v°/EPA
         United States
         Environmental Protection
         Agency
Office of Air Quality
Planning and Standards
Research Triangle Park. NC 27711
EPA-454/R-93-001
January 1993
         Air
        CONTINGENCY ANALYSIS
               MODELING
         FOR SUPERFUND SITES
          AND OTHER SOURCES

-------
                              EPA-454/R-93-001
CONTINGENCY ANALYSIS
         MODELING
  FOR SUPERFUND SITES
  AND OTHER SOURCES
        U.S. Environmental Protection Agency
        Region 5, Library ;PL-12J)
        77 West Jackson Boulevard, 12th Floor
        Chicago, IL 60604-3590
     Office Of Air Quality Planning And Standards
          Office Of Air And Radiation
       U. S. Environmental Protection Agency
        Research Triangle Park, NC 27711

             January 1993

-------
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 recommendation for use.
                                     EPA-454/R-93-001

-------
                                     TABLE OF CONTENTS

SECTION

1.0    INTRODUCTION AND EXECUTIVE SUMMARY                             1-1

2.0    IDENTIFICATION OF A REPRESENTATIVE RANGE                         2-1
       OF SOURCE TERMS

       2.1     Choice of Source Terms                                              2-1

               2.1.1   Telephone Survey                                            2-1
               2.1.2   Outcome of Telephone Survey                                 2-1

       2.2     Range of Possible Accident Scenarios                                  2-3

               2.2.1   Vessel Containing a Liquid                                    2-9
               2.2.2   Accident Scenarios Specifically Relevant to                      2-12
                      Superfund Sites

3.0    COMMUNICATIONS AND HOW TO USE THIS REPORT                      3-1

       3.1     Concerns Expressed During the Telephone Survey                        3-1
       3.2     Communications/Flow Charts ,                                      .3-1

               3.2.1   Accident Sequence Definition- General Discussion                3-1
               3.2.2   Accident Sequence Definition -                                3-9
                      Examples for Superfund  Sites
               3.2.3   Accident Sequence Definition -                                3-12
                      Examples for "Other" Sites
               3.2.4   Hydrogen Fluoride - an Interesting Case                        3-14

       3.3     Other Issues That Must be Addressed                                  3-15
       3.4     Conclusion                                                         3-17

4.0    BRIEF SURVEY OF AVAILABLE DISPERSION MODELS                      4-1

       4.1     Available Models                                                    4-1
       4.2     Why Atmospheric Dispersion Models Give Different Results               4-2
       4.3     Choice of Leak Size                                                 4-3
       4.4     Generic Issues in Atmospheric Dispersion Modeling                      4-4

               4.4.1   Atmospheric Stability Categories and Windspeed                  4-5
               4.4.2   Height at which Windspeed is Measured                        4-7
               4.4.3   Wind Direction                                              4-8
               4.4.4   AmbientTemperatures                                         4-8
               4.4.5   Relative Humidity                                            4-8
               4.4.6   Surface Roughness Length                                     4-8
               4.4.7   Toxicological Levels of Concern                               4-8
               4.4.8   Averaging Times                                            4-10
               4.4.9   Aerosolizatton                                              4-11

-------
                                TABLE OF CONTENTS (Continued)

SECTION                                                                         PAGE

5.0     SPILLAGE OF LIQUIDS ONTO SURFACES                                 5-1

       5.1     Spillage of a Liquid with Above Ambient Boiling Point                   5-1
               into a Diked Area - Acetone

               5.1.1   Description of Scenario                                        5-1
               5.1.2   Rate of Release of Liquid from Vessel                          5-1
               5.1.3   Behavior of Pool on the Ground                                5-2
               5.1.4   Calculation of Evaporation Rates                               5-3
               5.1.5   Equations in TSCREEN                                       5-5
               5.1.6   Duration of Evaporation                                       5-6
               5.1.7   Density of Mixture                                            5-6
               5.1.8   Input to SLAB                                               5-7
               5.1.9   Input to DEGADIS                                           5-10

       5.2     Spillage from Drum                                                   5-15
       5.3     Spillage of HF at 60°F                                                5-15

               5.3.1   Description of Scenario                                        5-15
               53.2   Calculation of Release Rate                    .                5-16
               5.3.3   Atmospheric Dispersion Model                                5-17

       5.4     Spillage of Cryogenic Liquid into a Diked Area                          5-17

               5.4.1   Choice of Example  - Release Rate a Function of Time            5-17
               5.4.2   Input to SLAB                                               5-19
               5.4.3   Input to DEGADIS                                           5-21

       5.5     Spillages  onto Water                                                  5-23
       5.6     Additional Considerations                                             5-23

               5.6.1   Heat Sources                                                 5-23
               5.6.2   Spreading Pool                                               5-24
               5.63   Advanced Modeling                                          5-24

6.0     JETS CONTAINING LIQUID AND VAPOR                                  6-1

       6.1     Emission Rate Formulae - Theory                                     6-1
               6.1.1    Gas Which Partially Condenses on Depressurization              6-2
               6.1.2    Saturated Liquid from Pressurized Storage                      6-6
               6.13    Subcooled Liquid from Pressurized Storage                      6-8
               6.1.4    Flow Chart                                                  6-10

-------
                               TABLE OF CONTENTS (Continued)

SECTION                                                                       PAGE

       6.2     HF at Elevated Temperature and Pressure                             6-10

               6.2.1    Description of Scenario                                       6-10
               6.2.2    Calculation of Release Rate                                   6-10
               6.2.3    Other Characteristics of the Source Term                       6-12
               6.2.4    Input to SLAB                                              6-14
               6.2.5    Input to DEGADIS                                          6-17

       6.3     One Ton Cylinder of Chlorine                                     .   6-22

               6.3.1    Release Description                                          6-22
               63.2    Input to SLAB                                              6-22
               633    Input to DEGADIS                                          6-25

       6.4     150 Ib Cylinder of Chlorine                                          6-28
       6.5     Miscellaneous                                                      6-29

               6.5.1    Jet Directed Downward                                      6-29
               6.5.2    Jet Emerging from a Long Pipe                               6-29
               6.5.3    Orifice Shape                                               6-30

7.0    VAPOR JET RELEASES                                                  7-1

       7.1     Vapor Release  Formulae - Theory                                    7-1

               7.1.1    Release Rate Estimates:  Leaks of Gas Directly                 7-2
                      from a Reservoir
               7.1.2    Release Rates:  Leaks of Gas from a Pipeline                   7-7
                      Attached to a Reservoir
               7.1.3    Flow Chart                                                 7-10

       7.2     Chlorine Vapor Release                                             7-10

               7.2.1    Description of Scenario                                       7-10
               7.2.2    Input to SLAB                                              7-12
               7.23    Input to DEGADIS                                          7-15

       73     Intermediate Sized Hole in the Vapor Space                            7-18

8.0    INSTANTANEOUS (PUFF) RELEASES                                     8-1

       8.1     Description of Release                                              8-1
       8.2     Input to SLAB                                                      8-1
       83     Input to DEGADIS                                                 8-4
       8.4     Small Cylinder  of Chlorine                                           8-7
                                              ill

-------
                               TABLE OF CONTENTS (Continued)

SECTION                                                                      PAGE

9.0     BUOYANT PLUME RELEASES                                           9-1

       9.1     Introduction                                                       9-1
       9.2     Incinerator Plume                                                  9-1

               9.2.1    Background                                                9-1
               9.2.2    Specification of Source Term for a Buoyant Plume               9-2
               9.23    Example of Source Term                                     9-7
               9.2.4    Atmospheric Dispersion Analysis                              9-8

       9.3     Fires at Ground Level                                               9-9

               9.3.1    Burning Pool                                               9-9
               9.3.2    Burning Tires                                               9-13
               9.3.3    Atmospheric Dispersion Modeling                             9-18

10.0    VAPOR RELEASES FROM MECHANICALLY DISTURBED SOIL           10-1

       10.1    Introduction                                                       10-1
       10.2    Equations for Emission Rates                                        10*3

               10.2.1  Average Long-Term Emission Rate                           10-3
               10.2.2  Average Short-Term Emission Rate                           10-3
               10.23  Simplified Average Short Term Emission Rate                  10-5
               10.2.4  Worst Case (Instantaneous) Emission Rate                     10-5

       10.3    Example                                                          10-6

               10.3.1  Description of Problem                                      10-6
               10.3.2  Total Emissions Potential for Site                             10-6
               1033  Average Short-Term Emission Rate                           10-6

       10.4    Atmospheric Dispersion                                             10-7


11.0    REFERENCES                                                           11-1
                                              IV

-------
                            LIST OF APPEND [CES

APPENDIX A EXAMPLE AND INTERPRETATION OF DEGADIS OUTPUT

APPENDIX B EXAMPLE AND INTERPRETATION OF SLAB OUTPUT

APPENDIX C INTRODUCTION TO BUOYANT PLUME RELEASES
APPENDIX D SIMPLIFIED SOURCE TERMS AND DENSITY CALCULATIONS
           FOR FLASHING LIQUID RELEASES

APPENDIX E LIST OF THOSE CONTACTED DURING THE
           TELEPHONE SURVEY

APPENDIX F SAFER ANALYSIS OF SCENARIOS FOR
           SUPERFUND SITES AND OTHER SOURCES

APPENDIX G HGSYSTEM

APPENDIX H CALCULATION OF MOLECULAR DIFFUSIVITY
           FROM BASIC PRINCIPLES

-------
                                         LIST OF TABLES

TABLE                             .                                              PAGE

4-1     Meteorological Conditions Defining Atmospheric Stability Categories              4-5
4-2     Relationship Between Atmospheric Stability Category,
        Surface Roughness Length z0 and Monin-Obukhov Length L                     4-6
4-3     Surface Roughness for Uniformly Distributed Ground Covers                    4-9

5-1     SLAB Input - Spillage of Acetone into a Diked Area                            5-8
5-2     DEGADIS  Input - Spillage of Acetone into a Diked Area                        5-11
5-3     Evaporation Rates as a Function of Time                                      5-19
5-4     SLAB Input - Spillage of Refrigerated Chlorine into a Diked Area                5-20
5-5     DEGADIS  Input - Spillage of Refrigerated Chlorine into a Diked Area           5-22

6-1     SLAB Input - Flashing Liquid Release of HF                                   6-15
6-2     DEGADIS  Input - Flashing Liquid Release of HF                               6-18
6-3     SLAB Input - Flashing Liquid Release of Chlorine from a One Ton Cylinder       6-23
6-4     DEGADIS  Input - Flashing Liquid Release of Chlorine                         6-26
        from a One Ton Cylinder

7-1     SLAB Input - Vertical Vapor Release of Chlorine                               7-13
7-2     DEGADIS  Input - Vertical Vapor Release of Chlorine                          7-16

8-1     SLAB Input - Puff Release of Chlorine                                         8-2
8-2     DEGADIS  Input - Puff Release of Chlorine                                    8-5

9-1     Polycyclic Aromatic Hydrocarbon Emission Factors                             9-15
        from Open  Burning Tires
9-2     Particulate Metals Emission Factors from Open Burning Tires                    9-16
9-3     Emission Factors for Organic Compounds from Open Burning of Tires            9-17
                                        LIST OF FIGURES
FIGURE                                                                          PAGE
2-1     Scenario Visualization                                                        2-4
2-2     Some Possible Releases at Super fund Sites                                     2-7
3-1     Scenarios Identification Flow Chart                                            3-2
6-1     Liquid Droplet Scenarios - Calculation Flow Chart                              6-11
7-1     Vapor Release Rate Calculations Flow Chart                                   7-11
9-1     Example of a Rotary Kiln Incineration System                                  9-3
10-1    Idealized Excavation Scenario                                                 10-2

-------
                        ACKNOWLEDGEMENTS

This report was prepared by Dina Christensen and Geoffrey Kaiser,
Science Applications International Corporation.  Jawad S. Touma was
the  Technical  Representative for  EPA.    Dr.  E.  D.  Chikhliwala,
EcoChem Technologies,  Inc.  provided  his  own time  for preparing
Appendix F.
                               via

-------
                     CONTINGENCY ANALYSIS MODELING
                                      FOR
                   SUPERFUND SITES AND OTHER SOURCES
1.0   INTRODUCTION AND EXECUTIVE SUMMARY

The  purpose  of this report is to provide information that will help  the Environmental
Protection Agency (EPA) develop guidance on contingency modeling  for Superfund sites
and other industrial sources. To this end, the possible range of different kinds of accidental
releases of hazardous vapors that might take place at such sites is first reviewed. These
scenarios are then used to illustrate how atmospheric dispersion models, including dense gas
models, should be applied. There is particular emphasis on the collection and calculation
of the input data that is needed for proper application of the models.

EPA has already developed a considerable amount of information on how to calculate the
characteristics of source terms and how to apply various dispersion models. Early guidance
was given in the "Guideline on Air Quality Models"(l) which describes screening and refined
air quality modeling techniques that focus on the six criteria pollutants  (particulate matter,
sulfur dioxide, nitrogen dioxide, carbon monoxide, ozone and lead). Subsequently, it was
recognized that there is an increasing need to provide models that specifically address the
impact of  toxic air pollutants. Such models deal with both heavier-than-air (dense) and
passive or neutrally buoyant (non-dense) releases.

To meet this  need, EPA published "A Workbook of Screening Techniques for Assessing
Impacts of Toxic Air Pollutants"® and then developed the TSCREEN personal computer
system(3) that utilizes concepts found in expert systems to implement the  scenarios described
in the workbook.  EPA also co-sponsored the development of the DEGADIS refined dense
gas model(4),  conducted  a  statistical  evaluation of seven dense gas models using  three
experimental  programs(S) and published guidance on  how to use the  DEGADIS, SLAB,
HEGADAS and AFTOX  modeis(6).   In  addition, guidance has  been developed on the
modeling  of  various  specific  scenarios,  including  (but  not limited  to)  incineration  at
Superfund sites00, the excavation of contaminated soil(8), the air stripping of contaminated
watert9) and the development of contingency plans using air monitoring(10).  The purpose of
                                       1-1

-------
the present work is to provide information that will help EPA develop further guidance that
supplements existing guidance.

In Section 2.0, there is a discussion of the kinds of release scenarios that are of interest at
Superfund and other industrial sites. In order to help determine this, a telephone survey
of interested EPA personnel was carried out.  Those contacted are listed in Appendix E.
The following  kinds of scenario were frequently mentioned  as being of interest:

      o     a range of accidents involving drums and cylinders, such as leaks, ruptures and
            fires

      o     accidents involving larger vessels that lead to evaporating pools; and

      o     accidents involving larger vessels containing gases liquified under pressure.

Section 2 goes  on to review the types of release scenarios (source terms) that span the range
of interests expressed above.  These scenarios are summarized in Figures 2-1 and 2-2.

In Section 3, the  subject of communications is discussed. This was also one of the topics
that was raised during the telephone survey, in which the concerns expressed were far from
uniform. However, there does seem to be an underlying thread of concern running through
what was said, namely how  can miscommunication between  the person requesting  the
analysis (the requestor) and the person performing the analysis (the analyst) be avoided?
Put another way,  how can these two individuals communicate in such  a way that there  is a
clear understanding of the scenario that is to be modeled and the information that  the
requestor requires? Other concerns raised included how to ensure that the source term is
realistic and the  model is used in such a way as to give credible answers.  Section 3 is
intended to address these concerns and also to show how this report should be used.

Section 3 addresses the communications problem by means of a flow  chart (Figure 3-1)
which is intended to help the requestor and the analyst accurately  define the scenario of
interest.  The  flow chart also directs the reader to specific sections  in this report where
various  scenarios are discussed in detail, except  in those cases where  complex physical
phenomena are  involved,  such as partial aerosolization,  in  which case  the  reader is
recommended to seek expert advice.
                                        1-2

-------
Section 3 also summarizes the scenarios that have been developed as examples for the
purposes of the present report.  Specifically, for Superfund sites, the following scenarios
have been chosen:

       o     spillage of a large quantity of acetone onto the ground.  This is meant to
             represent cases in which a liquid with a boiling point that is high relative to
             the ambient temperature, and a corresponding vapor pressure that is less than
             atmospheric, spills onto the ground and evaporates

       o     spillage of a liquid with a high boiling point from a drum

       o     various accidents involving cylinders, including vapor releases, liquid releases
             and large ruptures

       o     incineration at Superfund sites

       o     fire  involving the spillage of a liquid pool containing contaminated materials

       o     emissions from burning tires

       o     emissions from mechanically disturbed contaminated soil; and

       o     (for completeness) spillages onto water.

Various scenarios have also been chosen for "other" sites. Note, however, that many of the
phenomena discussed for the following scenarios are relevant to scenarios at Superfund
sites:

       o     spillage of refrigerated chlorine,onto the ground

       o     flashing liquid jet releases from chlorine storage, leading to scenarios in which
             the plume contains both vapor and liquid droplets (aerosol)

       o     puff releases from chlorine storage
                                        1-3

-------
      o     vapor releases of chlorine driven by high pressure from a storage vessel; and

      o     some scenarios involving hydrogen fluoride, which is an interesting case in
            which complicated thermodynamic and physical properties mean that there is
            a particularly wide range of possible source terms, depending on the way in
            which the material is released.

Section 3 also provides a check list of items that should be discussed by the requestor and
analyst.  These include what the output of the model should  be, the choice of toxicological
levels of concern and  meteorological considerations  that  may affect the atmospheric
dispersion.

Section 4 begins with a brief review of available atmospheric dispersion models. Of these,
two  were chosen for use in this report, DEGADIS and SLAB. The choice was somewhat
arbitrary.  The basis included such factors as  the resources available for this study, the
availability of the models and the familiarity of the authors with them.  For cases where the
use of the above models is not appropriate (e.g. when the initial source term is not denser
than air) the use of EPA's TSCREEN model is discussed.  Subsequently, an arrangement
was made with the authors of the proprietary computer model, SAFER, who ran a limited
number of scenarios as described in Appendix F.  Finally, two scenarios were run with the
relatively new model HGSYSTEM, see Appendix G.

Section 4 also discusses issues that are common to all dispersion modeling. These include,
for example, the choice of surface  roughness length, the choice of toxicological Levels of
Concern, the weather conditions and averaging times.

Sections 5 through 10 discuss in detail how several scenarios should be modeled in SLAB
and DEGADIS. The expectation is that, by following detailed discussions of the inputs to
each of the computer models, the reader  will come  to understand what is required to
produce a credible model of a particular scenario.  For example, he/she will learn that, even
with widely used  models, a great deal of work is required to cast the  input  into an
appropriate form  and that, in some cases, a considerable amount  of technical knowledge
about the input data is required. The reader should note that Sections 5 through 10 are not
organized in the  sequence  of the list of  scenarios described above.  Instead, they are
organized in groups for which the physical and chemical phenomena are similar:
                                       1-4

-------
       o     section 5 - vapors evaporating from liquids spilled onto the ground

       o     section 6 - flashing  liquid jets  that result  in an initial airborne release
             containing both vapors and liquid droplets (aerosols)

       o     section 7 - vapor releases driven out of a vessel by high pressure

       o     section 8 - puff releases

       o     section 9 - buoyant plumes; and

       o     section 10 - emissions from mechanically disturbed soils.

As noted above,  Section 5.0 is concerned with the spillage of liquids onto surfaces from
which evaporation subsequently occurs.  The first case is that of the spillage of acetone into
a diked area. This scenario has been chosen because it serves as a good  introduction to the
topic of evaporating pools, which are the likely mode of release for many high boiling point
liquids at Superfund Sites. A similar case considered is that of the small scale spillage from
a drum.  The principal difference from  the large scale spill is  the relative smallness of the
quantity released.

Another case considered in Section 5  is that of the spillage of hydrogen fluoride from a
storage vessel at  an industrial or "other" site. This case is chosen because the information
needed to develop the input data for  the computer  models is  more  elaborate for HF.
Finally, the spillage of refrigerated chlorine into a diked area is discussed as an introduction
to the  modeling of spillages of cryogenic liquids.

Section 5 continues with a brief "discussion of spillages onto water.  These are not treated
in detail but are included for completeness: the most important phenomena are summarized
and the reader is referred to  appropriate references. Chapter 5 concludes with discussions
of some additional  issues relating to evaporating pools for which this project did not allow
sufficient resources. These include additional sources of heat, spreading pools and advanced
numerical modeling techniques.
                                         1-5

-------
Section 6 is devoted to the subject of flashing liquid jets in which the initial source term
consists of a mixture of vapor and liquid droplets that is released directly from the leaking
vessel or pipework.  The  first example given is that of a release of HF from an orifice of
diameter 1/2" in a vessel containing HF at elevated temperature and pressure.  This is a
release  that  is very much characteristic  of those  expected at "other"  industrial  sites.
However, it is useful to begin with it because, by carefully considering it, the reader will
become familiar with many of the issues that must be addressed if such scenarios are to be
modeled realistically.

The HF example is followed by consideration of a flashing liquid  release from a one ton
cylinder of  chlorine at ambient temperature. Such one ton cylinders are  commonly found
at places such as water treatment facilities.  Finally, the specific example of a cylinder
containing 150 Ib of chlorine is examined as being particularly pertinent to the sort of case
that might need to be examined at a Superfund site.

Section 7.0 is devoted to a discussion of releases from the vapor space  of a vessel from
which the expected release will be pure vapor.  The specific example of chlorine storage is
chosen.

Section 8.0 considers puff releases.  The reader is shown how to  model the catastrophic
failure of vessels.  The specific examples chosen are those  of 1 ton and 150 Ib  chlorine
cylinders.

Section 9.0 introduces the subject of plume rise. Three examples are chosen, a) the plume
from  an incinerator at a Superfund site,  b) the plume  from a  burning  pool  of  liquid
containing toxic materials and c) a stack of burning tires.

Chapter 10 is devoted to a brief discussion of the issues involved  in modeling the plume
released when contaminated soil is accidentally disturbed.

Various backup appendices are provided. Appendix A gives an example of an output from
DEGADIS and Appendix B  gives an  example of an output from  SLAB.   Appendix C
consists of an introduction to  the important elements that must  be considered when
calculating  the atmospheric  dispersion of buoyant plumes.  This is meant to be  an
introduction for the reader who  is interested in pursuing some of the  more sophisticated
                                        1-6

-------
issues of plume rise, such as the suppression of lift-off of a buoyant plume initially on the
ground and the termination of plume rise.  These issues are often not considered in the
simpler  models.   Appendix D contains a discussion  of some  simplified  methods  of
calculating the densities of mixtures of air and aerosols.  Appendix E contains the list of
those contacted during the  telephone survey of EPA personnel. Appendix F contains the
outputs  of SAFER  analyses,  which were  kindly provided  by Dr  E.K. Chikhliwala  of
EcoChem Technologies, Inc.  The  subject  of Appendix  G is  the computer  model
HGSYSTEM.  Finally, Appendix H shows how to calculate molecular diffusivity, a quantity
that is required for the estimation of the rate of evaporation of a pool lying on the ground.

In conclusion, the reader should note that the emphasis in this report is on the source term
and in particular how to prepare inputs for the computer models that credibly represent the
scenarios in  question.  The SLAB, DEGADIS and HGSYSTEM model Users' Guides
adequately describe  the model assumptions and output.  For further information on the
proprietary model SAFER, the reader should contact EcoChem Technologies.
                                       1-7

-------
2.0   IDENTIFICATION OF A REPRESENTATIVE RANGE OF SOURCE TERMS

2.1   Choice of Source Terms

2.1.1  Telephone Survey

In order to identify a range of source terms that might be useful to potential readers, a
telephone survey was  conducted, encompassing several persons within EPA who had
expressed an interest in  this  project or who were thought to have  relevant skills and
expertise. The people who were contacted are listed in Appendix E.   In general, two
subjects were discussed:

      (i)   What kinds of releases should be included in the potential guidance?  This
            topic is discussed in Section 2.1.2.

      (ii)   What communications problems might arise between the person requesting
            a contingency analysis (the requestor)  and the analyst and how can the
            potential for miscommunication be minimized?  This topic is discussed in
            Section 3.0.

2.1.2  Outcome of the Telephone Survey

In the past there has been no systematic compilation of those kinds of source terms that are
potentially of most  use to those who have an interest in contingency modeling at Superfund
sites:  this seems to be largely because there is a diversity of concerns arising from the many
different chemicals that are likely to be encountered.  During the telephone survey, the
following or similar source terms were discussed (in no particular order):

      1.    Explosive detonation of HCN  cylinder/ or rupture due to polymerization.

      2.    Transfer of methyl mercaptans from a tank with tank failure or seal failure.

      3.    Spillage from a tank containing carbon disulfide.
                                       2-1

-------
      4.     Rupture of cylinders and/or tanks containing gas liquified under pressure such
             as chlorine, ammonia and hydrogen fluoride.

      5.     Contents of a drum catch fire: in general, fires of various types at a Superfund
             site

      6.     Rupture of a 300,000  gallon tank of dichlorobenzene with  damage  to and
             leaks from two adjacent tanks.

      7.     Acetone spill from a 25,000 gallon tank

      8.     Abandoned rail cars containing various materials leak.

      9.     Train derailment

      10.    Vapors released from  disturbed soil; and

      11.    Incineration at a Superfund site.

As expected, it became apparent that accidents involving drums and cylinders are important
to on-scene coordinators.  These are considered in the present document.

Spillage from larger vessels containing materials that are liquid at ambient pressure and
temperature were also frequently mentioned. Therefore, this document contains examples
of evaporation from pools spilled on the ground.

One of the more surprising aspects was how often chlorine and ammonia were mentioned,
not just in  the  context  of  small cylinders  but also  relating to  larger vessels  and  to
transportation accidents.  This means that the subject of flashing liquid releases needs to be
addressed. This is an area where the apparent interests of those who are concerned with
Superfund sites overlaps with the concerns of those at  "other" sites.

In summary, the discussions with the  personnel listed in Appendix E led to the identification
of the need to consider:
                                        2-2

-------
       o     A range of accidents involving drums and cylinders
       o     Accidents that lead to evaporating pools; and
       o     Accidents involving gases liquified under pressure.

The range of possible accident scenarios corresponding to the above categories is discussed
below.

22    Range of Possible Accident Scenarios

One of the first things that a contingency modeler learns is  that there is no single source
term and no single atmospheric dispersion model that will cover all of the different kinds
of releases that might take place at a Superfund or other site. The following examples are
intended to serve as surrogates for many of the more likely  accident scenarios  that might
be encountered at such sites.  These examples are organized  in relation to the physical and
chemical phenomena that need to be addressed.  See Section 3.0 for a discussion of how
they are organized with respect to Superfund or "other" sites.

It  is convenient to begin with one simple case that can be used to illustrate many of the
issues that confront a contingency modeler. This  is  the case of a storage vessel  or truck
containing a liquid, together with some associated pipework.  For example, there may be a
vessel on  a Superfund site containing a liquid that must be transferred into a truck and
transported offsite.   There is a possibility that there will be a  leak or a rupture during
preparation for transfer itself - for example, the failure of a valve when an operator attempts
to open it or the failure of the vessel or pipework in an area that may have been weakened
by corrosion.

As noted  above, it is convenient  to begin with the variety of possible leaks from such a
vessel  because, in  discussing the range  of  possible source  terms,  the reader  will  be
introduced to many of the issues and phenomena that a contingency analyst must consider,
either  at a Superfund  site or at an  industrial site where various  liquids are stored and
transferred.

Figure 2-1 contains a summary of the range of possible accident scenarios that might arise
in the  context  of a  vessel containing a  liquid. The sections in this  report where these
scenarios are discussed are given in parentheses in the third column.  Those sections
                                        2-3

-------
      DESCRIPTION OF
       RELEASE CASE
VISUALIZATION
   APPLICABILITY
(SECTION OF REPORT)
PHENOMENA/ISSUES
    ADDRESSED
1 A. Rupture in Vessel Containing
    Liquid With Above Ambient
    Boiling Point at Low Pressure
a)  Spillage Into Diked Area
b)  Unconfined Spillage
IB. Rupture in Vessel Containing
    Liquid With Above Ambient
    Boiling Point at High
    Pressure
                                   CONFINED OR
                                     UNCONFINED POOL
 Rupture in Vessel or
 Pipework While Preparing
 to Transfer Contents

 Accidents Involving
 Vessels Storing High
 Boiling Point Liquids on
 Industrial Sites

 Vessel Standing in Diked
 Area
      (5.1 and 5.2)
                         Undiked Vessel
                                 (5.6.2)
                         Rupture in Pipework While
                         Transfering Liquid Using
                         Nitrogen Pad

                         Truck Unloading
                         Operations
                                (4.2.9)
                                                       Calculation of Liquid
                                                       Release Rales - Static Head
                                                       Calculation of Evaporation
                                                       Rates From a Pool of Finite
                                                       Area
                               Calculation of Evaporation
                               Rates From an Unconfincd
                               Pool
                             •  Definition of Cases in
                               Which Pressure Driven
                               Releases Can Form
                               Aerosols
2.  • Rupture in Vessel Containing
    Refrigerated Liquids -
    Same Cases as 1 A and
    1 B Above
 Same As Above
 Ruptures During Storage
 or Transfer of Cryogenic
 Liquids
         (5.3)
Calculation of the Rate
of Evaporation of Cryogens
Spilled Onto Land
                                               Figure 2-S, Page 1
                                             Scenario Visuafinzaftion

-------
           PESCRIPTION OF
                                        VISUALIZATION
                                                               APPLICABILITY
                                                            (SECTION OF REPORT)
                              PHENOMENA/ISSUES
                                  ADDRESSED
to
Oi
     3A. Liquid Release From Wall
         of Vessel Containing Gas
         Liquified Under Pressure
3 B.  Liquid Release of Gas
    Liquified Under Pressure
    From Long Pipe (A or B)
     3C.  Liquid Release of Gas
          Liquified Under Pressure -
          Impingement Onto Surface
     3D. Catastrophic Failure of
         Vessel Containing Gas
         Liquified Under Pressure
                                    HIGH PRESSURE
                                   f/////sssr.
Rupture of Short Pipe or
Vessel Containing Gas
Liquified Under Pressure
     (6.2 - 6.4)
Rupture of Long Pipe
Containing Gas Liquified
Under Pressure
       (6.5.2)
                                                            Rupture of Vessel
                                                            Containing Gas Liquified
                                                            Under Pressure
                                                                   (6.5.1)
                                                            Catastrophic Failures
                                                                    (8.1)
Rate of Release of Liquids
Driven by High Pressure

Flashing

Flashing-Driven
Aerosoli/ation

Momentum Effects
Flashing in Pipe

Rate of Release of Two
Phase Mixtures
                               Droplet Recovery, Pool
                               Formation
                            •  Definition of
                               Characteristics of Puff
                               Sources
                                                     Figure 2-1, Page 2
                                                  Scenario Visuj?Si

-------
      DESCRIPTION OF
                                  VISUALIZATION
                         APPLICABILITY
                      (SECTION OF REPORT)
                                PHENOMENA/ISSUES
                                     ADDRESSED
4A. Small Hole in Vapor Space
    of Vessel Containing Gas
    Liquified Under Pressure
4B. Catastrophic Failure in
    Vapor Space
4C. "Intermediate" Sized Hole
    in Vapor Space of Vessel
    Containing Gas Liquified
    Under Pressure
See 3D
                                         2 PHASE
                                         RELEASE
        FOAMING
        OR
        FROTHING
• Rupture in Vapor Space
  of Tank During Transfer
  Operations

• Leaks in Pipes

• Relief Valve Discharges
           (7.2)

  See 3D
  Intermediate-sized Ruptures
  in Vapor Space of
  Pressurized Vessels
           (7.3)
                                                     Rate of Release of
                                                     Vapors Driven by
                                                     High Pressure

                                                     Momentum Effects
See 3D
Two Phase Flow Regimes
(Bubbly Row, Churn
Turbulent Flow, Droplet
Flow)
5.   Spillages Onto Water
                    •  Spill From Vessel Into
                       Diked Area That
                       Contains Water

                    •  Spillage Adjacent to
                       Waterway

                    •  Marine Transportation
                       Accidents
                               (5.5)
                                Calculations of Rate of
                                Evaporation of Liquids
                                Spilled Onto Water
                                              Figure 2-1, Page 3

-------
      DESCRIPTION OF
       RELEASE CASE
VISUALIZATION
   APPLICABILITY
(SECTION OF REPORT)
PHENOMENA/ISSUES
    ADDRESSED
6A. A Drum Containing a Liquid
    is ftmctured and the Contents
    Full lo (he Ground and
    Evaporate
611. A Cylinder Containing a
    Gas Liquified Under
    Pressure is Punctured
6C. A Cylinder Containing a
    Compressed Gas is
    Punctured
61). Catastrophic Rupture of
    Cylinder
6li. Contents of Drum Burn
                                    FLASHING
                                    LIQUID JET .
                                 INSTANT ANI-OUS. PUFF
                                                /
         BUOYANT
         VAPOR
         CLOUD
                                          .LID
                      •  Drum Removal Activities
                        at Superlund Site Puncture
                        Drum
                                 (5.4)
                        Cylinder Removal
                        Activities Puncture
                        Cylinder
                                 (6.4)
                        Cylinder Removal
                        Activities Puncture
                        Cylinder
                                 (7.2)
                        Cylinder Removal
                        Activities Cause Large
                        Rupture in Cylinder
                                 (8.0)
 Rupture or Removal of
 Lid From Drum Containing
 Pyrophoric Material
                              Evaporation From Small
                              Pools
                               Flashing Liquid Jets From
                               Small Containers
                               Vapor Jets From Small
                               Containers
                               Small Puff Releases
Buoyant Plumes
                                               Figure 2-2, Page 1
                                  Some Possi9>le Releases ;>! Stajperfiind Sites

-------
           PESCRIPTIONOF
            RELEASE CASE
VISUALIZATION
     APPLICABILITY
  (SECTION OF REPORT)
  PHENOMENA/ISSUES
      ADDRESSED
     7.   Vapor Releases From Soil
 VAPOR
 CLOUD/ /
                                            f
                                               /SOIL
  Activities at Superfund
  Site Disturb Contaminated
  Soil
         (10.0)
• Calculation of Source
  Terms From Disturbed
  Soil
oo
     8.   Stack Release From
         Incinerator
                                          HI  a
           lUOYANT
           VAPOR
           CLOUD
• Incineration Activities
          (9.2)
  Buoyant Plumes, Elevated
  Releases
     9.   Burning of Spilled
         Liquid Pools
                        Spills of Flammable
                        Waste From Storage in a
                        Vessel or Drums
                               (9.3)
                               Buoyant Plumes

                               Burning of Pools
     10.  Burning Tires
                        Burning Tires in Waste
                        Dump or Storage
                        Building
                               (9.4)
                               Buoyant Plumes

                               Rate of Release of
                               Heal from Burning
                               Tires
                                                  Figure 2-2, Page 2
                                      Some PosssuiSe ReBeases as Superfund Sines

-------
describe how to calculate the characteristics of the source terms. Figure 2-2 continues with
some additional accident scenarios that might specifically be of interest at a Superfund site.

2.2.1   Vessel Containing a Liquid

Consider the simple case of a vessel that contains a liquid  that must be transferred to
another vessel and then shipped off site.  The first question to address is whether the liquid
is at or below its atmospheric boiling point, or whether it is a gas liquified under pressure.

2.2.1.1       Liquid Below its Boiling Point

Suppose that there is a liquid  in a vessel with  a boiling point that exceeds the ambient
temperature.  Two cases  are considered: a) the driving pressure is  insufficient to cause
aerosolization and b) the driving pressure causes aerosolization.

A)     Spillage with no  Aerosolization (Case 1A, Figure 2-1)

In this case, the liquid spills onto the ground, where it may be confined in a dike or it may
spread over an unconfined area. The principal issue the analyst has to address here is the
rate of evaporation.  This depends, among others, on such factors as the wind speed, the
vapor pressure, the area of the dike, the  temperature of the ground and the  degree of
insolation.

B)     Spillage with Aerosolization (Case IB,  Figure 2-1)

If the pressure in the vessel is high enough, the jet of liquid  may emerge with a high enough
velocity for hydrodynamic forces to cause aerosolization.   Thus,  there will be an  initial
airborne source term consisting of liquid droplets, followed by evaporation from the pool
that remains behind. This scenario will provide  an introduction to circumstances in which
aerosolization could be  caused by the presence of a high driving pressure, such as might be
present if, for example, nitrogen pressure is being used to effect the transfer from the vessel.
This case will also serve  to introduce techniques for calculating the liquid release rate when
driven by high pressures in addition to the pressure arising from its own weight (the static
head).
                                         2-9

-------
2.2.1.2       Refrigerated Liquid in a Vessel (Case 2, Figure 2-1)

The next case to consider is that of a refrigerated liquid at atmospheric pressure and a
temperature that is lower than ambient, such as refrigerated chlorine or ammonia.

This liquid may  also be spilled with or  without  aerosolization  into a  dike  or into  an
unconfined area.  The principal difference between this and the foregoing is that the rate
of evaporation of the pool is driven by heat conducted from the surface on which the pool
is lying, rather than by mass transfer into the stream of air flowing above the pool. Thus,
this scenario will introduce the analyst to the modeling required to assess the consequences
of spillages of cryogenic materials.

2.2.1.3       Gas Liquified Under Pressure (Case 3, Figure 2-1)

Case 3 illustrates some of the potential ways in which liquid might be released from a vessel
containing a gas liquified under pressure.  For example, Case 3A  shows the outcome of a
puncture in the side of the vessel.  The liquid will be driven out by a combination of the
vapor pressure and the static head.  The key phenomenon to be borne in mind here is that
of flashing. Part of the liquid will immediately flash to vapor.  Of the remainder, some will
remain airborne as fine  liquid droplets and some will fall onto the ground, the relative
proportions being dependent on such factors as superheat. Thus, this scenario will introduce
the  analyst  to  the key phenomena associated  with  flashing, liquid  jets,  including
aerosolization, and the need to take into account momentum effects.

Cases 3B apply when the rupture takes place in a long  length of  pipework.  In that case,
there will likely be some flashing in the pipe. This strongly influences the predicted rate of
release. There will be further flashing and aerosolization once the  release emerges into the
atmosphere.

Case 3C applies when the emerging jet is directed at a  surface.  In this case, there is the
potential for  droplet recovery, leading to the formation of a pool on the ground, with only
a fraction of  the initially formed aerosol remaining airborne.

If there sh'ould be catastrophic failure of the vessel (Case 3D), there will be immediate bulk
boiling and the  formation of vapor and aerosol.   The resultant expansion will be highly
                                        2-10

-------
turbulent, so that an initial puff will be formed that will likely contain aerosol, vapor and
air. Thus, this source term can be used to illustrate how the contingency analyst sets about
modeling puff releases.

2.2.1.4       Releases from the Vapor Space of a Pressurized Vessel

One potential case is that of an escape of gas from a small hole in the vapor space of a
pressurized vessel (Case 4A, Figure 2-1).  In this context, the definition of a small hole  is
that R = a/A « 1, where a is the area of the hole and A is the area of the liquid surface.
For most storage vessels, the shearing off of a valve in  the vapor space leads to an orifice
with R  «1. In this case, vapor will emerge from the orifice.  The flow  is likely to be
choked (except for pressures of the order of only 1-2 atmospheres) - that is, the vapor
emerges at the speed of sound. Thus,  this case will enable the contingency modeler to
consider jet  releases of vapor.  It has application also  to the case of releases from relief
valves and to releases of lower velocity from stacks. In addition, the vapors in question may
be buoyant or may be denser than air.

If there is a catastrophic failure in the vapor space (Case 4B, Figure 2-1), the consequences
will be the same as for a catastrophic release in the liquid space (Case 3D, Figure 2-1).

There exists  a category of intermediate failures such as that illustrated in Case 4C on Figure
2-1, in which the hole can be categorized as neither small nor large. Considerable relevant-
work has been  done by  the  AIChE's  DIERS (Design  Institute  for  Emergency Relief
Systems) program.  Any one of up to four flow regimes through the orifice  is possible.

       •     Pure vapor - already discussed for a small hole

       •     Bubbly flow regime  in which the liquid  phase is continuous with  discrete
             bubbles

       •     Churn turbulent flow regime in which the  liquid phase is  continuous with
             coalesced vapor regions of increasing size; or

       •     Droplet flow regime in which the vapor phase is continuous with  discrete
             liquid droplets
                                        2-11

-------
The transition between these various flow regimes occurs with increasing volumetric vapor
flux and is also influenced by fluid characteristics and by the presence of impurities. It is
unlikely that, if such a scenario is envisaged, a typical on-scene coordinator or emergency
responder will have access to modelers or resources that will enable a quick atmospheric
dispersion calculation to be done.  Hence, in this case, expert assistance should be sought,
and this case will not be discussed in detail in the present  work.

2.2.1.5       Spillage onto Water (Case 5, Figure 2-1)

For completeness, the case of spillage onto water will be discussed, which will give guidance
to the analyst on how to calculate rates of evaporation when  the  important driving  forces
are the rate  of transfer of heat across the water/substance  interface and/or the generation
of heat if the substance is soluble in water.

2.2.2  Accident Scenarios Specifically Relevant to Superfund Sites

The accident scenarios discussed in Figure 2-1 are intended to give the reader guidance on
how to  tackle many of  the different kinds of releases  that  might be  encountered at
Superfund and other sites.  Several accident scenarios that are  specific to Superfund sites
are displayed on Figure 2-2.
2.2.2.1       Small Containers

The small containers most likely to be encountered on Superfund Sites are cylinders and
drums.  These can generally be  treated as small vessels, with the same phenomena being
important as in the case of large vessels, but with the small inventory being the principal
difference. Continuing the numbering scheme that was used on Figure 2-1, it is suggested
that the following scenarios will  be useful:

Case 6A      A drum containing a liquid is punctured and the contents fall to the ground
             and evaporate

Case 6B      A cylinder containing a gas liquified under pressure is punctured and the
             contents escape  through a small hole
                                        2-12

-------
Case 6C     A  cylinder containing a compressed gas ruptures  and the contents  escape
             through a small hole

Case 6D     A  cylinder containing compressed  gas or a gas  liquified under pressure
             ruptures catastrophically and the contents escape instantaneously, forming a
             puff

Case 6E     A drum catches fire and the contents burn.

2.222       Ground Disturbances

Case 7       On a Superfund site,  there is the potential for pockets of gas to be uncovered
             or for materials to be released from disturbed soils.

2.2.2.3       Stack Releases

Case 8       This case will address how to model  stack  releases, with the case of an
             incinerator being the chosen example.

222A       Burning Pools

Case 9       A flammable liquid containing toxic materials is spilled onto the ground and
             catches fire. This will serve as an introduction to the calculation of emission
             rates from burning pools.

2.2.2.5       Burning Tires

Case 10
             A pile of tires catches fire and toxic materials are emitted to the atmosphere.
                                        2-13

-------

-------
3.0   COMMUNICATIONS AND HOW TO USE THIS REPORT

3.1   Concerns Expressed During Telephone Survey

The issues raised by the EPA personnel who were surveyed by telephone were far from
uniform, but there does seem to be an underlying thread of concern running through what
was said, namely how can miscommunication between the person requesting the analysis
(the requestor) and  the person performing  the analysis (the analyst) be avoided?  Put
another way, how can these two individuals communicate in such a way that there is a clear
understanding of the scenario that is to be modeled and the information that the requestor
requires? Other concerns raised included how to ensure that the source term is realistic and
that the model is used in such a way as to give credible answers.  The following section is
intended to address these concerns and also to show how this report should be used.

3.2   Communications/Flow Charts

3.2.1  Accident Sequence Definition - General Discussion

The material in this section is intended to help the requestor and the  analyst clearly  to
define the accident scenario to be  modeled. This is done by a series of flow charts (Figure
3-1) which are to be interpreted as follows:

Beginning on Page  1 of Figure 3-1, identify which of the eleven cases displayed in boxes
beneath each other is most applicable to the scenario being investigated. The scenarios are
numbered with reference to Figures 2-1 or 2-2.

Page  1 of Figure 3-1 directs the reader to go to another page, or to go to  another section
of this report, or to  seek expert advice. On subsequent pages of Figure 3-1, the scenarios
are broken down into more classes, after which the reader  is directed to another section in
the report.  Once directed into another section of the  report, the  reader will find  there
advice on  how either to model the scenario in question or will be recommended to seek
expert advice.

If the  reader is advised to seek expert advice,  there are two possible reasons for this: either
the scenario is one that was not chosen as part of the present work, because of the
                                        3-1

-------
          NEED FOR
        CONTINGENCY
          ANALYSIS
         IDENTIFIED
* Refers to Page No. on Figure 3-1


a) Refers to scenario identification
  numbers on Figures 2-1 and 2-2
                                              1 *> SPILLAGE OF
                                                 LIQUID WITH
                                                 ABOVE
                                                 AMBIENT
                                                 BOILING POINT
                                               2.a) SPILLAGE
                                                  OF
                                                  CRYOGENIC
                                                  LIQUID
                                               3.»> PRESSURIZED
                                                  LIQUID
                                                  RELEASES
                                               4.»> PRESSURIZED
                                                  VAPOR
                                                  RELEASES
                                               5 a) SPILLAGES
                                                  ONTO
                                                  WATER
6.a> RELEASES
   FROM DRUMS
   OR
   CYLINDERS
                                               7»> VAPOR
                                                  RELEASES
                                                  FROM SOIL
                                                                          f  SOIL
                                               8 a) STACK
                                                  RELEASES
9.
    BURNING OF
    SPILLED
    FLAMMABLE
    LIQUID
                                               10.  BURNING
                                                  TIRES
                                        GO TO SECTION 5 5
                                        GO TO SECTION 10 0
                                        GOTOSECnON-9.2
                                        GO TO SECTION 9.3
                                        GO TO SECTION 9 4
                                               11. OTHER
                                          Figure 3-1, Page 1
                                 Scenario Identification Flow Chart
                                                  3-2

-------
a)

b)

c)
d)
e)
0

!.»> SPILLAGE OF
LIQUID WITH
AMBIENT
BOILING POINTW
i
age No. on Figure 3- 1
cnario identification
Figures 2- 1 and 2-2
tilled at a temperature that is
mospheric boiling point




TEMPERATURE
OF LIQUID WELL
BELOW BOILING
POINTc)

TEMPERATURE
OF LIQUID JUST
BELOW BOILING
POINTd>










A STATIC
lltAl)
ONLY

B HIGH
PRESSURE
IN VESSEL
OR PIPE

C STATIC
HEAD
ONLY

D HIGH
PRESSURE
IN VESSEL
OR PIPE
— 1™^>
— \^>
U^>
— 1^>>
	 / SEEK \ GOTO |
f EXPERT 1 SECTION 563"=) I
\^AD\ICL^y
/ SEEK \ (jo TO \
	 1 EXPERT SECTION 4 i 9 0 1

More than five degrees centigrade
l^ss than five degrees centigrade
Expert advice needed on heat balances
Expert advice needed on aerosolizaiion
                                                      Figure 3-1, Page 2
                                             Scenario Identification Flow Chart

-------

t


SPILLED LIQUID.
WELL BELOW
BOILING POINT.
STATIC HEAD
ONLY

SPILLED LIQUID.
WELL BELOW
BOILING POINT,
HIGH PRESSURE










A DIKED
AREA

B UNDIKED
AREA

C. DIKED
AREA

D. UNDIKED
AREA




(1)

(0

*   Refers 10 Page No. on Figure 3 -1
a)  Refers to scenario idenlificauon numbers on Figures 1-1 and 2-2
(I)  Key assumption: bigb pressure causes lilUe or no aerosolization
    when temperature of liquid is well below boiling point
                                                                                                                              SCENARIO
                                                                                                                            ! GO TO SECTIONS 5
                                                                                                                            I OR 5 3
J
                                                                                                                             SCENARIO lAb)a>

                                                                                                                           (GOTO SECTION 5 62
                                                                                                                              SCENARIO I Aa)»)

                                                                                                                           I  GO TO SECTIONS S.I
                                                                                                                           !  OR 5 3
                                                                                                                              SCENARIO I Ah)*)

                                                                                                                            GOTO SECTION 5621
                                                           Figure 3-1, Page 3
                                                  Scenario identification Flow Chart

-------
                           SPILLAGE OF
                           CRYOGENIC
                             LIQUID
                                                         STATIC
                                                         HEAD
                                                          HIGH
                                                        PRESSURE
                                                        IN VESSEL
*   Refers to Page No. on Figure 3-1

a)  Refers to scenario identification numbers on Figures 2-1 and 2-2

(1)  Fxpert advice required on aero&oli/ation
                                                                                      DIKED
                                                                                      AREA
                                                                                                                   SCENARIO 2A»)»)
                                                                                   B. UNDIKED
                                                                                       AREA
                                                                                                                   SCENARIO 2Ab)»)
                                                                                     SCENARIO 2B»
                                                                                       CONFINED OR
                                                                                         IINCONFINED POOL
| GO TO SECTION 4 2 «>
                                                         Figure 3-1, Page 4
                                                 Scenario Identification Flow Chart

-------
                                PRESSURIZED
                                  LIQUID
                                 RELEASE
V
O\
     *   Refers to Page No. on Figure 3-1
     a)  Refers to scenario identification
         numbers on Figures 2-1 and 2-2
     (I)  More than ten degrees centigrade
     (2)  Less than ten degrees cenugrade
     (3)  Expen advice needed on aerosolization
                                                        TEMPERATURE
                                                         WELL ABOVE
                                                        BOILING POINT (»
TEMPERATURE JUST
  ABOVE BOILING
      POINT(2)
                                                                                 A  HORIZONTAL
                                                                                        JET
                                                                                 B.  VERTICAL
                                                                                       JET
                                                                                        UP
                          C.   VERTICAL
                                JET
                               DOWN
                          D CATASTROPHIC
                               FAILURE
                          E.    OTHER
                                                                                                             CASIIQUFIED
                                                                                                            UNDER PMiSSUIU-
EXPERT  |  I GO TO SECTION 6 5
ADVICE*  >J
                                                                          SCENARIO 3A/Ba>
                                                                         GO TO SECTioNS 6;
                                                                         THROUGH 64
                                                                           SCENARIO 3B»>
                                                                         (GO TO SECTIONS 6 ijj
                                                                                                                                   SCENARIO 3C

                                                                                                                                   SCENARIO 3D
                                                                                                                                i(iOTOSECTIC)N80|
                                                               Figure 3-1, Page 5
                                                      Scenario Identification Flow Chart

-------
i
—I
VAPOR SPACE
RELEASE - GAS
  LIQUIFIED
   UNDER
  PRESSURE
          *   Refers to Page No. on Figure 3- 1

          a)  Refers to scenario identification
             numbers on Figures 2- 1 and 2-2
                                                                SMALL
                                                            B INTERMEDIATE
                             c   LARGE
                                                                                      )
           2PIIASE
\\\S  f  RELEASE
SSH/^
           .  FOAMING
             OR
             FROllilNCi
                                                                                       A
                                                                                     SCENARIO 4Aa>

                                                                                    iO TOSECT1ON72|
                                                                                                                  I GO TO SECTION 7 3JI
                                                                                   ! GO TO SECTION 8 1
                                                              Figure 3-1, Paged
                                                     Scenario Identification Flow Churl

-------
UJ
oo
                                          CYLINDERS
                                           DRUMS
             *  Refers lo Page No. on Figure 3-1

             a)  Refers lo scenario identification
                numbers on Figures 2-1 and 2-2
                                                                   A  FLASHING
                                                                       LIQUID
                                                                         JET
B.   VAPOR
      JET
                                                                   C. CATASTROPHIC
                                                                       RUPTURE
                                                                   D. SPILLAGE
                                                                    ONTO GROUND
                                                                       LIQUID
                                                                     EVAPORATES
E. PYROPHORIC
   MATERIAL
      FIRE
F. EXTERNALLY
   INITIATED
      FIRE
                                                                                          FLASHING
                                                                                          LIQUID JET
                   c
BUOYANT
VAPt>R
CLOUD
                                                     SCI-NARI()6B»)
                                                   i GO TO SI-CTJON 6 41
                                                                                                                        SCENARIO 6C»>
                                                                                                                      iGOTOSF£T10N72j
                                                                                                                        SCENARIO6D»)
                                                     SCENARIO 6A»>
  SCENARIO6E»)
| GO TO SECTION 9 3 jj
                                                                                                                        SCENARI()6Fa)
                                                              Figure 3-1, Page?
                                                     Scenario Ideiiiificution Flow Chart

-------
limitations on available resources, or the authors judge that the physical phenomena that
need to be addressed are so complex that they are outside the scope of this document.

3.2.2  Accident Sequence Definition - Examples for Superfund Sites

3.22.1       Spillage of Acetone from a Storage Vessel

In order to illustrate how this section should be used by the reader, assume that a situation
has been  identified in which it is known  that there  is a vessel containing acetone. At a
Superfund site, this may be known,  for example, by sampling the contents or by reference
to documentation that has been left  behind by  a  former operator  of the site.  This
corresponds to scenario #1 on page  1 of Figure  (3-1) where a  number of variations is
displayed.

The  boiling point of acetone is ~ 330  K,  which is  typically well above  the ambient
temperature: the temperature on a hot  day might be 305 K.  Therefore,  the reader chooses
the upper branch on p.2 of Figure (3-1) where the main question is whether the acetone is
under its own static head ( i.e. the pressure is due  to its own weight, box A on the figure)
or under an additional pressure from some other source (box B on the figure).

The conditions under which the acetone will  be under its static head only will be prior to
its being transferred from the vessel  in which it is stored.  It will also still be under the static
head in the vessel if it is being withdrawn by a pump. In this case,  the reader is directed to
go to "page 3 of Figure (3-1) where the choice is between spillage  into a diked or undiked
area.  The reader is then directed to Sections  5.1 or 5.3 for spillage into a diked area or to
section 5.6.2 for spillage onto an undiked area. Section 5.1 gives a detailed analysis of how
to calculate the rate of evaporation of acetone spilled  into a diked area and how to prepare
inputs for DEGADIS and SLAB.  Section 5.6.2 gives guidance on how to model a spreading
pool that  is  unconfined.

Returning to page 2 of Figure (3-1), the acetone may be under high  pressure (box B) - for
example, if a nitrogen pad is being used to transfer  it from the vessel. In this  case,  the
reader proceeds through page 3 of Figure (3-1) and on to various sections of this report just
as was illustrated for the case of the "static head only" as described above. On page 3, boxes
C and D differ from boxes A and B only by there being high pressure. This means that the
                                        3-9

-------
liquid will be driven out of the vessel at a higher rate than it would be for the case of the
static head  only.  At this point,  an important assumption is  made, namely that the
temperature of the material is low enough to preclude the formation of small liquid droplets
(aerosol) that arise as the outgoing jet shatters due to the action of hydrodynamic forces.
That is, aerosolization is not expected (see Section 4.2.9 for further discussion). Therefore,
once the pool has formed, there is no difference between the calculations for these cases
(boxes C and D on page 3 of Figure (3-1)) and those for the static head only case and the
reader is directed to the same sections in the report (5.1, 5.3 and 5.6.2).

The methods appropriate for a spill of acetone are  also appropriate for many other
materials likely to be found on a Superfund site. For example, the spillage and subsequent
evaporation of dichlorobenzene or methyl mercaptan will follow the same paths through
Figure (3-1). The methods are also appropriate for many high boiling point materials in
abandoned railcars.

3223,       Liquid Spillage from a Drum

Assume that there is a drum containing a liquid stored at a Superfund  site.  By definition,
the liquid has a vapor pressure  considerably below one atmosphere (i.e. a boiling point
considerably above ambient). It is not particularly important what the liquid is. The drum
may leak because it has been dropped or punctured in removal operations.

On page 1 of Figure (3-1), the closest scenario is #6.  The reader is directed to go to~ page
7, where the lower branch refers to drums and box D is the liquid spillage scenario, which
is discussed  in Section 5.2. There the reader will find that the principal  difference between
the calculations  used  for the drum case and those  used for a large spill of acetone  arise
because the total volume of the spill is much smaller.

3223       Accidents Involving Cylinders

Cylinders are used to supply small amounts of gases that are stored under high pressure,
such as chlorine.  Hydrogen fluoride is another example of a gas that is sometimes supplied
in cylinders. The telephone survey identified HCN in a cylinder as a cause for concern.
Thus,  it is  possible that cylinders will be left behind at Superfund  sites and must  be
removed, emptied or destroyed.
                                        3-10

-------
150 Ib is a typical size for a cylinder.  There are also larger, one ton cylinders such as are
sometimes used to supply chlorine for water treatment or HF for various industrial activities.

On page 1 of Figure 3-1, box #6 identifies releases from drums or cylinders and directs the
reader to p.7 where three alternative cylinder scenarios are shown: A) a flashing, liquid jet;
B) a vapor jet and C) a catastrophic rupture.

Assume that it is known that the cylinder contains chlorine, which will be liquified under
pressure.  Therefore, any of the three scenarios A, B or C is possible depending on whether
the rupture is located in the liquid or vapor space and on how big it is.  For example, if the
cylinder  is upright,  the concern might be that valving at the top could fail.  In this case,
there would be a vapor jet as discussed in Sections 7.2 and 7.3.  If the concern is that the
cylinder  might be dropped, then either case  A (liquid jet, Section  6.4) or case B  (total
rupture,  Sections 8.1 and 8.2) would be appropriate.

3.2.2.4       Fires  at a Superfund Site

There is a potential for fires at a Superfund  site.  Suppose, for example, that there  is a
storage vessel containing  mineral oils  in which polychlorinated biphenyls (PCBs)  are
dissolved. There is a leak from this vessel and the spilled pool ignites. This corresponds
to Box #9 on page 1 of Figure (3-1).  The reader is directed to go to Section 9.3, which
discusses how to calculate the rate of emission of toxic materials from a burning pool.

Fires can also involve drums. Two possible scenarios are shown on page 7 of Figure (3-1).
One consists of the burning of pyrophoric material in  a drum. The other is  similar to a pool
scenario, except that drums stand in the burning pool. These scenarios are  briefly discussed
in Section 9.3.

3.2.2.5       Incineration at a  Superfund Site

Incineration  is  often  an  option  employed   at  Superfund sites  to destroy volatile or
semivolatile organic materials in contaminated waste.  Item #8 on page 1 of Figure (3-1)
directs the reader to Section 9.2, which gives  an example of how to calculate the emission
rate  of toxic materials from an incinerator.
                                        3-11

-------
3.2.2.6       Mechanically Disturbed Soil

Material at a contaminated site must often be excavated prior to remedial treatment. Thus,
excavation itself may cause the release of toxic vapors. Item #7 on page 1 of Figure (3-1)
directs the reader  to  Section 10.0, which  contains a summary of a simple  model  for
predicting the rates of release during excavation activities.

3.2.2.7       Spillages onto Water

Spillages onto water are not highly probable at Superfund sites, but it is possible to envisage
circumstances in which they might occur if operations are being carried out close to a body
of water.  Item #5  on page 1 of Figure (3-1) directs the reader to  Section 5.5, where  the
issues involved in modeling such spills are discussed.

32.3  Accident Sequence Definition - Examples for "Other" Sites

In order to try to  illustrate some  of the types  of scenarios that may occur at "other"
industrial sites, chlorine has  been selected as  a vehicle for discussion.  Chlorine is  often
stored or transported in containers  at ambient temperature, in which case it is a liquid at
high pressure (typically ~  80 - 100 psig).  These containers range from 150 Ib cylinders
through 1 ton cylinders up to railcars that contain 80 tons  and are sometimes themselves
used as temporary storage vessels. As noted above,  150 Ib or one ton cylinders may also be
found at Superfund sites.

In addition,  chlorine is sometimes refrigerated  and stored at ambient pressure  and a
temperature of ~ 239 K (its boiling point). Very large quantities of chlorine can be found
in refrigerated storage,  e.g 10,000 tons.
3.2.3.1       Spillage of Cryogenic Liquid

Assume that a leak occurs in a vessel or its associated pipework in which there is chlorine
refrigerated at ambient pressure. This corresponds to item #2 on page 1 of Figure (3-1).
The reader is directed to page 4 of the figure, where  there  are two options.  Either the
pressure in the vessel is solely that due to the chlorine itself (the static head) or there is
                                        3-12

-------
additional pressure from other sources (e.g. a nitrogen pad for transfer purposes).

For the case of the static head, the liquid may spill into a diked or undiked area. For the
case of a diked area, Section 5.4 shows the reader how to calculate the rate of evaporation.
For the undiked case, Section 5.6.2 discusses the  issues involved if the spill is unconfined.

3.2.3.2       Pressurized Releases -  Flashing Jets

Imagine the case of a vessel containing chlorine under pressure at ambient temperature in
which  leakage occurs from the liquid space.  These releases are characterized by partial
flashing to vapor  and fragmentation of the liquid jet to form an  aerosol.  This  case
corresponds to item #3 on page 1 of Figure (3-1).  The reader  is directed to p.5  of the
figure where the first branch concerns whether the temperature is just above its boiling point
(e.g. < 10°C)  or well above its boiling point. The issue here is  whether the analyst can
assume complete aerosolization or not.  For chlorine the atmospheric boiling point is 239
K and a typical ambient temperature  may be in the range 0  - 30° C (273 - 303 K).
Therefore, for chlorine, the upper branch on page 5 of Figure (3-1) is appropriate, because
the  temperature is in the range 34 - 64 °C  above  the  boiling  point  and  complete
aerosolization is expected.

Five possible cases are considered including three jet  cases,  horizontal, vertical (up) and
vertical (down).  For the horizontal and vertical jet cases, the reader is directed to Sections
6.2 through  6.4 of which Section 6.3 discusses~a flashing jet from a one ton cylinder and
Section 6.4 considers a flashing jet from a 150 Ib cylinder. In addition, Section 8.1 considers
the case in which the rupture in the vessel  is so large that the contents are  essentially lost
instantaneously.

3.2.3.3       Vapor Releases  Driven by  High Pressure

For the case of a release from the vapor space of a chlorine vessel at ambient temperature
and  high pressure, item #4 on page  1 of Figure (3-1) directs the  reader to page 6,  where
there are three options, small, medium and large.

A "small" hole is one in which R = a/A « 1, where a is the  area of the hole and A is the
area of the liquid surface.  For a typical penetration into a vessel (e.g. 1" or 2" in diameter),
                                        3-13

-------
a/A is clearly very small and box A on page 6 of Figure (3-1)  applies.   The reader is
directed to Section 7.2, which discusses high pressure vapor releases.

If the release is  "catastrophic"  so that the  contents  of the  vessel are  lost  almost
instantaneously,  the case is  that of a  puff release,  see Section 8.1.  For releases of
"intermediate" size, complicated phenomena occur that are difficult to calculate, see Section
7.3.

3.2.3.4       Other Materials

The discussion above for chlorine is also appropriate for other materials such as ammonia,
which may also be found liquified under pressure or refrigerated at ambient pressure with
temperatures well  below those of the surrounding atmosphere.

3.2.4  Hydrogen Fluoride - an Interesting Case

Some of the scenarios considered in the present work have to do with the accidental release
of hydrogen fluoride (HF), which has a boiling point of about 20  °C (293 K).  HF  can be
found in circumstances in which its temperature is  well below its boiling point, just below
its boiling point, just above its boiling point or well above its boiling point.

For example, HF storage vessels are sometimes refrigerated so that the temperature  is (say)
  10 °C below its boiling point. In this case,  the methods used  to model it are similar to
those used for acetone (Section 3.2.2.1).  Following through Figure (3-1) leads the  reader
to Section 5.3, which considers a spillage of HF into a diked area.

In some industrial applications, HF may be at a relatively high temperature (say 100 °F) and
any accidental release may take the form of one of the jets shown on page 5 of Figure (3-1).
Specifically, Section 6.2 considers the case of a horizontal flashing jet of HF.

