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)

-------