NTJS
United States     Office of Air Quality      EPA-450/4-88-009
Environmental Protection  Planning and Standards     September 1988
Agency       Research Triangle Park NC 27711
Air
A WORKBOOK OF
SCREENING
TECHNIQUES FOR
ASSESSING IMPACTS
OF TOXIC AIR
POLLUTANTS
                      Of

-------
                                        EPA-450/4-88-009
A Workbook Of Screening Techniques  For
Assessing Impacts Of Toxic Air Pollutants
                             by.
                   TRC Environmental Consultants, Inc.
                      800 Connecticut Boulevard
                       East Hartford, CT06108

                   EPA Project Officer  Jawad S Touma
                      Contract No. 68-02-3886
                          Prepared for

                 U S. ENVIRONMENTAL PROTECTION AGENCY
                 Office of Air Quality Planning and Standards
                      Technical Support Division
                    Research Triangle Park, NC 27711

                         September 1988

-------
This report has been reviewed by the Office of Air Quality Planning and Standards, US EPA, and has been
approved for publication. Mention of trade names or commercial products is not intended to constitute
endorsement or recommendation for use. Copies of this report are available, for a fee, from the National
Technical Services, 5285 Port Roval Road, Springfield VA 22161.

-------
                               ACKNOWLEDGEMENTS








    This workbook was  prepared by Daniel J. McNaughton  and  Paul  M. Bodner  of




TRC Environmental  Consultants, Inc.  under  contract to  the  U.S.  Environmental




Protection  Agency,   Office of Air  Quality  Planning  and  Standards,  Source




Receptor  Analysis  Branch.   The work  was  directed by  Jawad S.  Touma,  Source




Receptor Analysis Branch.
                                     -in-

-------
                             TABLE OF CONTENTS

                                                                         Page

LIST OF FIGURES	       ix

LIST OF TABLES	       xi

1.0             INTRODUCTION	      1-1

2.0             SELECTION OF SCREENING TECHNIQUES FOR TOXIC AIR
                  CONTAMINANTS  	      2-1

    2.1           Release Categorization  	      2-1
    2.2           Limitations and Assumptions 	      2-2
    2.3           Technique Selection and Use	      2-3
    2.4           Determining Maximum Short-Term Ground Level
                    Concentration 	      2-8

3.0             SUPPORT DATA FOR SCREENING ESTIMATES	      3-1

    3.1           Meteorological Data	      3-1
        3.1.1       Wind Speed and Direction	      3-1
        3.1.2       Stability and Turbulence	      3-2
        3.1.3       Temperature	      3-4
        3.1.4       Atmospheric Pressure  	      3-4
    3.2           Chemical and Physical Parameters  	      3-4

4.0             SCENARIOS AND TECHNIQUES FOR RELEASE AND EMISSIONS
                  ESTIMATES	      4-1

    4.1           Continuous Particulate and Gaseous Releases from
                    Stacks	      4-3
        4.1.1       Particulate Matter  	      4-3
        4.1.2       Gases	      4-4
    4.2           Continuous Releases of Fugitive Dust  	      4-5
    4.3           Ducting Failures With Dust Releases	      4-6
    4.4           Flare Emissions	      4-7
    4.5           Continuous Gaseous Leaks from Tanks/Pipes 	      4-8
    4.6           Instantaneous Gaseous Releases from Stacks  ....     4-11
    4.7           Multiple Fugitive Continuous Gaseous Emission
                    Sources	     4-12
    4.8           Continuous Gaseous Emissions from Land Treatment  .     4-13
    4.9           Continuous Emissions from Municipal Solid Waste
                    Landfills	     4-15
    4.10          Continuous Emissions of Pesticides and Herbicides .     4-17
    4.11          Instantaneous Emissions Due to Equipment Openings .     4-18
    4.12          Evaporation from Quiescent or Aerated Surface
                    Impoundments  	     4-20
    4.13          Continuous Relief Valve Discharges (Two-Phase Flow)     4-24
    4.14          Instantaneous Relief Valve Discharges (Two-Phase) .     4-28
    4.15          Low Volatility Liquid Leaks from Pipes  	     4-29
    4.16          Low Volatility Liquid Leaks from Tanks  	     4-32
                                    -v-

-------
                                                                         Page

    4.17          High Volatility Liquid Leaks  from  Pipes  	     4-34
    4.18          High Volatility Liquid Leaks  from  Tanks  	     4-35

5.0             ATMOSPHERIC DISPERSION ESTIMATES   	      5-1

    5.1           Cloud Densities	      5-2
        5.1.1       Calculations to Determine the  Relative Density of
                      Instantaneous Releases   	      5-2
        5.1.2       Calculations to Determine the  Relative Density of
                      Continuous Releases 	      5-3
    5.2           Plume Rise Calculations	      5-6
        5.2.1       Mean Molecular Weight for Mixtures of  Gases .  .  .      5-6
        5.2.2       Flare Plume Rise	      5-6
        5.2.3       Buoyancy Plume Rise	      5-8
    5.3           Dispersion Parameters 	     5-10
        5.3.1       Horizontal and Vertical Dispersion Parameters
                      for Continuous Emission Releases  	     5-10
        5.3.2       Horizontal and Vertical Dispersion Parameters
                      for Instantaneous Emission Releases  	     5-13
        5.3.3       Horizontal and Vertical Dispersion Parameters
                      for Wake Effects	     5-16
    5.4           Buoyancy-Induced Initial Dilution  	     5-18
    5.5           Virtual Source Distances  	     5-19
        5.5.1       Virtual Distances for Area  Sources	     5-19
        5.5.2       Virtual Distances for Volume Sources  	     5-20
        5.5.3       Virtual Distances for Wake  Effects	     5-21
    5.6           Concentration Calculations   	     5-24
        5.6.1       Cavity Modeling	     5-24
        5.6.2       Heavy Gas Model - Instantaneous  Releases  ....     5-26
        5.6.3       Heavy Gas Model - Continuous Release  	     5-28
        5.6.4       Dispersion Model for Continuous  Releases  ....     5-31
        5.6.5       Dispersion Model for Instantaneous Releases .  . .     5-32

6.0             EXAMPLES	      6-1

    6.1           Continuous Gaseous Emissions  Prom  Stacks  	      6-1
        6.1.1       Building Cavity Example 	      6-1
        6.1.2       Near-Wake Example 	      6-3
        6.1.3       Far-Wake Example  	      6-5
    6.2           Fugitive Dust	      6-6
    6.3           Instantaneous Ejection of Particles from Ducts  . .      6-8
    6.4           Flare Emissions	     6-10
    6.5           Continuous Gaseous Releases from Tanks or Pipes . .     6-14
    6.6           Instantaneous Gas Releases   	     6-20
    6.7           Continuous Releases of Fugitive  Emissions 	     6-23
    6.8           Continuous Gaseous Emissions from Land Treatment  .     6-25
    6.9           Municipal Solid Waste Landfill   	     6-27
    6.10          Continuous Emissions from an Herbicide  	     6-29
    6.11          Equipment Openings  	     6-32
    6.12          Continuous Gaseous Emissions from Surface
                    Impoundments  	     6-34
                                   -vi-

-------
                                                                         Page

    6.13          Relief Valve Discharge (Two-Phase)  	    6-38
    6.14          Two-Phase Instantaneous Release 	    6-44
    6.15          Liquid Release from a Pipe	    6-47
    6.16          Low Volatility Liquid Releases from Tanks 	    6-49
    6.17          High Volatility Liquid Release from a Pipe  ....    6-51
    6.18          High Volatility Liquid Releases from Tanks  ....    6-53

REFERENCES      	     R-l

APPENDIX A      EMISSION FACTORS  	     A-l

APPENDIX B      GLOSSARY  	     B-l

APPENDIX C      FLOWCHARTS FOR WORKBOOK SCENARIOS 	     C-l

APPENDIX D      FLOWCHARTS FOR DISPERSION CALCULATIONS  	     D-l

APPENDIX E      AVERAGING PERIOD OF CONCENTRATION ESTIMATES 	     E-l

APPENDIX F      SELECTED CONVERSION FACTORS 	     F-l

APPENDIX G      CALCULATIONAL METHODS FOR DISPERSION PARAMETERS ...     G-l
                                   -VI1-

-------
                                LIST OF FIGURES

Figure No.                                                                 Page

   2-1    Release Scenarios 	     2-5

   5-1    Horizontal Dispersion Parameter (oy) as a Function of
            Downwind Distance and Stability Class (Continuous
            Releases)	    5-11

   5-2    Vertical Dispersion Parameter (az) as a Function of
            Downwind Distance and Stability Class (Continuous
            Releases)	    5-12

   5-3    Horizontal Dispersion Parameter (ay) as a Function of
            Downwind Distance and Stability Class (Instantaneous
            Releases)	    5-14

   5-4    Vertical Dispersion Parameter (az) as a Function of
            Downwind Distance and Stability Class (Instantaneous
            Releases)	    5-15
                                     -ix-

-------
                                 LIST OF TABLES

Table No.                                                                  Page

   2-1    Release Scenarios 	      2-4

   2-2    Wind Speed and Stability Class Combinations 	      2-9

   2-3    Calculation Procedures for Use with Various Emission Heights
            (Continuous Releases) 	     2-10

   2-4    Approaches for Maximum Concentrations 	     2-12

   3-1    Wind Profile Exponent as a Function of Atmospheric Stability      3-2

   3-2    Key to Stability Categories	      3-3

   3-3    Typical Physical and Chemical Property Parameters Used in
            Emission Modeling 	      3-5

   6-1    RVD Model Results:  Chlorine Gas Leak	     6-17

   6-2    RVD Model Results:  Chlorine Two-Phase Release  	     6-41
                                     -xi-

-------
1.0 INTRODUCTION




    This workbook provides  a  logical  approach to the selection  of  appropriate




screening  techniques  for  estimating  ambient concentrations  due  to  various




toxic/hazardous  pollutant  releases.   Methods used  in  the  workbook apply  to




situations where a  release  can be fairly well defined,  a condition  typically




associated with  non-accidental  toxic  releases.   The format of this  workbook is




built  around  a  series  of  scenarios  which may  be  considered  typical  and




representative of  the  means  by  which  toxic chemicals  become  airborne.   In




addition, suggestions are provided for modeling less typical  cases.




    Screening techniques are  simplified calculational procedures designed with




sufficient conservatism  to  allow  a determination of whether  a source:    1)  is




clearly not  an air  quality  threat or  2)  poses a potential threat which should




be  examined  with more  sophisticated  estimation techniques  or  measurements.




Screening  estimates  obtained  using this  workbook represent  maximum short-term




ground  level  concentration  estimates  from  a  meteorological  perspective.   If




the  screening  estimates demonstrate  that during these conditions the  ground




level  concentrations  are not  likely  to  be  considered  objectionable,  further




analysis  of  the source impact would  not  be  necessary  as  part  of the  air




quality  review  of  the  source.   However,  if screening   demonstrates  that  a




source may have  an  objectionable  impact, more detailed source  impact analysis




would be required using refined emissions and air quality models.




    Methods used in  this workbook should be  applied with caution.   Techniques




for estimating emissions are  evaluated  and  revised on  a continuing basis  by




EPA.  Thus the user  should  consult with EPA on the  most recent emission models




and emission factors.  Meteorological methods presented in this  report  reflect




guidance published  elsewhere,  and in particular the Procedures  for Evaluating




Impact of  Stationary  Sources  that includes the  PC-based model  SCREEN  (1988b)




and  the Guideline  on Air  Quality Models  (Revised)  (1986)  and Supplement A




                                      1-1

-------
(1987).   The  Regional  Modeling Contact  should  be  consulted as to the  present




status of guidance in air quality modeling.




    The  workbook  is   organized   into  six   sections   and  six   supporting




appendices.    Section 2  discusses  selection  of  screening techniques  and  the




general  approach  to  using the workbook.  Users are advised  to consult  this




section both  for  releases explicitly presented  in the workbook  and for  less




typical  releases.  The  section also  considers assumptions,  limitations  and




conservatism  of  estimates.    Section 3  describes  the   support   data   (i.e.,




meteorological  data  and  chemical  and physical  parameters)  needed  for  making




estimates.   Section 4 helps the user identify applicable  release  scenarios and




determine release and emission  rates.   In this workbook 18 release scenarios




have  been  selected  to  represent  situations  likely  to  be   encountered.




Section 5 of  the  report  guides the user  through all the  steps  required for




making  atmospheric  dispersion   estimates   once  atmospheric  emissions   and




contaminant characteristics  are known.  Section 6  provides an example  of the




emission  and  associated  dispersion  estimation  methods  for  each  specific




release scenario.




    Appendix A  discusses  currently  available  sources for  obtaining  emission




factors  that  can  be  used for  some of  the  scenarios.   Appendix B  provides  a




glossary of  terms  applicable  to  air  toxics   modeling.   Appendices C  and  D




contain flow  diagrams to be  used as a  guide in Sections 4 and 5 for selecting




applicable  emission  and dispersion  calculation methods.   Appendix  E  presents




methods   for   converting   concentrations   to   different   averaging   times.




Appendix F  provides  some  useful  unit  conversion  factors  applicable  to  the




workbook.   Appendix G provides methods  for  calculating  dispersion  parameters




as an alternative to using graphical methods.




    This workbook supersedes  information  in EPA  Report  EPA-450/4-86-11, Some




Applications of Models to Air Toxics Impact Assessments.




                                      1-2

-------
2.0 SELECTION OF SCREENING TECHNIQUES FOR TOXIC AIR CONTAMINANTS

    This workbook  attempts to account  for many  of the scenarios  expected to

produce toxic chemical releases to the atmosphere.



2.1 Release Categorization

    Selection  of  appropriate  techniques  for  screening   estimates  requires

categorization  of  the  toxic  chemical  release of  interest.   There are  three

overlapping categories which  should  be considered  when defining problems  for

screening:


    1) Physical  State  -  Gaseous  releases  to  the  atmosphere  can,   in
       general, be  simulated  using techniques developed for  criteria  air
       pollutants unless the gas is dense, is highly  reactive,  or rapidly
       deposits  on  surfaces.   Additional   source   modeling   must   be
       performed if  the release  is  liquid,  aerosol  or  multi-phased  to
       determine the  state of the  material as  it becomes available  for
       dispersion in air.

    2) Process/Release Conditions  - Knowledge of  the circumstances under
       which chemicals  are released  helps to  determine  both state  and
       dispersive characteristics.  For  example,  location  of a leak in a
       pressurized liquefied gas storage tank will  determine  if a release
       is  liquid or  gas  and if  source  modeling is  required  prior  to
       dispersion estimates.

    3) Dispersive Characteristics  -  Techniques for  pollutant  dispersion
       estimates are  categorized by  terms such  as  instantaneous  versus
       continuous,  or point versus area or volume  releases.   To  complete
       dispersion estimates,  this  final  characterization  is required  at
       some point in concentration calculations.


The primary emphasis  in this  workbook  is to provide the  user with screening

techniques  for estimating  short-term,  ground  level  concentrations of  toxic

chemicals  released  to  the atmosphere.   However,  in order  to  do this,  the

workbook  also   provides  assistance  to  the user  in  formulating  the  release

conditions.  In framing each  problem, the user must  reason the  path  required

to complete the concentration estimate.
                                      2-1

-------
2.2 Limitations and Assumptions

    Methods  included  in  this  workbook  are  intended  to  provide  simplified,

conservative dispersion estimates for  situations  which may represent  extremely

complex  release  scenarios.    As  such,   the   methods   are  limited  in  their

applicability.   Some of these limitations are  as  follows:


    •  Screening techniques  provided  are  intended for  use on  small  to
       mid-scale non-accidental releases.

    •  All  techniques  assume  that  the   toxic air  contaminant  is  non-
       reactive and non-depositing.  Thus  these  screening methods are not
       applicable for reactive gases and  settleable particles.

    •  Conditions  resulting   in  worst   case  concentrations   cannot  be
       uniquely defined in instances where gases  released  are  denser  than
       air  or  where meteorological conditions  affect source  estimates.
       For example, in the  case  of evaporation,  the highest source rates
       are  related to  high  wind  speeds.   High wind  speeds,  however,
       result in more dilution which acts to lower concentrations in air.

    •  Time  dependent  emissions   cannot   be  simulated  with  a  simple
       screening technique.   Techniques  provided  assume   steady  releases
       for a specified period.

    •  Methods are not provided for the following phenomena:

       -  heavy gas releases influenced by obstructions to flows

       -  non-vertical jet releases of heavy or passive gases

       -  the  influence  of  aerosol evaporation,  deposition,  surface or
          radiational  cloud   heating  and   exothermic   or   endothermic
          reactions on dispersing clouds

    •  All release calculations assume thermodynamically  ideal conditions
       for gas and liquid flows.

    •  Pasquill-Gifford   dispersion  parameters   for   continuous  plumes
       (Turner, 1970) are assumed to represent hourly average conditions.

    •  Building wake  effects  calculations  are based  on  methods  used in
       the  Industrial  Source  Complex  Model  (EPA,  1987c)  and  cavity
       calculations are  based on  methods  in Procedures (EPA,  1988b).
       Simple  rectangular  building geometries are assumed in  both cases.
       When  a  selection of  wake effects  is  made,  calculations  are not
       performed for receptors within three building heights downwind.

    •  Complex and elevated terrain effects are not considered.
                                      2-2

-------
    •  Sources  in  rural  and  urban  settings can  use  the same  procedures
       presented here.   Dispersion  parameters  for urban settings are  not
       presented.  They may be obtained from other EPA publications.

    •  Screening  techniques  are  applicable to  short-term  concentration
       estimates.
2.3 Technique Selection and Use

    Eighteen most prevalent  release  scenarios  were selected for this  workbook

and  are  grouped  in  Table 2-1  according  to  three  categories:   particulate

matter, gases and liquids.   For some of  the  release scenarios,  subcategories

have  been  added.   Table 2-1  also  provides   a  convenient  look-up  table  to

indicate  relevant   sections  in  this  report   that are  associated  with  each

release  scenario.    Figure 2-1  provides  a  graphical   illustration  of  each

release  scenario.    Appendix  C  guides  the  user   through  techniques   for

estimating  release  and   emission   rates.    Appendix D  guides   the  user  in

selecting the atmospheric dispersion estimates for these scenarios.  The first

step  in  analyzing  any  scenario  is,   therefore,  to consult  the  appropriate

release flowchart in Appendix C.