For HF in  storage that is directly  affected  by ambient  conditions, its temperature  can
fluctuate above and below the boiling point during a single day. If the temperature is just
above the boiling point,.the reader will arrive at the lower  branch on page 5 of Figure (3-1).
The reader should seek expert advice because of uncertainty about how much aerosolization
there will be and how much liquid will remain behind as  a pool on the ground.
                                        3-14

-------
If the temperature of the HF is slightly below its boiling point, the reader will arrive at the
lower branch of the flow chart on page 2 of Figure (3-1). There is a need to seek expert
advice because a) there is the potential for partial aerosolization and b) sources of heat that
are neglected in some simpler treatments of evaporation (such as that described in Section
5.1) cannot be neglected in this case.  An explanation of where to look is given in Sections
5.6.3 and 4.2.9.

3.3   Other Issues that Must Be Addressed

There are a number of other issues that the requestor and the analyst must discuss in order
to ensure  that the results are in a useful form for decision making.

      o      What size of leak should be considered?  See  Section 4.3.

      o      What combinations of windspeed  and atmospheric stability category should
             be considered?  For example, the  requestor may wish to know the results in
             a "worst case" weather condition and also .in a  condition that is favorable for
             dispersion (i.e. a condition in which an on-scene-coordinator might choose to
             perform an operation in order to minimize the  consequences should anything
             go wrong). The height  at which the windspeed is measured should be taken
             into account, if known.  If not,  a  height  of 10 m should  be assumed.  See
             Section 4.4.1 for further discussion.

      o      Which wind  directions should be considered -  see Section 4.4.3.

      o      What is the  ambient temperature? What is the temperature of the ground,
             if different?  See Section 4.4.4.

      o      Has the temperature of  the  material  in the  vessel prior to release  been
             specified?

      o      The issue of site  surface roughness is important because, for some heavy
             vapor dispersion models, there is an upper bound on the  surface roughness
             lengths  for which  the  model is valid.   For  example, the authors  of the
             computer models SLAB and DEGADIS recommend against using a surface
                                       3-15

-------
      roughness length higher than 0.1 m. This is an issue that may have to be
      discussed with the model  originator. Assuming a smaller roughness length
      than is the case at the site in question is likely to lead to some overprediction
      of the distances  of concern because smaller surface roughness leads to less
      rapid dilution as the vapor cloud travels downwind.  See Section 4.4.6.

o     Should the site be characterized as rural or urban? This question relates to
      how the material disperses once the vapor cloud has evolved out of the phase
      where initial density or momentum effects are important.   In this passive
      vapor (non-dense-gas) stage,  available heavy vapor models switch  to  a
      traditional  Gaussian approach.   The  parametrizations of the standard
      deviations (i.e. plume width and height as a function of downwind distance
      and atmospheric stability)  differ greatly between rural and urban areas, with
      a great deal more dilution  occurring in the latter case.  The requestor should
      be  aware that several of the  current generation of heavy vapor dispersion
      models do not have an easily applicable urban option for the region of passive
      dispersion, in which case they-will overestimate distances of concern on urban
      sites.

o     Which levels of concern (LOCs) are to be considered? This is a controversial
      issue.  EPA has data bases that are applicable to some of the  Superfund
      scenarios,  such  as   IRIS  and  HEAST.   Emergency  Response Planning
      Guidelines (ERPGs) have been issued  by the American Industrial Hygiene
      Association for some materials. There  is the possibility  of using the IDLH
      (Immediately Dangerous to Life or Health) concentration or some agreed
      fraction  thereof.  It may be necessary  to contact  experts in  the  field.  See
      Section 4.4.7.

o     Is the level of concern to be regarded as a maximum concentration to which
      people might be exposed,  irrespective of the exposure time?  For example,
      ERPGs are currently defined for exposure times of one hour. What should
      be  done for shorter durations  of plume passage?  Should  something like
      Haber's  law be used, which essentially says that the dosage - the product of
      exposure time and average concentration - is what causes  a given health
      effect? This is a controversial issue and  the requestor may choose to be safely
                                 3-16

-------
            conservative  for releases  of short duration by assuming the LOG for  an
            exposure  time of  one hour.   However,  this can  lead to considerable
            overestimates of distances of  concern and there  may be circumstances in
            which the decision maker might wish to  try to understand  how much
            conservatism there  is - e.g. if the conservative answer leads to a decision to
            evacuate a very large number of people.

      o     What output does the requestor wish to see? For  example, are these simply
            distances  of  concern (DOCs)  in each weather condition or  are  complete
            contours of constant LOG desired? At what heights does the requestor wish
            to see the DOCs and LOCs?

      o     What averaging time(s) should be used in the model?  See Section 4.4.8 for
            a discussion.  This is also a subtle question.

      o     What is the maximum downwind distance to which the calculations should
            extend?

3.4   Conclusion

The requestor who uses the above checklist (Section 3.3) and defines the source term using
Figure (3-1) should  avoid most  of  the potential  for miscommunication that  can arise
between requestor and analyst.
                                       3-17

-------
4.0   BRIEF SURVEY OF AVAILABLE DISPERSION MODELS

4.1   Available Models

A review of the capabilities, strengths and limitations of all of the models that are available
for modeling accidental releases of toxic or flammable vapors is beyond the scope of this
report.  Useful references  that review a good deal of relevant material include a recent
comparative study by EPA(5).  The models considered there include AIRTOX, CHARM,
DEGADIS, SLAB, TRACE, FOCUS and SAFEMODE.  Of these, DEGADIS and SLAB
are available in the public domain and are used in this study.

SLAB treats four types of sources: a) an evaporating pool; b) a horizontal jet release; c) a
vertical stack or jet release and d) an instantaneous (puff)  or short duration evaporating
pool release.  Thus, SLAB  is suitable for all of the denser-than-air releases considered in
this report with the exception that it does not simulate time varying releases, in which case
approximations must be made  (such as taking an average rate of release).

DEGADIS is able to simulate  elevated jet releases of dense gases, provided that these are
vertically oriented. It can  also be used to model heavy vapor releases at ground level,
including continuous releases, varying release rates and puff (instantaneous releases).  Thus,
DEGADIS is suitable for all  of the source terms modeled below with the exception of
horizontally oriented jets.

In addition, during the course  of this work an arrangement  was made with the authors of
the proprietary model SAFER, who have analyzed some of the accident scenarios listed in
Section 3. Their results  are included in Appendix F.  SAFER contains a comprehensive
array of models that can address any of the scenarios described in this report. However, in
Appendix F, examples are  given for four scenarios only, a continuous release of flashing
liquid chlorine, a puff release of chlorine, a spillage of refrigerated chlorine onto the ground,
and a spillage of acetone onto the ground.

Finally, HGSYSTEM is  a  new model that was sponsored  by a consortium of industrial
companies with an interest  in hydrogen fluoride (HF).  It is  among the most sophisticated
of publicly available models and was chosen for use with two scenarios, see Appendix G.
The scenarios that HGSYSTEM can model are listed in Table 1 of Appendix G.
                                       4-1

-------
Some of the accident scenarios below do not lead to heavier-than-air releases.  In such
cases, the  use  of models such as TSCREEN or the  Industrial Source Complex (ISC)(3S)
model is discussed.  .

4.2    Why Atmospheric Dispersion Models Give Different Answers

The user of atmospheric dispersion and contingency models soon discovers that, even when
they are nominally solving the same problem, the answers can be quite different.  To explain
in detail all of the differences between SLAB,  DEGADIS,  SAFER and HGSYSTEM is
beyond  the scope of the present report, because the  focus of the present work is on the
preparation of input for the models and not on explanation of the output.  The calculated
differences may  themselves differ from scenario to  scenario.  However, the following
provides general guidance of which the reader ought to be aware:

      i)    There is inherent uncertainty in all contingency models and a difference of a
            factor of two between the results predicted by different models should  not be
            surprising. If such differences arise, it is likely that various parameters in the
            models have also been assigned different values while remaining within the
            bounds of uncertainty.

      ii)    Unless the input to the different models has been carefully prepared  by the
            same analyst, there may be input assumptions that lead  to large differences
            in the answers.  A particular example is the choice of LOG - for example,
           - whether it has been adjusted for exposure time or not.

      iii)   The large differences that may arise because of the selection of an urban or
            a rural dispersion model have already been discussed above.

      iv)    Some models  can  accept  time-varying  release rates (e.g.  DEGADIS,
            HGSYSTEM) while  others cannot (e.g. SLAB).

      v)    There are specific modeling differences.   For example, for jet releases,
            DEGADIS can only  accept a vertical orientation whereas SLAB can  accept
            a vertical or horizontal orientation and HGSYSTEM can  accept  any
            orientation.
                                       4-2

-------
             As another example, HGSYSTEM has some sophisticated new features.  It
             has a new entrainment model for the initial slumping phase of a heavy vapor
             release that allows the user to assume higher surface roughness lengths than
             the 0.1 m  upper bound discussed above  for SLAB and DEGADIS.  The
             higher  roughness leads to extra dilution and predicted concentrations  in
             HGSYSTEM can be  as  much as a  factor  of five  lower than those  in
             DEGADIS. In addition, HGSYSTEM has a state-of-the-art algorithm for the
             longitudinal spreading of transient releases due to the action of wind shear.
             This tends  to make HGSYSTEM predict lower peak concentrations together
             with larger exposure times for  transient or puff releases than do the other
             models.

In summary, the user of  the models should assess differences in answers as  follows:

      a)     If predicted distances of concern lie within a factor of two, the answers are
             within the  expected range of uncertainty.

      b)     Differences in input assumptions should be carefully investigated.

      c)     If neither a) nor b) proves satisfactory, it will be necessary to make a detailed
             study of the differences in the algorithms  contained in  the computer models.
4.3    Choice of Leak Size

One of the most critical factors among those governing the rate of release is the size of the
leak.  The analyst has to decide how large the area is likely to be.  Some examples are as
follows:

There may be a gasket rupture, in which case it is conventional to assume the total loss of
gasket material between two bolts. For  example, for a flange in a pipe of diameter 2" with
4 bolts, the fraction of the circumference that lies between two  bolts has length rd/4 =
1.571". A typical gasket thickness is 1/8" so that A = (1.571)(l/8) = 0.196 in2 - i.e. almost*
equivalent to a hole of diameter 1/2".
                                       4-3

-------
Another rough rule of thumb is that a serious leak from a valve due to loss of valve packing
has an equivalent diameter of about 1/2".

Another potential accident consists of a leak through a pump seal.  Conservatively, the total
loss of a seal will leave an annulus around the pump shaft through  which material can leak.
For example, a pump shaft of diameter 1" with clearance 0.03"  has effective area A =
(T)(l)(0.03) = 0.094 in2, which has an effective diameter of 1/3" (when converted to an
equivalent circular orifice).

Overall, there are a number of potential accident scenarios for  which the orifices have
effective  diameters  in the  1/4" - 1/2" range.  A leak diameter  of 1/2" is a  reasonable
surrogate for these.

The potential for larger leaks can be assessed by looking at vessel penetrations. In many
cases  these are 1" or 2" in diameter. For old vessels at Superfund sites, inspection of the
vessel may reveal a corroded area that might fail and  A can be estimated from the size of
this area.

4.4   Generic Issues in  Atmospheric Dispersion Modeling

Before proceeding to develop atmospheric dispersion  modeling input data for each source
term and each atmospheric dispersion model, it is convenient to discuss the choice of certain
parameters or the treatment of certain phenomena that are  common to most, if not all,
models and are independent of the source terms.  These include:

      o      Atmospheric stability category

      o      Windspeed

      o      Wind direction

      o      Ambient temperatures

      o      Relative humidity
                                        4-4

-------
      o      Surface roughness length

      o      Toxicological levels of concern

      o      Averaging times;  and

      o      Aerosolization.

A good basic introduction to many concepts such as atmospheric stability, surface roughness
length and turbulence can be found in Reference (11).

4.4.1  Atmospheric Stability Categories and Windspeed

4.4.1.1       Background

The atmospheric stability category is  intended to be a rough  indication of the degree of
turbulence in the atmosphere - that is, in the present context,  an indication  of the
effectiveness of the atmosphere in diluting an accidentally released vapor cloud.

Table 4-1 shows a simple scheme that relates stability category A -F to widely observed
quantities such as the degree of insolation, the cloud cover and the windspeed6
Daytime Insolation

Strong
A
A-B
B
C
C

Moderate
A-B
B
B-C
C-D
D

Slight
B
C
C
D
D
Nighttime
Cloudiness4/8
__
E
D
D
D

<3/8
....
F
E
D
D
a)  Fraction of sky covered by clouds
                                        4-5

-------
Atmospheric stability category A corresponds to a hot summer's day when there is a great
deal of convective  turbulence  in  the  atmosphere.  Atmospheric Stability  Category D
generally occurs when there is a brisk windspeed and considerable mechanical turbulence
is present.  Atmospheric Stability Category F corresponds to the case where there is little
turbulence of any kind in the  atmosphere, such as may occur on a still winter's night.

Atmospheric stability is sometimes expressed in terms of the Monin-Obukhov length,
L = (-PsCpu.3T)/(kgH), where:

      o      pa is the atmospheric density
      o      Cp is the  specific heat of dry air at constant pressure
      o      u, is the friction velocity
      o      T is the surface temperature
      o      k is Von Karman's dimensionless constant  ~ 0.4
      o      g is the acceleration due to gravity
      o      H is the vertical flux of sensible heat to the ground.  H is negative (upward)
             in unstable conditions (A-C), zero in  neutral conditions (D) and positive in
             stable conditions (E,F).

Some atmospheric dispersion models ask for values of L.  There is a simple relationship
between  L, the surface roughness length z0 and the atmospheric stability category,  taken
from Table V.I of Reference(4):

           Table 4-2.  Relationship Between Atmospheric Stability Category,
              Surface Roughness Length z0 and Monin-Obukhov Length L
             Atmospheric Stability            Monin-Obukhov
                   Category                   Length (m)
                   A                        L = -11.4z0ai
                   B                        L = -26.0z0an
                   C                        L = -123z00'30
                   D                        oo
                   E                        L = 123z0°-30
                   F                        L = 26.0z0°-17
                                        4-6

-------
4.4.1.2       Choice of Weather Conditions

For scenarios in which heavier-than-air  effects  are not important, it is possible to use
TSCREEN, which  has a built-in range of windspeeds with associated stabilities.  At each
distance  downwind,  the  model identifies the  windspeed  and  atmospheric  stability
combination that gives the maximum predicted concentration for the scenario in question.

For the dense gas models discussed in this report, the worst case weather conditions are not
always obvious and must be determined  by trial and error.  However, it is generally too
onerous a  task to provide runs for the full range of weather conditions  that may  be
encountered. In general, a useful idea of the range of possible  outcomes of  accident
scenarios can be obtained by looking at three stability category /windspeed combinations -
A with a low windspeed(e.g. 1.5 m/s), D with a moderate windspeed  (e.g. 5 m/s) and F with
a low windspeed (e.g. 1.5 m/s). The category A condition will often give the highest ground
level  concentrations for elevated releases where the plume centerline does not return  to
ground level, while the category F condition will often be the worst case for ground level
releases.  The category D condition, by contrast, is more like an average condition.  In the
SLAB and DEGADIS examples given in this report, the case of stability category F with a
windspeed of 1.5 m/s is chosen for illustrative purposes.

The computer models discussed in the present report generally ask the reader to provide
the stability category and windspeed. It is an easy matter to substitute weather conditions
other than those described above if the user wishes to do so.
4.4.2  Height at which Windspeed is Measured

If there is a meteorological station at the site, then the height at which the windspeed is
measured, Z0 should be that of the lowest velocity measurement on the tower.  If there are
no local measurements, Zr should take on a default value of 10 m.  Again, the computer
models described below specifically request the user to provide a value of this parameter,
which can easily be changed.
                                        4-7

-------
4.4.3  Wind Direction

The models discussed in the present work do not require that the direction of the wind be
specified.  After the analysis has been completed,  contours of constant concentration can
then be overlaid on maps of the neighborhood in any direction required by the requestor.
Some more sophisticated models may directly plot the contours on maps, in which case the
specific instructions in the model users' manual should be followed.

4.4.4  Ambient Temperatures

There is no one-to-one correspondence between the ambient atmospheric temperature and
the atmospheric stability category.  If site specific measurements are available showing
average temperatures for each stability category, these should be used. However, reasonable
default values  are as follows: Category A, 305 K; Category D, 288 K; Category F, 278 K.

4.4.5   Relative Humidity

Again, there is no precise correlation between atmospheric stability category and relative
humidity.  In the absence of any information, an RH of 50% could be used. More generally,
Category A is  usually a  dry condition, Category F is often humid and category D can lie
between.  Thus, for default values, the following values of RH can be used:  Category A,
25%; Category D, 50%;  Category F, 75%.  Site specific data should be used if available.

4.4.6  Surface Roughness Length

The  intensity of mechanical turbulence at a site is  dependent on the surface roughness
length, z0.  Table (4-3) gives examples of the surface roughness for  different surfaces(I2).
similar table appears in  Reference (6).

4.4.7  Toxicological Levels of Concern

The  atmospheric dispersion analyses result in contours delineating  "Levels  of Concern"
(LOCs).  For  short term  releases,  one  of  the most recent sources of LOCs are the
Emergency Response Planning Guidelines (ERPGs), which have been made available for
some thirty substances by the American Industrial  Hygiene Association/13'
                                        4-8

-------
                    Table 4-3.  Surface Roughness for Uniformly
                             Distributed Ground Covers
      Surface
Roughness (m)
Ground Cover Height (m)
Ice or smooth mud
 flats
Snow
Sand
Soils
Short grass
Mowed grass
Long grass
Agricultural crops
4 m high buildings
 with lot areas of
 2,000 m2 and a
 50 m2 silhouette
20 m high buildings
 with lot areas of
 8,000 m2 and a
 560 m2 silhouette
100 m  high buildings
 with lot areas of
 20,000 m2 and a
 4,000 m2 silhouette
0.00001

0.00005 - 0.0001
0.0003
0.001 - 0.01
0.003 - 0.01
0.002 - 0.007
0.04 - 0.1
0.04 - 0.2
0.05
0.7
1.0
      0.015 - 0.03
      0.25 - 1.0
      0.40 - 0.2
             4
             20
             100
The ERPG-3 is the maximum-airborne concentration below which it is believed that nearly
all individuals could be exposed for up to one hour without experiencing or developing life-
threatening health effects.  The ERPG-2 is the  maximum airborne concentration below
which it is believed that nearly all individuals could be exposed for up to one hour without
experiencing  or developing irreversible or other serious health effects or symptoms which
could impair an individual's ability to take protective action.  The ERPG-1 is the maximum
airborne concentration below  which it is believed that nearly all  individuals  could  be
exposed for up to one hour without experiencing other than mild, transient adverse human
health effects or without perceiving a clearly defined  objectionable odor.


Note  that the ERPGs are defined  for a period of one hour.  For releases of different
duration, rules such as Haber's law may apply. Haber's law states that a given health effect
                                        4-9

-------
will be observed if the product of exposure time and concentration (i.e. dosage) remains
constant. However, Haber's law may not apply to all chemicals.  In seeking to extrapolate
LOCs to exposure times other than those for which they are defined, the advice of an expert
in toxicology should be sought.

Other possibilities for short term LOCs include the IDLH (Immediately Dangerous to Life
or Health) or  some fraction thereof,  such  as  one tenth.   The National Institute  of
Occupational Safety and Health  (NIOSH) is the organization that publishes IDLH values.
For more conservative estimates of short-term LOCs, the Occupational Safety and Health
Administration's (OSHA's) Permissible Exposure Limit - Time Weighted Average (PEL-
TWA),  or the  American Conference of Government  Industrial  Hygienists' (ACGIH's)
Threshold Limit Value - Time Weighted  Average (TLV - TWA), or some fraction thereof
such as  a hundredth, are sometimes used, see for example Ref. (8).

If cancer or other effects due to long term exposure are a concern, data on LOCs can be
obtained from EPA's Integrated Risk Information System (IRIS) or EPA's Health Effects
Assessment Summary Tables (HEAST), which are updated quarterly.(14>

4.4.8  Averaging Times

The  concept of the averaging time is often one of the most confusing for the user.  Here,
two averaging times are discussed, that for meteorological purposes and the exposure time
for toxicological purposes.

In meteorological modeling, the  averaging time is used to adjust the effective width of the
plume to take account of meander in the  mean wind direction. If a release is of prolonged
duration Tr and the wind direction is nominally unchanging during that period,  the action
of large scale turbulent eddies will cause the  plume to be broader than it would be for a
release  of shorter duration.   One of the simplest approaches (valid for the traditional
Gaussian- approach) is to allow  the width Wp of the plume to be  a function of Tr: Wp is
increased to Wp(Tr/Te)p where Te is the duration  of release in the experiments from which
the values of Wp were derived(15).  A typical value for p is 0.25 and Tc ~ 3 minutes. This
kind of averaging is usually already incorporated into meteorological models.
                                       4-10

-------
With respect to the LOCs, the averaging time is also important, as has been seen above
where the ERPGs are defined for an exposure time of one hour. If the duration of release
is also one hour, then the averaging times for meteorological and toxicological purposes are
approximately the same. However, if the duration of release and/or the duration of cloud
passage at a particular point are much smaller than one hour, then the meteorological and
toxicological averaging times are not the same. In fact, as was discussed above, the problem
then becomes one  of adjusting LOCs for exposure times that  differ from those for which
they were defined: expert advice should be sought.  Further guidance on how to address this
problem is given in later sections of this report for the specific  cases  of SLAB and
DEGADIS inputs.

4.4.9  Aerosolization

The problem of aerosolization has received considerable attention in the past  few years.
When a flashing liquid jet, or a liquid jet at or just below its boiling point  driven by high
pressure, emerge into the  atmosphere, liquid droplets  may become and remain airborne.
This phenomenon  is  known as "aerosolization."  The  fraction of the initial liquid that
aerosolizes is a function both of the degree of superheat and of  the pressure in excess of the
vapor pressure in  the vessel.   The degree of  superheat  is  the difference between the
temperature of the liquid and its atmospheric boiling point.

For the purposes of this report, the examples given in Section 5 fall into the category of
releases that have no associated  aerosolization and the examples given in Section 6 fall into
the category of releases for which the assumption of 100% aerosolization is  appropriate.
If there is doubt, conservative answers can be obtained by assuming 100% aerosolization.
Alternatively, expert advice should be sought.
                                       4-11

-------
5.0   SPILLAGES OF LIQUIDS ONTO SURFACES

5.1   Spillage of Liquid with Above Ambient Boiling Point into a Diked Area - Acetone

This scenario corresponds to #1A of Figure 2-1. This scenario has been chosen because it
is an example of how to model spillages of high boiling point liquids onto the ground with
subsequent evaporation and formation of a vapor cloud. See Section 3.2.2.1 for a further
discussion.

5.1.1   Description of Scenario

A storage vessel stands in a dike. It contains a liquid at atmospheric pressure and ambient
temperature. The liquid has a boiling point that is higher than the ambient temperature so
that its vapor pressure  is below one atmosphere.

Several potential scenarios could lead to a spillage of liquid into the dike. For example, if
the vessel is at a Superfund site, an old and corroded valve could fail when the operator
tries to turn it and the  contents will spill into the diked  area. Gasket failures could occur
or,  if there is a pump, there could  be  a pump seal leakage.  There could be damage if
mechanical equipment  being used nearby accidentally collides with pipework or the vessel
itself. In any event, it  is assumed that there is a leak and the liquid spills into the dike,
where it  lies and evaporates slowly.  The  driving  force for the leak in this  example  is
assumed  to be only a small static head that does not cause aerosolization.  The source term
is then that caused by evaporation  of the pool.  For the purposes of this example, it  is
assumed  that the dike has an area of 400 m2 (20 m x 20 m). This is not an unusual size for
a diked area at a chemical facility.

5.1.2  Rate of Release  of Liquid from Vessel

The rate  of release  of liquid from an orifice is  given by Bernouilli's formula(16>:

             Q =  c.A.pL (2.(p - Pa)/pL  + 2g.h) m                                (5-1)

where       Q is the rate of release (kg/s)
             c is a  constant, normally set to 0.6
                                        5-1

-------
             A is the area of the orifice (m2)
             p is the absolute pressure in the vessel (Pa)
             pa is the atmospheric pressure (l.OlxlO5 Pa)
             g is the acceleration due to gravity (9.81 m/s2)
             h is the static head (m), and
             PL is the liquid  density in the vessel (kg/m3).

The above formula is suitable for calculating the rate of pure liquid flow through a circular
orifice.   There are  variations  on the formula for  a non-circular orifice.   Eq. (5-1) is
consistent with the corresponding formula that is programmed into TSCREEN(3), which in
turn is derived from work sponsored by the EPA and prepared by Spicer(4I) - see Section
5.1.5 below.

As  an example, suppose the static head is h  = 3 m.   The density of liquid acetone is 791
kg/m3. (Note that a good reference for the standard properties of materials, such as density
and specific heat,  is to be found in Reference (46), "Data Compilation - Tables of Properties
of Pure Compounds.") Since there is no additional source of pressure, p = pa.  Assume that
there  is a leak with  an  effective diameter 1/2". Then A  =  (3.142)(l/4)2 =  0.196 in2 =
1.2x10^ m2.  Using Eq. (5-1), the initial rate of release is :

             Q = (0.6)(1.2xlO^)(791){(2)(9.82)(3)}"2
               = 0.437  kg/s.

If the hole has a diameter of 2", the  area A is 4.8X10"4 m2 and Q = 1.75  kg/s.  See Section
4.3  for guidance on selecting hole sizes.

5.1.3   Behavior of Pool on Ground

As  the liquid spills from the vessel, it will spread across the ground.  There are several
possibilities - for  example:

       a)    the spill is so rapid that a surrounding diked  area is covered almost at once
             and the subsequent evaporation takes much longer than  the spill from the
             vessel
                                         5-2

-------
      b)     The pool spreads until the rate of evaporation just equals the rate of spillage.
             This can happen for a small rate of spill within a diked area, or for any rate
             of spill with an undiked area; or

      c)     The contents of a vessel or  drum spill very rapidly and the pool continues to
             spread after the spillage from the vessel has ceased.

For simplicity, the case of rapid coverage  of a diked area with a subsequent relatively long
period of evaporation is considered below.   Discussion of more complex cases is  given
below.

5.1.4  Calculation of Evaporation Rates

There is considerable literature on the  evaporation rate of  liquids  from pools on  the
ground(16).  For slowly evaporating pools,  the following approximation is often used:

             Q0 =  kg.Ap.pvp.M/(R.Tp)                                          (5-2)

where        Q0 is the rate of evaporation (kg/s)
             Ap is the area of the pool (m2)
             Pvp is the vapor pressure (Pa)
             M is  the molecular weight  (kg/kg-mol)
             R  is the gas constant (8314 J/mol/K),  and
             Tp is the temperature of the pool  (K).

The parameter kg is the mass transfer coefficient, given by the formula

             k, =  Dm.N3h/d                                                   (5-3)

where        Dm is  the molecular diffusivity of the vapor in air (m2/s)
             d is the effective pool diameter (m), and
             N3h is  the Sherwood number, given by

             N,h =  0.037 (kJDJlfl((ud/kjai - 15,200)                          (5-4)
                                        5-3

-------
where        k^ is the kinematic viscosity of air (m2/s), and
             u is the windspeed at a height of 10 m (m/s).

Note that this formula neglects other potential sources of heat such as insolation.  The
reader should consult standard texts for the treatment of such sources, see Section 5.6.1.

The formula is applied as follows:

i)      The temperature of the pool is 278 K (assumed to be a typical ambient temperature
       during atmospheric stability category F conditions).

ii)     The vapor pressure pvp of the acetone  over the pool is given by:

             Log10(pvp)  = exp(-0.2185A/Tp + B)
             from which pvp  = l.OSxlO4  Pa.                                    (5-5)

       This  equation is taken from the Handbook of Chemistry and  Physics(I8), which
       contains a table of estimates of A and B for  a great number of substances.

iii)    The area A of the pool is 400 m2  as stated above.

iv)    The molecular weight of acetone is 58.

v)     As stated above, the gas constant  R is 8314 kg/kg-mole

vi)    The effective diameter of the pool d is 20 m, the square root of the area.

vii)    The windspeed u at a height of 10 m  is  1.5 m/s because the calculations here are
       being done in atmospheric stability category  F with a low windspeed of 1.5 m/s

viii)   The  kinematic viscosity of air k,,, is  l.lxlO"5 m2/s from standard texts, see, for
       example, Ref. (17).

ix)    Dm.the molecular diffusivity of acetone in air, is l.lxlO"3 m2/s from Reference  (17).
                                         5-4

-------
x)     Sample Rate of Evaporation in a Windspeed of 1.5 m/s at 278 K

             Nsh = 0.037(1. lxlO-5/1.10xlO-5)I/3.X       (Eq. (5-4))
             X   = ((1.5x20/l.lxlO-5)0'8 - 15200)
             Nsh = 4646

             kg   = (l.lxlO'5).(4646)210 = 2.555xlO'3   (Eq. (5-3))

             Q0  = (2.555xlO-3).(400).(1.05xl04).(58)/((8314).(278))
                = 0.27 kg/s = 38 Ib/min  (from Eq. (5-2)).

Note that, from above, the rate of spillage from a 2" hole is predicted to be 1.48 kg/s.  This
greatly exceeds the predicted rate of evaporation, 0.27 kg/s, showing that the assumption
that the whole diked area is rapidly covered is a reasonable one.

5.1.5  Equations in TSCREEN

TSCREEN^'40 implements the following equtions:

       Q =  6.94xlO'7(l + 0.0043[TP -  273.15]V'73AMpvp/Pvh                     (5-6)

where  Tp is  the temperature of the pool (K)
       u is the windspeed at a height  of 10 m (m/s)
       A is the area of the pool  (m2)
       M is  the molecular weight (kg/kg-mol)
       pvp is the vapor pressure  of the material in the pool, in the present case acetone,
       (Pa), and
       pvh is the vapor pressure of hydrazine (Pa) (i.e. Eq. (5-6), for reasons of calculational
       convenience, calculates the evaporation rates of all materials by comparison with the
       rate for hydrazine).

In the  above equation, [Tp - 273.15]' is taken to be zero if Tp - 273.15 < 0.  pvp is evaluated
using the Clausius - Clapeyron equation:
       Pvp =  101325 exp[(XM/R)[(l/Tb) - (1/TP)]]                               (5-7)

                                         5-5

-------
where X is the latent heat of vaporization of acetone at the normal boiling point (S.OlxlO5
      R is the gas constant (8,314 J/kmol/K), and
      Tb is the boiling point of the material in the pool (330 K for acetone).

The vapor pressure of hydrazine is given by:

      pvh = exp [76.858 - 7245.2/Tp - 8.22 ln(Tp) + 0.0016557TP]                (5-8)

Using Eqs (5-6) through (5-8) gives a predicted rate of evaporation for acetone in a 400 m2
diked area at a temperature of 278 K of 0.65 kg/s, 2.4 times that calculated in Section 5.1.4.
Thus, it appears that the equations in TSCREEN are conservative. The reader wishing to
be consistent with other EPA guidance should use these methods.

5.1.6  Duration of Evaporation

Suppose that there  is a one long ton  - 1,000 kg of acetone left in the vessel. If the spill
were to evaporate at a constant rate of 0.27 kg/s, evaporation would be complete in almost
exactly one hour, assuming that no mitigating measures are available.

In general, the duration of release must be assessed on a case by case basis, taking account
of the quantity released, the rate of evaporation, the availability of mitigating measures and
potential operator actions.

5.1.7   Density of Mixture

As calculated above, the predicted vapor pressure of the acetone at 278 K is about l.OxlO4
N/m2, or about one tenth  of one atmosphere. Therefore, each cubic meter above the pool
contains  0.9 m3 of air and 0.1 m3 of acetone.  The density of air at 278 K is 1.27 kg/m3.
The density  of acetone  at 278  K is 2.549 kg/m3.   Therefore, the overall density  is
(0.1)(2.549)  + (0.9)(1.27)  = 1.398 kg/m3. Thus, the density difference is (1.398 - 1.27)/1.27
- 10%, sufficient to ensure initial denser-than-air behavior (experience shows that denser-
than-air behavior persists down  to density differences as low as 0.1%).  That is why it is
appropriate to use SLAB  or DEGADIS.
                                         5-6

-------
5.1.3  Input to SLAB

The Users' guide to SLAB is to be found in Ref.(19).  The input to SLAB for the acetone
release described above is given in Table (5-1).

Line 1: IDSPL is the spill source type.  IDSPL = 1 for an evaporating pool.

Line 2: NCALC is a numerical substep parameter.  The code developer recommends using
NCALC  = 1.  However, NCALC can be increased if numerical  stability problems are
encountered.

Line 3: WMS is the molecular weight of acetone in kg/mole WMS; = 0.058.

Line 4: CPS is the specific heat at constant pressure, CPS  = 1450 J/kg/K.

Line 5: TBP is the boiling point of acetone, 330 K.

Line 6: CMEDO is the liquid mass fraction in the initial airborne cloud, which in this case
is 0 because only vapor evaporates from the pool on the ground.

Lines 7 - 9: DHE =  5.01E 05 , CPSL  =  2,058 and  RHOSL = 791 are the heat  of
vaporization at 293 K (J/kg), the specific heat of liquid acetone (J/kg/K) and the_ liquid
density (kg/m3) respectively.

Lines 10.11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure of acetone:

            Ps = Pa.exp(SPA - SPB/(T + SPC))                              (5-9)

Ps is the saturated vapor pressure, Pa is the ambient pressure (= LOlxlO5 Pa). Table 2  of
the Users' guide contains some values of SPB and SPC, but not for acetone. When these
values are not known, the user's guide recommends default values of SPB = -1 and SPC =
0.  T is the ambient temperature, 278 K. The code then uses the Clapeyron equation  to
define the value of SPB. When the source is pure vapor, as it is in the present case, and
the temperature of the cloud does not drop below  the boiling point, this default is always
                                       5-7  '

-------
     Table 5-1.  SLAB Input -
Spillage of Acetone into a Diked Area
1
1
0.058
1450.
330.
0.0
5.01E 05
2058.
791.
-1
0.
278.
0.27
400.
3600.
0.
0.0
1800.
l.OE 04
1.
0.
0.
0.
0.1
10.
1.5
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
QTIS
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
                5-8

-------
adequate because neither the saturation pressure nor any of the liquid properties will be
used in the SLAB  calculation.  However, a value for all of these properties must be
specified in the input whether they are used or not.

Lines 12 -  17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature of the released material, 278 K.  QS is the rate of release, 0.27 kg/s.
AS is the effective area of the source, 400 m2. TSD is the duration of release, 3600 s. QTIS
is zero except when IDSPL = 4, when it specifies the  mass in a puff release.  HS.is the
height of release, taken to be Om (close to ground level).

Line 18: TAV is the exposure time.  This is set equal to 1800 seconds to be consistent with
the definition of the LOG, in this case assumed to be the IDLH/10 (note - SLAB does not
explicitly ask for this value) which is valid for  an exposure time of 30 minutes. See p.52 of
the Users' guide for caveats on  the use of this quantity, particularly if TAV > >  TSD.  '

Line 19: XFFM is the maximum downwind extent of the calculation.  It may be necessary
to determine this by trial and error.  A value of 10 km (l.OxlO4 m) should be adequate for
many applications.

Lines 20-23: ZP(I) allows the user to specify up to four heights at which the concentration
is calculated as a function of downwind distance. ZP(1) is taken to be 1.6 m (head height).
The remaining ZP(I) are zero, which means that SLAB only considers the first height.

Lines 24 - 29  allow the user to specify meteorological conditions.  ZO is the surface
roughness length, which is set to 0.1 m for this example. In the User's Guide, the authors
of SLAB caution against using too  high a value  of ZO.  ZA is the  height at which the
windspeed  is measured   (10 m). UA is the windspeed  at height ZA (1.5 m/s). TA is the
ambient temperature (278 K).   RH  is the relative humidity (75%). STAB is the stability
class (6 or  F).  The user can easily change the windspeed or atmospheric stability  class if
he/she wishes  to use different weather conditions..

Line 30: TER  < 0 terminates the run.
                                        5-9

-------
Note that SLAB does not have inputs for levels of concern. That is, SLAB does not provide
contours of constant concentration. The user must postprocess the output to obtain the
distances or contours of concern.

5.1.9  Input to DEGADIS

The DEGADIS Users' manual is  available as Reference (4), which is the EPA-sponsored
version. An additional source of information on DEGADIS is given in Ref.(20), which was
sponsored by the Gas Research Institute. The input to DEGADIS is described in Table
(5-2), which is  the free format input to the module of DEGADIS that models heavy vapor
releases at ground level.  The table was generated by using the interactive mode of input
to DEGADIS.  References (4) and (5) adequately describe this interactive mode. The
easiest way for a user to become familiar  with  DEGADIS  is to experiment with the
interactive mode of input.

Lines 1-4 of the table allow the  user to input four lines of title.

LineJ requests a value of the windspeed UO at a height of ZO m. As in the SLAB example,
these take on  values of  1.5 m/s  and 10 m.  Line 5 also requires a value of the surface
roughness length ZR,  which is 0.1 rn for all of the examples  in this report.  If the user
wishes  to use different weather conditions, he/she can readily  change UO.

Line 6:  ISTAB is the atmospheric stability category, in this case F (6).  The user can easily
change ISTAB to 1,2,3,4 or 5 if stability categories A through E respectively are required.

Line 7:  OODIST is the  distance  downwind from the source at which the DEGADIS
calculations start. OODIST is zero for a ground level  release. AVTIME is the averaging
time for plume meander, set equal to 3600 seconds because the duration of release is 3,600
seconds, see Section 5.1.6.

Line 8: DELTA and BETA are coefficients  in the  expression  for the horizontal standard
deviation  in the passive phase:
                                       5-10

-------
                                            Table 5-2.   DEGADIS Input-
                                       Spillage of Acetoue into • Diked Area
       tONI INI I) SPILL INTO A DIK ED AREA
       STEADY STATE ACETONE SPILL, POOL DIAMETER 10
                1000
.10
 1.30
  6
 00    3600.00
7.74E-02  .90     17.52
 .17    .97    50.00
        278.00     I 00   4.00E-03  75.00
         0  278.00
         0  .00
         0  .00
       ACL
L/,      5808     278.00    2.55
~      1450.    1.00
~      .15    2.00L-05 1.60
        2.30E-05
         00
          4
         00     .70     11.20    0.10
        1600.   .70     11.20    0.10
        3601.    .00    .0000    0.10
        3602.    .00    .0000    0.10
       I I I I I I
       8 MAY-1992 17:24: 7.63
        .70     20.00    7.85
             278.00
             278.00
              278.00
              278.00
                                       1.00
                                       LOO
                                       1.00
                                       1.00
TITLE I                                                 I
TITLE 2                                                 2
TITLE 3                                                 3
TITLE 4                                                 4
UO, ZO, ZR                                              5
ISTAB                                                   6
OODIST, AVTIME                                        7
DELTA, BETA, RML                                      8
SIGX  COEFF, SIGX POW, SIGX  MIN_DIST                9
TAMB, PAMB, HUMID                                    JO
ISOFL, TSURF                                           11
I1ITFL, HTCO                                            12
IWTFL, WTCO                                           13
GAS_NAME                                             14
GAS MW, GAS_TEMP, GAS_RHOE                        15
GAS CPK.GAS  CPP                                     16
GAS_UFL, GAS_LFL, GAS_ZSP                           17
CCLOW                                                  18
GMASSO                                                19
NT      '                                               20
PTIME(I), ET(I),  RIT(I), PWC(I), PTEMP(I), PFRACV(I)       21
PTIME(2), ET(2),  RIT(2), PWC(2), PTEMP(2), PFRACV(2)       22
PTIME(3), ET(3),  RJT(3), PWC(3), PTEMP(3), PFRACV(3)       23
PTIME(4), ET(4),  RIT(4), PWC(4), PTEMP(4), PFRACV(4)       24
CHECK I, CHECK2, AGAIN, CHECK3.CHECK4.CHECK5      25
TINP                                                    26
ESS, SRCLEN, SRCB                                      27

-------
            cry = DELTA.XBETA                                             (5-10)

where the values of DELTA and BETA are given on p.32 of Ref.(20):

      Stability Category         DELTA     BETA
A
B
C
D
E
F
0.423B
0.3 13B
0.210B
0.136B
0.102B
0.0674B
0.9
0.9
0.9
0.9
0.9
0.9
where B = (AVTIME/600)0'2 accounts for plume meander during prolonged releases. Note
that the above values of DELTA are characteristic of dispersion at rural sites.

RML is the Monin-Obukhov length (see Section 4.2.1.) which is either directly calculated by
the interactive data input module  (as is the case here) or input by the user.

Linej?  consists of parameters that describe the spreading of the plume along the wind due
to the  action of atmospheric  turbulence.   The x-direction  dispersion coefficient  SIG-X-MIN-DIST             (5-11)

where X is the distance downwind of the source. This expression is explained on p.(34) of
Ref.(20):
                                      5-12

-------
      Stability Category      SIGX-COEFF  SIGX-POW  SIG-X-MIN-DIST
Unstable (A,B,C)
Neutral (D)
Stable (E,F)
0.02
0.04
0.17
1.22
1.14
0.97
130m
100m
50m
Line 10: TAMB is the ambient temperature, taken to be 278 K, which lies within the likely
range for atmospheric stability category F. PAMB is the ambient pressure, taken to be one
atmosphere ( the results are not particularly sensitive to this parameter).  HUMID consists
of two numbers, the absolute humidity (4.0xlO~3 kg water/kg air) and the relative humidity,
which is set to 75% as being consistent with a cool, relatively humid night such as might
occur under  category F weather conditions.

Line 11: TSURF is the surface temperature, set equal to the air temperature in the present
example, 278 K. ISOFL is a number generated by the DEGADIS interactive input routine
in answer to  the question, is this an isothermal spill?  The answer is no in this case, leading
to ISOFL  =  0.  "No" here  implies that the  effect of heat sources,  such as heating by the
ground or  the heat entering the system as heat is entrained, is taken into account.

Lines 12 and 13: IHTFL, HTCO, IWTFL and WTCO are also generated by the interactive
method of preparing input for DEGADIS.  IHTFL is generated in answer to the question,
is heat transfer included?  Again, the answer  is no in this case  and DEGADIS generates
both IHTFL = 0 and HTCO = 0.  IWTFL is generated in answer to the question, is water
transfer to be included in the  source?  The answer is no so that IWTFL = 0 and WTCO
= 0 as generated by DEGADIS.

Line 14:  GAS-NAME is a three character identifier chosen by the user.

Line 15:  GASMW  is  the molecular weight of  the  gas, 58.08.  GAS-TEMP is the
temperature of the released gas (278 K) and GAS-RHOE is the density of acetone at that
temperature (2.55 kg/m3).
                                       5-13

-------
Line 16: GAS-CPK and GAS-CPP are parameters in DEGADIS' formula for specific heat
at constant pressure.  In the formulation chosen here,  1,450 J/kg/K is the specific heat of
acetone and GAS-CPP = 1 ensures that this value is chosen independent  of temperature.
There are more sophisticated options in DEGADIS, allowing temperature dependence of
the specific heat; these options are not discussed here.

Line 17:  GAS-UFL is the upper concentration of interest.  It is labeled "UFL" because the
model was originally developed to handle flammable vapors. Here, it is arbitrarily set equal
to 0.15.  GAS-LFL corresponds to the lower flammable limit or, for a toxic gas, the LOG.
For acetone, the LOG is taken to be 2.0xlO"5 ppm, which in this case  is the IDLH/10 and
is valid  for an exposure time  of  30 minutes.  Finally, ZSP is the  height  at which the
concentration is measured, taken to be  1.6 m (head height).

Line 18:  CCLOW is the lowest  concentration of interest  (in kg/m3).  The interactive
module  of DEGADIS prints out a suggested value  (in  this case, 2.3xlO"5 kg/m3).  The user
is free to change this value if so desired.  It is  a lower level of concentration below which
the computer model terminates its calculations.

Line 19:  CMASSO ( = 0) is the initial mass over the pool and can be used, for example, for
puff releases.  It is zero for a continuous release.

Line 20:  NT specifies the number of times at which release parameters are required. For
a release at a constant rate-over a finite duration NT=4, see  the discussion of lines 21-24
for an explanation.

Lines 21-24 specify release rate parameters at four times:

PTIME(l)  = 0 (release starts): PTIME(2) = 3,600 s.  If PTIME (2)  = 60,000 s, a steady
state release is assumed). If PTIME(2) < 60,000, a transient release of duration PTIME(2)
is assumed.  PTIME(3) and PTIME(4) are always  1 and 2 seconds respectively more than
PTIME(2)  and are required by DEGADIS to indicate that the  release  has terminated.
Thus, for a steady state release or for a release at a fixed rate for a finite time, NT=4. The
array ET contains the rate of release (0.27 kg/s up to 3,600 s and 0 thereafter).  The array
RIT contains the radius of the pool, taken to be 11.2  m (the radius of a pool of area 400
m2). The array PTEMP contains the pool temperature  (278 K). The array PWC(I) contains
                                       5-14

-------
the initial mole fraction of the released material. In this case, the mole fraction above the
pool is 0.1, see Section 5.1.6.

Lines 25-27 are always generated by the interactive data entry module and are not further
discussed here. Line 27 only appears for the case of a steady state release and gives the
rate of release, source length and source half width (ESS, SRCLEN and SRCB respectively).

5.2   Spillage from Drum

In this case, it is assumed that a drum containing a liquid is punctured and the liquid spills
on the ground.  It is unlikely that there will be a dike to contain the spill. A simplified and
likely conservative approach is to assume that the whole contents of the vessel spill at  once
and spread to a depth h of one centimeter. If the volume released is V m3,  the area A
covered by the spill is then V/0.01 m3 (from the formula V = Ah).

The rate of evaporation Q0 can then be  calculated using Equations (5-1) through (5-3). If
the density of the liquid is pL, the liquid regression rate  is r  = Q0/(/oLA) m/s and the
duration of release td is h/r  = 0.01/r.  Once these quantities have been determined, input
for the computer models can be generated in exactly the same way as is described in Section
5.1.

5.3   Spillage of HF at 60 °F

5.3.1   Description of Scenario

This example is deliberately chosen to illustrate a case in which finding appropriate values
of input parameters is more difficult. The specific example chosen is one relevant to "other"
sites in which anhydrous hydrogen fluoride is in a refrigerated storage vessel: a brine cooling
loop  keeps its temperature at 60 °F, about 10 °F below its boiling  point.   Note  that
anhydrous  hydrogen fluoride contains no water and this  is how it differs from aqueous
hydrogen fluoride, which is a solution of HF in water.  In this hypothetical example, a pump
withdraws  the   HF   from  the   vessel   and  transfers  it   to   a  reactor   where
hydrochlorofluorocarbons (HCFCs) are produced. These are refrigerants that have a small
effect on the ozone layer relative to the chlorofluorocarbons (CFCs), the use of which is to
be phased out by  international agreement.
                                        5-15

-------
The vessel and the associated pipework may leak for a number of reasons  - e.g. gasket
failure, corrosion or puncture by an external agent such as a fork lift truck or a vessel. The
assumption is that  the leak is large enough to fill the diked area quickly.

5.3.2  Calculation of Release Rate

As noted above, the chosen example is one in which there is a leak in a refrigerated storage
vessel containing anhydrous HF.  The HF spills into  a hypothetical diked  area that is
assumed to be 10 mxlO m =  100 m2.  Eqs. (5-2) through (5-4) are applied as follows.

i)     The temperature of the pool is 60 °F as stated above.

ii)    The vapor pressure pvp of the HF over the  pool is given by:

             pvp = exp(-3030/Tp + 21.9) = 9.1xl04-Pa, see  Eq.  (5-5).

iii)    The area A of the pool is 100 m2 as stated above.

iv)    The molecular weight M of HF at 60 °F and atmospheric pressure is about 70 kg/kg-
      mol because of the oligomerization effect (that is, it exists as a mixture of HF, (HF)2,
      (HF)6 and (HF)8).  However, wind tunnel experiments on the evaporation of HF
      from pools  show that there  is very little evidence of heavier-than-air vapor effects.
      It is possible, therefore, that the HF evaporates from the pool as a monomer and
      does not immediately form  (HF)2, (HF)6  and (HF)8.  However,  for the. sake  of
      conservatism in  Eq. (5-2), where the rate of evaporation of HF is proportional to M,
      the molecular weight is taken to  be 70.  This is an assumption that might require
      further consideration at a later date.

v)    As stated above, the gas constant R is 8,314 kg/kg-mol

vi)    The effective diameter of the pool d is 10 m, the square  root of the area.

vii)   The windspeed  u at a height of 10 m is 1.5 m/s  (as above).

viii)  The kinematic viscosity of air ^ is l.lxlO"5 m2/s from standard texts.
                                        5-16

-------
ix)     Dm,  the  molecular diffusivitv of HF in air  is ~  10~5 m2/s, see Appendix H or
      Reference (46).

x)    Sample Rate of Evaporation in a Windspeed  of 1.5 m/s

             Nsh = 0.037(l.lxlO-5/9.84xlO-6)1/3.X             (Eq.(5-4))
             X   = ((1.5xlO/l.lxlO-5)a8 - 15200)

             Nsh = 2520

             kg   = (9.84xlO-6).(2520)/10 = 2.487xlQ-3  (Eq.(5-3))

             Q0 = (2.487xlO-3).(100).(9.1xl04).(70)/((8314).(289))
                = 0.66 kg/s = 87 Ib/min      (from Eq. (5-2)).

5.3.3   Atmospheric Dispersion Model

As  noted above, vapor clouds  from HF evaporating from pools in wind tunnels do not
exhibit denser-than-air behavior. Therefore, it is appropriate to use the Gaussian dispersion
model with a line source  width  of 10 m.   Far downwind,  this line  source becomes
indistinguishable from  a  point source.   TSCREEN  would be a suitable  atmospheric
dispersion model for this case.

5.4   Spillage of Cryogenic Liquid into Diked Area

5.4.1  Choice of Example - Release Rate a Function of Time

This example will generally be  relevant to "other" sites rather than  to superfund sites and
consists of the case of a refrigerated chlorine vessel standing within a diked area. There is
a leak for reasons that are similar  to  those discussed above  for  acetone or  HF.  It  is
assumed that the leak is large enough to fill the diked area quickly. In this case, the  rate
of evaporation is driven by the rate at which heat can be transfered into the liquid from the
surface on which the pool lies.  The  rate of evaporation is given by:
                                        5-17

-------
             Q0 = (k.(Tf - T))/((ast)°-5H,)                                (5-12)


where       ks =  thermal conductivity of the material on which the pool lies (W/m/K)
             Tg =  temperature of the ground (K)
             T =  temperature of the liquid pool (K)
             a, = thermal diffusivity of the soil (m2/s)
             t = time after spill  (s), and
             HL = latent heat of vaporization of spilled material (J/kg).
Thus, this example is one  in  which the release rate varies  as a function of time - it  is
proportional to t'1'2.

Eq.  (5-14)  differs  from Eq.  (5-2)  because two  completely different cases  are  being
considered.  Eq. (5-2) refers to the evaporation of a liquid at ambient temperature where
the vapor pressure is less than one atmosphere.  In that case, the wind governs the rate of
evaporation. Eq. (5-14) is for a liquid with a temperature that is far below ambient. In this
case, evaporation is driven by the rate at which heat is conducted into the pool from the
surface beneath.

i)     The thermal conductivity or, of surfaces such as concrete and soil is generally of the
       order of 2 W/m/K.  For insulating concrete, this value can be considerably smaller.

ii)     The temperature of the ground Tg is 278 K as in all of the examples in this report.

iii)    The temperature of the liquid pool T is assumed to be that of chlorine at its boiling
       point, 239 K. Therefore, Tg -T = 39 K.

iv)    The typical value of the thermal  diffusivitv of the underlying surface is 10~6 nr/s.

v)     The latent heat of vaporization of chlorine  is 2.88x10* J/kg.

vi)    The coefficient of f0-3 in Eq. (5-14) is (2)(39)/{(10-6)°'5(2.88xl05). =  16.62. Thus, the
       predicted rate of evaporation after 300 seconds is 0.96 kg/s.  The rate of release as
                                        5-18

-------
           Table 5-4.   SLAB Input -
Spillage of Refrigerated Chlorine into a Diked Area
1
1
0.07091
498.1
239.1
0.
287840.
926.3
1574.
1978.34
-27.01
239.1
0.783
100.
1800.
0.
0.
3600.
l.OOE + 04
1.
0.
0.
0.
0.1
10.
1.5
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
cms
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
                      5-20

-------
Lines 10.11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure of chlorine:

             Ps = Pa.exp(SPA - SPB/(T + SPC))

P3 is the saturated vapor pressure, Pa is the ambient pressure (= l.OlxlO5 Pa), a value for
SPA is specified in the code and the values of SPB (1978.34) and SPC (27.01) are given in
the SLAB Users' Guide. T is the ambient temperature (K).

Lines 12 - 17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature of the released material, 239 K. QS is the rate of release. Note that
SLAB  cannot accept a time varying release rate.  In  principal, SLAB could be  rerun for
differing release rates over short periods of time and then the results postprocessed to give
average concentrations over the whole duration of release.  This has not been attempted
here, but rather an average value  QS = 0.783 kg/s over  1,800 s has been chosen.  AS is the
effective area of the source, 100  m2.  TSD is the duration of release, 1800 s.  HS is the
height  of release, 0 m. As described before, QTIS  =  0  except for a puff release.

Line 18: TAV is the exposure time.  This  is set equal to the exposure time corresponding
to the LOC, in this case  3600 seconds because ERPGs are available for chlorine.

5.4.3  Input to DEGADIS

The input to DEGADIS is given in Table (5-5). The following text explains only items that
were not previously explained.

Line 12: IHTFL is one because heat transfer is included.

Lines 14 -  18: give the properties  of the gas: CL2 (GAS-NAME), 70 (GAS-MW), 238.7 K
(the assumed temperature of the vapor above the pool), 3.67 kg/m3 (the density of chlorine
vapor at this temperature, calculated using the perfect gas  law).  The specific heat  at
constant pressure is 484 J/kg/K (GAS-CPK) and choosing GAS-CPP to be unity ensures
that CPK remains constant at all temperatures. The GAS-UFL and the GAS-LFL are the
LOCs,  in this case the ERPG-3 for C12, 20 ppm and the ERPG-1, 3 ppm. GAS-ZSP is the
                                      5-21

-------
                                     Table 5-5.   DEGADIS Input -
                                Spillage of Refrigerated Chlorlue into a Diked Area
CRYOGENIC LIQUID SPILL ONTO GROUND
CHLORINE SPILL INTO DIKED CONCRETE AREA
TRANSIENT SIMULATION

1.50 10.00
6
.00 1800.00
7.74E-02 .90 1
.17 .97 50.<
278.00 1.00
0 278.00
1 .00
0 .00
CL2
70.91 238.70
484.20 1.00
2.00E-05 3.00E-06
9.32E-06
.00
9
.00000 16.62
300.00 .9600
600.00 .6790
900.00 .5540
1200.0 .4800
1500.0 .4290
1800.0 .3920
1801.0 .0000
1802.0 .0000
F FFFFF
11 -MAY- 1992 18:33:
.10


7.52
X)
4.00E-03




3.67

1.60



5.60
5.60
5.60
5.60
5.60
5.60
5.60
.000
.000

28. 7





75.00










.00
.00
.00
.00
.00
.00
.00
.00
.00


















238.70
238.70
238.70
238.70
238.70
238.70
238.70
238.70
238.70


                                          1.00
                                          1.00
                                          1.00
                                          1.00
                                          1.00
                                          1.00
                                          1.00
                                          1.00
                                          1.00
TITLE I                                                 1
TITLE 2                                                 2
TITLE 3                                                 3
TITLE 4                                                 4
UO, ZO, ZR                                              5
ISTAB                                                  6
OODIST, AVTIME                                        7
DELTA, BETA, RML                                      8
SIGX_COEFF, SIGX POW, SIGX_MIN_DIST                9
TAMB, PAMB, HUMID                                    10
ISOFL, TSURF                                           11
IHTFL, HTCO                                            12
IWTFL. WTCO                                           13
GAS_NAME                                             14
GAS_MW, GAS_TEMP, GAS_RHOE                        15
GAS_CPK, GAS_CPP                                     16
GASJUFL, GAS_LFL, GAS_ZSP                           17
CCLOW                                                 18
GMASSO                                                19
NT                                                     20
PTIME(I), ET(I), RIT(I), PWC(1), PTEMP(l), PFRACV(I)       21
PTIME(2), ET(2), RIT(2), PWC(2), PTEMP(2), PFRACV(2)       22
PTIME(3), ET(3), RIT(3), PWC(3), PTEMP(3), PFRACV(3)       23
PTIME(4), ET(4), RIT(4), PWC(4), PTEMP(4), PFRACV(4)       24
PT1ME(5), ET(5), R1T(5), PWC(5), PTEMP(5), PFRACV(5)       25
PTIME(6), ET(6), RIT(6), PWC(6), PTEMP(6), PFRACV(6)       26
PTIME(7), ET(7), R1T(7), PWC(7), PTEMP(7), PFRACV(7)       27
PTIME(8), ET(8), RIT(8), PWC(8), PTEMP(8), PFRACV(8)       28
PTIME(9), ET(9), RIT(9), PWC(9), PTEMP(9), PFRACV(9)       29
CHECK I, CHECK2, AGAIN, CHECK3.CHECK4.CHECK.5      30
TINP                                                    31

-------
height  at  which concentrations are  calculated, i.e. head  height  (1.6 m).   CCLOW is
calculated by DEGADIS as a recommended input value.

Lines 20 - 29 specify the release rates ET(I) for the times PTIME(I) from Table 5-3.  The
radius is taken to be 5.6 m for a diked area of 100  m2. PWC, PTEMP and PFRACV have
already been explained.

5.5   Spillages onto Water

This is  one of the source terms for which the available resources did not permit the detailed
development of an example. The models used for calculating the rates of evaporation of
chemicals onto water differ somewhat from those used for releases onto land. The principal
differences arise in the behavior of the chemical immediately after release and in the rate
of evaporation. For example, a cryogenic liquid, such as LNG, released onto water boils
rapidly, at almost  a  constant rate.  Other materials may  interact with the water  (e.g.
anhydrous ammonia).   In such cases, chemical specific models need to be developed.

A review of models for evaporation following spillages onto water has been given by Raj(22).
The specific example of ammonia is covered in Reference (23).

5.6   Additional Considerations

There  are additional items to be considered that are not part of the  above discussion of
specific source  terms.  They generally cover areas where the requestor and /or analyst
should seek  expert advice.

5.6.1  Heat Sources

The evaporation equation (5-2) contains the assumption that the rate of evaporation is slow
and that the temperature of the pool remains constant. This  will certainly not always be so
and a comprehensive model should take account of the  following sources of heat:

      o      the rate of convective  heat transfer from the air to the pool in cases where the
             temperatures of the pool and the air differ.  The rate is proportional to the
             temperature difference
                                       5-23

-------
      o      the  rate of radiative heat transfer from the air to the pool.  This effect is
             proportional to the difference in the fourth power of the temperatures.

      o      solar radiation

      o      transfer of heat from the dike walls

      o      heat lost due to evaporation of the pool; and

      o      heat added as additional liquid enters the pool from the vessel.

A clear and concise discussion of these heat sources may be found in Ref. (24).

5.6.2  Spreading Pool

The examples given above are all for  a confined pool.  If there is no diked area, the pool
will spread and the evaporation rate becomes a function of time.  The reader is referred to
Reference  (22) for further details.  An earlier, "classical" reference is that by Shaw and
Briscoe.(25)

5.6.3  Advanced Modeling

The modeling described above is considerably simplified. As noted above, it does not cover
heat sources such as solar radiation or spreading pools.  In general, the vaporization rate
is a function of the wind-speed, the pool temperature, the pool size, the properties of the
liquid under consideration and the thermal properties of the environment. The problem
simplifies in two limits: the "boiling cryogen"  limit where the vaporization rate is entirely
controlled  by  the heat flux into the pool and the "evaporation" limit  where  the pool is
essentially at ambient temperature and mass transfer is controlled by the rate at which the
air streaming over the pool can remove the vapor.  In the first case (see Section 5.4) the
vaporization rate is independent of the windspeed. In the second case (see Sections 5.1 and
5.3) it depends strongly on the windspeed.

In general,  and in particular for substances which boil close to ambient temperature, neither
of these two limits is appropriate. In that case, there are a number of coupled linear
                                        5-24

-------
differential equations that need to be solved and numerical modeling is required. The issues
involved are discussed in Reference (26).
                                        5-25

-------
6.0    JETS CONTAINING LIQUID AND VAPOR

These releases  correspond  to  the Scenarios labeled  #3  in  Figure 2-1.   They  are
characterized by a liquid jet with subsequent flashing to vapor and aerosol. In addition, this
category of releases contains vapor jets  in which there is condensation  accompanying
expansion from the pressure in the reservoir to atmospheric pressure, see Section (7.1). The
cases considered here  differ  from the pure vapor releases in Sections 5 or 7 in that the
airborne release consists of both vapor and liquid droplets.

Section (6.1) contains  a  detailed theoretical discussion of various  release formulae. This
discussion follows closely the methods developed by Spicer <36), which were sponsored by the
EPA and have been incorporated into TSCREEN.  Sections (6.2) through (6.4) then contain
specific examples, including, first, consideration of an orifice of diameter 1/2" in a vessel
containing HF at elevated temperature and pressure (Section (6.2)). This is a release that
is very much characteristic of those to be expected at  "other" industrial sites.  However, it
is useful to begin with it because, by carefully considering it, the reader will become familiar
with many of the issues  that must be addressed if  such scenarios are to be modeled
realistically.

The HF example is followed by a liquid  release from a cylinder containing  one  ton of
chlorine at ambient temperature (Section(6.3)). Such one ton cylinders are commonly found
at such places as water treatment facilities.   Next,  the  specific  example of a cylinder
containing 150 Ib of chlorine is examined as being particularly pertinent to the sort of case
that might need to be considered at a Superfund site  (Section(6.4)).

6.1    Emission Rate Formulae  - Theory

This section is intended to describe how to model the atmospheric dispersion of momentum
jets  that contain a  mixture of vapor and fine liquid droplets (two-phase  leaks).   The
following subsections  show how to estimate the release rate for three scenarios:

       (i)     a gas  release that partly condenses when depressurized

       (ii)    a saturated liquid release from pressurized storage; and

       (iii)   a subcooled liquid  release from pressurized storage.

-------
For all of the cases considered in this section, the following assumptions apply: a) any vapor
phase obeys the perfect gas law;  b) the pressure and temperature of the reservoir remain
essentially constant;   c)   for two-phase flows, all released  liquid is assumed to travel
downwind with negligible rainout near the source.

6.1.1         Gas which Partially Condenses on Depressurization

The  procedure described herein applies to a continuous release of gas that partially
condenses during depressurization.  The screening  procedures for determining whether
condensation occurs are described in Section (7.1)  The required input information  is as
follows:

             AO    area of reservoir hole or opening (m2)
             A,    flow area representing reservoir conditions (m2). In the case of a leak
                   from a tank, A, -» oo and j8 ( = Ag/A,) -» 0. In the case of a small leak
                   from the wall of a pipe, AI is  the area of the pipe's cross-section
            •Cp    specific heat of gas at constant pressure (J/kg/K)
             D0    equivalent diameter of hole opening (D0 = (A^/ir)112 )(m)
             Dp    pipe diameter (as appropriate) (m)
             Lp    pipe length (as  appropriate) (m)
             M    gas molecular weight (kg/kg-mol)
             pa     ambient pressure  (N/m2 (Pa))
             pv     vapor pressure  (N/m2(Pa))
             p     pressure in reservoir (N/m2(Pa))
             R     gas constant (8314 J/kg-mol/K)
             Tb    boiling point of released material (K)
             Tr    release temperature (K)
             T!    temperature in reservoir (K)
             7     Cp/Cv, where Cv is the specific heat of the gas at constant volume
             hL    heat of vaporization at the atmospheric boiling point (J/kg)
             pL    liquid density (kg/m3)

The calculations of emission characteristics should proceed as follows:
                                        6-2

-------
A.    Calculate Choked Flow Pressure p« from:


            I: =  <_l_)T/cr-i>
            PI     Y+l

      where p. is the choked flow pressure.

