pre-release, at-release, and post-
release conditions. The conditions prior to the release can determine the possible
conditions at the time of release. In turn, the at-release conditions can determine the
possible conditions occurring after the release. The primary pre-release conditions are:
Phase (solid, liquid, or gas);
Temperature: and
• Pressure.
Figure F-l shows a flowchart describing the high-level partitioning of the release
class by the pre-release phase state. The analysis cannot continue unless one of the state
variables is known. If one or two state variables are known, the other variables can be
estimated through the use of the flowcharts in Figures F-2. F-3, and F-4.
F-l
-------
One or two state
parameters
unknown. Use
Figures F-2, F-3,
andF-4
Need for
Modeling
Analysis
Initial state
(temperature, pressure,
and phase) of material
known?
artially known
Analysis cannot be
performed without at
least one state
parameter. Seek
expert advice.
Figure F-1. Scenario/Release Class Identification
-------
T = storage temperature
P = storage pressure
P3 = triple point pressure
T3 = triple point temperature
Assume material is
at ambient
temperature or T, if
known, and ambient
pressure
(-Solid
s ambien
temperature
-------
Assume material is a
compressed gas at
ambient temperature.
Determine contained
quantity and use ideal
gas law to find P
(=nRT/V).
How many
parameters (phase,
T, P) are known?
Phase is
known
One parameter
must be known
Material is either a
compressed gas (ideal
gas law) or a pressurized
liquid (vapor-liquid line).
Storage mass and volume
could be used to
determine phase by
comparing to liquid and
vapor densities.
Assume T equal to
ambient
temperature. Find
P from vapor-liquid
curve on
triple-point
diagram.
rombient
temperature
Yes
Material is likely a
refrigerated liquified
gas. P can be found
from the vapor-liquid
line on a triple-point
diagram.
Material is either a refrigerated liquified
gas or a low-volatility liquid (liquid
phase). T can be found from the
vapor-liquid line on a triple-point diagram.
If boiling point is larger than ambient
temperature, material is a low-volatility
liquid (use vapor-liquid Ine on
triple-point diagram). If boiling point is
less than ambient temperature,
material is either a compressed gas
(ideal gas law) or a pressurized liquid
(vapor-liquid line). Storage mass and
volume could be used to determine
phase by comparing to liquid and
vapor densities.
Figure F-3. Phase, temperature, and pressure determination for a stored liquid or gas with one parameter known
-------
matena
gas or
liquid?
Find point on triple-point
diagram corresponding to
known P, T. Region will
indicate the phase
Material is a
compressed gas.
Determine
storage quantity
and use ideal gas
law to find P
(=nRT/V)
Yes
Is
T
-------
Figure F-2 attempts to determine whether the pre-release state is a solid. If it Is
a solid but no other state information is known, the material may be assumed to be at
ambient temperature and pressure. If the temperature is known and phase is unknown, the
temperature can be compared to the melting point of the material. If the temperature is
below the melting point, it can be assumed that the material is a solid. If the material is
not a solid, the flow passes to Figure F-3 for determining whether the pre-release state is
liquid or gas. (Figures F-3 and F-4 refer to triple-point diagrams A triple-point diagram
for a particular compound represents the equilibrium relationship between all three phases
[liquid, gas. and solid] at any given pressure and temperature.)
Conditions at the time of a release can be determined given the pre-release
conditions. The at-release conditions are those parameters normally supplied to an air
dispersion model. The conditions at the time of release normally required or assumed
include:
Phase;
Emission rate;
• Temperature; and
Density.
Determining the at-release conditions requires calculation using the pre-release conditions
and thermodynamic properties of the material being released.
The post-release conditions refer to what may happen to an emission as it leaves the
source or after it has dispersed in the atmosphere. Normally, concentration impacts are the
major concern at a Superfund site. However, if the emitted material is flammable and an
ignition source is present, a fire or explosion may occur. If large quantities are
released, the effects from thermal radiation and/or Shockwaves associated with a fire or
explosion may be of sufficient concern to be considered.
F-6
-------
A release class is a combination of pre- and at-release conditions. The two primary
guidance documents containing discussion and flowcharts for determining release classes are
Contingency Analysis Modeling For Superfund Sites and Other Sources.1 and Guidance on the
Application of Refined Dispersion Models to Hazardous/Toxic Air Pollutant Releases.2
Once a release class has been determined, a model can be selected for simulation of
the release. The calculations performed in the referenced documents are not designed for a
specific model Rather, the calculations provide a number of parameters, all or some of
which may be needed by a specific model.