    Steps in using  the workbook are as  follows:


    1. Select the  release scenario  from Figure 2-1  and Table 2-1  which
       most   closely   resembles   the   release   of   concern.    Release
       descriptions   provided  in  the report  sections and  noted  in  the
       table provide selection guidance.

    2. Follow the instructions given in the report section  describing  the
       release  of  concern.   The  flowchart  referenced  either guides  the
       user through emission calculations  in  Section 4 or  indicates  that
       emission  estimates can be  obtained  from a  separate  source  of
       emission factors (Appendix A).

    3. The  final step  of  the  applicable release  flowchart  indicates  the
       appropriate  dispersion  flowchart  in Appendix D  which  guides  the
       user in  determination of ambient concentrations.  The  user should
       proceed to the  flowchart keeping in mind the dispersion  modeling
       category  specified   for  the   release   scenario  of   concern   in
       Table 2-1  (i.e.,   instantaneous   or  continuous  emissions  from  a
       point,  area,  or volume  source).   The dispersion  model  category is
       used at a decision point in the  dispersion flowchart.
                                      2-3

-------
































8
M

rH <
CN H
O
w co
m W
< CO
EH <

•j
g


































co G
rH O
ft-H
mri
U
X CO
W W
C
O 4->
rH -H U
(0 CO rO
•rt U J2
4-> 0) U
•H ft 5
C co O
M -H rH
P Pn
4->
C U
O (0
•H x;
Wf »
(ui
w 5

'e -H"
w £
G
o c
•H O
CO -H
CO i i
• iH r i
n (j
P d)
Cn >i
G U
•H O
rH tj>
Q) Q)
T3 4-)
O 03
2 U













co
QI
t^l
H

CO
en
rd
CO
CO
a








ca
to
CO
rH rH
ro co
•H a
•H 4-4
C 0
M

U
O




rH CN 00 ^ rH IT)
VO U3 W5 i£> VD VO





CN <* kO r~ rH i-H
III III
Q Q Q Q Q P







f~H 0*3 CO ^T LO ^D
iii iii
o o o o o o



rH CN]
•
CN C*J ^7* I-H UJ
«„«, „„„



ft < > ft ft ft
U U M U O O










V
en
ft
co e
4J 3
G ft
0)
> -
ca
- CO
CO ft
CO -H
> ft
ro -
.2
MH G
4-> CO (0
ca co -H H
Q 3 co ••
>H aw
0) 'H i
> rd k <0
•H fe CQ ca co
Jl! 4-> CO ^ >-J
U -H 4J U CJ
(0 tji U fO tO CO
4-133 rH 4J (0






CO
4-1
m
•-H U
3 co
U 4J
•H 4-1 en
4-J fO CU
ij S en
rO (0
ft O


O rH
VD r*- 00 CT» rH rH
v£> VD v£> M3 VD vo


IT)
1
Q

1 * 1 1 1 1
Q «tf Q Q Q Q
1
P




O rH CN]
1 1 1 1 1 1
CJ U U O O U




O rH

„, „ ^ ^ <* ^



CU 
M cj c3 c3 o H




co
G
o
•H
CO
CO
•H
tl

rH
rH
en -H
1 i u) l^_j
C C T3
CO O G
> -H CO tO
eo G iJ
» en O
to .H -H co ca
co € en 4-> co
> fl en in T3
rH -H (0 "H CO
(0 Q) S !2 O C7»
> > fl .H G
IM 4-1 4-1 -H U G
CO CO -H C rH CO CO
U -H CTI 0) O K ft
o rH 3 e t>3 ^ c5
CQ co fa 4-) ca
CQ Qi (0 rH CO 4-1
cu 0) 0) ro TJ C
U «• rH U ft -H 0)
ft en ft H -H u g
£ i! .H O -H ft
0 O 4-1 T3 -H 4J 'H
U ro rn G C CO 3
4-1 3 (0 3 CO C71
W S J S ft W


















CN CN ro •*
rH rH rH rH
vfi vO ^ v^





^ ^ rH VO
III 1
P P P P






n ro «* in
III I
O U 0 O




CN CN CO 
30) Q) r-H
(5* < co ca ro
*•—'* %w^ ^ fO ^
rH JiJ
w w m cu M-I
-M -M ^ QJ
C G O -H
co co IH 5 r-n
e e co H co
•6 -3 .H a
3 3 0) -^ eo
o o a co 3
CU ft co O
g g en ro co
M M 3 X G
o ft m
CO CO 3 4J
U U C O G
(0 rO *rt .£ rO
4^ 14H 4-> 4-> 4-1
t-i i-i G «-* en
330 G
t/l CO O M









ca

•H

Cr*
•H
J


ID ID r~ oo
rH rH rH rH
VO ^ VD VD





^ ^ rH rH
1 III
P P P P






vo r*^ oo o^
i iii
o o o o




LT) VD f**" CO

^ <* ^ ^



6< ft ft
o o o





•O 0) "^ >1 r>1
0) 4-1 >t 4J 4-»
N -H 4-1 "H -H
•H C "H rH rH
U -H rH -rl -H
3 -4-1 -H 4J 4J
CO -P It! CO
CO >• tO rH rH
Q) CO rH O O
U 3 0 > >
ft O >
13 £ £,
G G S O> Cn
O -H O -H -H
C 4-> rH X! £
i»-^ C *•— •• "^^* *****
o
ca u ca ca ca
CO M CO .*
ft •• C ft C
•H •^'^ rfQ t-H fO
ft >i H ft EH
4J
1 -H 1 | I
CO 6? -H CO CO CO
CO J< 4J J O^ rH 0^ (J^ tj^
•H -H .H 'H
J J J J












































































3
O
CQ CO
3 G
O rd
3 4->
G G CO
•H 4J rd g
4J G to 4J 3
G -H co en rH
SO u G O
ft < M >

1 1 1 1 1

O ft < M >

•K

2-4

-------
STACK RELEASES OF GASES
 OR PARTICLES (4.1.1, 4.1.2)
                                     FUGITIVE DUST EMISSIONS (4.2)
PARTICULATE MATTER
DUCT FAILURES (4.3)
FLARE EMISSIONS (4.4)
   GAS LEAKS (4.5)
    INSTANTANEOUS
GASEOUS EMISSIONS (4.6)
            FIGURE 2-1.  RELEASE SCENARIOS
             (Section Number in Parentheses)
                              2-5

-------
            fit
             _n
 MULTIPLE FUGITIVE EMISSIONS (4.7)
   LAND TREATMENT (4.8)
MUNICIPAL SOLID WASTE LANDFILL (4.9>
PESTICIDES/HERBICIDES (4.10)
       EQUIPMENT OPENING (4.11)
SURFACE IMPOUNDMENTS (4.12)
            FIGURE 2-1.  RELEASE  SCENARIOS (Continued)



                    (Section Number in Parentheses)
                                   2-6

-------
                                                    „.
                                                 O ° O  0°00°01



                                                       °
                                              00
                                                   oo ° oo
                                            ••*##•##*
                                             o  o°o o°o o°o o°o
                                              ? oo ° oo ° oo  <•
RELIEF VALVE - CONTINUOUS (4.13)
RELIEF VALVE - INSTANTANEOUS (4.14)
 LIQUID LEAK - PIPE (4.15)
                                       LIQUID LEAK - TANK (4.16)
HIGH VOLATILITY PIPE LEAKS (4.17)
 HIGH VOLATILITY TANK LEAKS (4.18)
          FIGURE 2-1.  RELEASE  SCENARIOS (Continued)




                (Section Number In Parentheses)



                               2-7

-------
    4.  The concentrations  obtained are  representative  of the  averaging
       times  identified.   Procedures  to convert  concentrations  to  other
       averaging times are described in Appendix E.


2.4 Determining Maximum Short-Term Ground Level Concentration

    In modeling air toxic  releases,  a reasonable degree of assurance is needed

that  the  maximum  short-term  ground  level  concentration  estimate  from  a

meteorological  perspective   is  obtained.   This  maximum   concentration  is

selected  from  those  concentrations  calculated  using  the  range of  stability

classes and wind speeds in Table 2-2.

    The first choice is to use  all six stability classes  and their associated

wind speeds.   To reduce  the  number of  calculations,  the second choice  is to

use a  subset  of these meteorological conditions  associated with the emission

release  condition for the  scenario  of  concern.   While  these subsets  will

generally provide  maximum concentration  estimates,  the user may wish  to use

the entire range  of  conditions  in Table 2-2 to  provide greater  assurance that

the maximum concentration is always calculated.


    A. Continuous Passive Releases

       Concentration   estimates   for   continuous   passive   (non-dense)
       releases  should be made  by  using  the  applicable procedures  in
       Table 2-3.

    B. Instantaneous Passive Releases

       Concentration   estimates   for   instantaneous   passive  (non-dense)
       releases should be made for the following conditions:

       I.  Ground level releases:   use stability class  F  and 1  m/s wind
           speed

       II. Elevated  releases:   use  stability  classes  A,  C,  and  F each
           with  1  m/s  wind  speeds.   Calculate concentrations  at the
           greater  of  the distance  at  which  the  vertical  dispersion
           coefficient, az, equals H/V2~ (where H = release height) or the
           fenceline distance.   Select  the maximum of  these  concentration
           estimates.
                                      2-8

-------
                 TABLE 2-2

WIND SPEED AND STABILITY CLASS COMBINATIONS
                             10-m Wind Speed
                                  (m/s)
Stability
Class
A (very unstable)
B (moderately unstable)
C (slightly stable)
D (neutral)
E (slightly stable)
F (moderately stable)
1
*
*
*
*
*
*
2
*
*
*
*
*
*
3 4 5 8 10 15 20
*
* * *
*****
* * * * * * *
* * *
* *
                    2-9

-------
                                  TABLE 2-3

                 CALCULATION PROCEDURES FOR USE WITH VARIOUS
                    EMISSION HEIGHTS (CONTINUOUS RELEASES)
Height of Emission Stability Classes Wind Speed (m/s)*
I. Stack Height > 50 m**
II. 10 m < stack height < 50 m**
III. Stack height < 10 m and some
plume rise
IV. Stack height < 10 m and no
plume rise; all ground
level sources
V. Wake effects exist
Cavity effects exist ***

A
C
A
C
F
C
F
F
C
F
-

1 and 3
1, 3, 5, 8 and 10
1 and 3
1, 3, 5, 8 and 10
1, 3 and 4
1, 3, 5, 8 and 10
1, 3 and 4
1
1, 3, 5, 8 and 10
1, 3 and 4
10 for critical speed
1 for dilution in
concentration equation
  * Use 10  m wind  speed  adjusted  to  stack height  using  the  equation  in
    Section 3.1.1 with the  exponents  as shown  in  Table  3-1.

 ** For these sources,  the user  should determine the  distance to  fenceline,
    Xfc,  and the distance to final plume rise, Xf.  If:

    Xfc > Xf  final  plume   rise   is   used.    (The  distance  to   the   maximum
              concentration is found  at oz  = H/VT".  Where,  H =  release  height,
              and az = vertical plume  dispersion).   For each stability  class,
              the distance  is determined using  a wind  speed  of  1  m/s.   The
              greater  of  this distance or the  distance to  the  fenceline  is
              used in  the estimate.

    Xfc < Xf  transitional   plume   rise  is used,  and  the  calculation,:  nust
              iterate  over  several distances  from Xfc  to Xf, usually at 100 m
              intervals.
***
    Cavity effects are assumed independent of stability classes.
                                     2-10

-------
    C. Dense Gas Releases

       I.  For a continuous  dense  gas release,  the  maximum ground  level
           concentration  usually  occurs  under  light  wind  speed,  stable
           atmospheric  conditions,  but  this  may  vary  depending  upon
           receptor   location   relative    to   initial   dilution   and
           gravitational spreading of the heavy gas cloud  (spreading  is  a
           function of  initial  release  size and density).   Because of the
           complexity of making dense gas  calculations,  hand calculations
           are  impractical.   The  use  of  the  EPA Relief  Valve  Discharge
           (RVD) model is recommended (see  Section 5.6.3).   This  model is
           not dependent  on stability.   After running  the  model  using  a
           range of  wind  speeds,   the  maximum concentration is  estimated
           at  the  farther  of  the  plume  touchdown  distance  and  the
           distance to the fenceline.

       II. For  instantaneous  dense  gas  releases,   a  simple  model  is
           provided  in  Section 5.6.2.   As  a  conservative  approximation,
           elevated  releases  should be  simulated using the ground level
           model.   The  maximum  concentration  release  is  obtained  by
           assuming Gaussian  dispersion and using stability class  F with
           1 m/s wind speed.
    The  user  should  carefully  examine  each  release  scenario  and  use  the

appropriate   approach   from   those   listed   above   to   determine   maximum

concentrations.   There  is  no  universal  approach  to  use  for  all  cases.

However, Table 2-4  summarizes approaches  that  may be  used for  the  scenarios

described in the workbook.
                                     2-11

-------




































•^

o
*— (
2
2
O
2
0
T
' 2^

UJ l-l
—1 X
CO 
rtj U
O C
t- o
Q. < 1-1
< E
-o i
OJ -r- 1-1
T3X < l-i
c re M
O I-H
U < n
Cy
ex.
< 1-1
OJ
0) C
Q OJ
> Q
1—
OJ
VI OJ
re >
OJ .r-
1 — I/I
OJ V)
o£ re
O-
OJ
E
o
>
CD
U ro
!- Ol
0 <
l/l


3;
o
0.























0
t.
re
c
01
u


CD
VI
re
Oi

CD
at



















X








X










X


X

X






X













X X





VI

u
re

in

e
O
S-
4-

VI
CD
VI
re
o>

CD


VI
3
O
CD

re


•a
c
re

CD

re

3
U
+J
i.
re
a.

VI
3
O
C

+J
c
o
.
"—
X


X


X


X






X X






X


X

X X






X XXX










X


X X XX XX

















VI
I/I ***
CD U
a. re

a. i/i

vi vi vi E
3 01 -* O
O VI C i-
re re M-
0) Ol h-
> r— VI
•r- CD E 01
-4-> O£ O I/I
.,- i. re
Ol -i-> H- O)
3 VI r—
U_ 3 VI CD
O •"* ry
4- re
O -C CD VI
*> —13
l/l*i- O
01 2 vi 01
VI 3 VI
re vi o re
Ol CD l/l Oi O
r— 1- C I/I
CD 3 o re vi

• i- VI O
vi re vi vi CD
3 Li. .1- 3 C
o E o re
3 Ol LU 3 4->
C C C C
.,- .,- CD •<- re
-U -U 1- -U -P
c u re c vi
0 3 r- 0 C
(_> a ij- c_) M
.... .
CNJ ro <3" in ^JT>
















X X


















X X





X




X










VI
01
u
L.
3
a


c
o

vi
• r—
g


I/I
3
O
CD
vi
re
o

VI
2
O
3
C

4J
C
o
(_)

CD
>
.JJ
•r-
Ol
3
U_
Ol
a.

4-*

3
.
f-
















X X


















X X










X X







VI
J^
• r>
l+_
-a
c
m
_i
^»
C CD

E t!
*> re
re 2
Ol
J- -o
1— ••-

•o o
c 
re

"re
E a.
o •.-
I- U
4- ..-
C
VI 3
C £
0
••- e
VI O
VI 1-
• r- l+-
E
UJ I/I
c
I/I O
3 •.-
O VI
01 vi
VI •!-
re E
CD UJ

VI I/I
3 3
0 O
3 3
CZ C

1 > [ i
C C
O 0
CJ O
, .
CO O*>
X








X






X











X






X X





X




X X


X











I/I I/I
CD Ol
T3 C

u c
•r- Ol
ja a.
>- o
01
X -i->
c
"O CU
c E
re a.

VI 3
01 0-
-0 LU

'o o
• r- +J
^j
I/I Ol
CU 3
0- O

l+- I/I
o c
o
I/I -f-
C VI
O VI

'vi i
VI LU
>p.
E vi
LU 3
O
vi 01
3 C
o re
3 *>
c c
.•- re
-u -u
C VI
0 C
<_> l-l
.
° ^



X












X X











X






X X










X


X X




VI
c
Ol
E
TO
c
3 — .
o 2
a. o
E i—
1-H LL.

Oi CD
U VI
re re
t- a.
3 1
1/1 O
2
•a i-
OJ —

A I/I
i- CD
cu 01
< i-

L. J=
o u
VI

c a
CD
U CD
VI >
CD i—
•^ re
3 >
o-

E Oi
1- i^
H- Ol
ac
c
O I/I
•r- 3
H-> O
re 3
1. C
O •!-
a. -u
re c
> o
LU O
.
CNJ ro
X















X











X






X













X X











f^.
CU
VI
re
JC
a.
i
2
I—
^_^

VI
Oi
Ol
re
-C
u
I/I
.F>
o

CD
>
r~
re


4-
OJ
.f—
^_
Ol
ae.

VI
3
O
Ol
c
re
c
re
i *
VI
c
1-H
.
2
















X X


















X X










X X
























VI Wl
0) -X
Q- c
UJ
Q. t—
E E
O O
L. La
l«- M-

VI I/I
^C ^
re TJ
Ol CU


T3 -O

3 3
o- o-

_J _l
^i ^>
4-> 4-1
*l— *^
p— ^a—
4J 4J
re re
'o 'o


2 2
O O
—1 _l
. .
LO 1C
















X X


















X X













X X




















VI I/I
Ol _*
0. C
•r- re
0. I—
E E
O O
w. &-
4- 4-

VI VI
-ae Ja£
re re
Qj ll
^J ^J

•a -a
•»— .^
3 3
o- o-
• p. .«~,
-J —1
>) X
4J 4^
.f_ .!_
("• P"1
•r- >r-
re re
"o "o
> >

-C -C
01 ai
X X
. .
— 00





















































































CU
>
• F-
CU I/I
> VI
• r- («
vi a.
VI V.
re vi
a. 3
x o
VI CU
3 C VI
o re re
3 *> Ol
c c
••-re *
4J -U I/I
c v> c
o c cu
u ••- -a
1 l l
< CQ O
«
2-12

-------
3.0 SUPPORT DATA FOR SCREENING ESTIMATES



    Simulations   for   air   toxic   releases   require   information   on   the



meteorological  conditions  at  the  time of  release  as  well  as physical  and



chemical parameters describing the materials being released.