B.    Choked Flow: for choked flow, estimate the discharge temperature Tr, the discharge
      density pr and the emission rate Q, using the following procedures (for subcritical
      flow, go to C below):

      a)    For pure components, estimate T. (the temperature which corresponds to p«)
      from the Clausius-Clapeyron equation:

          p.  = 101325 exp(^(-L-_L))                                 (6_2)


      For multicomponent mixtures, the value of T» can be estimated using the methods
      of Sandier (38) (not discussed further here).

      b)    Estimate Properties at Choked Flow Conditions:  assuming isentropic conditions,
      estimate the vapor fraction x. at choked flow conditions:
          x,  = 1  * --(M Cpln(T1/T.)-Rln(p1/p.))                       (6-3)
                    MhL

      Using x, from Eq. (6-3), estimate the enthalpy change (H, - H,) and the density p.
      as follows:

          (H,  -H.) = Cp(T,-T.) + hL(l-x.)                                   (6-4)
                    RT       1-x
          P.  = [*.(—-rj) - (	-)]•'                                          (6-5)
                                       6-3

-------
c)    Estimate Emission Rate: the formula for the emission rate Q is based on work
by Lees(40):

     Q=AoP. [2(0.85)(  "")]1*                                     (6-6)
where the factor of 0.85 is included to account for irreversibilities in the flow, based
on work by Levitt(39), and the term 4fLp/Dp accounts for the pressure drop between
the reservoir and the hole opening.  Use f = 0.0045.

d)    Estimate  Discharge  Temperature  and  Density:   estimate the  discharge
temperature Tr after depressurization.  If a condensed phase is present, Tr will be
given by:

    pt»  101325 exp(^(_L-J-))                                      (6-7)

Using this estimate  of T^ the released vapor fraction xr is given by:
    xr = x. +  Cp(T. -Tr)/hL                                             (6-8)

If 1 > ^ ^  0, the above estimate of Tr is valid and the density of the discharged
material is given by:
             RT      1-x  .,
           r  P,M      PL

If xr <  0 or xr > 1, the condensed contaminant phase that was present at the choked
conditions p. and T. is no longer present and the release  now consists of a pure
vapor, without  any condensed phase, in which case the discharge temperature and
density are estimated as follows:

    T=T. +ht(l-x.)/Cp                                            <6-l0)


          p M
                                                                       (6-H)
             r

and XT = 0.

                                  6-4

-------
C.    Subcritical  (non-choked) Flow;    For subcritical  flow, estimate the  gas/liquid
      discharge temperature T^ the discharge density pr and the emission rate E as follows:

      a)    Estimate 7/  For pure components, estimate Tr from  the Clausius-Clapeyron
      equation:
                          h.M  1     i
           Pa = 101325 exp(-L_(JL- _L))                                    (6.12)
                            K   lb    lr

      For multicomponent mixtures, Tr can be determined by the methods of Sandler(37)
      (not discussed here).

      b)    Estimate Properties at Discharge Conditions; Assuming isentropic behavior, the
      vapor fraction at flow discharge conditions is:

           xr  = 1 *  ^(MCpln(T,/Tr) - Rln(p/pJ)                           (6-13)
                     MhL

      Using xr  from Eq. (6-13), the change in enthalpy (H, - Hr) and the density pr can be
      estimated from the  following two equations:

           (H^H)  »Cp(VT,) + hL(l-xr)                                   (6-H)
                   RT      1  x
          Pr = [\(-TJ)  + (-^)r'                                           (6-15)
                   P.M       PL

      c)    Estimate Emission Rate:  based on work presented by Lees(40), the emission
      rate is given by:
          Q - AoPr&(0.8S)()]'fl                                    (6-16)
      where the 0.85 is included to account for irreversibilities in the flow, see work by
      Levitt(39), and the term 4fLp/Dp is included to account for the pressure drop between
      the reservoir and the point of discharge.  Use f = 0.0045.
                                        6-5

-------
D.    Example: Spicer gives an example of a saturated chlorine vapor leak driven by a
      pressure of 2.586xl06 Pa through an orifice of diameter 10.16 cm.

6.1.2         Saturated Liquid from Pressurized Storage

This subsection applies to a continuous release  of a pressurized liquid  stored  under
saturated conditions. The required input information is as follows:

             AO    area of discharge orifice (m2)
             Cp    vapor specific heat at constant pressure (J/kg/K)
             CpL   liquid specific heat (J/kg/K)
             DO    equivalent diameter of hole or opening (= (Ao/ir)1'2) (m)
             Dp    diameter of pipe (as appropriate) (m)
             Le    (empirical)  pipe  length  required  to establish  equilibrium  flow
                   conditions (0.1 m)
             Lp    Length of pipe (as appropriate)
             M    Molecular weight (kg/kg-mol)
             pa    Ambient pressure (Pa)
             p     reservoir pressure (Pa)
             R     gas constant (8314 J/kg-mol/K)
             Tb    atmospheric boiling point temperature (K)
             Tr    discharge temperature (K)
             T!    Temperature in reservoir
             hL    latent heat of vaporization at the atmospheric boiling point (J/kg) .
             pL    liquid density at the normal boiling point (kg/m3)

The procedures for estimating emission characteristics are as follows:

A.    Estimate Discharge Temperature;  the discharge temperature Tr can be estimated
      from the Clausius-Clapeyron equation:
                          htM  1    1
           p, = 101325 exp(-L_(JL-JL))                                     (6-17)
                            K   *b   Ar

      Note  that,  for a  pressure of one atmosphere =   101,325  Pa, Tr  =  Tb.   For
      multicomponent mixtures, the discharge temperature can be calculated  using the

                                        6-6

-------
      methods of Sandler(37) (not discussed here).

B.    Vapor Mass  Fraction  after Depressurization:   the  vapor mass  fraction after
      depressurization is given by:

                                                                            (6-18)
      If 1 S x, > 0, Eq.(7'18) is valid and the reader should proceed to Step C below.
      If xr > 1 or  xr < 0, seek expert advice - the solution is unphysical.

C.    Release Rate: The release rate can be calculated from the formulae of Fauske and
      Epstein(34):  if VLe ^ 1  (where L, = 0.1 m),

                    h.Mp     T,  ..,
           Q  • V-^*^                                            (6-19)
                               pL
where
N =
R(hLMp)2
                                          VL*

      Here, C is the discharge coefficient, which takes on a value of 0.6
      If Lp /Le > 1:
      where F represents the effect of friction in the pipe: F2 = 1/(1 + 4fLp/Dp) with f
      0.0015.
                                       6-7

-------
D.    Discharge Density:  the density after depressurization is:
                   RT      1 -x
           Pr ' W-rrj)  + (	-)]"                                          (6-22)
                   P,M       PL

E.    Example:  Spicer(36) gives an example of a saturated liquid chlorine leak from a
      reservoir at pressure 2.586xl06 Pa through an orifice of diameter 10.16 cm.

6.1.3   Subcooled Liquid from Pressurized Storage

This application is  similar to that described in  Section (7.1.2) except that  the  liquid is
subcooled -that is, the pressure is less than the saturated vapor pressure at the temperature
of storage.  The pressure and temperature in the reservoir are assumed not to change over
the duration of release.  The required input information is as follows:

             AQ    area of discharge  orifice (rn)
             Cp    specific heat of vapor at constant pressure (J/kg/K)
             CpL    specific heat of liquid at constant pressure (J/kg/K)
             D0    equivalent diameter of opening (D0 = (\/ir)m) (m)
             Dp    diameter of pipe (as appropriate) (m)
             Lp     length of pipe (as appropriate) (m)
             M     molecular weight  (kg/kg-mol)
             pa     ambient pressure  (Pa)
             p     pressure in reservoir (Pa)
             R     gas constant (8314 J/kg/K)
             Tb    atmospheric boiling point (K)
             Tr     discharge temperature  (K)
             TI     temperature in reservoir (K)
             hL    latent heat of vaporization at the atmospheric boiling point (J/kg)
             PL     liquid density  at  the atmospheric boiling point (kg/m3)
The calculation is performed as follows:
                                        6-8

-------
A.    Estimate the Discharge Temperature T. from the Clausius-Clapeyron equation:

           pa =  101325 exp(^l (JL -1))                                      (6-23)
                            R   Tb  Tr

Note  that,  for pa =  one atmosphere =  101,325  N/m2, Tr =  Tb.  For multicomponent
mixtures, the method of Sandler(37) (not discussed here) should  be used.

B.    The Vapor Fraction after Depressurization is calculated from the following equation:

                                                                            (6-24)
      If 1  > XT  > 0 then Eq. (7-25) is valid and the reader should proceed to step C. If
      XT > 1 or  xr < 0, seek expert advice, because  the solution is unphysical.

C.    The Emission Rate O  can be calculated using the work of Fauske and Epstein (42):

           Q  =  A [2c2(p-Pl>L  + JlL(^Z)2]I/2                             (6.25)
      where CpL is the specific heat of the liquid at constant pressure and

           plv-  101325 exp(^(JL-JL))                                    (6-26)


      and where c = 0.6 is the discharge coefficient.  F  represents the  effect of pipe
      friction:

           F2 = l/(l+4fL/DJ                                                (6-27)
      Here, f = 0.0015.

D.    The Discharge Density or is given by:
           Pt = l*,        +       ]                                            (6-28)
                   P.M      PL
                                        6-9

-------
E.    Example;  Spicer gives an example  of a  discharge of subcooled  liquid chlorine
      through an orifice of diameter 10.16  cm from a reservoir in which  the pressure is
      2.586xl06 N/m2 and the temperature is 298.15 K.

6.1.4  Flow Chart

A flow chart that illustrates the steps outlined above is given on Figure (6-1).

62    HF at Elevated temperature and Pressure

This case is included as  an example  of a  complex scenario from  a site other than a
Superfund site and, as noted above, it also illustrates some of the issues that need to be
addressed when the thermodynamics is complex.  This case corresponds to scenario #3A
of Figure 2-1.

6.2.1  Description of Scenario

It is assumed that there  is a vessel in which there is HF at an elevated temperature of 100
°F and a pressure of 200 psig.  The possible failures here are much the same as already
discussed in Section 5.1. For the present case, it is assumed that there is a leak of effective
diameter 0.5".

622  Calculation of Release Rate

The rate of release of liquid from an orifice is given by Bernouilli's formula06':

             Q = c.A.pL (2.(p - PJ/pL + 2g.h)  "2                         (6-29)

where       Q is the rate of release (kg/s)
             c is a constant, normally set to 0.6
             A is the area of the orifice (m2)
             p is the absolute pressure in the vessel (Pa)
             pa is the atmospheric pressure  (l.OlxlO5 Pa)
             g is the acceleration due to gravity (9.81 m/s2)
                                        6-10

-------
           Vapor Kclease
           Scenario from,
           Fig. 3-1. p. 6
           Scanano 4A
o\
                                       Continue with
                                       Vapor Release
                                         Secc. 7.1
  Droplet
Formation?
 Sect. 7.1
  Calculate
Choked Row
 Pressure P *
  Eq.(6-l)
                                                                                           Choked How
                                                                                           Source Term
                                                                                          Characteristics
                                                                                          from Eqs (6-2)
                                                                                          through (6-11)
Prepare Source
Term Input for
Vapor Release
   Models
                                                 Subcruical Row
                                                  Source Term
                                                 Characteristics
                                                 from Eqs (6-12)
                                                 through (6-16)
 Prepare Input for
Dispersion Models,
   wtlh Liquid
   Droplets in
   Source Term
          Liquid Droplet
          Release Scenario
          from Fig. 3-1. p. 5
          Scenario 3A/D
                                                Saturated
                                                Subcoolcd
                                 Source Term
                                Characteristics
                                from Eqs (6-17)
                                 through 6-23)
                                 Source Term
                                Characteristics
                                from Eqs (6-24)
                                through (6-29)
                                                         Prepare Input
                                                         for Dispersion
                                                            Models
                                                                          Figure 6-1
                                                    Liquid Droplet Scenarios - Calculation I'loxv Chart

-------
             h is the static head, the difference in level between the orifice and the liquid
             surface (m), and
             pL is the liquid density in the vessel (kg/m3).

The above formula (which is the same as Eq. (5-1)) is suitable  for calculating the  rate of
pure liquid flow through a circular orifice.  It is also the same as Eq. (6-19) for Le/Lp = 0.
There are variations on the  formula for a non circular orifice and for cases where there is
flashing in the pipe upstream of the orifice. See Section 6.1.2 for further discussion.  Values
to be assigned to the variables in Bernouilli's formula are as follows:

      For the case of a half inch orifice, the area A is (3.142)(l/4)2 = 0.196 in2 = 1.2xKT*
      m2.

      The density of liquid HF is - 1,000 kg/m3

      The operating pressure in the vessel is 200 psig = 1.374xl06 Pa.  This is well above
      the  saturated vapor pressure of  HF at the operating temperature  of  100° F. For
      other applications to liquids such as ammonia or chlorine  stored under their own
      vapor pressure, the pressure in the vessel can be obtained from standard curves of
      vapor pressure versus temperature.

      The static head h is generally of the order of a few meters (say 5).   In the present
      case, this makes  a very small contribution to the total rate of release.

Applying the above values to Bernouilli's formula gives a rate of release of 3.7 kg/sec ( ~
60  gpm).   This  rate  will persist for some time before  there is a significant  decrease of
pressure in the vessel.

6.2.3 Other Characteristics of the Source Term

Velocity and Orientation

The velocity of release is given by (Q/A ) = 3.7/1.2x10^/1000 = 30 m/s. For the purposes
of the present example, it is assumed that the release is horizontal in the direction of the
wind (the most conservative assumption).
                                        6-12

-------
Flashing and Vaporization

The liquid will immediately flash and part of it will vaporize.  The fraction that flashes to
vapor is given by the formula(16):

             Fv  = Cp (T, -Tb)/hL                                        (6-30)

where        Cp is the heat capacity of the liquid averaged over the temperature range T,
             to  Tb (J/kg/K)
             Tt  is the initial temperature of the liquid (K)
             Tb is the atmospheric boiling point of the liquid (K)
             hL is the latent heat of vaporization  of the liquid (J/kg), and
             Fv  is the fraction of the liquid that is vaporized.

For the present example, the initial temperature of the liquid is  100 °F (310 K).   The
atmospheric boiling point of HF is 292.7 K.  The  heat of vaporization is 208 kJ/kg.  The
specific heat of  the liquid  is - 2500 J/kg/K.  Substituting these  values into the above
formula gives Fv ~ 0.2.

The balance of the HF (80%) remains  liquid at the boiling point of HF.  From large scale
HF experiments, it is clear that this liquid does not fall to the ground, but remains airborne
as a fine aerosol. Thus, the initial release consists of 80% liquid droplets and 20% vapor
at the atmospheric boiling point of HF.

Duration of Release

Generically, the  duration of the  release  is hard to predict because it depends on site specific
mitigating equipment and procedures.  For example, there may be manual valves that can
be closed in order to isolate the leak. In such a case, the operators  would first have to put
on cumbersome protective clothing.  It might  therefore be  10 - 20 minutes  before the
release  could be terminated.  In the present case, a duration of release of 20  minutes  =
1,200 seconds has been assumed.
                                        6-13

-------
6.2.4  Input for SLAB

The input to SLAB for the HF release described above is given in Table (6-1).

Line 1; IDSPL is the spill source type. IDSPL =2 for a horizontal jet.

Line 2: NCALC is a numerical substep parameter. The code developer recommends using
NCALC  =  1.  However, NCALC can be increased if numerical stability  problems are
encountered.

Line 3: WMS is the molecular weight of HF in kg/mol, WMS  = 0.02.

Line 4: CPS is the specific heat at constant pressure, taken from a table in the SLAB Users'
manual.  CPS =  1450.

Line 5: TBP is the boiling point of HF, 293 K.

Line 6: CMEDO is the liquid mass fraction, which was calculated above to be 0.8.

Lines 7  - 9:DHE  = 373,200, CPSL = 2,528 and RHOSL  ~  1,000 are  the  heat of
vaporization at 293 K (J/kg), the specific heat of liquid HF (J/kg/K) and the liquid density
(kg/m3) respectively. Their values are taken from Table  2 of the  SLAB Users' Guide.

Lines 10.11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure  of HF:

            Ps = Pa.exp(SPA - SPB/(T + SPC))

Ps is the saturated vapor pressure, Pa is the ambient pressure (= l.OlxlO5 N/m2), a value for
SPA is specified in the code and the values of SPB (3404.51) and  SPC( 15.06) are given in
the Users' Guide. T is the ambient temperature (K).

Lines 12 - 17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature of the released material, 293 K.  QS is the rate of release, 3.7 kg/s.
AS is the effective area of the source, 0.018 m2. For a flashing liquid jet, the SLAB User's
                                       6-14

-------
 Table 6-1.    SLAB Input -
Flashing Liquid Release of HF
2
1
0.02
1450.
293.
0.8
3.73E 05
2528.
1000.
3404.51
15.06
293.
3.7
0.018
1200.
0.
5.0
3600.
l.OE 04
1.6
0.
0.
0.
0.1
10.
1.5
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
QTIS
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
            6-15

-------
Manual recommends the formula AS = (pL)(A)/(pm) where pL is the liquid density, A is the
orifice area and pm is the density of the vapor/droplet mixture after flashing. Application
of this formula gives AS  =  0.018 m2.  QTIS = 0 for a release that is not a puff. TSD is the
duration of release, 1200 s. HS is the height of release, 5 m.

Line 18: TAV is the exposure time.  This value  is conservatively  set equal to one  hour
because the LOCs for HF are the ERPG-3 and the ERPG-2, which are valid for exposure
times of one hour.

Line 19: XFFM is the maximum downwind extent of the calculation.  It may be necessary
to determine this by trial and error.  A value of 10 km (1.0xl04m) should be adequate for
many applications.

Lines 20-23: ZP(I) allows the user to specify up to four heights at which the concentration
is calculated as a function  of downwind distance.

Lines 24 - 29 allow the user to  specify meteorological conditions.  ZO  is the surface
roughness length, which is here set to 0.1 m.  In  the User's Guide, the authors  of SLAB
caution against using too high a value of ZO. ZA is the height at which  the windspeed is
measured  (10 m).   UA is the windspeed at height ZA (1.5 m/s).  TA is the  ambient
temperature (278 K). RH  is  the relative  humidity (75%). STAB is the stability class  (6 or
F).  All of these inputs  have  been discussed previously.

Line 30:  TER < 0 terminates the run.

Note that SLAB does not have inputs for levels of concern. That is, SLAB does not provide
contours of constant concentration.  Note also that SLAB does  not have the capability of
accepting  input such as the  ordered triples described below for DEGADIS.  Therefore,
SLAB does not take account of the oligomerization of HF. What SLAB does do  is to take
account of  the  initial  presence  of the  liquid  droplets of  HF and  allows   separate
evaporation/condensation  of both HF and water vapor as the  plume dilutes.  For many
applications,  this will be  quite adequate, especially  if low  levels of concern are  being
considered.
                                       6-16

-------
6.2.5   Input to DEGADIS

DEGADIS can be run in one of two modes, a jet release or a pool simulation release.  The
table below shows the input to the jet release module of DEGADIS.  This differs from the
pool type releases that were described in Section 5 because there are some differences  in
the free format input file. The input for DEGADIS is shown on Table (6-2).

Lines 1-4 of the table allow the user to input four lines of title.  Note that the second line
above states that the release is being modeled as a vertical jet. This is because DEGADIS
does  not allow the  user to simulate a horizontal jet.  For the present example, it is not
expected that there  will be a significant impact on the calculation of contours at  low levels
of concentration such as 20 ppm, see the discussion of GASUL, GASLL and NDEN below.

Line 5 requests a value of the windspeed UO at a height of ZO m.  For the purposes of the
present example, the category F case is chosen with a windspeed UO = 1.5 m/s.  The value
of ZO is 10 m unless specifically stated otherwise.

Line  6 requires a value of the surface roughness length ZR.  As noted in previously, the
value has been limited to 0.0.1 m on the advice of the model developers.

Line 7: INDVEL is an index which determines the method of calculation of the velocity
profile in the atmosphere. If INDVEL = 1, the  computer program accesses default values
of the profile and of the Monin-Obukhov length based on the stability category ISTAB. In
"the present example, the default  mode is chosen  by setting  INDVEL equal  to unity,
choosing the F stability category (ISTAB = 6, see above) and setting the Monin-Obukhov
length RML to 0. RML is then recalculated by the computer program.

Line 8: TAMB is the ambient temperature, taken to be 278 K, which lies within the likely
range for atmospheric stability category F. PAMB is the ambient pressure, taken to be one
atmosphere (the results are not particularly sensitive to this parameter). RELHUM is the
relative humidity, which is set to 75% as being consistent with a cool, relatively humid night
such as might occur under category F weather conditions.
                                       6-17

-------
                  Table 6-2.    DEGADIS Input -
                  Flashing Liquid Release of HF
FLASHING LIQUID HF RELEASE
VERTICAL JET SIMULATION
1.5  10.
0.1
1 6 0.
278. 1. 75.
278.
HF
20
3600.
293,
4.17E-05 1.7E-05  1.6
0 0.0  0.0
10
0.0  0.0 1.27
0.0125  0.01  1.27
0.025 0.02 1.27
0.125  a. 13  1.44
0.25 0.26 1.55
0.32 0.42 1.69
0.59 1.11 2.21
0.74 1.89 2.83
0.88 3.47 4.16
1.0  13.7 13.7
3.7
5.0  0.106
1200.
50.
TITLE 1
TITLE2
TITLES
TITLE4
UO,ZO
ZR
1
2
3
4
5
6
INDVEL, ISTAB, RML  7
TAMB, PAMB, RELHUM 8
TSURF
GASNAM
GASNMW
AVTIME
TEMJET
GASUL, GASLL, ZLL
INDHT, CPK, CPP
NDEN
ORDERED TRIPLES
ERATE
ELEJET, DIAJET
TEND
DISTMX
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
                                  6.-18

-------
Line 9: TSURF is the surface temperature, set equal to the air temperature in the present
example.

Line 10:  GASNAM is a three character identifier chosen by the user.

Line 11:  GASMW is  the molecular weight of the gas.  As has been discussed above, HF
oligomerizes at high concentrations and its molecular weight is effectively much higher than
that of the monomer (20).  However, HF oligomerization is  taken into account in the
discussion of NDEN below and here it is adequate to set the molecular weight equal to 20.

Line  12:  According to  the Users'  Manual, AVTIME is  the averaging  time  for the
calculation of the width of the cloud.  Here it is set equal to the duration of release, 1800
s.

Line  13:  TEMJET is the temperature of the release. As described above, this  is the
atmospheric boiling point of HF, 293 K.

Line 14:  GASUL and GASLL are the upper and lower concentration levels (or "Levels  of
Concern" (LOCs)) for which DEGADIS prints out contours of constant concentration. For
the purposes of  the present example,  the ERPG-3(50 ppm or 4.17xlO"5 kg/m3)  and the
ERPG-2 (20 ppm or 1.7xlO"5 kg/m3) are chosen as the levels of concern.  Note  that the
ERPGs are defined for a period of one hour.

ZLL is the height at which the concentration contours are calculated and is taken to be 1.6
m (head height).

Line  15:  INDHT is  used to include heat transfer  in the DEGADIS computation.  Heat
transfer is not included for INDHT = 0. In the present case, INDHT is set  equal to zero
because heat transfer is already included under the NDEN entry, see below, by introducing
"ordered triples."  In  this  case, CPP and CPK are not used by the computer program and
so  are arbitrarily set  to zero.   For INDHT  = 1, heat transfer is included  and the heat
transfer coefficient is calculated by  DEGADIS. CPP and CPK are then used to calculate
the heat capacity as a function of temperature according to a correlation that is included in
DEGADIS.  If a constant heat capacity is required, set CPP = 0 and CPK to the desired
specific heat at constant pressure (J/kg/K).
                                       6-19

-------
Lines 16 - 26:  NDEN is used to specify the contaminant density profile.  There are three
alternatives for NDEN.  If NDEN = -1, the model treats the contaminant as if it were an
ideal gas with  a molal heat capacity equal to that of air.  Water condensation effects are
ignored.

If NDEN = 0,  the model treats the contaminant as if it were an ideal gas with heat capacity
indicated by CPK and CPP.  The effects of condensation of water are taken into account
as appropriate.  This option is suitable for use with vapors that do not have complex
thermodynamic properties when mixed with air.

If NDEN > 0,  NDEN specifies the number of triples which follow in the next NDEN lines.
The triples are used to specify the contaminant concentration as a function of density, based
on adiabatic  mixing with ambient  air.   The  ordered triples represent  (in  order) the
contaminant mole fraction, the contaminant concentration (kg/m3) and the mixture density
(kg/m3).

Clearly, the ordered triples are  intended to  be  used when the plume has  complex
thermodynamic properties, as is the case with HF. However, these properties then have to
be calculated off line. The two key factors to remember as far as HF is concerned is, first,
that  it oligomerizes.  At high concentrations, it can  be represented  by a  mixture of
monomer, dimer and hexamer  and an effective molecular weight close to that of chlorine.
As air is entrained, it disassociates.  This is an endothermic reaction and  tends to keep the
cloud cold and denser-than air.  Second, HF interacts exothermically with water vapor and
forms fine liquid  droplets.  This effect, in moist air,  tends to raise the temperature and
reduce the density of the cloud.

Models that are available for calculating the thermodynamic properties of air/HF mixtures
include those by Clough et al.(30) and Schotte(31). The NDEN  array in the table above was
calculated by  the authors of this report using  the  model of Clough et al.  Readers are
advised to seek expert help in calculating this NDEN array.

Note that, when the HF mole fraction is unity,  the density is very high, 13.72 kg/m3. This
is because of the presence of liquid droplets, as described above.  When  the mole fraction
is 0.025,which  corresponds to an air/HF mass mixing ratio of about 60, the cloud ceases to
be denser than air because of the exothermic reaction with water vapor.  For even smaller
                                       6-20

-------
mole fractions, the mixture in fact becomes slightly buoyant, but this has been neglected in
the present work.  The slight amount of buoyancy that is generated will not cause the plume
to lift off the ground.

Note also that one recommended way to avoid the labor of having to calculate 10-20
triples for the NDEN array is to provide simply the first and last lines(6).  The code then
extrapolates the cloud density linearly between the two.  As can be seen, for this particular
example of HF in a humid atmosphere, this would lead to an overestimate  of the cloud
density for mole fractions that are less than about 0.025 (equivalently,  mass mixing ratios
in excess of 60  or so). Hence, DEGADIS would preserve the denser than air cloud for
greater mass mixing ratios and greater distances downwind than would  in fact be the case.
This cautionary note  is introduced to encourage the reader to look skeptically at  any
approximations that are made  in order to ascertain whether they are in  fact physically
reasonable.

Line 27:  ERATE is the rate of release, which is calculated above  to be 3.7 kg/s.

Line 28:  ELEJET is the height of release, arbitrarily set to 5 m.  DIAJET is the  diameter
of the jet.  The actual orifice diameter in the present example is only 0.5".  However, the
jet immediately expands because of flashing and its density decreases from  ~ 1,000 kg/m3
to 13.72 kg/m3.  With a mass flow rate  of 3.7 kg/s, the volume flow rate  is 3.7/13.72 = 0.26
m3/s. Making the simplifying assumption that the velocity remains constant at 30 m/s as
flashing takes place (see above), the effective  orifice area increases to 0.26/30  = 0.0089 m2.
This corresponds  to an effective diameter of 0.106 m or about 4".

Line 29:  TEND is when the release ends, i.e. after 20 min or 1,200 seconds.  With TEND
> 0, DEGADIS chooses the transient mode  once the initial jet phase is over.

Line 30:  DISTMAX is the maximum distance between points  in the JETPLU output  (m)
and is arbitrarily set to 50 m.
                                       6-21

-------
6.3     One Ton Cylinder of Chlorine

6.3.1  Release Description

One ton cylinders of chlorine are often used at water treatment facilities, where the system
is usually very simple, namely a pipe off the top of the vessel through a regulator and then
through pipework to a station where the chlorine is mixed with water. The types of releases
that can take place might consist of, for example, a shearing off of the regulator, leaving an
unobstructed flow to the atmosphere, or a leak in the vessel itself, perhaps caused by impact
by an external agent. For the purposes of this example, it is  assumed again that there is a
hole of diameter 3/8" in  the vessel through which  liquid  emerges at a conservatively
assumed  constant  rate until the vessel  is empty.  3/8"  is a typical size for pipework
connected to  one ton chlorine cylinders.

Using Eq. (6-29), the  release rate calculated for Chlorine at 278 K is 1.96  kg/s.  Using
Eq.(6-30), the flash fraction is 0.17 and the liquid aerosol fraction is 0.83. It is assumed that
all of this remains airborne as fine droplets. It is assumed that the resulting aerosol/vapor
mixture is at the atmospheric boiling point of chlorine, 239 K. There is assumed to be 1 ton
~  1,000 kg in the cylinder, so that the duration of release  is 1,000/1.96 -510 seconds.

6.3.2  Input to SLAB

The input for SLAB is  contained in Table (6-3).

Line 1: IDSPL is the spill source type. IDSPL =2 for a horizontal jet.

Line 2: NCALC is a numerical substep parameter. The code developer recommends using
NCALC  = 1.  However, NCALC  can be increased  if numerical stability problems are
encountered.

Line 3: WMS is  the molecular weight of chlorine in kg/mole, WMS = 0.07.

Line 4: CPS is the specific heat at constant pressure, taken from a table in the SLAB Users'
manual.  CPS =498
                                       6-22

-------
                Table 6-3.    SLAB Input -
Flashing Liquid Release of Chlorine from a One Ton Cylinder
2
1
0.07091
498.1
239.1
0.83
287840.
926.3
1574.
1978.34
-27.01
239.1
1.96
0.009
510.
0.
5.
3600.
l.OOE + 04
1.
0.
0.
0.
0.1
10.
1.5
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
QTIS
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
                          6-23

-------
Line 5: TBP is the boiling point of chlorine, 239.1 K

Line 6: CMEDO is the liquid mass fraction, which was calculated above to be 0.83.

Lines 7 - 9:  DHE =  287,840, CPSL  = 926.3 and  RHOSL ~ 1,574 are the heat of
vaporization at 293 K (J/kg), the specific heat of liquid chlorine (J/kg/K) and the liquid
density (kg/m3) respectively.  Their  values are taken from Table 2 of the  SLAB Users'
Guide.

Lines 10.11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure of chlorine:

             P,  = Pa.exp(SPA - SPB/(T + SPC))

Ps is the saturated  vapor pressure, Pa is the ambient pressure (=  1.01E 05 N/m2), a value
for SPA is specified in the code and  the values of SPB (1978) and SPC (-27.01)  are given
in the Users'  Guide.  T is the ambient temperature (K).

Lines12 ^17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature of the released material, 239 K.  QS is the rate of release, 1.96 kg/s.
AS is the effective area of the source, 0.009 m2. TSD is  the  duration of release, 510 s.
QTIS is zero  except in the case of a  puff release. HS is the height of release, 5  m.

Line 18: TAV is  the exposure time.  As recommended in the User's Guide, its value is set
to that which is appropriate for the LOG that is to be considered. In this case, that value
is 1 hr = 3,600 seconds because the  LOCs for chlorine are the  ERPG-2 and the  ERPG-3,
which are defined for an exposure time  of one hour.

Line 19: XFFM is  the maximum downwind extent of the calculation. It may be necessary
to determine  this by trial and error.  A value of 10 km (1.0xl04m) should be adequate for
many applications.

Lines 20-23: ZP(I) allows the user to specify up to four heights at which the concentration
is calculated as a function of downwind  distance.
                                       6-24

-------
Lines 24 - 29 allow the user to  specify meteorological conditions.  ZO is the surface
roughness length, which is set to 0.1 m as in previous examples. ZA is the height at which
the windspeed is measured (10 m). UA is the windspeed at height ZA (1.5 m/s). TA is the
ambient temperature (278 K). RH is  the relative humidity (75%).  STAB is the stability
class (6 or F).

Line 30: TER < 0 terminates the  run.
6.3.3   Input to DEGADIS

The input for DEGADIS is given on Table (6-4).

Lines 1 - 4 of the above table allow the user to input four lines of title.  Note that the
second line above states that the release is being modeled as a vertical jet. As noted earlier,
this is because DEGADIS does not allow the user to simulate a horizontal jet.

Line 5 requests a value of the windspeed UO at a height of ZO m. As in previous examples,
ZO takes on a value of 10 m.  However, the windspeed is 5 m/s, considerably larger than
the 1.5 m/s used in previous examples. This was done after telephone conversations with
the model  developer, Dr T. Spicer.  DEGADIS could not be  made to run at the lower
windspeed and Dr Spicer advised  that this is because, when the initial density is initially high
(as when there is  a large fraction of liquid droplets), the plume simulated by DEGADIS is
likely to fall back onto the source, in which case it will fail to run. A way of overcoming this
problem is to assume a higher initial windspeed.

Line 6 requires a value  of the surface roughness length ZR, which is 0.1 m for all of the
examples in this report.

Line 7: INDVEL is an index which determines  the method of calculation of the velocity
profile in the atmosphere. If INDVEL = 1, the computer program accesses default values
of the profile and of the Monin-Obukhov length based on the stability category ISTAB. In
the present example, the default mode is chosen by setting  INDVEL equal to unity,
choosing the F stability category  (ISTAB = 6,  see above) and setting the Monin-Obukhov
length  RML to 0.  RML is then recalculated by the computer program.
                                       6-25

-------
                       Table  6-4.    DEGADIS Input -
           Flashing Liquid Release of Chlorine from a One Ton Cylinder
FLASHING LIQUID CL2 RELEASE
VERTICAL JET SIMULATION
LARGE SCALE: 1,000 kg, 0.5 inch orifice

5.  10.
0.1
1 6 0.
278.  1. 75.
278.
CL2
70
3600
239
2.00E-05 3.0E-06  1.6
0 0.0  0.0
10
0.0  0.0 1.27
0.0125  0.04 1.33
0.025  0.08  1.39
0.125  0.47  1.84
0.25  1.07 2.40
0.32  1.48 2.78
0.59  3.94 5.07
0.74  6.57 7.52
0.88  11.27  11.90
1.0  20.82 20.82
1.96
5.0  0.11
510.
50.
TITLE 1
TITLE 2
TITLE 3
TITLE 4
UO, ZO
ZR
INDVEL, ISTAB, RML
TAMB, PAMB, RELHUM
TSURF
GASNAM
GASMW
AVTIME
TEMJET
GASUL, GASLL, ZLL
INDHT, CPK, CPP
NDEN
ORDERED TRIPLES
ERATE
ELEJET, DIAJET
TEND
DISTMX
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20-
21
22
23
24
25
26
27
28
29
30
                                   6-26

-------
Line 8: TAMB is the ambient temperature, taken to be 278 K, which lies within the likely
range for atmospheric stability category F.  PAMB is the ambient pressure, taken to be one
atmosphere (the results are not particularly sensitive to this parameter). RELHUM is the
relative humidity, which is set to 75% as being consistent with a cool, relatively humid night
such as might occur under category F weather conditions.

Line 9: TSURF is the surface temperature, set equal to the air temperature in the present
example, 278 K.

Line 10: GASNAM is a three character identifier chosen by the user.

Line 11:  GASMW is the molecular weight of the gas.

Line  1.2:  According  to  the Users'  Manual, AVTIME is the  averaging  time  for the
calculation of the width of the cloud, which is taken to be equal to the exposure time for
the chlorine  LOCs, 3600 s.

Line 13:  TEMJET is the temperature of the release, 239 K.

Line 14:  GASUL and GASLL are the upper and lower concentration levels (or "Levels  of
Concern"  (LOCs)) for which DEGADIS  prints out contours of constant concentration.
These are taken to be the ERPG-3 and the  ERPG-2, namely 20 ppm and 3  ppm, leading
to  mole  fractions 2xlO"5 and 3xlO"6  respectively.  ZLL is  the  height  at which the
concentration contours are calculated and is taken  to be 1.6 m (head height).

Line  15:   INDHT is used to include heat transfer  in the DEGADIS computation.  Heat
transfer is not included for INDHT = 0. In the present case, INDHT is set equal to zero
because heat transfer is already included under the NDEN entry, see below, by introducing
"ordered triples."  In this  case,  CPP and CPK are not used by the computer program and
so are arbitrarily set to zero.  For INDHT =1, heat  transfer is included and  the heat
transfer coefficient is calculated by DEGADIS.  CPP and CPK are used to calculate the
heat capacity as  a function  of temperature according to a correlation that is specified  in
DEGADIS.  If a constant heat capacity is required, set CPP = 0 and CPK to the desired
specific heat at constant pressure (J/kg/K).
                                       6-27

-------
Lines 16 - 26:  NDEN is used to specify the contaminant density profile.  There are three
alternatives for NDEN.  If NDEN = -1, the model treats the contaminant as if it were an
ideal gas with  a molal heat capacity equal to  that of air.  Water condensation effects are
ignored.

If NDEN = 0,  the model treats the contaminant as if it were an ideal gas with heat capacity
indicated by CPK and CPP.  The effects of condensation of water are taken into account
as  appropriate.  This option  is  suitable for use with vapors that  do not have complex
thermodynamic properties when mixed with air.

If NDEN > 0,  NDEN specifies the number of triples which follow in the next NDEN lines.
The triples are used to specify the contaminant concentration as a function of density, based
on adiabatic  mixing with ambient air.  The ordered triples represent (in  order) the
contaminant mole fraction, the contaminant concentration (kg/m3) and the mixture density
(kg/m3).   For chlorine, these ordered triples have been calculated using the method
described in Section 3 of Appendix D, which contains detailed examples.

Line 27:  ERATE is the rate of  release, which is calculated above to be 1. 96 kg/s.

Line 28:  ELEJET is the height of release, arbitrarily set to 5 m. DIAJET is the diameter
of the jet.  The  actual orifice diameter in the present example  is only 0.5".  Here, the
effective diameter is larger because of initial plume expansion during the flashing process.

Line 29: TEND is when the release ends, after 510 seconds. With TEND > 0, DEGADIS
chooses the transient mode once the initial jet phase is over.

Line 30:  DISTMAX is the maximum distance between points in the JETPLU  output (m)
and is arbitrarily set to 50 m.

6.4    150 Ib  Cylinder of Chlorine

The case of the 150 Ib cylinder of chlorine  is similar to that for the 1 ton cylinder, with the
exception that the cylinder is predicted to empty in 35 seconds. Therefore, the only change
in the above SLAB input is that AS (line 14) should be reduced to 35 s.  In addition, HS
(Line 17) could  be set to zero for a cylinder at ground level.  In the DEGADIS input,
                                       6-28

-------
TEND (line 29) is reduced to 35 s and ELEJET is reduced to zero. This scenario could
equally well be modeled as a puff release, see Section 8.

6.5    Miscellaneous

The following section is devoted to a discussion of two items that require further study, a
flashing liquid jet that is directed downward and a  flashing jet from a long pipe.

6.5.1   Jet Directed Downward

This corresponds to Case 3C on Figure (2-1). When a jet containing an aerosol is directed
downward, there is  recovery  of a significant  fraction  of  the initially  airborne aerosol.
Experiments with ammonia, for example, show that this fraction can be as high as 75%.<32).
Thus the source term will be a mixture of an initial airborne cloud with some remaining
aerosol  together with a cloud from  an evaporating pool.  The reader is advised to seek
expert advice in this case.

6.5.2   Jet Emerging from a Long Pipe

If the pipe from which the liquid jet emerges is long, there will be some flashing before the
material is released to the atmosphere.  In this case,~the flow at the outlet will be two phase,
with liquid and vapor in thermal equilibrium. The flow may still -be choked at the outlet,
so  further flashing will occur in the  atmosphere.  Equilibrium flow results in smaller
discharge rates than does pure  liquid flow.  Therefore, the  cases described above may
represent conservatively high flow rates for some release scenarios.

In addition to liquid flow and equilibrium flow, there is a continuum of cases for which flow
at the outlet is two phase but  not in thermal equilibrium.  These  cases have  intermediate
flow rates.

The question of rate of release has been addressed by Wheatley (33), whose article constitutes
a useful source of references.  Another helpful  reference is that by Fauske and Epstein'341,
see Section 6.1.2 and Eqs (6-19) through (6-29) for further discussion.
                                        6-29

-------
6.5.3   Orifice Shape

The  shape of the orifice will also influence  the  predicted rate of release.   The work
described above is appropriate for a circular orifice. If the orifice has a different shape -
for example, if it is a long and narrow rectangle, the constant c0 in  Eq. (6-1) will be
considerably reduced.
                                        6-30

-------
7.0    VAPOR JET RELEASES

The purpose of this  chapter is to describe  how to calculate  emission characteristics and
atmospheric dispersion  model inputs  for vapor jets.  Section (7.1)  contains  a detailed
theoretical discussion of various release formulae. Section (7.2) contains a specific example,
that of chlorine released from a small hole in the vapor space of a  vessel that contains
chlorine liquified under pressure.  The reader who is not interested in theoretical details can
move directly on to Section (7.2), which is  self-contained.  Section (7.3) contains a brief
discussion of holes in the vapor space that are potentially large enough for there to be two-
phase emissions.

7.1    Vapor Release Formulae - Theory

This chapter is intended to describe how to model the atmospheric dispersion of pure vapors that
are driven through an orifice by internal pressure.  This section begins with a discussion of the
issues involved in calculating the release rates and other characteristics of the source term. The
following  subsections (7.1.1 and 7.1.2) show how to estimate the release rate of a gas for two
scenarios:

       (i)     a release of gas from an orifice directly in the wall of a reservoir; and

       (ii)    a release from a pipe attached to a reservoir.

For all of the releases considered in Section 7.1 it is assumed that (a) the released material is
a vapor under the stated conditions and (b) that the pressure and temperature of the gas in the
reservoir are essentially constant.

Note that the work in Sections (7.1.1) through (7.1.3) draws heavily on the release rate work
that has been performed by Spicer(36> under the sponsorship of EPA and  is  intended  to be
consistent  with that work. Spicer's work has been incorporated into the Workbook of Screening
Techniques for Assessing the Impacts of Toxic Air Pollutants (Revised)(3) and the accompanying
TSCREEN model.
                                          7-1

-------
7.1.1  Release Rate Estimates: Leaks of Gas Directly From a Reservoir

This section applies to a release of gas through a hole or opening in the wall of a reservoir in
which the pressure and temperature remain constant.  The released material must be an ideal gas
at the reservoir conditions, during the depressurization process and after depressurization.

The above assumptions are  clearly simplifications.   The assumption  that the release rate is
constant is conservative because a) for a vessel containing vapor only, the pressure in the vessel
declines as the release  proceeds, thus causing a  reduction in the rate  of release and b)  for a
vessel containing both vapor and liquid, the escaping vapor is replaced as the liquid boils.  This
boiling in turn causes the liquid to cool (autorefrigeration) and leads to a corresponding reduction
of pressure in the vapor space.  If the assumption that pressure and temperature are constant
leads to  predictions  of distances of concern that seem overly large  (e.g.  there may be the
potential need to evacuate large numbers of people) it will be necessary for the requestor to seek
additional advice and/or use a contingency model  that can accept a time-varying rate  of release.

The ideal gas law states that pV = RT, where V is the volume of one mole of the  material at
temperature T and pressure p. R is the gas constant. For a non-ideal gas,  the actual behavior
is expressed by a standard text-book equation of the form pV = RT(1 + B/V + C/V2	).
The quantities  B, C, - -, are known as virial coefficients.   Values for the second  virial
coefficient B can be found for many substances in Reference (44),  "Data Compilation - Tables
of Properties of Pure Components,"  which  has  been published by the American Institute  of
Chemical Engineers  (AIChE).  If B/V is large compared to unity, the ideal gas assumption
breaks down. In general, for most commonly encountered vapors, contingency modelers do not
question the ideal gas assumption.

If the reservoir contains both vapor and liquid, the hole must be in the vapor space of the tank.
Possible applications include a gas  leak from a tank, a small gas leak from the wall of a large
pipe, or a gas discharge from a pressure relief valve mounted on a tank.

The procedures discussed in this subsection show the reader how to calculate the release rate and
the  density  and temperature after depressurization.   The input required  for the  calculation
includes:
                                          7-2

-------
              AQ    area of reservoir hole or opening (for guidance on how to choose values
                    for AO, see Section 4.3).
              A,    flow area representing reservoir conditions (m2). For a leak from a tank,
                    AI-* oo (and |8  = 0,  see below). For a leak from a pipe, At is the cross
                    sectional area of the pipe.
              Cp    the specific heat of the released gas at constant pressure (J/kg/K)
              D0    equivalent diameter of reservoir hole or opening (D0 =  (Ao/x)"2)
              hL    the heat of vaporization at the boiling point Tb  (J/kg)
              M    molecular weight of the released gas  (kg/mol)
              pa     ambient pressure (Pa)
              pv     vapor pressure as a function of temperature (Pa)
              p     reservoir pressure (Pa)
              R     gas constant (8314 J/kg-mol/K)
              Tb    boiling point at atmospheric pressure  (K)
              Tc    critical temperature (K)
              Tr    discharge temperature (K)
              T,    reservoir temperature (K)
              ft     (Ao/A,)1/2
              7     Cp/Cv where Cv = Cp - R is the specific heat at constant volume (J/kg/K)
              p}     density of gas inside  reservoir (kg/m3)

The calculational procedure is as follows:

A.     Choked Flow Pressure:   determine whether the flow is choked or not by first estimating
       the choked flow pressure p, from;
            p,  -
                      (Y+D
       If p, >  pa, then the flow is choked or critical (go to B below).  If p, <  pa, then the
       flow is subcritical (go to C below).

B.     Choked Flow:  for choked flow, estimate a) the gas temperature T,, when the pressure
       is p.; b) the emission rate Q (kg/s); and c) the discharge temperature Tr - see paragraphs
       a), b) and c) below, respectively.
                                         7-3

-------
      a)     Estimate T». from the equation:

            T.. = 2T,/ (Y+D                                                  (7-2)

      Before proceeding further, it is necessary to enquire whether condensation will occur by
      comparing T., with the  (pseudo) critical temperature T0.   If T« > Tc, there is no
      condensation and Eq. (7-2) applies. If T,. ^ T0, first calculate the vapor pressure of the
      gas at -temperature T« using the Clausius Clapeyron equation:


            pv = l.OlxlO5  exp{({h,M)/R)  (l/Tb-l/T..))          '         (7-3)

      If pv(T«)  > p,, then condensation does not occur during depressurization arid Eq. (7-2)
      is valid.  If pv(T«)  < p.,  then the release should be considered as two phase at the
      critical pressure and the appropriate procedures are given in Section  (6.1).  Note that,
      for some releases, even if there is liquid present at choked conditions, there may not be
      liquid when the final expansion to atmospheric pressure takes place. Section (6.1) shows
      the reader how to determine if this is the case.

      For multicomponent  mixtures, the  mixture dewpoint temperature  at  p,  should be
      calculated.  This  can be done using methods described by Sandier*37* (not further
      discussed here).  If T,. is less than the dewpoint temperature, then condensation occurs
      and the reader should proceed  to Section (6.1).

b)    Estimate the emission rate Q from Perry and Chilton(38):

           Q =  CQA0[pp1Y(2/ (Y+l))<^1>'CY-D]1/2                           (7-4)

      Taking into account the ideal gas law,  which gives p} = pM/RT, Eq.(7-4) can be
      rewritten  in a  form in which it is frequently found in  textbooks(16>17):
           Q =  G0A0p[(yM/RT) (21  (y*i))
-------
c)    Estimate the Discharge Temperature :  Spicer recommends the following formula:

           Tr = T^l-0.85 ( (Y-l)/ (Y+D )]                                (7-6)

      based on the assumption that the expansion from reservoir to choked conditions occurs
      adiabatically and irreversibly.   However, the  expansion from choked  conditions  to
      atmospheric pressure is assumed to occur adiabatically but not reversibiy.  Based on
      work by Lewitt^ the irreversible, adiabatic expansion is 85 % efficient. After estimating
      Tr, the reader should proceed to step D below.


C.    Subcritical Flow

a)    Estimate the  Emission Rate Q:   Perry's handbook gives  the  rate of discharge  for
      subcritical flow as:

           Q = KYAofep, (p~pa)]1/2                                          (7-7)

      where K = GO (1 - /?*)1/2 (= c0 for 18 = 0) and

           Y = l-[(p-p)/ (pY)](0.41 + 0.354)                           (7-8)
          .    = l-0.4l[(p-pa)/ (py)] for P  =  0

      A frequently encountered alternative formulation is(I6'17):

       Q = c0A0(2pp(Y/(Y-l)[(p/p)2^-(p/p)(^i)/Y])V2                (7_9)
      Spot hand calculations show that Eq.(7-7) is somewhat conservative with respect to
      Eq.(7-9)
                                        7-5

-------
b)    Estimate Discharge Temperature Tr: the discharge temperature is given by:

                                                                             (7-10)
       where

           a =  (l/2Cp) (QR/fpMA,,))2                                     (7-11)


       This formula has been taken by Spicer from work by Lees(40).

       The above estimate for Tr must be checked to see if there will be condensation. If Tr >
       Tc  (the (pseudo) critical temperature) Eq. (7-10) is valid.  If Tr  :£ Tc, the  following
       procedure for single component parameters should be applied: first calculate  the vapor
       pressure pv at temperature Tr using the Clausius-Clapeyron equation (7-3).  Ifpv(Tr) <
       pa, then condensation occurs during the process of depressurization and the release should
       be treated as two phase, see  Section (6. 1).  For multicomponent contaminants, estimate
       the dewpoint of the mixture at pressure pa using the methods of Sandier*3^ (not addressed
       here). If Tr is less than the dewpoint, there will be condensation and the reader should
       proceed to Section (6.1).

D.     Discharge Density: pr can be calculated from the discharge temperature using the ideal
       gas law:
           pr = paM/(RTr)-                                           •       (7-12)

E.     Examples :   Spicer^ gives examples of the applications of steps A through D above to
       the following:

             Air leak  (reservoir pressure of l.OlxlO6  Pa; hole of diameter 5.25 cm; choked
             flow)
             Air leak (reservoir pressure of 1 .82X105 Pa; hole of diameter 5.25 cm; subcritical
             flow), and
             Chlorine gas leak (reservoir pressure of 6. 89x10* Pa; temperature 320 K; hole of
             diameter 2.8 cm; choked flow).
                                         7-6

-------
The reader is directed to Reference (36) for details.  Section 7.2 below contains the case of a
chlorine gas leak from a reservoir at a pressure of 4.3xl05 Pa, with choked flow through a hole
of diameter 3/8" (0.0096 m).

7.1.2  Release Rates:  Gas Leaks from a Pipeline Attached to a Reservoir

This section applies to a continuous release of a gas through a  long pipe from a reservoir in
which temperature and pressure are assumed to remain constant.  The flow through the pipe is
assumed to be adiabatic.  The released material  must be an ideal gas at the .conditions in the
reservoir, during the depressurization  process and after depressurization.  If the reservoir
contains vapor and liquid, the pipe must be attached to the vapor space.

The procedure described below provides estimates of the gas release rate  and its density and
pressure after depressurization. The input information required is as follows:

             AO     area of orifice (m2)
             DO     equivalent diameter of hole (m);  D0  =  (Ao/T)1/2
             Dp     pipe diameter (m)
             Lp     pipe length
             M     molecular weight  (kg/mol)
             Ns     number of pipe elbows
             pa     ambient pressure (Pa)
             pv     vapor pressure (Pa)
             p      absolute pressure in reservoir (Pa)
             R     gas constant (8314 J/kg-mol/K)
             Tc     critical temperature (K)
             T,     reservoir temperature (K)
             7
The procedures to be followed are as follows:

A.     Pipe Friction Loss:  The frictional loss N is estimated as(38):

                 4fL
            N = - ? + 0.5 +  0.75Ne +  (0.5)                            (7-13
                    p
                                          7-7

-------
In this equation, the first term represents skin friction losses in the pipe with f — 0.0045. The
second term provides for friction losses in  the reservoir to pipe connection.  The third term
accounts for frictional losses in  any pipe elbows.   The fourth  term (written in parentheses)
should only be included if D(/DP < 0.2:  otherwise, it should be neglected.

B.    Discharge for Choked Flow.  First,  estimate the rate of discharge if the flow is choked
      (the validity of this assumption will be checked later), in which case the speed of the gas
      at the exit approaches that of sound.  For choked flow, the following equations must be
      solved simultaneously for the Mach number M2 at the pipe entrance and a dimensionless
      parameter Y2:

           " Y   =  1  + Jl-LlLM2
            X2    X  + - - - "2                                            (7-14)
                                     -  (JL  - I)' + YN  - 0
                                         M,2                                   (7~15
After M2 and Y2 are known, the mass flux G is obtained from:

     G  = p^CM
                       RT,
                            2                                                   (7-16]
       The discharge pressure must now be checked to see if it is indeed choked:

                      RT,    2
            P3  = G  (	l(_f_))1/2                                         (7-17)
                       MY   Y + i
       If p3 < pa or p3 >  p1} then the flow is not choked and the reader should proceed to step
       C below.  If P]  > p3 > pa then the flow is choked and the discharge rate is given by Q
                The discharge temperature is given by Tr = 2T,/(7 + 1).
C.     Discharge for Subcritical Flow:  For subcritical flow, the gas discharge pressure at the
       end of the pipe approaches that of the atmosphere.  The following eight equations must
       be solved simultaneously for the  mass flux, G; the Mach numbers at the pipe entrance
       and exit, respectively, M2 and M3; the temperature at the pipe entrance and exit, T^ and
       Tr; and two dimensionless flow parameters, Y2 and Y3:

                                          7-8

-------

                      RT,
           M  = -1 (122)1/2                                                (7-22)
            2   P2   YM


                        P1/2                                                (7-23)
                p3
                                                                             (7-24
                   T1
           Tp     Y2
                                                                             (7-25
      The above equations have been written into TSCREEN and are solved numerically there.
      The solution to these equations must be checked particularly to ensure that M3  <  1 , p
       > p2 and T! > T2. If these conditions are not met, seek expert advice.

D.    Check Discharge Temperature Tj if Tr is greater than the pseudo-critical temperature
      Tc, the equation above used to estimate Tr applies.  If not, the following procedure is
      recommended.   For single component contaminants, evaluate the contaminant vapor
      pressure at temperature Tr, pv. If pv < pa, then contaminant condensation occurs during
      the process of depressurization and the release is two phase: the reader should proceed
      to Section (6.1).

                                         7-9

-------
      For multicomponent contaminants, estimate the dewpoint temperature of the mixture at
      pressure pa using the method described by Sandier'38'.  If Tr is less than the dewpoint
      temperature,  then  condensation occurs during the process  of depressurization and  the
      reader should proceed to section (6.1).

E.    Discharge Density:  the discharge density p, may be estimated from the equation p{ =
      PaM/(RT3)

F.    Examples: The following examples are contained in the workbook:

             Air leak (reservoir pressure of 1.101x10' Pa, 10 m of pipe of diameter 5.25 cm)
             Air leak (reservoir pressure of 1.824xl06 Pa, 10 m of pipe of diameter 5.25 cm)

7.1.3 Flow Chart

Figure 7-1 is a flow chart that summarizes the discussions in Sections (7.1.1) and  (7.1.2).


7.2  Chlorine Vapor Release

7.2.1 Description of Scenario

This example is  relevant  to the case of a gas liquified under pressure.  There is a hole in  the
vapor space which could  be caused for any of the reasons  previously discussed - impact from
an external agent, failure at a corroded area, valve or gasket failures, for example.

This corresponds to  scenario 4A on Figure 2-1. It is assumed that there is a storage vessel such
as a one ton cylinder  containing chlorine at  278 K.  For some reason such as the  rupture of a
pipe, there is a leak from the vapor space. In this case,  it is assumed that the orifice is 3/8" in
diameter because this is a typical size for pipework from a one ton cylinder. Chlorine vapor jets
vertically out of the orifice.
                                         7-10

-------
  From
Fig. J-l.p.6
Scenario 4 A^

Emission Kale
Q
b|S(7-4).(7-S)



Discharge
Temperature
anil Density
Eys (76).
(7-7U7 13)



NO
Subcnlicat Discharge
Flow Rale and
Ei)»(7 I9)lhru
<7-26)


\
f1
                                                                                                        Corukii-
                                                                                                         solion?
                                                                                                         Section
                                                                                                        (7.1 2D)


l^epare
Iiipul For
Dupersioii
MotleU
                                                                                                         fkoplel
                                                                                                        Formal ion
                                                                                                      Go to Seel (6-1)
                                                                    Figure 7-1
                                                  Vapor Release Rale Calculations Flow Chart

-------
The analyst must first compare  p/pa with ((-y+l)/2)'l'/(Y'1), where pa is the ambient pressure,
l.OlxlO5 Pa, p is the absolute pressure in the vessel (Pa) and 7 = CP/CV is the ratio of specific
heats, 1.308 for chlorine. This comparison is the same as that made in Eq. (7-1).  For chlorine
at 278 K, the vapor pressure is 4.323X105 Pa  (from standard vapor pressure equations) so that
p/pa is equal to 4.32.  ((y+l)!2^n equals 1.84, so that p/pa  > ((7+l)/2)Y/(-y-') and the flow
is choked.

The quantity T., in Eq. (7.2) is  equal to (2)(278)/(2.308) = 240 K.  This is well below the
critical temperature Tc of chlorine (450 K) so that T., < Tc.  From Eq.(7-3), the vapor pressure
corresponding to temperature T,,  is l.OSxlO5 Pa < p..  Therefore, there is condensation and the
reader should proceed to Eq.(6-2), from which  the temperature corresponding to the critical
pressure, p, can be calculated to be T, = 261 K.  T, and p, are the temperature and pressure at
which the material  actual emerges from the orifice. The properties of the gas at these conditions
are as follows: a) The vapor fraction x. is 0.962 (from Eq.(6-3)); b) the density p* is 7.88 kg/m3
(from Eq. (6-5)); and c) the predicted rate of release Q is 0.111 kg/sec (from Eqs (6-4) and
(6-6)).

From Eq. (6-7), the discharge temperature Tr equals the atmospheric boiling point, 239.1 K and
the corresponding released vapor fraction xr is 1.00088 (Eq. (6-9)).  Since xr > 1, the solution
is unphysical and Eq. (6-10) applies, giving Tr  = 239.6 K.  The corresponding density pr is
3.549 kg/m3 from  Eq. (6-11).

Thus, from above, the predicted  flow rate is 0.111 kg/s.  In practice, as noted previously, this
rate of flow would begin to decline immediately because, as vapor escapes, the pressure in the
vessel declines.  This causes more vapor to evaporate, cooling the liquid in the vessel, thus
reducing the pressure.  This causes the  release  rate to decrease.   However, to illustrate a
conservative calculation, the release rate was assumed to remain constant for 20 minutes. By
this time, in practice,  the release rate would have been much reduced.  If the results of such a
conservative calculation show unacceptable distances to the LOG, it will be necessary to repeat
the calculation with more careful attention to the  variation of release rate as a function of time.

7.2.2  Input to SLAB

The input to SLAB corresponding  to the above scenario is contained in Table (7-1).
                                          7-12

-------
    Table 7-1.  SLAB Input -
Vertical Vapor Release of Chlorine
3
1
0.07091
498.1
239.1
0.
287840.
926.3
1574.
1978.34
-27.01
239.6
0.111
1.6E-04
1,200.
0.
5.
3600
l.OOE + 04
1.6
0.
0.
0.
0.1
10.
1.5
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
QTIS
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
- 16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
              7-13

-------
Line 1: IDSPL is the spill source type.  IDSPL = 3 for a vertical jet.

Line 2: NCALC is a numerical substep parameter. The code developer recommends using
NCALC  = 1.   However, NCALC can be increased if numerical stability problems are
encountered.

Line 3: WMS is the molecular weight of chlorine in kg/mole, WMS = 0.07.

Line 4: CPS is the specific heat at constant pressure, taken from a table in the SLAB Users'
manual.  CPS = 498 J/K/Kg.

Line 5: TBP is the boiling point of chlorine, 239.1 K.

Line 6: CMEDO is the liquid mass fraction, which is 0 because the release is  pure vapor.

Lines 7 - 9: DHE =  287,840, CPSL  = 926.3 and RHOSL ~  1,574  are  the  heat of
vaporization at  293 K  (J/kg), the specific heat of liquid chlorine (J/kg/K) and the liquid
density (kg/m3) respectively.  Their  values are taken from Table 2 of the SLAB Users'
Guide.

Lines 10,11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure  of HF:

            P, = Pa.exp(SPA - SPB/(T + SPC))                        (7-26)

P3 is the saturated vapor pressure, Pa is the ambient pressure (= l.OlxlO3 N/m2), a value for
SPA is specified in the code and the values of SPB (1978.34) and SPC(-27.01)  are given in
the Users' Guide.  T is the  ambient  temperature (K).  Eq.(7-26) with the  given values of
SPB and SPC is an empirical fit to the  experimental data on chlorine  vapor pressure.

Lines 12 - 17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature of the released  material, 239.6 K. QS is the rate of release, 0.111
kg/s. AS is the effective area of the source,  1.6x10"* m2. This is derived as follows:  the
density at choked  conditions (from above) is 7.88 kg/m3. The material  exits the hole at this
density and then  expands to atmospheric pressure, at  which its density is 3.549 kg/m3.
                                       7-14

-------
Therefore, assuming that the  exit velocity remains unchanged,  the  effective area must
increase by the ratio 7.88/3.549 = 2.22. The orifice area is 7.86xlO'5 m2. Therefore, the
effective area is 2.22x7.86xlO'5  = 1.6xl(T* m2.  HS is the height of release, 5 m.

Line 18: TAV is the exposure time. As noted previously, it is set equal to the exposure time
for the LOG, 3600 s.

Line 19: XFFM is the maximum downwind extent of the calculation.  It may be necessary
to determine this  by trial and error.  A value of 10 km (l.OE 04 m) should be adequate for
many applications.

Lines 20 - 23: ZP(I) allows the user to specify up to four heights at which the concentration
is calculated as a  function of downwind distance.

Lines 24  - 29  allow the user to  specify meteorological conditions.   ZO is the surface
roughness length,  which is set to 0.1 m as in previous examples. ZA is the height at which
the windspeed is measured (10 m).  UA is the windspeed at height ZA (1.5 m/s). TA is the
ambient temperature (278 K).  RH is  the relative humidity (75%). STAB is the stability
class (6 or F).

Line 30: TER  < 0 terminates the  run.

7.2.3 Input to DEGADIS

The input to DEGADIS corresponding to the above case is contained in Table  (7-2).

Lines 1 - 4 of the  above table allow the user to input four lines of title.

Line 5 requests a value of the windspeed UO at a height of ZO m.  As in previous examples,
these take on values of 1.5 m/s and 10 m.

Line 6 requires a  value of the surface  roughness length ZR, which is  0.1 m for all of the
examples in this report.
                                       7-15

-------
                      Table 7-2.  DEGADIS Input -
                     Vertical Vapor Release of Chlorine
CL2 VAPOR RELEASE
VERTICAL JET SIMULATION
LARGE SCALE: 3/8" orifice

1.5 10.
0.1
1 6 0.
278. 1.  75.
278
CL2
70
1200
239.6
2.00E-05  3.00E-06  1.6
0 498.1 0.0
0
0.111
5.0 1.6E-04
1200.
50.
TITLE 1                 1
TITLE 2                 2
TITLE 3                 3
TITLE 4                 4
UO, ZO                   5
ZR                     6
INDVEL, ISTAB, RML     7
TAMB,  PAMB, RELHUM   8
TSURF                  9
GASNAM                10
GASMW                 11
AVTIME                 12
TEMJET                 13
GASUL, GASLL, ZLL      14
INDHT, CPK, CPP         15
NDEN                   16
ERATE                 17
ELEJET, DIAJET         18
TEND                   19
DISTMX                 20
                                 7-16

-------
Line 7:  INDVEL is an index which determines the method of calculation of the velocity
profile in the atmosphere. If INDVEL = 1, the computer program accesses default values
of the profile and of the Monin-Obukhov length based on the stability category ISTAB.  In
the present  example,  the  default mode is chosen by setting INDVEL equal to unity,
choosing the F stability category (ISTAB = 6, see above) and  setting the Monin-Obukhov
length RML to 0.  RML is then recalculated by the computer program.