DETERMINING THE RELEASE CLASS OF A STORED LIQUID
For material stored in the liquid phase, the material's boiling point provides an
indication of the types of release classes that can be expected to result from a given
release scenario. Further determination of the release class, specifically the amount of
flashing of a pressurized liquid release, can also be made. A "flash diagram" may be used
to determine the approximate fraction of liquid that will flash to vapor during a release.
A flash diagram is constructed by use of the equation shown below.
where:
F = flash fraction (dimensionless);
Cpl = liquid heat capacity at Ts (J/kg K);
Ts = storage temperature (K):
Tb = normal boiling point (K);
X = heat of vaporization at Tb (J/kg);
J = Joule;
K = Kelvin; and
kg = kilogram.
F-7
-------
When Ts is below Tb, the flash fraction is set to zero. The flash fraction cannot be larger
than 1. When Ts is greater than the temperature given by the equation.
T - — * T.
S si O
the flash fraction should be set to 1.
The value of the flash fraction can be used to indicate whether the release is two-
phase (liquid and gas) Further, a comparison of the boiling point of the chemical with the
ambient temperature can be made to determine whether the release is a high-volatility or
low-volatility spill. If the boiling point is lower than the ambient temperature, the
release should be considered one of high volatility. If the boiling point is higher than
the ambient temperature, a low-volatility release is assumed. Determining whether the
liquid release is two-phase, high-volatility or low-volatility allows the user to choose the
appropriate liquid release option in TSCREEN3 (i.e., the user must know what the release
class is before proceeding with the screening analysis).
To be conservative, the two-phase release may be assumed to lead to a totally
suspended mixture of gas and liquid droplets. No liquid pool is assumed to form and then
evaporate. The high-volatility release may be assumed to have the liquid immediately
vaporize into gas upon release. The low-volatility release may be assumed to form a pool
that then evaporates. No flashing and no aerosol formation are assumed to occur.
figure F-5 illustrates the determination of a liquid release class from a container.
DETERMINING THE RELEASE CLASS OF A STORED GAS
For material stored in the gas phase, it must first be determined if there is choked
flow. A calculation is made to define the critical pressure that is then compared to the
ambient pressure.
F-8
-------
The calculation for determining if there is choked flow is:
p. Y*I
where:
p. = the critical pressure (Pa):
Pa = pascal (a unit of pressure)
Y = the ratio of the gas phase specific heat at constant pressure to that at
constant volume: and
ps = storage pressure (Pa).
F-9
-------
Tb = Normal boiling point
F = Flash fraction
Two-phase release
High-volatility spill
Assume the emission
temperature is Tb.
Calculate F.
No
x'ls Tb belowx.
ambient
No
Low-volatility spill
Figure F-5. Determination of a liquid release class from a container
-------
If p. is greater than or equal to the ambient pressure (pa), then the flow should be
considered choked. If p. is less than pa. then the flow is not choked.
The next step for determining the class for a stored gas release involves
determination of a reference temperature (Tref) and pressure (pref) indicative of the
conditions at the hole. The assumption is made that the release is two-phase, so that these
reference values can be checked against the chemical's properties for consistency. If all
calculated values are internally consistent, the release is two-phase. If there is an
inconsistency, the release is a single-phase gas release.
The method of determining the reference temperature and pressure depends on whether
or not the flow is choked. If the flow is choked, then the reference pressure is equal to
the critical pressure, and the temperature at choke conditions (T.) must be determined. This
calculation is derived from the Clausius-Clapeyron equation given as:
AM
P. - P. exp --
where:
pa = ambient pressure (Pa):
A = heat of vaporization at Tb (J/kg);
J - joule
K = Kelvin
Kg = Kilogram
kmol = kilomole
Pa = pascal (unit of pressure)
M = molecular weight (kg/kmol):
R - gas constant = 8314 J/kmol K;
Tb - normal boil-ing point (K): and
T. - temperature at choked conditions (K).
F-ll
-------
The above equation can be solved for T., which results In the following equation.
Tb AM p
The values o^ p. and T, could then be set to reference values pref and Tref.
If the flow is unchoked. the reference pressure (pref) is equal to the ambient
pressure (pa). The reference temperature (Tref), is set to the unchoked release temperature,
(Trel), which for unchoked flow is equal to Tb The value of T. is derived in the same manner
as for the choked flow but the value of p. is equal to pa. With p. equal to pa. the above
equation for T. reduces to T, = Tb.
Once the reference temperature and reference pressure have been determined, they can
be used to determine whether the release is two-phase (i.e.. whether any condensation occurs
during the release). This is done by performing two checks to determine whether the release
is single-phase. If both checks prove negative, the release is two-phase. The first check
is to compare the reference temperature to the critical temperature of the chemical. If the
reference temperature is greater than the critical temperature, the release is single-phase.