3.1 Meteorological Data



    Computational  procedures  for  estimating  concentrations   require  data  on



wind  speed and  direction,  temperature and  atmospheric  pressure.  These  data



are normally collected at National Weather Service stations  and some  military



installations on an hourly basis.   Stability and turbulence  parameters  can be



estimated from cloud data  as described below.   A record of these is  available



from   the  National   Climatic  Data   Center,   Asheville,    North   Carolina.



Meteorological data are  sometimes  recorded at air quality monitoring  sites at



existing  plants.   The  use  of  the on-site  data with proper  quality  assurance



procedures  as  described  in  On-site  Meteorological   Program  Guidance   for



Regulatory Modeling Applications (EPA,  1987d) is preferred.






    3.1.1 Wind Speed and Direction



    Wind  speed  and direction  data  are required  to  estimate  short-term  peak



concentrations.   Wind speed is  used to determine (1)  plume dilution,  (2) plume



rise  and  (3)  mass transfer in  evaporation models.  These  factors,  in turn,



affect   the   magnitude   of,   and   distance   to,   the   maximum  ground-level



concentration.



    Most wind data are  collected near ground level.   The wind  speed  at release



height can be estimated by using the following power law equation:



                           P
                    /   h  i
             u =
                  * i -
                      zl
                                      3-1

-------
where:
    u  = the wind speed (m/s)  at  release  height  h  (m),
    MI = the wind speed at the anemometer height z^  (m), and
    p  = the stability-related exponent from Table 3-1.
                                  TABLE 3-1

         WIND PROFILE EXPONENT AS A FUNCTION OF ATMOSPHERIC STABILITY

                                                     Rural
         	Stability Class	Exponent	

                          A                           0.07
                          B                           0.07
                          C                           0.10
                          D                           0.15
                          E                           0.35
                          F                           0.55
    The wind  direction is an  approximation  for the direction of  transport  of

the plume.   The variability of  the direction  of  transport over  a period  of

time is a major factor in estimating ground-level  concentrations averaged over

that time period.



    3.1.2 Stability and Turbulence

    Stability categories,  as  depicted in Tables 3-1 and 3-2, are  indicators  of

atmospheric turbulence.  The stability  category at any given time depends upon

thermal  turbulence  (caused  by  heating  of   the  air  at  ground  level)  and

mechanical  turbulence  (a  function  of  wind  speed  and  surface  roughness).

Stability is  generally estimated by a  method  given  by  Turner   (1970),  which

requires  information  on  solar  elevation angle,  cloud  cover,  cloud  ceiling

height, and wind speed (see Table 3-2).

    The solar  elevation angle  is a  function  of the time of year  and  the time

of day, and  is presented  in charts in  the  Smithsonian  Meteorological  Tables

(List, 1968).   The  hourly weather observations of the National Weather Service


                                      3-2

-------
                                    TABLE 3-2

                           KEY TO STABILITY CATEGORIES

Surface Wind
Speed at a
Height of 10m
(m/sec)
< 2
2-3
3-5
5-6
> 6

Day

Incoming Solar Radiation**

Strong

A
A-B
B
C
C
(Insolation)
Moderate

A-B
B
B-C
C-D
D

Slight

B
C
C
D
D
Night*
Thinly Overcast
or
> 4/8 Low
Cloud Cover
_
E
D
D
D


< 3/8
Cloud
Cover
_
F
E
D
D
The neutral class  (D)  should be assumed for all  overcast  conditions during  day
or night.

*   Night is defined as the period from one hour before sunset to one hour  after
    sunrise.

**  Appropriate insolation categories may  be  determined through  the use of  sky
    cover and solar elevation information as follows:
 Sky Cover
Solar Elevation
Angle > 60°
Solar Elevation     Solar Elevation
Angle < 60°         Angle < 35°
But > 35°           But > 15°
 4/8 or Less or
 Any Amount of
 High Thin Clouds
   Strong
  Moderate
Slight
 5/8 to 7/8 Middle
 Clouds (7000 feet to
 16,000 foot base)
   Moderate
  Slight
Slight
 5/8 to 7/8 Low
 Clouds (less than
 7000 foot base)
   Slight
  Slight
Slight
                                      3-3

-------
include  cloud cover,  cloud  ceiling  height,  and  wind  speed.   Methods  for




estimating atmospheric stability categories  from  on-site data are provided in




EPA modeling guidelines.




    Friction velocity  (u^)  represents mechanical turbulence  due to wind flow




over  the  surface  and  is   used  in calculating  release  Richardson  number.




Friction  velocity is  a  function  of stability,  decreasing  with  increasing




stability.   An approximation  of  friction  velocity under  neutral  stability




conditions and assuming a roughness length of 1  cm is:




       u^ = 0.06u




    where u is the wind speed (m/s) at a  height  of 10 m.






    3.1.3 Temperature




    Ambient air temperature must be  known in order  to calculate the amount of




rise of a buoyant plume and  to calculate  evaporation rates.








    3.1.4 Atmospheric Pressure




    Atmospheric pressure data are  used in calculating gas  and  liquid  release




rates from storage and process vessels and pipes.








3.2 Chemical and Physical Parameters



    Numerous   chemical   and  physical   properties   of  chemicals   and  spill




substrates  are required  to use  some of the  emission  estimation  techniques




presented.   A  list  of  typical  physical and  chemical  properties  and  their




typical units as used  in  this workbook is shown  in  Table 3-3.   The  complexity




and  diversity  of  chemical  and  physical  behavior  of the  many  air  toxic




substances make  it critical  that  the correct  input parameters  are obtained.




These  parameters  can  be  found  in  compendiums  of  physical  and  chemical




characteristics.  Three of the more comprehensive sources of information are:





                                      3-4

-------
                          TABLE 3-3

      TYPICAL PHYSICAL AND CHEMICAL PROPERTY PARAMETERS
                  USED IN EMISSION MODELING
        Parameter Name
Typical Units
Boiling point at ambient pressure

Specific heat of liquid
   - at constant pressure
   - at constant volume

Specific heat of vapor
   - at constant pressure
   - at constant volume

Molecular weight

Latent heat of evaporation

Vapor pressure

Vapor density
Liquid density

Specific gravity (S.G.)

Constituent diffusivity
  (diffusion coefficient)
   - in air
   - in water
   - in oil

Henry's Law constant

Solubility in water

Net heating value
cal/g-mole °K
cal/g-mole °K
cal/g-mole °K
cal/g-mole °K

  g/g-mole

 cal/g-mole

     atm
    cm^/s
    cm2/s
    cm2/s
atm-m^/g-mole

  g-mole/m^

 cal/g-mole
                             3-5

-------
    Beilstein,  1987:   Handbook of Organic Chemistry,  Springer-Verlag,  New
    York.

    Green,  D.,  1984:   Perry's  Chemical   Engineer's  Handbook,   Sixth
    Edition, McGraw-Hill, New York.

    Verschueren,  K.,  1983:   Handbook  of  Environmental Data  on  Organic
    Chemicals.  Van Nostrand Reinhold Company,  New York.


The user should be  cautioned that a  characteristic "constant" used  in  modeling

may have  different values  depending on the reference  from  which the parameter

was obtained.
                                      3-6

-------
4.0 SCENARIOS AND TECHNIQUES FOR RELEASE AND EMISSIONS ESTIMATES

    Techniques for estimating air toxics emissions must be  capable  of treating

a  large  variety of  potential  release scenarios.   This section  is  intended to

help the user  identify  the applicable  release  scenario and  determine  release

and  emission rates  and to  guide  the  user  through remaining  calculations  to

arrive at a  concentration  estimate.   Eighteen release  scenarios  are  presented

in this workbook.   Descriptions  of similar releases are provided, and the user

is advised to  review these descriptions if an obvious  choice is not apparent

in Table 2-1.

    Since so many  varied processes  and sources have  the  potential  for  toxic

chemical releases,  the  eighteen  scenarios  cover a small percentage of possible

release, emission,  and dispersion combinations.   Techniques  in this  section

estimate  emissions  to  the  atmosphere  after  providing guidance  on  release

calculations.    In  these   cases    and  in  all   other   applications,   the

characterization of  emissions  is  a  critical  step which is best  met  through a

complete  and  accurate  measurement  program.    In  practical   applications,

measured data are  seldom available and the user  is left to techniques such as

those  presented  in  this  section,  data from existing inventories,  emission

factors, or process specific material balance estimates.

    Some of the  numerous sources of  existing data  are  permit and registration

files, technical  literature, and  SARA Title III reporting forms.  A new data

source  summarizing regulatory  data   is the  National   Air  Toxics  Information

Clearinghouse  (NATICH)  and  Data  Base.   Information  on NATICH is  available

through EPA Regional Air Toxics Contacts and:


                      Pollutant Assessment  Branch (MD-12)
                     U.S.  Environmental Protection  Agency
                       Research Triangle Park, NC  27111
                         (919)  541-0850  FTS  629-0850
                                      4-1

-------
    For  some  sources,  mass  balances  are  used  to  estimate   releases  when




conservative assumptions concerning quantities of  input and output streams are




made.   The  amounts  entering and/or  leaving  a  process  can  be  measured  or




estimated.  A  mass balance  can  then be performed on  the  process as a whole or




on the  subprocess.   For  processes  where material reacts  to  form a product  or




is  significantly  changed,   use  of  mass  balance  may be  too  difficult  for




estimating emissions  and the use of emission factors  may be  more appropriate.




    When measured or plant  specific data are unavailable, the  user is advised




to  review  emission  factors  developed  for specific processes.   Appendix  A




provides a description of sources of emission factors.
                                      4-2

-------
4.1 Continuous Particulate and Gaseous Releases from Stacks




    Similar  Releases;   Continuous emissions  of particulate  matter and  gases




from  building  vents,  vertical  stacks and pipes, or  conventional  point sources




when  emission  flowrates  and  temperature  are  known.   Combustion  sources,




chemical  reactors,  and some process vents are typical emission  sources that




emit  pollutants  through  stacks.   These releases may also  be  due  to a process




failure such as a rupture disk release or failure of control equipment.








    4.1.1 Particulate Matter




    Continuous  emissions of  particulate  from stacks  are analyzed  beginning




with  Flowchart C-l.   EPA recommends  that emission rates  from such  sources  be




determined through  source  testing using EPA Reference  Methods  (40  CFR Part 60




Appendix  A)  or  process  calculations.   If source-specific  emissions  are  not




available,  representative  emission  factors  can  be  substituted.   Emission




factors are available  for  individual toxic compounds (Appendix A,  items  1,  2,




and  3).    Otherwise,   factors  determined  by  compiling  extensive  source test




results using  EPA  Reference  Methods  are  reported  in AP-42.  Toxic components




of  emissions  can be  determined  using the Air  Emission Species Manual,  Volume




II, Particulate Matter Species Profiles (Appendix A, item 4).




    Once  emissions  of  the  toxic  pollutants  are  determined,  this  release




scenario  represents a  case  for which ground  level  concentration  estimates can



be  made   using  specific  dispersion  calculations  outlined  in   Flowchart  D-2




(Appendix D).   Specific equations to be applied  can be  found  in the  report




sections  referenced in each flowchart.  For  point  sources,  calculations  begin




with  determining if  cavity or  wake  analysis is  applicable as  outlined  in




Flowchart D-2.   If  the plume is in  the  cavity or the wake region, then cavity




and wake impacts must be determined.   If the plume is outside of  the  cavity or




wake  regions,  then basic  point source  techniques are applied  as  shown  in




                                      4-3

-------
Flowchart D-3.   Section  6.1  provides  an example  of  this scenario.   Maximum




concentrations are obtained as shown in Table 2-3.








    4.1.2 Gases




    Continuous  emissions  of  gases  from  stacks are  analyzed  beginning  with




Flowchart C-5.  Emission  factors  are available for individual  toxic compounds




(Appendix A,  items  1,  2,  and 3).   Otherwise,  total VOC emission rates  can  be




obtained from AP-42  in a  similar  manner as  discussed  in Section 4.1.1.  Toxic




components of these emissions can  be determined using the  Air  Emission Species




Manual, Volume I,  Volatile Organic Compound Species Profiles (Appendix A).




    Dispersion techniques  for continuous gaseous emissions  (Flowchart  D-l) use




similar techniques  to  modeling  particulate emissions  except that cloud density




calculations are used to determine if dense gas effects may be  applicable.  If




the gas is not dense, passive point source techniques  apply (Flowchart D-2).
                                      4-4

-------
4.2 Continuous Releases of Fugitive Dust




    Similar Releases;   Any fugitive dust  from process  losses,  generation  by




mechanical action in material handling  or windblown dust.








    These fugitive  dust releases  are  generalized  area  emissions  originating




from a  surface or  collection  of  small,  poorly quantified point  sources.   As




indicated in  Flowchart  C-2,  emissions  are either user-specified or  calculated




with  representative  emission  factors.   Emission  factors for  fugitive  dust




emissions are  typically found  in AP-42  and  are assumed  to be independent  of




wind speed  for this  workbook.   Toxic  components  can be  determined  using the




Air Emission  Species Manual,  Volume  II, Particulate  Matter  Species  Profiles




(Appendix A).   Dispersion calculations  for a  continuous  area source  release




are outlined in Flowchart D-4.   For screening,  particle  settling is  assumed to




be  insignificant.    Virtual   distances   are  determined  for  calculation  of




dispersion parameters,  and concentrations  are then  calculated for the  area




source.   Maximum concentration is obtained as shown in Table 2-4.
                                      4-5

-------
4.3 Ducting Failures With Dust Releases




    Similar  Releases:   Instantaneous bursts  of  dust  particles  due  to  duct




failure   (e.g.,   pneumatic  conveyor   line   failures),   line   disconnection,




isolation joint failure,  or other types  of equipment openings.









    Flowchart  C-3  indicates  that  emission  estimation  techniques  are  not




available  for  duct  failures  and  user  specification  is  required.   Limited




information  on  powder  releases  is  available  in  the  technical  literature.




Crude estimates  of  release  amounts can be  made based on  transfer  line  rates




and  time  for   equipment   shutdown  and  equipment  capacity.    Modeling  for




dispersion is shown in Flowchart D-6.  If possible, the user should  attempt to




estimate  the  initial  cloud  dimension  resulting  from  dilution due  to  the




mechanical action of the release.  If initial dilution can not be estimated by




the  user,  conservative  concentration  estimates  can  be  obtained  using  an




instantaneous point  source simulation  (also  Flowchart  D-6).   For  screening,




particle settling is  assumed insignificant.   Maximum concentration is obtained




as shown in Table 2-4.
                                      4-6

-------
4.4 Flare Emissions




    Flares are  used as  a  control device  for a variety of  sources.   As such,




flares  must  comply  with  requirements  specified  in  40  CFR  60.18.    Once




emissions  are  vented  through the  flare,  a  minimum  98%  reduction  of  all




components  of  the  flare  must  be  achieved.    Therefore,   the  user  should




calculate  the  process  emissions  that feed  into a  flare  and multiply  this




number by  0.02  to achieve  a  conservative estimate  of emissions  emitted  from




the flare.   After emissions  are  determined  (Flowchart  C-4),  the user  should




calculate  total plume rise  from flares (Flowchart D-7).   Flame tip  height  is




calculated and  added  to the physical stack height to  account  for the distance




between  the  flare  outlet   and the  flame  tip.    Total  heat release  rate  is




calculated prior  to  use in  buoyant plume  rise equations.   Continuous point




source dispersion techniques are then used to determine  dispersion parameters,




buoyancy induced dispersion, and receptor concentrations (Flowchart D-3).
                                      4-7

-------
4.5 Continuous Gaseous Leaks  from Tanks/Pipes

    Similar Releases;  Continuous  gaseous emissions  due  to visible  (usually)

holes or  openings  in tanks,  pipes,  or  flanges  (e.g.,  at pipe  connections,

valves,  pumps, and compressors).



    Emissions  due  to continuous  gaseous leaks  from tanks  or  pipes can  be

estimated using the following procedures,  as  outlined in Flowchart C-6:


    •  Input:

       Pa  -  atmospheric pressure (dynes/cm^)
       ?t  -  tank or pipe pressure (dynes/cm^)
       A   -  hole or pipe area (cm^)
       MW  -  molecular weight (g/g-mole)
       Tt  -  tank absolute temperature  (deg.  K)
       K   -  ratio  of specific  heat at constant  pressure  to  specific
              heat at constant volume
       pv  -  vapor (gas)  density (g/cm^)
       Xj  -  mole fraction of each constituent  in vapor
       YI  -  weight fraction of each constituent in vapor


    •  Limitations/Assumptions:

       -  Does not simulate time dependent release rates

       -  C
-------
    Mean specific heats  (calculate  for both constant  pressure,  cp,
    and constant volume, cv)
                n
       cmean
    The ratio of specific heats is:

       K =   C
             cv mean
    A typical value of K at atmospheric pressure is 1.5.


 2)  Determine  if   the  maximum  (critical)  release  rate   is  to  be
    calculated by  evaluating  the  pressure  ratio  at  the  release
    point:

    If:
                                    use subcritical  rate equation
                                    use  critical  rate  equation
                   £t

   A  typical  value for  the  right hand side  is  2.0.


 3) For  critical  flows,  calculate:

   Release  rate  (g/s) .
                                   /K+l

                     V D. D   / 1  \
      <3v =  Cd  A
K Pt Pv / 2
                             \K+1,

4) For subcritical flows:

   Release rate (g/s)
      qv =  Cd  A  / KPt Pv   '  2
Output',

Vapor venting rate (qv) in g/s for use in dispersion models.

                               4-9

-------
    After  emissions  are  determined,  point  source  dispersion  techniques  are



applied  (Flowchart  D-l),  including  the determination  of plume  density  and



dense gas concentrations,  if applicable.   For non-dense gas releases,  possible



cavity or  wake  effects  are  examined (Flowchart D-2).   Note,  if the user cannot



obtain reliable  data to  use  in these  equations,  then  the  use  of  emission



factors (Appendix A)  is  suggested as an alternative.
                                      4-10

-------
4.6 Instantaneous Gaseous Releases from Stacks




    Similar  Releases:   Instantaneous gaseous  vent  releases or  gas  leaks  and




relief  valve or  rupture disk  discharges  which are  of short duration (e.g.,




less than 5  minutes).   These  releases may arise from  process  upsets, chemical




reactor process failures or equipment opening or purges.








    Screening methods are  not available for  estimating emission  rates,  plume




rise and  downwash effects  for  this  release  type  (Flowchart  C-7).   Emissions




estimates  for  this  scenario are  generally  process  specific  and  must  be




specified by the user.  Limiting estimates  of emissions can  be determined by




considering  the  capacity of the source under  consideration.   For example,  the




maximum amount  of  gas  released  from a reactor would be the reactant or product




amount.