Line 8:  TAMB is  the ambient temperature, taken to be 278 K, which lies within the likely
range for atmospheric stability category F.  PAMB is the ambient pressure, taken to be one
atmosphere ( the results are not particularly sensitive to this parameter).  RELHUM is the
relative humidity, which is set to 75% as being consistent with a cool, relatively humid night
such as  might occur under category F weather conditions.

Line 9:  TSURF is the surface temperature, set equal to the air temperature in the present
example, 278 K.

Line 10: GASNAM is a three  character identifier chosen by the user.

Line 11: GASMW is the molecular weight of the gas.

Line  12:  According  to  the  Users'  Manual, AVTIME is the averaging time for the
calculation of the width of the cloud, which is taken to be equal to  the duration of release
in this example.

Line 13: TEMJET is the temperature of the  release.

Line 14: GASUL and GASLL are the upper and lower concentration levels (or "Levels of
Concern" (LOCs)) for which  DEGADIS prints out contours of constant concentration.
These are  taken to be the ERPG-3 and the ERPG-2, 20 ppm and 3  ppm, or mole fractions
of 2.0xlO'5  and 3.0xlO~6 respectively. ZLL is the height at which the  concentration contours
are calculated and is taken to be 1.1.6 m (head height).

Line 15: INDHT is used to include heat transfer in the DEGADIS computation.  Heat
transfer is  not included for INDHT =  0.  For INDHT = 1, heat transfer is included and the
heat transfer  coefficient is calculated by  DEGADIS. CPP and CPK are used to calculate
                                      7-17

-------
the heat capacity as a function of temperature according to a correlation that is specified
in DEGADIS. If a constant heat capacity is required, set CPP = 0 and CPK to the desired
specific heat at constant pressure (J/kg/K), 498.1 for chlorine.

Line 16:   NDEN is used to  specify the  contaminant density profile.  There are three
alternatives for NDEN.  If NDEN = -1, the model treats the contaminant as if it were an
ideal gas with a molal heat capacity equal to that of air.  Water condensation effects are
ignored.

If NDEN = 0, the model treats the contaminant as if it were an ideal gas with heat capacity
indicated by CPK and CPP.  This option is suitable for use with vapors that do not have
complex thermodynamic properties when mixed with air.

Line 17:  ERATE is the rate of release, which is calculated above to be 0.054 kg/s.

Line 18:  ELEJET is the height of release, arbitrarily set to 5 m.  DIAJET is the diameter
of the jet.  The actual  orifice diameter in the present example is only 3/8". The orifice
diameter specified  above takes account  of  the  initial expansion  of the  plume,  seethe
explanation for line 14  of the  SLAB input in  Table 7-1.

Line 19:  TEND is when the release ends, i.e. after 20 minutes or  1,200 seconds.   With
TEND > 0, DEGADIS chooses the transient mode once the initial jet phase is over.

Line 20:  DISTMAX is the maximum distance between points in the JETPLU output (m)
and is arbitrarily set to 50 m.

7.3 Intermediate Sized Hole in the Vapor Space

There exists a category of intermediate failures such as that illustrated in #4C, Figure (2-1),
in which the hole can be categorized as neither small nor large. Considerable relevant work
has been done by  the AIChE's DIERS (Design Institute for  Emergency Relief Systems)
program.  Any one  of up to four flow regimes through the orifice is possible:
                                       7-18

-------
      •     Pure vapor - already discussed for a small hole
      •     Bubbly flow regime in  which the liquid phase  is continuous with discrete
            bubbles
      •     Churn  turbulent flow regime in which the liquid phase is continuous with
            coalesced vapor regions of increasing size
      •     Droplet flow regime in which the vapor phase is continuous with discrete liquid
            droplets

The transition between these various flow regimes occurs with increasing volumetric vapor
flux and is also influenced by fluid characteristics and by the presence of impurities.  It is
unlikely that, if such a scenario is envisaged, a typical on scene co-ordinator or "emergency
responder will have access to modelers or resources that will  enable a quick atmospheric
dispersion calculation to be done. Hence expert assistance should be sought, and this case
is not be discussed in detail in the present work.
                                       7-19

-------
8.0   INSTANTANEOUS (PUFF) RELEASES

8.1   Description of Release

This release corresponds  to the catastrophic failure of a vessel containing chlorine under
pressure. The sequence of events in this case is that a major rupture of a vessel or cylinder
occurs, releasing the whole  contents almost at once.  Some percentage of the  chlorine
flashes to vapor, and the remainder  is fragmented and remains airborne  as fine liquid
droplets. This initial flashing process is highly turbulent and much air is entrained almost
immediately. Experiments and observations of the outcome of actual accidents suggest that
the initial mass of air entrained can equal about ten times the initial mass of chlorine. This
mixture is taken to  be the starting point for the SLAB or  DEGADIS calculations. The
initial characteristics of the puff are calculated as described in Section 2.1 of Appendix D.

In this particular case, the rupture of a one ton cylinder of chlorine is considered. The
spontaneous rupture of such a vessel is very unlikely:  it generally corresponds to a worst-
case scenario. The  case of a small ruptured cylinder is discussed in Section 8.4.

8.2   Input to SLAB

The input to SLAB is displayed on  Table 8-1.  The  principal" differences from previous
SLAB  inputs are that a) this is a puff release and b) the initial  source already has  air
entrained into it by  assumption.

Line 1: IDSPL is the spill source type. IDSPL =4 for a puff.

Line 2: NCALC is a numerical substep parameter.  The code developer recommends using
NCALC  =  1.  However, NCALC can be increased  if numerical stability  problems are
encountered.

Line 3: WMS is the molecular weight of the initial chlorine/air mixture in kg/mole. The
reader should remember  that the initial puff consists of 1,000 kg of chlorine  and  10,000 kg
of air,  so that chlorine only takes up 9.1% by mass or 9.1x28.9/70 =  3.76% by volume.
Therefore,  the effective  molecular weight is WMS = (70)(0.0376) + (28.9)( 1-0.0376) =
30.45.
                                        8-1

-------
Table 8-1.  SLAB Input-
Puff Release of Chlorine
4
1
0.03045
954.
239.1
0.
287840.
926.3
1574.
1978.34
-27.01
239.1
0.
444.
0.
11000.
0.
3600
l.OOE + 04
1.
0.
0.
0.
0.1
10.
1.5.
278.
75.
6.
-1.
IDSPL
NCALC
WMS
CPS
TBP
CMEDO
DHE
CPSL
RHOSL
SPB
SPC
TS
QS
AS
TSD
QTIS
HS
TAV
XFFM
ZP(1)
ZP(2)
ZP(3)
ZP(4)
zo
ZA
UA
TA
RH
STAB
TER
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
          8-2

-------
Line 4: CPS is the effective specific heat at constant pressure. Here again, a composite gas
consisting of 9.1% of C12 and 90.9% of air has an effective specific heat at constant pressure
of (0.091)(498.1) +  (0.909)(1,000) -  954 J/Kg/K.

Line 5: TBP  is the  boiling point of chlorine, 239.1 K.  SLAB does not make use of this
quantity because the initial puff has been defined to consist of vapor.

Line 6: CMEDO is the liquid mass fraction, which, for the composite air/chlorine mixture
is zero because all of the chlorine droplets have been evaporated by the  air.

Lines 7 - 9:  DHE  =  287,840, CPSL. =  926.3 and RHOSL ~  1,574  are  the  heat of
vaporization at 293  K (J/kg), the  specific heat of liquid chlorine (J/kg/K) and the liquid
density (kg/m3) respectively.  Their  values are taken from Table 2 of the SLAB Users'
Guide. These quantities are not made use of by SLAB because the initial puff is entirely
vapor.

Lines 10.11: SPB and SPC are parameters that go into the formula for the saturated vapor
pressure of chlorine:

             P3 = Pa.exp(SPA - SPB/(T + SPC))

Ps is the saturated vapor pressure, Pa is the ambient pressure (=  1.01E 05 N/m2), a value
for SPA is specified in the code and the values of SPB (1978.24) and SPC(-27.01) are given
in the Users' Guide. T is the ambient temperature (K).  These quantities  are also not used
in the present case.

Lines 12 - 17 specify spill parameters, which have already been discussed in detail above.
TS is the temperature  of the released material, 239 K.   QS is the rate of release, 0 for a
puff.  AS is the effective area of the air/chlorine source, 444 m2 from Eq. (2-2) of Appendix
D. TSD is the duration of release, 0. QTIS is the initial mass in the puff, 11,000 kg, being
made up of 1,000 kg of chlorine and 10,000 kg of air.  HS is the effective height of release,
0 m.

Line  18: TAV is the exposure time. As discussed previously for chlorine, it is set equal to
one hour.
                                        8-3

-------
Line 19: XFFM is the maximum downwind extent of the calculation.  It may be necessary
to determine this by trial and error. A value of 10 km (1.0xl04m) should be adequate for
many applications.

Lines 20 - 23: ZP(I) allows the user to specify up to four heights at which the concentration
is calculated as a function of downwind distance.

Lines 24 - 29 allow the user to specify meteorological conditions. ZO is the surface roughness
length, which  is  set to 0.1  m as in previous  examples. ZA  is  the height at which  the
windspeed is measured (10 m).  UA is the windspeed  at height ZA (1.5 .m/s). TA is  the
ambient temperature (278 K).  RH is the relative humidity (75%).  STAB is the stability
class (6 or F).

Line 30: TER  <  0 terminates the run.

8.3     Input to DEGADIS

The input for DEGADIS is displayed on Table 8-2.

Lines 1 - 4 of the table allow the user to input four  lines of title.

Line 5  requests a value of the windspeed UO at a height of ZO m.  As in previous examples,
these take on  values of 1.5 m/s and  10 m. Line 5 also requires a value of the surface
roughness length ZR, which is 0.1 m for all of the examples in this  report.

Line 6 requests a value for the stability category, 6 (F) in this case.

Line 7:  OODIST is the distance downwind  of the source  at which  the DEGADIS
calculations start.  AVTIME is the averaging  time  for plume meander, set equal to  the
exposure time  for the chlorine LOG, 3600 seconds.

Line 8: DELTA, BETA and RML are explained in  Section 5.1.4.

Line 9: consists of parameters that describe the spreading of the plume along the wind due
to the  action of atmospheric turbulence.  They are described in Section 5.1.4.
                                       8-4

-------
                                     Table 8-2.   DEGADIS Input
                                     Puff Release  of Chlorine
INSTANTANEOUS  PUFF RELEASE
1,000 kg CHLORINE
  1.50    10.00
  6
  .00    3600.00
.10






£











7.74E-02 .90 17.52
.17 .97 50.00
278.00 1.00 4.00E-03 75.00
0 278.00
1 .00
0 .00
CL2
30.44 238.70 3.67
954. 1.00
5.400E-04 8.100E-05 1.60
l.OOOE-04
11000
3
.00 .00 11.89 3.76E-02 239.00
1.00 .00 .00 3.76E-02 239.00
2.00 .00 .00 3.76E-02 239.00
FT F F F F
19-MAY-1992 14:48:50.40













1.00
1.00
1.00


                                       TITLE
                                       TITLE
                                       TITLE
                                       TITLE
                                                                                 SIGX MIN D1ST
UO, ZO, ZR
ISTAB
OODIST, AVTIME
DELTA,  BETA, RML
SIGX COEFF,   SIGX_POW,
TAMB, PAMB,  HUMID
ISOFL, TSURF
IHTFL, HTCO
IWTFL, WTCO
GAS NAME
GAS MW,  GAS TEMP,  GAS RHOE
GAS CPK,   GAS CPP
GAS_UFL,   GAS_LFL,   GAS_ZSP
CCLOW
GMASSO
NT
PTIME(l),  ET(1), R1T(1),  PWC(l), PTEMP(l), PFRACV(l)
PTIME(2),  ET(2), R1T(2),  PWC(2), PTEMP(2), PFRACV(2)
PTIME(3),  ET(3), R1T(3),  PWC(3), PTEMP(3), PFRACV(3)
CHECK1,  CHECK2, AGAIN,  dHECK3,CHECK4,CHECK5
TINP
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

-------
Line 10: TAMB is the ambient temperature, taken to be 278 K, which lies within the likely
range for atmospheric stability category F.  PAMB is the ambient pressure, taken to be one
atmosphere (the results are not particularly sensitive to this parameter). HUMID consists
of two numbers, the absolute humidity (4.0xlO~3 kg water/kg air) and the relative humidity,
which is set to 75% as being consistent with a cool, relatively humid night such  as might
occur under category F weather conditions.

Line 11: TSURF is the surface temperature, set equal to the air temperature in the present
example, 278 K. ISOFL is a number generated by the DEGADIS interactive input routine
as discussed in Section 5.1.4.

Lines 12 and 13: IHTFL, HTCO, IWTFL and WTCO are also generated by the interactive
method of preparing  input for DEGADIS  as discussed in Section 5.1.4.

Line 1.4:  GASNAM is a three character identifier chosen by the user.

UnemJ5:   GASMW  is  the effective molecular weight of the gas/air mixture,  30.44  as
explained in the discussion of the SLAB  input.  GAS-TEMP is the temperature of the
released gas (239 K) and GAS-RHOE is  the density of the air/chlorine  mixture at that
temperature (1.56 kg/m3, see Section 2.1 of Appendix D).

Line 16: GAS-CPK and GAS-CPP are parameters in DEGADIS' formula for specific heat
at constant pressure.  In the formulation chosen here, 954 J/kg/K is the specific heat of the
initial air-gas mixture as  discussed during the development "of the SLAB input. GAS-CPP
= 1 ensures that this value is chosen independent of temperature.

Line 17:  GAS-UFL is the upper concentration of interest.  It is labeled  "UFL" because the
model was originally developed to handle  flammable vapors.  Here,  it corresponds to the
ERPG-3 for chlorine, 20 ppm. However, as noted above, the  initial  puff consists of 1,000
kg of chlorine and 10,000 kg of air, so that  the chlorine only makes up 9.1% of the total  by
mass or 9.1x28.9/70 = 3.76% by volume.  Therefore, the upper limit should be increased
from 20 ppm to 20/0.037 = 540 pprn =  5.4X10"4 mole fraction. Thus, the input data is being
manipulated to ensure that the model recognizes that, when the concentration of chlorine
is actually 20 ppm, the concentration of the initial composite air/gas  mixture is in fact 540
ppm.  Similarly, the chlorine LOG of 3 ppm is manipulated to be equivalent to a composite
                                       8-6

-------
LOG of 80 ppm, which is equivalent to a mole fraction of ~ S.lxlO"5. Finally, ZSP is the
height at which the concentration is measured, taken to be 1.6 m (head height).

Line 18:  CCLOW is the lowest concentration of interest (in kg/m3). and is taken to be
l.OxlO"4 kg/m3 (-40 ppm), somewhat below the composite ERPG-2 described above.

Line 19:  CMASSO ( = 11,000 kg) is the initial mass.

Lines 20-23:  NT specifies the number of times at which release parameters are required.
For this puff example, the DEGADIS input requirements are  arbitrarily satisfied by 3 points
with release rate zero. Note that the fourth column (PWC)  reproduces the initial dilution
over the source.  The radius of 11.89 m is that of the source  mixture of air and chlorine as
described in Eq. (2-1) of Appendix D.

Lines 24-25 are always generated by the interactive data entry module and are not further
discussed here.

8.4   Small Cylinder of Chlorine

In order to consider a puff release of chlorine, all that is necessary is to replace the initial
mass in the DEGADIS and SLAB outputs above by 150 Ib of chlorine plus 1500  Ib of air,
which comes to a total of 1650 kg (GMASSO, line 19 in the DEGADIS input and QTIS,
line 16 in the SLAB  input).  In addition, the  initial radius (RIT(l) in line 21  of the
DEGADIS input) falls to a fraction (1650/11000)l/3  = 0.53 of  its original value.  Thus,
RIT(l) in line 21 of the DEGADIS input becomes (11.89)(0.53)  = 6.3 m.  In addition, AS
(line 14 of the SLAB input) becomes ir(6.3)2  - 125 m.
                                       8-7

-------
9.0    BUOYANT PLUME RELEASES

9.1    Introduction

The purpose of this section is to introduce the reader to ways of calculating the air impacts of
buoyant plumes.  This is done by discussing three examples, namely a) an incinerator on a
Superfund site; b) a burning pool containing PCBs; and c) a stack of burning tires.  For the
interested reader, Appendix C contains a discussion of the elements  of plume rise modeling.
EPA has two models that contain plume rise algorithms, TSCREEN for short term, worst-case
ambient concentration estimates and the Industrial Source Complex Model (ISCLT) for long-term
average concentrations.

9.2    Incinerator Plume

9.2.1  Background

Incineration is  frequently employed at  Superfund sites in  order to destroy Volatile Organic
Compounds (VOCs) in contaminated waste.  An incinerator plume (Scenario #8, Figure 2-2) is
one of the simpler cases considered in the present work because it corresponds to an isolated
point source.

The items that are needed to  specify the characteristics  of the source term for a buoyant plume
from an incinerator are:

       o     the rate of emission from the stack Q in g/s for each toxic  material  in the plume

       o     stack data: a)  stack exit temperature (K); b) stack exit velocity (m/s); c) stack
             inside diameter (m); d) stack height (m).

Reference (7) contains a detailed  discussion  of screening  procedures for  estimating the  air
impacts of incineration using a rotary kiln at Superfund sites. For the convenience of the reader,
a summary of the approach taken in Ref. (7) is included.
                                          9-1

-------
9.2.2  Specification of Source Term for a Buoyant Plume

Figure 9-1 shows a schematic drawing of a typical  rotary kiln incineration  system that  is
designed to process soils and liquids that are contaminated by volatile and semivolatile organic
compounds. Organics are volatilized in the kiln and exit with the hot gases into the secondary
combustion chamber (SCC) where destruction is completed.  Cyclones upstream of the SCC trap
large paniculate matter.  A water quench reduces gas temperature, and a packed tower scrubber
provides  primary  removal of acid gases.  An ejector scrubber removes fine paniculate matter
and additional acid gases before release through the stack.

Figure 9-1 shows  typical input rates of liquid and solid (1,000 Ib/hr and 6 tons/hr respectively)
and  the rate  of heat input (35 mm  BTU/hr).  The stack height (12 m) and the stack exit
temperature (160°F = 344 K) are also  specified.  For Superfund site incinerators, a typical
value of stack internal diameter is 0.5 m  in a range  0.3 - 1  m  and a typical stack exit velocity
is 10 m/s in a range 7-20  m/s.  Thus, the  stack parameters have already been specified.
Generally, these would be expected to  be  available from  calculations performed during the
design of the incinerator.

The  emissions of toxic materials depend,  of course, on the nature and quantities of contaminants
in the waste.   Waste characterization data are usually obtained from the remedial investigation
(RI)  and any treatability studies that  may have been conducted.  Generally, it is necessary to
consider emissions of organic compounds, metals, acid gases and paniculate matter. In addition,
the  organic  compounds  can  be  divided  into  three groups,  total  hydrocarbons  (THCs),
polychlorihated biphenyls  (PCBs) and dioxins.  THC compounds are considered to consist of
volatile and semivolatile  organic compounds  that are not PCBs or  dioxins.  Reference (7)
provides worksheets that enable the user to calculate emission rates for all of the above.

Organic  Compounds

If the concentration C0 of one  of the group  of organic  compounds is known in ppm and if the
feed rate  of the waste to the  incinerator is FR Ib/h, then the feed rate FR,, of the organic
compounds of interest is given by

              FR,, =  (FR)(C0)(10-6) Ib/h                                    (9-1)
                                          9-2

-------
Stack
Height 12m
                    Liquid
                    IQQOlb/hr
                    Soil
                    6_Tons/hr
                                    Incinerator
                                    35mm BTU/hr
 Rotary Kiln

1400-1 SOOT
                                            T:
                                                              o
                                              Clean" Soil
                                                      Packed
                                                      Tower
                                                     Scrubber
  Secondary
 Combustion
  Chamber

2100 - 2400T
                                                                       Air Pollution
                                                                      Control System
                   Figure 9-1. Example of a rotary kiln incineration system.
                                           9-3

-------
In Eq. (9-1), 10~6 is a conversion factor that adjusts units.

Emissions of these organic compounds then depend on the destruction and removal efficiency
(DRE)  of the  incinerator.   The exact  DRE depends  on the gas residence  times  and the
temperatures in the incinerator combustion chambers. This is  generally measured during trial
bums of the incinerator. However, if this information is not available, RCRA standards require
a DRE of 99.99% for each principal organic hazardous constituent (POHC).  POHCs are
organic indicators chosen for the trial burn and  are a subset of THCs. Therefore, a 99.99%
DRE may be chosen for THCs. In addition, RCRA standards require a DRE of 99.9999% for
dioxins and  the Toxic Substances  Control  Act (TSCA) standards  also require a DRE of
99.9999% forPCBs.

The  emission rate from  the  stack for the organic compounds that are not removed by the
incinerator is ER,, where:

             ER,, = (FR0)(1 - DRE0(%)/100)(0.126)     g/s                 (9-2)

where 0.126 is a conversion factor from Ib/hr to g/s.

Metals

Metals may be present with organics  in soils or other solid wastes and in liquid waste fuels.
EPA has identified ten toxic metals that may pose a hazard to human health and the environment
when released  in  incinerator emissions:  antimony, arsenic,  barium, beryllium, cadmium,
hexavalent chromium, lead, mercury,  silver and cadmium.  Assume that metal  m is present in
the waste in concentration Cm ppm and that the feed rate of the waste is FR Ib/hr. Then the feed
rate  of the metal is FRn, where:
                  = (FR)(CJ(10^) Ib/h.                                  (9-3)

Metals in the feed will either remain with the solids and be discharged in the bottom ash or they
will be carried out of the rotary kiln and SCC by the combustion gases.  The percentage  that
becomes airborne and travels into the air pollution control device (APCD) is known as the
partitioning factor (PF).  EPA has determined conservative values of PFs from experimental
tests.  The PFs given in Ref. (7) are as follows: a) if the feed is liquid, PFs = 100% for all ten
                                         9-4

-------
metals at all incinerator temperatures; b) if the feed is solid and the solid temperature (generally
100 °F - 400 °F lower than the combustion gas temperature) is 1600 °F, the PFs are all 100%
except for barium (50%), beryllium (5%) and chromium (5%); c) if the feed is solid and the
solid temperature reaches  2,000  °F, the PFs are all  100% except  for beryllium (5%) and
chromium (5%).

It follows that, for metal m, the rate at which the metal leaves the combustion chamber, ER,,,,
is given by:

                  = (FRJ(PF/100)(0.126) g/s.                            (9-4)
The fraction of FJR,,, that is removed by the APCD, CEn, is dependent upon the design. EPA
has compiled a table of efficiencies for different APCD devices (Ref. (7), Table 6).  CE,,, varies
from as little as 0 up to 99%, depending on the metal and upon the design of the APCD. The
final emission rate of metal m from the stack, ER,,, 3 is given by:
                   = (ERJd-CEJIOO)  g/s.                              (9-5)

Acid Gases

The presence of halogenated organics and/or sulfur in the waste feed can cause the formation
of acid gases during incineration.  The acid gases of primary interest are hydrogen chloride
(HC1), hydrogen fluoride (HF), hydrogen bromide (HBr) and sulfur dioxide (SO2). By knowing
the concentration and molecular weights of the compounds containing acid forming elements in
the waste, the aggregate concentration CA of each element A can be determined by:

                CA  = EJI,  (CA)i)(AWA)/(MWA;   ppm                   (9-6)


where CA ; is the concentration of compound i containing acid forming element A
      AWA is the atomic weight of the acid-forming element
      MWA>i is the molecular weight of compound i that contains acid-forming element A.

If the ultimate analysis of waste samples has been performed, CA will be available from the data
and Eq.(9-6)  need not be used.
                                         9-5

-------
The feed rate FRA of each acid forming element is then given by

             FRA  = (FR)(CA)(10-6) Ib/h.                                   (9-7)

The rate of formation of acid gases from halogenated waste is calculated by assuming that the
total mass of each acid forming element combines with hydrogen in stoichiometric proportions
to form the acid gases. Likewise, it is assumed that all sulfur compounds combine with oxygen
to form SO2.  The mass of gas formed per unit mass of element, RA, is  1.013 for bromine,
1.028 for chlorine,  1.053 for fluorine and 1.998 for sulfur.  The rate ERA at which acid gas A
enters the APCD is given by:

             ERA  = (FRJCRJCO.ne) g/s.                                 (9-8)

Hazardous waste incinerators are  equipped with acid gas scrubbers in order to control  HC1
emissions.  Typical control efficiencies CEA reported for wet scrubbers are 99% for HC1 and
HF,  90%+ for SO2 and  unknown  for HBr.

The rate of emission of acid gas containing element A from the stack, EA)S is given by:

             ERA)S = (ERJO-CE^lOO) g/s.                               (9-9)

Paniculate Matter

Under existing RCRA standards, the incinerator paniculate matter emissions must not exceed
a stack concentration of 0.08 gr/dcsf  (180 mg/dcsf) corrected to 7% oxygen in the  stack gas
(50% excess air).  Thus,  the allowable emission rate of PM, ERpM is calculated by the equation:

              ERPM = (0.08 grains/dscf)(Q0)(0.00108)  g/s                   (9-10)

where Q0 is the gas  flow rate at 7%  oxygen in the stack gas and 0.00108 is  the conversion
factor from grains/min to g/s.
                                         9-6

-------
9.2.3         Example of Source Term

Ref. (7)  contains a detailed worksheet and an example of the use of the above calculational
techniques.  The following is a simplified version of that example.

The site  area to be remediated by incineration contains 66,600 tons of soil contaminated with
272 ppm of PCBs,  0.08% (800 ppm) of chlorine (from the ultimate analysis) and 778 ppm of
lead.  Emission rates will be calculated based on average feed conditions assuming that soils are
blended before incineration.

The proposed incineration system works at 35 Btu/h with a feed rate  capacity of 6 tons/hr for
soils with a moisture content of 10% or less.  The APCD and general configuration are shown
in Figure 9-1.  The ejector  scrubber is considered equivalent to  a  venturi scrubber  with a
pressure  drop of 25" of water.  The kiln is to be operated at  a temperature of 1600 °F and the
SCC at 2200 °F.  The system will be operated 24 hours per day for 6 days per week and will
take 540 days to  complete the remediation.   The stack  is  to be located  300 meters  from the
fenceline and will be 8 meters tall.

The feed rate for PCBs  (Eq. (9-1)) is:

              FRpcB =  (12,000 lb/hr)(272 ppm)(ia6) = 3.264 Ib/h

The PCB emission rate from the stack is (Eq. (9-2)):
                    = (3.264 lb/h)(l - 99.9999/100)(0.126)
                     = 4.113xlO-7g/s.
The feed rate for lead (Eq.(9-3)) is:

             FRp,,  = (12,000 lb/h)(778 ppm)(10^) = 9.336 Ib/h.

The rate at which lead enters the APCD (Eq. (9-4)) is:

                   = (9.336 lb/h)(100/100)(0. 126) = 1.176g/s
                                          9-7

-------
where, as noted above, the PF for lead is always 100%.  According to Table 6 of Ref. (7), a
typical value for the removal efficiency of a venturi scrubber is CEpb = 96%.  The rate at which
lead is emitted from the stack is given by (Eq.  (9-5)):

             ERp,,,, = (1.176 g/s)(l - 96/100) =  0.047 g/s.

The feed rate for chlorine (Eq. (9-7)) is:

             FRa  =  (12,000 lb/h)(800 ppm)(lQ-6) =  9.6 Ib/h.

The rate at which HC1 enters  the APCD  (Eq.  (9-8)) is:

             ERHCI = (9.6 Ib/h)(l.028)(0.126) =  1.244 g/s.

As noted above, a wet scrubber will typically remove 99 % of the HC1. The release rate of HC1
from the stack (Eq.  (9-9) is:

             ERHa,3  = (1-244 g/s)(l - 99/100) =  0.0124 g/s.

Thus, the quantities needed for input to the atmospheric dispersion analysis are 4.113x10~7 g/s
of PCBs, 0.047 g/s  of lead and 0.0124 g/s of HC1.  As described above, the stack height is 8
m, the stack internal diameter is 0.5 m, the stack exit velocity is 10 m/s and the stack gas exit
temperature is 344 K.

9.2.4  Atmospheric Dispersion Analysis

This case can be run with a  Gaussian  dispersion model that  contains a plume rise  option.
TSCREEN is suitable  for determining the short-term ambient atmospheric concentrations.  The
data given above are sufficient to define the source term. TSCREEN also requires a definition
of terrain (flat in a rural area)  and a specification of building dimensions for the calculation of
downwash or wake effects.  In this case,  there are no buildings nearby.  For this particular
problem, TSCREEN predicts  a one-hour dispersion factor at 300 m downwind (the fence line)
Of 197.5 /xg/m3 per  g/s released.
                                          9-8

-------
Thus, the one hour ambient concentration of PCBs in air at the fenceline is (197.5)(4.113xlO"7)
= 8.12xlO'5 /Ag/m3.  Similarly, that for lead is 9.29 ^g/m3 and that for HC1 is 2.46 /xg/m3.

9.3    Fires  at Ground Level

In this  section,  fires at ground  level are considered:  first,  a  burning liquid spill of  PCB
contaminated waste (Section 9.3.1) and, second,  burning tires (Section 9.3.2). Section 9.3.3
describes the  use of TSCREEN to model these scenarios, including a discussion of uncertainties.

9.3.1  Burning Pool

In this section, a burning liquid spill of PCB contaminated waste is taken as an example.  Ref.
(27) is a source of information on this kind of spill and the work in this section draws heavily
on that reference.

The calculation of the characteristics of the source term requires the following:

       a)    estimation of the heat of combustion and the liquid density of the mixture that is
             spilled

       b)    determination of the surface area of the spill, the burn rate and the duration of the
             fire (see below)

       c)    calculation of the combustion  product rates  of formation or the rate at which
             unburned toxic materials become airborne (see below); and

       d)    estimation of the heat output of the  fire for the purpose of performing plume rise
             calculations.
9.3.1.1       Characteristics of Pool Fires

The Burn Rate

From Ref.  (27), the burn rate BR (kg/min) of a pool of area A m2 is given by:

                                          9-9

-------
             BR = (A)(pL)(BV)/1000                                       (9-11)

where PL  is the liquid density (kg/m3)  and BV is  the burn velocity (mm/min.).  Note that
conservatisms can be introduced into Eq. (9-11) by choosing conservative burn velocities.  Burn
velocities  for individual chemicals are given in  a publication by the U.S. Coastguard(28).  A
weighted  average burn velocity can  be obtained  by multiplying the burn velocities of the
components by their mass fraction in the mixture. A burn velocity of zero may be assumed for
PCBs because they  will only burn in the presence of other flammable materials.  If data are
unavailable, the following correlation  can be used(27):

             BV = (0.076)(HC)/(HV) mm/min                               (9-12)

where Hc  is the heat of combustion (J/kg) and 1^ is the latent heat of vaporization (J/kg).
In calculating the burn rate, Ref.(27) considers three cases:

i)     Burn rate = release rate

      This case is  appropriate for an undiked spill or for a case in which the application of
      Eq.(9-ll) to a diked area would  give a bum rate  that exceeds the release or spill rate.
      In  such cases, it can be assumed that the spilled liquid spreads just far enough for the
      burn rate to equal  the spill rate.   In  this case,  the area A  does not  enter  into the
      calculation.  The spill rate  equals the burn rate and  the duration  of release equals the
      duration of the spill.

ii)    Spillage of a Small Mass of Liquid

      If a mass of liquid is spilled that is too small to cover the diked area to a  depth of 1 cm,
      the assumption is made that the liquid spreads until it has a depth of 1 cm.  Thus, if the
      mass spilled is M kg, the volume V spilled is simply the mass divided by the density, V
       =M/ PL m3-  The area occupied by the spill is then the volume divided by the depth, A
       =  V/0.01 m2. The burn rate can then be calculated from Eq. (9-11) and  the duration of
       release tR is  simply M/BR.
                                          9-10

-------
iii)    Rapid Spillage of a Large Mass of Liquid

      Assume a rapid spill of liquid M that fills the diked area A to a depth D = M/{(PL)(A)}
      m. The burn rate can then be calculated from Eq. (9-11) and the duration of release is
      tR =  (D)/(1,OOOBV).

Product Rates of Formation

Once the bum rate is known, the product rates of formation can readily be calculated from a
knowledge of what fraction of the burning material forms each kind of product. Ref. (27) comes
to the following conclusions on the basis of a review of the literature on PCB fires:

      o     PCBs: one weight percent of the PCBs involved in a fire are entrained without
             decomposition

      o     polychlorodibenzofurans (PCDFsV.  0.5 weight percent of the PCBs  involved in
             a fire are converted to 2,3,7,8-tetrachlorodibenzofuran

      o     potychlorodibenzodioxins (Peeps'): 0 weight percent of the PCBs involved in a
             fire are converted  to PCCDs:  however,  0.01  weight  percent  of  chlorinated
             benzenes  (if present as in  some commercial  products  containing  PCBs)  are
             converted to 2,3,7,8-tetrachlorodibenzo-p-dioxin

      o     phosgene: 0 weight percent of the hepta- or less chlorinated biphenyls involved
             in a fire form phosgene; 0.8 weight  percent of the octa- or higher  chlorinated
             biphenyls are converted  to phosgene

      o     hydrogen chloride: the chlorine content of the PCBs that is not entrained and does
             not form PCCDs, PCDFs or phosgene is converted to  hydrogen chloride in a fire.

The above estimates are intended to be  realistic  deductions from experiment;  the authors of
Reference (27) do not appear to have deliberately sought  conservatism.
                                         9-11

-------
Heat Output

The heat output Qh in J/s is given by the product of the burn rate BR in kg/s and the latent heat
of combustion H (j/kg).  This quantity can then be used in the plume rise calculations.

9.3.1.2      Example

Consider the case of a commercial product known as Aroclor 1260.  This  has the following
weight composition of PCBs: 12% C12H5C15, 38% CnH4Cl6, 41%  C12H3C17, 8% C12H2C18 and
1 % CijHCL,.  Assume that this is present in a mixture of 1 weight percent Aroclor and  99
weight percent transformer oil.  This  means that the following masses of PCBs are present in
1 Kg of Aroclor: 0.0012 kg C12H5Cl5,  0.0038 kg C12H4C16, 0.0041  kg C12H3C17, 0.0008  kg
C,2H2C18 and 0.0001  kg C^HCL,.  Ref.(27) gives a liquid density pL of 788 kg/m3 and a bum
velocity BV of 4.0 mm/min.  This mixture is present  in an outdoor storage tank that fails
catastrophically, leading to 50 m3 of liquid being spilled over a diked area of 200 m2. The pool
ignites and burns.

From Eq. (9-11), the burn rate BR is given by:

             BR =  (200)(788)(4)/(1000)
                = 10.5 kg/s.

The PCBs are present at 1 weight percent = 0.105 kg/s.  According to the rules given above:

       o      1% of the  PCBs become entrained without decomposition at a rate of 0.00105
              kg/s = 1.05 g/s.

       o      0.5% of the PCBs are converted to PCDFs at a rate of 0.5 g/s.

       o      0.8% of the octa- or higher chlorinated biphenyls are converted into phosgene.
              These  highly  chlorinated PCBs  constitute  9%  by mass of the  total  PCBs.
              Therefore,  the rate of conversion of PCBs to phosgene is (0.09)(0.008)(0.105) =
              7.2x10-' kg/s = 0.072  g/s.
                                         9-12

-------
      o      The total mass M of chlorine that is present in 1  kg of Aroclor can readily be
             shown to be 0.006 kg.  This follows by taking the above stated masses of PCBs
             in 1 kg of Aroclor and multiplying them by the mass fraction of chlorine in each
             PCB. The mass fractions are as follows:  0.54 for C12H3C13, 0.59 for C12H4C16,
             0.625 for C12H3C17, 0.657 for C12H2C18 and 0.684 for C12HC19.   These mass
             fractions are calculated as follows: CI2H5C15 has a molecular weight of 12x12  +
             5x1 + 5x35 = 324, of which the chlorine makes up 5x35 =  175. Therefore, the
             ratio  is 175/324 = 0.54.   The fractions for  the other PCBs are calculated
             similarly.  The mass of chlorine liberated per kilogram of Aroclor is therefore
             0.54x0.0012 + 0.59x0.0038 + 0.625x0.0041 + 0.657x0.0008'+ 0.684x0.0001
             = 0.006 kg.  This forms (0.006)x(molecular weight of HCl)/(atomic weight  of
             chlorine) = (36)(0.006)/(35) = 0.00619 kg HCl/kg of Aroclor.  The emission rate
             of HC1 is therefore  (rate of burning of Aroclor  = 10.5 kg/s)(mass of HCl/kg of
             Aroclor = 0.00619) = (10.5)(0.00619)  ~ 65 g/s.

The  heat of combustion of oils is typically around 5xl07  J/kg(28), so that the  rate of heat
production is predicted to be (rate of burning of Aroclor = 10.5 kg/s)x(heat  of combustion)  =
(10.5)(5xl07) = 5.25 x 108 J/s.

The total mass spilled is (50 m3)(788 kg/m3) = 39,400 kg.  The burn rate  calculated above is
10.5  kg/s or  (10.5)x(60)  =  630 kg/min so that the duration of release is  (total mass
spilled)/(burn rate) =  39,400/630 - 62  min.

See Section 9.3.3 for details on how to run this scenario  in TSCREEN.

9.3.2  Burning Tires

Considerable concern has been expressed about the topic of burning  tires.  The following
information is needed in order to generate information for dispersion modeling:

             what compounds are emitted when tires burn?
             what are the emission factors (mg of compound per kg of tire burned)?
             what is the rate  of burn (kg/s)?
             what is the amount of heat released per unit mass (J/kg)?
                                         9-13

-------
A search has been carried out to determine the best available answers to these questions.  The
summary given below shows that there are still considerable uncertainties, particularly in the
calculation of the rate of burn.

9.3.2.1      Compounds and Emission Factors

The EPA publishes and routinely updates a document entitled "Compilation of Air Pollution
Emission Factors"  (AP-42).  Radian Corporation recently undertook a study<42) of emission
factors for a) open burning  of scrap tires; b) open burning of non-agricultural waste; c) open
burning of inorganic agricultural waste; and d) open burning of organic agricultural waste.  The
purpose was to provide additional emissions factors for inclusion in (AP-42).

The Radian study found one earlier experimental study of emissions from open air  burning of
scrap  rubber tires'4^ (steel belted tires were not tested).  Experiments  were performed for both
large  "chunks"  of tires and  smaller slices or "shredded" tires.

The experimental work indicates that emissions from burning scrap tires are dependent on the
bum rate of the tire, with a  greater potential for emissions at a lower bum rate when the tire is
smoldering,  rather than when it is burning out of control.  Oxygen transport controls the rate
of burn.  Gaps  between the tires provide the major avenue for oxygen transport.

Compounds  emitted from scrap tires burning in the open include particulate matter, aromatic
hydrocarbons, alkenes, dienes, sulfonated compounds and nitrogenated hydrocarbons.  These
compounds can conveniently  be grouped into the  following categories: organic compounds.
polycyclic aromatic hydrocarbons (PAHs) and particulate metals.  Tables  (9-1) through (9-3)
give the emission factors that were extracted from the experiments in  Ref. (43).  These factors
are weighted averages over two days of burning and could be higher for a prolonged,  slow burn.

9.3.2.2       Heat Release per Unit Mass

Ref.(44) gives  the heat of combustion of rubber tires as 34 MJ/kg and makes reference to
"Tire  Storage," Loss Prevention Data Sheet 8-3 issued by the Factory Mutual System.
                                         9-14

-------
                Table 9-1.  Polycyclic Aromatic Hydrocarbon
                Emission Factors From Open Burning Tiresa

                        Emission Factor Rating:  D
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(A)pyrene
Benzo(B )fluoranthene
Benzo(G,H,I)perylene
Benzo(K)fluoranthene
Benz(A)anthracene
Chrysene
Dibenz(A,H)anthracene
Fluoranthene
Fluorene
Indeno( 1 ,2,3-CD)pyrene
Naphthalene
Phenanthrene
Pyrene
Chunkb»c
mg
kg tire
718.20
570.20
265.60
173.80
183.10
36.20
281.80
7.90
48.30
54.50
42.30
43.40
58.60
0.00
28.00
35.20
Ib
1000
tons tire
1436.40
1140.40
531.20
347.60
366.20
72.40
563.60
15.80
96.60
109.00
84.60
86.80
117.20
0.00
56.00
70.40
Shredded***
mg
kg tire
2385.60
568.08
49.61
115.16
89.07
160.84
100.24
103.71
94.83
0.00
463.35
189.49
86.38
490.85
252.73
153.49
Ib
1000
tons tire
4771.20
1136.17
99.23
230.32
178.14
321.68
200.48
207.43
189.65
0.00
926.69
378.98
172.76
981.69
505.46
306.98
aReference 42.
bO.OO values indicate pollutant was not found.
cValues are weighted averages.
                                     9-15

-------
                    Table 9-2. Participate Metals Emission
                      Factors From Open Burning Tiresa

                         Emission Factor Rating: C
Tire Condition:
Pollutant
Aluminum
Antimony
Arsenic
Barium
Calcium
Chromium
Copper
Iron
Lead
Magnesium
Nickel
Selenium
Silicon
Sodium
Titanium
Vanadium
Zinc
Chunkb
mg
kg tire
3.07
2.94
0.05
1.46
7.15
1.97
0.31
11.80
0.34
1.04
2.37
0.06
41.00
7.68
7.35
7.35
44.96
Ib
1000
tons tire
6.14
5.88
0.10
2.92
14.30
3.94
0.62
23.61
0.67
2.07
4.74
0.13
82.00
15.36
14.70
14.70
89.92
Shredded5
mg
kg tire
2.37
2.37
0.20
1.18
4.73
1.72
0.29
8.00
0.10
0.75
1.08
0.20
27.52
5.82
5.92
5.92
24.75
Ib
1000
tons tire
4.73
4.73
0.40
2.35
9.47
3.43
0.58
15.99
0.20
1.49
2.15
0.40
55.04
11.63
11.83
11.83
49.51
Reference 42.
b Values are weighted averages.
                                     9-16

-------
                     Table 9-3. Emission Factors For Organic
                    Compounds From Open Burning Of Tires2
                            Emission Factor Rating:  C
Tire Condition:
Pollutant
1,1 hipheayl, methyl
Ihfluocene
1- methylnaphthalene
2-methyinaphthalene
Acenaphihalene
Benzaldehyde
Benzene
Benzodiazine
BenzoAiran
Benzochiopheao
Benzo(B)thioDhene
Benzsisothiazoie
Bipheoyl
Butadiene
Cyanobenzene
Cyclopentadiene
Dihydroindene
Oiniethyl benzene
Dimethyl hexadiene
Dimelhyl naphthalene
Dimethyldihydro iodene
Ethenyl, dimethyl benzene
Ethenyl, methyl benzene
Ethenyi benzene
Ethenyl cyclobexene
Ethenylmethyl benzene
Ethyenylmethly benzene
Ethyl, methyl benzene
Ethyl benzene
Ethynyl, methyl benzene
Ethynyl benzene
Heptadiene
Hexahydro izepinone
Indene
Ijocyano benzene
Isocyano naphthalene
Limonene
Methyl, ethenyl benzene
Methyl, melhylethenyl benzene
Methyl, methylethyl benzene
Methyl benzaldehyde
Methyl brazeo*
Methylcyclohex«M
Methyl haxadiena
Methyl indene
Methyl, methylethyl benzene
Methyl naphthalene
Methyl, ptcpyi benzene
Methyl thicphene
Methytene indene
Methylethyl benzene
Phenol
Propenyi, methyl benzene
Propenyl naphthalene
Propyl benzene
Styiea*
Tetnmethyl benzene
Thiophene
Tricfalorofluacomethane
Thmethyl benzene
Trimethyl naphthalene
Chunkb
me Ib
kg tire
12.71
191.27
299.20
321.47
592.70
223.34
1526.39
13.12
40.62
10.31
50.37
0.00
. 190.08
117.14
203.91
67.40
9.82
323.58
6.22
35.28
5.02
11.50
12.48
539.72
4.85
103.13
0.00
79.29
138.94
459.31
259.82
6.40
64.35
472.74
283.78
10.75
48.11
21.15
35.57
109.69
0.00
1129.80
3.91
15.59
50.04
11.76
144.78
0.00
4.39
30.37
41.40
337.71
0.00
26.80
19.43
618.77
0.00
17.51
138.10
19SJ9
0.00
1000 tons tire
25.42
382.54
598.39
642.93
1185.39
446.68
3052.79
26.23
81.24
20.62
100.74
0.00
380.16
234.28
407.62
134.80
19.64
647.16
12.44
70.55
10.04
23.01
24.95
1079.44
9.70
206.26
0.00
158.58
277.87
918.62
519.64
12.79
128.69
945.48
567.55
21.51
96.22
42.30
71.13
219.39
0.00
2229.60
7.83
31.18
100.07
23.52
289.56
0.00
8.78
60.75
82.79
675.41
0.00
53.59
38.87
1237.53
0.00
35.02
276.20
391.18
0.00
Shreddedb
ms Ib
kg tire
0.00
315.18
227.87
437.06
549.32
322.05
1929.93
17.43
0.00
914.91
0.00
151.66
329.65
138.97
509.34
0.00
30.77
940.91
73.08
155.28
27.60
196.34
21.99
593.15
89.11
234.59
42.04
223.79
335.12
345.25
193.49
42.12
764.03
346.23
281.13
0.00
2309.57
67.05
393.78
1385.03
75.49
1395.04
33.44
102.20
286.68
114.33
122.68
30.14
10.52
58.91
' 224.23
704.90
456.59
0.00
215.13
649.92
121.72
31.11
0.00
334.80
316.26
1000 tons tire
0.00
630.37
455.73
874.12
1098.63
644.10
3859.86
34.87
0.00
1829.82
0.00
303.33
659.29
277.95
1018.68
0.00
61.53
1881.83
146.15
310.57
55.20
392.68
43.98
1186.31
178.22
469.19
84.07
447.58
670.24
690.50
386.98
84.24
1528.05
69147
562.25
0.00
4619.14
134.10
787.56
2770.07
150.98
2790.08
66.88
204.40
573.36
228.66
245.37
60.28
21.03
117.82
448.46
1409.80
913.18
0.00
430.26
1299.84
243.44
62.22
0.00
669.59
63152
•Reference 42.
D0.00 values indicate (be pollutant was not found.
Values are weighted averages.
                                       9-17

-------
9.3.2.3       Rate of Burning

Recently, experiments have been carried out at the Lawrence Livermore Laboratory on the
burning of tires(43). There were four tests, with 48, 9, 12 and 12 tests respectively for tests
1 through 4.  The average rate of burn was 0.028, 0.021, 0.0214 and 0.02 kg/s respectively.
The peak rate of bum was 0.03, 0.0233,  0.03 and 0.04 kg/s respectively.  It is notable that
these  numbers are almost independent of the number of tires involved. In discussions with
the person who carried out the experiments, it became apparent that an extensive literature
search had failed to reveal any better data. Clearly, although the  above data appear to be
the best that are available, there are still considerable uncertainties, including the need to
confirm whether there is indeed only a  week dependence of burn rate on the number of
tires.

9.3.2.4       Example

For the purposes of estimating plume rise, a low rate of heat release is conservative (i.e. a
low rate of burn translates into a low height of rise and relatively  high ground level
concentrations.  The estimated rate of heat  release at  a  bum rate  of 0.02  kg/s  is
(0.02)x(34,000,000)  = 680,000 J/s = 0.68 MW.

For the purposes of estimating emissions, a higher burn rate is the most conservative.  For
example, taking the emission factor for  acenaphthene from Table (9-2) gives an emission
rate of (0.04 kg/sec)x(718.2 mg acenapthene/kg) = 28.7 mg/sec = 0.0287 g/s. This release
rate is suitable for the burning of whole tires, as  was the case in the experiments. The burn
rate data given above are suitable for chunks of tires, not for shredded tires.
9.3.3   Atmospheric Dispersion Modeling

Both of the above scenarios (burning pool and burning tires) can be approximately modeled
in TSCREEN, because they are similar to a scenario that already exists in TSCREEN,
namely a gaseous release from a flared source (Section  4.2.1  of Reference (3)).  That
scenario requires as input the emission rate (e.g 0.072 g/s of chlorine from the burning pool
or 0.0287 g/s of acenapthene from burning tires), the total rate of heat release, Hr (5.25x10*
J/s for the burning pool or 6.8X105 J/s for the burning tires) and the release height above
                                        9-18

-------
the ground,  Hs (a nominal 1 m for both scenarios,  for example).  TSCREEN takes this
information  and calculates an effective height of the flare tip Hsl, where:
             Hsl  = Hs + 4.56xlO-3(Hr/4.1868)a478                        (9-13)

The concentrations downwind are then calculated using the elevated Gaussian model with
an allowance for further plume rise.

The  use of  TSCREEN  to model the two ground  level scenarios is. a considerable
simplification. Both are area sources, whereas Eq. (9-11) and the subsequent use of the
Gaussian model apply to a point source. This could lead to an overestimation of the height
of rise and therefore to an underestimation of ground level concentrations close in and an
overestimation of ground level concentrations far away from the source. However, the point
source assumption neglects initial dilution by air drawn in over the area source  and thus, in
this respect, tends  towards conservatism.

The uncertainty in the emission rates is at least as large as the uncertainty in the dispersion
calculations, particularly in the case of the burning tires where the mass being burned is not
known exactly.  For the  pool case,  moisture  content is also an important source of
uncertainty.

In summary, the  prediction of the dispersion of materials from  fires on the ground is
difficult.  The method outlined above is a considerable simplification and refinement of the
model may be possible in  the future.
                                       9-19

-------
10.0   VAPOR RELEASES FROM MECHANICALLY DISTURBED SOIL

10.1 Introduction

Excavation and removal of soils contaminated with Volatile Organic Compounds (VOCs)
is a common practice  at Superfund sites.   Excavation and removal may itself be  the
preferred remediation option, or it may be a necessary step in a remediation  approach
involving treatment. The purpose of this Section is to acquaint the reader with the methods
that are available with which to estimate the air impacts of the excavation of contaminated
soil.

A description of recent EPA work on this subject is to be found in References (8) or (29).
First, the model assumes an idealized excavation procedure in which soil is scooped out of
a pit in rectangular blocks which retain their shape and are typically of dimension Imx2
mxlm. The  scoops are then stacked in rectangular piles. Over a period  of one hour,  150
cubic meters of soil may be removed and stacked in a pile of dimensions 5 mxlO mx3 m,
leaving a pit  that is 1m deep and of area 15 mxlO m. (Note that Ref. (8) performs sensitivity
studies on some of these variables.)  Figure (10-1) displays this  idealized excavation
scenario.

Ref.(8) considers emissions from:

      o      exposed waste in the evacuation pit

      o      material as it is dumped from the excavation bucket; and

      o      waste/soil in short-term disposal piles.

There are several approaches to the estimation of the rate  of emission  of VOCs during
excavation. The best method is to measure the emissions during full-scale or pilot-scale  soil
handling activities.  The next best method is to estimate the emissions using predictive
equations  with site-specific inputs. If site-specific inputs are not available,  a  conservative
estimate can be made using default values for the parameters. Ref. (8) gives equations for
the following:
                                       10-1

-------
                                Time Between Scoops is
                               Approximately 40 Seconds.
Figure 10-1. Idealized Excavation Scenario
                10-2

-------
             o     average long-term emission rate
             o     short-term emission rate (both detailed and simplified); and
             o     worst-case (i.e.maximum) instantaneous emission rate.

The equations are derived in Appendix (A) of Ref.(8). They are reproduced below to give
the reader an overview of the  variables  to  which values must be assigned  in order to
calculate emission rates.

10.2   Equations for Emission Rates

10.2.1 Average Long-Term Emission Rate

A simple  check on the  total long term emissions potential of the site can be made by
dividing the total mass of a given contaminant to be removed by the expected duration of
the clean-up:

             ER = (Sv)(C)(/3)(l)/tR                                    (10-1)

where        ER is the average emission rate (g/s)
             Sv is the volume of contaminated soil to be excavated (m3)
             C is the average contaminant concentration (/tg/g)
             j3 is the bulk density of the soil (g/m3)
             1 (unity) is a constant (g/10Vg*106cm3/m3); and
             tR is the duration of remediation (s).

10.2.2 Average Short-Term Emission Rate (Detailed Model)

Ref.  (8) gives a detailed model for the average emission rate
(ER g/s) from excavation:

             ER = ERre + ERDIFF                                     (10-2)
where ERpg is the emission rate from the soil pore space, and ER^^ is the emission rate
from diffusion.
                                       10-3

-------
     is given by the following equations.

                  = (P)(MW)(106)(Ea)(Q)(ExC)/{(R)(T)}               (10-3)
where        P is the vapor pressure of the contaminant (mm Hg)
             MW is the molecular weight of the contaminant (g/g-mol)
             Ea is the air-filled porosity (dimensionless)
             Q is the excavation rate (m3/s)
             106 is a conversion factor from m2 to cm 2
             ExC is the soil to gas exchange constant (dimensionless)
             R is the gas constant (62,631 mm Hg-cm3/g-rnol/K); and
             T is the  temperature (K).

Equation (10-3) is based on the assumption that the soil pore gas is saturated with  the
compound of interest If this is not the  case, then Eq.(10-3) may overpredict the emission
rate.  The output of Eq. (10-3)  should be multiplied by  the  duration  of  release and
compared to the total mass M of contaminants in the soil:

             M = (C)(SV)(106) g                                        (10-4)

If Eq.  (10-3) gives  a value that  exceeds M/3,  then the  following equation should be
substituted for Eq. (10-3):

             ERre =  (M)(0.33)/(tsv)                                    (10-5)

where tsv is the time in seconds taken to excavate a given volume of soil.

ERDtFF is given by the following equations:

             ERD(FF = (C)(10,000)(SA)/(B)                             (10-6)

             B = (EJ/UK^Xk,)} + [(Tt)/{(De)(Keq)}]1/2                 (10-7)

where the symbols not already defined above are as follows:
                                        10-4

-------
             SA is the area of the emitting surface (m2)
             C is the concentration  of the contaminant in soil (g/m3)
             10,000 is a conversion  factor from m2 to cm2
             K^ is the equilibrium coefficient (dimensionless)
             kg is the gas-phase mass transfer coefficient (cm/s); and
             De is the effective diffusivity in air (cm2/s).

10.2.3 Simplified Average Short Term Emission Rate

The equations for ERre and ERDffF can be simplified as follows (as shown in Appendix A
of Ref. (8)):

             ERre =  (P)(Q)(0.98)                                     (10-8)

where P and Q are defined above and 0.98 is a conversion factor (g/mm  Hg/m3): and

             ERD(FF  = (C')(10,000)(SA)/(B)                            (10-9)

             B = (1.22)(106)(C')/(P) + [(1.79)(109)(C')/(P)]1/2           (10-10)

where the symbols have already been defined above except for C': = (C)/(/3)/(106) /*g/g,
the concentration of the contaminant in soil.  1.22xl06 is a conversion factor (cnr-s-mm
Hg/g) and 1.79xl09 is another conversion factor (s2-cm-mm Hg/g).  Note, however, that the
soil concentration term C in this equation is now  in /tg/g rather than in g/cm3 as was the
case above.

10.2.4 Worst-Case (Instantaneous) Emission Rate

The worst-case instantaneous emission rate occurs when  the exposed surface area is at a
maximum and immediately after a  bucket load of soil is dumped onto the storage pile.  This
emission rate can be  approximated  by  considering the case in which a pure chemical is
exposed to the atmosphere,  in which  case the maximum emission rate ER^^ can be shown
to be:

                    = (kg)(P)(MW)(SA)(10,000)/{(R)(T)}               (10-11)

                                       10-5

-------
10.3  Example

The reader is referred to Ref. (8) for detailed discussions of values for the variables above.
However, a brief example follows.

10.3.1 Description of Problem

Assume that 2 m3 of soil are removed per scoop in a block of dimension Imx2 mxlm.  75
scoops per hour are moved  (Q =  150 m3/hr  =  0.042 m3/s).  The excavation pit has
dimensions 10 mxl5 mxl  m and, after 1 hour, the storage pile has dimensions 5 mxlO mx3
m.  The total surface area exposed (pit plus pile) is 290 m2. See  Figure (10-1).

Assume that the site has approximately 10,000 m3  of soil contaminated by chloroform at
concentrations C in soil of 0.1 pg/g. The bulk density of the soil, £, averages 1.5 g/cm3. The
removal is expected to take 20 days (tR = 1.728xl06 s).

10.3.2 Total Emissions Potential for the Site

Using equation (10-1),  Sv =  10,000 m3, C =0.1 /*g/g, 0 = 1.5 g/cm3 and tR =  1.728xl06 s:

             ER = (10,000)(0.1)(1.5)(l)/(1.728xl06)
               =  8.68x10^ g/s

10.3.3 Average Short Term Emission Rate

The rate of emission from the pore space, ERps is calculated from Eq. (10-8)  in which the
vapor pressure P of chloroform is 208 mm Hg (p.26 of Ref. (8)) and Q =  0.042 m3/s:

             ERps = (208)(0.042)(0.98) = 8.56 g/s

The total  amount of chloroform predicted  to be emitted   over  1 hour  (3,600  s)  is
(8.56)(3,600)  = 30,820 g.  However, the total mass of the contaminant present in the soil
is given by the total mass of soil excavated in one hour (150 m3) times the bulk density (1.5
g/cm3 = 1.5xl06 g/m3) times the concentration of the chloroform in soil (0.1 jttg/g = 10"7
g/g): this product  is  22.5 g.   Therefore, Eq. (10-1) overpredicts and, as described above,
                                        10-6

-------
defaults to Eq.(10-5) in which M is 22.5 g, and tsv is one hour = 3,600 s:

                ps =  (22.5)(0.33)/(3,600)  = 2.08xlQ-3 g/s.
The emission rate due to diffusion is calculated from Eqs (10-9) and (10-10) in which C =
10'7 g/g, SA = 290 m2, and P = 208 mm Hg. Substituting these values into Eqs (10-9) and
(10-10) gives:
             ERD1FF = (10-7)(10,000)(290)/B
             B = (1.22xl06)(10-7)/(208) + [(1.79xl09)(10-7)/(208)]1/2

             ERDIFF = 0.269 g/s

Thus, the total predicted release rate is 0.269 + 0.002 = 0.271 g/s.

10.4  Atmospheric Dispersion

The small rates of release calculated above mean that the scenario should be run with the
Gaussian dispersion model,  specifically EPA's TSCREEN model for short-term (1 hour)
worst-case ambient conditions. EPA's Industrial Source Complex (ISCLT) can be used for
long term concentrations.

To run TSCREEN, the following assumptions and input variables are needed:

      o      The combined emission rate for the excavation pit and the storage pile  is
             0.271 g/s

      o      The excavation pit and the pile are sufficiently close together to allow the
             user to assume that the area of the emissions source is equal to the combined
             horizontal areas of the pit and storage pile

      o      the terrain  is flat without any nearby structures; and

      o      downwash or building wake effects are not applicable.
                                        10-7

-------
Running  TSCREEN  for scenarios  representing  ground level  area  sources,  such as
TSCREEN scenario #2.9, shows that the maximum hourly ambient air concentration at (for
example) a distance of 400 m is approximately 760 ^eg/m3.  This can then be compared to
action level ambient air concentrations.
                                     10-8

-------
11.0   REFERENCES

1.     U.S. Environmental Protection Agency, "Guidelines on Air Quality Models," EPA-
      450-4-78-027R (NTIS PB 86-245248), Research Triangle Park,NC (1988).

2.     U.S. Environmental Protection Agency, "A Workbook of Screening Techniques for
      Assessing the Impacts of Toxic Air Pollutants," EPA 450/4-87-013 (NTIS PB 87-
      227452), Research Triangle Park, NC (1988): "Workbook on Screening Techniques
      for Assessing Impacts  of Toxic Air Pollutants  (Revised),"  EPA 454/R-92-024,
      Research Triangle Park, NC (1992).

3.     U.S. Environmental Protection Agency, "User's Guide to TSCREEN, a  Model for
      Screening Toxic Air Pollutant Concentrations," EPA 450/4-89-013 (NTIS PB 91-
      141820), Research Triangle Park, NC (1988):

4.     Spicer, T. and Havens, J.A., "  User's Guide for  the DEGADIS 2.1 Dense Gas
      Dispersion  Model,"  EPA-450/4-89-019, United States  Environmental  Protection
      Agency, Research Triangle Park (November, 1989).

5.     U.S. Environmental Protection Agency, "Evaluation  of Dense  Gas Simulation
      Models," EPA 450/4-90-018, Research Triangle Park, NC (1988).

6.     U.S. Environmental Protection  Agency, "Guidance on the Application of Refined
      Dispersion  Models for  Air Toxics  Releases,"  Source  Receptor Analysis Branch,
      Technical Support Division, Research Triangle Park, NC (1991).

7.     U.S. Environmental Protection Agency, "Screening Procedures for Estimating the Air
      Impacts of  Incineration at Superfund Sites," EPA  Contract No. 68-02-4466 (Work
      Assignment No. 91-77) JTN 803770-077-02, Research Triangle Park,  NC (1991).

8.     U.S. Environmental  Protection Agency,  "Estimation  of  Air  Impacts for  the
      Excavation  of Contaminated Soil,"  EPA 450/1-92-004, Research Triangle Park, NC
      (1992).
                                      11-1

-------
9.     U.S. Environmental Protection Agency, "Estimation of Air Impacts for Air Stripping
      of Contaminated Water," EPA 450/1-91-002, Research Triangle Park, NC (1991).

10.    U.S. Environmental Protection Agency, "Contingency Plans at Superfund Sites Using
      Air Monitoring," EPA 450/1-90-005, Research Triangle Park, NC (1988).

11.    "Atmospheric Science and Power Production," (Darryl Randerson, Ed.), DE84005177
      (DOE/TIC-27601),  published  by the National Technical Information Service,
      Springfield, VA (1984).

12.    Pielke, R.A., "Mesoscale Meteorological Modeling," Academic Press (1984).

13.    American  Industrial  Hygiene  Association,  "Emergency  Response Planning
      Guidelines," Akron,  Ohio.

14.    U.S. Environmental Protection  Agency,  "Health  Effects Assessment Summary
      Tables," Washington, D.C. (Updated Quarterly).

15.    American Meteorological Society Workshop, "Stability Classification Schemes and
      Sigma  Curves  - Summary of  Recommendations,"  Bulletin of the  American
      Meteorological Society 58 (1977)  pp 1305 - 1309.

16.    Useful summaries of some of the basic formulae to be used for calculating release
      rates can be found in "Guidelines  for Chemical Process Quantitative Risk Analysis,"
      Center for Chemical Process Safety of the American Institute of Chemical Engineers,
      New York, New York (1989).

17.    "Workbook of Test Cases for Vapor Cloud Source Term Dispersion Models,"  Center
      for Chemical Process Safety of the American Institute of Chemical Engineers, New
      York (1987).

18.    "Handbook  of Chemistry and Physics" (Robert  C.  Wease,  Ed.), CRC  Press,
      Cleveland, OH (54th Edition, 1973 - 1974).
                                       11-2

-------
19.    Ermak,  D.L. "User's  Manual for the SLAB Model, An Atmospheric Dispersion
      Model for Denser-than-Air Releases," Lawrence Livermore Laboratories (1989).