If not, the second check must be performed. In this check, if the vapor pressure of the
chemical (which must be externally calculated from the chemical data) at the reference
temperature is greater than the reference pressure, then the release is single-phase.
Otherwise, the release is two-phase. Figure F-6 illustrates the determination of a gas
release class from a container.
DETERMINING THE RELEASE CLASS OF A SOLID
Figure F-7 describes the logic in determining the release class from a solid. There
are only two classes considered here: vapor release from soil and from burning tires. The
document Contingency Analysis Modeling For Superfund Sites and Other Sources1 contains
guidance on calculating emissions from these sources.
F-12
-------
Tref = Reference temperature
Pref = Reference pressure
Tc = Critical temperature
Calculate temperature at
choked flow. Call it Tref.
Pref is choke pressure.
«_Yes
Calculate discharge
temperature. Call it Tref.
Pref is ambient pressure.
vapor
pressure at
Tref>Pref?
Yes.
.Yes-
Gas release
Gas release
Figure F-6. Determination of a gas release class from a container
-------
Calculate emission
rate
vapor
release from
soil?
Not referenced. Seek expert]
advice.
Yes
1
Calculate emission
rate
Figure F-7. Determination of a release class from a solid
-------
REFERENCES
1. U.S. Environmental Protection Agency, 1993. Contingency Analysis Modeling for
Suoerfund Sites and Other Sources. EPA-454/R-93-001. Office of Air Quality Planning
and Standards, Office of Air and Radiation, Research Triangle Park, North Carolina.
2. U.S. Environmental Protection Agency, 1993. Guidance on the Application of Refined
Dispersion Models to Hazardous/Toxic Air Pollutant Releases. EPA-454/R-93-002.
Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina.
3. US. Environmental Protection Agency, 1994. User's Guide to TSCREEN. A Modgl_ for
Screening Toxic Air Pollutant Concentrations._(ftgvi_sed).. EPA-454/3-94-023. Office of
Air Quality Planning and Standards, Research Triangle Park, North Carolina.
F-15
-------
"EPA-454/R-95-OC3
Air/Superfund National Technical Guidance
Study Series - Volume V-Procedures for Air
Dispersion Modeling At Superfund Sites
- , -_ -.
Radian Corporation
8501 Mo-Pac Boulevard
Austin, Texas 78759
•=•••, MN". AGEr.'i t.A!:r AN' A::?E.T
U.S. Environmental Protection Agency
Technical Support Division
Research Triangle Park, NC 27711
P E : r : C.T ' A E "•:'.'
February 1995
-. FEf^!:.",.-, ,..f-M.:--- • •_:-.
r FE?C w::: ':-~,-^:zf-~: • >••:- 'FT N
1 FiiO-RAK ELEKCr" '
EPA Contract No. 68-D30033
'.). TYPE OF PEPO^r At.1l PEP'.CI i.\ . EP E:
Final Report
., :=> :;- fi'. /-,-:•
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
EPA Work Assignment Manager: Jawad S. Touma
This manual is the fifth in a five-volume series dealing with air pathway
assessments at hazardous waste sites. It is an update of the air dispersion modeling
discussion in the original Volume IV of this series. Air pathway assessments involve
complex procedures requiring the use of professional judgment. This manual provides
for flexibility in tailoring the air pathway analysis to the specific conditions of
each site. It offers technical guidance for use by a diverse audience including EPA
Air and Superfund Regional and Headquarters staff, State Air and Superfund program
staff, Federal and State remedial and removal contractors, and potentially responsible
parties in analyzing air pathways at hazardous waste sites. It is written to serve the
needs of individuals having different levels of scientific training and experience in
designing, conducting, and reviewing air pathway analyses. Because assumptions and
judgements are required in many parts of the analysis, the individuals conducting air
pathway analyses need a strong technical background in air emission measurements,
modeling, monitoring, and risk assessment. Remedial Project Managers, On-Scene
Coordinators, and the Regional Air program staff, supported by the technical expertise
of their contractors, should use this guide when establishing data quality objectives
and the appropriate scientific approach to air pathway analysis.
KE> WORDS AND DOCUMENT ANALVCIi
*! DE"^MPTQP:7
Air Pollution
Superfund
Air Pathway Analysis
Dispersion Modeling
IK tllSTPIBUTlON STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN D1DEI' TERM."
li SECURITY CLAGL (Rtfpojr)
Unclassified
20. SECUPITY CLASS CPdye,'
Unclassified
c C'OCATI Fl- . 1/Gx ouf.
21 NO OF PA:E-
170
22 PMCE
ZP* rozm 2220-1 (»»T. 4-77)
------- |