    Dispersion  estimates can be  obtained  by  applying  the  procedures outlined




in Flowchart D-6.   If  the  release  is  not dense  (based  on Richardson number




criteria),  instantaneous point  source  dispersion  techniques  are applied to




obtain  concentration  estimates.   If,  however,   the  instantaneous  puff  is




determined to be dense, concentrations are determined  after  including initial




gravitational spreading.  A coarse screening estimate for the  effects of dense




instantaneous gaseous releases  can be performed assuming that the releases are




undiluted and at ground level as indicated in Flowchart D-6.




    Screening  techniques  are  available  for  simulating   only  spreading  and




dispersion  from a  low momentum,  ground-level  dense  release.   Descent  of




elevated heavy  gas clouds  and high momentum associated with many instantaneous




releases tend to provide significant initial dilution.  This  dilution acts to




reduce concentrations and  density.   As a conservative estimate,  elevated heavy




gas dispersion from instantaneous releases is assumed to be at ground level.
                                     4-11

-------
4.7 Multiple Fugitive Continuous Gaseous Emission Sources




    Similar  Releases;   Releases  from  any continuous  area or  volume  source




where  the  emissions  are  released  uniformly  over  the  area   or  the  area




represents  a  collection  of  small  sources   poorly quantified  in  terms  of




location (e.g., multiple  vents  on large manufacturing buildings, fugitive  VOC




sources in refineries or chemical process manufacturing plants).








    Fugitive gaseous emissions resulting from collections of small  sources and




gaseous  area source  emissions  of  different  types  (e.g., process equipment,




valves etc.) are modeled  using  techniques  shown in Flowchart C-8.  The  use of




EPA  fugitive emission  factors  for  selected  equipment  are found  in the EPA




report  Fugitive Emission Sources  of  Organic  Compounds  (Appendix  A).   For




selected air toxics, fugitive  factors  are also found in  Appendix A (items 1




and  3).   Often, areas  of fugitive  emissions  can be  specified for elevated




releases  such  as  manufacturing facilities where  substantial  numbers of hood




and  vent sources are found on  the  roof and  fugitive emissions  identified in




mass  balances  are  suspected from  ventilation  sources.   In these cases, the




area of  release can be considered  as a volume source using  a  characteristic




height  such  as  a building height.  Dispersion calculations for continuous area




and  volume   source  releases  are  outlined   in  Flowcharts   D-4  and  D-5,




respectively, with  relevant  calculations referenced by report section for each




procedure.   Receptor concentrations are  estimated after determining  horizontal




virtual distances and corresponding dispersion parameters.
                                      4-12

-------
4.8 Continuous Gaseous Emissions from Land Treatment

    Similar Releases:   Landfarms;  application of a volatile material to soil.



    Land  treatment emissions  are  modeled  using  the  techniques  outlined  in

release  Flowchart  C-9.   The  emissions  equation  is  a  simplification of  the

Thibodeaux-Hwang  Emission  Model,  assuming  ground-level  application  of  the

waste,  more  rapid   diffusion   through  the   oil  layer,  and   vapor-liquid

equilibrium between the oil layer and pore spaces.


    •  Input:

       D     -  diffusivity of organic component in air (cmr/s)
       A     -  land treatment surface area (cm^)
       hp    -  depth of soil penetration by waste sludge (tilling depth)
                (cm)
       t     -  elapsed time since waste application (s)
       P     -  vapor pressure of the constituent (atm)
       MWoii -  average molecular weight of the sludge (g/g-mole)
       M     -  total oil application rate (g/cm^)
       ppm   -  grams of organic component per million grams of waste  oil
                (g/106g)
       R     -  gas constant (82.06 cm^ atm/g-mole • K)
       T     -  gas temperature (K)


    •  Limitations/Assumptions:

    -  Waste is a sludge consisting of organics in oil.

    -  Methods  are a  simplification  of  the  Thibodeaux-Hwang  Emission
       Model (Thibodeaux and Hwang, 1982).

    -  Assumes  no  subsurface  injection,  slower  diffusion  of   organic
       component  through  air-filled pore  spaces  than  through  the  oil
       layer, and  vapor-liquid equilibrium  between the air  in the  pore
       spaces and the oil layer.

    -  Assumes that Raoult's Law applies.

    -  Effective diffusivity  can  be  assumed to be 40%  of pure component
       diffusivity.


    •  Procedure:

       Determine the  average  emission  rate  over  the  entire area, E,  in
       g/s:

                                     4-13

-------
                           0.5
            . DP MWoil M  \                     ,
      E  =  (             ) '  ppm •  A •  2 x 1CT6
            I  5 hpRT t


Output:

   Emission rate, E (g/s) from a land treatment site

-  Land  treatment  area  for use  in  determining  virtual  distances
   for dispersion

Dispersion  of land  treatment  emissions  is  simulated as  an area
source (Flowchart D-4).
                               4-14

-------
4.9 Continuous Emissions from Municipal Solid Waste Landfills

    Gaseous emissions  from municipal  solid waste landfills may be greater than

those from  properly maintained,  capped  hazardous waste  landfills.   Therefore

this  section  presents  how  gaseous  emissions  from  municipal  solid  waste

landfills may  be estimated.   This information  is  from  the draft  background

document  for  proposal of  air regulations for municipal  solid  waste landfills

(EPA, 1988a).   This document  explains how  emissions  can  be  estimated  using

either  (1)  an emission factor based  on  the  amount of refuse in  a landfill or

(2)  sampling  data  (e.g.,  field  measurements  of  the  gas  flow  rate  and

composition).   The  emission factor is based on measuring the amount of VOC per

ton of  landfilled  waste using  data provided by  California's South  Coast  Air

Quality  Management  District.    The  total  VOC  emissions  determined by  this

procedure can be  speciated  using a  profile  from  the  Air Emission  Species

Manual,   Volume  I,   Volatile  Organic  Compound  Species  Profiles  (Appendix  A,

item 4).  There  are a  number of  factors  contributing to  the  variability  in

gaseous   emissions   from   municipal   solid waste  landfills   (e.g.,   waste

composition,  landfill  moisture  content,  age of  refuse,  pH and  alkalinity of

landfill,  amount   of   buried  waste,  climate,   and  physical   and  operating

characteristics  of  landfill).   The greatest sources  of  uncertainty are  the

type  and amount of  waste buried  in a  landfill.  Use  of  sampling data  is

strongly  recommended (as  described in the  EPA  background  document  for  draft

proposed  regulations for  municipal solid waste  landfills).  However,  the  use

of an emission factor is  considered  appropriate as  a simple  screening  tool,

and this approach is described below,  as outlined in Flowchart C-10.


    •  Input

       M -  amount of refuse in place in a landfill (millions of tons)

         -  either  the average  annual precipitation at the  landfill site
            or the state in which the landfill is located


                                      4-15

-------
    •  Limitations/Assumptions:

       -  Applicable to municipal solid waste landfills.

       -  Provides an  average  VOC emission  rate.   (To obtain  the  amount
          of  individual toxic  constituents,  the  concentration of  the
          individual  toxic  constituents   is  needed.   The   background
          document  for  the  proposed  regulations  provides  the  range  in
          concentration of toxic constituents  that has been measured from
          landfills nationwide.)

       -  Emission rates are assumed  to be steady-state,  with  no seasonal
          or diurnal variation.  However,  the  effect  of  precipitation  on
          emission rate  is accounted for  using an empirical  correlation
          based on measured  data  for  20 landfills.  (Refer to  background
          document   for   draft   proposed    regulations  for  further
          information.)


    •  Procedure

       1) If  the  landfill  site   averages   less  than   23    inches   of
          precipitation per  year (or,  in  the absence  of local  data,  if
          the landfill  is  located  in  the  States  of AZ,  CA,  CO, HA,  ID,
          MT, NV, MM,  ND,  SD,  UT,  or  WY),  then use the following equation
          to determine the emission rate, E (g/sec):

             E  =  (0.4 g/s/million tons) M


       2) If  the   landfill   site   averages   23   inches  or  more   of
          precipitation per  year (or,  in  the absence  of local  data,  if
          the landfill is not in one of the states noted  above),  then use
          the  following  equation  to  determine  the  emission  rate,  E
          (g/sec):

             E  =  (1.0 g/s/million tons) M


    •  Output

       E = Average VOC emission rate (g/s)


    Dispersion of  emissions  from a  landfill  is  simulated  as  an area  source

(Flowchart  D-4),  involving  determination  of  dispersion parameters based  on

virtual  distances  before  concentrations  can  be  calculated at  each  receptor

location.
                                     4-16

-------
4,10 Continuous Emissions of Pesticides and Herbicides




     Emissions  resulting from  the  volatilization of  applied  pesticides  or




herbicides  are modeled  using  area  source  techniques.   Generally,  screening




level methods  are not  available  for  release  estimates;  therefore,  emission




rates must  be  user-specified  as  indicated in  emission  Flowchart  C-ll.   The




best sources  of  information  are  technical  literature searches  and  contacts




with agricultural research  stations.   Area  source  dispersion techniques,  as




outlined in Flowchart D-4, are used.   These involve determining virtual  source




distances,  dispersion   parameters,   and  estimated  concentrations  at  each




receptor location.
                                     4-17

-------
4.11 Instantaneous Emissions  Due to Equipment  Openings




      Similar  Releases;  Any puff or  burst type  release  with short  duration




emissions resulting from the  opening of equipment after processing  (e.g.,  coke




ovens  or  chemical   reactors),   from  routine  sampling  of  product  during




processing or gaseous emissions from disconnected lines.








     Sourjes of  this  type  are modeled as instantaneous point or  volume sources




of  gaseous  emissions due  to the  momentum of  their  release.   Emissions  can




either be estimated  on a source-specific basis by the user  or calculated from




representative  emission  factors,  as  shown  in  Flowchart   C-12.    Emissions




estimates are  available in  AP-42 for some batch operations. VOC profiles are




also  available to  identify   toxic  components  (Appendix  A,   item   4).   Simple




estimates of  emissions from  failed  or disconnected transfer  lines or similar




sources  can  be calculated from the gas  volume  between  the  break point  and




nearest shutoff valve.




    Receptor concentrations  are  calculated  as  outlined in  Flowchart  D-6.   A




density  check  determines whether  the  cloud is negatively buoyant  or passive.




If passive, dispersion is simulated as a volume source with  initial dimensions




dependent  on  the circumstances  of  release.   If volume  dimensions  are  not




known,  conservative  concentration  estimates can be  obtained by  assuming  a




point source release.




    If   the  cloud  is  negatively  buoyant,  dense   gas  concentrations  are




calculated.  Screening techniques are available  for  simulating  only  spreading




and dispersion from a low momentum, ground-level  release.  Descent  of  elevated




heavy gas  clouds  and high momentum associated with many instantaneous  releases




tend  to provide  significant  initial dilution.   This  dilution acts  to  reduce




concentrations and density.   A conservative screening estimate for the  effects
                                      4-18

-------
of  heavy  gas,  instantaneous  releases  can  be performed  assuming  that  the




releases are undiluted (i.e., point  sources)  and at ground level  as  indicated




in Flowchart D-6.
                                     4-19

-------
4.12 Evaporation from Quiescent or Aerated Surface  Impoundments

     Similar  Releases;   Waste  lagoons  and other  impoundments  with  emissions

resulting from the evaporation of volatile chemicals  from liquid mixtures.



     Emission rates  from  well-mixed aqueous waste  in surface impoundments  are

modeled  using  techniques  outlined  in Flowchart  C-13.  Branching  occurs  to

provide  techniques  for  both  quiescent  and  aerated  impoundments.    Emission

estimates  account for  volatilization  solely,  with other  removal  mechanisms

assumed to be negligible.   Inputs, assumptions, and  calculation procedures  for

emission estimates are summarized below.


    •  Input:

       Co   -  initial concentration of the chemical  in the  waste (g/m^)
       S    -  chemical solubility in water (g-mole/m^)
       Vp   -  pure component vapor pressure (atm)
       H    -  Henry's Law Constant (atm-m^/g-mole)
       A    -  area of impoundment (m^)
       F    -  fetch (linear distance across the impoundment)  (m)
       D    -  depth of waste in impoundment (m)
       Q    -  volumetric flow rate of the waste (m^/s)
       t    -  time after disposal (for impoundments  with no outlet flow)
       POWR -  aerator power (horsepower x number of  aerators)
       f    -  fraction of impoundment aerated
       Mi   -  number of aerators


    •  Limitations/Assumptions:

       -  Equations  are  simplifications  of  methods in EPA,  1987a  for
          quiescent  surface  impoundments  with and  without flow  and for
          aeration basins.

       -  Simplified  by  assuming  a  wind  speed  of 5 m/s,  constituent
          diffusivity in water of 10~^  cm2/s,  and  constituent diffusivity
          in air of 0.10 cnr/s.

       -  Assumes waste is well mixed in impoundment.

       -  Assumes removal  entirely  by volatilization,  with no loss due to
          biodegradation, seepage, or adsorption.

       -  Assumes waste is aqueous, with no separate organic phase.
                                     4-20

-------
•  Reference:

   -  EPA (1987a)



•  Procedure:

   1)  Calculate the equilibrium constant, K6g, using



               Keq = 40.9H

      If H is  not  available,  then it  can be approximated as:


               H=    VP
   2) Determine the gas-phase  mass transfer coefficient,  kg,  in m/s
      using

         kg = 1.26 x 1CT2  A-°-055



   3) Calculate the  liquid phase  mass transfer  coefficient,  k]_, in
      m/s using the appropriate equation

      i)    if F/D < 14.0

            ki = 2.92  x 10~6

      ii)   if 14.0 <  F/D  < 51.2

            kx = 6.84  x 10~8  (F/D)  +  3.35  x  10~6

      iii)  if F/D > 51.2

            k  = 6.85  x 10~6
   4) Determine the  overall mass  transfer  coefficient,  Kq,  in m/s
      using:


         *,.t  1    +    *     Vl
                1         g   eq

   5) Determine  the   equilibrium   or  bulk   concentration   in  the
      impoundment,  CT,  in g/m^  using
         CL =
                 QCO
               Kg A + Q

      For aerated impoundments,  skip to step  7.


                                 4-21

-------
       6)  Calculate  the  source  emission rate,  E,  from  the impoundment
          (g/s)
              E  =
Kg CL
          For disposal  impoundments with  no  outlet  flow,  use the  same
          steps  1 through 4 above  and the following  equation:
             E =  AD co  [i - exp <-Kgt/D)]
       7)  Steps 7 through 11 are  for aerated impoundments only.  For  the
          aerated impoundment  calculate the  turbulent liquid-phase mass
          transfer coefficient,  k}a:

             kla  = 0.2623 (POWR / A  f)


       8)  Calculate the  turbulent gas-phase  mass  transfer  coefficient,


             kga  = 0.021 (POWR /Hi  )°'4


       9)  Calculate the overall  turbulent mass transfer coefficient,  K^:
                        kla       Keq kg<

      10) Determine  the mass  transfer  coefficient  resulting  from  the
          quiescent and turbulent components:

             K  =  Kt f  +  (1-f) Kq


      11) Emissions are obtained by calculating:

             E  =  K CL A

    •  Output

       -  source emission rate, E, from the impoundment (g/s)

       -  area  of  impoundment,  A  (m^),  for use  in  virtual  distance
          calculations


    Dispersion from  a surface  impoundment is simulated  as a  continuous area

source  with  initial  dimensions  equal  to  those  of  the  impoundment  (see
                                     4-22

-------
Flowchart D-4).  Virtual  distances  are used to determine dispersion parameters



which  are  input  to  the  continuous  point  source  dispersion  equation  to




determine receptor concentrations.
                                     4-23

-------
4.13 Continuous Relief Valve Discharges (Two-Phase Flow)

     Similar Releases;  High  pressure  superheated liquid releases,  continuous

two-phase releases, two-phase releases  from stacks.



     A  pressure  relief event  is considered  to  be  the  venting  of  process

equipment  (i.e.,   reactor   vessels,  columns,  storage  tanks)  through  rupture

disks,  safety relief valves, manual,  or emergency vents.  An event  may  consist

of one pressure relief  device discharge or a  series  of  pressure relief device

discharges  stemming  from   the  same  set  of   circumstances.   Pressure  relief

emissions  can  be  either  liquids  or  gases  or  some  combination  of the  two.

Other two-phase releases can  result  from tank leaks  (see Section  4.18).   Pure

vapor continuous  discharges  are simulated  using  techniques  in  Section  4.5.

Parameters for relief valve discharges  are specified  from plant  designs.  Steps

in calculating two-phase releases are as follows (Flowchart C-14):


    1)  Calculate the fraction of liquid flash vaporized:

       •  Applicability:

          Calculation of  the fraction  of liquid  flash  vaporized  in the
          depressurization of a pressurized liquefied gas.


       •  Inputs

          Ts -  storage or  line temperature of liquid (deg. K)

          TJ-J -  boiling temperature at ambient pressure (deg. K)

          Cp -  specific heat at constant pressure  (erg/(g deg. K))

          L  -  latent heat of vaporization (erg/g)


       •  Limitations/Assumptions:

          -  The method does not allow for aerosol  "rain-out".
                                      4-24

-------
   •  References:

      -  World Bank (1985)

      -  Wallis (1969)


   •  Procedure:

      The fraction of liquid flashed to vapor is given by:


            F = CP  '^L^

   •  Output:

      Fraction  of  a  release  which  is  vaporized  to  be  used  in
      calculations of cloud density


2) Calculate the mean density of the mixture:


   •  Applicability:

      Cloud density of a release containing a liquid aerosol.


   •  Input:

      P! -  liquid density (g/cm^)

      Pv -  vapor density (g/cm^)

      F  -  mass fraction of vapor


   •  Limitations/Assumptions:

      -  The method assumes suspension of liquid droplets.


   •  Reference:

      -  Wallis (1969)


   •  Procedure:

      Calculate the mean mixture density by:


              Pm  =            l
                       (F/PV)
                                 4-25

-------
   •  Output:

          Density of a vapor/liquid aerosol cloud (g/cm^)


3) Calculate the two-phase outflow:


   •  Applicability:

      Two-phase  (vapor/liquid  aerosol)  release  rates  resulting  from
      major  leaks   in   the   vapor  space  of  a  containment  for  a
      pressurized superheated liquid.