20.    "LNG Vapor Dispersion Prediction with the DEGADIS Model," Topical Report GRI
      89/0242, Gas Research Institute, Chicago, IL (1990).

21.    Bird, R.B., W.E. Stewart and E.N. Lightfoot, "Transport Phenomena," John Wiley &
      Sons (1960).

22.    Phani K. Raj, "Chemical Release/Spill Source Models - A Review," International
      Conference  and Workshop on  Modeling  and Mitigating the Consequences of
      Accidental Releases of Hazardous Materials (New Orleans, 1991), Published by the
      American Institute of Chemical Engineers, New York (1991): "Models for Cryogenic
      Liquid Spill Behavior on Land and Water," J. Haz. Mat. 5 (1981) 111 - 130.

23.    Raj, P.K., Hagopian, J. and Kalelkar, A.S., "Prediction of the Hazards of Anhydrous
      Ammonia Spillage onto Water," CG-D-74-74, prepared for the use of the U.S.
      Coastguard by Arthur D.Little, Inc., Cambridge, MA (1974).

24.    Cavanaugh, T.J., II, J.H. Siegell and K.W.Steinberg, "Simulation of Vapor Emissions
      from Liquid Spills," Paper Presented at the 85th Annual Meeting & Exhibition of the
      Air and Waste Management Association, Kansas City, Missouri (June, 1992).

25.    Shaw, P. and F. Briscoe, "Evaporation from Spills of Hazardous Liquids on Land and
      Water,"  United Kingdom Atomic Energy Authority Report SRD R120 (1978)

26.    Webber, D.M.  and  SJ.Jones,  "A  Model  of Spreading,   Vaporizing  Pools,"
      International Conference on Vapor Cloud Modeling (Boston, 1987): published by the
      Center for Chemical Process Safety of the American Institute of Chemical Engineers,
      New York (1987).

27.    Zamjec, E.R. and Chin C. Chao, "Consequence Analysis of PCB-Containing Liquid
      Pool Fires," Presented at the AIChE Spring Meeting, Orlando,  FL (March, 1990).
                                      11-3

-------
28.    U.S. Coastguard, "Chemical Hazardous Response Information System (CHRIS),
      Hazardous Chemical data," Washington, D.C. (1984).

29.    Eklund, B., S. Smith, J.F. Durham and J.S. Touma, "Estimation of VOC Emissions,
      Ambient Air  Concentrations  and  Health  Effects  from  the   Excavation  of
      Contaminated Soil," Presented at the 85th Annual Meeting & Exhibition of the Air
      and Waste Management Association, Kansas City, Missouri (June, 1992).

30.    Clough, P.N., D.R. Grist and C.J. Wheatley, "The Mixing of Anhydrous Hydrogen
      Fluoride with Moist Air," in the Proceedings  of the International' Conference  on
      Vapor Cloud Modeling, John Woodward,  Ed.,  American Institute  of Chemical
      Engineers (November  1987).

31.    Schotte,  W.,  "Fog Formation of HF  in Air,"  Industrial  Engineering Chemical
      Research 26(1986)300  .

32.    Resplandy, A., "Etude  Experimentale des Proprietes de PAmmoniac," Chim. Ind. -
      Gen. Chim. 102 (1969) 691.

33.    Wheatley, C.J., "Discharge of Liquid Ammonia to Moist Atmospheres - Survey of
      Experimental Data and Model  for  Estimating  Initial Conditions in Dispersion
      Calculations," United Kingdom Atomic Energy Authority  Report SRD/HSE/R4K)
      (1987).

34.    Fauske, H.K. and  M.  Epstein, "Source Term Considerations in Connection with
      Chemical Accidents and Vapor  Cloud Modeling," Presented at the  International
      Conference on Vapor Cloud Modeling, November 2-4,  1987, (Sponsored by the
      Center for Chemical Process Safety, the American Institute of Chemical Engineers
      and the U.S. Environmental Protection Agency).

35.    U.S. Environmental Protection Agency, "User's  Guide for the Industrial Source
      Complex (ISC2) Dispersion Models." EPA-450/4-92-008, Research Triangle Park. N'C
      (1992).
                                       11-4

-------
36.    Spicer, T.O., "Supplement to: A Workbook of Screening Techniques for Assessing
      Impacts of Toxic Air Pollutants," Preliminary Report Prepared for the Office of Air
      Quality Planning and Standards, U.S. Environmental Protection Agency, Research
      Triangle Park, NC (July, 1992).

37.    Sandier, S.I., "Chemical and Engineering Thermodynamics," 2nd Edition, John Wiley
      and Sons,  New York (1989).

38.    Perry, R.H., D.W.Green and J.O. Maloney, "Perry's Chemical Engineers' Handbook,"
      6th Edition, McGraw-Hill, New York (1984).

39.    Lewitt, E.H., "Thermodynamics Applied to Heat Engines,"  5th Edition, Sir  Isaac
      Pitman and Sons, London (1953).

40.    Lees, P.P., "Loss Prevention in the Process Industries,", Butterworths, London (1980).

41.    Spicer, T.O., "A Screening Procedure to Estimate the Release Rate of Low Volatility
      Liquids from Tanks and Pipes," Prepared for the Office of Air Quality Planning and
      Standards, U.S. Environmental  Protection Agency,  Research Triangle  Park,  NC
      (May, 1992).

42.    "Emission Factor Documentation for AP-42, Section 2.4, Open Burning," Prepared
      by Radian Corporation for EPA under Contract No.  68-DO-0125, Office of  Air
      Quality and Planning Standards, Research Triangle Park, NC (May, 1992 - Draft).

43.    "Characterization of Emissions from the Simulated Open Burning of Scrap Tires,"
      Prepared by Acurex Corporation for the EPA, Office of Air  Quality and Planning
      Standards, Research Triangle Park, NC (1989).

44.    Yung,  D. and J.R.Mehaffey,  "Fire  Resistance Requirements for Rubber-Tire
      Warehouses," Fire Technology, May 1991, pp 100-112.

45.    Hasegawa, H., Lawrence  Livermore National Laboratory,  Privae Communication
      (November 12, 1992).
                                      11-5

-------
46.    Design Institute for Physical Property Data, "Data Compilation - Tables of Properties
      of Pure Compounds," T.E.Daubert and R.P.Danner (Editors), American Institute of
      Chemical Engineers,  New York, NY (1985).
                                       11-6

-------
        APPENDIX A
EXAMPLE AND INTERPRETATION
            OF
      DEGADIS OUTPUT

-------
DEGADIS Output File

Example DEGADIS output files are provided for both a transient and steady state release
simulation. Numerically indexed explanatory notes are given for each main section of the
output listings.
                                      A-2

-------
Tha following la • lilting of tha output for a transient ralaaaa aimulation of chlorina.

Output Listing lotea

Vertical Vapor Chlorina Jat Transient Release Simulation


1.      Tha data and  time  tha aimulation wai input and run ia raportad.

2.      Tha Input .supplied by tha uaar  ia tapaatad for documantation of tha  aimulation.   Tha  tltla  block,  atmoapharic  condition! and adiabatic mixing
        of tha contaminant ralaaaa ia shown.

3.      Tha contaminant gaa properties, aa apacifiad by tha user, ara documantad.
                                                        t
4.      Tha input aourca characteristics ara diaplayad.   Tha  aourca  input data point* section  liata tha initial mass in tha cloud, tha contaminant mass
        rata, aourca  radius, contaminant maaa fraction, tamparatura, and anthalpy.

5.      Numarical parameters and calculation flags uaad by DEGADIS  it  liatad.  Thaaa valuaa ara  aat by tha intaractiva modula of DEGADIS (JETINT) and
        tha numaclcal paramatar fllaa Includad with tha coda.  Tha latt thraa linaa in this aaction ara »«t by tha uaar and document if tha aimulation
        la laotharmal and  whathar haat  and water tranafar ara includad.

6.      A Hating of tha calculatad aourca paramatera aa a function of time  ia  provldad hara.  Includad ara tha  aacondary gaa radiua, height, aourca maaa
        flux (Q, the contaminant  mole fraction  (Mole frac C), the
        gaa mixture danaity, temperature, and the Richardaon  number  baaed on  the  cloud apreading velocity (Rich Ho.).  The aource calculation ends when
        all the primary and aecondary aource gaa haa been taken up in the atmoapharic flow.

7.      Documentation ia provldad to indicate if x-direction diaperalon correction wa»  Included  in  the aimulation.   If applied.  Identification of the
        conatanta uaad in  tha x-diraction dlaparaion correction ia diaplayad.

8.      Tha concentretion  field ia given for different timea  after apill initiation.   Tha timea given in thla  output Hating are default valuea aet by
        the numerical parameter file EXAMPLE.ER3.  To chooae  different timea at which tha concentration field  la output, change the appropriate valuea
        in HUH_NAME.E83 and execute OEG3 again.

9.      For eech time Increment, the downwind portion of the  calculations la  Hated.   Tha eleven columns contain the following information:

                Column 1  - Distance  downwind of  the aource;
                Column 2  * Mole  fraction;
                Column 3  - Contaminant  concentration;
                Column 4  - Mixture denaity;
                Column 5  - Gamma -  (p'pt)ICt. where p  - contaminant  denaity,  p. - denaity  of tha air,  and Cc -  contaminant concentration;
                Column 6  - Temperature  of the mixture on-the centerline of the gas cloud  at ground  level;
                Column 7  - Half  width,  the  contour ahape paramatar  b;
                Column 8  - S,, a contour shape parameter;
                Column 9  - Sr, a contour shape parameter;
                Column 10- Width  from the centerline to  the user  apeciflad lower level  of Interest  at the  user  specified height;  and
                Column 11- Width  from the centvrllne to  the user  specified upper level  of interest  at tha  user  specified height.

10.     A report of the mass of contaminant between the upper and  lower level of  intereat  la provided  for each time increment.  The mass of contaminant
        above the lower level of interest is also given.


                                                                        A-3

-------
0**e****************

 ***************


 Date input on
 Source progrem run on

0          TITLE BLOCK
UOA_DEOADIS   MODEL   OUTPUT   --   VERSION   2.1

          *************** 25-JUN-1992 10:59:10.41  ***************

       25-JUN-1982 10.58: 8.  0
       25-JUN-1992 10:59:10.41
 CL2 VAPOR RELEASE
 VERICAL JET SIMULATION
 LARGE SCALE: 1,000 kg; 2" critic*
«=  1


0
0
0




0

0



Wind velocity «t reference height
Reference height
Surface roughness length
Pa«ijuiU Stability class
Honln-Obukhov length
Gaussian distribution constants
Specified averaging time
Oeltey
Betey
Wind velocity power lew constant Alphs
Friction velocity
Ambient Temperature
Ambient Pressure
Ambient Absolute Humidity
Ambient Relative Humidity















Adlebetlc Mixing: Hole fraction CONCENTRATION










0



.00000
.05568
. 12093
.198*8
.29214
.40753
. 55320
.74286
1.00000
Specified Gas Properties:
Molecular weight:
Release temperature:
kg/m«*3
.00000
. 17086
.37113
.60910
.89654
1.25065
1.69769
2.27973
3.39696



1.50 B/S
10.00 m
.100 B
F
17.5 m

1200.00 s
.07742
.80000
.44905
.07195 m/s
.278.00 K
1.000 atm
4.009E-03 kg/kg BOA
75.00 X
OF C GAS DENSITY
kg/m**3
1.26655
1 . 36690
1.48451
1.62427
1.79308
2.00105
2.26358
2.60541
3.39896

70.000
251.00 K















Enthalpy
J/kg
.00000
.00000
.00000
.00000
. 00000
.00000
.00000
.00000
.00000


















Temperature
K
278.00
278.00
278.00
278.00
278.00
276.00
278.00
276.00
251.00



                                                                                                                                              «= 3
                                                                         A-4

-------
          Density at releaae temperature end ambient preaaure:
          Average heat capacity:
          Upper mole fraction contour:
          Lower mole fraction contour:
          Height for laopletha:
 3.3990    k$/m**3
 .00000    J/kg K
2.00000E-05
3.00000E-06
 1.6000    m
 Source input data point*

                Initial meaa in cloud:    .00000
                                                                                                                                              «= 4
  Time

    a
.00000
662.00
663.00
664.00
Contaminant
 Maaa Rate
  1.5100
  1.5100
  .00000
  .00000
Source Radiua

      m
  19.686
  19.696
  .00000
  .00000
                                                          Contaminant
                                                         Maaa Fraction
                                                        kg contain/kg mix
                                                           3.62986E-03
                                                           3.62986E-03
                                                           3.62966E-03
                                                           3.62986E-03
          Temperature

               K
            278.00
            278.00
            276.00
            278.00
Enthalpy

  J/kg
.00000
.00000
.00000
.00000
0     Calculation procedure for ALPHA:  1
0     Entrainment prescription for FBI:  3
0     Layer thickneaa ratio uied for average depth:     2.1300
0     Air entrainnent coefficient uaed:  .590
0     Gravity (lumping velocity coefficient uaed:  1.150
0     NOH laothermal calculation
0     Beet tranafer not included
0   * Water tranafer not included
                                                  CALCULATED SOURCE PARAMETERS
                                                                                                                                              <= 6
Time
aec
. 000000
3.77626
10.6993
•17.6226
36.9309
56.2382
92.9689
129.699
149.042
168.386
363. 045
Gaa Radlua
m
19.6955
20.0691
21.4156
23.1726
28.6984
34.8794
45.7382
55.5388
59.4749
60.9302
60 0111
Height
m
1.100000E-OS
.792351
2.06330
3.05618
4.64252
5.17819
4.97146
4.14046
3 69906
3.51850
3.46528
Qatar
kg/n>**2/e
2.489397E-04
2.452237E-04
2.329614E-04
2.192248E-04
1.867220E-04
1.650777E-04
1.423013E-04
1.321261E-04
1.301492E-04
1.300341E-04
1.337128E-04
SZU-L/2.)
m
3.35682
3.40271
3.56532
3.77162
4.40255
5.00345
5.96491
6.69385
6.94315
7.02180
6.90624
Mole free C
1.501292E-03
1.477573E-03
1.399892E-03
1.313851E-03
1.115344E-03
9.887324E-04
8.663783E-04
8.264913E-04
8.26B289E-04
8.326102E-04
8.637744E-04
Denaity
kg/m**3
1.26926
1.26922
1.26906
1.26892
1.26856
1.26834
1.26812
1.26804
1.26804
1.26805
1.26811
Temperature
K
276.000
278.000
276.000
278.000
278.000
278.000
278.000
278.000
278.000
278.000
278.000
Rich No
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
                                                                         A-5

-------
   658.484
                 59 8289
                               3.45473
                                            1.343098E-04   6.88599
                                                                        8.686844E-04   1.26812
                                                                                                     278.000
                                                                                                                   .756144
687. 4IU
724.000
7U2.348
708.022
827 591
847. 4U1
857 860
8(>0.0tt2
8UO . 1 50
54. B/67
46.5404
36.0291
24 5/54
13 8486
5.95016
1.58070
.612091
.601683
3.16929
2 68807
2 . 08066
1.41924
.790743
.343585
9.126908E-02
3.454256E-02
3.058242E-02
1
1
9
0
6
6
5
5
5
227504E-04
081975E-04
. 413527E-05
126567E-05
. 997320E-OS
.OB9663E-05
.405815E-05
. 157967E-05
. 148170E-05
6.80812
6.45340
5.72198
4.61030
3.22167
1.84430
.749643
.391354
.386510
7.
5.
4.
3.
3.
2.
2.
2.
2.
403055E-04
980279E-04
794928E-04
861275E-04
149216E-04
642615E-04
296968E-04
179467E-04
175158E-04
1.26789
1.26763
1.26742
1.26725
1.26712
1.26703
1 . 26697
1.26695
1.26695
278.
278.
278.
278.
278.
278
278
278.
278
000
,000
.000
000
.000
.000
.000
.000
.000
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
.756144
   8B0.71B       .601883        .000000      5.116335E-05
0     Sorted vatuea foi ••ch (pacified tin*.
0     X l>iiaction collection waa applied.
           Coefficient:         .17000
           Power;               .97000
           Minimum Dl«t«nc»:    50.000     o
                                                           .386519
                                                                        2.166874E-04   1.26694
                                                                                                     278.000
                                                                                                                    .000000
                                                                                                                                        «= 7
Time aftei beginning of apill   171.0000
         Mole   Cuncentreti
       Ki action
     (m)

   2J5.
   2tit>
   2UU.

  *335.
       1 04UE-03  3.21UE-03
       7.B82E-04  2 419E 03
       4.142E 04  1.271E 03

       1 075E-04  3 300E-04
ana 1 ty

f/B**3)
1.2684
1 2680
1.2673
Genraa


.587
.587
.587
Ta
-------
   291.
             7.518E-04  2.307E-03
                                     1.2679
                                                .587
                                                           278.
                                                                      52.2
                                                                                 4.86
                                                                                            23.0
                                                                                                       105.
                                                                                                                  84.8
326.
363.
401.
441.
483.
5
V
2
1
5
994E-04
364E-04
716E-04
394E-04
178E-05
1.839E-03
1.339E-03
8.336E-04
4.279E-04
1.589E-04
1.2676
1.2673
1.2670
1.2668
1.2666
.587
.587
.587
,587
.587
278.
278.
278.
278.
278.
54.0
53.8
52.0
46.7
23.0
4.94
5.10
5.57
6.52
8.60
27.8
32.0
36.3
40.0
38.0
117.
124.
128.
124.
86.2
104.
108.
109.
100.
58.3
 For tha ULC of   2.00000E-03 mola parcant,  and th« LLC of   3.00000E-04  nola parcant:

 Tha D«a> of contaminant batwaan tha ULC and LLC la:   2.4413     kg.
 Th* ma** of contaminant abova tha LLC la:     262.13     kg.
0 Tlma aft-ar baginntng of (pill
0 Dlatanca Mola Concentration
Fraction
(m)
220.
251.
263.
318.
354.
392.
432.
473.
515.
559.
604.
650.

1
9.
a
7.
5.
4
3.
2
1
1
6.
3.

113E-03
827E-04
493E-04
058E-04
711E-04
534E-04
S39E-04
.684E-04
900E-04
212E-04
7S5E-05
101E-05

-------
Tine after beginning of aplll
Distance Hole Concentration
Fraction
(•) (kg/n**3)
2.
2.
2.
2.
2.
2.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
S84E+03
65BE+03
735Et03
8HE+03
888E+03
865E+03
043E+03
122E+03
202E+03
282Et03
362Et03
444E+03
526E+03
608E+03
691E+03
775E+03
859Et03
844Et03
030E+03
116E+03
202E+03
290E+03
377E+03
466E+03
555E+03
644E+03
734E+03
825E+03
916E+03
3
4
4
5
6
6
7
7
a
a
a
a
a
7
7
7
6
6
6
5
5
5
4
4
3
3
3
2
2
. 273E-06
. 008E-06
.763E-06
.498E-06
. 198E-06
. 819E-06
. 342E-06
.731E-06
.016E-06
. 172E-06
.208E-06
. 153E-06
.022E-06
.8176-08
. 583E-06
.280E-06
. 957E-06
.6126-06
.255E-06
.868E-06
. 482E-06
. 074E-06
. 663E-06
.241E-06
.825E-06
.406E-06
.003E-06
. 609E-06
.240E-06
1.
1.
1.
1.
1.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
2.
1.
1.
1.
1.
1.
1.
1.
1.
8.
a.
6.
004E-05
230E-05
462E-OS
687E-05
802E-05
083E-05
253E-05
373E-05
460E-05
508E-05
519E-05
S02E-OS
462E-05
399E-Q5
327E-05
234E-05
135E-05
028E-OS
820E-05
a01E-05
682E-05
557E-05
431E-05
302E-05
174E-05
045E-05
216E-06
007E-06
8736-06
2169.000 aec
Denaity Gamma
(kg/n**3)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.2666
.529
.581
.585
.585
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.586
.585
.585
.585
.585
.584
.584
.582
 T«np»r«tur«
    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
    278.

    278.
    278.
                                                                     Half
                                                                     Width
                                                                      (o)

                                                                     .000
                                                                     .000
                                                                     12.7

                                                                     37.9
                                                                     56.9
                                                                     71.7

                                                                     83.3
                                                                     91.5
                                                                     96.0

                                                                     96.4
                                                                     95.4
                                                                     94.4

                                                                     93.3
                                                                     92.3
                                                                     91.2

                                                                     90.1
                                                                     89.0
                                                                     87.9

                                                                     86.6
                                                                     85.4
                                                                     83.9

                                                                     81.0
                                                                     76.4
                                                                     69.9

                                                                     61.6
                                                                     51.0
                                                                     35.7

                                                                     14.9
                                                                     .000
 Sz

 (a)

46.3
41.2
38.1

37.9
38.1
36.3

38.6
39.0
39.6

40.3
41.1
41.9

42.7
43.5
44.4

45.2
46.0
46.8

47.6
48.4
49.3

50.2
51,2
52.2

53.3
54.4
55.7

57.4
60.5
                                                                                            Sy
                                                                                           128.
                                                                                           151.
                                                                                           178.

                                                                                           200.
                                                                                           216.
                                                                                           229.

                                                                                           240.
                                                                                           249.
                                                                                           256.

                                                                                           261.
                                                                                           265.
                                                                                           270.

                                                                                           275.
                                                                                           279.
                                                                                           284.

                                                                                           288.
                                                                                           293.
                                                                                           297.

                                                                                           302.
                                                                                           306.
                                                                                           310.

                                                                                           314.
                                                                                           317.
                                                                                           318.

                                                                                           318.
                                                                                           316.
                                                                                           311.

                                                                                           301.
                                                                                           287.
  Width at t-
3.000E-0
      (in)

     36.2
     80.3
     132.

     192.
     240.
     278.

     309.
     333.
     348.

     356.
     361.
     363.

     364.
     364.
     363.

     360.
     356.
     351.

     344.
     335.
     324.

     307.
     285.
     255.

     216.
     161.
1.60 m to:
2.000E-03moleZ
  (m)
For th» ULC of   2.00000E-03 moU percent,  and the LLC of   3.00000E-04 mole  percent:

                                                                kg.
The maia of contaminant betwaen the ULC and LLC la:
The rnaas of contaminant above the LLC ia:     459.65
459.65
  kg.
                  A-8

-------
Output Llatlu* lots*

Acetone Confined Spill Steady State Release Simulation



1.      Th« data and tint*  tha •inulation waa input and run 1* report ad.                                                   >

2.      Tha input  auppliad by tha  uaar  la  rapaatad for documentation of tha  aioulation.   Tha  title block,  aUnoapheric conditiona and adiabatic mixing
        of tha contaminant ralaaaa is ahown.

3.      Tha contaminant gaa properties, aa apecifled by the uaer, are documented.

4.      The input  source characteristics are displayed.   The  source  input data points section  lists  the initial mass in the cloud, the contaminant mass
        rate, source radlua, contaminant masa fraction, temperature, and enthalpy.  For a ateedy  state releaae,  there is  no initial mass in the cloud
        and the source parameters  are held constant for an arbitrarily Ions time period.

5.      numerical  parameters and calculation flags used by DEGADIS  is  listed.  These values ere  set by the interactive module of DEGADIS (DE6INP) and
        tha numerical parameter files included with the code.  The lest three lines in this section are eat by tha user end document if the simulation
        is isothermal and  whether heat  and water transfer are included.

6.      A Hating of the calculated source  parameters as a function of time  ia provided here.  Included are tha secondary gas  radius, height, source mass
        flux (QlUr). the vertical concentretlon distribution parameter at tha downwind edge (SZ(x-L/2)), the contaminant mole  fraction  (Mole free C), tha
        gaa mixture density, temperature, and the Richardson number based on the cloud spreedlng velocity (Rich No.).   For e steady state release, the
        aource calculations terminate once the parameter* no longer change with time.

7.      A summary  of tha steady state primary and secondary aource ia provided.

6.      The downwind portion of'the calculations is listed.  The eleven columns contain tha following information:

                Column  t  ~ Distance downwind of  tha source;
                Column  2  - Mole  fraction;
                Column  3  - Contaminant  concentration;
                Column  4  - Mixture density;
                Column  5  - Gamma • (p"P.)/Cc, where p  •  contaminant  density,  p. ~ density  of the air,  and Cc - contaminant concentration;
                Column  6  - tempereture  of  the mixture on tha centerllne of the gas  cloud  at ground level;
                Column  1  - Self width,  tha contour shape parameter  b;
                Column  8  - S,, a contour shape parameter;
                Column  9  - S,, e contour shape parameter;
                Column  10- Width from the  centerline  to  the uaar apecified lower level of intereat at the  user  specified  height; and
                Column  11- Width from tha  centeriine  to  the user specified upper level of interest et the  user  specified  height.

        Output continues until the centerline ground level concentration ia below the user specified  lower level-of interest.

0.      A report of the maas of contaminant between the  upper and lower level of interest  is provided  for each time increment.  The mass of contaminant
        above the  lower level of interest  is slso given.
                                                                         A-9

-------
0*******************
 Data input on
 Source program tun on

0          TITLE BLOCK
UOA_DEGADIS   MODEL   OUTPUT

          *************** 26-JUH-1982  9:31:  4.60


       26-JUH-1982  9:30:52.47
       26-JUN-1992  9:31: 4.60
 CONFINED SPILL INTO A DIKED AREA
 STEADY STATE ACETONE SPILL. POOL DIAMETER 10m
                                                                                     V E 8 S I  O H   2.1
                                                                                                                 *******************
                                                                                                                      ***************
                                                                                                                                              «= 1
      Reference height
      Surface roughneaa length
      Pasqulll Stability claaa
      Monin-Obukhov length
      Friction velocity
      Ambient Temperature
      Ambient Pressure
      Ambient Absolute Humidity
      Ambient Relative Humidity

      Adiabatic Mixing:
'erence height

ingth
ilaaa
i
in constants
icified averaging time
Deltay
Betay
law constant Alpha



kidity
lidity















Mole fraction CONCENTRATION

.00000
.00676
.02290
.04689
.08244
.12619
.18301
.25294
.33574
.43037
.54491
.67955
kg/m**3
.00000
.01717
.05776
.11708
.20288
.30510
. 43266
, 58208
.74904
.92777
1.12862
1.34547
1.50
10.00
.100
F
17.5

1200.00
.07742
. 90000
.44905
.07195
278.00
1.000
4.009E-03
75.00
•/a
B
m

m

s



a/a
K
atm
kg/kg BDA
Z
OF C GAS DENSITY













kg/m**3
1.26655
1.27159
1.28349
1.30088
1.32604
1.35602
1.39342
1.43723
1.48618
1.53859
1.S9749
1.66107















Enthalpy
J/kg
.00000.
.00000
.00000
. 00000
. 00000
.00000
. 00000
.00000
.00000
.00000
.00000
.00000















Temperature
K
278.00
278.00
278.00
278.00
278.00
278.00
278.00
278.00
278.00
278.00
278.00
278.00
                                                                        A-10

-------
                             .817*1
                                                1.54856
                                                                    1.72062
                                                                                       .00000
                                                                                                          278.00
      Specified But Properties:

          Molecular weight:                                        58.080
          Release temperature:                                     278.00    K
          Density at releaae temperature and ambient  praaaure:      1.7920    kg/m**3
          Avactga heat capacity:                                   59*.97    J/kg K
          Upper nole fraction contour:                             .15000
          Lower mole fraction contour:                            2.00000E-OS
          Height for laoplatba: *                                   .00000    m
 Source input data point*

                Initial mass in cloud:
                                                                                                                                             «= 4
        Time
      .00000
      60230.
      60231.
      60232.
                      Contaminant
                       Maaa Rata
                          kg/.
                        .20100
                        .20100
                        .00000
                        .00000
                                          .00000
                                        Source Radiua
5.0000
S.OOOO
.00000
.00000
  Contaminant
 Maaa Fraction
kg contain/kg mix
    .90000
    .90000
    .90000
    .90000
Temperature

     K
  278.00
  278.00
  278.00
  278.00
Enthalpy

  J/kg
.00000
.00000
.00000
.00000
0     Calculation procedure for ALPHA:  1
0     Enfcrainment prescription for FBI:  3
0     Layer thickness ratio uaad for average depth:     2.1500
0     Air entrainment coefficient uaad:  .590
0     Gravity (Lumping velocity coefficient uaed:  1.150
0     NO* laotharmal calculation
0     Heat tranifer not included
0     Water transfer not Included
                                                                                                                                             «= 5
                                                  CALCULATED SOURCE  PARAMETERS
                                                                                                                                             «= 6
     Time
     aac
                Gaa Radiua
   602.300       5.00000
   1806.90       5.00000
OSource strength (kg/s) :
 Equivalent Primary aource length (m)

 Secondary aourca concentretlon (kg/m**3)
Height
m
.000000
.000000
th (m) :
Qatar
kg/m**2/a
2.5592UE-03
2.559211E-03
.20100
10.000
SZU-L/2.)
m
. 528482
.528482
Equivalen
Equivalen
                                                                      Mole  free  C
                                                                                       Danaity
                                                                                       kg/m**3
                                               .17518
                              7.085218E-02   1.31792
                              7.085218E-02   1.31792
                              imary source radius (ml  :
                 Equivalent  Primary source half-width  (m)

                 Secondary aource SZ (ml  :
                                           Temperature
                                                K

                                             278.000
                                             276.000
                                                  5.0000
                                                  3.9270

                                                  .52848
                                                                                                                  Rich  No.
                                                                                                                   .000000
                                                                                                                   .000000
                                                                        A-ll

-------
 Contaminant flux rate:   2.55921E-03
 Secondary source mate fraction*... contaminant;    .132924
            Enthalpy:    .00000          Denalty:     1.3178
                                                                         .86361
 Secondary aourc* length [•)
                                               10.000
Secondary cource half-width [ml
                                                                                                          3.9270
0 Diatanca     Ho la   Concentration Daniity  Temperature   Half        Sz         Sy     Width at z-   .00 • to:
             Fraction                           '           Width                       2.000E-03moleZ  1S.O    molaX
     (•)                (kt/m**3)  (k»/o.**3)     (K)         (m)        «n>         (m)        (n)        (a)
                                                                                                                                             «= 8
s.oo
5.10
5.85
8. 02
10.2
13. 5
16.8
23.2
33.7
57.1
80.5
120.
160.
200.
240.
280.
320.
360.
400.
440.
460.
520.
560.
600.
640.
680.
720.
7.
7.
6.
5.
3.
085E-02
046E-02
696E-02
268E-02
806E-02
2.248E-02
1.
5.
3.
1.
6.
3.
2.
1.
1.
9.
7.
6.
5.
4.
4.
3.
3.
3.
2.
2.
2.
402E-02
743E-03
153E-03
160E-03
402E-04
320E-04
126E-04
S16E-04
158E-04
245E-05
618E-OS
430E-05
531E-OS
830E-05
270E-05
814E-05
436E-OS
119E-05
850E-05
619E-05
418E-05




9.
5.
3.
1.
8.
2.
1.
8.
5.
3.
2.
2.
1.
1.
1.
1.
1.
8.
8.
7.
7.
6.
6.
175
174
166
131
538E-02
670E-02
550E-02
4S9E-02
017E-03
853E-03
630E-03
4S2E-04
412E-04
865E-04
850E-04
354E-04
940E-04
637E-04
406E-04
230E-04
067E-04
711E-05
750E-05
942E-05
257E-05
668E-05
157E-05
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.32
.32
.32
.31
.29
.28
.28
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
.27
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
278.
3.93
3.88
4.64
7.12
9.29
11.9
13.9
17.1
19.1
22.2
23.8
25.1
25.7
25.9
25.6
25.5
25.2
24.8
24.3
23.7
23.1
22.5
21.9
21.2
20.5
IS. 8
19.1
.528
.516
.445
.376
.386
.460
.565
.877
1.20
2.04
2.81
4.00
5.07
6.07
' 7.01
7.91
8.77
9.59
10.4
11.2
11.9
12.6
13.4
14.0
14.7
15.4
16.1
.000
.243
.755
1.63
2.36
3.34
4.24
6.24
7.99
12.1
15,5
20.6
25.1
29.2
33.0
36.6
39.9
43.2
46.3
49.3
52.2
55.0
57.7
60.4
63.0
65.5
68.0
3.93
4.57
6.78
11.7
15.8
20.8
24.7
32.0
37.1
46.5
52.7
59.7
64.3
67.4
69.5
70.8
71.4
71.4
70.9
70.0
68.6
66.7
64.3
61.4
58.0
53.8
48.7
                                                                        A-12

-------
II
o e
II   a

H   ~
e a   w
S S   o
t
     i
           *• o o>
           o o o
X 10 N

ie o ro


sis
i  i i
o o o
5*   •
 §   "
r?   s
"I   r
.5   I
           w M to
           •*j ^ -w
           ex a cc
     o
     o
     o
     e
     n
           O> ^ CB


           ie >i *-
           CB -W Ol


           o u «j
     o
     c»
            N N>


            CO K

-------
        APPENDIX B
EXAMPLE AND INTERPRETATION
            OF
       SLAB OUTPUT

-------
SLAB Output File

The SLAB output file consists of three general categories of information:

       •     problem description,
            instantaneous spatially-averaged cloud properties, and
            time-averaged volume fraction.

An example output file is provided  to illustrate  each category.  Numerically indexed
explanatory notes are given for each main section of the output.
                                        B-2

-------
Output Listing tot**

Vertical Vapor Chlorine J«t Release Simulation
1.
2.
3.
5.
6.
A list of tba various user specified input parameters to define th« simulation ia provided.  This information
ia organized into six aaetions: problem input, releaaa gas properties, apill eharactariatics, field parameters,
ambient meteorological propartiaa, and additional parameters that describe the ipill scanazio.  Hot* that some
of th«  problam input paramatars  may b« changed  by tba cod*  in  ordar to ba  consistant with  SLAB  modeling
assumptions.   Tha  valuas listad  in  tha  othar fiva iactions giva  tha  paramatar valuas actually used  in tha
simulation.

Tha Instantaneously spatially averaged cloud  parameters  giva tha raaults of  tha dispersion calculation phase
of tha  simulation.  Theae intermediate  result*  axe tha  solution* of tha  spatially  avatagad  conservation
equations  (plume or puff), tha equation  of stata  (ideal  gas  law),  and  tha  length and  width equations.   Cloud
meander or time-averaging effects are not included.   In steady  state dispersion mode, during a finite duration
release (t < TSD), apatial averaging  is  over  tha  crosswind  plan* of tha cloud. •  In the puff dispersion mode,
after tha  releaae haa terminated  (t > TSD),  spatial averaging  i* over  tha  entire volume of the cloud.   Refar
to the SLAB User's Manual for further discussion.

The instantaneous spatially averaged cloud property reaults  are listed  in two groupa as a function of downwind
distance and ar* described by tha following cloud parameters:
         (Group  1)
         (Group 2)
                 x
                 zc
                 h
                 bb
                 b
                 bbx
                 bz
                 cv
                 rho
                 t
                 u
                 ua
                 cm
                 cmv
                 cmda
                 cmw
                         ug
                         w
                         V
                         vz
downwind distance (m)
profile center height (m)
cloud height (m)
cloud half-width (m)
half-width parameter (m)
cloud half-length (m)
half-length paramatar (m)
volume fraction of emission
density (kg/or1)
temperature (K)
downwind cloud velocity (m/s)
height averaged ambient wind spaed (m/s)
mass fraction of emission
     fraction of emission vapor
     fraction of dry air
     fraction of water
naas fraction of water vapor
gravity flow velocity, Z-direction (m/s)
grsvity flow velocity, Y-direction (m/s)
gravity flow velocity, X-direction (m/s)
vertical entrainment velocity (m/s)
crosswind horizontal entrainmant velocity (m/s)
downwind horizontal entrainment velocity (m/s)
The time; averaged volume fraction is tha last of tha  three general  categories  of  information provided by the
output file and is generally of primary interest  to the analyst.  From these volume  fractions, the time averaged
concentrations of tha modeled  release can be determined (volume fraction is converted to concentration in parts
per million by multiplying the volume fraction by one million).

Tha time- averaged volume concentration output is  divided into three sections: concentration contour parameters,
concentration in tha z-ip(I) plane, and maximum center Line concentration.  All of the reaults presented at a Cram
th« viewpoint  of an observer located at tha  downwind distance z,  crosswind distance  y from the mean'cloud
centerline, and height s above tha ground.

Tha equation that should be used for dispersion  calculations  ia provided.   The  diapersion calculations can  ba
performed in one of three ways:  steady state plume made, tranaient puff mode, or a combination of both modes.
Depending on the release type, the time averaged volume concentration  is calculated by:

1) steady state plum* mode (x i* independent variable),

e(x,y,z.t) - ec(x) * (erf{x«)-erf(xb)) » (erf(ya)-erf(yb)) • (e*p<-*a*za)+exp(-zb*zb)),

where
                    c(z,y,z,t)
                           concentration (volume fraction) at (x.y.x.t)
                                                         B-3

-------
                         x        - downwind distance (CD)
                         y        • ccoaswind horizontal distance (m)
                         z        - hal«ht (m)
                         t        • tin* (s)

                         erf     • error function
                         xa      - 
                         xb      - (x-xc-bx)/(sr2*batax)
                         ya      • (y+b)/(ar2*betac)
                         yb      - 
                         •xp     • exponential function
                         za      • (z-ze)/(sr2*sig>
                         zb      • 
-------
problem input
idapl -
ncalc •
«na •
cps
tbp •
cmedO •
dhe •
cpsl •
rhosl »
3pb -
spc
ts -
V
aa •
tsd -
qtii -
ha
tav «
XfflB "
Zp(l> "
zp(2) -
zp(3) -
zp(*> -
zO •
za •
ua -
ta •
rh
stab •
3
1
.070910
498.10
239 . 10
.00
2878*0.
926.30
1374.00
1978.3*
-27.01
251.00
1.51
.00
662.
.00
S.OO
1200.00
10000.00
1.00
.00
.00
.00
. 100000
10.00
1.50
278.00
75.00
6.00
                                  (only displays values to tiro decimal places)
 releaae  gaa properties

 molecular weight  of source gaa  (kg)
 vapor heat capacity, conat.  p.   (J/kg-k)
 teoperature  of source  gaa  (k)
 danaity of source gaa  (kg/m3)
 boiling point  temperature
 liquid  oaaa  fraction
 liquid  heat  capacity (j/kg-k)
 heat of vaporization (J/kg)
 liquid  source  density  (kg/m3)
 saturation pressure constant
 saturation preasure constant (k)
 saturation pressure constant (k)
MDS -
cpa •
ta •
rhos •
tbp •
omedO"
cpal -
dhe •
rhosl-
spa •
spb •
spc -
7.0910E-02
4 . 9810E+02
2.5100E+02
3.4429E+00
2.3910E-H32
0 . OOOOE+00
9.2630E+02
2.8784E-KI3
1.3740E+03
9.3278E-KJO
1.9783E-1-03
-2.7010E+01
 spill characteristics

  spill type
  oaaa source rate (kg/a)
  continuous source duration (s)
  continuous source oaas (kg)
  instantaneoua source m»*t (kg)
  source area (o2)
  vertical vapor velocity  (n/s)
  source half width (a)
  source height (m)
  horizontal vapor velocity (m/s)
- idspl-
tsd
qtca
qtis
as
ws
bs
hs
us
   1.5100E+OQ
   S. S2QOE-H)2
   9.9962E-MJ2
   0.OOOOE-MJO
     4300E-03
"  9.8338E+01
•  3.3334E-02
-  5.0000E+00
•  0.OOOOE+00
       -  k.
 field paracMters

  concentration averaging time (s)
* tav  •  1.2000E-K13
                                                         B-5

-------
 mixing layer height (a)
 rn.-r4n.nii downwind diStrSCe (O)
 concentration measurement height (a)
 - has  •  2.60002+02
 - xffln -  1.00002+04
 - zpU)-  1.00002+00
 - zp<2>-  O.QOOOE-KIO
 - zp<3)-  0.00002+00
 - zp(4)-  O.OOOOE+00
ambient meteorological properties

 molecular weight at ambient 
-------
4.282+01
5.07E-MJ1
6.021+01
7 . 182+01
8.392+01
1.03E+02
1.242+02
1.492+02
1.79E+02
2.16E+02
2.61E+02
3 . 15E+02
3 . 82E+02
* . 67E+02
5.73E+02
7 . 09E+02
8.81E+02
1.102+03
1.38E+03
1.742+03
2.19E+03
2.77E+Q3
3.512+03
* . 452+03
S.65E+03
7.16E+03
9.08E+03
1.132+04
1 . 462+04
1.84E+04

z
1 . OOE+00
1.342+00
2.08E+00
2.63E+00
3.172+00
3.712+00
4.232+00
4 . 80E+00
5.342+00
5.88E+00
6.422+00
6 . 56E+00
6.732+00
5.93E+00
7 . 172+00
7.47E+00
7 . 83E+00
8.26E+00
8.79E+00
9.432+00
1.022+01
1.122+01
1.23E+01
1.37E+01
1.54E+01
1.742+01
1.99E+01
2.29E+01
2.632+01
3.09E+01
3.832+01
4.282+01
3.072+01
6.022+01
7.182+01
8.392+01
8.43E-01
4.012-01
3.012-01
2.452-01
2.09E-01
1.83E-01
1.63E-01
1.472-01
1.35E-01
1.24E-01
1.15E-01
1.08E-01
9.96E-02
9.28E-02
8.71E-02
8.242-02
7 . 83E-02
7 . 482-02
7.172-02
6.90E-02
6.652-02
8.432-02
6.232-02
8.03E-02
5.88E-02
5.732-02
5.592-02
5.462-02
5.3*2-02
5.23E-02

em
1. OOE+00
6.44E-01
3.11E-01
1.67E-01
1.012-01
6.7*2-02
*.78E-02
3 . 55E-02
2.74E-02
2.18E-02
1.77E-02
1.77E-02
1.772-02
1.77E-Q2
1.78E-02
1.78E-02
1.73E-02
1.74E-02
1.74E-02
1.732-02
1.72E-02
1.702-02
1.68E-02
1 . S6E-02
1 . 64E-02
1.812-02
1.382-02
1.342-02
1.302-02
1.432-02
1.372-02
1.102-02
7 . 932-03
3. 922-03
4.392-03
3.632-03
2.82E+00
2.37E+00
2.382+00
2.302+00
2.662+00
2.862+00
3 . 10E+00
3.39E+00
3.71E+00
4 . 082+00
4.31E+00
4 . 99E+00
3.822+00
6.39E+00
7.31E+00
8.41E+00
9.892+00
1.122+01
1. 292+01
1.492+01
1.732+01
2.002+01
2.312+01
2.862+01
3 . 07E+01
3.34E+01
4.08E+01
4.70E+01
5.412+01
8.212+01

emv
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1.002+00
-1.002+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
1.77E-02
1.77E-02
1.77E-02
1.77E-02
1.762-02
1.762-02
1.7SE-02
1.74E-02
1.74E-02
1.73E-02
1.722-02
1.70E-02
1.682-02
1.562-02
1.84E-02
1.612-02
1.382-02
1.34E-02
1.302-02
1.432-02
1.372-02
1.102-02
7.932-03
3.922-03
4.392-03
3.632-03
2.40E+01
3.89E+01
3.232+01
6.432+01
7.61E+01
8.74E+01
9.83E+01
1.092+02
1.20E+02
1.312+02
1.43E+02
1.34E+02
1.632+02
1 . 762+02
1.872+02
1.98E+02
2.10E+02
2.222+02
2.332+02
2.48E+02
2.63E+02
2.80E+02
2.992+02
3.202+02
3 . 44E+02
3.71E+02
4 . 03E+02
4 . 38E+02
4 . 792+02
3.26E+02

coda
0 . OOE+00
3.332-01
6.36E-01
8.302-01
3.932-01
9.292-01
9.482-01
9.61E-Q1
9.692-01
9.74E-01
9.78E-01
9.782-01
9.782-01
9. 782-01
9.782-01
9.782-01
9.792-01
9.792-01
9.792-01
9.792-01
9.792-01
9.792-01
9.792-01
9.792-01
9.802-01
9.802-01
9.802-01
9.812-01
9.812-01
9.822-01
9.822-01
9.332-01
9.382-01
9.902-01
9.912-01
9.922-01
2.25E+00
3 . 83E+00
4 . 84E+00
3.94E+00
6 . 97E+00
7.97E+00
8.93E+00
9.87E+00
1.08E+01
1.17E+01
1.26E+01
1.35E+01
1.4*2+01
1.32E+01
1.60E+01
1.682+01
1.73E+01
1.822+01
1.882+01
1.93E+01
2. 012+01
2.082+01
2.14E+01
2.20E+01
2.26E+01
2.322+01
2.37E+01
2.43E+01
2.48E+01
2.33E+01

cam
0 . OQE+QO
1.44E-03
2.78E-03
3. 362-03
3 . 622-03
3 . 762-03
3. 842-03
3 . 89E-03
3 . 92E-03
3.94E-03
3 . 962-03
3.96E-03
3 . 96E-03
3 . 96E-03
3.96E-03
3 . 962-03
3.96E-03
3 . 96E-03
3 . 962-03
3 . 962-03
3.962-03
3.962-03
3.962-03
3.962-03
3 . 962-03
3 . 962-03
3.972-03
3.972-03
3.972-03
3.972-03
3.972-03
3.982-03
4 . 002-03
4.012-03
4.012-03
4.012-03
3.93E+01
4.67E+01
5.37E+01
6.66E+01
7.98E+01
9.38E+01
1.13E+02
1.39E+02
1.67E+02
2.02E+02
2.44E+02
2.95E+02
3.26E+02
3 . 63E+02
4 . 09E+02
4.67E+02
3.38E+02
6.27E+02
7.39E+02
8.79E+02
1.03E+03
1.27E+03
1.34E+03
1 . 38E+03
2.30E+03
2.82E+03
3.46E+03
4.24E+03
3.20E+03
6.37E+03

cmwv
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-l.OOE+QO
-1.002+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
3.98E-03
3 . 96E-03
3 . 962-03
3.96E-03
3 . 96E-03
3 . 962-03
3 . 962-03
3.96E-03
3.962-03
3.962-03
3 . 962-03
3 . 962-03
3 . 96E-03
3 . 962-03
3 . 962-03
3 . 96E-03
3.972-03
3 . 972-03
3.972-03
3.972-03
3.972-03
3 . 982-03
4.002-03
4.012-03
4.012-03
4.012-03
3.93E+01
4.67E+01
5.372+01
S.66E+01
7.982+01
9.38E+01
1.13E+02
1.392+02
1.672+02
2.02E+02
2.44E+02
2.95E+02
3.00E+02
3.052+02
3 . 082+02
3.12E+02
3.14E+02
3.17E+02
3 . 192+02
3 . 202+02
3.212+02
3.222+02
3.23E+02
3.232+02
3 . 24E+02
3 . 24E+02
3.24E+02
3.24E+02
3 . 25E+02
3.252+02

we
9. 362+01
6.74E+01
3.432+01
4.462+01
3.822+01
2.892+01
2.222+01
1.612+01
1.04E+01
5.06E+00
0 . OOE+00
-8.7QE-03
-1.922-02
-3.19E-02
-4.73E-02
-6.582-02
-8.822-02
-1.15E-01
-1.47E-01
-1.86E-01
-2.332-01
-Z. 882-01
-3.54E-01
-4.32E-01
-5.24E-01
-6.322-01
-7.372-01
-9. 032-01
-1.082+00
-1.292+00
-3.912-01
-4.902-02
-1.332-02
-5.452-03
-2.732-03
-1.372-03
4.312-03
3.232-03
2.422-03
1.882-03
1.492-03
1.20E-03
9.79E-04
7.99E-04
8.322-04
5.332-04
4.352-04
3.332-04
2.662-04
1.972-04
1.432-04
1.032-04
7.332-05
3.162-05
3.592-05
2.46E-05
1.682-05
1.132-03
7 . 362-06
5.01E-06
3.312-06
2.172-06
1.412-06
9.192-07
5. 962-07
3 . 862-07

v«
0. OOE+00
0 . 002+00
0 . 002+00
0.002+00
0.002+00
0 . 002+00
0.002+00
0 . OOE+00
0 . 002+00
0 . 002+00
0 . 002+00
0 . OQE+QO
0 . 002+00
0 . 002+00
0 . OOE+00
0. 002+00
0 . 002+00
0 . 002+00
0 . OOE+00
0 . 002+00
0.002+00
0 . 002+00
0.002+00
0 . 002+00
0 . 002+00
0 . 002+00
0 . 002+00
0 . 002+00
0.002+00
0 . 002+00
2.232+00
1.322+00
1.302+00
9.482-01
7.232-01
3.722-01
1.28E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+OQ
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.27E+00
1.272+00
1.27E+00
1.27E+00

u«
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . 002+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+OQ
0 . 002+00
a . OOE+OO
0 . 002+00
0 . 002+00
0 . 002+00
a . 002+00
0 . 002+00
0 . 002+00
0.002+00
0.002+00
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.782+02
2.782+02
2.782+02
2.78E+02
2.78E+Q2
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.782+02
2.78E+02
2.782+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02
2.78E+02

w
-1. OOE+00
- 1 . OOE+00
-1. OOE+00
-1.002+00
-1. OOE+00
-1. OOE+OQ
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
2.09E-02
8.31E-04
6.322-04
6.33E-04
6.33E-04
S.37E-04
S.39E-0*
6.432-04
6.482-04
6.552-04
6.642-04
8.772-04
S.95E-04
7.212-04
7.812-04
8.222-04
9.202-0*
1.102-03
1.502-03
3 . 062-03
3.962-02
3.452-02
3 . 992-02
2.322-02
2.132-02
1.722-02
3.55E-01
8.09E-01
8.03E-01
8.04E-01
8.04E-01
8.06E-01
8.122-01
8.21E-01
8.33E-01
8.49E-01
8.69E-01
8.92E-01
9.23E-01
9.60E-01
1. OOE+00
1.05E+00
1 . 10E+00
1.162+00
1.22E+00
1.28E+00
1.35E+00
1.42E+00
1.49E+00
1 . 56E+00
1.63E+00
1.702+00
1.782+00
1 . 85E+00
1.93E+00
2. OOE+00

V
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1.002+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
-1. OOE+00
3.63E-02
3.62E-02
3.62E-02
3.62E-02
3.62E-02
3.S2E-02
3.62E-02
3 . 62E-02
3.62E-02
3.82E-02
3.62E-02
3.S2E-02
3.62E-Q2
3.S1E-02
3.S1E-02
3.60E-02
3.57E-02
3.46E-02
3.27E-02
2.94E-02
2.11E-02
1.35E-02
1.37E-02
1.292-02
1.222-02
1.172-02
3 41F-01
7 99F-01
7 97E-01
7 97E-01
8.00E-01
8.0AE-01
8.11E-01
3.21E-01
8.34E-01
8.51E-01
3.71E-01
8.94E-01
9.24E-01
9.60E-Q1
1. OOE+00
1.05E+00
1.10E+00
1.18E+00
1.22E*00
1.28E+00
1.35E+00
1.42H-00
1 49E100
1 . 56E+00
1.63E+00
1.70E+00
1.78E+00
1.85E+00
1.93E+00
2. OOE+00
«= 4
vx
0 OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
O.OOE-i-00
0 OOE-M30
0 OOE+OQ
0. OOE+00
3.46E-01
3.46E-01
3.46E-01
3 46E-01
3 46E-01
3.46E-01
3.46E-01
3.46E-01
3.46E-01
3.45E-01
3 45E-01
3.45E-01
3.45E-01
3 44E-01
3.44E-01
3.42E-01
3.39E-01
3.33E-01
3.14E-01
2.34E-01
1. 73E-01
1.61E-01
1.56E-01
1.55E-01
1.572-01
B-7

-------
1.032+02
1.24E+02
1.492+02
1.792+02
2.162+02
2.612+02
3 . 152+02
3 . 822+02
4 . 672+02
5.73E+02
7.092+02
3.812+02
1.10E+03
1.382+03
1.742+03
2.192+03
2.77E+03
3 . 512+03
4 . 452+03
5.652+03
7.162+03
9.082+03
1 . 152+04
1.462+0*
1.8*2+0*

2.952-03
2.402-03
1.962-03
1.602-03
1.312-03
1.072-03
6 . 702-0*
6.532-0*
4.822-0*
3.522-0*
2.53E-0*
1.80E-0*
1.272-0*
8.802-05
6.032-05
*. 112-05
2.772-05
1.852-05
1.232-03
8.112-06
3.322-06
3.472-08
2.262-06
1.462-06
9.472-07

2.952-03
2.402-03
1.962-03
1.602-03
1.312-03
1.072-03
8.702-0*
6.532-04
* . 822-0*
3.522-0*
2.532-0*
1.802-0*
1.27E-0*
8.802-03
6.052-03
4.112-05
2.772-05
1.832-03
1.232-05
8.112-06
3.322-06
3.472-06
2.262-06
1.462-06
9.472-07

9.932-01
9.942-01
9.9*2-01
9.9*2-01
9.932-01
9.932-01
9.932-01
9.952-01
9.952-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01
9.962-01'
9.962-01
9.962-01
9.962-01

4.022-03
4.02E-03
4.021-03
4.022-03
4.022-03
4.022-03
4.03E-03
4.032-03
4 . 032-03
4.032-03
4.032-03
4.032-03
4.032-03
4 . 032-03
4.032-03
4 . 032-03
4.032-03
4 . 032-03
4 . 032-03
4.032-03
4.032-03
4.032-03
4.032-03
4 . 032-03
4.032-03

4 . 02E-03
4.02E-03
4 . 022-03
4.022-03
4.022-03
4 . 022-03
4.032-03
4 . 032-03
4.032-03
4 . 032-03
4.03E-03
4.03E-03
4 . 03E-03
4.03E-03
4 . 032-03
4.03E-03
4.03E-03
4 . 032-03
4.03E-03
4.032-03
4 . 032-03
4.032-03
4.032-03
4.032-03
4 . 032-03

-9.622-04
-6.20E-0*
-4.152-0*
-2.862-0*
-2.022-0*
-1.452-0*
-1.072-0*
-9.172-05
-6.262-05
-4.332-05
-3.032-03
-2.152-05
-1.54E-05
-1.122-03
-8.212-06
-6.082-06
-4.532-06
-3.402-06
-2.562-06
-1.9*2-06
-1.472-08
-1.112-06
-8.412-07
-6.352-07
-4.782-07

4.602-01
3.752-01
3.082-01
2.552-01
2.132-01
1.802-01
1.522-01
1.202-01
9.462-02
7.312-02
3.992-02
4.812-02
3.902-02
3.192-02
2.632-02
2.192-02
1.822-02
1.532-02
1.292-02
1.092-02
9.222-03
7.832-03
6.622-03
5.612-03
4.752-03

0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0.002+00
0 . 002+00
7 . 932-02
6.292-02
4 . 972-02
3 . 92E-02
3.08E-02
2.42E-02
1.902-02
1.482-02
1.152-02
8.882-03
6.812-03
5.192-03
3.94E-03
2.98E-03
2.252-03
1.692-03
1.272-03
9.SOE-0*
7 . 12E-0*

1.452-02
1.272-02
1.142-02
1.052-02
9.872-03
9.022-03
8.462-03
7.9*2-03
7.442-03
6.962-03
6.51E-03
6.09E-03
5.70E-03
5.34E-03
5.012-03
4 . 702-03
4.412-03
4.152-03
3 . 902-03
3.662-03
3.442-03
3.222-03
3.022-03
2.812-03
2.622-03

1.132-02
1.Q9E-02
1.06E-02
1.04E-02
1.022-02
1.01E-02
l.OOE-02
l.OOE-02
1.01E-02
1.02E-02
1.04E-02
1.06E-02
1.08E-02
1.10E-02
1.12E-02
1.14E-02
1.1SE-02
1.16E-02
1.17E-02
1.17E-02
1.16E-02
1.142-02
1.12E-02
1.09E-02
1.06E-02

l.SOE-0
1.S4E-0:
1.6SE-01
1.75E-01
1.81E-01
1.382-01
1.96E-01
2.04E-Q1
2.14E-01
2.25E-01
2.36E-01
2.47E-Q1
2.59E-01
2.70E-Q1
2.81E-01
2.91E-01
3.01E-01
3.09E-01
3.17E-01
3.24E- 01
3.3i£ 01
3.36E 01
3.40E- 01
3.42£-01
3.44E-01
«= 5
time «v«Et*«d (tav « 1200. •) voluaw concentration:   concentration contour parametera

c - ec(x) • (»rf(xa)-er*(xb)> • («r£.
5.
6.
6.
8.
6.
7.
7.
7.
a.
8.
9.
xc(t)
OOE+00
542+00
082+00
632+00
172+00
71E+00
252+00
302+00
342+00
382+00
422+00
562+00
732+00
932+00
172+00
472+00
332+00
262+00
792+00
432+00
bx(t) betax(t)
0.
S.
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
5.
5.
5.
S.
S.
8.
7.
7.
OOE+00
102-01
022+00
532+00
0*2+00
552+00
062+00
57E+00
062+00
592+00
102+00
222+00
382+00
372+00
802+00
082+00
412+00
822+00
322+00
922+00
0.
t
a.
i.
i.
2.
2,
2.
3.
3.
4.
4.
It.
4.
4.
4,
5.
3.
3,
6.
OOE+00
. 16E-03
.32E-03
.25E-02
, S6E-02
.08E-02
.50E-02
.912-02
.33E-02
. 74E-02
. 16E-02
,27E-02
.39E-02
.55E-02
,73E-02
.96E-02
.24E-02
,57E-02
.98E-02
.47E-02

-------
1.
1.
1.
1.
1.
1.
1.
2.
2.
3.
3.
4.
5.
6.
7.
8.
1.
1.
1.
1.
2.
2.
3.
3.
4.
5.
7.
3.
1.
1.
1.
2.
2.
3.
4.
5.
7.
02E+01
121+01
23E+01
37E+01
54E+01
74E+01
99E+01
29E+01
65E+01
09E+01
63E+01
28E+01
07E+01
02E+01
18E+01
59E+01
03E+02
24E+02
49E+02
79E+02
16E+02
61E+02
1SE+02
82E+02
67E+02
73E+02
09E+02
81E+02
10E+03
38E+03
74E+03
19E+03
77E+03
31E+03
43E+03
65E+03
18E-MJ3
9.08E+03
1.
1.
1.
13E+0*
46E+04
84E+04
2
2
2
2
2
2
2
2
2
2
1
1
7
5
3
3
2
1
1
1
1
a
7
5
4
3
3
2
2
1
1
1
1
1
9
8
7
7
6
5
5
. 13E-02
. 1*E-02
. 13E-02
. 13E-02
. 17E-02
. 18E-02
.20E-02
.23E-02
.26E-02
.30E-02
. 69E-02
. 11E-02
. 30E-03
. 19E-03
. 92E-Q3
. 06E-03
.43E-03
. 98E-03
. 61E-03
.31E-03
.07E-03
. 79E-04
.21E-04
.912-04
. 83E-04
. 99E-04
.31E-04
. 78E-04
.32E-Q4
.98E-04
.69E-04
. *6E-0*
.27E-04
. 12E-0*
. 92E-03
. 33E-03
. 88E-03
.03E-05
.31E-03
. 64E-03
.04E-03
4.
4.
4.
4.
4.
4.
4.
4.
4.
4.
8.
2.
3.
4.
3.
8.
7.
a.
9.
1.
1.
1.
1.
1.
1.
1.
1.
i.
1.
1.
1.
2.
2.
2.
2.
76E-01
76E-01
76E-01
78E-01
76E-01
76E-01
76E-01
76E-01
78E-01
77E-01
81E-01
2SE+00
63E+00
84E+00
9*E+00
97E+00
97E+00
93E+00
87E+00
Q8E+01
17E+01
28E+01
33E+01
44E+01
32E+01
60E+01
68E+01
75E+01
82E+01
88E+01
9SE+01
01E+01
081+01
14E+01
20E+01
2.26E+01
2.
2.
2.
2.
2.
321+01
37E+01
43E+01
48E+01
33E+01
2.
2.
2.
2.
2.
2.
2.
2.
2.
3.
3.
1.
2.
3.
3.
4.
5.
3.
6.
S.
7.
8.
8.
9.
1.
39E+00
42E+00
45E+00
49E+00
54E+00
60E+00
68E+00
78E+00
91E+00
07E+00
34E+00
39E+01
2SE+01
02E+01
72E+01
39E+01
04E+01
69E+01
33E+01
97E+01
61E+01
27E+01
93E+01
83E+01
03E+02
1.11E+02
1.
1.
1.
1.
1.
1.
1.
2.
2.
2.
3.
3.
4.
4.
3.
18E+02
27E+02
37E+02
48E+02
62E+02
78E+02
97E+02
20E+02
49E+02
83E+02
23E+02
70E+02
2SE+02
89E+02
63E+02
1
1
1
1
1
1
1
1
9
8
1
6
4
3
2
2
1
1
1
1
1
1
1
9
9
8
a
7
7
7
6
6
8
6
6
5
3
5
3
3
5
.S1E+01
.S9E+01
.S7E+01
.S4E+01
.49E+01
.43E+01
.33E+01
. 18E+01
. 82E+00
. 19E+00
.62E+00
.43E-01
.01E-01
.01E-01
.4SE-01
.09E-01
.83E-01
.63E-01
.47E-01
. 35E-01
.24E-01
. 13E-01
.08E-01
. 96E-02
.28E-02
. 71E-02
.24E-02
. 83E-02
. 48E-02
. 17E-02
. 90E-02
. 63E-02
. 43E-02
.23E-02
.03E-02
. 88E-02
. 73E-02
. 39E-02
. 46E-02
.34E-02
.23E-02
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
2.
2.
Z.
2.
3.
3.
4.
4.
5.
8.
7.
8.
9.
1.
1.
1.
1.
2.
2.
2.
3.
3.
38E+00
38E+00
38E+00
38E+00
38E+00
38E+00
40E+QO
44E+00
32E+00
69E+00
58E+00
14E+00
13E+00
20E+00
30E+00
42E+00
5SE-H)0
70E+00
87E+00
06E+00
29E+00
34E+00
82E+00
19E+00
63E+00
17E+00
81E+00
53E+00
*2E+00
43E+00
S9E+00
94E+00
13E+01
33E+01
S3E+01
77E+01
04E+01
33E+01
71E+01
12E+01
S8E+01
1.
1.
1.
1.
1.
1.
2.
2.
2.
3.
4.
5.
7.
1.
1.
1.
2.
2.
3.
3.
4.
3.
7.
7.
a.
9.
1.
1.
1.
I.
i.
2.
2.
3.
3.
4.
3.
8.
7.
9.
1.
03E+01
16E+01
29E+01
45E-HU
64E+01
88E+01
16E-M)1
S1E+01
95E+01
53E+01
38E+01
78E+01
69E+01
01E+02
29E+02
84E+02
07E+02
38E+02
20E+02
93E+02
81E+02
83E+02
09E+02
36E+02
25E+02
34E+02
07E+03
23E+03
42E+03
581+03
94E+03
29E+03
71E+03
22E+03
83E+03
58E+03
49E+03
39E+03
93E+03
55E+03
15E+04
1.
1.
1.
1.
1.
1.
1.
2.
2.
3.
3.
4.
5.
6.
7.
8.
1.
1.
1.
1.
2.
2.
3.
3.
4.
3.
7.
8.
1.
1.
1.
2.
2.
3.
4.
5.
7.
9.
1.
1.
1.
02E+01
12E+01
23E-HH
37E+01
54E+01
74E+01
99E+01
29E+01
65E-MJ1
09E+01
63E+01
28E+01
07E-HJ1
02E+01
18E+01
59E+01
03E+02
24E+02
49E+02
79E+02
16E+02
61E+02
13E+02
82E+02
67E+02
73E+02
09E+02
81E-HJ2
10E+03
38E+03
74E+03
19E+03
77E+03
S1E+03
45E+03
65E+03
16E+03
08E+03
13E+04
46E+04
84E+04
8
9,
1.
1.
1,
1.
.85E+00
.34E+00
.06E+01
19E+01
.33E+01
,34E+01
1.77E+01
2.
2.
2.
3.
3.
4.
5.
6.
7.
9.
1.
1.
1.
2.
2.
2.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
.06E+01
40E+01
81E+01
32E+01
93E+01
67E+01
37E+01
86E+01
98E+01
38E+01
15E+02
39E+02
67E+02
02E+02
44Z+02
95E+02
OOE+02
03E-H32
08E+02
12E+02
14E+02
17E+02
19E+02
20E+02
21E+02
22E+02
23E+02
23E+02
24E+02
24E+02
24E+02
24E+02
23E+02
23E+02
7.
7.
3.
9.
1.
1.
1.
1.
1.
2.
2.
3.
3.
4.
5.
6.
7.
9.
1,
1.
1.
1.
2.
7.
1.
1.
2.
2.
3.
3.
4.
5.
7.
8.
1.
1.
1.
1.
2.
3.
3.
. 06E-02
. 79E-02
.S7E-02
73E-02
10E-01
26E-01
45E-01
68E-01
96E-01
30E-01
71E-01
21E-01
31E-01
54E-01
44E-01
31E-01
82E-01
*1E-01
13E+QO
37E+00
65E+00
99E+00
41E+00
33E+01
UE+02
55E+02
OOE+02
52E+02
12E+02
35E+02
73E+02
79E+02
11E+02
72E+02
07E+03
32E+03
62E+03
99E+Q3
44E+03
OOE+03
68E+03
tim« averaged (t«v • 1200.  »>  volvm* concentration:   concentration  In the  z •   1.00 plane.
                                                                                                            « 3
downwind
distance
x (m)
l.OOE+00
1.34E+00
2.08E+00
2.S3E+QQ
3 . 17E+00
3.71E+00
4.23E+00
4 . 80E+00
3.34E+00
5.88E+00
8.42E+00
6.36E+00
8.73E+00
S.93E+00
7.17E+00
7.47E+00
7 . 83E+00
3.26E+OQ
3.79E+00
time of
max cone
(s)
3.31E+02
3 . 32E+02
3 . 32E+02
3.33E+02
3.33E+02
3.34E+02
3.34E+02
3.33B+02
3.38E+02
3.38E+02
3.37E+02
3.37E+02
3.37E-MJ2
3.37E+02
3.38E+02
3 . 38E+02
3 . 38E+02
3 . 39E+02
3.39E+02

cloud
duration

6.
8.
8.
8.
8.
8.
8.
8.
8.
8.
6.
8.
8.
8.
8.
8.
8.
8.
a.
(s)
52E+02
62E+02
62E+02
62E+02
S2E+02-
S2E+02
82E+02
82E+02
621+02
821+02
82E+02
82E+02
52E+02
82E+02
82E+02
S2E+02
821+02
S2E+02
82E+02
effective
half width
bbc (m)
3.34E-02
4.29E-01
8.23E-01
1.22E+00
1.62E+00
2.01E+00
2.41E+00
2.81E+00
3.20E+00
3.SOE+00
3 . 99E+00
4 . OOE+00
*.01E+00
4 . 02E+00
4.03E+00
4 . 04E+00
4 . 08E+00
4.07E+00
4 . 10E+00
avera«« concentration (volume fraction) at (x.y.z)
y/bbc-

0.
0.
0.
0.
0.
0.
0.
0.
8.
1.
1.
1.
1.
1.
1.
1.
1.
2.
3.
0.0
OOE+00
OOE+00
OOE+00
OOE+00
OOE+00
OOE+00
OOE+00
OOE+00
01E-43
90E-33
32E-29
33E-29
34E-29
38E-29
44E-29
36E-29
78E-29
19E-29
03E-29
y/bbe-
0.3
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
0 . OOE+00
4.13E-43
1.31E-33
9.07E-30
9.13E-30
9.23E-30
9.48E-30
9.91E-30
1.07E-29
1.22E-29
1.30E-29
2.08E-29
y/bbc-
1.0
0 . OOE+00
0. OOE+00
O.OOE+OQ
0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
1.33E-43
4.24E-36
2.94E-30
2.96E-30
3.00E-30
3.07E-30
3. 222-30
3.48E-30
3.97E-30
+.39E-30
8.78E-30
y/bbc-
1.5
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
2.10E-44
6.30E-37
4.31Z-31
4.34E-31
4.60E-31
4.71E-31
4.93E-31
3.34E-31
8.08E-31
7.49Z-31
1.04E-30
y/bbc-
2.0
0. OOE+00
0. OOE+00
0 . OOE+00
O.OOE+00
0 . OOE+00
O.OOE+00
0 . OOE+00
0 . QOE+00
l.*OE-45
4.71E-38
3.27E-32
3.29E-32
3.33E-32
3.41E-32
3.37E-32
3.36E-32
*.<»OE-32
3.42E-32
7.31E-32
y/bbc"
2.5
O.OOE+00
O.OOE+00
O.OOE+00
0. QOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.61E-39
1.12E-33
1. 12E-33
1. 14E-33
1.16E-33
1.22E-33
1.32E-33
1.50E-33
1.35E-33
2.57E-33
                                                        B-9

-------
9 . 432+00
1.022+01
1. 122+01
1.232+01
1.372+01
1.S4E+01
1.74E+01
1.99E+01
2.292+01
2.652+01
3.092+01
3.632+01
4.282+01
5.072+01
6.022+01
7.182+01
8.592+01
1.032+02
1.242+02
1.492+02
1.792+02
2. 16E-t-02
2.612+02
3 . 152+02
3.32Z-KI2
* . 872+02
5.732+02
7 . 091+02
8.312+02
1. 102+03
1.38E+03
1.74E+03
2.192+03
2.77E-MJ3
3 . 512+03
4.452+03
5 . 852+03
7 . 181+03
9.08E+03
1.132+04
1.462+04
1.842+04
3 . 402+02
3.412+02
3.422+02
3.432+02
3.442+02
3.46E+02
3.48E+02
3.51E+02
3.34E+02
3 . 582+02
3 . 63E+02
3 . 68E+02
3.75E+02
3 . 83E+02
3.93E+02
4.062+02
4.202+02
4 . 38E+02
4 . 80E+02
4 . 87E+02
5.192+02
5.582+02
6.052+02
8.62E+02
7.36E+02
8.232+02
9.34E+02
1.07E+03
1.23E+03
1.42E+03
1.66E+03
1.94E+03
2.29E+03
2.71E+03
3.22E+03
3.83E+03
4 . J8E-MD3
5.492+03
6.592+03
7.93E+03
9.552+03
1.13E+04
8.62Z+02
.82E+02
.822+02
.622+02
.822+02
.82E+02
.822+02
.622+02
.822+02
6.622+02
6.622+02
6.622+02
6.622+02
6 . 622+02
6.622+02
6.622+02
8 . 622+02
6.622+02
6.622+02
6.622+02
6.622+02
6.622+02
6.622+02
6.622+02
7.07E+02
7.37E+02
8.17E+02
8.892+02
9.762+02
1. 082+03
1.212+03
1.372+03
1.562+03
1.802+03
2.082+03
2.422+03
2.832+03
3.312+03
3.89E+03
4 . 382+03
3.402+03
6.382+03
4
4
4
4
4
4
4
4
4
3
3
9
2
3
3
6
7
8
9
1
1
1
i
i
i
i
i
2
2
2
2
2
3
3
3
4
4
3
6
7
8
9
13E+00
17E+00
.212+00
.27E+00
. 34E+00
.42E+00
. 33E+00
. 672+00
. 84E+00
. 06E+00
.34E+00
. 63E+00
.422+01
.912+01
.252+01
.472+01
. 64E+01
. 77E+01
. 89E+01
. 10E+02
.212+02
.322+02
. 44E+02
. 36E+02
. 67E+02
. 80E+02
.922+02
. 06E+02
.202+02
.372+02
. 582+02
. 812+02
. 082+02
.422+02
. 822+02
.312+02
. 902+02
. 60E+02
.41E+02
.372+02
. 482+02
.732+02
4
1
3
1
1
6
7
1
3
6
3
5
4
2
2
1
1
1
9
7
6
5
4
3
2
2
1
1
. 99E-29
. 062-28
.292-28
.742-27
.962-28
.242-23
. 882-23
.122-19
.312-13
.102-10
.262-05
.332-03
.092-03
. 752-03
. 042-03
.612-03
.322-03
. 09E-03
. 092-04
. 572-04
.302-04
.222-04
.312-04
. 552-04
.862-04
.272-04
. 782-04
.372-04
1.042-04
7
3
4
2
1
1
8
5
3
2
1
a
5
. 822-03
.74E-05
. 13E-OS
. 902-05
. 992-05
. 33E-03
. 73E-06
. 62E-06
. 562-06
.242-06
. 402-06
. 69E-07
. 40E-07
3.432-29
7.312-29
2.262-28
1. 192- 27
1.342-26
4.292-25
3.422-23
7.722-20
2.412-15
4.192-10
3.622-05
3.662-03
2.812-03
1.892-03
1.402-03
1.112-03
9.062-04
7.302-04
6.242-04
3.202-04
4.332-04
3.502-04
2.962-04
2.44E-04
1.962-04
1.562-04
1.222-04
9.432-05
7 . 182-05
5.372-03
3.94E-05
2.84E-OS
2.002-03
1.372-05
9.162-06
6.002-06
3.862-06
2.452-06
1.54E-06
9.592-07
5.972-07
3.712-07
1.112-29
2.372-29
7.342-29
3 . 882-28
4.372-27
1.392-25
1.76E-23
2.502-20
7.832-18
1.36E-10
1.17E-03
1.19E-03
9.122-04
6.132-04
4.552-04
3.602-04
2.942-04
2.442-04
2.032-04
1.692-04
1.402-04
1.162-04
9.622-05
7.912-05
6.372-05
5.062-05
3.972-05
3.062-05
2.33E-03
1.742-05
1.282-05
9.222-06
6.482-06
4.442-06
2.97E-06
1.952-06
1.232-06
7.952-07
4 . 992-07
3.112-07 •
1.94E-07
1.212-07
1.712-30
3.54E-30
1.132-29
5.952-29
6.692-28
2.132-26
2.702-24
3.842-21
1.202-16
2.092-11
1.802-06
1.822-04
1.40E-04
9.402-05
6.972-05
5.522-05
4.512-05
.3.732-05
•3.112-05
2.592-05
2.152-05
1.792-05
1.472-05
1.212-05
9.772-06
7.762-06
6.082-06
4.702-06
3.572-06
2.68E-06
1.96E-Q6
1.412-06
9.942-07
6.822-07
4 . 362-07
2.992-07
1.922-07
1.222-07
7.632-08
4.782-08
2.972-08
1.352-08
1.242-31
2.63E-31
8.152-31
4.31E-30
4.85E-29
1.54E-27
1.952-25
2.78E-22
8.702-18
1.312-12
1.30E-07
1.322-05
1.01E-05
6.812-06
5.05E-06
4.002-06
3.27E-06
2.702-06
2.25E-06
1.38E-06
1.562-06
1.292-06
1.072-06
8.79E-07
7.08E-07
3.62E-07
4.40E-07
3.402-07
2.59E-07
1.94E-07
1.422-07
1.022-07
7.202-08
4.942-08
3.30E-08
2.16E-08
1.39E-08
8.332-09
5.54E-09
3 . 462-09
2.15E-09
1.34E-09
4.22/-33
8.9««-33
2.78E-32
1.472-31
1.66E-30
5.28E-29
6.66E-27
9.49E-24
2.972-19
5.172-14
4.462-09
4.522-07
3.46E-07
2.32E-07
1.73E-07
1.362-07
1.122-07
9.25E-08
7 70E-08
6.42E-38
5.33E-08
4.432-08
3.55E-08
3.01E-08
2.422-08
1.922-08
1.512-08
1.162-08
8.86E-C9
5.S3E-09
4.36E-QS
3.50E-OS
2.47E-C";
1.59E 0'.
1. 132-OS
7 42E-10
4.762-10
3.02E-1D
1.892-10
1.182-10
7.36E- ! I
4 57E-1J
tima avaragad (tav - 1200.  *)  voluma concentration:   maxiimm  concentration  (voluma  fraction)  along  cantatlina.
downwind
diatanea
X (0)
1.002+00
1.3*2+00
2.082+00
2.632+00
3.172+00
3.712+00
4.232+00
4 . 802+00
5.342+00
5.382+00
6.422+00
8.362+00
6.732+00
6.932+00
7.172+00
7.472+00
7.332+00
8.262+00
8.792+00
9.432+00

hvight
x (a)
5.002+00
9.932+00
1.182+01
1.312+01
1.412+01
1. 442+01
1.342+01
1.382+01
1.612+01
1.832+01
1.832+01
1.832+01
1.632+01
1.632+01
1.632+01
1.832+01
1.632+01
1.632+01
1.622+01
1.622+01
majcimum
concentration
e<*,0.*)
8.472-01
4.452-01
1.632-01
7.922-02
4.612-02
3.002-02
2.102-02
1.352-02
1.192-02
9.432-03
7.662-03
7.642-03
7.632-03
7.612-03
7.592-03
7.572-03
7.34E-03
7.302-03
7.432-03
7.402-03
tima of
max cone
(a)
3.312+02
3 . 322+02
3.322+02
3.332+02
3.33E+02
3.34E+02
3.34E+02
3.352+02
3.362+02
3.362+02
3.372+02
3.372+02
3.372+02
3.372+02
3.382+02
3.382+02
3.382+02
3.392+02
3.392+02
3.402+02
cloud
duration
(s)
6 . 322+02
6.322+02
6.622+02
6 . 622+02
6 . 622+02
6.822+02
8.822+02
3 . 322+02
6. 622+02
6.622+02
6.622+02
6.822+02
6.622+02
6.622+02
6.622+02
6.622+02
3.322+02
3.622+02
6.622+02
6.622+02
                                                       B-10

-------
1.12E+01
1.23E+01
1.37E+01
1.5*E-M)1
1.74E+01
1.99E+01
2.29E+01
2.65E-t-01
3.09E+01
3.63E+01
4.28E+01
5.07E-MJ1
6.02E+01
7.18E+01
8.59E+01
1.03E+02
1.24E+02
1.49E+02
1.79E-MJ2
2.16E+02
2.81E+02
3 . 15E+02
3 . 82E+02
4 . 67E+02
S.73E-MJ2
7.09E+02
8 . 81E+02
1.10E+03
1.38E+03
1.74E+03
2.19E+03
2.77E-M)3
3.311*03
4.4SE+03
3.652+03
7 . 16E-M)3
9 . 08E+03
1.13E+04
1 . 46E+04
1.84E+04
1.59E+01
1.37B+01
1.3*8+01
1.49E+01
1.43E+01
1.33E+01
1.18E+01
9.62E+00
8.19E-I-00
6.39E-01
O.OOE+00
0 . OOE+00
O.OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
0 . OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
0 . OOE+00
0. OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
0. OOE+00
0 . OOE+00
0. OOE+00
0. OOE+00
7.23E-03
7 . 16E-03
7.04E-03
6.90E-03
8.73E-03
6.33E-03
8.29E-03
6.01Z-03
5.69E-03
3.33E-03
3.33E-03
3 . 86E-03
2.821-03
2.15E-03
1.88E-03
1.34E-03
1.08E-03
8.73E-0*
7 . 08E-04
3.74E-04
4.66E-04
3.78E-04
3.00E-04
2.36E-04
1.83E-04
1.40E-04
1.06E-04
7.91E-03
3.79E-03
4.16E-OS
2.92E-03
2.00E-03
1.34E-03
8.73E-06
3.63E-06
3.37E-08
2.24E-06
1.40E-06
8.69E-07
3.40E-07
3 . 42E+02
3.43E+02
3 . 44E+02
3.46E+02
3.48E+02
3.31E+02
3.34E+02
3.38E+02
3.63E+02
3 . 88E+02
3.73E+02
3.83E+02
3.93E+02
4.06E+02
4 . 20E+02
4.38E+02
4 . 60E+02
4 . 87E+02
3.19E+02
3.38E+02
8.03E+02
8.62E+02
7.36E+02
8.23E+02
9.34E+02
1.07E+03
1.23E+03
1.42E+03
1.S6E+03
1.94E+03
2.29E+03
2.71E+03
3.22E+03
3 . 83E+03
4.38E+03
3.49E+03
6.39E+03
7 . 93E+03
9.33E+03
1.13E+04
6 . 62E+02
6.62E+02
6 . 82E+02
6 . 62E+02
6 . 62E+02
6.82E+02
8.82E+02
6.S2E+02
S . 62E+02
6 . 82E+02
6.82E+02
6.62E+02
6.62E+02
8.62E+02
6.62E+02
6.S2E+02
6.62E+02
6 . 82E+02
8 . 62E+02
6 . 82E+02
8.62E+02
S.62E+02
7.07E+02
7 . 37E+02
8.17E+02
8 . 89E+02
9.76E+02
1.08E+03
1.21E+03
1.37E+03
1.36E+03
1.80E+03
2.08E+03
2.42E+03
2.83E+03
3.31E+03
3 . 89E+03
4 . 38E+03
3 . 40E+03
8.38E+03
B-ll

-------
              APPENDIX C
INTRODUCTION TO BUOYANT PLUME RELEASES

-------
                          TABLE OF CONTENTS

Section                                                             Page


1.0    BACKGROUND	 C-l

2.0    MODE OF RELEASE  	 C-l

3.0    LIFT-OFF	':	 C-4

4.0    PLUME TRAJECTORY	 C-4

5.0    GROUND LEVEL CONCENTRATIONS UNDER A RISING PLUME ... C-6

6.0    TERMINATION OF PLUME RISE 	 C-8

      6.1   Termination of Plume Rise in Neutral Conditions  	C-9
      6.2   Termination of Plume Rise in Stable Conditions  	 C-10

7.0    PASSIVE DISPERSION 	 C-10

8.0    REFERENCES	 C-ll


                            LIST OF FIGURES

Figure                                                              Page

C-l   Typical History of Plume Rise 	 C-2

C-2   Some Conceivable Modes of Release of a Buoyant Plume  	 C-3

C-3   Simulation of a Dense Plume in a Water Tank	 C-7
                                   11

-------
                                  APPENDIX C

               INTRODUCTION TO BUOYANT PLUME RELEASES

1.0    BACKGROUND

The purpose of this section is to introduce any reader who may be interested to the subject
of plume rise and to summarize the issues that must be considered by the requestor and the
analyst when a buoyant plume is released into the environment.  These are:

(i)    Definition of the mode of release,

(ii)    "Lift-Off —  the behavior of a buoyant plume in a turbulent building wake (if the
      plume arises other than from an isolated point source),

(iii)   Plume trajectory,

(iv)   Ground level concentrations under a rising plume,

(v)    Termination of plume rise, and

(vi)   Finally,  passive dispersion.

As an example,  Figure  C-l  shows a typical history  of plume rise for a release into  a
turbulent building wake.  A good  introductory discussion has been compiled by Briggs(1).
(Note:  reference numbers refer to Section 8.0 of this Appendix).

2.0    MODE  OF RELEASE

Some potential modes of release are shown in Figure C-2.  These include an isolated point
source (A - e.g., an incinerator stack).  Figure C-2(B) shows two alternatives for a short
stack: a) the plume rises in a low windspeed, and b) the plume is mixed into the turbulent
building wake  in a high windspeed. Figure C-2(C) shows direct leakage from the face of
the building into a turbulent wake. Figure C-2(D) shows how a high momentum vertical
jet may escape the wake  of  the building,  while  Figure C-2(E) shows that a randomly
                                       C-l

-------
                                                    INVERSION LJD
n
  RELATIVELY
     LOW
CONCENTRATION
                                                                                            HIGHER ^	
                                                                                         CONCENTRATION
                                     Figure C-1. Typical History of Plume Rise.

-------
    A. TALL STACK - ISOLATED
       POINT SOURCE
n
    D. VERTICAL JET
B.  SHORT STACK
   (a) PLUME BEHAVIOUR - LOW
   WINDSPEED. (b) WAKE
   BEHAVIOUR. BRISK WIND
E.  RANDOMLY ORIENTED JET
                                                                                                  ^
C.  LEAKAGE FROM FACE OF
   BUILDING INTO TURBULENT
   WAKE
                                                                                                    r
F.  LARGE ARIiA SOURCE AFTER
   SPILLAGE ONTO GROUND OR
   AFIRE
                               Figure C-2. Some Conceivable Modes of Release
                                           of a Buoyant Plume

-------
oriented jet may be trapped in the wake. Finally, Figure C-2(F) shows a buoyant plume
evaporating from a spillage on the ground or rising from a fire.

3.0   LIFT-OFF

The question, what happens to a buoyant plume emerging from an area source or one that
is mixed into a turbulent building wake, has not yet been satisfactorily answered and major
uncertainties still remain. However, Briggs(2) has put forward some simple ideas based on
the consideration of a Richardson number
                                                                      (3-1)
where g is the acceleration due to gravity, H is the depth of the plume, u, is the friction
velocity, pa is the density of the air and Ap = (pa-p), p being the density of the plume. The
quantity gHAp is the square of a buoyancy induced velocity,  u., on the other hand, is a
velocity that  is typical  of the rate of  spread of passive plumes.  If Lp < 1, it is to be
expected that the dilution  of the plume is dominated by atmospheric turbulence and the
plume should behave passively. If, on the other hand, Lp > 1, the plume should rise off the
ground leaving little or no  residual gas at  ground level. For  intermediate values of Lp the
plume should behave in some intermediate fashion.

Briggs(2) used heuristic  arguments to obtain a critical  value of Lp above which the  plume
essentially lifts off cleanly:   this was Lp =  2.5. Subsequently, Meroney'3' looked at the rise
of a buoyant plume released at ground level, using smoke visualization in a wind tunnel.
On the basis of these data Lp ~ 20-30 is a more appropriate figure.

4.0   PLUME TRAJECTORY

There are over a hundred plume rise models in the literature. Comprehensive reviews have
been given  by Briggs(t) and readers  with sufficient  stamina may  work  back from  his
references.  In neutral  conditions, Briggs' formula for the center-line of a buoyant, rising
plume is given by
                                    C-4

-------
                            Ah = 1.6F1/3x2/3LJ-'  m                       C4"1)

where Ah is the height above the point of release, F is the buoyancy parameter, x is the
distance travelled downwind, and U is the velocity of the plume.

F is given by

                           F - (gQ/*Cpp.T) mV3                       (4-2)


where g is the acceleration due to gravity (9.81 m/s2)
      Q is the rate of heat release accompanying the plume (J/s)
      Cp  is  the specific heat of air at constant pressure (J/Kg/K)
      pa is the density of air (kg/m3); and
      T is the temperature of the air (K).

If Q is in  megawatts, it can be shown that

                               F~8.9QmV3                           (4-3)


In the treatment of Briggs, the data used to test Eq. (4-1) are taken from the rise of plumes
from power  station chimneys which are 100 m or more in height.  The velocity U is either
that at the final height of plume rise or that averaged over the whole of the plume depth
from the source height to the top of the observed  plume.

In stable conditions, a rising, bent-over continuous plume follows the trajectory

                                                                        (4-4)
where z is the height of rise (m)
      F is the quanfity defined in Eq.(4-3) (mY3)
      i8T is the square of the Brunt-Vaisalla frequency (s2), and is related to the gradient
        of potential temperature 9 in the atmosphere by j3T = (g/T)(d0/dz); and
                                        C-5

-------
      /?e is the entrainment coefficient (dimensionless)


Eq. (4-4) reduces to Eq. (4-1) for  &"\< I When  $^\~ jthe plume reaches its final height,

which is given by Briggs as

                             Ah = 2.6(F/U/3T)I/3                        (4-5)
The experimental data are consistent with values of the constant on the right hand side of
Eq. (4-5) being in the range 2.3 to 2.9.. The same comments apply to the height at which
U is measured as for the neutral case.

5.0   GROUND LEVEL CONCENTRATIONS UNDER A RISING PLUME

For a well  established  plume  rising  into the atmosphere, there is  no doubt that the
concentration beneath it is exceedingly small.  Consider the analogy with a salt water plume
falling through a tank of pure water.  A reservoir of salt in solution in water was carried on
a trolley which moved above a long trough containing water, see Figure C-3. The relatively
dense salt solution was  allowed to flow into the  trough through  a tube, so  simulating a
plume emerging into a  cross wind.  The salt plume was made visible by shining a light
through it. It was very noticeable that, should a wisp of salt solution attempt to escape from
the plume  (as  at point A on  Figure C-3) it was  literally "snatched back"  (as at  point
B). Caution should be exercised in assuming that this observation applies to the case of a
buoyant  plume  in the  atmosphere,  since in that  case both the plume itself and the
surrounding atmosphere are turbulent, whereas in Figure  C-3 the water is not  turbulent.
It is reasonable to assume, however, that as long as the "vigor" of the plume  generated
turbulence  exceeds  that of the  atmosphere,  the rate of growth of the plume  radius r is
entirely determined by the actions of turbulence generated within the plume itself according
to an entrainment law such as r=(3ez, see below.

Returning to Figure C-3, when a salt sensitive probe was moved across the plume, there was
a  very sharp concentration boundary:  that  is, the radial concentration distribution was
nearly "top-hat".
                                        C-6

-------
  '/  '/  V  '/  V
///./.  /.  /.  .
                                 j^/J^-TURBULENT SALT WATER PLUME"  ^  ^
'   /  ' /   /  /  /  /N5
/   /   /    /   /   /  Xs
 /   /   /    /   /   /   /
 x/  x/  ^/   x/    y   '/
                                                         /   /   /   /
                                                         /-   ^    x    x
             Figure C-3. Simulation of a Dense Plume in a Water Tank.

-------
One simple way of simulating this effect in a conservative way is to assume that the radial
concentration profile of a rising plume x as a function of distance downwind (x), acrosswind
(y), and  above ground  (z),  is Gaussian,  with radius  r  defining the  10% concentration
contour.  That is

                                        2ir
-------
(i)    The action  of atmospheric turbulence - the "vigor" of the turbulence  within the
      plumes decreases until it is less than the corresponding "vigor" in the atmosphere.

(ii)   Stable temperature gradients - the plume rises until its temperature equals that of
      the surroundings, give or take a few damped oscillations.

6.1   Termination of Plume Rise in Neutral Conditions

The problem encountered in trying to define a final height of rise in neutral  conditions is
that the vast majority of plume rise observations show the plume still rising at the greatest
distances of observation. It follows that terminating plume rise in neutral conditions usually
involves the postulation of some conservative criterion, and since there is an infinite number
of possibilities, it  is not surprising that  a large number of examples are  found in the
literature.

It is the action of atmospheric turbulence that terminates plume rise, and the  quantity that
has been described as the "vigor" of the plume turbulence is given a precise meaning by
equating it to the turbulence energy dissipation rate, for which an  approximate  expression
within the plume is

                                          z                             (6-1)
where 77 is a constant and w=dz/dt.  Outside the plume,
                                                                        (6-2)
where k is Von Karman's constant.  It can be shown that, for plumes rising several hundred
meters into the atmosphere, equating the two quantities gives a final height of rise

                               Ah~300F/U3                           (6-3)
For very buoyant plumes, the quantity Ah may exceed the height of the inversion- lid 1
(Figure C-l). Typical values of 1 vary from a few hundred meters to over a kilometer. If
                                        C-9

-------
the upper edge of the plume  touches  the inversion lid, its rise can conservatively be
terminated at this point.

For further discussion of the mechanisms for terminating plume rise in unstable and neutral
conditions, consult the  reference by Briggs(28).

62   Termination of Plume Rise in Stable Conditions

This has already been discussed and is given in Eq. (4-5). In stable conditions, the plume
rise is effectively terminated directly above the source.

7.0   PASSIVE DISPERSION

The simple picture given on Figure C-l implies that plume rise terminates abruptly and that,
thereafter, the plume behaves passively.  A straightforward way of taking this into account
is as follows.

Suppose that plume rise has terminated at a distance xb downwind.  The plume will then
have a  radius rb,  its centre-line will  be at a height hb and the spatial distribution of
concentration will be described by Eqs. (5-1) and (5-2) with
Pr -
                                                                        (7-1)
For distances exceeding xb, the atmospheric dispersion continues to dilute the plume
so that
                       7) .
where

                                                                         (7-3)
                                    C-10

-------
                           ^(x-x,) = <£ + ^(x-xj                      (7-4)

and
Here, ay(x-xb) and 
-------
                  APPENDIX D

SIMPLIFIED SOURCE TERMS AND DENSITY CALCULATIONS
          FOR FLASHING LIQUID RELEASES

-------
                           TABLE OF CONTENTS


Section                                                                Page


1.0    PURPOSE 	    D-l

2.0    SIMPLIFIED TREATMENT OF FLASHING LIQUID RELEASES	D-l

      2.1   Puff Release 	D-l
      2.2   Continuous Release	D-2

3.0    SIMPLIFIED CALCULATION OF ORDERED TRIPLES  	D-3

      3.1   Purpose	D-3
      3.2   Visualization	D-3
      3.3   Calculations  	D-4

           3.3.1  Initial Density When Mf = 1	D-4
           3.3.2  Mole Fraction Me When All Liquid Droplets Just Evaporate . .  D-4
           3.3.3  Calculations for Mf > Me	D-5
           3.3.4  Calculations for Mf < Me 	D-5
                                    u

-------
                                  APPENDIX D

          SIMPLIFIED SOURCE TERMS AND DENSITY CALCULATIONS
                       FOR FLASHING LIQUID RELEASES

1.0    PURPOSE

The purpose of this Appendix is to provide guidance on how to simplify the calculation of
source term parameters and of the density of air/gas/aerosol mixtures in cases where there
might otherwise be the need for complex calculations.  The calculations outlined below can
quickly be encoded in a computer program or performed by hand.  The methods described
below were used to define the initial conditions for  puff releases in Chapter 8  and the
ordered triples for jet releases in Chapter 6.

2.0    SIMPLIFIED TREATMENT OF FLASHING LIQUID RELEASES

The following treatment is suitable for materials such as chlorine or ammonia which are
initially  at room temperature with a high degree of superheat:

(i)    When such materials flash, it is experimentally observed that they generally remain
      completely airborne as a mixture of vapor and droplets.

(ii)   It is also known from  many experiments that the initial jetting and flashing  phase
      causes the entrainment of the order of ten times as much air by mass as the  initial
      release.

(iii)   Simplifying Assumption: this air evaporates all of the liquid  droplets and  leaves a
      mixture at the boiling point Tb; this mixture can then be used  as a starting point for
      the atmospheric dispersion calculations.

2.1    Puff Release

The purpose  of this section  is to define the  initial parameters for a puff release.  This
method was used to define the initial conditions for the puff release of chlorine described
in Section 8.1. The following quantities are needed:
                                       D-l

-------
      Total mass of material released     :    M kg

      Total mass of air entrained         :    10M kg (by assumption)

      Temperature of mixture            :    Tb(K)  (boiling  point of  material  -  by
                                            assumption)

      Density of air at Tb                :    p^ kg/m3 = pa(Tb)

      Density of released material at Tb   :    pb kg/m3 = pg(Tb)

Assume pab and pb can be obtained from the densities at ambient temperature by using the
perfect gas law, or look them up in an engineering, chemistry, or physics handbook.

It can then easily be shown that:

      Initial volume occupied, V          :    M/pgb  +  10M/pab m3

      Initial density of puff, p;            :    11M/V

Assume air moves  with mean velocity u and that the initial mass M is initially stationary.

      Initial velocity of puff U;  "          :    10u/ll (conservation of momentum)

This puff can then  be used as initial input to a dispersion modeling code, which will model
the subsequent entrainment of air, and heating by air  entrainment and (possibly) by the
ground.

If the initial puff dimensions are needed, a further simplifying assumption is that the puff
is cylindrical with the radius equal to the height.
                                    D-2

-------
                              r  = h =
3V
4x
                                                                        (2-1)
Which gives the effective area covered by the cylinder as
                                  A = XT'
                               (2-2)
Example

Chlorine initially at 278 K is released as a puff from a one ton ~ 1,000 kg cylinder in which
it is kept as a liquid under pressure.
      Total mass of material released
      Total mass of air entrained
      Temperature of mixture
      Density of air at 239 K
      Density of chlorine vapor at 239 K
       Initial volume occupied
      Initial density of puff
      Initial velocity of puff
      From Eq. (2-1), r  = h = 11.89 m
          1,000 kg

          10,000 kg (by assumption)

          239 K, the boiling point of chlorine

          1.477 kg/m3

          3.578 kg/m3

          1,000/3.578 + 10,000/1.477
          7,050 m3

          11,000/7,050
          1.560 kg/m3
                                                   1.36 m/s in a wind of 1.5 m/s.
                                        D-3

-------
      From Eq. (2-2), A =  T(11.89)2 = 444 m2

These results are used in Section 8.1 as input to the SLAB and DEGADIS models.

22   Continuous Release

The purpose of this section is to define initial volume flow rate, density and velocity of a
continuous flashing release at ground level.

Assume an orifice of diameter A(m2) leading out of a vessel at  gauge pressure p  (Pa).
Using Bernouilli's formula, the velocity of release V; is


                              YJ =c/2p/pL  m3                        (2'3)

(neglecting static head) where c is a constant (~0.6) and pL is the liquid density (kg/m3).

The mass rate of release is M = ApLv-, kg/s                             (2-4)

As above, assume that air is entrained at a rate 10M so that  the initial mass flux is  11M
kg/s, the initial volume flux is V  = M /pgb + 10M pab, the initial density is 11M /V (kg/m3)
and the initial temperature is Tb (K).

The initial velocity of the puff uf is given by

                          U; = (vsta + 10Mu)/llM(m/s)      .      (2-5)

(which is a simple application of conservation of momentum).  This should enable the user
to define the initial starting conditions for a continuous, horizontal, ground level (non-jet)
release.  This continuous case has  not been used in the present work, but is included  for
completeness.
                                        D-4

-------
3.0    SIMPLIFIED CALCULATION OF ORDERED TRIPLES

3.1    Purpose

Suppose that an initial flashing process leads to a proportion fv of vapor and l-fv of liquid
droplets. Assume that fv has been calculated independently, either as a simple flash fraction
or by a more sophisticated model that takes account of droplet rainout.  Assume also that
this initial mixture is at temperature Tb.

The  object is to calculate "ordered triples" consisting of mole fraction Mf, density p, and
concentration X-  These ordered triples are those required by the DEGADIS model, see Table
6-4 of Section 6; lines 17-26 of that table were calculated by the method described below.

3.2    Visualization
   X
   i
        Ta~^
(A)
The above figures show what happens if air is mixed into the vapor/aerosol mixture defined in
Section 3.1.  Mf declines from 1 to zero as the amount of air is increased. As shown on Figure
(A) above, small amounts of air evaporate some but not all of the liquid droplets, leaving the
temperature at Tb. At some mole fraction M., all of the liquid droplets have evaporated, but the
temperature is  still Tb.   For smaller values of Mf (greater amounts of air), the temperature
increases towards that of the surrounding air. Meanwhile, as air is mixed in, the concentration
                                        D-5

-------
X decreases to zero and the density p decreases to that of the ambient air, pa(TJ, see (B) above.
As noted above, the object is to calculate p and x as a function of Mf.

3.3    Calculations

The calculations proceed as follows:

3.3.1  Initial Density When Mf = 1
As described in Section 3.1, every kilogram of material that is initially airborne consists of fv
kg of vapor and (l-fv) kg of liquid.  The density of the vapor is pg(Tb) and of the liquid is pL,
so the volume occupied by the vapor  is V=fv/pg(Tb) +  (l-fv)pL.   The overall  density is
p(Mf=l)=(fv+l-fv)/V=l/V  kg/m3  and   the corresponding  concentration   is  x(Mf=l)
=p(Mf=l)= 1/V kg/m3. Mf=l, p(Mf=l), and x(Mf=l) are the "ordered triple."

Example

Chlorine initially at 278 K flashes to vapor and aerosol.  From Eq.  (6-2) and Section 6.2. 1 of
the main body of this report, fv = 0.17 and l-fv = 0.83.  The density of chlorine vapor at its
boiling point is 3.578 kg/m3 (see above).  The density of liquid chlorine is 1,574 kg/m3 (see
Table 5-4 in Section 5.4.2 of the main body of this report).  The volume occupied by 1 kg of
chlorine is V = 0.17/3.578 +  0.83/1574 = 0.048 m3. The overall  density p(Mf=l) is 1/V =
20.82 kg/m3 = x(Mf=l).  Therefore, the ordered triple is (1, 20.82,  20.82). This appears as
line 26 on Table 6-4 of Section 6.2.3 of the main  report.

3.3.2  Mole Fraction M, When All Liquid Droplets Just Evaporate

Let us assume that mass mK of air just evaporates (l-fv) kg of liquid droplets.

                                (Ta-Tb)Cp. mae  = (l-Oh,                  (3-1)

where T,  is the temperature of the air (K), Cpl is the specific heat  at constant pressure of air
(J/Kg/K) and Iv, is the latent heat of vaporization of the liquid droplets (J/kg).
                                          D-6

-------
                                  vv    ,      ,    r   •  •               /•* o\
                      m   = -   kg per kg of emission.            (y-^)
                            (Ta-Tb)Cpa

mae kg/air corresponds to  Nae=miie/Mwa moles,  where Mwa is the molecular weight of air
(kg/mole).   1  kg of released material corresponds  to Ng=l/Mwg moles, where  M^ is the
molecular weight of released material (kg/mole).  The mole fraction Me (see (A) above) is then
given by

                                         N
                                 Me  = - L_                            (3-3)
Example

For the case of chlorine initially at 278 K, Eq. (3-2) gives

      mae = (0.83)(287,840)/{(278-239)(l,000)} = 6.126 kg

(where 287,840 is the latent heat of vaporization of chlorine from the SLAB User's manual and
1,000 is the specific heat at constant pressure for air). Nae = 6. 125/0.029 = 21 1 moles of air
and Ng = 1/0.07 =  14.2  moles of chlorine.   From Eq.  (3-3), Me =  14.2/(14.2  +  211)  =
0.063.

3.3.3 Calculations for M, > M,

This corresponds to the case where a small amount of air is added (i.e., the mole fraction lies
to the right of M, on Figure (A)).

Let Nf moles of air  correspond to a mole fraction Mf on Figure (A) above.  An equation similar
to (3-3) applies:

                                 Mf = Ng/(Ng + Nf)                    (3-4)

Rearranging Eq.  (3-4) and  remembering that Ng =  1/M,,, it follows that:
                                        D-7

-------
                                                                           (3-5)
The corresponding mass of air is obtained by multiplying Eq. (3-5) by the molecular weight of
air Mwa:
                                              ,                           (3.6)
The amount of liquid droplets evaporated will be fe (kg) where
all at temperature Tb.

             The volume occupied is :

             V=myPa(Tb) + (fv+fe)/pg(Tb) + (K-fJ/ft. m3                  (3-8)

             The density p(Mf) is (1+MJ/V kg/m3

             The concentration x(Mf) IS l/v kg/m3

             Mf, p(Mf) and X(Mf) give the "ordered triple."

Example

Consider again the case of chlorine initially at 278 K.  From Eq. (3-6):

       01^ = (0.029)(0.75)/{(0.07)(0.25)} = 1.242 kg.


                                         D-8

-------
From Eq. (3-7):

       fe = (278-239)(l,000)(1.242)7(287,840) = 0.168.

From Eq. (3-8):

       V = (1.242)7(1.477) +  (0.17 + 0.165)7(3.578)
         + (1 - 0.17 - 0.168)7(1,574) = 0.935 m3

so that p(Mf) = 2.242/0.935 = 2.40 kg/m3
and    x(Mf) = 1/0.935 =  1.07 kg/m3.

The ordered triple is thus (0.25, 1.07, 2.4) and appears on line 21 of Table 6-4 of Section 6.2.3
of the main body of this report.

3.3.4  Calculations for M, < M.

This corresponds to the case in which a relatively large amount of air is added (i.e., the mole
fraction  lies to the left of  Me on Figure (A)).   Analogous to  Eq. (3-6), the mass of air
corresponding to  Mf mole fraction is:


                              maf = -J!!	  k§                         (3'9)
                                      Mwg - Mf

From Section 3.3.2, Eq. (3-2), the mass of air required to evaporate all of the liquid droplets
is
and mae <
                                          D-9

-------
Let the final temperature of the mixture be T,  which is found by solving the heat  balance
equation
                     (marmae)Cpa(Ta-T) = (T-Tb)mae Cpa +  (T-Tb)Cpg        (3-11)

where Cpg is the specific heat at constant pressure of the released material
                       T -   bcp.pg.rp.a                  (3
                                    KA.+CJ

The volume occupied by the air/material mixture is

                                V = m>a(T) + l/pg(T)                   (3-13)
from which jo(Mf) = (1 +mlf)/V and x(Mf) = 1/V. Mf, p(Mf) and x(Mf) then give the "ordered
triple."

Example

Taking once more the case of chlorine at 278 K, consider Mf = 0.025.  From Eq. (3-9):

       mrf = (0.029)(1 - 0.025)/{(0.07)(0.025)} = 16.16 kg.

Eq. (3-10) has already been solved (see Section 3.3.1 of this appendix) to give m.,c = 6.126 kg.
Eq. (3-12) gives:

       T =  {(239)[(6.126)(1,000) + 498.1] +
        + [(16.16 - 6.126)(1,000)(278)]}/{(16.16X1,000) + 498.1)}
        = 262.5 K

where 498.1 is the  specific heat of chlorine at constant pressure from Table 2 of the SLAB
User's guide.

Using the perfect gas law,  pa(262.5) = 1.345 kg/m3 and pg(262.5)  = 3.258 kg/m3.  From Eq.
(3-13):
                                        D-10

-------
       V =  16.16/1.345 + 1/3.258 = 12.32 m3

and p(Mf) = 17.16/V = 1.39 kg/m3 and x(Mf) =  1/V = 0.081 kg/m3. Therefore, the ordered
triple is (0.025, 0.08, 1.39) which appears on line 19 of Table 6-4 of Section 6.2.3 of the main
body of this report.
                                        D-ll

-------
        APPENDIX E

  LIST OF THOSE CONTACTED
DURING THE TELEPHONE SURVEY

-------
                                 APPENDIX E

       LIST OF THOSE CONTACTED DURING THE TELEPHONE SURVEY

As was noted in Section 2.1.1, a telephone survey of interested or potentially interested EPA
personnel was conducted in order to obtain advice on the following:

      a)    Which accident release scenarios would it be most useful  to include  in the
            guidance that EPA is developing on contingency modeling for superfund and
            other sites?

      b)    What potential communications problems may arise between those requesting
            a contingency analysis and those performing it?

The following people work for EPA unless otherwise stated.

            Kristen Harvey (Roy F. Weston)        (908) 906-3484
            Bob Cibulski                         (908) 321-6746
            Al Cimorelli                          (215) 597-6563
            Dave Guinnup                        (919) 541-5368
            Jerry Garman                         (202) 260-7767
            Steve Gilrain               .          (214) 655-6710
            Charles Hall                          (312) 353-2213
            Mark Hansen                         (214) 655-6582
            Jerry Heston                         (215) 597-7915
            Norm Huey                          (303) 293-1760
            Steve Jarvela                         (215) 597-7915
            Bill Keffer                           (913) 551-7000
            George Moein (University of Virginia)  (804) 982-5252
            Jim Mullins                          (214) 655-2273
            Joe Padgett                          (919) 541-5589
            Tom Pritchett                         (908) 321-6724
            Ann Schober                         (214) 655-6710
            Joe Tikvart                           (919) 541-5562
            Joe Touma                           (919) 541-5381
                                      E-l

-------
            Appendix F



Examples of SAFER9 Inputs and Outputs

-------
TABLE OF CONTENTS
No.
1.0
2.0
3.0
4.0
5.0
5.1
5.2
5.3
5.4
5.5
6.0
6.1
7.0
7.1
7.2
7.3
7.4
8.0
8.1
8.2
8.3
9.0
10.0
11.0

TITLE
INTRODUCTION
METHODOLOGY
SCENARIO DESCRIPTION
CHEMICAL AND PHYSICAL PROPERTIES
TOXICOLOGICAL CRITERIA
Concentration
Dose
Toxic load
Lethal exposure potential
Toxicological criteria for chlorine
RELEASE RATE ESTIMATION
Model run and results
SOURCE TERM FOR DISPERSION
Pool evaporation
Aerosol formation
Initial air entrainment
Source term for dispersion for the chlorine cylinder scenario
ATMOSPHERIC DISPERSION
Release scenario inputs
Meteorological data inputs
Results from the dispersion of the chlorine cylinder scenario
SCENARIO 2 - CHLORINE PUFF RELEASE
SCENARIO 3 ~ REFRIGERATED CHLORINE SPILL
SCENARIO 4 - ACETONE SPILL
REFERENCES
PAGE
F-1
F-2
F-3
F-3
F-4
F-5
F-5
F-5
F-5
F-6
F-6
F-9
F-10
F-10
F-10
F-11
F-1 2
F-1 2
F-12
F-1 3
F-1 5
F-1 7
F-1 9
F-26
F-30
       F-ii

-------
LIST OF FIGURES
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
TITLE
Sequence of procedures in consequence analysis
Inputs for release rate estimation for the chlorine cylinder scenario
Flow rate as a function of time for the chlorine cylinder scenario
Source term for dispersion
Release scenario definition for the chlorine cylinder scenario
Meteorological data required by the program (for all scenarios)
Centertine values of the lexicological criteria as a function of the
downwind distance for the chlorine cylinder scenario
Release scenario definition for atmospheric dispersion of the
chlorine puff scenario
Centeriine values of the lexicological criteria as a function of
downwind distance for the chlorine puff scenario
Inputs for evaporation rate estimation for the refrigerated chlorine
spill
Evaporation rate for the refrigerated chlorine spill
Release scenario definition for atmospheric dispersion of vapors
for the refrigerated chlorine spill
Centeriine values of the lexicological criteria as a function of
downwind distance for the refrigerated chlorine spill
Inputs for evaporation rate estimation for the acetone spill
Evaporation rates from the acetone pool
Release scenario definition for atmospheric dispersion of the
acetone vapors
Centeriine concentration as a function of downwind distance for
the acetone spill
PAGE
F-2
F-7
F-9
F-10
F-13
F-14
F-15
F-17
F-19
F-20
F-24
F-24
F-25
F-28
F-28
F-29
F-29
LIST OF TABLES
No
1
2
3
4
5
TITLE
Chemical and physical properties of chlorine
Toxicological criteria for chlorine
Summary of downwind distances of the lexicological criteria for the
chlorine cylinder scenario
Chemical and physical properties of acetone
Toxicological criteria for acetone
PAGE
F-4
F-7
F-16
F-26
F-27
     F-iii

-------
1.0 INTRODUCTION
In this appendix, we will study how to simulate accidental release scenarios relevant to
Superfiind sites and other sources using the SAFER® model (a chemical hazard analysis
and emergency planning software package  available from DuPont SAFER Emergency
Systems,  Westlake Village, California). The main document has several scenarios and
extensive discussion regarding the inputs required to model  these  scenarios. In this
appendix we will simulate the following four scenarios :

      Scenario 1 - Chlorine cylinder
      A flashing liquid chlorine release occurs from a Vi" orifice in a 1 ton cylinder. It is
      assumed that no  flashing occurs  before  the chlorine is released  into  the
      atmosphere. The initial temperature of the vessel is 293 °K.
      This release corresponds to the scenario discussed in  Section 6.2  of the main
      document.

      Scenario 2 - Chlorine puff
      The catastrophic failure of a 1 ton cylinder results in a puff release. The  initial
      temperature of the vessel is 293 °K.
      This release corresponds to the scenario discussed in  Section 8.1  of the main
      document.

      Scenario 3 - Refrigerated chlorine spill
      A refrigerated chlorine leak is large enough to very rapidly fill up a diked area. The
      initial temperature of the chlorine is 239 °K. The ground  temperature  is 278 °K. It
      is assumed that the spill occurs on a concrete diked area of 100 m.
      This release corresponds to the scenario discussed in  Section 5.4  of the main
      document.

      Scenario 4 - Acetone spill
      A large acetone spill from a tank very rapidly fills up  a diked area. The initial
      temperature of the acetone is  278 °K. The ground temperature is 278 °K. It is
      assumed to that the spill occurs on a concrete diked area of 100 m.
      This release corresponds to the scenario discussed in  Section 5.1  of the mam
      document.

      Meteorology for all scenarios
      For atmospheric  dispersion, a  stability class of F with a wind speed of 1.5 m/s,
      ambient temperature of 278 °K and surface roughness of 0.1m is assumed.
                                     F-l

-------
In order to illustrate the methodology we will discuss scenario 1 in detail. Accompanying
the set up of inputs for this scenario, we will provide discussion regarding variables which
are required by the program. Thus sections 2 through 8 provide general information about
the program and details regarding scenario  1. For the other three scenarios, we discuss
only those aspects of the program which require special attention and lay emphasis on the
inputs required to simulate these scenarios. Thus, sections 9 through 11 provide the inputs
and simulation results for the other three scenarios.
2.0  METHODOLOGY
In order to evaluate the consequences of a hazardous release, several sub-tasks have to be
performed. The sequence of procedures involves obtaining a scenario definition, defining
the physical and chemical properties, the lexicological criteria, estimating the release rates,
calculating the source term for dispersion,  atmospheric dispersion and finally calculating
dose, toxic load and lethal exposure potential (see Fig. 1).

The SAFER* program has several modules. Examining the nature of the scenarios and the
basic purpose of the project, it was determined that the TRACE   module would be used
for release rate estimation and defining the source term for dispersion. The source term for
dispersion was then used by the Consequence Analysis module for atmospheric dispersion
and subsequent calculations.


0
i
P


Scenario
description

ose, toxic load
Si lethal exposure
otential


^


Physical and
chemical
properties

Atmospheric
dispersion





Toxicologies!
criteria
,

Release
rate
estimates
>
Source t
for atmo
dispersic
r
erm
spheric
m


Fig. 1 : Sequence of procedures in consequence analysis
                                       F-2

-------
3.0 SCENARIO DESCRIPTION
In this stage of the process, we try to obtain a definition of the scenario. Specifications
regarding  the containment variables (tank /  pipe geometry and process variables)  are
defined along with the failure mode (e.g. catastrophic failure or hole rupture etc.). For
some  scenarios, additional information regarding diked area may also be available. More
often  than not, the scenario definition tends to be fairly general and often the analyst (the
person who interprets a real-life scenario description into a form such that the computer
model can simulate)  who runs the scenarios has to make judgments regarding variables.
Usually these  judgments are made  from past experience, case  studies and / or by
conferring with the appropriate plant personnel or emergency  responders.  During our
simulations for the four scenarios this will become apparent, as the  complete scenario
definition will not be supplied and we will make assumptions in order to obtain results
from the models.

A flashing liquid chlorine release occurs
from a Vin orifice in a 1 ton cylinder. It is
assumed that no flashing occurs before the
chlorine is released into the atmosphere.
The initial temperature of the vessel is 293
°K. For atmospheric dispersion, a stability
class of F with a wind speed of 1.5 m/s,
ambient temperature  of 278 °K and
surface roughness of O.lm is assumed.

This release corresponds to the scenario discussed in Section 6.2 of the main document.
4.0 CHEMICAL AND PHYSICAL PROPERTIES
In order to run any simulation for a
particular chemical, the program first
requires the user to define specific
chemical/physical properties. The required
chemical and physical  properties of
chlorine  are defined in Table 1.
                                      F-3

-------
1   Molecular weight (gin/mole)
2   Critical pressure (atm)
3   Critical volume (cmA3/mole)
4   Critical temperature (°K)
5   Normal boiling point (°K)
6   Surface tension of liquid (dyne/cmA2)
7   Liquid viscosity at NBP (cP)
8   Liquid density at NBP (gm/cmA3)
9   Liquid density at 5 °K below NBP
10  Temp when vap press is 400 mmHg (°K)
11  Enthalpy of satd. liquid (cal/mole) at
    NBP
    10 °K above NBP
    20 °K above NBP
12  Cp coefficients (cal/gmol-°K) of gas phase
    where Cp=A+B*T+C*TA2+D*TA3
                70.91
                76.00
               124.00
               417.00
               238.70
                25.60
                 0.49
                 1.56
                 1.58
           :    225.40

                 0.00
               161.00
               327.00

A=6.432   B=0.008
C=-.924e-5  D=0.00
           Table 1  : Chemical and physical properties of chlorine
The chemical and physical properties to be defined  include the molecular weight, the
critical properties (pressure, volume and temperature), the normal boiling point,  surface
tension, liquid viscosity, liquid density, vapor pressure data, enthalpy data and specific
heat of the gas phase. The program comes equipped with a large library of commonly used
chemicals. Additionally, the program has features by which new chemicals can be added.
Several standard references (e.g. [1],[2]) provide the values of the variables which are
required by the program while defining a new chemical.
5.0 TOXICOLOGICAL CRITERIA
To evaluate the consequences of the
release, the program requires the user to
define one or more sets of toxicological
evaluation criteria. These toxicological
criteria relate to the maximum
concentration,  dose, toxic load and lethal
exposure potential at a fixed location. The

program evaluates the concentration versus time profile for different receptors and uses
this profile to  calculate the other three criteria (i.e. dose, toxic load and lethal exposure
potential).
                                      F-4

-------
5.1 Concentration
The concentration limits are usually determined by the final goal of the study. In order to
evaluate toxicity hazards, concentration limits like IDLH (Immediately Dangerous to Life
and Health), STEL (Short Term Exposure Limits), ERPG (Emergency Response Planning
Guideline) and EEGL (Emergency Exposure Guidance Limit) have been used. Associated
with each one of the concentration limits, there is an averaging time (IDLH - 30 min.,
STEL-15 min. etc.).

5.2 Dose
The dose represents the amount of chemical to which  one is  exposed over a period of
time. For our application, the dose will be expressed in units of PPM-MIN.
                                   Dose = JCdt
5.3 Toxic load
The toxic load is similar to the dose, but additionally it takes into account  the specific
toxicity of the chemical. The toxic load is generally evaluated as :

                                Toxic load = Jcndt

where n is commonly referred to as the toxic exponent. The value of n has been calculated
for several chemicals (see [3] and [4]). For our application, toxic load will have the units
of PPM -MIN where n is the chemical specific toxic exponent.

5.4 Lethal Exposure Potential
The lethal- exposure potential  is evaluated by using the probit model ([5]).  The  probit
model provides a convenient methodology by which one can evaluate the fraction of an
average population  which will have a particular response (e.g. odor, serious toxicity,
lethality) for a specific  toxic load exposure. The basic assumption of this model is that the
response plotted  against the logarithm of concentration or logarithm of exposure time has
a cumulative normal distribution [4]:


                 Probit = a + b * Ln(Toxic Load)
                                             .j  Probit-5
                 Lethal Exposure Potential = -_  Jexp (-u2  / 2) du


where a and b are the chemical specific probit constants (usually evaluated by a regression
analysis of the lethality data) and n is the appropriate toxic exponent.
                                      F-5

-------
5.5 lexicological  criteria for chlorine
Since chlorine is a very widely used chemical it is possible to define  all   toxicological
criteria. The concentration limits are defined as ERPG-1, ERPG-2, and ERPG-3.  The
ERPGs (Emergency Response Planning Guidelines) are limits which have been defined by
the American Industrial Hygiene Association [6] and are commonly used to assess the
impact of accidental releases. It is to be noted that the ERPG limits  represent 1 hour
exposure numbers.

For our specific application (i.e. Superfund sites) the ERPG-2 and the ERPG-3 represent
levels of most interest (i.e. they represent concentration limits where preventive measures
would be activated). However, due to the requirement of 3 levels  by  the program, we
define the additional third level to be ERPG-1.

The dose limits were defined from  the ERPG limits. Thus, the dose  limits effectively
represent an equivalent  "ERPG dose". The limits are obtained by multiplying each of the
ERPG limits by the exposure duration (i.e. 60 minutes).

The toxic load limits were defined using the methodology for the "Dangerous Toxic Load"
proposed by the Health & Safety Executive,  UK [7].  Using the values  suggested  in [7],
the toxic load medium limit was assigned to be the HSE Dangerous Toxic Load. The low
and high limits are arbitrarily assigned to be  some multiple of the HSE Dangerous toxic
load (divided by 5 and multiplied by 5).

There are several references  which provide probit constants for chlorine ([4],  [8]-[12]).
For this particular simulation, we used the probit constants proposed by Withers and Lees
[8]. Additionally, the lethal exposure potentials were assigned to be 20%, 50% and 80%
(these limits would produce  the LC-20, LC-50, LC-80  curves —  LC-x represents the
Lethal Concentration for x percent of an average population  for a particular exposure
period).

The complete toxicological criteria for chlorine are shown in Table 2.
6.0 RELEASE RATE ESTIMATION
In order to define the release rate as a
function of time from the tank, the
program requires an extensive set of
inputs. These inputs relate the tank
contents, tank geometry and dimensions,
hole geometry and dimensions and the
meteorological conditions. For the
particular chlorine cylinder scenario the
input screen is as illustrated in Fig. 2.
                                      F.-6

-------
CONCENTRATION
Comment
Concentration limit low [PPM]
Concentration limit medium [PPM]
Concentration limit high [PPM]
DOSE
Comment
Dose limit low [PPM-MIN]
Dose limit medium [PPM-MIN]
Dose limit high [PPM-MIN]
TOXIC LOAD
Comment
Toxic load exponent (n)
Toxic load low [PPMAn-MIN]
Toxic load medium [PPMAn-MIN]
Toxic load high [PPMAn-MIN]
PROBITS
Comment
Probit exponent (n)
Probit constant (a)
Probit constant (b)
Lethal exposure potential limit low (%)
Lethal exposure potential limit medium (%)
Lethal exposure potential limit high (%)
Table 2 : Toxicological
Status: ON
ERPGs
1.00
3.00
20.00
Status: ON
ERPG Dose
60.00
180.00
1200.00
. Status: ON
HSE Danger. Toxic Ld
2.00
21600.00
108000.00
540000.00
Status: ON
Withers & Lees
2.00
-8.29
0.92
20.00
50.00
80.00
criteria for chlorine
      TnttK  KUPTUHE SCEMilHIU  :            -,

 1J  : Tank  Rupture File; Hame;   Tank  Rupture Scenario Bescrlptlon;;::            •'••' 1
 3 >•.;;..' Chcntitra 1 Mumhuir     	•:.:,:.•::: v•.:•:,. ;.. ,.•': ,.••;:.''.'."::...-;. :.v..;.:'..:" "...:... '.;:::'.-;;..':.:..
 'IJ   CJ>en»t c»1 Hamc
 5i   Temperature: Inside The Tank                         td
 O>   Mass  Of Chemical                    top-tlonal*;
 7>   Liquid Level Inside  The  Tank      {optional}
 U}   1'rcssure  Inside: The  Tank   if of gases onlij>        •
 9>   Tank  Tyjie  ;i   -Cl-recta
                    : £3=cyJ:  «jcr-tleal *»si»liertcal:      j:
1Q>   Tank  LeTistliW    Cf ti:  fr.aQOGei    t!3  Tank Hei{jht
12J:  TanR: Uidth;      (rt)5;       O.    13)  Tank Dl«metc»
I'll:.  Wall: Thielmess: Cf t>::; : :'::.'/Q:.O3;x'-;,:;:. -•  .•'•,.:.',.. :..,. .  v ,,::.v,: .:..v:..;:
ii>)   Hale: Type     ll=ctrc:i swootht  Z«reet!::;-s«iwothJ';:;.;"':};::;/-'.iv'.;;:;
i                  :   O—eirex jayijed  -t—reet.  jaifjied} •  :  : :V
IG>   DIscliarcfcs Coci;r     -       :  0.    17)  Hoie ntametcw
18>   Holtr  Uiath      Cft>:         Q.    ig >•• Hulc Height
:o>   Hole  Center HeIflh-t ftbove: Tank Bottom
                                                               1 TOM CUt  CYLIMDEH
                                                              •••     •::.'• ••. -••••••• &.- • -•• •-.:.-:.
                                                                        CHLOHINK
                                       Z~cyl- horizontal>s
Jl>   Wind  Sjieed
                                           22J nmbierit Temp  CdegFJ•;
          Enter- $>eIection;: number' or HETUMtt? to;- exit
Fig.  2: Inputs for release rate estimation of the chlorine cylinder scenario
                                             F-7

-------
Some of the inputs are self-explanatory. The  particular inputs which  require special
attention are the following :

     Input no.                                 Comment

        3,4          Specify the chemical used for the simulation.
                     For  our simulation, chlorine is identified as number 6 by the
                     program's chemical library.

         5           Specifies the tank temperature.
                     For our simulation this is 68 °F (293 °K).

        6-8          Specifies the contents of the tank. Any one of the three inputs is
                     to be defined.  The program estimates the other two variables
                     depending  upon  the tank  geometry   and chemical/physical
                     properties.
                     For  our simulation we specify  the total mass (2000 Ib.) within
                     the cylinder.

        9-14         Geometry and tank dimensions.  Depending upon the input 9, one
                     or more inputs  10-13 should be defined. For example, if input 9 is
                     selected to  be 2  (cylindrical horizontal)  then the  user  should
                     define input 10 (tank length) and input 13 (tank diameter).
                     For our simulation \ve choose a standard 1 ton chlorine cylinder
                     (6.8ft long and 2.5ft diameter).

       15-20         Specify the  geometry, dimension and location of the hole within
                     the tank. One or more inputs 17-19 should be defined depending
                     upon input  15.  For example,  if input  15  is chosen to be  1
                     (circular smooth)  then the user needs to  define input  17 (hole
                     diameter). If input 16 is left undefined  (=0) then the program
                     assigns a default value of the  discharge coefficient (depending
                     upon input 9). For gas releases,  the default value of the discharge
                     coefficient  is  0.61. For liquid  releases,  the  values are  0.65
                     (circular smooth),  0.622  (circular jagged),  0.625  (rectangular
                     smooth) and 0.828 (rectangular jagged).
                     For  our simulation we choose a circular smooth hole which is
                     l/2n (0.04ft) in diameter. Additionally, the hole is located at the
                     bottom of the tank

       21,22         Wind speed and ambient temperature.
                     For  our simulation, the wind speed is 3.35 mph (1.5 m/s) and
                     ambient temperature is 41 °F (278 °K).
                                       F-8

-------
6.1 Model run and results
When the above release scenario is executed, the program indicates that the "single-phase
liquid flow" and "sonic gas phase" equations were used. Reference [13]-[15] outline the
equations used by the program.

Additionally, the program performs a transient flow rate calculation where the release rate
is estimated as a function of time. This is performed  by assuming a sequence of "pseudo-
steady states". Knowing the flow rate at a particular instant, the rate is kept constant for a
time interval which is dependent upon an incremental spilled mass. Subsequently, the new
state of the system (temperature, pressure, liquid/vapor breakup)  is calculated and a new
release rate is estimated ([13], [14]).

It is observed that the initial flow rate is 6.97 Ib./s and gradually decays to 5.86 Ib./s after
294 seconds. The complete tank empties in about 341 seconds resulting in an average flow
rate of 5.84 Ib./s. Additionally,  the program indicates that the liquid flashes resulting in a
21% vapor fraction and a temperature of-30 °F .
     Fig.  3 : Flow rate as a function of time for the chlorine cylinder scenario
                                       F-9

-------
7.0 SOURCE TERM FOR DISPERSION
It is at this stage that the program
calculates the evaporation rate from a
pool, performs flash calculations and
models phenomena close to the source —
aerosol formation and initial air
entrainment (Fig. 4).
R«|«MM rat*
avnmotm
\
(
                                      Initial air entrained
                                                            Source term
                                                               for
                                                            dispersion
                      Fig 4 : Source term for dispersion
7.1 Pool evaporation
The evaporation  rate from  the  pool  is  calculated  by taking  into account  the  heat
conduction from the ground, mass convection due to the ambient wind and by calculating
the heat input from the solar radiation. Heat conduction from the ground becomes the
driving force  for cryogenic liquid spills (refrigerated chlorine, ammonia, LNG etc.). The
convection is  important for normal boiling liquids (those that have a boiling point above
typical ambient temperatures). The solar radiation can become an important source of heat
input for low volatility liquid pools. The algorithms used to calculate the evaporation  rates
are described  in [13] and follow similar approaches outlined in [16]-[20].