   •  Inputs:

      A   -   Release area (cm^)
      Pm  -   mixture density (g/cm^)
      PI  -   line or storage pressure (dynes/cm^)


   •  Limitations/Assumptions:

          Assumes homogeneous flow with components in equilibrium.

          Pressure   is   assumed   to  drop  to   atmospheric   pressure
          immediately on release.  This  is  conservative for  higher
          pressure releases.

          C$,  the coefficient of discharge, is assumed to be 0.8.


   •  References:

      -   World Bank (1985)

          Hunsaker and Rightmire (1947)


   •  Procedure:

      Calculate the discharge rate: (g/s)
qm  =  0.76 A  Vp
                                    m
   •  Output :

      Total discharge rate, g^, in g/s for the liquid/vapor mixture
                                  4-26

-------
    Screening estimates can  be obtained using Flowchart D-l assuming  that the




density used  is  that of the mixture.   The  mass  emission rate is equivalent to




the total  release  rate  from the valve  or  tank.   Density checks  are  performed




to  determine  whether  the resultant  plumes may  be affected  by gravitational




spreading.  The  dispersion of  the  resultant cloud  assumes  no  change  of  state




from the  initial value.  Continuous  two-phase release events are  some  of the




most  frequent  scenarios  in  which  dense  gas  effects   may  be   indicated.




Estimates of  plume  density  will determine if  these  effects are  important.  If




the plume  is  dense,  worst  case estimates  are made using  the  RVD model  for




which  a  range of  wind speeds  should be  input.   If  the plume   is not  dense,




point source  techniques are  used to  simulate  dispersion (Flowchart D-2).   If




the plume  is  in  the cavity,  different procedures may be necessary depending on




receptor location.   If  the  plume is not  in the  cavity  or  wake, normal  point




source techniques apply.




    Initial dilution of  high  pressure  releases  should   be  considered,  but




techniques are not  currently available.  In addition, neither fallout nor the




evaporation of droplets is not included.
                                     4-27

-------
4.14 Instantaneous Relief Valve Discharges (Two-Phase)




     Similar  Releases;    This  class  represents  any  instantaneous  two-phase




pressure relief event from relief valves or pressurized tank lines  or vessels.








     Relief valve or  rupture  disk discharges and other two-phased  releases are




generally  site  specific  and  emissions  must  be specified  as  indicated  in




Flowchart  C-15.   Atmospheric  dispersion  screening  estimates  are  made  using




instantaneous dispersion modeling techniques.




    Dispersion  modeling  follows  a  determination  of whether the  cloud  is




affected by negatively buoyant  forces as shown in Flowchart  D-6.   The density




determination  is  made  based  on  the  mean  cloud  density  (see  equation  in




Section 4.13), and subsequent simulations of dispersion are based on the total




cloud mass.   If the  puff is  not dense,  instantaneous  point  source techniques




apply.
                                      4-28

-------
4.15 Low Volatility Liquid Leaks from Pipes

       Similar   Releases;    Continuous   low   volatility   liquid  leaks   from

connectors,  flanges,  transfer  lines  or  pumps;  finite  liquid  releases  from

disconnected transfer lines.



    Liquid  releases  can  occur  in  the  form  of finite  spills  or  continuous

leaks.  For  these  screening  calculations,  it  is assumed that finite  spills are

small and will  form a pool with  a 1 cm depth.   A steady-state  evaporation  rate

is   estimated  and  is   conservatively   assumed  to  persist   throughout   the

dispersion   averaging   period.    For  continuous   releases,  Flowchart   C-16

indicates  that  if  liquid leak  rates are  unknown,  the  leak  rate  equals  the

maximum flow rate in the pipe.   The released liquid is then assumed  to  pool  in

an  area  which is  the lesser of  the unbounded  pool  spread area,  impoundment

boundary  or the   area  over  which  the   liquid  spreads  before  reaching   the

impoundment boundary  (i.e.,  pools  not contained in the impoundment).  Finally,

the   steady   state  emission   rate   is   calculated.    Equations    for   these

calculations are provided below.


    •  Input:

       MW  -  molecular  weight (g/g-mole)
       Ac  -  area for confined  releases  (m^)
       Tp  -  pool surface  temperature  (deg.  K) assumed  equal  to  ambient
              temperature
       u   -  wind speed (m/s)
       qi  -  liquid release rate (g/s)
       P   -  vapor pressure (dynes/cm^)  at surface temperature Tp
       V   -  finite liquid release volume (m^)


    •  Limitations/Assumptions:

          The  model  is  steady  state  and  applicable  to  single  phase
          releases.

       -  Phase change of superheated liquids  is not considered.

       -  Evaporation begins after pool  formation.

                                      4-29

-------
   -  Minimum pool  depth is  1  cm.

   -  Pool spreading  reaches  a steady-state  when the  liquid  release
      rate equals  the evaporation  rate.
•  References:

   -  NOAA (1988)


•  Procedure:

   1) Calculate the  intermediate  parameter B  consolidating terms  of
      the model:

                              A    u°-78  MWO-67 p
              B  =  1.54 x 1(T4
                                         TP

   2) Estimate liquid  release  amount or  rate.   For finite  releases
      use  a   known  volume  or  estimate the  released  volume as  the
      volume   of  the  disconnected  or   failed  transfer  line.    For
      continuous liquid  leaks,  assume  that  the liquid  release  rate,
      qj, is  equal  to the mass  flow rate in the pipe or line.
   3) Calculate the area of the pool.   For finite releases:

              A = 100 V

      where:   V = liquid release volume (nH)
      Proceed to step 4.

      For continuous  releases,  the area  of the  evaporating pool  is
      calculated as  the smaller of  the impoundment area or the area
      at which evaporation across the pool equals flow into the pool.
              A = min.
   4) The steady state emission rate qv (g/s) is given by:

              qv  =BAO-94


•  Output:

   -  Steady state emission rate (g/s)

   -  Pool  area (m^)

                                 4-30

-------
    Dispersion  from  the  pooled  liquid  is  simulated  using  continuous  area




source  techniques  (Flowchart  D-4).   Low  volatility liquids  are expected  to




pool  in  a ground  level  area  source  from which  emissions are  generated  by




evaporation.    Pool  dimensions  are  used  to  determine virtual  distances  and




dispersion parameters  for area  source modeling  of passive  releases.
                                     4-31

-------
4.16 Low Volatility Liquid Leaks from Tanks

      Similar  Releases:    Low  volatility  liquid  leaks  from  containment  or

reactor vessels (i.e., leak below the liquid level).



     As indicated  in  release Flowchart C-17, liquid  release  rates  from a tank

are computed  and used as  input to  the  evaporation  model,  where pool  spread

area  and  a   steady  state  emission  rate  are  estimated.   Equations  for  the

calculation are given below.



    1) Calculate Liquid Release Rate

       •  Input:

          P! -  liquid density (g/cm^)
          A  -  hole or puncture area (cm^)
          Pa -  atmospheric pressure (dynes/cm^)
          Pt -  tank pressure (dynes/cm^)
          H  -  height of the liquid column above the hole (cm)
          C<3 -  coefficient of discharge


       •  Limitations/Assumptions:

          -  Does  not simulate  time dependent  release  rates for  tanks
             with decreasing pressure.

          -  The coefficient of discharge varies between 0.6  and  1.0  as a
             function  of  release  geometry  and  Reynolds  number.   For
             screening purposes, assume C^ = 0.8.


       •  References:

          -  Environmental Protection Service (1985)

          -  Hunsaker and Rightmire (1947)


       •  Procedure:

          Calculate the liquid flow rate by:
                                      4-32

-------
   Flow rate (g/s)
= 0.8 A P    1960
                               H +  2 (Pt ~ Pa>
2) Proceed to apply  the  Evaporation Model from Section 4.15 for
   continuous releases beginning  in step 1 using the  value  for
   q  derived above.
                           4-33

-------
4.17 High Volatility Liquid Leaks  from Pipes




     Small  pipe leaks of  highly  volatile  liquids  are assumed,  for  screening




purposes, to boil off instantaneously, resulting in a plume which is  simulated




with  a  continuous  point  source   dispersion model.   Emission  rate  (g/s)  is




assumed to equal the pipe flow rate,  as indicated in release Flowchart C-18.




    Dispersion  of  the  plume  is   simulated  with a  continuous  point  source




dispersion model (Flowchart D-2).   For screening estimates, it  is assumed that




the  leak is in the cavity  zone and  plume rise calculations are  not  required.




In addition, due  to the  small leak size,  it  is assumed  that  the  release  has




low momentum and  is passive rather than negatively buoyant  (i.e., not a dense




gas).
                                      4-34

-------
4.18 High Volatility Liquid Leaks from Tanks

      Small  tank leaks  of  highly  volatile   liquids  are  simulated using  an

equation  to  calculate  liquid  release  rates  and an  assumption  of  instant

evaporation,  resulting  in  an emission  rate   equal  to  the release  rate  (see

release Flowchart C-19).  The emission rate  calculation procedures  are  given

below.  (Note, in the  special case of pressurized releases from moderate leaks

below  the   liquid   level   of  the  tanks,  techniques   used   are   those   in

Section 4.13.)


    •  Applicability:

       Calculation  of  release  rate of a  liquid from  a storage  tank  or
       other vessel  with a small leak below the liquid level.


    •  Input:

       PI  -  liquid density (g/cm^)
       A   -  hole or puncture area (cnr)
       Pa  -  atmospheric pressure (dynes/cm^)
       Pt  -  tank pressure (dynes/cnr)
       H   -  height of the liquid column above the hole (cm)
       C(j  -  coefficient of discharge


    •  Limitations/Assumptions:

       -  Does not  simulate time  dependent  release rates  for  tanks with
          decreasing pressure.

       -  The coefficient  of discharge varies between 0.6 and 1.0  as  a
          function  of   release   geometry  and  Reynolds  number.    For
          screening  purposes, assume C^ = 0.8.


    •  References:

       -  Environmental Protection Service (1985)

       -  Hunsaker and Rightmire (1947)


    •  Procedure:

       Calculate  the liquid flow rate by:
                                     4-35

-------
          Flow rate (g/s)
= Cd  A    P!    1960
                                             H +  2 (?t ~ Pa>
    •  Output :




       Liguid discharge rate in g/s for use in the dispersion model.






    Dispersion  of  the  plume  is  conservatively  simulated  with  a  continuous




point source dispersion model  (Flowchart D-2).  For screening estimates, it is




assumed that the leak is  in the  cavity zone  and plume rise  calculations  are




not reguired.   In  addition,  due to the small leak size, it is assumed that the




gas release has low momentum and is passive rather than dense.
                                      4-36

-------
5.0 ATMOSPHERIC DISPERSION ESTIMATES




    This  section provides  screening  techniques  and equations  to  determine




receptor  concentrations  resulting from  toxic  releases.   Methods  are provided




to  examine  and  estimate  the  impacts  of  cloud density,  plume rise,  initial




dilution, and  atmospheric  dispersion on downwind concentrations.   This section




is  designed to  be used  in  conjunction  with  the   dispersion  flowcharts  in




Appendix  D.   Prior to using this section, the  user  should estimate emissions




using the release  scenario  descriptions and emission estimating techniques  of




Section  4 along  with the  flowcharts in  Appendix C.  The  final  step  of  the




appropriate  release  scenario flowchart  (Appendix  C) will  direct  the  user  to




the first step of the appropriate dispersion flowchart and subsection below.




    Dispersion  estimates  for  continuous  and  instantaneous   emissions  are




provided  by  the  techniques  described  in  this   section.   Averaging  times




represented  in  the  estimates  are  determined  by the  dispersion  parameters




used.  It  is assumed  that  continuous estimates will result  in hourly average




concentrations and that  instantaneous estimates represent  peak concentrations




averaged  over  periods  of  less  than  a  minute.    The   user  is  directed  to




Appendix E if approximations to other averaging periods are required.
                                      5-1

-------
5.1 Cloud Densities

    5.1.1 Calculations  to  Determine  the  Relative  Density  of  Instantaneous
          Releases

    •  Applicability:

       All instantaneous releases
    •  Input

       pv  -  density of the gas or vapor (g/cm^)
       mn  -  mass of each constituent (g)
       Mn  -  molecular weight of each constituent (g/g-mole)
       u   -  wind speed (m/s)
       Ts  -  temperature of the material released (deg.  K)
       Ta  -  ambient temperature {deg.  K)
       Ms  -  molecular weight of material  released (g/g-mole)
       Vi  -  volume released  (m^)  (calculated  using  density and  release
              amount)
       pa  -  density of air (g/cm^)
    •  Limitations/Assumptions:

           It is assumed  that neutrally  and positively buoyant  releases
           will be  simulated with passive dispersion models.

           Releases which are negatively  buoyant may disperse as  passive
           materials if atmospheric turbulent  energy exceeds or dominates
           buoyancy effects.
    •  References:

           Havens and Spicer (1985)

           Briggs (Randerson, 1984)


    •  Procedure:

       1)  Density  calculations  begin  with  an  estimate  of  molecular
           weight for gas mixtures


             MS =    £mn
                  £ (mn/Mn)

       2) If:
              Ms        28.9
                                      5-2

-------
      dispersion  is not  affected by  negative  buoyancy effects  and
      passive techniques can be used.
   3) If the release Richardson number


         Ri = 2,722  /    MS Ta     - A    Vi1/3    > 30
                        28.9 Ts        /     u2


      then  effects  of  the  negative  buoyancy  should be  considered.
      Otherwise  passive  techniques  can  be  used.    Note  that  the
      following   density   ratios  may  be   substituted ,  for   the
      parenthetical expression above:

          Pv - Pa \  or  /Pv  _
            Pa     /      \ Pa



5.1.2 Calculations to Determine the Relative Density of Continuous Releases

•  Applicability:

   All continuous releases


•  Input:

   mn  -  mass of each constituent (g)
   MH  -  molecular weight of each constituent (g/g-mole)
   u   -  wind speed (m/s)
   Ts  -  temperature of the material released (deg. K)
   Ta  -  ambient temperature (deg. K)
   d   -  effective diameter (m)
   V   -  volume emission  rate  (m^/s)  specified  or  calculated  from
          mass release rate and density
   Ms  -  molecular weight of material released (g/g-mole)


•  Limitations/Assumptions:

   -  It is  assumed that  neutrally  and  positively buoyant  releases
      will be simulated with passive dispersion models.

   -  Releases which  are  negatively buoyant  may disperse  as  passive
      materials  if  they  are  small  enough  such  that  atmospheric
      turbulent energy exceeds or dominates buoyancy effects.


•  References:

   -  Havens and Spicer (1985)

   -  Briggs (Randerson, 1984)
                                  5-3

-------
•  Procedures:

   1) Density calculations begin with an estimate of molecular  weight
      for gas mixtures
         Ms =
              E (mn/Mn)

   2) If:
          Ms        28.9


      dispersion  is  not  affected by  negative buoyancy  effects  and
      plume rise  and passive modeling  procedures should be  applied.
      Otherwise, proceed to Step 3.
  '3) The release Richardson number is calculated:
         Ri =  2,722
                         Ms Ta     _ , \    V
                        28.9 Ts        /  u3d
      where V is the volume release rate which is  either  specified or
      calculated  from  the  mass  release  rate.   The  volume  rate  is
      obtained  by multiplying  the mass  release  rate by  the  molar
      volume at the  release  temperature and dividing by the molecular
      weight.
         V =
           _  gv
      where  MV is  the molar  volume  (m^/mole)  and  gv is  the  mass
      release  rate  (g/s).   Alternatively, if the  vapor (gas)  density
      is  known,  the  volume rate  is  the mass  rate divided  by  the
      density.

      The diameter, d, in the equation is a  scale  length or effective
      diameter.   For  gaseous  releases,   it  is  taken  to be stack or
      vent  diameter.   For  high  volatility  liquid releases,  it  is
      taken   to   be  the  diameter  of   a   circular  plane  situated
      perpendicular to the  direction of material  transport  by  the
      wind away from the leak or
         d =  _ _
      The parenthetical  terms in  the  Richardson number  equation can
      be replaced by the density ratios

                          /      N
                     or  / Pv  -1
            Pa           I Pa

                                  5-4

-------
if densities of the vapor (gas) and air are known.

If Ri  > 30,  the  effects  of  the  negative  buoyancy should  be
considered.    Otherwise,   plume   rise  and  passive   modeling
procedures should be applied.
                            5-5

-------
5.2 Plume Rise Calculations

    5.2.1  Mean Molecular Weight for Mixtures of Gases

    •  Applicability:

       Plume  rise must  be  calculated for  a  complete  effluent  stream.
       This calculation  provides a mean  molecular weight  to  be used  in
       determining buoyancy flux.


    •  Input:

       mn -  mass of each constituent (g)

       Mn -  molecular weight of each constituent (g/g-mole)


    •  Procedure:

       Compute the mean molecular weight in (g/g-mole) by the following


             Mc =    E m"
                    (mn/Mn)
    5.2.2  Flare Plume Rise

    •  Applicability:

       Continuous flares


    •  Input:

       V  -  volume release rate from the flare (m^/s)
       fi -  volume fraction of each component of the flare gas
       hs -  stack height (m)
       HJ -  net heating value of each component (cal/g-mole)


    •  Limitations/Assumptions:

       -  Plume  rise  for  continuous  flares  must  be  calculated  using
          special  techniques  to account  for radiational  heat losses and
          flame bending in the wind.

       -  55 percent of  the  total heat  output  of the  flare is radiated
          and unavailable for plume rise.

       -  The  flare  flame  is  assumed  to  be  tilted  45  degrees  from
          vertical.
                                      5-6

-------
   -  Source  height  for  the flare  consists  of  the  physical  stack
      height and the height from flare outlet to flame tip.
•  References:

   -  Beychok (1979)

   -  Leahey and Davies (1984)


•  Procedures:

   The EPA regulates the  design of flares used as  control  devices in
   40 CFR  60.18.   In  the regulation, minimum  values of net  heating
   value for the combusted  gas  and exit velocity for  flares,  and air
   and  steam  assisted  flares are  specified.  It should  be confirmed
   prior to calculating  flare plume rise that the  flare  is permitted
   under this regulation.

   1) Calculate  the total  heat  release  rate  from  the  flare  gas
      combustion by:

                                   n
         Qt (cal/s)  =   44.64 V    E   fj_  Hi
      where the  summation is  over  the n components of  the  flare gas
      stream,  fi is  the volume  fraction,  and Hi  is the  net  heating
      value of each component.