7.2 Aerosol formation
Aerosol formation is specified to the program by using  two different modes [21]:

       OManual mode
       The user specifies the aerosol to flash mass ratio
                                     F- 10

-------
       ii)Default mode
       By using the default option, the program calculates the aerosol fraction. For low-
       flashing  fraction, the program splits the inlet stream into a flashed vapor and a
       liquid pool. For extremely high flashing, the program splits the inlet stream into a
       flashed vapor and liquid droplet aerosol. For intermittent flashing fractions, the
       program would split the inlet stream  into a partial liquid droplet aerosol  and
       remnant pool liquid.

       Let   F = Flashing fraction (determined thermodynamically)
             A = Liquid droplet (aerosol) fraction
             G = Pool liquid fraction

       By normalizing with the total inlet stream
             F+A+G=1

       Define the two cutoff flashing fractions : Fl (=0.05) and F2 (=0.25) where FKF2
       ForFF2
                    A=1-F and G=0
       ForFK=F<=F2
                    A=(1-F2)*(F-F1)/(F2-F1) and G=(1-F1)*(F2-F)/(F2-F1)

       Note,  the  above  scheme  is  approximate and to be  used only  when other
       information is not available. When specific information is available about a release
       scenario, it may be better to specify the actual aerosol fraction by using the Manual
       mode for Aerosol formation (i.e. Section 7.2, option i).

7.3 Initial air entrainment
The initial air entrainment is associated with the amount of air that is entrained into the
cloud during a catastrophic release or a pressurized release and fairly close to the source.
The program allows two different modes:

       OManual mode
       In this selection, the user specifies, the air to total chemical mass ratio.

       ii) Default mode
       In this mode, the program performs a set of iterative calculations and entrains just
       sufficient air to evaporate all liquid aerosol droplets. Usually, this is the state at
       which the cloud of maximum initial density of and minimum cloud temperature is
       produced.  The algorithms used for calculating  the  amount of air entrained are
       described in [22] and follow similar work outlined in [23].
                                      F-ll

-------
7.4 Source term for dispersion of the chlorine cylinder scenario
Our particular scenario involves the pressurized release of chlorine resulting in a flash. It is
generally expected that there would be minimal pool formation under such conditions and
even in the case where  there was  pool formation,   the chlorine would evaporate very
rapidly due to heat conduction. Thus for our scenario, it will be assumed that no pool is
formed, there is a flashing vapor stream of 1.23 Ib./sec (21% of 5.84 Ib./sec) and the
remaining 4.61 Ib./sec (79% of 5.84 Ib./sec) will form liquid aerosol droplets. Further, we
use the default initial air  entrainment option and entrain just sufficient air to evaporate all
liquid aerosol.
8.0 ATMOSPHERIC DISPERSION
The Consequence Analysis program is
designed for doing detailed planning and
evaluating the consequences of accidental
releases. The atmospheric dispersion
model accounts for dense gas dispersion
and takes into account gravity slumping
followed by subsequent transition to a lean

gas (Gaussian) model. A detailed review of the dispersion algorithms in this program is
beyond the scope of this study but details are provided in [13] and obtainable on request.
The basic dispersion routines have been compared to field test data and evaluated in
independent studies (e.g. [24]).

8.1 Release scenario inputs
In order to run a simulation in the Consequence Analysis module,  we  have to define a
release scenario.  The release scenario was defined using the values developed in Section
6.4. The specific inputs for our scenario are illustrated in Fig. 5.

Some explanations  regarding the inputs on the release scenario screen:

    Input                                 Comment

      1        Type of release, i.e. single-phase or two-phase
              For our scenario, we select two-phase.

      2       Release rate type, i.e. 1 instantaneous,  2=continuous.
              For our scenario we choose continuous.

      3        Chemical Identification number.
              Chlorine is identified as number 6 in the program's chemical library.
                                      F-12

-------
NOTE:
               Duration of the release.
               For our scenario this value is calculated to be 341 seconds.

               Initial phase of the release (l=liquid, 2=gas or 3=two-phase).
               For our scenario we choose two-phase . Note, we choose the two-phase
               option because in the following entries we will specify the vapor and
               liquid aerosol fractions and ask the program to calculate the initial air
               entrained  required to just evaporate all liquid aerosol.

               Liquid release mass rate - this specifies the aerosol fraction.
               For our scenario,  this was determined to be 4. 61 Ib./sec.

               Gaseous release mass rate — this specifies the initial vapor fraction (may
               have originated from flashing).
               For our scenario,  this was determined to be 1.23 Ib./sec.

               Initial release  temperature.
               For our scenario we choose -30 °F.

               We did not explicitly choose an initial air entrainment value (discussed
               in Section 6.3). Hence,  the program uses the Default mode ii)  and
               entrains just sufficient air to evaporate all liquid aerosol.
       nUDiFY  fflHflflETEB UftLUtS
      MENU   130 :   SCENARIOS
      NEXT MENU :      BO
      ITEtt      DESCBIPTION
                                                     SELKCTIUM 13 UF <21 =
                                              SELECTION  r SUPEH1B
                                              FIELD CODE ;    19
                                              :v U»LUE -:Y;-'.:----:: '•-:-: :- UNITS:
               Two-phase madel!V-;.':'S;.il'.: ""•''-.-.
               Release rate type- •!!>•:
               Chemical ID • number: I'•'.•; • •
               Duration   • •  -,-•••.•.-., •••^•:,.:..''• '!":::,'.'
               :.Phase  •• •:-.«S :-i-x •^•••-•.'-<:A • ;•;•,••• :;y:™;'>i;.
               Lirju id;: release mass:: r*te:>:::
               Gaseous: release: mass: rate:
               He 1 ease; temperature •::   : ::
                                                       .QQQQQQ
                                                      2.860000
                                                      &.OGQ00Q
                                                    341.800000
                                                      3.QQQ00Q
                                                      4.603898:
                                                      1.230G0Q
                                                    -3Q.eoooa0
     Fig. 5: Release scenario definition for the chlorine cylinder scenario
8.2 Meteorological data inputs
The program requires that  data regarding the wind speed, wind direction, ambient
temperature, solar radiation, relative humidity and stability class be defined.
                                       F-13

-------
      MENU   2O :  STflMDflBD COMBH
      NEXT MENU t       1
      ITEM     UESCHIPTICm
         i      Manual Met Flag
               Man in Obhukhau length
               WIND  IS FROM
               Wind speed
               Ambient temperature
         »      Stability class Umrz 4 uert)
         ?      Surface roughness ;
         I-     Insolation
         »      Helntfup TiumifHtii
,^-T-.-^ SELECTION
 SELECTION   : EPfl
 FIELD  CODE  :
   UALUE
          ,000000
          .QQ0Q00
     SUBMENU 684
        3.345000
       •11.000000
        &. 000000
          .330000
          .000000
UNITS
Fig. 6:  Meteorological data required by the program (used for all scenarios)
Some comments regarding meteorological inputs are :
    Input                                  Comment
      1         Type of meteorology (Real-time or Manual mode).
               Since we are in the Planning mode \ve use Manual Meteorology (i.e.
               the user will input all meteorological variables).

      2         Monin-Obhukhov length.
               We use the default option, by which the program calculates the Monin-
               Obhukhov length as a Junction of surface roughness and stability class.

      3         Wind direction

      4         Wind speed - the value specified should be the measured value at 10 m
               above the ground.
               For our simulation, we choose 3.35 mph.

      5         Ambient temperature.
               For our simulation we specify 41 °F.

      6         Stability class (Very Unstable A=l, B=2, C=3, Neutral D=4, E=5, Very
               Stable F=6).
               For our scenario we specify 6.

      1         Surface roughness.
               For our scenario we specify a value of 0.33 ft.

      %         Insolation.
               For  this  scenario  we  choose   no  solar   radiation.  Note,   the
               meteorological conditions which we specify would most probably occur
               at night-time (F stability, 1.5 m/s) and hence this value is chosen to be
               zero.
                                      F-14

-------
      9        Relative humidity.
               For our scenario \ve choose a value of 50%.

8.3 Results from  the dispersion of the chlorine cylinder scenario
The program displays the impact of the
release in terms of cloud footprints
(cumulative area swept by the cloud) and
has options to evaluate indoor/outdoor
concentrations and the effect of wind
direction. In this section we will simply
examine the most basic of outputs, i.e.
evaluate the toxicological criteria

(concentration,  dose, toxic load and lethal exposure potential) as a function of downwind
distance and also evaluate cloud footprints in terms of the four basic criteria.

In Fig. 7 we illustrate the variation of the centerline values with downwind distance. It is
observed that the ERPG-1 level  (1 PPM) reaches 17 miles, the ERPG-2  level  (3 PPM)
reaches 10 miles and the ERPG-3 level (20 PPM) reaches 3.7 miles.
i ooe+n
1.00E+10
I.OOEfOl
i.ooE+oa
1.00E-K07
t.OOE + 04
1.00C*09 .
1.0OE + 04
1.00E*03
1.00E + 02
t.OOE + 01
i.ooe»oo
1





















































LMtMl Cxp<

^^^



OM

mn>

Mnr«

X,



• II




Po

V
>


fM -HI
•*»,
-«.


witM (

to

\
\

Ml

"•>
^

»)




^s


»«N
•>v










".







s»


-*.
m *


too
0«OTHOTM4 DlMWIM 4M«




-------
Dangerous Toxic Load which is primarily defined for land-use planning around industrial
facilities reaches 1.2 miles. Lastly, it is observed that lethal concentration for 50% of an
average population would be achieved approximately 0.5 mile downwind.  It should be
emphasized that this simulation creates a theoretical worst-case situation, where people
are not sheltered and  do not take mitigating measures. As such, in a real-life situation
depending upon actual circumstances and the course of events different consequences may
be expected.
      TOXICOLOGICAL
         CRITERION
            COMMENT
DOWNWIND
 DISTANCE
  (MILES)
     Concentration (PPM)
              1
              3
             20
        ERPG -1 for chlorine [6]
        ERPG-2 for chlorine (6]
        ERPG -3 for chlorine {6]
    17.0
    10.1
    3.7
      Dose (PPM-MIN)
             60
             180
            1200
                                          5.3
                                          2.9
                                          0.9
   Toxic load (PPMA2-MIN)
           21,600
           108,000
           540,000
 HSE Dangerous Toxic Load for chlorine [7]
    2.0
    1.2
    0.7
   Lethal Exposure Potential
            20%
            50%
            80%
LC-20 for scenario using probit constants [8]
LC-50 for scenario using probit constants [8]
LC-80 for scenario using probit constants [8]
    0.6
    0.5
    0.3
    Table 3: Summary of downwind distances for different toxicological
                    criteria for the chlorine cylinder scenario
                                      F-16

-------
9.0 SCENARIO 2 - CHLORINE PUFF  RELEASE

SCENARIO DESCRIPTION
The catastrophic failure of a 1 ton cylinder results in a puff release. The initial temperature
of the vessel is 293 °K.
This release corresponds to the scenario discussed in Section 8.1 of the main document.

CHEMICAL AND PHYSICAL PROPERTIES
The properties of chlorine which were defined in section 4.0 are used.

TOXICOLOGICAL CRITERIA
The toxicological criteria for chlorine defined in section 5.5 are used.

RELEASE RATE ESTIMATION
For this scenario, there are no specific calculations which have to be performed to estimate
the release rate / amount. It will simply be assumed that the entire contents of the 1  ton
cylinder is released to form an instantaneous puff.

SOURCE TERM FOR DISPERSION AND RELEASE SCENARIO INPUTS
Since the initial cylinder is at 293 °K and the ambient temperature is 278 °K, there will be a
flashing vapor fraction of'X/21% (see section 7.4). Thus we will specify an amount of 430
Ib. flashed vapor and 1570 Ib. of airborne liquid aerosol (droplets). The entire system will
initially be assumed at the boiling point (-30 °F).  Further because  of the catastrophic
release, we will use the assumption that just sufficient air is initially entrained to evaporate
all liquid aerosol. The release scenario  is defined in Fig. 8. An explanation of the inputs is
as follows:
       tlQDlEV PftRftflETER UftLUES
      MENU  130 ;  SCENARIDS
      NEXT MEMU :     8O
      ITEW     DESCRIPTION
               Two-phase nradet  : ;
               He 1 ease rate type; ;   ;   :  ':;
               Chemical ID:^number;
              •• Durat-1 onSfVy;-yy.y'-vyy;yy: yyy yyyyy:: • •••;•••;•;•;•;•;
               Phase  ....••::: • y y" ,..'.'••.- .;•-.-.:, •.':•  ;..?.v'..:•''.
               Liquid release mass: amount:
               Gaseous release mass  amount;
               Heleaseitemperaturey/yyyyy :;;;;•:
       • SELECTION  16  Of 21  '
 SELECTION   :  SUPERZft
 FIELD^' CODE ;    13    	:.;:;,:'-:T.-:;::
   UrtLUE              UNITS
.:'::; •. :'•::'::.;::-;-: -.'   . QQQGQQ      ••: •••• :::.;:v:"
:;oy-:--:-;• -i-;-'-.:: H;  Z.OOGQQQ  V  ;• ': :-:'- "•':-;;:
•. ?!.:.::.::;::; ;•:v •:::;:::  &. OQQQQQ  :    ::3;;;-
•yy:•::•:• yx:';y::::y'•• .QOOQQQCyyySHiy sec;vV*
-v-y:-•-•--.••••:'' 3.QQQQQ&
     1578.000000  •   1
       439.000000     1
       -33,000Q8Q;;  i  deqF
 Fig. 8 :  Release scenario definition for atmospheric dispersion of chlorine
                                 puff scenario
                                      F-17

-------
    Input                                  Comment

      1        Type of release, i.e. single-phase or two-phase
              For our scenario, we select two-phase

      2        Release rate type, i.e. l=instantaneous, 2=continuous.
              For our scenario \ve choose instantaneous.

      3        Chemical Identification number.
              Chlorine is identified as number 6 in the program's chemical library.

      4        Duration of the release.
              Since this is an instantaneous release, the duration is set to be 0.

      5        Initial phase of the release (l=liquid, 2=gas or 3=two-phase).
              For our scenario we choose two-phase. Note, we choose the two-phase
              option because in the following entries we will specify the vapor and
              liquid aerosol fractions arid ask the program to calculate the initial air
              entrained required to just evaporate all liquid aerosol.

      6        Liquid release mass amount — this specifies the aerosol fraction.
              For our scenario, this was determined to be 1570 Ib.

      1        Gaseous release mass amount - this specifies the initial vapor fraction
              (may have originated from flashing).
              For our scenario, this was determined to be 430 Ib.

      8        Initial release temperature.
              For our scenario we choose -30 °F.

   NOTE:     As before, we did not explicitly choose an initial air entrainment value
              (discussed in Section 6.3). Hence,  the program uses the Default mode
              ii) and entrains just sufficient air to evaporate all liquid aerosol.


RESULTS FROM THE DISPERSION OF THE CHLORINE PUFF SCENARIO
The results  of the above defined  release  scenario are illustrated in Fig. 9.  Of special
interest, is the fact that the toxic load and dose go through a maximum (fairly close to the
source)  while  the  concentration  has  a  monotonic  decreasing  behavior.  This  Is
characteristic of instantaneous heavy  gas releases.  Qualitatively, this behavior can be
explained as follows. The toxic load and dose are functions of the concentration and the
total  exposure time  for the release.  For instantaneous heavy gas releases,  the gravity
slumping adds to the  growth of the radial  cloud dimensions leading  to an increased
exposure time. This  increase in the exposure time overrides   the decrease  in cloud
                                      F-18

-------
concentration  which  occurs  due  to  dispersion.  Hence  the  resulting  product  (of
concentration and time) has an increasing - decreasing behavior.
1. OOE* II
r. OOE* 10
1.00£*O9
t.OOE»O8
i.ooe*07
1. OOE* 08
1.0OE.O5
t.OOE»O4
1.0OE»O3
1. OOE* 01
1. OOE tOO

JB— -



^*




*=



^




.••




tf




^




*
•




.-•




-•'
• • •






... 	




.,- 	
1 . HMMM
Cdfwcntfi

UlhalE

*S




»m,
""5
M«n

WO.

S




•«.
X
(pp

uf*


s



•»
•»
Ml

Pn


V



s
^».


t«ntia

10



I
^

3o.. (PPM-
"\
"-v^

<%>




s


h^^
^«






T
S


x






»»
s


s






0
S


0
>w
^




Lama
^


"•
«.
N





(PPM- 2-1

^
>
.^^
<

\



MM


\

rv









s
s,
•s








V

s










^
s<









S
s











*
sN
^
























































too tooo loooo 100000
0»»ii»l~< Olxno. (mwt
Fig. 9: Centerline values of the toxicological criteria as a function of
downwind distance for the chlorine puff scenario
10.0 SCENARIO 3 - REFRIGERATED CHLORINE SPILL

SCENARIO DESCRIPTION
A refrigerated chlorine leak is large enough to very rapidly fill up a diked area. The initial
temperature of the chlorine is 239 °K. The ground temperature is 278 °K.  It is assumed
that the spill occurs on a concrete diked area of 100 m .
This release corresponds to the scenario discussed in Section 5.3 of the main document.

CHEMICAL AND PHYSICAL PROPERTIES
The properties of chlorine which were defined in section 4.0 are used.

TOXICOLOGICAL CRITERIA
The toxicological criteria for chlorine defined in section 5.5 are used.
                                    F-19

-------
RELEASE RATE ESTIMATION
For this scenario, the precise containment configuration (tank or pipe) is not defined. The
scenario basically requires, that a large release has occurred such that a diked are of 100
m  is  filled with chlorine  liquid. Hence for our simulation, we will assume that  an
instantaneous spill of 10,000 Ib. (5 tons) of refrigerated chlorine at -31°F.  This amount of
chlorine will spread rapidly and fill up the diked area.


SOURCE TERM FOR DISPERSION
For this scenario, it will be assumed that the complete release amount is "dumped" into the
pool and the evaporation rate from the pool is the source term for dispersion. In order to
model  this evaporation process, the program requires inputs which are displayed in Fig.
10.
An explanation of the various inputs is as follows:
                                       1)  File name :    SUFEB5
                                              REFRIGCLZ SPILL
                                              CHLOHIME
                                                   100QQ.
                                                      O.
                                                   •:: -31,
      RELEftSE SCENARIO SETUP           1) File name :
 2)  Scenario description  ;                   REFRIG Ct
 3)  Chemical reference number  :      6       CHLOHIME
 5)  Release type  tl=cont>: 2=tnst. 3=tranX :;   ;;; 2
 6): Phase of chemical    Temperature of chemical  (de^IT)       :         -3
1O)  Elroation of  release  (ft)
tl)  Uertical velocity.-.;..(ft/sec)-.
12)  Horizontal velocity (ftxsec)
13)   Initial radius (ft>
14)  ftir/Chewlcaf  woie rvitia Clnltlal dlluttonl
15)  Maximum pool  area ••:•.. fterosoi air entrainment tll^tlanual J,. 2=Default)
:1)  ftip/Chewical  mass ratio    (ttZO. Manual option only)
:2)  Substrate   CO=W,1-C,IZ-ftsai11,3=SDso il,4=SMsoiIi
3)  Temperature? of;; substrate- (degF):  ;: V ;: ; :.       :
lommand  :              •';'.•••!  '     -'  :"" •'-•'••':•;'•..    ': '  :'.::
,'omment: : Enter Selection  number or- BETURH to exit:
                                                             1076.39
                                                           o.oazaea
                                                               Q.1S
                                                                       (gas)
dig)
dig)
tliq)
tiiq)
tliq)
tliq)
Clltj)
 Fig.  10: Inputs for evaporation rate estimation for the refrigerated chlorine
                                      spill
  Input no.                                 Comment

     1,2       Specifies the file name and an  associated description regarding the
               scenario
     3,4       Specify the chemical used for the simulation.
               As mentioned previously,  chlorine is identified as number  6 by the
               program's chemical library.
                                       F-20

-------
  5        Specifies   the   type   of  release   (1 instantaneous,  2=continuous,
           3=transient, i.e. time-varying).
           For our simulation we specify an  instantaneous  release due to the
           "dumping" of chlorine into the pool..

  6        Specify the phase of release (1 =liquid and 2=gas)
           For our simulation, we specify a liquid release.

  1        Specify the release size or release rate. Depending upon the selection
           made for input 5, we have a release rate (for steady continuous releases)
           or the total amount spilled (for instantaneous releases).
           For our simulation we will specify the total amount spilled as 10,000
           Ib.

  8        Specify the duration of release. For instantaneous releases, this variable
           is set to be zero. For steady  continuous releases, specify the total time
           over which the spill occurs.
           For  our simulation,  we  specify 0  since  we  have an instantaneous
           release.

  9        Specify the temperature of the release.
           For our simulation we specify -31 °F.

10-14      Specify the initial source characteristics related to GAS releases. These
           variables  define  the  elevation  of release, the initial velocity (both
           horizontal  and  vertical),  the initial  radius (hole size) and  the initial
           dilution.
           For  our simulation,  we have  a liquid amount that  is spilled.  Thus
           variables JO-14 are ignored and set to be 0.

  15        Specify the maximum  pool area (area contained within the dike).
           For our simulation, this is set to be 1076ft2 ( 1-/00 m2 ).
                                   F-21

-------
 16        Specify the minimum pool depth.
           For our simulation, we choose the default value of 0.032 inch (^1 cm).
           NOTE: Variables 15 and 16 are used in conjunction to determine the
           dimensions of the  pool. The program takes into  account the radial
           growth of the pool. The pool is allowed to grow until a maximum pool
           radius (determined by the  maximum  allowable pool area) is reached.
           Simultaneously, the minimum  pool depth constraint  should also  be
           satisfied. Thus the pool  keeps  on "growing" as long as the minimum
           pool depth is maintained and the spill amount/rate exceeds the pool
           evaporation  rate.  In scenarios  where  the spilling has  ceased or  in
           scenarios where the pool evaporation rate exceeds the spill rate, the
           pool may "contract" while maintaining the minimum pool depth.

 17        Specify the pool albedo (fraction with a value from 0  to  1). The pool
           albedo  represents  the  reflectivity of the  pool  surface.  (1 -albedo)
           specifies the fraction of solar radiation that is incident upon the pool
           surface and which is absorbed by the  pool liquid. The pool albedo is a
           function of the pool surface (shiny, dark, transparent, opaque etc.) and
           the angle of the incident radiation. By default, the program assumes an
           pool albedo  of 0.15. Thus in the  default mode, 85 % of the incident
           solar radiation is actually absorbed by the pool liquid.

           The albedo of a plane water surface typically varies as:

           Elevation of
           sun (degrees)         90     50     30 '   20     10     5      0

           Albedo of water     0.020  0.025 0.060 0.134   0.348  0.584  1.0

           For our simulation, we will use the default value ofO. 15.
1 8-2 1      These variables are used to estimate the aerosol  liquid  fraction and the
           initial air entrainment. A detailed discussion of these inputs  is given  in
           section 7.2 and 7.3.
           For our simulation we choose default values (2= Default) for inputs 18
           and 20. Note, in our case,  we specify a temperature slightly less than
           the  boiling point so that no flashing occurs and the complete spilled
           amount is "dumped" into the diked area to form a liquid pool.
                                   F-22

-------
     22        Specify the substrate on which the pool is formed. The characteristics of
               the substrate (density, specific heat,  thermal conductivity, and thermal
               diffusivity) are used for calculating the  heat conduction flux from  the
               substrate into the pool. The program has the following options:

               #      Substrate       Density Specific heat   Thermal Cond.  Thermal Diff.
                                    kg/m3  Joule/kg-°K   Watt/m-°K    m2/sec
1
2
3
4
Concrete
Average soil
Sandy dry soil
Sandy moist soil
2300
2500
1650
1750
961.4
836.0
794.0
1003.2
0.92
0.96
0.26
0.59
4.16 e-7
4.59 e - 7
1.98e-7
3.36e-7
               For our simulation, we choose the substrate to be concrete.

     23        Specify the temperature of the substrate. This variable is also used for
               calculating the heat conduction flux from the substrate into the  pool.
               For  certain substrates  under some  meteorological  conditions  (e.g.
               asphalt substrates on a hot summer day, or snow/ice of a "sunny" day)
               the temperature of the substrate may be significantly different than the
               ambient atmospheric temperature.
               For  our simulation we will specify the substrate  temperature to be
               identical to the ambient atmospheric temperature i.e. 41 °F.
When the program simulates the above specified pool evaporation scenario, it produces an
evaporation rate as a function of time. This rate is illustrated in Fig. 11. It is observed that
the rate has a peak value of ^7 Ib./sec initially and then decays to % 0.7 Ib./sec at the end
of an hour. Note, by very definition of the scenario,  we have basically "designed" the
scenario such that heat conduction from the ground is the driving force. It is expected that
for such a situation,  the heat flux is inversely  proportioned to the square-root of the
elapsed time ( Q a 1 /1'/2 where Q is the heat conduction flux and t is the elapsed time).

The Consequence  Analysis module of the program does not presently allow the user to
specify  a time-varying source term for dispersion. In  such  a case, one can follow  two
different approaches :

       OChoose the maximum release rate that occurred over the total evaporation period
       as the source term for dispersion.  This assumption would lead to  the "worst-case"
       results for the simulation. In our case, this would result in choosing 7 Ib./sec as the
       source term for dispersion.

       ii)Choose an average release rate  that occurred over the total evaporation period
       as the source term for dispersion.  In our case, it is observed that within the first 1
                                       F-23

-------
       hour, /\/4420 Ib. are evaporated leading to an average evaporation rate of 1.23
       Ib./sec for the first hour.

For further simulations, we will choose the latter of the two options (average evaporation
rate) and let 1.23 Ib./sec for 1 hour be the source term for dispersion.

Note: The scenario definition did not specify the total evaporation time associated with
the pool. Often, when an accidental release  of chemical occurs that results in a pool,
mitigation measures (such as application of a foam) may be applied to diminish or stop
the vapor emissions. For this particular scenario, we have assumed that within one hour,
mitigation measures would be effective to stop all vapor emissions from the pool.
          Fig. 11:  Evaporation rate for the refrigerated chlorine spill
Subsequently, we define the inputs which are used for the atmospheric dispersion of the
chlorine vapors from the pool. The inputs are displayed in Fig. 12.
       MODIFY PflHflnETER UflLUES
      MENU  13O :  SCENARIOS
      NEXT MENU :     QO
      ITEtt     DESCRIPTION
               Single-phase model
               ChemicaI ID  number r    :
               Duration
               Release rate type
               Phase      ••-..-.•-.... : -.-.••.••":".'••• ::,'
               Total chemical mass rate-
               Re lease temperature
       SELECTION 13 OF Zl:.=
SELECTION  :  SUPERS
FIELD CODE :     19
  VALUE              UNITS
         ,000000
        6.QQQ0QQ
     3609.000000
        1.000000
        2.000000
        1.230000
      -38.000080
: ib/secs
 degF
  Fig.  12: Release scenario definition for atmospheric dispersion of vapors
                       for the refrigerated chlorine spill
                                       F-24

-------
An explanation of these variables is as follows:
    Input

      1
                              Comment
Type of release, i.e. single-phase or two-phase
For our scenario, we select single-phase.

Chemical Identification number.
Chlorine is identified as number 6 in the program's chemical library.

Duration of the release.
For our simulation, this is set to be 3600 seconds.

Release rate type, i.e.-l=instantaneous, 2=continuous.
For our scenario we choose continuous. Note, though the initial release was
a instantaneous release (of liquid), the actual source term for dispersion is a
continuous vapor release (due to pool evaporation).

Initial phase of the release (l=liquid, 2=gas or 3=two-phase).
For  our  scenario  we choose gas  (to  represent  the vapors from  the
evaporating pool).

Total chemical mass rate.
For our scenario, this was determined to be 1.23 Ib./sec.

Initial release temperature.
For our scenario we choose -30 "F.
RESULTS FROM THE DISPERSION OF THE REFRIGERATED CHLORINE SPILL
The results of the dispersion calculations are illustrated in Fig. 13.
1.00C + 14
1.00C*13
i.ooe-t-tt
LOW 1.11
LOW I- 10
1 OOC+0*
1.00E*07
1.00C-MM
1.00C+0*
l.OOf+M
i.oac*o3
1. ODE +02
1.001+01
1.006*00

^^^






••••MB





••m






•••





^






vg





Nx






= -i







"''^
NJ




"""X
'"^



UtM&m
III


v,
A


•fM-
*•-,
"+,


•ml



k,






"

KM
>•
V


>M
«



•M




S(





•>•



to






''s.
^S

^.^^

'"N^.









SI


"«J

^J














Tort
XI


^

•»•


s

1*

^








• LOM
*.,

^ t

•I,
k"«

100







1 IPTM-?*

''v
"^
"^>J

"l
\
1000







am


"\
Y*i

««*











s
"S












k,
5

r'«













N
^ I
*s
'i.^
























































10000 1















30000
Fig. 13 : Centerline values of the toxicological criteria as a function of
downwind distance for the refrigerated chlorine spill
                                         F-25

-------
11.0  SCENARIO 4 - ACETONE SPILL

Scenario description
A large acetone spill from a tank very rapidly fills up a diked area. The initial temperature
of the acetone is 278 °K. The ground temperature is 278 °K. It is assumed to that the spill
occurs  on a concrete diked area of 100 m  .
This release corresponds to the scenario discussed in Section 5.1 of the main document.

Chemical and physical properties
The standard library of chemicals provided with the program had the predefined properties
of acetone. These are outlined in Table 4.
1   Molecular weight (gm/mole)                                        44.05
2   Critical pressure (atm)                                              71.00
3   Critical volume (cmA3/mole)                                       140.00
4   Critical temperature (°K)                                          469.00
5   Normal boiling point (°K)                                          283.50
6   Surface tension of liquid (dyne/cmA2)                                 26.84
7   Liquid viscosity at NBP (cP)                                         0.28
8   Liquid density at NBP (gm/cmA3)                                     0.88
9   Liquid density at 5 °K below NBP                                     0.89
10  Temp when vap press is 400 mmHg (°K)                             267.90
11  Enthalpy of satd. liquid (cal/mole) at
    NBP                                                              0.00
    10 °K above NBP                                                 189.00
    20 °K above NBP                                                 403.00
12  Cp coefficients (cal/gmol-°K) of gas phase
    where Cp=A+B*T+C*TA2+D*TA3                   A=-1.796   B=0.053
                                                     C=-0.300e-4  D=0.00
           Table 4 :  Chemical and physical properties of acetone
TOXICOLOGICAL CRITERIA
Since acetone is comparatively less hazardous than chlorine, it was determined that for the
simulation it would be sufficient to simply examine the concentration levels. ERPG limits
were not available for acetone (as of July, 1992). It was decided to use the IDLH level as
the "concentration limit high". The lexicological criteria are defined in Table 5.
                                     F-26

-------
    CONCENTRATION                                         Status: ON
    Comment                                             Limit high is EDLH
    Concentration limit low      [PPM]                                 10.00
    Concentration limit medium  [PPM]                                100.00
    Concentration limit high     [PPM]                              20000.00

    DOSE                                                     Status: OFF
    Comment                                                         	
    Dose limit low      [PPM-MIN]                                    0.00
    Dose limit medium  [PPM-MIN]                                    0.00
    Dose limit high      [PPM-MIN]                                    0.00

    TOXIC LOAD                                              Status: OFF
    Comment                                                         	
    Toxic load exponent (n)                          .                 0.00
    Toxic load low      [PPMAn-MIN]                                  0.00
    Toxic load medium  [PPMAn-MIN]                                  0.00
    Toxic load high      [PPMAn-MIN]                                  0.00

    PROBITS                                                  Status: OFF
    Comment                                                         	
    Probit exponent  (n)                                                0.00
    Probit constant  (a)                                                0.00
    Probit constant  (b)                                                0.00
    Lethal exposure potential limit low    (%)                            0.00
    Lethal exposure potential limit medium (%)                            0.00
    Lethal exposure potential limit high    (%)                            0.00
                 Table 5 : Toxicological criteria for acetone
Note: For this particular scenario, we have turned the status "OFF" for the dose, toxic
load and probits. Thus the program only produces concentration isopleths and does not
calculate the impact with respect to the other toxicological criteria.

RELEASE RATE ESTIMATION
Similar to scenario 3, the precise containment configuration is not defined. Hence we will
assume a large release (10,000 Ib.) that will very rapidly fill up a diked area of 100 m  .
Additionally, for this scenario we will assume the initial chemical temperature to be the
ambient temperature (41 °F).
                                    F-27

-------
SOURCE TERM FOR DISPERSION
The release scenario is defined  in Fig. 14. The logic for defining the different inputs is
similar to that used for defining the inputs for the refrigerated chlorine spill (see section
10.0).
                                          U File name :   SuTEJH
                                                flCETOME LIQ SPILL
                                                ftCETQHE
                                          an)'" :- ••'-'  2  -  "•'- '	'• -:
      KEL£f)3£ SCLNfllilU SETUP           U  File name  :
 2)   Scenario description :•••-'•                 flCETOME  LI
 31   Chemical reference number :     122       ftCETOME
 5J   Release type  (l=cont, 2=inst, 3=tran) :-••'-/.  2
 fr)   Phase of cliemicn 1  (1=1 iqi 2=gas)     !'•••-.•   1
 7)   Release size   Clbor Ik/sec)          :       1000Q
 0)   Duration of release  Csec)            :           Q
 3)   Temperature of chemical   (dcgF)       :           41
10)   Elevation of  release  (ft)
11)   Uertical velocity  (ft/sec)
12)   Horizontal velocity (ft/src)
L3)   Initial raaius (ft)
H)   rtir/CJiemic«l  mole: ratio tinitial dilution)
15)   flaxiraum pool  area Cft»«2)
l&)   tliniraum pool  depth (ft>
17)   ftlheda {reflectivity] of pool (0.0-1.0)
18)   fterosoI formation       
-------
- 1-lUDlFV  PftHftMETER UftLUES
 MENU  130 :  SCENARIOS
 NEXT MENU :     OO
 ITEM     DESCRIPTION
    1     Single-phase model
         Chemical  ID  number    :
         Duration          : :\
         Release rate type
     • '.  ;...• Phase\' '•/'.'  •'.-  •.  • :>;;V: '';;'•
         Total chemical mass rate
    7    Release temperature ;:;
                                                    — SELECTION 18 OF 21 -
                                               SELECTIOM  :  SUPER*
                                               FiELD CODE;:;.:;-•;.;::••:: ia•,'-:::-^':-^
                                                 UALUE              UMITS
                                                        .GOQQGQ
                                                     1ZZ.GQQ000
                                                    36Q9.QOOOQO     sec
                                                       1.000000
                                                       2,000000
                                                        .26100Q     lb/-sec
                                                      41.GOQOGO     degE
  Fig. 16: Release scenario definition for atmospheric dispersion of vapors
                              from acetone spill
RESULTS FROM THE DISPERSION OF THE ACETONE SPILL
The results of the dispersion calculation are illustrated in Fig. 17. It is observed that the
IDLH concentration of 20,000 PPM is attained
   i.ooE+oa
   1.00t+05
   i.ooe+04
   1.00E+03
   1.00E + 02
   1.00E+01
   1.0OE+00
                                          CanCMlVMian (PPM)
                                         100
                                                         1000
 Fig.  17: Centerline concentration as a function of downwind distance for
                               the acetone spill
                                      F-29

-------
REFERENCES

[1] Reid, R. C., Prausnitz, J. M. and Poling, R.,The Properties of Gases and Liquids, 4th
edition, McGraw-Hill, New York, New York, 1987.

[2] Daubert, T. E., and Danner, R. P., Data compilation tables for properties of pure
compounds, Design Institute for Physical Property Data, American Institute of Chemical
Engineers, New York, New York, 1985.

[3] Guidelines for  Chemical   Process Quantitative Risk Analysis, Center for Chemical
Process Safety (CCPS) of the American Institute of Chemical Engineers, New York, New
York, 1989.

[4] ten Berge, W. F. et al, Concentration-Time Mortality Response relationship of irritant
and systematically acting vapors and gases, J. Haz. Matl., 13, 301-309, 1986.

[5] Finney, D. J., Probit Analysis, Cambridge University Press, London, U.K. 1977.

[6] Emergency Response Planning Guidelines for Chlorine, American Industrial Hygiene
Association, Akron, Ohio, 1988.

[7] Toxicology of  substances  in relation to major hazards, Chlorine, Health and Safety
Executive, London, UK, 1990.

[8] Withers, R. M. and Lees, F. O., The assessment of major hazards:  The lethality of
Chlorine. Part 2, Model of toxicity to man, J. Haz. Matl., 12, 283-302, 1985.

[9] Eisenberg et al,  Vulnerability model, A simulation for assessing damage resulting from
marine spills, NTIS report AD-A015-245, 1975.

[10] Perry, W. W.  and Articola, W. P., A study to modify the vulnerability model of the
risk management system, NTIS report AD-A084-214, 1980.

[11]  COVO steering committee, Risk Analysis of six potentially hazardous  industrial
objects in the Rijnmond area, A pilot study, Riedel,  Dordrecht, 1982.

[12] Marshall V. C., The prediction of human mortality from chemical accidents with
special reference to the lethal toxicity of Chlorine, J. Haz. Matl., 22, 13-56, 1989

[13] Description of modeling techniques for hazardous chemical releases, DuPont SAFER
Emergency Systems, Westlake Village, California, March, 1989.

[14] Belore, R., et al., A computer model for predicting chemical leak rates from damaged
storage tanks, Environment Canada, 1986.
                                     F-30

-------
[15]Delhaye, J. M., Instrumentation in Two-Phase Flow and Heat Transfer in the Power
and Process Industries, Hemisphere Publishing, Washington, 1981.

[16]Bird, R. B. et al, Transport Phenomena, John Wiley, New York, New York, 1981.

[17]Fleisher, M. T., Mitigation of chemical spills, Shell Development Company, Houston,
Texas, 1980.

[18]Mackay, D. and Matsugu, R. S., Evaporation rates of liquid hydrocarbons on land and
water, Canadian J.Chem.Eng., Vol. 51, 434-439, 1980.

[19]Shaw, P. and Briscoe, F., Evaporation rates from spills of hazardous liquids on land
and water, UKAEA-SRD-R100, U. K.,1978.

[20]Treybal, R. E., Mass Transfer Operations, McGraw-Hill, New York, New York,
1968.

[21JTRACE Users Manual,  DuPont SAFER Emergency Systems,  Westlake Village,
California, 1987.

[22]Chikhliwala, E. D., and Hague, W. J., Specialized techniques for modeling the unique
phenomena exhibited  in  HF  releases, Proceedings of the International  Conference  on
Vapor Cloud Modeling sponsored by CCPS - AIChE, Boston, 955-973, 1987.

[23]Haddock, S. R. and Williams, R J., J. Chem. Tech. Biotech., Vol. 29, 655, 1979.

[24]Zapert, J. G. et al, Evaluation of Dense Gas Simulation Models, EPA-450/4-90-018,
Research Triangle Park, North Carolina,  1991.
                                     F-31

-------
APPENDIX G
 HGSYSTEM

-------
                          TABLE OF CONTENTS

1.0 INTRODUCTION 	G-3

2.0 FLASHING LIQUID RELEASE OF HF  	G-6
     2.1 Input for HGSYSTEM  .	G-6
           2.1.1 Input for the HFPLUME Module	G-6
           2.1.2 Input for the HEGADAS-T Module	G-8
     2.2 Example of Output	G-14

3.0 CHLORINE VAPOR RELEASE	G-22
     3.1 Input Description 	:	G-22
           3.1.1 Input for the PLUME Module	 G-22
           3.1.2 Input for the PGPLUME Module	 G-24
     3.2 Example of Output	G-29

4.0 REFERENCES	,	G-33
                             LIST OF TABLES

1.  Modeling Capabilities of HGSYSTEM  	G-5
2.  HGSYSTEM Input to HFPLUME - Flashing Liquid Release of HF  	G-ll
3.  HGSYSTEM Input to HEGADAS-T - Flashing Liquid Release of HF	G-12
4.  HEGADAS-T Output - Flashing Liquid Release of HF	 G-15
5.  HGSYSTEM Input to PLUME - Vapor Jet Release of Chlorine  	G-27
6.  HGSYSTEM Input to PGPLUME - Vapor Jet Release of Chlorine	G-28
7.  PGPLUME Output - Vapor Jet Release of Chlorine	G-30
                            LIST OF FIGURES

1.  Available Model Combinations in HGSYSTEM	G-4
2.- Sequence of HGSYSTEM models for a pressurized release of HF	G-10
3.  Sequence of HGSYSTEM models for a pressurized release of non-reactive gas.  G-26
                                   G-2

-------
1.0  INTRODUCTION

The computer model HGSYSTEM and its associated manuals(I"3) were generated by the
Industry Cooperative HF Mitigation/Assessment program. This ad-hoc industry program
was started in 1987 in order to study and test techniques for mitigating accidental releases
of hydrogrn fluoride (HF) and alkylation unit acid and  to allow better estimation of the
impacts of such releases on the surrounding population.

Because of this focus, HGSYSTEM is a software package of atmospheric dispersion models
with a heavy emphasis on the  release and dispersion behavior of  hydrogen fluoride.
However,  because HGSYSTEM was  developed from  an  existing  heavy vapor  model,
HEGADAS(4), there are also several models for the release and dispersion of ideal gases.
The models in HGSYSTEM are  stand-alone computer programs which may either be run
individually,  or  alternatively  may be  linked  together  to simulate  the behavior  of  an
accidental release of material into the atmosphere.  A schematic of the available model
combinations in  HGSYSTEM is shown in Figure  1.   Note  that HGSYSTEM  cannot
currently model all of the cases considered in the main body of this report, such as flashing
liquid jets of materials other than HF. The details on HGSYSTEM installation and model
execution is documented in References 1 and 3. For a complete description of the technical
bases underlying each HGSYSTEM model,  the reader is directed  to the HGSYSTEM
Technical Reference Manual®.

The types of releases HGSYSTEM is capable of modeling are listed  in Table 1. Included
in this table are the standard input files, provided with the HGSYSTEM software package,
used to model the desired release type.  Two postulated releases  were chosen from these
capabilities for comparison with  previous DEGADIS and SLAB simulations:  1)  flashing
liquid release of HF and 2) chlorine vapor release.
                                      G-3

-------
Source: Reference 3.
        Figure 1. Available Model Combinations in HGSYSTEM
                                G-4

-------
Model
HFSPILL

EVAP





HFFLASH
HFPLUME
PLUME
HEGADAS-S



HEGADAS -T



PGPLUME

HFJET
Standard
Input files
STLIQUID.HLI
STVAPOUR.HLI
STANDST1 . EVI
STANDST2 . EVI
STANDST3 . EVI
STANDTR1 . EVI
STANDTR2 . EVI
STANDTR3 . EVI
SKELETON. HFI
SKELETON. HP I
SKELETON. PLI
STPOOLNO.HSI
STPOOLHF.HSI
STBRKNO.HSI
STBRKHF.HSI
STPOOLNO.HTI
STPOOLHF.HTI
STBRKNO.HTI
STBRKHF.HTI
STANDNO.PGI
STANDHF.PGI
SKELETON. HJ I
Type of problem being analysed
transient spillage of HF liquid
transient spillage of HF vapour
steady-state evaporation of boiling pool on water
steady-state evap. of non-boiling pool on land
steady-state evap. of non-boiling HF pool on land
transient evaporation of boiling pool on water
transient evaporation of non-boiling pool on land
transient evap. of non-boiling HF pool on land
flashing of HF
flashing, jet-flow and near-field dispersion of KF
jet- flow and near- field dispersion of ideal gas
steady HEGADAS run from pool (ideal gas)
steady HEGADAS run from pool (HF gas)
steady HEGADAS run from breakpoint (ideal gas)
steady HEGADAS run from breakpoint (HF gas)
transient HEGADAS run from pool (ideal gas)
transient HEGADAS run from pool (HF gas)
transient HEGADAS run from breakpoint (ideal gas)
transient HEGADAS run from breakpoint (HF gas)
elevated passive dispersion (ideal gas)
elevated passive dispersion (HF-gas)
flashing and jet- flow of HF
Source: Reference 3.
             Table 1. Modeling Capabilities of HGSYSTEM
                                  G-5

-------
2.0 FLASHING LIQUID RELEASE OF HF

These releases correspond to the scenarios labeled #3 in Figure 2-1 of the main report.
They are characterized by a liquid jet with subsequent flashing to vapor and aerosol.  A
complete description of flashing liquid jets is given in Section 6.0, along with the simulations
made with the DEGADIS and SLAB atmospheric dispersion modeling computer codes. In
this appendix, HGSYSTEM will be used to model this release.

The model combination used to simulate the flashing liquid release of HF in this example
is HFPLUME and HEGADAS-T (refer to Figure 1; HFSPILL was not needed since the
"spill" parameters have already been identified in Section (7.2)  of the main report). The
sequence of HGSYSTEM models used for pressurized releases of HF is depicted in Figure
2.  From this figure one can see that two HGSYSTEM models are used to describe this type
of release. HFPLUME is the initial model and the subsequent model is determined from
executing the HFPLUME module.  In this example, the transition is made from HFPLUME
to HEGADAS-T, which is a model for the dispersion of transient releases of heavy vapors.

This case corresponds to scenario #3A of Figure 2-1 in the main report.  It is assumed that
there is a vessel in which there is HF at an elevated temperature of  100 °F (37.78 °C) and
a pressure of 200 psig (14.61 atm). For the present case, it is assumed that there is a leak
of effective diameter 0.5 inches (0.0128 m).

2.1 Input for HGSYSTEM

The input to HGSYSTEM for the HF release described above is given in Tables 2 and 3.
Table 2 is the standard input file (SKELETON.HPI) for the simulation of a flashing liquid
release of HF using the HFPLUME module.  An intermediate  file is generated from the
execution of HFPLUME and is the subsequent input file for the HEGADAS-T module.
This file is used to model the transition made from HFPLUME to HEGADAS-T stages (see
Figure 2) and is shown in Table 3.

2.1.1  Input for the HFPLUME Module

The input for the HFPLUME module of HGSYSTEM is contained in Table 2.  Where
possible, these data are the same as those used for the DEGADIS run in Table (7-2).

Line 1: TRES is  the liquid HF storage temperature  (C).

Line 2: PRES is  the liquid HF (absolute) storage pressure (atm).

Line 3: DMDTHF is the release point mass flow rate  (kg/s).

Line 4: DEXTT is the rupture diameter of the vessel or pipe break  (m).
                                      G-6

-------
Line 5:  ZEXIT is the release height above level ground (m).

Line 6:   PHISTK is the  release angle to  the  horizontal, taken to be 90° because the
corresponding DEGADIS run can only handle a vertical release.

Line 7:  DURATION is the duration of the release (s).

Line 8:  ZO is the height at which the windspeed is measured (m).

Line 9:  UO is the ambient wind-speed (m/s). See the note at the end of this list of input
parameters.

Line 10:  AIRTEMP is the atmospheric temperature (C).

Line 11:  AIRPRESS is the ambient (absolute) pressure at the release height (atm).

Line 12:  RHPERC  is the atmospheric humidity (%).

Line 13:  ZR is the surface roughness (m).

Line 14:  PQSTAB is the Pasquill/Gifford classification. See the note at the end of this list
of input parameters.

Line 15:  XLST is a termination criterion.  The run of HFPLUME is terminated at this
distance downwind.

Lines 16-19: These lines form the far-field transition criteria MATCH datablock.  RULST
is  the transitional excess  velocity ratio; RELST is  the  last  required entrainment ratio;
RGLST is the transitional buoyancy  effect; and RNLST is the entrainment ratio for passive
advection.  The values given in lines 16 - 19 are all default values provided in the Users'
Manual.

Note: in the course of developing this example, the authors attempted to run a case with
a  low windspeed (1.5 m/s)  and atmospheric stability category F  and  obtained several
warning  messages.   It was  concluded, after a number of  sensitivity  studies,  that the
HFPLUME model,  as used in the present  work, runs into numerical problems with low
windspeeds or very stable weather conditions. To overcome this problem requires a detailed
knowledge of the numerical algorithms in HGSYSTEM. The resources available for this
project did not allow the authors to develop this understanding.  If the reader encounters
similar problems, he/she should contact the model developers. For the present case, the
model was run in neutral  stability (D)with a moderate windspeed (5 m/s).
                                       G-7

-------
2.12 Input for the HEGADAS-T Module

The input for the HEGADAS-T module of HGSYSTEM is contained in Table 3.  Most of
these data were  generated by the HFPLUME module.

Lines 1 and 2: These lines are the CONTROL datablock parameters.  This block controls
the generation of concentration contour-data by HEGADAS-T. ICNT is the contour control
flag and ISURF is the heat transfer flag. The values ICNT=0 and ISURF = 3 are provided
by the HFPLUME output. ISURF = 3 means that surface heat transfer is included. For
further details on ICONT, ISURF and other control parameters, consult Chapter 8 of the
User's Guide.

Lines 3-8:  These lines specify the ambient  conditions.  ZAIRTEMP is  the height of
temperature measurement (m); AIRTEMP is the air temperature (C); ZO is the height of
wind-speed measurements (m);  UO is the  ambient wind-speed (m/s); RHPERC is the
ambient air humidity (%); and TGROUND is the ground surface temperature (C).

Lines 9-12:  This data block contains the data relating the mechanisms  of dispersion
appropriate for a given passive limit formulation,  and to a specified concentration averaging
time. ZR is the ground surface roughness (m);  PQSTAB is the Pasquill/Gifford stability
class; AVTIMC is the concentration averaging time; and CROSSW is the form of sigma-y,
the crosswind standard deviation, see Section 7.3.1 in the technical reference manual for full
details.

Lines 13-18: Define the physical composition of  the released anhydrous HF.  THERMOD
is a flag implying release of anhydrous HF; CPGAS is  the HF monomer (isobaric) specific
heat (J/mol/C);  MWGAS is the HF monomer molecular weight (g/mol); TEMPGAS is the
gas temperature  (C); and HFLIQFR is the liquid  mass-fraction (calculated by HFPLUME).

Line 19:   Identified the downwind distance  form source  to the transition plane from
HFPLUME to HEGADAS-T (m).

Lines 20-22: These lines define the release history as determined by HFPLUME at the
transition point  ("breakpoint") between the near-field plume model HFPLUME and the
time-dependent heavy gas dispersion model  HEGADAS-T. TSTPOOL identifies the start
time for transition data (s); TSTEPR is the  time step between successive BRKDATA
records (s); and BRKDATA contains data at transition points ( the cloud  half-width in
meters, the equivalent HF monomer mole-concentration and the mass-flow rate through the
transition plane  in kg/s).

Line 23: TSTAR defines the output time (s) to  record data for report  presentation.

Warning: the user must be aware that the choice  of TSTAR is far from trivial. The default
values provided by HFPLUME in this particular  case sent the plume far too far downwind:


                                     G-8

-------
effective step lengths were too great to enable the user to make an accurate determination
of where the plume concentration falls below LOCs. It was therefore necessary to rerun the
model with smaller TSTAR steps.  This kind of iterative model running is a characteristic
feature of HEGADAS-T and failure to recognize that it is necessary is one of the most
common pitfalls associated with its use. The only way to become  confident that the choice
of TSTAR is sensible is to practice with the input.

Lines 24-  27:  Comprise  the data necessary to control the output  listing  generated by
HEGADAS-T.  XSTEP is  the arithmetic progression step-length (m);  CU is  the inner
contour concentration (kg/m3); CL is the  outer contour concentration (kg/m3); the values
given in the current example are the ERPG-3 (inner) and ERPG-2  (outer) for chlorine,
adjusted for a 20  minute exposure time  using Haber's law. CAMIN is  the last required
mass-concentration (kg/m3).
                                       G-9

-------
           -»• wind

              HFPLUME
            (urborn* plume)
    PGPLUUE
(puatre. =i«T»ted plume)
                        (round
              •wind
                                 HFPLUME
                        HEGADAS
                        urborae
               ~FSr'LL
              H." source
              sc-.l rates
HFPLUME
Jet release.
elevated plume
dispersion.
dense plume
touchdown
and slumping


HFFLASH
HF flash
conditions
on release
	 •*•
\

PGPLUME
Passive far- field
dispersion of
elevated plume

HEGAOAS-S
HEGAOAS-T
Steady or transient
ground— level
dispersion of
dense/trace gas
      Source: Reference 1.

Figure 2.  Sequence of HGSYSTEM models for a pressurized release of HF.  A transition
is made from HFPLUME to either PGPLUME (elevated dispersion) or HEGADAS (ground-
level dispersion).  If HFPLUME is not run, HFFLASH is used to set the flash data needed
by HEGADAS.
                                      G-10

-------
       Table 2.  HGSYSTEM Input to HFPLUME -  Flashing Liquid Release of HF
                HFPLUME standard input file SKELETON.HPI
TITLE   Flashing HF Release
RESERVOIR

   TRES
   PRES
PIPE
   DMDTHF
   DEXIT
   ZEXIT
   PHISTK
   DURATION
37.0
14.61
3.7
0.0128
5
90.00
1200
AMBIENT CONDITIONS
   ZO
   UO
   AIRTEMP
   AIRPRESS
   RHPERC

DISP

   ZR
   PQSTAB

TERMINAT

   XLST

MATCH

   RULST
   RELST
   RGLST
   RNLST
10.0
5
15
1.00
50.
0.1
D
1000.
 .1
 .3
 .3
 .1
                          * RESERVOIR FLUID THERMODYNAMIC STATE
DEG. CELSIUS
ATMOSPHERES
                ABSOLUTE TEMPERATURE
                ABSOLUTE PRESSURE
* PIPE EXIT-PLANE  (CHOKE-FRONT) CONDITIONS
*
* KG/S   "       DISCHARGE RATE
* M             EFFECTIVE ORIFICE DIAMETER
* M             HEIGHT ABOVE (LEVEL) GROUND
* DEGREES       RELEASE DISCHARGE ANGLE
* S             RELEASE DURATION (<0 FOR STEADY)
*
* ATMOSPHERIC AMBIENT CONDITIONS
*
* M             REFERENCE HEIGHT
* M/S           WIND VELOCITY AT HEIGHT ZO
* CELSIUS       AIR TEMPERATURE
* ATMOSPHERES   AMBIENT PRESSURE
* PERCENT       RELATIVE HUMIDITY
              DISPERSION DATA
              M
                 SURFACE ROUGHNESS PARAMETER
                 PASQUILL STABILITY CLASS
  JET/PLUME DEVELOPMENT TERMINATION CRITERIA

  M              LAST REQD. DOWNWIND DISP.

  MATCHING CRITERIA FOR HEGADAS/PGPLUME

                 LAST REQD. ABS. VALUE OF UJET/UAMB-1
                 LAST REQD. JET/(JET+HEG) ENTRAINM.
                 MAX. BUOYANCY EFFECT FOR ADVECTION
                 MAX. BUOY. EFF. FOR PASS.  DISPERSION
                                                             9
                                                             10
                                                             11
                                                             12
                                                             13
                                                             14
                                                                           15
                                                             16
                                                             17
                                                             18
                                                             19
                                                  G-ll

-------
     Table 3.  HGSYSTEM Input to  HEGADAS-T - Flashing Liquid Release of HF
TITLE Flashing HF Releasa
         Input file  for the  (transient) heavy gas
          advection  model HEGADAS-T. The file is
       generated by  the near field dispersion model
       HFPLUME.  It incorporates all the breakpoint
        data generated by FLUME together with such
         additional  variables and flags needed to
      ensure physical consistency. In addition, the
       file contains variables needed to complete a
       viable input  file suitable for submission to
       HEGADAS-T; Such additional data are prefixed
        by an asterisk (*) and should be physically
      and contextually sensible, but may be changed
    at the user's discretion. Such data may also be
     overwritten by  the addition of keywords to the
      HEGADAS-T partial input file under HGSYSTEM.
 CONTROL * HEGADAS  Control Flags Datablock.
   ICNT-
   ISURF-
               * flag controlling contour generation (-).
               * flag indicating plume/ground heat  transfer  (-).
 AMBIENT * Ambient  Atmosphere Datablock.
   EAIRTEMP-
   AIRTEMP-
   Z0>
   UO-
   RHPERO
   TGROUND-
  14.4
  15.0
  10.0
  5.00
  50.0
  15.1
height of temperature measurement
ambient (air)  temperature  (C).
height for wind-speed measurement
ambient wind-speed (m/s).
atmosphere relative humidity  (Z).
ground surface temperature (C).
(m).
(m).
 DISP * Pasquill/Gifford  Dispersion Data.
   ZR-      0.100
   PQSTAB-    D
   AVTIMC-  1200.
   CROSSW-    2, ,
                * surface roughness height (m).
                * Pasquill/Gifford stability class.
                * concentration averaging time (s).
                * type of  formula (-).
 GASDATA * Released HF-gas  Datablock.
   THERMOD-
   CPGAS-
   MWGAS.
   WATGAS-
   TEMPGAS-
  2
 29.1
 20.0
O.OOOE-01
 19.5
   HFLIQFR- 0.862
thermodynamie model flag  (-).
specific heat HF monomer  J/mol/C.
molecular weight HF monomer  (g/mol).
fraction water "picked up" from  ground  (-).
plum* temperature "immediately post flash" (C).
liquid mass-fraction "post flash"  (-).
 TRANSIT * Location of Transition Datablock.
   DISTS-
             119.
                            *  downwind distance from release point (m).
                                                            9
                                                            10
                                                            11
                                                            12
                          13
                          14
                          15
                          16
                          17
                          18
                                                                                         19
                                                 G-12

-------
Table 3. HGSYSTEM Input to HEGADAS-T - Flashing Liquid Release of HF (continued)
  TIMEDATA 0,2 * Transient breakpoint Datablock.
        NOTE: The TIMEDATA datablock comprises in
        addition to the TSTPOOL and  TSTEPR keywords
        a sequence of BRKDATA records containing
       information regarding the transition state at
       times TSTPOOL+TSTEPR, TSTPOOL*2*TSTEFR	
        The source DURATION is here  1.200E+03s; the
       time required for flow establishment at the
      transition plane is correspondingly   22.7    s.
          The data for each BRKDATA record are (in
       order of occurrence), the cloud half-width (m),
       mole-concentration of (equivalent) HF-monomer,
         and the mass-flow of anhydrous HF (kg/s).
TSTPOOL= -1.27
TSTEPR- 48.0
BRKDATA= 12 . 1
BRKDATA" 12 . 1
BRKDATA- 12 . 1
BRKDATA= 12 . 1
BRKDATA= 12 . 1
BRKDATA- 12 . 1
BRKDATA= 12 . 1
BRKDATA- 12 . 1
BRKDATA= 12 . 1
BRKDATA= 12 . 1
BRKDATA- 12 . 1
BRKDATA- 12 . 1
BRKDATA= 12 . 1
BRKDATA- 12 . 1
BRKDATA- 12.1
BRKDATA- 12.1
BRKDATA- 12 . 1
BRKDATA- 12.1
BRKDATA- 12 . 1
BRKDATA- 12 . 1
BRKDATA- 12.1
BRKDATA- 12 . 1
BRKDATA- 12 . 1
BRKDATA- 12.1
BRKDATA- 12 . 1
CALC * Output times
TSTAR= 200 .
TSTAR= 500
TSTAR= 800
TSTAR= 1100
TSTAR= 1400
TSTAR= 1700
TSTAR= 2000
* start time (s; zero breakpoint data).
* time step between BRKDATA records (s).
1.737E-03 3.70 * breakpoint data at 46.7 s.
1.737E-03 3.70 * breakpoint data at 94.7 s.
1.737E-03 3.70 * breakpoint data at 143. s.
1.737E-03 3.70 * breakpoint data at 191. s.
1.737E-03 3.70 * breakpoint data at 239. s.
1.737E-03 3.70 * breakpoint data at 287. s.
1.737E-03 3.70 * breakpoint data at 335. s.
1.737E-03 3.70 * breakpoint data at 383. s.
1.737E-03 3.70 * breakpoint data at 431. s.
1.737E-03 3.70 * breakpoint data at 479. s.
1.737E-03 3.70 * breakpoint data at 527. s.
1.737E-03 3.70 * breakpoint data at 575. s.
1.737E-03 3.70 * breakpoint data at 623. s.
1.737E-03 3.70 * breakpoint data at 671. s.
1.737E-03 3.70 * breakpoint data at 719. s.
1.737E-03 3.70 * breakpoint data at 767. s.
1.737E-03 3.70 * breakpoint data at 815. s.
1.737E-03 3.70 * breakpoint data at 863. s.
1.737E-03 3.70 * breakpoint data at 911. s.
1.737E-03 3.70 * breakpoint data at 959. s.
1.737E-03' 3.70 * breakpoint data at 1.007E+03S.
1.737E-03 3.70 * breakpoint data at 1.055E+03S.
1.737E-03 3.70 * breakpoint data at 1 . 103E+03S .
1.737E-03 3.70 * breakpoint data at 1.151E+03s.
1.737E-03 3.70 * breakpoint data at 1.199E+03s.
Datablock.
* geometric sequence of output times (s)
* geometric sequence of output times (s).
* geometric sequence of output times (s).
* geometric sequence of output times (s).
* geometric sequence of output times (s).
* 'geometric sequence of output times (s).
* geometric sequence of output times (s).
20
21
22

























23






GLOUD * Output control datablock.
XSTEP= 50.0
CU- 4.17E-05
CL» 1.S7E-05
CAMIN= l.QE-05
* (fixed) output step-length (m) .
* inner contour concentration (kg/m3).
* outer contour concentration (kg/m3).
* last reouired mass concentration HF (kg/m3)
24
25
26
27
                                                G-13

-------
2.2  Example of Output

The final output generated by the HGSYSTEM modeling effort describing the flashing
liquid release of HF is shown in Table 4. Specifically, this file was obtained from running
the HEGADAS-T module.  As can be seen, the first page reproduces the input data from
Table 3 together with some additional default values  of control  parameters (such  as
BLMODEL and  ICSCOR).   The  second  page reproduces information  on the time
dependence of various parameters at the "breakpoint" (i.e. the point at which the transition
from HFPLUME to HEGADAS-T occurs). The third page of Table 4 begins with reference
to "observers." The observers travel with small portions of the release and begin their travel
at various intervals during the times specified on page 2 of Table 4.  They observe the
dispersion of the small portions of the release; the final results are generated by summing
the concentrations seen by each observer. HEGADAS-T chooses the number of observers
to ensure adequate numerical accuracy.  The observer concept is used to track transient
releases.  The reader who is  interested in more details should consult the  HGSYSTEM
documentation.

The remainder of the  HGSYSTEM output consists of "snapshots" generated at each of the
times generated in the TSTAR array from the input file. At each TSTAR time, the  output
consists of a table giving:

      The distance x downwind in meters
      CONC, the concentration of HF monomer in vol% on the plume axis at x
      SZ, the vertical dispersion coefficient at x (m)
      SY, the crosswind dispersion coefficient at x (m)
      MIDP, the half-width b of the flat portion of the concentration profile at x (m)
      YCU, YCL, the crosswind distances to the isoconcentration contours for the upper
      and lower levels of concern (m)
      ZCU, ZCL, the vertical distances to the isoconcentration contours for the upper and
      lower levels of concern (m)
      CA, the ground level concentration on the axis at x (kg/m3)

The  (YCU,ZCU) and (YCL,ZCL) outputs allow the  user to generate snapshots  of the
isoconcentration contours for transient releases at different values of TSTAR.  The TSTAR
array must include at least one time for which all values of YCU, ZCU, YCL and ZCL are
zero - the user then knows that the plume has travelled far enough downwind for it to have
diluted below the levels of concern. In the current example, TSTAR  = 1,400 s was the last
time at which the "snapshot" shows non-zero  values of these  parameters. For TSTAR  =
1,700 and TSTAR = 2,000 (not shown on table 4) the values of ZCU, YCU, ZCL and YCL
were  all  zero.   If the last TSTAR in the input array still has non-zero values of these
variables, the model should be rerun with higher values of TSTAR. More closely  spaced
values of TSTAR can also be used to obtain any desired accuracy for the maximum
downwind distance at which levels of concern might be seen.
                                      G-14

-------
           Table 4.  HEGADAS-T Output  - Flashing Liquid Release of HF
HTMAIN
DATE 17/11/92
                 HEGADAS-T PROGRAM C  VERSION  NOV90     )
                          STANDARD REPORT  FILE
                                                                             PAGE     0
                                                                              TIME 09-32
                                  ««   Flashing HF Release
                                 HEGADAS-T  INPUT DATA
OUTPUT CODE                    ICNT
SURFACE-TRANSFER CODE         ISURF
GAS-BLANKET FORMULATION     BLMODEL
CLOUD-SHAPE CORRECTION CODE   ICSCOR
AIR TEMP.AT HEIGHT ZAIRTEMP AIRTEMP  =    15.000
REF. HEIGHT FOR AIR TEMP.   ZAIRTEMP  =    14.400
RELATIVE HUMIDITY            RHPERC  =    50.000
WIND VELOCITY AT HEIGHT ZO        UO  -    5.0000
REFERENCE HEIGHT FOR WIND VEL.    ZO  =    10.000
EARTH-S SURFACE TEMPERATURE TGROUND  -    15.100
SURFACE ROUGHNESS PARAMETER
PASQUILL STABILITY CLASS
AVERAG. TIME FOR CONC.MEAS.
MONIN - OBUKHOV LENGTH
TYPE OF FORMULA FOR SIGMA_Y
        with parameters:

CONST. IN GRAV.  SPREADING LAW
CONST. IN GRAV.  SPREADING LAW
TYPE OF FORMULA FOR SIGMA_X
        with parameters:
    ZR
PQSTAB
AVTIMC
 OBUKL
 MODSY
 DELTA
  BETA
    CE
    CD
 MODSX
 ASIGX
 BSIGX
0.10000
    D
 1200.0
l.OOOOOE+05
    2
9.18959E-02
l.OOOOOE-04
 1.1500
 5.0000
    3
 10.000
0.10000
	> CONTROL data block: control parameters
 (no output of cumulative cloud data)
 (only heat transfer, no water vapor)
 (new, non-oscillatory formulation)
 (correction  included)
	> AMBIENT data block: ambient data
 CELSIUS
 M
 Z
 M/S
 M
 CELSIUS
	> DISP data block: dispersion data
 M

 SECONDS
 M
 (Briggs  formula)

 M**(-l)
 (Chatwin/WilsonX
TEMPERATURE OF EMITTED GAS  TEMPGAS  -    19.500
SPECIFIC HEAT OF EMITTED GAS  CPGAS  -    29.100
MOLECULAR WEIGHT OF EM.  GAS   MWGAS  »    20.000
.?ICKED-UP WATER BY EM. GAS   WATGAS  =   O.OOOOOE-01
HEAT GROUP IN HEAT FLUX QH   HEATGR  =    24.000
THERMODYNAMIC MODEL         THERMOD.=       2
::NITIAL LIQUID IN HF        HFLIQFR  =   0.35200
TIME-DEPENDENT RECORD DATA:   ITYPBR =       2
- skip increment for reading  INCRT =       0
- start time for data       TSTPOOL =   -1.2700
- time step between read records  DT =    48.000
- number of records read       NTYD =      25
OUTPUT STEP LENGTH            XSTEP  -    50.000
CA AT WHICH CALC.  IS STOPPED   CAMIN  =   l.OOOOOE-05
UPPER CONCENTRATION LIMIT        CU  =   4.17000E-05
LOWER CONCENTRATION LIMIT        CL  =   1.67000E-05
PSTS OF TIME FOR CLOUD CALC. TSTAR = 200.00
500.00
800.00
1100.0
1400.0
1700. 0"
2000.0
SEC
SEC
SEC
SEC
SEC
SEC
SEC
 rULL BREAKPOINT AT            DISTS  =    119.00

 HND PROFILE EXPONENT         ALPHA  =   0.28820
 •RICTION VELOCITY             USTAR  *   0.44413
 JR TEMP  AT GROUND LEVEL	TAP  =    15.000
                      	>  GASDATA data block: gas data
                      CELSIUS
                      J/MOLE/CELSIUS
                      KG/KMOLE
                      (MOLAR FRACTION)

                      (hydrogen fluoride)
                      (MASS FRACTION)
                     	>TIMEDATA data block: source/breakpoint data
                      (breakpoint data:  B_aff, CONC, GSFLOW)

                      S
                      S
                      (see list of breakpoint  data below for  record data read)
                     	>  CLOUD data block:  control of cloud output
                      M
                      KG/M3
                      KG/M3
                      KG/M3
                                        >  CALC data block, control of output times







                                        >TRANSIT data block(s): breakpoint data
                      M
                      M/S
                      CELSIUS
                                                              G-15

-------
               Table 4. HEGADAS-T Output - Flashing Liquid Release of HF (continued)
HTMAIN
DATE 17/11/92
HEGADAS-T PROGRAM ( VERSION NOV90 )
STANDARD REPORT FILE
PAGE i
TIME 09:32
«« Flashing HF Release »»

TIME-DEPENDENT
TIME -
TSTPOOL
(S)
7.70
69.3
131.
193.
254.
316.
377.
439.
501.
562.
624.
686.
747.
309.
871.
932.
994.
1.055E+03
1.117E+03
1.179E+03
1.240E+03
CONC CLOUD
(Z VOL.) HALF -WIDTH

2.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0,
0,
0,
0,
0
0
0,
2

788E-02
174
174
174
174
174
174
,174
,174
,174
,174
,174
.174
.174
,174
.174
.174
.174
.174
.174
. 788E-02
(M)
1.94
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
12.1
1.94
CLOUD
HEIGHT
(M)
82.2
19.8
19.8
19.8
19.8
19.8
19.8
19.8
19.8
19.8
19.8
19.8
19.3
19.8
19.8
19.8
19.8
19.8
19.8
19.8
32.2
DATA AT FULL
GAS
FLOW
BREAKPOINT X - 119.
SZ
(M)
RIB
M

TMP
(C)

CA
(KG/M3)
(KG/S)
0.594
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
0.
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
.70
594
88.9
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
21.4
88.9
2.96
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.44
2.97
14.
14.
14.
14.
14,
14,
14.
14
14.
14,
14
14.
14
14
14
14
14
14
14
14
14
,7
3
,3
,3
.3
.3
,3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.3
.7
2.362E-04
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
1.474E-03
2.362E-04
Observer-dispersion data set at breakpoint; total CPU
                                              15 seconds
                                                  G-16

-------
                  Table 4.  HEGADAS-T Output - Flashing Liquid  Release of HF (continued)
HTMAIN
DATE 17/11/92

HEGADAS-T PROGRAM ( VERSION NOV90 )
STANDARD REPORT FILE
«« Flashing HF Release »»
PAGE
TIME 09'

•32

Observer-release frequency  • 32; maximum value  over all times of [mean error in observer  concentration]/[peak  cone.] = 0.516
Observer-release frequency  = 16; maximum value  over all times of [mean error in observer  concentration]/[peak  cone.] = 0.152
Observer-release frequency  =  8; maximum value  over all times of [mean error in observer  concentration]/[peak  cone.] = 0.139
Observer-release frequency  =  4; maximum value  over all times of [mean error in observer  concentration]/[peak  cone.] = 3.426E-02
Convergence tolerance OBSEPS = 5.000E-02 is  satisfied

Cloud shape correction performed

Observer-dispersion data set for  41 observers;  total CPU =  313 seconds
DISPERSION DATA AT TIME = 200.0
DISTANCE
(M)
150.
200.
250.
300.
350.
400.
450.
500.
550.
600.
650.
700.
750.
800.
850.
900.
950.
l.QOOE-t-03
1.050E+03
1.100E+03
1.150E+03
1.200E+03
1.250E+Q3
1.300E+03
1.350E+03
1.400E+03
CONC
(%
3
0
0
0.
9.
8.
7.
6,
5.
4.
4.
3.
3,
2.
2.
2.
1.
1.
1.
9,
7.
5,
3.
2.
1.
6.
VOL.)
. 190E-02
.145
.135
.116
.896E-02
.668E-02
. 476E-02
.373E-02
,505E-02
. 855E-02
.286E-02
.763E-02
.297E-02
. 887E-02
.514E-02
. 165E-02
. 834E-02
.520E-02
.226E-02
. 557E-03
.151E-03
, 085E-03
.399E-03
. 108E-03
.198E-03
168E-04
SZ
(M)
14.
15.
13.
12.
11.
12.
12.
13.
13.
14.
15.
16.
17.
18.
19.
20.
22.
24.
27.
29.
39.
50.
62.
73.
75.
77.


9
2
9
7
8
3
7
2
8
6
4
2
2
4
6
8
7
9
1
4
5
3
0
3
7
9
SY
(M)
12.3
19.2
26.4
33.6
40.5
46.3
52.1
58.0
63.4
68.4
73.4
78.4
34.8
93.5
102.
111.
117.
122.
128.
133.
135.
137.
139.
141.
147.
153.
MIDP
(M)
6.11
6.57
7.04
7.51
7.86
7.69
7.53
7.36
6.91
6.02
5.13
4.25
3.32
2.32
1.31
0.313
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.QOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
YCU
(M)
22.9
42.0
55.1
67.2
78.1
86.2
93.6
100.
105.
110.
113.
116.
120.
127.
132.
135.
134.
130.
122.
108.
82.6
24.6
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
SEC
ZCU
(M)
24.2
39.1
35.3
31.1
27.7
27.8
27.7
27.3
27.3
27.7
28.0
28.1
28.3
28.6
28.6
28.2
28.1
27.3
25.3-
21.3
18.4
3.51
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01

YCL
(M)
26.6
46.5
61.4
75.3
88.0
97.8
107.
115.
123.
128.
134.
139.
146.
155.
164.
172.
175.
175.
172.
167.
153.
134.
103.
36.4
O.OOOE-01
O.OOOE-01

ZCL
(M)
33.0
47.0
42.7
37.9
34.1
34.5
34.6
34.6
35.0
36.0
36.8
37.5
38.3
39.5
40.4
40.9
42.3
43.4
43.3
41.9
48.1
48.7
38.7
8.92
O.OOOE-01
O.OOOE-01

CA
(KG/M3)
2.708E-04
1.232E-03
1.147E-03
9.842E-04
8.404E-04
7.359E-04
6.345E-04
5.406E-04
4.663E-04
4.116E-04
3.633E-04
3.189E-04
2.794E-04
2.446E-04
2.130E-04
1.834E-04
1.553E-04
1.287E-04
1.038E-04
8.092E-05
6.054E-05
4.306E-05
2.878E-05
1.785E-05
1.014E-05
5.222E-06
                                                           G-17

-------
Table 4. HEGADAS-T Output - Flashing Liquid Release of HF (continued)
DISPERSION DATA AT TIME - 500.0
DISTANCE
CM)
150.
200.
250.
300.
350.
400.
450.
500.
550.
600.
650.
700.
750.
800.
850.
900.
950.
l.OOOE+03
1.050E+03
1.100E+03
1.150E-KI3
1.200E+03
1.250E+03
1.300E+03
1.350E+03
1.400E+03
1.450E+03
1.500E+03
1.550E+03
1.600E+03
1.650E+03
1.700E-HJ3
1.750E+03
1.800E+03
1.850E+03
1..900E+03
1.950E-H)3
2.000E-HJ3
2.050E+03
2.100E-H33
2.150E+03
2.200E+03
2.250E+03
2.300E-1-03
2.350E+03
2.400E+03
2.450E+03
2.500E+03
2.550E+03
2.500E+03
2.650E+03
2.700E+03
CONC
(Z
0,
0,
0,
0,
9,
8,
7,
6.
5,
4
3,
3,
3.
2
2,
2,
1.
1.
1.
1.
1.
9,
8.
3.
7
6.
6,
5,
5.
4,
4
4.
3.
3.
3,
3
3
2
2
2
2
2
2
2
1
1
1.
1
1
1
1,
1
VOL.)
.155
.147
.135
.114
.875E-02
. 409E-02
.085E-02
.086E-02
.301E-02
.594E-02
. 984E-02
.490E-02
.077E-02
. 710E-02
.380E-02
.085E-02
.824E-02
.597E-02
. 404E-02
.243E-02
. 108E-02
.941E-03
.964E-03
, 124E-03
.394E-03
.759E-03
.201E-03
.709E-03
.273E-03
.885E-03
.538E-03
.226E-03
. 946E-03
.692E-03
. 463E-03
.253E-Q3
.063E-03
. 888E-03
. 728E-03
. 581E-Q3
. 446E-03
.320E-03
.204E-03
.096E-03
. 995E-03
. 900E-03
. 810E-03
.725E-03
.642E-03
.562E-03
.484E-03
. 406E-03
SZ
(M)
19.
16.
14.
12.
12.
12.
12.
13.
14.
14.
15.
16.
17.
18.
20.
21.
23.
24.
26.
28.
29.
31.
32.
34.
36.
37.
39.
40.
42.
43.
45.
47.
48.
50.
51.
53.
54.
56.
57.
59.
60.
62.
63.
65.
66.
63.
69.
71.
72.
74.
75.
77.


4
7
0
0
2
4
7
3
1
9
7
8
8
8
0
6
3
9
5
1
7
3
9
5
1
7
2
a
4
9
5
0
6
1
7
2
7
3
3
3
3
3
8
3
3
3
3
2
7
2
6
1
SY
(M)
8.37
17.6
26.9
35.7
42.6
49.6
56.5
62.9
69.0
75.2
81.3
88.4
95.6
103.
109.
115.
121.
127.
133.
139.
145.
151.
156.
162.
168.
174.
179.
185.
191.
196.
202.
207.
213.
218.
224.
229.
235.
240.
246.
251.
257.
262.
267.
273.
278.
283.
288.
294.
299.
304.
309.
314.
MIDP
(M)
9.36
8.69
8.03
7.55
7.74
7.93
8.12
7.63
6.82
6.01
5.20
3.80
2.39
0.983
O.OOOE-01
O.OOOE-01
O.OQOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
YCU
(M)
24.9
41.2
57.0
70.8
81.6
91.4
100.
107.
113.
118.
123.
128.
132.
135.
137.
139.
139.
138.
136.
134.
130.
126.
121.
115.
107.
97.7
86.0
71.1
49.7
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
-O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
SEC
ZCU
(M)
50.8
43.2
35.6
29.1
28.7
28.0
27.2
27.2
27.6
27.9
27.9
28.2
28.5
28.5
28.5
28.8
28.7
28.2 '
27.5
26.5
25.3
23.8
22.1
20.1
17.9
15.4
12.5
9.24
5.26
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01

YCL
(M)
26.9
45.3
63.4
79.4
92.1
104.
115.
124.
132.
139.
146.
154.
161.
167.
173.
177.
181.
184.
187.
189.
190.
192.
192.
193.
193.
193.
192.
191.
189.
187.
184.
181.
177.
173.
168.
182.
156.
148.
140.
130.
119.
105.
89.0
67.0
29.0
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01

ZCL
(M)
61.0
52.0
43.0
35.5
35.2
34.8
34.2
34.6
35.6
36.4
37.0
38.0
39.0
39.8
40.7
42.1
43.3
44.2
44.7
45.1
45.4
45.5
45.4
45.2
44.8
44.3
43.6
42.8
41.8
40.7
39.5
38.1
36.6
34.9
33.1
31.1
28.9
26.6
24.1
21.4
18.4
15.2
- 11.6
7.39
2.01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-Q1
O.OOOE-01

CA
(KG/M3)
1.320E-03
1.246E-03
1.150E-03
9.667E-04
8.388E-04
7.139E-04
6.011E-04
.5.162E-04
4.495E-04
3.894E-04
3.377E-04
2.957E-04
2.607E-04
2.296E-04
2.016E-04
1.766E-04
1.545E-04
1.353E-04
1.189E-04
1.053E-04
9.384E-05
8.417E-05
7.589E-05
6.878E-05
6.260E-05
5.722E-05
5.249E-05
4.833E-05
4.464E-05
4.135E-05
3.841E-05
3.578E-05
3.340E-05
3/125E-05
2.931E-05
2.754E-05
2.592E-05
2.445E-05
2.309E-05
2.185E-05
2.070E-05
1.964E-05
1.866E-05
1.774E-05
1.688E-05
1.608E-05
1.532E-05
1.460E-05
1.390E-05
1.322E-05
1.256E-05
1.190E-05
                              G-18

-------
Table 4. HEGADAS-T Output - Flashing Liquid Release of HF (continued)
DISPERSION DATA AT TIME - 800.0
DISTANCE
CM)
200.
250.
300.
350.
400.
450.
500.
550.
600.
650.
700.
750.
800.
850.
900.
950.
l.OOOE+03
1 . 050E+03
1 . 100E+03
1.150E+03
1.200E+03
1.250E+03
1.300E+03
1.350E+03
1.400E+03
1.450E+03
1.500E+03
1.550E+03
1.600E+03
1.650E+03
1.700E+03
1.750E+03
1.800E-K13
1.350E+03
1.900E+03
1.950E+03
2.000E+03
2.050E+03
2.100E+03
2.150E-H33
2.200E+03
2.250E+03
2.300E+03
2.350E+03
2.400E+03
2.450E+03
2.500E+03
2.550E+03
2.500E-H33
2.650E+03
2.700E+03
2.750E+03
2.800E+03
2.850E-HJ3
2.900E+03
2.950E+03
:i.OOOE-K)3
:i . 050E-HJ3
3 . 100E-HJ3
CONC
(2
VOL.)
2.756E-07
9.045E-02
0.114
9.727E-02
8.189E-02
7.023E-02
6.074E-02
5.204E-02
4.478E-02
3.916E-02
3.452E-02
3.036E-02
2.661E-02
2.331E-02
2.040E-02
1.787E-02
1.570E-02
1.
1.
1.
9.
8.
8.
7.
6.
6.
5.
5.
4,
4.
4.
3.
- 3.
3.
3.
3.
2.
2.
2.
2.
2.
2.
2.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
.3871-02
.232E-02
. 100E-02
874E-03
907E-03
074E-03
.352E-03
722E-03
168E-03
681E-03
247E-03
862E-03
517E-03
208E-03
929E-03
577E-03
449E-03
241E-03
052E-03
878E-03
720E-03
573E-03
438E-03
314E-03
199E-03
092E-03
993E-03
901E-03
315E-03
735E-03
660E-03
590E-03
524E-03
462E-03
404E-03
350E-03
298E-03
250E-03
204E-03
160E-03
119E-03
080E-03
SZ
(M)
10.9
12.7
12.5
12.4
12.2
12.7
13.5
14.2
15.0
15.9
16.9
17.8
19.0
20.5
22.0
23.4
25.0
26.6
28.2
29.3
31.4
33.0
34.6
36.2
37.8
39.3
40.9
42.5
44.0 .
45.6
47.1
48.7
50.2
- 51.8
53.3
54.8
56.4
57.9
59.4
BO. 9
62.4
63.9
65.4
66.9
68.4
69.3
71.3
72.8
74.3
75.7
77.2
78.7
80.1
81.6
83.0
84.5
85.9
87.4
38.8
SY
(M)
22.1
28.6
36.0
43.4
50.3
57.6
64.2
70.8
77.3
83.2
89.2
95.1
101.
108.
115.
121.
128.
134.
139.
145.
151.
157.
163.
168.
174.
180.
185.
191.
197.
202.
208.
213.
219.
224.
230..
235.
241.
246.
251.
257.
262.
268.
273.
278.
283.
289.
294.
299.
304.
309.
315.
320.
325.
330.
335.
340.
345.
350.
355.















0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0,
0
MIDP
(M)
6.12
6.68
7.25
7.81
8.37
8.11
7.50
6.88
6.20
4.91
3.63
2.34
1.36
.960
.561
.162
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
. OOOE-Q1
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
YCU
(M)
0. OOOE-01
55.5
71.1
82.8
93.5
102.
109.
116.
121.
125.
128.
131.
133.
136.
137.
138.
137.
136.
133.
130.
126.
121.
114.
107.
97.0
85.2
70.0
48.0
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
O.QOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
SEC
ZCU
(M)
0. OOOE-01
29.0
30.5
28.9
27.3
27.2
27.5
27.6
27.7
28.0
28.3
28.4
28.5
28.8
28.8
28.5
28.0
27.3
26:4
25.2
23.7
22.0
20.0
17.8
15.2
12.4
9.02
4.98
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01

YCL
(M)
0. OOOE-01
62.7
79.8
93.5
106.
117.
126.
135.
143.
149.
154.
160.
165.
171.
176.
180.
184.
186.
189.
190.
192.
193.
193.
193.
193.
192.
191.
189.
187.
184.
181.
177.
173.
168.
162.
155.
148.
139.
130.
118.
105.
88.0
66.0
28.0
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01

ZCL
(M)
0. OOOE-01
35.9
37.2
35.6
34.0
34.2
35.0
35.6
36.2
37.2
38.2
38.9
39.9
41.3
42.4
43.3
44.0
44.7
45.1
45.4
45.5
45.4
45.1
44.8
44.2
43.5
42.7
41.8
40.6
39.4
38.0
36.5
34.8
32.9
30.9
28.8
26.5
23.9
21.2
18.2
15.0
11.4
7.23
1.89
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
O.OOQE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01

CA
(KG/M3)
2.340E-09
7.685E-04
9.711E-04
8.259E-04
6.950E-04
5.959E-04
. 5.152E-04
4.413E-04
3.796E-04
3.319E-04
2.925E-04
2.572E-04
2.255E-04
1.975E-04
1.728E-04
1.513E-04
1.329E-04
1.174E-04
1.043E-04
9.315E-05
8.360E-05
7.541E-05
6.835E-05
6.224E-05
5.690E-05
5.222E-05
4.809E-05
4.442E-05
4.115E-05
3.824E-05
3.562E-05
3.326E-05
3.113E-Q5
2.919E-05
2.743E-05
2.583E-05
2.436E-05
2.302E-05
2.178E-05
2.064E-05
1.958E-05
1.861E-05
1.771E-05
1.687E-05
1.609E-05
1.536E-05
1.469E-05
1.405E-05
1.346E-05
1.290E-05
1.238E-05
1.189E-05
1.142E-05
1.099E-05
1.058E-05
1.019E-05
9.822E-Q6
9.475E-06
9.145E-06
                              G-19

-------
Table 4. HEGADAS-T Output - Flashing Liquid Release of HF (continued)
DISPERSION DATA AT TIME - 1100.
DISTAHCE
(M)
' 150.
200.
250.
300.
350.
400.
450.
500.
550.
600.
650.
700.
750.
800.
850.
900.
950.
1.000E-HJ3
1.050E+03
1.100E+03
1.150E+03
1.200E+03
1.250E+03
1.300E+03
1.350E+03
1.400E+03
1.450E+03
1.500E+03
1.550E+03
1.600E+03
1.650E+03
1.700E+03
1.750E+03
_1.800E+-03
1 . 850E+03
1.900E+03
1.950E+03
2.000E+03
2.050E+03
2.100E+03
2.150E+03
2.200E+03
2.250E+03
2.300E+03
2.350E+03
2.400E+03
2.450E+03
2.500E+03
2.550E+03
2.600E-I-03
2.650E+03
2.700E+03
2.750E+03
2.800E+03
2.850E+03
2.900E-I-03
2.950E+03
3 . OOOE+03
3.0SOE+03

(Z
1
7
0
0
9
8
7
6
5
4
3
3
2
2
2
2
1
1
1
1
1
9
8
3
7
6
6
5
5
4
4
4
3
3
3
3
3
2
COHC
VOL. )
.365E-12
.185E-02
.136
.114
.599E-02
.237E-02
.095E-02
.026E-02
.136E-02
.463E-02
.907E-02
.413E-02
.981E-02
.611E-02
.289E-02
.007E-02
.762E-02
.553E-02
.375E-02
.221E-02
.090E-02
.781E-03
.828E-03
.006E-03
.293E-03
.669E-03
. 122E-03
.639E-03
.211E-03
.830E-03
.•488E-03
. 182E-03
.906E-03
.656E-03
.429E-03
.223E-03
.035E-03
.863E-03
2.705E-03
2
.560E-03
2.427E-03
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.303E-03
. 189E-03
.083E-03
. 985E-03
.893E-03
. 808E-03
.728E-03
.654E-03
. 584E-03
.519E-03
.457E-03
.399E-03
.345E-03
.294E-03
.246E-03
.200E-03
.157E-03
. 116E-03
SZ
(M)
11.3
14.1
13.4
12.7
12.1
12.2
12.8
13.5
14.1
15.0
16.0
16.9
18.0
19.3
20.7
22.0
23.5
25.1
26.7
28.4
30.0
31.5
33.1
34.7
36.3
37.9
39.5
41.1
42.6
44.2
45.7
47.3
48.3
50.4
51.9
53.4
55.0
56.5
58.0
59.5
61.0
62.5
64.0
65.5
67.0
68.5
70.0
71.4
72.9
74.4
75.9
77.3
78.8
80.2
81.7
83.1
34.6
36.0
87.5
SY
(M)
14.1
20.6
28.3
36.1
43.8
51.0
57.9
64.8
71.6
77.9
84.1
90.3
96.7
103.
110.
116.
122.
128.
134.
140.
146.
152.
157.
163.
169.
174.
180.
186.
191.
197.
203.
208.
214.
219.
225.
230.
236.
241.
247.
252.
257.
263.
268.
273.
279.
284.
289.
294.
300.
305.
310.
315.
320.
325.
330.
335.
341.
346.
351.

















0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
MIDP
(M)
5.42
6.14
6.85
7.56
8.28
8.31
7.95
7.59
7.16
6.00
4.34
3.67
2.66
1.36
1.06
.252
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
YCU
(M)
0. OOOE-01
39.9
58.5
71.5
33.8
93.9
103.
110.
117.
122.
126.
129.
132.
135.
137.
138.
138.
137.
136.
133.
130.
125.
120.
114.
106.
96.1
34.0
68.4
45.4
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01 '
0. OOOE-01
0. OOOE-01
0. OOOE-01
o.oooE-ai
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
SEC
ZCU
(M)
0. OOOE-01
30.3
34.1
31.0
28.1
27.3
27.5
27.5
27.4
27.8
28.1
28.2
28.4
28.8
28.8
28.6
28.4
28.0
27.3
26.3
25.1
23.5
21.8
19.8
17.6
15.0
12.1
8.69
4.57
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01

TCL
(M)
0. OOOE-01
45.2
65.2
80.2
94.6
107.
118.
127.
137.
144.
150.
156.
162.
168.
173.
177.
181.
184.
187.
189.
191.
192.
193.
193.
193.
193.
192.
190.
189.
186.
184.
181.
177.
172.
167.
161.
155.
147.
139.
129.
117.
103.
86.4
63.7
21.7
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01

ZCL
(M)
0. OOOE-01
38.0
41.1
37.8
34.7
34.0
34.6
35.0
35.3
36.4
37.4
38.1
39.1
40.4
41.5
42.3
43.2
44.1
44.3
45.2
45.4
45.5
45.4
45.1
44.7
44.2
43.5
42.6
41.7
40.5
39.3
37.9
36.3
34.6
32.8
30.8
28.6
26.2
23.7
21.0
18.0
14.7
11.0
6.82
1.27
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
a. OOOE-OI
0. OOOE-01
0. OOOE-01
0. OOOE-01

CA
(KG/M3)
1.159E-14
6.101E-04
1.157E-03
9.649E-04
8.149E-04
6.992E-04
6.020E-04
5.111E-04
•'4.355E-04
3.783E-04
3.312E-04
2.892E-04
2.525E-04
2.212E-04
1.939E-04
1.700E-04
1.493E-04
1.316E-04
1.164E-04
1.034E-04
9.227E-05
8.281E-05
7.474E-05
6.778E-05
6.174E-05
5.646E-05
5.183E-05
4.774E-05
4.411E-05
4.088E-05
3.799E-05
3.540E-05
3.306E-05
3.095E-05
2.903E-05
2.728E-05
2.569E-05
2.424E-05
2.290E-05
2.167E-05
2.054E-05
1.950E-05
1.S53E-05
1.763E-05
1.680E-05
1.602E-05
1.530E-05
1.463E-05
1.400E-05
1.341E-05
1.285E-05
1.233E-05
1.184E-05
1.139E-05
1.095E-05
1.054E-05
1.016E-05
9.791E-06
9.447E-06
                              G-20

-------
              Table 4.  HEGADAS-T Output - Flashing Liquid Release of HF (concluded)
DISPERSION DATA AT TIME « 1400.
DISTANCE
(M)
l.OOOE+03
1.050E+03
1.100E+03
1.150E+03
1.200E-H)3
1.250E-HJ3
1.300E-HJ3
1.350E+03
1.400E-HJ3
1.450E-H33
1.500E+03
1.550E+03
1.600E+Q3
1.650E+03
1.700E+03
1.750E+03
1.800E+03
1.850E+03
1.900E-H)3
1.950E+03
2.000E-I-03
2.050E+03
2.100E+03
2.150E+03
2.200E+03
2.250E+03
2.300E+03
2.350E+03
2.400E+03
2.450E+03
2.500E+03
2.550E+03
2.600E+03
2.S50E+03
2.700E+03
2.750E+03
2.800E+03
2.850E-I-03
2 . 900E+03
2.950E+03
3.QOOE+03
3.Q50E+03
3.100E+03
CONC
(*
1.
1.
2.
3.
It.
5.
5.
5.
5.
5.
5.
4.
4.
4.
4.
3.
3.
3.
3.
3.
2.
2.
2.
2.
2.
2.
2.
1.
1.
1.
I.
1.
1.
	 -1
1.
1.
1.
1.
I.
1.
1.
1.
1.
VOL. }
132E-03
912E-03
842E-03
788E-03
612E-03
221E-03
582E-03
716E-03
S75E-03
511E-03
272E-03
992E-03
698E-03
405E-03
125E-03
863E-03
620E-03
398E-03
195E-03
010E-03
840E-03
684E-03
540E-03
408E-03
286E-03
173E-03
068E-03
971E-03
881E-03
796E-03
717E-03
643E-03
574E-03
510E-03
449E-03
391E-03
338E-03
287E-03
239E-03
193E-03
151E-03
110E-03
072E-03
SZ
(M)
65.0
67.2
59.1
50.9
42.7
35.7
36.7
37.7
38.7
39.7
41.3
42.8
44.4
46.0
47.5
49.1
50.6
52.1
53.7
55.2
56.7
58.2
59.7
61.2
62.7
64.2
65.7
67.2
68.7
70.2
71.7
73.1
74.6
76.1
77.5
79.0
80.5
81.9
83.4
84.8
86.3
87.7
39.1
SY
(M)
111.
117.
127.
137.
146.
156.
162.
168.
175.
181.
187.
192.
198.
204.
209.
215.
220.
226.
231.
237.
242.
247.
253.
258.
263.
269.
274.
279.
285.
290.
295.
300.
305.
311.
316.
321.
326.
331.
336.
341.
346.
351.
356.
MIDP
(M)
0,
0
0.
0
0
0
0
0
0
0
0.
0.
0.
0
0,
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.
0.
0.
0.
0,
0,
0
0
0
0
0
0
0
. OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
. OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
.OOOE-01
YCU
(M)
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
37.5
57.3
65.0
65.8
60.6
48.6
22.2
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0, OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
0. OOOE-01
O.OOOE-01
0. OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
SEC
zcu
(M)
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
3.93
7.30
8.59
8.48
7.26
5.11
1.50
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-31
O.OOOE-01
O.OOOE-01
O.OOOE-01

TCL
(M)
O.OOOE-01
O.OOOE-01
76.7
110.
135.
154.
165.
174,
180.
183.
185.
185.
184.
182.
180.
176.
171.
166.
160.
154.
146.
137.
127.
115.
101.
83.6
59.5
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
0. OOOE-01
O.QOOE-01
O.OOOE-01
O.OOOE-Q1

ZCL
(M)
O.OOOE-01
O.OOOE-01
27.0
36.5
37.6
35.0
37.8
39.5
40.3
40.5
40.7
40.4
39.8
38.8
37.5
36.0
34.3
32.5
30.5
28.3
25.9
23.3
20.5
17.5
14.2
10.5
6.14
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01
O.OOOE-01

CA
(KG/M3 )
9.586E-06
1.619E-05
2.406E-05
3.207E-05
3.904E-05
4.420E-05
4.725E-05
' 4.839E-05
'4.804E-05
4.665E-05
4.463E-05
4.226E-05
3.977E-05
3.729E-05
3.492E-05
3.270E-05
3.064E-Q5
2.876E-05
2.704E-05
2.548E-05
2.404E-05
2.272E-05
2.150E-05
2.038E-05
1.935E-05
1.839E-05
1.751E-05
1.668E-05
1.592E-05
1.520E-05
1.453E-05
1.391E-05
1.332E-05
1.278E-05
1.226E-05
1.178E-05
1.132E-05
1.089E-05
1.048E-05
1.010E-05
9.739E-06
9.396E-06
9.071E-06
Xloud data set; total CPU
                      337 seconds
                                               G-21

-------
3.0  CHLORINE VAPOR RELEASE

These releases correspond to the scenarios labeled #4 in Figure 2-1 of the main report.
They are characterized by a gas liquified under pressure.  A complete description of vapor
releases is given in Section 7.0, along with the simulations made with the DEGADIS and
SLAB atmospheric dispersion modeling computer codes. In this appendix, HGSYSTEM will
be used to model this release.

The model combination used to simulate the vapor release of chlorine in this example is
PLUME and PGPLUME (refer to Figure 1). The sequence of HGSYSTEM models used
for pressurized releases of non-reactive gases such as chlorine is depicted in Figure 3. From
this figure one can see that two HGSYSTEM models are used to  describe this type of
release.  PLUME  is the initial model  and the subsequent  model  is determined  from
executing the PLUME module.  In this example, the transition is made from PLUME to
PGPLUME.

This case corresponds to scenario #4 A of Figure 2-1 in the main report. It is assumed that
there is a storage vessel containing chlorine at 278 K (-34 °C). For some reason such as the
rupture of a pipe, there is a leak from the vapor space.  In this case, it is assumed that the
orifice  is 0.01 m in diameter. Chlorine vapor jets vertically out of the orifice.

3.1  Input Description

The input to HGSYSTEM for the chlorine release described above is  given in Tables 5 and
6.  Table 5 is the standard input file (SKELETON.PLI) for the  simulation of a jet vapor
release of an ideal gas, chlorine in this case, using the PLUME module. An intermediate
file is generated from the execution of PLUME and is  the subsequent input  file for the
PGPLUME module.  This file is used  to model the transition  made from  PLUME to
PGPLUME stages (see Figure 3) and is  shown in Table 6.

3.1.1 Input for the PLUME Module

The input for the PLUME module of HGSYSTEM is contained in Table 5.

Line 1: TEMPGAS is the initial temperature of the chlorine at the point of release.

Line 2: MFGAS is the mole fraction of the pollutant, in this case 100% because the chlorine
is released as a pure vapor.

Line 3: MFH20 is the mole fraction of water vapor in the release, taken to be zero.

Line 4: MWGAS is the molecular weight of the released gas in g/mol (70 for chlorine).

Line 5: CPGAS is the specific heat of chlorine at constant pressure, 34.86 J/mol/C.

                                      G-22

-------
Line 6: DMDT is the discharge rate of 0.054 kg/s, see Section (7.2).

Line 7: DEXTT is the actual diameter of the orifice, 3/8" = 0.01 m.

Line 8: ZEXTT is the discharge height, arbitrarily taken to be 5 m.

Line 9: PHISTK  is the orientation of the release, assumed to be vertical for the purposes
of comparison with the DEGADIS input in Table (7-2).

Line 10: DURATION is the duration of release, taken to be 20 minutes (see Section (7.2)).

Line 11: ZO is the reference height at which the windspeed is measured (10 m).

Line 12: UO is the windspeed, 5 m/s.

Line 13: AIRTEMP is the ambient temperature, 5 °C.

Line 14: AIRPRESS is the ambient pressure, 1 atmosphere.

Line 15: RHPERC is the ambient relative  humidity, assumed to be zero in this example.

Line 16: ZR gives the surface roughness length of 0.1  m as assumed for all applications in
this report.

Line 17: PQSTAB gives the stability category, D in this case.

Lines 18  through  23:  these give  user-supplied criteria for terminating  the PLUME
calculation.  Negative values  imply  that the criteria are ignored.  Any keywords that are
omitted are assigned the notional value -1. SLST is the distance along the plume axis at
which the plume calculations are terminated. DLST specifies the maximum plume diameter
of interest (e.g. when the plume touches the ground). ZLSY specifies the maximum height
of interest (e.g. when the plume  encounters an inversion lid).  DXLST is the maximum
distance downwind at which PLUME  calculations are desired. ULST is a plume velocity
at which calculations will be terminated. BETLST is the lowest concentration of interest
(in volume percent). The values of the variables in lines 18 through 23 are default values
from the PLUME standard input file.
Lines 24 through 27 contain default criteria for transition to the passive dispersion model
PGPLUME or to the heavy vapor model HEGADAS.  RULST examines the difference
between the plume's velocity and the ambient windspeed.  RELST examines  the relative
values of entrainment velocities for jet dispersion versus those for  heavy gas entrainment
and/or  entrainment in the passive mode. RGLST compares the  relative magnitude of
buoyancy induced velocities and advection velocities. RNLST is another parameter that


                                      G-23

-------
compares entrainment rates prior to transition to passive advection.  See Reference (3),
Appendix 2 for further details.

The output from the PLUME model defines the input for the PGPLUME model, which
follows.

3.1.2  Input for the PGPLUME Module

The input for the PGPLUME module of HGSYSTEM is contained in Table 6.

Line 1: CPGAS is the specific heat of chlorine vapor at constant pressure, 34.9 J/mol/K.

Line 2: MWGAS is the molecular weight of chlorine (70 g/mol).

Line 3: GASFRAC is the released mole fraction of chlorine (1.0).

Line 4: WATFRAC is  the released mole fraction of water vapor (0).

Line 5: DXPLUME is the horizontal distance at which the transition from PLUME to
PGPLUME took place (m).

Line 6: ZPLUME is the height of the plume center-line  at the point of transition (m).

Line 7: DPLUME is the plume diameter at the point of transition  (m).

Line 8: PHIPLUME is the orientation of the plume at the  transition point (2.96°).

Line 9: UREL is the difference  between the plume's speed and the windspeed on the axis
at the transition point.

Line 10: CMASS is the average concentration across the plume at the  transition point
(kg/m3).

Line 11: RREL is the difference in density between the plume and its surroundings (kg/m3).


Line 12: DURATION is the effective duration of release  (i.e. the time taken for the plume
to pass through a vertical plane at the transition point).

Line 13: AIRTEMP is  the ambient temperature (°C).

Line 14: AIRPRESS is the ambient pressure (1 atmosphere).

Line 15: PHPERC is the ambient relative humidity (%).


                                     G-24

-------
Line 16: UATM = 4.63 m/s is the windspeed on the plume axis at the transition point.

Line 17: RATM is the ambient atmospheric density  (kg/m3).

Line 18: ZR is the surface roughness length (0.1 m).

Line 19: PQSTAB is the atmospheric stability category (D).

Line 20: AVTTMC is the concentration averaging time, set equal to the duration of release
(1,200 s.).

Line 21: XFIRST is the downwind distance at which PGPLUME calculations start (set equal
to DXPLUME).

Line 22: STEP is the length of successive arithmetical steps at which PGPLUME performs
calculations.

Line 23: NSTEP is the number of steps of length STEP that PGPLUME takes.

Line 24: FACTOR is the scale factor for taking steps after NSTEP equal steps have been
taken (i.e. if the arithmetic sequence of stepping ends at distance XN  downwind, the  next
step is to a distance  1.2xN ).

Line 25: XLAST is the greatest downwind distance of interest to the analyst (m).

Line 26: VFLAST is the lowest concentration of interest to the analyst (ppm).
                                      G-25

-------
                                      PGPLUME
                           crouad
                                                            HEGADAS
                                                    — "I
                           (round
                           PLUME
                      J«'. release.
                      elevated plume
                      dispersion.
                      dense plume
                      touchdown
                      and slumping
    PGPLUME
Passive far-field
dispersion of
elevated plume
                                             HEGAOAS-S
                                             HEGAOAS-T
                                          Steady or transient
                                          ground—level
                                          dispersion of
                                          dense/trace gas
      Source: Reference 1.
Figure 3. Sequence of HGSYSTEM models for a pressurized release of non-reactive gas.
A transition is made from PLUME to either PGPLUME (elevated dispersion) or HEGADAS
(ground-level dispersion).
                                       G-26

-------
          Table 5.  HGSYSTEM Input to PLUME - Vapor Jet Release of Chlorine
                 PLUME    standard input file SKELETON.PLI
TITLE   Vapor  chlorine release
GASDATA
TEMPGAS =
MFGAS
MFH20
MWGAS
CPGAS
PIPE
DMDT
DEXIT
ZEXIT
PHISTK
DURATION -
-34.0
100.0
0.0
70.0
34.86

0.054
0.01
5.00
90.00
1200
AMBIENT CONDITIONS
ZO
UO
AIRTEMP =
AIRFRESS -
RHPERC
DISP
ZR
PQSTAB
TERMINAT
SLST
DLST
"ZLST
DXLST
ULST
BETLST =•
MATCH
RULST
RELST
RGLST
RNLST
10.0
5.0
5.00
1.00
0.00

0.10
D

-1
-1E6
-.35
-500
-0.1
IE- 7

.1
.3
.3
.1
                          * PHYSICAL PROPERTIES OF GAS
                          *
                          * CELSIUS        TEMPERATURE OF  POLLUTANT
                          * PERCENT        MOLE FRACTION OF  POLLUTANT
                          * PERCENT        MOLE FRACTION OF  WATER
                          * s/mol          MOLECULAR WEIGHT  POLLUTANT
                          * J/mol/C        ISOBARIC SPECIFIC HEAT
                          *
                          * PIPE EXIT-PLANE (CHOKE-FRONT)  CONDITIONS
                          •ft
                          * KG/S           DISCHARGE RATE
                          * M              EFFECTIVE ORIFICE DIAMETER
                          * M              HEIGHT ABOVE (LEVEL) GROUND
                          * DEGREES        RELEASE DISCHARGE ANGLE
                          * S              RELEASE DURATION  (<0 FOR STEADY)

                          * ATMOSPHERIC AMBIENT CONDITIONS
                          * M
                          * M/S
                          * CELSIUS
                          * ATMOSPHERES
                          * PERCENT
REFERENCE HEIGHT
WIND VELOCITY AT HEIGHT  ZO
AIR TEMPERATURE
AMBIENT PRESSURE
RELATIVE HUMIDITY
                          * DISPERSION DATA
                          *
                          * M              SURFACE ROUGHNESS PARAMETER
                          *                PASQUILL STABILITY CLASS
                          *
                          * JET/PLUME DEVELOPMENT TERMINATION CRITERIA
                          *
                          * M              LAST REQD. DOWNWIND DISP.
                          * M              LAST REQD. PLUME DIAMETER
                          * M              LAST REQD. PLUME CENTROID RISE HI.
                          * M              LAST REQD. HORIZONTAL DISPLACEMENT
                          * M/S            LAST REQD. (MEAN) PLUME VELOCITY
                          * PERCENT        LAST REQD. POLLUTANT CONCENTRATION
                          *
                          * MATCHING CRITERIA FOR HEGADAS / PGPLUME
                          *
                          *                LAST REQD. ABS. VALUE OF UJET/UAMB-1
                          *                LAST REQD. JET/(JET+HEG) ENTRAINM.
                          *                MAX. BUOYANCY EFFECT FOR ADVECTION
                          *                MAX. BUOY. EFF. FOR PASS. DISPERSION
                                                       g
                                                       10
11
12
13
14
15
                                                       16
                                                       17
                                                       18
                                                       19
                                                       20
                                                       21
                                                       22
                                                       23
                                                       24
                                                       25
                                                       26
                                                       27
                                                  G-27

-------
       Table 6.  HGSYSTEM  Input to PGPLUME - Vapor Jet Release of Chlorine
TITLE Vapor chlorine release
        Input file for the (steady-state) passive
          adveotion model PGPLUME. The file is
       generated by the near  field dispersion model
         PLUME. It incorporates all the breakpoint
        data generated by PLUME together with such
         additional variables and flags needed to
      ensure physical consistency. In addition,  the
       file contains variables needed to complete a
       viable input file suitable for submission to
        PGPLUME; Such additional data are prefixed
        by an asterisk (*) and should be physically
      and contextually sensible, but may be changed
    at the user's discretion. Such data may also be
     overwritten by the addition of keywords to  the
        PGPLUME partial input file under HGSYSTEM.
GASDATA * released gas composition datablock.
CPGAS= 34.9 * pollutant specific heat (J/mol/C).
MWGAS= 70. Q * pollutant molecular weight (g/mol).
GASFRAO 1.00 * release mole-fraction pollutant (-).
WATGAS- O.OOOE-01 * release mole-fraction watar-vapor (-).
GEOMETRY * plume geometry at matching datablock.
DXPLUME= 7.34 * matching plane displacement (m) .
ZPLUME= 7,11 * oenttoid height above ground (m) .
DPLUME= 3.04 * near-plume (effective) diameter (m).
PHIPLUME= 2.96 * plume axis orientation (degrees).
STATE * plume dynamic/thermodynamic state.
UREL- -6.251E-02 * plume relative velocity (m/s).
CMASS- 1.632E-03 * nearfield mass-concentration (kg/m3).
RREL» 9.583E-04 * plume (mean) excess density (kg/m3).
DURATION- 1.200E+03 * (steady) release duration (s).
AMBIENT * ambient atmosphere datablock.
AIRTEMP- 5.03 ambient (air) temperature (C) .
AIRPRESS- 1.000 ambient (absolute) pressure (atm).
RHPERO O.OOOE-01 ambient (relative) humidity (Z).
UATM- 4.63 wind-speed at centroid height (m/s).
RATM* 1.27 ambient atmosphere density (kg/ra3).
DISP * Pasquill/Gifford dispersion data.
ZR" 0.100 * ground surface roughness (m) .
PQSTAB- D * Pasquill/Gifford stability class (-).
AVTIMC=» 1200. * concentration averaging time (s).
TERMINAT * output control datablock.
XFIRST- 7.34 first required downwind distance (m) .
STEP- 100. arithmetic series step- length (m).
HSTEP" 10 maximum number of (arithmetic) steps (-).
FACTOR- 1.20 scale factor for geometric series (-).
XLAST- 1.001E+04 last required downwind distance (m).
VFLAST- 1.00 last required mole concentration (mm).

1
2
3
4

5
6
7
8

9
10
11
12

13
14
15
16
17

18
19
20

21
22
23
24
25
	 26 	
                                               G-28

-------
3,2  Example of Output

The final output generated by the HGSYSTEM modeling effort describing the vapor jet
release of chlorine is shown in Table 7. Specifically, this file was obtained from running the
PGPLUME module. The output is in the form of tables of mole-concentration at several
heights above  the  ground and distances  off-axis, repeated  for a  number  of distances
downwind (a few examples only are shown).  On each page of the output, the downwind
displacement is given immediately under the heading "cross-section data".
                                      G-29

-------
Table 7. PGPLUME Output - Vapor Jet Release of Chlorine
Output from PGPLUME Version HOV90
Cross-Section Data:
downwind displacement: 7.340E-03km;
peak mole-concentration gas: 2.86E-02 Z;
plume averaging time: 20. mins;
transverse plume "width": 2.3 m;
maximum concentration height: 7.1 m;
Title: Vapor chlorine release
Cross-Section Data:
peak excess-velocity: -4.6
peak excess-density: 0.51
peak mass-concentration: 0.87
vertical plume "height": 0.93
section centroid height: 7.1
Date: 17/11/92 Time: 08:26
Virtual Source Data:
cm/s; downwind displacement: -15.
g/m3; height above ground: 7.1
g/n>3; source mass-flux gas: 5.41E-02
m; release duration: 20.
m; matching achieved: "Perfect" match
m;
m;
kg/s;
mins ;
! Table of Mole-Concentration gas (ppm) at several !
!heights (z)(m) above (level) ground, and at several!
! distances (y)(m) measured horizontally off-axis: !
i !

! O.OOOE-01 0.458
i
! O.QOOE-01! 1.124E-10 1.102E-10
i i
! 6.18 ! 173. 170.
i i
! 6.41 ! 216. 211.
i ;
! 6.65 ! 252. 247.
i ;
' 6.88 ! 277. 272.
i
! 7.11 ! 286. 280.
! 7.34 ! 277. 272.
i i
! 7.57 ! 252. 247.
t ;
! 7.81 ! 216. 211.
i i
! 8.04 ! 173. 170.
i i
!
! height (m) above !
! (level) ground. !
i t
Near-field Matching Data:
nean plume velocity-excess: -6.2
nean plume density-excess: 0.96
nean plume concentration gas: 1.6
affective plume "diameter": 3.0
ilume downwind displacement: 7.3
>lume centroid height: 7.1
flume cross-sectional area: 7.3
lean plume inclination: 3.0

0.917

1.038E-10

160.

199.

233.

256.

264.
256.

233.

199.

160.



1.38 1.83 2.29

9.388E-11 8.162E-11 6.817E-11 5

145. 126. 105.

180. 157. 131.

211. 183. 153.

231. 201. 168.

239. 208. 173.
231. 201. 168.

211. 183. 153.

180. 157. 131.

145. 126. 105.



2.75

.471E-11

84.4

105.

123.

135.

139.
135.

123.

105.

84.4



3.21 3.67 4.13 ! 	
!
4.218E-11 3.125E-11 2.224E-11!
!
65.1 48.2 34.3 '
i
81.0 60.0 42.7 !
!
94.7 70.; 49.9 !
i
104. 77.0 54.8 !
i
107. 79.5 56.5 !
104. 77.0 54.8 !
!
94.7 70.1 49.9 '
i
81.0 60.0 42.7 '
i
65.1 48.2 34.3 !
i


t
t
r
t
t
r
t
t
!
i
i
t
t
i
i
t
i
i
i
i
t
! horizontal off-axis !
! displacement (m). !


cm/s;
g/m3;
g/m3;
tn;
m;
in;
ra2;
degrees

Atmosphere Conditions:
atmosphere density: 1.3
atmosphere temperature: 5.2
atmosphere pressure: 1.0
relative humidity: O.OOE-01
ambient wind-speed: 4.6
surface roughness: 0.10
Pasquill/Gifford class: D



kg/m3;
C;
atm;
Z;
m/s;
m;
(-);

!
Pasquill/Gifford Matching Data:
peak excess-velocity: -4.6
peak density-excess: 0.51
peak concentration gas: 0.87
peak mole concentration gas- 2.86E-02
peak concentration height: 7 1
plume averaging time : 20 .
transverse plume "width": 2.3
vertical plume "height" : 0 . 93
i

cm/s ;
g/m3,
g/m3;
I;
m;
mins :
m;
m.
                        G-30

-------
                                Table 5.   PGPLUME Output  - Vapor  Jet Release  of Chlorine  (continued)
Output from PGPLUME    Version HOV90
                                               Title:  Vapor chlorine release
                                                                                        Date: 17/11/92
                                                                                                           Time: 08:26
Cross-Section Data:

downwind displacement:        8.808E-03km;
peak mole-concentration gas:  2.54E-02 Z;
plume averaging time:          20.     mins;
transverse plume "width":      2.4     m;
maximum concentration height:  7.1     m;
Cross-Section Data:

peak excess-velocity:    -4.1
peak excess-density:     0.46
peak mass-concentration: 0.78
vertical plume "height": 0.98
section centroid height:  7.1
                                                  Virtual Source  Data:

                                           cm/s;  downwind displacement:       -15.      m;
                                           g/m3;  height above ground:          7.1      m;
                                           g/m3;  source mass-flux  gas:        5.41E-02 kg/s;
                                           m;     release duration:             20.      nuns;
                                                                                 m;
                                                                                        matching achieved:  "Perfect" match
 !  Table of Mole-Concentration gas (ppm) at several !
 Iheights (z)(m) above (level) ground, and at several!
 ! distances (y)(m) measured horizontally off-axis:  !



i
i
! O.OOOE-01!
t
! 6.13
!
! 6.37
r
! 6.62
!
! 6.36
i
! 7.11
I
! 7.36
I
! 7.60
!
! 7.85
i
! 8.09
i
!
i
T
t
!
i
!
i
I
i
i
I
i
1
i
i
!
!
!
!


O.OOOE-01

2.280E-09

154.

192.

224.

246.

254.

246.

224.

192.

154.



0.487

2.235E-09

151.

188.

220.

242.

249.

242.

220.

188.

151.



0.974

2.105E-09

142.

177.

207.

227 .

235.

227.

207.

177.

142.



1.46

1.905E-09

129.

160.

187.

206.

212.

206.

187.

160.

129.



1.95

1.656E-09

112.

139.

163.

179.

185.

179.

163.

139.

112.



2.44

1.383E-09

93.5

116.

136.

149.

154.

149.

136.

116.

93.5



2.92

1.110E-09

75.0

93.4

109.

120.

124.

120.

109.

93.4

75.0



3.41

8.559E-10

57.9

72.0

84.2

92.5
•
95.4

92.5

84.2

72.0

57.9


! height (m) above !
! (level) ground. !
i

i








3.90

6.340E-10

42.9

53.3

62.4

68.5

70.7

68.5

62.4

53.3

42.9



4.38 !
) i
4.513E-10! !
t ,
30.5 i '
i i
38.0 ! !
I 1
44.4 ' '
t t
48.8 ! !
i t
50.3 t !
i t
48.8 ' !
i t
44 .4 ' !
i i
38,0 ' '
t t
30.5 ! i
t i
i
! horizontal off-axis !
! displacement (m) '
!
t
Near-field Matching Data:

mean plume velocity-excess:   -6.2
mean plume density-excess:    0.96
mean plume concentration gas:   1.6
effective plume "diameter":    3.0
plume downwind displacement:   7.3
plume centroid height:         7.1
plume cross-sectional area:    7 3
mean plume inclination:        3.0
         Atmosphere Conditions:

cm/s;    atmosphere density:      1.3     kg/m3;
g/m3;    atmosphere temperature:  5.2     C;
 g/m3;   atmosphere pressure:     1.0     atm;
m;       relative humidity       O.OOE-01 %;
m;       ambient wind-speed:      4.6     m/s;
m;       surface roughness:       0.10     m;
m2;      Pasquill/Gifford class:   D      (-);
degrees;
                                         Pasquill/Gifford Matching Data:

                                         peak excess-velocity:        -4.6     cm/s;
                                         peak density-excess:         0.51     g/m3;
                                         peak concentration gas:      0.87     g/m3.
                                         peak mole concentration  gas-  2.36E-02 7,,
                                         peak concentration height:    71     m;
                                         plume averaging time:         20.     mins:
                                         transverse plume "width":    2.3     m;
                                         vertical plume "height":     0.93     m.
                                                              G-31

-------
Table 5.   PGPLUME Output - Vapor Jet Release o£ Chlorine (concluded)
Output from PGPLUME Version NOV90
Cross-Section Data:

downwind displacement: 0.598 km;
peak mole-concentration gas: l.OOE-04 Z;
plume averaging time: 20. mins;
transverse plume "width": 50. m;
maximum concentration height: O.OOE-01 m;
Title: Vapor chlorine
Cross-Section Data:
release

peak excess-velocity: -0.28 cm/s;
peak excess-density: 1.79E-03 g/m3;
peak mass-concentration: 3.05E-03 g/m3;
vertical plume "height": 19. m;
section centroid height: 16. m;
Date: 17/11/92 Time: 08:26
Virtual Source Data:
downwind displacement: -15.
height above ground: 7.1
source mass-flux gas: 5.41E-02
release duration: 20.
matching achieved: "Perfect" match

m;
m;
kg/s;
nuns ;
! Table of Mole-Concentration gas (ppm) at several !
'.heights (z)(m) above (level) ground, and at several!
! distances (y)(m) measured horizontally off -axis: !
i i


! O.OOOE-01 10.00 20.0
i
! O.OOOE-01! 1.00 0.980 0
i I
! 2.37 ! 0.993 0.974 0
i i
! 4.74 ! 0.974 0.954 ' 0
i t
! 7.11 ! 0.942 0.923 0
t t
! 11.9 ! 0.846 0.829 0
I j
! 16.6 ! 0.720 0.706 0
i t
'21.4 ! 0.580 0.568 0
i i
! 26.2 ! 0.442 0.433 0
I j
! 31.0 ! 0.319 0.312 0
; !
! 35.7 ! 0.217 0.213 0
! ! -
i
! height (m) above !
! (level) ground. !
i i
Near-field Matching Data:
mean plume velocity-excess: -95.
mean plume density-excess: 0.96
mean plume concentration gas: 1.6
effective plume "diameter": 3.0
plume downwind displacement: 7.3
plume centroid height: 7.1
plume cross-sectional area: 7.3
mean plume inclination: 3.0

.923

.917

.899

.869

.781

.664

.535

.408

.294

.200




30.0 40.0 50.0

0.835 0.726 0

0.830 0.721 0

0.813 0.707 0

0.787 0.684 0

0.707 0.614 0

0.601 0.523 0

0.484 0.421 0

0.369 0.321 0

0.266 0.231 0

0.181 0.158 0



.607

.603

.591

.571

.513

.437

.352

.268

.193

.132



60.0

0.487

0.484

0.474

0.458

0.412

0.350

0.282

0.215

0.155

0.106



70.0 80.0 90.0 ! 	
i
0.375 0.278 0.198 !
i
0.373 0.276 0.197 '
I
0.365 0.271 0.193 !
i
0.354 0.262 0.186 !
i
0.318 0.235 0.167 !
i
0.270 0.200 0.142 \
i
0.218 0.161 0.115 !
i
0.166 0.123 8.745E-02!
i
0.120 8.858E-02 6.305E-02!
t
8.147E-02 6.036E-02 4.296E-02!
r
-

i
'
;
!
1
1
1
1
1
1
1
1
1
1
1
I
1
r
i
!
1
1
! horizontal off-axis
' displacement (m) .


cm/s;
g/m3;
g/m3;
m;
m;
ro;
m2;
degrees:

Atmosphere Conditions :
atmosphere density:
atmosphere temperature:
atmosphere pressure:
relative humidity:
ambient wind-speed:
surface roughness:
Pasquill/Gifford class:



1.3
: 5.2
1.0
O.OOE-01
5.5
0.10
: D



kg/m3 ;
C;
atm;
%;
m/s;
m;
(-);

!
Pasquill/Gifford Matching Data:
peak excess-velocity: -93.
peak density- excess: 0.51
peak concentration gas : 0 87
peak mole concentration gas: 2.86E-02
peak concentration height: 7.1
plume averaging time: 20.
transverse plume "width": 2.3
vertical plume "height". 1.1


cm/s ;
g/m3,
S/m3,
• ;
m;
nuns ;
m;
m.
                              G-32

-------
4.0  REFERENCES
1.     Witlox, H.W.M., "HGSYSTEM: Dispersion Models for Ideal Gases and Hydrogen
      Fluoride, Tutorial and Quick-Reference Guide",  External Report TNER.90.007,
      Thornton Research Centre, Shell Research, Chester, England (May, 1991).

2.     McFarlane, K., Prothero,  A., Puttock, J.S., Roberts, P.T., and  Witlox, H.W.M.,
      "Development and Validation of Atmospheric Dispersion Models for Ideal Gases and
      Hydrogen  Fluoride,  Part  I: Technical  Reference  Manual",  External Report
      TNER.90.015,  Thornton  Research  Centre,  Shell  Research,  Chester, England
      (November, 1990).

3.     Witlox, H.W.M., McFarlane, K., Rees, F.J., and Puttock, J.S.,  "Development and
      Validation of Atmospheric Dispersion Models for Ideal Gases and  Hydrogen
      Fluoride,  Part II: HGSYSTEM Program  User's  Manual",  External Report
      TNER.90.016,  Thornton  Research  Centre,  Shell  Research,  Chester, England
      (November, 1990).

4.     Colenbrander, G.W. and J.S.  Puttock, "Decription of the HEGADAS  Model  for
      Dense Gas Releases," External Report TNER.90.022, Thornton Research Centre,
      Shell Research, Chester, England (1989).
                                     G-33

-------
             APPENDIX H

CALCULATION OF MOLECULAR DIFFUSIVITY
        FROM BASIC PRINCIPLES

-------
                                  APPENDIX H

                 CALCULATION OF MOLECULAR DIFFUSIVITY
                           FROM BASIC PRINCIPLES
The equation for the rate of evaporation of a liquid with an above ambient boiling point
from a pool lying on the ground contains the molecular diffusivity Dm of the diffusing species
in air, see Eq. (5-3) in Section 5.1.3 of the main body of the report.  If D,,, is not available
from data compilations, it can be calculated by using the Chapman-Enskog kinetic theory
of gases as explained by Bird et al.(21) (the reference is to be found in Section 11):

            Dm = 0.001853.((Tb)3(l/Ma +  l/Mb))ai/B cm2/s                    (1)

            B = Pa(Sab)2.Oab                          .                       (2)

The following is an example for  a pool of HF spilled on the ground at a temperature Tb of
60 °F =  298 K (see Section 5.3). Ma is the molecular weight of air (28.9). Mb( = M) is the
molecular weight of HF, taken here to be 70 because HF vapor oligomerizes (associates into
(HF)2, (HF)6 and  possibly  (HF)8).  pa is the atmospheric  pressure, l.OlxlO5  Pa  =  1
atmosphere.

The quantity S^ is expressed as the  arithmetic mean of two  other quantities: Sab = (sa +
sb)/2, measured in  angstroms.  For air,  Bird et al. give a value of sa equal to 3.617.  For HF,
sb is not tabulated  by Bird et al. and has to be calculated from the critical temperature Tc
(K) and the critical pressure Pc(atmospheres) using a formula that is also given by Bird et
al.:

            sb =  2.44(TC/PC)"3                                               (3)

For HF, T0 =  461 K and  Pc  = 64 atmospheres so that sb =  2.44(461/64)1/3  = 4.712
angstroms. Hence, Sab = (3.617 + 4.712)/2 = 4.16 angstroms.

The quantity Oab is a  complicated function of another quantity eab/k which in turn is the
geometrical mean  of two other quantities  ea/k and eb/k, where k is Boltzmann's constant.

            e^/k =  (e,eb)°-Vk                                              (4)

For air, Bird et al. assign a value of 97 to ea/k.  For HF, eb/k = 355. Therefore eab/k =
(355x97)a5 =  186.

There is a table in Bird et al. (Table B-2) that relates O^ to kTb/ert. For kJJs^ = 289/186


                                        HI

-------
=  1.557, the table gives Oab - 1.180. Combining all of the above gives B = (1)(4.16)2(1.180)
=  20.46 and Dm = (0.001853)((2893)( 1/28.9 + 1/70))0-5 = 0.0984 cm2/s = 9.84xlO'6 m2/s.

The above method is applicable when Dm is not available from data sources.  A typical
default value for Dm for many materials is ~ 10"3 m2/s.
                                         H2

-------
                                    TECHNICAL REPORT DATA
                  (Please read Instructions  on reverse before completing)
1. REPORT NO.
   EPA-454/R-93-001
                                                             3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE

   Contingency Analysis Modeling  for Superfund
   Sites and Other  Sources
                                   5. REPORT DATE
                                     January  1993
                                   6. PERFORMING ORGANIZATION CODE
7. AUTHOR
-------