   2) Calculate the vertical flame tip height, hf (meters)

         hf  =  4.56 x 10~3  Qt°'478


      and the  effective release height before plume rise (meters) as

         hse  =  hs  +  hf


   3) Calculate buoyancy flux (m^/s3) based on the heat release rate:

         F  =   1.66  x  10~5  Qt


•  Output

   -  hse, flare flame tip height (m)

   -  F, buoyancy flux
                                  5-7

-------
5.2.3  Buoyancy Plume Rise

•  Applicability:

   Plume rise for continuous sources with buoyant releases


•  Input:

   mn  -  mass of each constituent (g)
   Mn  -  molecular weight of each constituent {g/g-mole)
   F   -  buoyancy flux (m^/s^)
   x   -  downwind distance (m)
   hs  -  stack height (m)
   Ta  -  ambient temperature (K)
   u   -  wind speed (m/s)
   Ts  -  stack gas temperature  (K)
   Ms  -  mean molecular weight  of effluent (g/g-mole)
   Vs  -  stack exit velocity (m/s)
   d   -  diameter of stack (m)

•  Reference

   -  Briggs (Randerson, 1984)


•  Procedures

   1) If calculating  plume  rise for a flare, use  the flux determined
      in Section  5.2.2 and  start with  Step 3.   For other  sources,
      compute the mean molecular weight by:


            MS  =     l mn
                        Mn

   2) If calculating plume rise from a flare, use  the flux determined
      in  Section 5.2.2  and  start  with  Step  3.  For  other sources,
      calculate  the  buoyancy  flux,  F (mVs^), using  either  a mixture
      or single component molecular weight and the following:


            F = 2.45 V.d'f (TS/MS) " (Ta/28'9>
                                 (Tg/Mg)

      If F is negative  (or  zero),  the plume is  assumed to be passive
      and  release  height, hs,  should be- used as  the effective plume
      height,  H,  in  dispersion  calculations  (i.e..  Ah  =  0  and
      continue with Step 6).

      If  F is positive,  the  plume  is  buoyant  and  plume  rise can be
      calculated with the procedures that follow.
                                  5-8

-------
3) Determine the distance to final plume rise (meters):

      neutral and unstable conditions:

                   (49  F0.625            for F < 55 m4/s3

                  119  F°-4              for F > 55 m4/s3

      stable conditions:

         xf  =  2.0715 u s-°-5

   where

                A0
          9.81  _
                Az
      s = 	


        A0
   and  —equals 0.02 K/m and 0.035 K/m for E and F stability,
        Az

   respectively.


4) If the  receptor distance,  x,  is less than  Xf,  calculate plume
   rise (meters) for neutral and unstable conditions as:

                      F0.33 X0.667
         Ah  =  1.6
                            u
   For x > xf, substitute   x = xf  in the above equation.
5) Estimate the effective plume height (meters) as

              H  =  hs  +  Ah

   For flares, the stack height hs is defined as hse.
                               5-9

-------
5.3 Dispersion Parameters

    5.3.1 Horizontal and  Vertical Dispersion  Parameters  for  Continuous
          Emission Releases

    •  Applicability:

       Parameters estimating  the  horizontal  and vertical dispersion of  a
       plume for use in Gaussian dispersion models.
    •  Input:

       x   - downwind distance (m)  for point and area sources

       Xy  - horizontal virtual  distance  plus receptor distance  (m),  for
             use with area and volume sources

       xz  - vertical virtual  distance plus  receptor  distance  (m),  for
             use with volume sources

       A-F - stability class


    •  Limitations/Assumptions:

       -  The parameters  were  developed from data collected over  10 min.
          periods but have been used extensively to  provide  concentration
          estimates for one hour averaging periods


    •  Reference:

       -  Turner (1970)


    •  Procedure:

       1) Figures  5-1  and 5-2 provide  estimates of  dispersion parameters
          versus downwind distance (i.e.,  source-receptor distance)  for
          each  stability  class.   Mote  that  downwind distance  in  the
          figures is given in kilometers.

       2) Virtual  point  source distances  for each stability class can be
          determined  by  first  locating  the  value  of  the  dispersion
          parameter  for  each  stability  class  and  then  locating  the
          initial distance on the x axis.

       3) Dispersion  parameters   for  modeling  can   be   determined  by
          selecting  the  value  of  the  dispersion  parameter  for  each
          receptor distance and stability.
                                      5-10

-------
10000
 5000
                 0.5    1             5     10

                      DOWNWIND DISTANCE, km
50    100
     FIGURE 5-1.  HORIZONTAL DISPERSION PARAMETER (cfy) AS  A
     FUNCTION OF DOWNWIND DISTANCE AND  STABILITY CLASS
     (Continuous Releases; Turner, 1970)
                                5-11

-------
5000
1000
   0.1
0.5     1             5     10

      DOWNWIND DISTANCE, km
                                                      50    100
      FIGURE 5-2.  VERTICAL DISPERSION PARAMETER (
-------
   Output:

   °y az   ~  horizontal  and vertical  continuous  plume  dispersion
               parameters  for  each  stability  class   and  receptor
               distance (m)
5.3.2 Horizontal  and  Vertical  Dispersion  Parameters  for  Instantaneous
      Emission Releases
   Input:

   x   - downwind distance to the receptor (m)
   A-P - stability class
   Limitations/Assumptions:

   -  Parameters  are  given  for  three  stability  classes.    It  is
      assumed that the unstable  class includes PG classes A  -  C, the
      neutral category  includes  PG  class  D, and the  stable  class
      includes PG classes E and F.

   -  Dispersion  parameters  represent  instantaneous  peaks   of  one
      minute duration.
•  Reference:

   -  Petersen (1982)


•  Procedure:

   1) Figures  5-3 and  5-4  provide instantaneous dispersion parameters
      relative  to  the puff  center  for  each  stability  class  and
      downwind distance  (i.e.,  source-receptor  distance).   Note that
      downwind distance in the figures is given in kilometers.

   2) Horizontal dispersion parameters,  represented  by ax and av, are
      assumed  to be equal and are given by 0r in the figures.

   3) Dispersion  parameters  can  be determined  by selecting  a  value
      for each downwind distance and stability class.


•  Output:

   -  Vertical,   crosswind,  and  downwind dispersion  parameters,  az,
      Oy, and  ox, for each downwind tosition of the puff (m)
                                 5-13

-------
 5000
 1000
  500
  100
£  50
0)
*••
0)
E
   10
   0.5
   0.1
        '      /
             X

          X
               inn
                Illl
                                                          _
                                                      X    X
                                  '  /     =

                                           /—




                          x
                           X



     0.01
0.05  0.1       0.5   1          5    10


         DOWNWIND DISTANCE, km
                                                       50  100
   FIGURE 5-3.  HORIZONTAL DISPERSION PARAMETER (
-------
 5000
 1000
  500
  100
CO
k
0)
*-

0)

E
  0.5
  0.1
    0.01      0.05   0.1        0.5   1
5    10
              50   100
                         DOWNWIND DISTANCE, km





    FIGURE 5-4.  VERTICAL DISPERSION PARAMETER (dz) AS A


    FUNCTION OF DOWNWIND  DISTANCE AND STABILITY CLASS


    (Instantaneous Releases; Petersen, 1982)
                              5-15

-------
5.3.3 Horizontal and Vertical Dispersion Parameters for Wake Effects

•  Applicability:

   Continuous emission point sources


•  Inputs:

   hs    -  stack height (m)
   hfc    -  building height (m)
   W     -  building width (m)
   L     -  building length (m)
   hjn    -  momentum plume rise (m) at x = 2h^ (see Section 5.6.1)
   A-F   -  stability


•  Limitations/Assumptions:

   For stack to building height ratios  (hg/h^)  <  1.5,  EPA recommends
   the  use  of  the Schulman and  Scire  method  in the  ISC  model.
   However,  this  method  is  not  simple  enough   to   use  in  this
   workbook.  The Huber and Snyder technique is used here.


•  Reference:

   -  Huber  and Snyder  technique  derived from  the ISC  model  (EPA,
      1987c)


•  Procedure :

   1) Determine the maximum projected building width:

         hw = (L2 + W2)1/2

   2) Determine  if  the  plume  will  be  affected  by  the  wake  by
      comparing  the  plume  height (hs +  hn,) to the depth  of  the wake
      region:

      If:
      Wake  effects  are not  significant and  need not  be considered.
      Proceed to Flowchart D-3.

   3) Determine if the receptor is located in the cavity  region:

      If:
            x < 3 ha
                                  5-16

-------
      where:
         ha =
                     for squat buildings (hw >

                     for tall buildings (hw < hj-,)
      then  the  downwash  concentrations cannot  be  estimated  with
      screening  techniques.   Otherwise,  proceed  with  the  wake
      analysis.

4) Calculate plume dispersion parameters:

           0.7 ha + 0.067 (x - 3ha)  for  3ha < x < 10 ha
      °z =\
           az (x + xvz)  for  x > 10ha
   If:
   then
                1.2 h
           0.35 ha + 0.067 (x - 3ha)  for  3ha < x < 10ha
           (x +
                         for  x > 10h
Otherwise
                is obtained from Section 5.3.1.
   Xyy and  xvz  are  the horizontal  and vertical  virtual  distances
   for wake effects as calculated in Section 5.5.3.
5) Use  dispersion parameters  from  4) to  calculate concentrations
   using Section 5.6.4.
Output:
   and oz for wake effects.
                              5-17

-------
5.4 Buoyancy-Induced Initial Dilution

    •  Applicability:

       Turbulent motion  of  buoyant plumes  and associated  entrainment  of
       air  provides  initial  dilution  of  dispersing  clouds.    Pasguill
       (1976) provides a method  for  computing the effect of this dilution
       by modifying the  dispersion parameters used in continuous  sources
       models.


    •  Input:

       aZ' °y   ~  Dispersion  parameters   used  in  modeling   (m)   from
                   Section 5.4.1(a)

       Ah    -  Distance dependent buoyant plume rise (m)


    •  Limitations/Assumptions:

          For small sources the  effect may be negligible.

       -  For most  elevated sources,  neglect of this  step results  in a
          less conservative result (i.e., lower predicted concentrations).


    •  Procedure:

       1) Calculate the modified horizontal dispersion parameter

              ay = 
-------
5.5 Virtual Source Distances

    5.5.1  Virtual Distances for Area Sources

    •  Applicability:

       Continuous  or   instantaneous  releases   of   pollutants   from  area
       sources or  initial  clouds of pollutants which cover a  significant
       near-source area.


    •  Input:

       W - area source width (m)

       A-F - stability class


    •  Limitations/Assumptions:

       -  An area  source can  be represented by a large  plume or puff with
          initial  spreading or  dispersion similar  to  that which  would
          occur with  a point source  located  at some  distance upwind  of
          the source.

       -  Centerline  concentrations near  the   area  source will  be  very
          conservative.


    •  Reference:

       -  Turner (1970)


    •  Procedure:

       1) Determine the width of the area source and estimate  the  initial
          horizontal dispersion parameter by:

             oyo = W / 4.3

       2) Determine the  downwind  distance corresponding  to  the  initial
          horizontal  dispersion parameter  using either  Figure 5-1 (for
          continuous releases) or 5-3  (for  instantaneous  releases).   This
          distance  is the  virtual  point  source  distance,  xv.   Convert
          this distance to meters.

       3) For both instantaneous and  continuous sources, dispersion can
          be simulated by using the distance,  in meters:

             Xy = {xr  + xv)

          where xr is  the  distance to each receptor from the  area source
          center.   In  the  calculation  of  dispersion  parameters,  the
          distance xv is used for determining ay.

                                     5-19

-------
   4)  In comparisons  of  the  source/receptor  distance   relative   to
      position of maximum  concentrations,  the  receptor distance,  xr,
      should be used.
   Output:

   -  Virtual  point source distance and  total  distance to be used  in
      determining plume spreading parameters  for area  sources.
5.5.2  Virtual Distances for Volume Sources

•  Applicability:

   Continuous/instantaneous  ground   level  releases  of   pollutants
   which,  as  a   result  of   release   characteristics  or   initial
   dispersion, can be  represented as a cloud with  significant  volume
   at the start of dispersion away from a source location.
   Input:

   W   - cloud or source width
   Hv  - cloud depth (m)
   A-F - stability class
   Limitations/Assumptions:

   -  A  pollutant  cloud  with initial  spreading or  dilution  can  be
      represented by  a  point source  located sufficiently  far  upwind
      such that the  simulated puff or plume is bounded by the initial
      cloud dimensions.

   -  Centerline concentrations  within 10  side  widths of  the  volume
      source will be conservative.
   Reference:

   -  Turner (1970)


   Procedure:

   1) Use the techniques  of  Section  5.5.1  to determine  a horizontal
      virtual source distance, xv, for use with the volume source.

   2) Determine  the  depth  of  the  volume  source and   the  initial
      vertical dispersion parameter by:

         azo = Hv / 2.15
                                  5-20

-------
   3) From  Section  5.3.1  (for  continuous  releases)  or  5.3.2  (for
      instantaneous   releases),   determine   the  downwind   distance
      corresponding  to  the  initial  vertical  dispersion  parameter.
      This  distance  is  the  vertical  virtual point source  distance.
      xvz-

   4) For  both  instantaneous  and   continuous  sources,  dispersion
      estimates are made  by determining  the vertical  and  horizontal
      parameters using the distances

         Xy = (Xj- + X^)


      for the horizontal parameter, and
      for the vertical parameter.
   Output :

   -  Virtual  source  distances  and total  distances  for  determining
      plume  and  puff  spread  parameters  to  be  used in  dispersion
      models for volume sources.
5.5.3  Virtual Distances for Wake Effects

•  Applicability:

   Continuous emission point sources


•  Inputs:

   hs   -  stack height (m)
   h]-,   -  building height (m)
   W    -  building width (m)
   L    -  building length (m)
   t^   -  momentum plume rise (m) at x = Zhj-) (see Section 5.6.1)
   u    -  wind speed (m/s)
   A-F  -  stability


•  Limitations/Assumptions:

   For stack to building  height  ratios (hg/hj^) <  1.5,  EPA recommends
   the  use  of the  Schulman  and  Scire  method  in the  ISC  model.
   However,  this  method  is  not  simple  enough   to   use   in  this
   workbook.   The Huber and Snyder technique is used here.
                                 5-21

-------
•  Reference:

   -  Huber and  Snyder technique  derived from  the  ISC model  (EPA,
      1987c)
•  Procedure:

   1) Determine the maximum projected building width:

           1^ = (L2 + W2)1/2
   2) Determine  if  the  plume  will  be  affected  by  the  wake  by
      comparing the plume height  {hs + 1^) to  the  depth of  the wake
      region:
      If:
                       2.5 hb

                       1.5 hw + hb
                    , or
      Wake  effects  are not  significant and need not  be  considered.
      Proceed to Flowchart D-3.
   3) Determine if the receptor is located in the far wake region:

      If:
            x > 10 ha

         where:
                      hb for squat buildings (hy, >

                      hy, for tall buildings (h^ <

      then  proceed  with  step  4.   Otherwise,  wake  effects virtual
      distances are  not  needed  (proceed with near  wake calculations
      in Section 5.3.3).
   4) Calculate   the  horizontal   and   vertical  plume
      parameters at x = 10ha:

         oy = 0.82 ha
                                             dispersion
            = 1.2 h,
         where:
ha =
                         for squat buildings  (h^,  >

                         for tall buildings  (h^, <


                                  5-22

-------
5) Determine  the  downwind distance  corresponding to  the  enhanced
   horizontal   dispersion   parameter   using   either   Figure 5-1
   (continuous  releases)  or  5-3  (instantaneous).  Similarly,  use
   Figures 5-2 or  5-4  to determine  the distance  corresponding  to
   the  enhanced vertical dispersion parameters  for  the  stability
   category of interest.

6) Use  the  horizontal  virtual  distance,  xvy,   and   the  vertical
   virtual distance, xvz, when  calculating wake  effects dispersion
   parameters for the far wake using Section 5.3.3.
Output:
       and xvz for wake effects
                              5-23

-------
5.6 Concentration Calculations

    5.6.1  Cavity Modeling

    •  Applicability:

       Concentrations  from  continuous point  sources  trapped within  the
       recirculation zone  in the  lee  of a  building.
       Input:

       hfc  -  height of building (m)
       Lfc  -  lesser dimension (height or projected width of building)(m)
       Ta  -  ambient air temperature (K)
       Ts  -  stack exit temperature  (K)
       vs  -  stack exit velocity (m/s)
       d   -  stack diameter (m)
       uc  -  critical wind speed (m/s)
       hs  -  stack height (m)
       W   -  crosswind building dimension (m)
       A   -  cross-sectional area of building  normal to wind (m2)
       u   -  wind speed (m/s)
       L   -  alongwind building dimension (m)
    •  Limitations/Assumptions:

       -  The  model  simulates  cavity   concentrations   with  a  uniform
          distribution.
    •  Reference:

       -  EPA,  1984  (Regional  Workshop  on  Air  Quality  Modeling:
          Summary Report)
    •  Procedure:

       1) Compare  the  stack height to  the cavity height.   Calculate the
          cavity height hc (m):

             hc =  hb + 0.5(Lb),
          If the  stack height  is greater  than  or  equal  to the  cavity
          height,   no further  cavity  analysis  is  required.   Proceed  to
          perform  the  wake  effects  analysis  (Section 5.5.3).   If  the
          stack height is less than the cavity height, proceed to Step 2.
       2) Estimate  the  momentum  plume  rise   for  neutral  atmospheric
          conditions.  First compute the momentum flux, Fm (m4/s2):
                = 

                                     5-24

-------
3) Next, compute the momentum plume rise hm (m)
             3Fm*x

             b2 u^

      where:  b = (1/3 + uc/vs).

      x = downwind distance (m).   Use x = 2hj-,.

      Assume a critical wind speed uc = 10 m/s


4) Compute the plume height, Hp (m):

      ^p  =  hg + hj^
   If  the  plume  height is  greater than  or  equal  to  the  cavity
   height calculated in Step 1, then no further  cavity  analysis is
   required.      Proceed    to   the    wake    effects    analysis
   (Section 5.5.3).  If the plume  height is  less than  the  cavity
   height,  proceed to Step 5.
5) Estimate the downwind extent of the cavity.  Compute  the cavity
   length (xr), measured from the lee side of the  building (m):

   For short buildings (L/h^Z):
           1.0 +

      where:  a = -2.0 + 3.7 (L/hb)-°-33 and

              b = -0.15 + (0.305) (L/hb)-°-33

   For long buildings (L/hj-, > 2):
      xr =    1.75{W)
           1.0 + 0.
25(W/hb)
   Next, compare the  cavity  length to the closest  distance  to the
   plant property  line.   Consider only  plant  property to  which
   public access is precluded.   If the cavity does not exceed this
   distance, then it  may  be  assumed that  cavity effects will not
   impact   ambient  air,   and   no  further  cavity   analysis  is
   required.  Proceed  to the  wake effects analysis.   If the  cavity
   extends  beyond plant property, proceed to step 6.
6) Estimate impacts within the cavity.   "Worst  case" concentration
   impacts  (C)  can  be  estimated  by  the following approximation
   (g/m3):

                              5-25

-------
         where:

               Q = emission rate  (g/s),

               A = cross-sectional  area  of  building normal  to  wind
                   (n)2)  equals  W  x  h^ and

               u = wind  speed (m/s).

      For  u,   choose the  lowest   wind speed  likely  to  result  in
      entrainment of most pollutants into the cavity.   If no  data are
      available from which the minimum  speed can  be estimated,  assume
      a worst  case wind  speed of  1  m/s.

      Since  both   the    cavity   concentration   and   cavity   length
      calculations depend  on  building  orientation relative  to  the
      wind, it  is  advisable  to  repeat  the calculations  using  two
      orientations of the  building; one  with  the minimum  dimension
      alongwind  and  the  other with maximum dimension alongwind and
      choosing the highest concentration.
5.6.2 Heavy Gas Model - Instantaneous Releases

•  Applicability:

   The model  described  is  applicable  to instantaneous ground  level
   releases of  dense  gases  in  flat  terrain.   The  model  neglects
   evaporation of  condensation of  drops and  vapors  in clouds,  heat
   and mass  exchange  with  underlying  surface,  radiation  flux  and
   chemical reactions.


• Inputs:

   Ro -  initial cloud radius  (m)
   u  -  wind speed (m/s)
   Vo -  initial cloud volume  (m^)
   Pa -  density of air (g/m^)
   Pq -  density of gas (g/m^)
   xr -  downwind receptor distance (m)

•  Limitations/Assumptions:

   Gases which are  heavier  than air  disperse  under the influence of
   buoyant  and turbulent forces.   Models for  simple cases  of  gases
   much heavier  than air on  flat  surfaces  have  been developed  and
   perform  well  in   the   nearfield  where  buoyant  forces  clearly
   dominate  spreading  due   to  atmospheric  turbulence.    Additional
   research   is   required  for   determining   concentrations  in  the
   transition zone  between  dominance by buoyant forces and turbulent
   mixing.

                                  5-26

-------
   The  model  presented  consists  of two  components,  a  negatively
   buoyant  cylindrical  spreading  model  and  a  passive  dispersion
   model.  The  initial  spreading  of the  dense cloud  is  controlled  by
   gravitational effects, and  the  cloud  essentially slumps under  the
   influence  of gravity.   Slumping  is  terminated by entrainment  of
   air.  This results  in a  dense gas cloud which  hugs  the ground  as
   it  travels  downwind.   Eventually the  cloud  of  dispersing  gas
   becomes  dilute,  and  atmospheric  turbulence  dominates  the  cloud
   growth.   Van Ulden  (1974)   recommends  ending  the slumping  phase
   when  the  frontal  velocity  becomes less  than  twice  the  friction
   velocity.    At   this   point,    the   maximum   spread  radius   is
   calculated.  This radius is  approximate since the  simple  spreading
   model  of Van Ulden  neglects any  entrainment during the  slumping
   phase.

   To  approximate  dilution due  to  entrainment,   a  relationship  is
   defined which estimates  the volume to be added  to the  cloud as  a
   result  of edge  entrainment.   This  total  volume  is  then  used  to
   calculate the cloud  height  at the  termination  of spreading.   The
   entrainment  parameter for  edge  spreading  is   given by  Cox  and
   Carpenter  (1980).   Gravitational  spreading   is  assumed  to  occur
   instantaneously  and   to  be  centered   at   the   initial   source
   location.  The resultant  cloud  is assumed to be  cylindrical  with
   dimensions   given   by  the  maximum   spread  radius  and   height
   determined from the spread volume.

   Dispersion under  the control of atmospheric  turbulence  is  assumed
   to  be  represented  by  a passive  Gaussian  volume  source  model.
   Virtual  distances  are  calculated using  expressions  similar  to
   those presented in Section 5.5.2.
•  Procedure:

   Steps in the heavy gas simulation are shown in Flowchart D-6.

   1) The  maximum  spread  radius  (m)   is  given  by  the  following
      equation using information  on  the initial cloud  size,  density,
      and wind speed:
                14.7     Pg-Pa   V0
                _  / _ - _
                 u   V    pa
      V0  is  determined  from  the  initial  mass   released  and  cloud
      density.
   2)  The total cloud volume  (V-p)  is calculated based on  the  maximum
      radius and an approximate equation representing the  final  cloud
      volume due to edge entrainment (V):
         V = V0
                   Rmax
                           1.2
                 \  Ro
         VT = V0 + V
                                 5-27

-------
   3) Cloud  height  can  then  be  calculated  assuming  a  cylindrical
      shape:

                  VT
         H  =
      In the  case  of  very  heavy  gases,  the  potential  exists  for
      unrealistically   thin   clouds   to   be   formed.    To   bound
      calculations  in  this  instance,  a minimum  depth of  5  cm  is
      specified.   If the  calculated cloud  height is less than  5 cm,
      this  height  is  used  in further  calculations  and the  maximum
      radius is recalculated.
   4) The emission amount (g)  for passive calculations  is  obtained by
      multiplying  the  initial  cloud  density  by  the  initial  cloud
      volume.

         Qt =  v0 Pg


   5) Virtual   distances  for   passive   dispersion   estimates   are
      calculated with dispersion parameters using:

         xvz  = (9.3 H)  i-64

              X~~ t V  -4- V    J« 1?   \
          z   ~~ *xr r vz  *Tnax'

      and

         xvy  = (23.26 Rmax'

              X— / *r  ^ V   \
          y   ~ vAr ~ Avy'

      Calculations  proceed  as  in  Section  5.6.5   using  the  mass
      emission  amount   (Qt)'   wind  speed,  virtual  distances  and
      stability.
5.6.3 Heavy Gas Model - Continuous Release
   Applicability:

   The   model   described   is   applicable   to   short-term   ambient
   concentration  estimates  resulting  from  continuous  elevated dense
   gas releases occurring at a release height of 10 m or  more in flat
   terrain.   Concentration  estimates  are   applicable  for  downwind
   distances of no more than one kilometer.
                                 5-28

-------
   Inputs:

   Q    - contaminant emission rate (kg/s)
   vs   - stack exit velocity (m/s)
   Ts   - stack exit temperature (K)
   hs   - stack height (m)
   d    - stack diameter (m)
   u    - wind speed at stack height  (m/s)
   y    - contaminant concentration in stack exit gas  (percent  volume)
   po   - exhaust gas density (kg/rn-^)
   M    - exhaust gas molecular weight (g/  g-mole)
   Mc   - contaminant molecular weight(gXg-mole)
   x    - receptor distance (m)
   p    - wind speed profile  exponents (see section 3.1)
   pa   - ambient air density (kg/irH)
   Ta   - ambient temperature (K)
   C(x) - centerline   ground   level    contaminant  concentration   at
          downwind distance,  x (ug/m^)
•  Limitations/Assumptions

   The Relief  Valve  Discharge  (RVD)  model is  a  screening  technique
   applicable to  denser-than-air gaseous  releases.   The name  of  the
   model would  imply that  it  is only  applicable  to pressure  relief
   valves.   However,  this  is  not the case and the  model is  applicable
   to screening analysis  of elevated dense gas releases.  The  model
   is based on  wind  tunnel data and empirical relationships developed
   by Hoot, Meroney,  and  Peterka (1973) using heavy gas  tracers  in a
   non-turbulent environment.  Since  the  wind tunnel experiments most
   closely    represent    stable    atmospheric    conditions,     the
   concentrations that occur  under  unstable  or neutral  conditions may
   be significantly  overestimated.    The  model simulates plume  rise
   and  descent  under  the  control  of  buoyant  forces  to provide  an
   estimate  of   plume  trajectory   to  the   touchdown  point   and
   concentration  at  the   point  of  touchdown.   The  dense  gas  plume
   rises at first due  to  initial upwind momentum from  the  stack,  but
   then  sinks  due  to  its  excess  density.   Eventually  the  plume
   centerline   strikes    the    ground   surface.     For   screening
   calculations, the maximum concentration should  be that  calculated
   at the  point of  plume touchdown.   If this estimate indicates  a
   problem, more  refined  techniques such as the DEGADIS model  should
   be used.
•  Procedure:

   The RVD  model  is available  on  a  PC compatible  diskette  from:
   Source Receptor Analysis Branch,  MD-14,  USEPA, RTP, NC 27711.  The
   use  of  this  computer  model  is  suggested  for  accuracy.   The
   following is an abbreviated  version of the RVD model for use as an
   illustrative tool  only.

   1) Calculate  the   release  Richardson  number  R^  as described  in
      Section  5.1.2.   If  R^  >  30,  then  dense  gas  effects  are
      considered.   Proceed to step 2 below.

                                 5-29

-------
2) Calculate Froude number,  Fr:


                   Po \\°-5
      Fr = u
                 Po-Pa
                 9.8 d
3) Calculate plume rise,  Ah:

                     u pay
4) Calculate dilution ratio at maximum plume height:

                                 1.85
      R = 5.67 x 10'
                         -
                       yd2uM\/Ah
5a) Calculate plume molecular weight at maximum plume height:

         ..   _   M + 29 (R-l)
         MH  =  	
5b) Calculate density difference at maximum plume height:
      A =
- 1
            29
   If A  < 0.005,  use  continuous  dispersion  model for  non-dense
   gases. Section 5.6.4.
6) Calculate distance at plume touchdown, XT:
      XT =
               "
                 + 0.56 d Fr
               /VsPa\
                                                            0.5
7) Calculate concentration at plume touchdown,

                                    -1.95
      C(xT) = 3.1 x 109
                        ud2 \  d
8) Downwind  concentrations   after  plume   touchdown  depend   on
   critical distance, xc:

               .          K     ,-1.538
               '2.045 x 105 • Mc\
      xc = XT
                    C(xT)
                               5-30

-------
      If xc < XT, let xc = XT-

      If receptor distance x < xc:
                         \-0.65
         C(x) = C(xT)
               (-T
               \XT/
If receptor distance x > xc:

                    -0.65    -1.7
   C(x) = C(xT)
                       M     /
                       *T/     \xc
      Model  results  should  be  reviewed  first to  determine  if  the
      plume  touches  down  within  1  kilometer.   If  it  does,  then  an
      estimate of touchdown  distance  and  concentration is  available.
      If touchdown  is  not indicated  within the first kilometer,  the
      model is performing beyond its scope.
5.6.4 Dispersion Model for Continuous Releases

•  Applicability:

   Simulations  of  dispersion  from  continuous   point   sources   and
   continuous area or  volume sources through  application of  virtual
   point source  techniques.  May also be applied  to sources with wake
   effects due to building downwash.


•  Inputs:

   Q       -  emission rate (g/s)
   °z' °y  ~  continuous    vertical    and    horizontal   dispersion
              parameters (m)
   u       -  wind speed (m/s)


•  Limitations/Assumptions:

   -  Method provides  centerline maximum concentrations with  downwind
      distance for specific input conditions.


•  Reference:

   -  Turner (1970)


•  Procedure:

   Concentrations in g/m^  are  provided  at each downwind  distance,  x,
   by the following equation.


                                 5-31

-------
      C (x)  =
              ir oz Oy u
exp   -0.5
                                           H
   The effective dispersion parameters  incorporate, where  necessary,
   initial dispersion  due  to wake  effects  as  well as area or  volume
   source  releases.    Concentrations   estimated  represent  one  hour
   average values.
5.6.5 Dispersion Model for Instantaneous Releases

•  Applicability:

   Simulations  of  dispersion  from instantaneous  point  sources  and
   continuous volume  or  area  sources  through application  of virtual
   point source techniques.
   Input:

   Qt      -  release amount (g)
   H       -  effective source height (m)
   az' ay  ~  instantaneous   horizontal   and   vertical   dispersion
              parameters (m)
•  Limitations/Assumptions:

   -  Maximum ground level concentrations  are provided for the center
      of each instantaneous  puff at selected downwind locations.

   -  Full surface reflection is assumed.

   -  Crosswind and downwind dispersion are assumed to be equal.

   -  The downwind  position  of the puff is  determined by multiplying
      wind speed times travel time.


•  Reference:

   -  Petersen (1982)


•  Procedure:

   1) Use  of   the  PUFF  model   (Petersen,   1982)   will   simplify
      calculations.
   2) Hand calculations are provided by:

               0.127 Qt                    „
         C = 	     exp (-0.5 (H/az)2)
                                  5-32

-------
The  technique  is  strictly  valid  only  if  travel  time  to  the
receptor   from  the   source   exceeds   the   release   duration.
Otherwise, an unrealistically high concentration will result.
                             5-33

-------
6.0 EXAMPLES

    This section  provides examples  for the  release  scenarios  identified  in

Table 2-1   and  Section 4.    These   examples   illustrate   the  solution   to

mathematical equations  used in  the  text  and may  not  represent  the  maximum

short-term   ground  level   concentration   estimate   from  a   meteorological

perspective.  To obtain  these  maximum concentrations,  the  user should  follow

the procedures outlined in Section 2.4.



6.1 Continuous Gaseous Emissions from Stacks

    Scenario:  Hydrocyanic  Acid (HCN)  is  released from a vent  stack.   Hourly

maximum concentration estimates are required.



    Discussion;  This example  represents a continuous stack release of a vapor

or,  similarly, particulate  matter  at  near  ambient  conditions.    The  only

difference  for particulate matter is  that plume  density checks would  not be

performed.   In the example, flow  is  under  the influence of a  nearby building.

Emission  rates are  determined  as  specified  in Section  4.1.2 and shown  in

Flowchart C-5.  Emission rates must  be calculated  from process parameters  or

determined  from representative emission factors.  In  this case, emissions  are

specified.   Emission factors are  also available (see  Appendix A).   Dispersion

procedures are outlined beginning in Flowchart D-l.



    6.1.1  Building Cavity Example

    Source Parameters

    MHCN =  27 9/g-mole
    Ta   =  298 K
    Ts   =  298 K
    d    =  0.1 m
    V    =0.14 m3/s
    xrec =  distance to property line = 25 m
    Q    = 9.3 x 10~4 g/s

                                   6-1

-------
hg   = 16 m
hjj   = height of building = 19 m
L    = alongwind building dimension = 19 m
W    = crosswind building dimension = 19 m
Ljj   = lesser dimension = 19 m
v    = stack exit velocity = 4V/(ird2) =17.8 m/s
Sample Estimates

   1) (Flowchart  D-l  and  Section  5.1.2).   Because  HCN  (molecular
      weight 27)  is  the primary constituent (besides air)  in the gas
      stream,  and  its  concentration  is  low  compared  to  air,  the
      overall molecular  weight  of  the  gas stream can be  approximated
      by that of  air.   Alternatively,  mean density calculations would
      show the  gas  stream to have a mean  molecular weight  which is
      slightly less than air.

         Ms = 28.9 g/g-mole
   2) Determine  if  dispersion  is   affected   by  negative  buoyancy
      (Section 5.1.2).

          Ts      298       Ta
          Ms      28.9     28.9

      Negative  buoyancy is  not  a  factor  and  point  source  passive
      dispersion techniques apply (Flowchart D-2).
   3) Determine if plume is in the cavity (Section 5.6.1)

      •  Compare stack  height  to cavity height  (Note  that LJ-, = L = W
         for this square building):

         hc = 19 m + 0.5 (19 m) = 28.5 m

         hs =  16  m < hc;  therefore,  the stack release height  is in
         the cavity
      •  Calculate  the  momentum flux and plume  rise (Note that stack
         exit velocity  is  obtained from the volumetric  flow rate and
         the stack diameter):

                 / 298 K \ /       m  \2         2              .   ,
            Pm = /	H  17.8 	]     (O.lm)  /4  = 0.79 m4/s2
                   298 K / \      sec
                        3  (0.79 m4/s2) 2  (19 m)
                   (0.33 + 10 m/s/17.8 m/s)2  (10 m/s)2
1/3
                                                         =  1.0 m
                                6-2

-------
          *  Compute plume height and determine if plume is in cavity:

                Hp = 16 m + 1 m = 17 m

                Hp < hc; therefore plume is in cavity


       4} Determine if the receptor is in cavity.

          •  Compute the downwind extent of the cavity
                     =  19 m/19  m =  1;  therefore  use  the  equation  for
                short buildings:

                           [-2.0  + 3.7 (19/19)-°-33] 19
                xr = 	  = 28.0 m
                     1.0 + [-0.15 + 0.305 (19/19)-0-33] (19/19)
          For receptor = 25 m,  receptor is in cavity  region.   Use cavity
          model  (Section 5.6.1).   The  cavity  calculation  would not  be
          necessary if the receptor were outside of the plant boundary.
                A = (hb) (L) = 19 m x 19 m = 361 m2

                       9.3 x 10~4 g/s

                    1.5 (361 m2) (1 m/s)
       9.3 x 10~4 g/s               CQ           o
C = 	 = 1.7 x 10~6 g/m3 = 1.7 ug/m3
    6.1.2 Near-Wake Example

    In  the  previous  example cavity  calculations  were  required.   If  in  the

example

          hs = 28 m

    and

          x = 100 m,

an  alternative  path  of calculations  is necessary,  resulting in  a near-wake

region concentration estimate.  Prom step 3 above:
       hs = 28

       Hp = 28 + 1.2 = 29.2

       !!._, > 23.5  (the  cavity height).  Therefore, the plume is not in the
       cavity and the user should follow Section 5.5.3.
                                   6-3

-------
Sample Estimates

1) liw = ((19)2 + (19)2)1/2 = 26.9 m

   Since:  hs+hm=28m+l.lm=29.1m

   and:

                (2.5 (19 m), or

                1.5 (19 m) + 19 m,

   wake effects are significant.


2) Compare receptor distance to the wake region.

      x = 100 m < 10 (19 m)

   therefore, near wake equations are used.


3) In  Section 5.3.3,  all parameters through  step 2  are calculated.
   In step 3:

      x = 100 m > 3 (19),

   therefore, the receptor is located in the near wake region.


4) Calculate plume dispersion parameters:

      x = 25 m < 10 (19)

   therefore:

      oz = 0.7 (19 m) + 0.067 (100 m - 3 (19 m)) = 16.18 m

   and,

      hs + hn, = 29.1 m <  1.2 (28 m)

   therefore:

      ay = 0.35 (19 m) +  0.067  (100 m - 3  (19 m)) = 9.53 m
5) Calculate concentration using Section 5.6.4:

                         9.3 x 10~4 g/s
      C  (100 m) = 	   exp
                  3.14  (9.53 m)  (16.18 m)  1 m/s

                = 3.81  x 10~7 g/m3 =38.1  ug/m3
                                6-4
-0.5
      29.1
      ,16.18,
             2

-------
    6.1.3 Far-Wake Example

    The far-wake  region  is  defined as receptor distances beyond  10  ha or,  in

this  example,  x  > 190  m.   This  section  demonstrates  a  calculation  for  a

receptor at 200 m from the source.



    Sample Estimates

    1) The far-wake  determination  is made  in Section 5.5.3, Step 3.   In
       Step 4,  dispersion parameters are  calculated  along with  virtual
       distances:

          oy = 0.82 (19 m) = 15.6 m

          az = 1.2 (19 m) - 22.8 m

       giving virtual distances using Figures 5-1 and 5-2 (stability  F):

          Xyy = 440 m

          xvz = 2200 m


    2) In Section 5.3.3, check  the  location of the  receptor  relative to
       the cavity  zone.   The dispersion  parameters are determined  using
       virtual distances, x^,  xvz  (stability F):

          ay (200 m + 440 m) =  22 m

          az (200 m + 2200 m) = 24  m
    3) Calculate concentration (Section 5.6.4):

                           9.3 x 10~4 g/s
          C (200 m) =
                      3.14 (22 m) 24 m (1 m/s)
exp
     -0.5
          '29.2
            24 m
                    = 2.68 x 10~7 g/m3 =26.8 ug/m3
                                    6-5

-------
6.2 Fugitive Dust

    Scenario;   Concentration estimates  at  the  fenceline   are   required   for

arsenic  emissions  resulting  from wind  erosion from  a pile  of  flyash at  a

secondary lead smelter blast furnace.



    Discussion;    This   example   demonstrates   calculation  of   particulate

emissions from storage piles and use of particulate matter profiles to  study  a

specific  toxic   chemical.    Maximum  concentration  estimates   are   normally

obtained using the procedures described in Section 2.4.  Worst case estimates,

in this  case,  use  conservative  assumptions and deviate from those discussed in

Section 4.2 since maximum emissions are wind speed dependent.


    Source Parameters

    Ash pile:   height -3m
                diameter - 10 m

    Distance to boundary 100 m


    Sample Estimates

    I) The  fugitive  dust  scenario  is  presented  in  Section  4.2,   and
       estimates  follow  Flowchart  C-2.   Fugitive  emissions  for  this
       scenario  are  not  directly  available.    Emissions   factors   for
       aggregate storage  are  available  in AP-42 as are particulate matter
       profiles  (Appendix A).  For  this  example,  the profiles  indicate
       that arsenic makes up 0.3 percent of fine particles (less than 2.5
       microns)  emissions mass.   The  aggregate  storage  emission  factor
       for windblown dust is:

                                             (365-p)
          E (kg/day/hectare)  =  1.9 (s/1.5) 	  (f/15)
                                               235

       where:

          s - percent silt content
          p - number of days with more than 25 mm of precipitation
          f - percent of time wind exceeds 5.4 m/s

       Since   the   factor   is   not  directly   applicable,   conservative
       assumptions  are made  that  20  percent of wind  exceeds  5.4 m/s, no
       days have precipitation in excess of 25 mm and  the  silt content is
       50 percent.   The calculated emission  rate  is  131.2 kg/day/hectare

                                    6-6

-------
   or 0.00015  g/(s  m^).  Since  0.3  percent of this mass  is  arsenic,
   the emission rate  is 4.55 x 10~^ g/(s m^)  over 78.5 m^ (irr^ area)
   or 3.58 x 10~5 g/s.
2) The  pile   height   is   relatively   low  (3  m)   in   height,   and
   conservative  estimates  will  result  if  area  source  techniques
   (Flowchart  D-4)  are  used.   The  initial  horizontal  plume  spread
   parameter is given by (Section 5.5.1):

      <7y  =  10 m/4.3  =  2.3 m


3) Meteorological conditions  resulting in  maximum  concentrations for
   a ground level area source  are  low  wind speed (1  m/s),  stable (F)
   conditions.    These are  assumed  for  conservatism and provide  a
   virtual distance  from Figure 5-1 of  approximately:

      xv = 74 m
4) Using  this  distance  with  the  minimum  receptor  distance,  the
   horizontal and vertical dispersion  parameters (Section  5.3.1)  are
   determined from Figures 5-1 and 5-2:

      Oy (174 m) =  6.6 m

      oz (100 m) =  2.3 m
5) Maximum  estimated concentration  is  obtained  from  Section 5.6.4
   (the exponential term equals unity because H is zero):

                3.58 x 10~5 g/s                    -TOO
      C =  	  =  7.5 x ID"7 g/m3 = 0.75 ug/m3
           3.14 (6.6 m)  2.3 m (1 m/s)
                               6-7

-------
6.3 Instantaneous Ejection of Particles  from Ducts

    Scenario;     A   failure    of   a   pneumatic   conveyor   system   carrying

3,3-dichlorobenzidine powder from a  spray dryer lasted  5 minutes.   Estimates

are  required  for 15-minute  average  concentrations at  receptors greater  than

100 m downwind of the source.



    Discussion:    The  scenario  represents a  class of  possible releases  from

various types of gas-solid  conveyance  systems  or reactor  failures.   Common

causes of this type  of  release are duct failures due  to  abrasion or failure of

flexible connectors.   Short  duration  events  can be simulated as  instantaneous

passively  dispersing  puffs   (i.e.,   all   mass  was  released  instantaneously

(within  one minute)).   The  effect   of  this  assumption  is  a  conservative

concentration estimate.   In  general, powders  emitted by this  type  of release

will consist of  relatively large  particles (order of  10 microns) which would

be  subject  to  gravitational  fallout.   Since the  screening  techniques neglect

deposition and fallout,  conservative  concentration  estimates are expected.


    Source Parameters

    release height = 10 m
    conveyance rate = 2 kg/s
    duct diameter = 0.305 m


    Sample Estimates

    1) Section 4.3 and Flowchart C-3  indicate that  duct  failure emissions
       are  typically  user  estimated  and that  dispersion  calculations
       follow point source procedures in Flowchart  D-6 (a point source is
       assumed because  no indication of  initial  dilution  dimensions are
       provided in the problem).  The release scenario would  result  in an
       initially  high  rate of  emissions  which decreases  rapidly as line
       pressure decreases, as  in  a  pipeline blowdown.   One example  of a
       user calculation is  given by assuming  that total  emissions (0,^)
       consist of that material  which  would normally be conveyed  in a 5
       minute period, i.e.:

                             /60 s\                    _
          Qt  =   (5 minutes) j 	 J 2 kg/s  =  6.0 x 105 g
                             \rnin /
                                   6-8

-------
2) Since the release  is  at 10 m, the distance to maximum ground level
   concentration (Section  2.4)  is approximately  that  at  which az  =
   H/>5~or  7  m.   Distances   to  maximum  concentration  for  each
   stability are determined from Figure 5-4:

          Stability      Distance (m)     az  (m)      ov (m)
                                                      y
          unstable            30            73
          neutral            240            7         10
          stable            3500            7         30

          (unstable)        (100)         (10)       (10.5)

   The  maximum  concentration  occurs where  the product  Oy az2  is  a
   minimum which, for  this  case, is under unstable conditions.   Since
   the distance  to  maximum concentration  for unstable conditions  is
   within  the  minimum   receptor   or  fenceline  distance  (100  m),
   dispersion parameters for 100 m were also determined.
3) With  the  unstable  dispersion parameters  considered, the  maximum
   concentration occurs with the minimum product of oz Oy2,  or during
   neutral conditions.  Prom Section 5.6.5:

              0.127 Qt                      .
      C =	   exp (-0.5 (H/oz)2)

              az  0y2


             0.127 (6.0 x 105 g)                          ,
        =  	   exp (-0.5 (10 m/7 m)2)
               7 m (10 m)2

        =  39.2  g/m3

   The resultant peak  15  minute average concentration at the receptor
   can be found using techniques in Appendix E:

               (900 s) (1 m/s)
      N =  	  =  45  giving A = 1
                  2 (10 m)

                  A - 0.5
      F =  	  =  0.028
                (0.3989) 45


   The concentration is then:

      Cavg  =  °-028 <7-85
                               6-9

-------
6.4 Flare Emissions

    Scenario;   A  gas  is  sent  to  an  elevated  flare  to  be  burned.   For

simplicity,  it  is  assumed that the flare is permitted.   The  gas is a  mixture

with  one toxic  component.   The  gas stream  is  made  up  of  methane,  ethane,

carbon dioxide  and benzene.  Maximum one-hour  concentrations  are required  for

benzene assuming 98% reduction  efficiency of the flare.



    Discussion;  Flare problems are done in two parts, an emission  calculation

and dispersion  modeling.   Toxic emissions for  permitted  flares  are  reduced to

2% of  the  potential  emissions  based on a  required  control  efficiency  of  98%.

Flare  problems  are similar  to stack examples  except  that there are  buoyancy

flux reductions  associated  with radiative  heat  losses and a need  to  account

for  flame  length  in  estimating  plume  height.  Estimates   of  concentrations

require calculations of heat flux, flame length, plume rise and dispersion.


    Source Parameters

    Fi  =  gas composition (volume fraction):
           - methane - 0.50
           - ethane  - 0.098
           - carbon dioxide - 0.40
           - benzene - 0.002
    hs  =  flare height -  32 m
    V   =  flow  rate to flare - 6.58 nrVs


    Sample Estimates

    The  flare  scenario  is presented in Section  4.4 and  Flowchart  C-4.  Steps

in calculations are as follows:
    1) Determine the  emission  rate of benzene  from the  volume fraction,
       molar  volume,   flow  rate,  and molecular weight.   The volume  of
       benzene is the fraction times flow rate:

          V(benzene)  =  0.002 (6.58 m3/s) = 0.013 m3/s
                                   6-10

-------
       Mass  emission rate  after controls  is given  by  determining  the
       number  of  moles   in  the  benzene  fraction  and multiplying  times
       molecular weight (the gas is assumed to be at  standard conditions)
       considering the control efficiency:

              (0.013 m3/s) (78.1 g/g-mole) (0.02)
          Q = 	  =  0.9 g/s
                       0.0224 m3/g-mole

    2) Calculate the  total heat  release  from  the  flare  (Section  5.2.2).
       In this  example, carbon dioxide is not combustible and is  assumed
       not to affect  flame heat.   Total  heat generated  by  the flame  is
       determined  using  mole  fractions,  molar  flow  rate,  and heats  of
       combustion  for methane,  ethane,   and  benzene  (see  references  of
       phys i ca1 cons tant s).

         Qt = (44.64 g-mole/m3) 6.58 m3/s [0.5(191,760 cal/g-mole)  +

              0.098 (341,260 cal/g-mole) + 0.002 (780,922 cal/g-mole)]

            = 3.84 x 107 cal/s


    3) Compute the flame tip height, Flowchart D-7 (Section 5.2.2):

                         .           _  0.478
          hf = 4.56 x 10~3 (3.84 x 107)
             = 19 m

    4) Calculate the effective release height before plume rise:

          hse = 32 m + 19 m  =  51 m


    5) Determine   the  buoyancy  flux   from  the   heat   release   rate
       (Section 5.2.3):

          F  = 1.66 x 10~5 (3.84 x 107)

             = 637.4  m4/s3


    6) Buoyancy   plume   rise   calculations   begin   at   step   3   of
       Section 5.2.3.  The distance  to  final  plume  rise  for  unstable
       conditions and F > 55 mVs3 is:

          xf  =  119 (637.4)°'4= 1575 m


    The  high buoyancy  of this  release  makes  determination  of  the  maximum

concentration very complex.   The distance to final plume  rise at  1568 m would

in most  applications  be  well  beyond the facility  fenceline  distances.   As  a

                                   6-11

-------
result, the potential exists for the maximum concentration to occur  during  the

transitional plume  rise  stage which makes the concentration  calculations  very

complex.  That is, plume rise  is  a function of wind  speed, downwind distance,

and  stability,   concentration  is  a function  of  dispersion  parameters,  wind

speed  and height  of release,  and  dispersion  parameters  are  a  function  of

stability   and   downwind   distance.   The  method  of  determining   maximum

concentration  available  in  Section  2.4  requires  numerous   iterations   to

determine the maximum concentration.   In this  instance, it  is recommended that

refined modeling,  such  as  ISC,  be  used  to determine  maximum  concentrations.

An  alternative  to  refined  modeling  or  iterative  solutions  is  the  very

conservative assumption that  the  maximum plume height  equals the  stack height

modified by the flame length.

    The following steps are included to demonstrate  calculations  for  a single

receptor  distance  (1 km)  and arbitrary meteorological condition  (5 m/s and B

stability)   typically   required   in  an   iteration   to  determine   maximum

concentration.
    7) The  receptor distance  is  less   than  xf  and  is  used  to  estimate  a
       transitional plume rise (step 4 of Section 5.2.3):


                     (637.4)°-33(1000)°-67
          Ah =  1.6  	     = 276 m
                           5 m/s

       Effective plume height is given by:

          H = 276 m + 51 m  =  327 m


    8) Dispersion  parameters  (Flowchart  D-3  and   Section  5.3.1)   for
       receptors at 1000 m are:

          Oy (1000 m) = 158 m

          az (1000 m) = 110 m
                                   6-12

-------
9) The  effects  of  buoyancy  induced  dispersion  (Section  5.4)  are
   calculated by:


                     2            ?,°-5
      oy  =  [(158 m)  + (276/3.S)2]      = 177 m

                     2            .  0.5
      az  =  [(110 m)  + (276/3.S)2]      = 135 m


10) One-hour concentrations can then be  calculated at the receptor by
    (Section 5.6.4):


                     0.9 g/s              r                 ,,
                                       exp[-0.5(327 m/135 m)2]
           3.14 (5 m/s)(135 m)(177 m)


        =  1.28 x 10~7 g/m3 = 0.128 ug/m3
                               6-13

-------
6.5 Continuous Gaseous Releases from Tanks  or Pipes

    Scenario;   In  this example chlorine gas is released from the  vapor  space

of a pressurized tank through a 2.8 cm diameter hole.



    Discussion:

    This scenario  represents  a  continuous  gaseous  release  from a  pressurized

vessel.  The eguations used in the calculations are  shown in Section 4.5.


    Source Parameters

    Chlorine gas:- molecular weight = 70.9  g/g mole
                 - temperature = ambient =  283 K
                 - pressure = 6.89 x 10& dynes/cm2
                 - ratio of specific heats  = 1.35
    minimum receptor distance = 100 m


    Sample Estimates

    Calculations   for  this   release   are   guided   by  Section  4.5   and

Flowchart C-6.  Steps in the calculations are as follows:


    1) Release calculations:

       Because  the release  is single  component,  mean  density and mean
       specific heat  ratios  are  not required.  Calculations begin  with a
       determination   of   the  emission  rate  from   the  leaking  tank.
       Selection of the equation for release rate depends on the  ratio of
       tank to atmospheric pressure:

          _ = (6.89 x 106 dynes/cm2 / 1.01 x 106 dynes/cm2)
          pa
             = 6.8

       to be compared to the conditional value:

                      1.35
          (1.35 + 1}\(1'35 1}
                                = 1.86


       Since the pressure  ratio is greater than  the  conditional value of
       1.86 generated using  the specific heat ratio of 1.35, the critical
       flow equation is used after calculating vapor density by:
                                   6-14

-------
               M Pt
          Pv =
               R* T


                70.9 g/g-mole .(6.89 x 106 dynes/cm2)

                (8.31 x 107 dyne-cm/g-mole K)   283 K


       and:
                                            =  0.0208 g/cm3
                                                                         1.35+1
qv = 0.8 (6.16 cm2)/ (1.35)6.89 x 106 dynes/cm2 •  (0.0208 g/cm3)/  2   ^1.35-1

                                                                ,1.35-fl/
   = 1261 g/s


       Prior  to density  determination (Flowchart  D-l),  the volume  flow
       rate and exit velocity  from the tank  must  be determined.   Volume
       flow  rate  can  be  determined  from  the  vapor  density  at  tank
       conditions  or,  as  in this  case,   from the  molecular  weight  and
       molar volume:

                 1261 g/s              m3      1.01 x 106 dynes/cm2   283
          v _ 	 . 0.0224 	 • 	 • 	 K
              70.9 g/g-mole          g-mole   6.89 x 106 dynes/cm2   273

            =0.06 m3/s

       Exit velocity is calculated from the leak area

                0.06 m3/s

                    A).028\2
Vs = 	   =97.5 m/s
               3.14
    2) Chlorine  is  substantially  more  dense  than  air  which  can  be
       confirmed in the  first  step of Flowchart D-l  (Section 5.1.2).   In
       the second step, a determination is made of whether  the density is
       sufficient  such  that  buoyant  effects  will  dominate  turbulent
       mixing  in the  atmosphere.   This  is  done  using  the  Richardson
       number:

                      70.9   \    0.06 m3/s
          Ri = 2722 f 	  -11
                      28.9    / (I m/s)3 0.028 m

             = 8,477

       The  value is  well in  excess of 30  indicating the  importance of
       heavy gas modeling and the RVD model (Section 5.6.3) is used.
                                   6-15

-------
    Table 6-1  provides  results for  this example  beginning  with a  listing  of




model inputs for the version available in August,  1988.   The second  portion  of




the  output  identifies those cases in which the model is applicable.   In  this




section,  a   "0"  indicates  that  the  release   is  passive  and  the   model  is




inapplicable,  a  "1"  indicates that  the gas  is  influenced by  gravitational




effects and  a "2"  indicates  that the  meteorological condition  identified  is




not  likely  to  occur.  The  determination  of  whether the gas  is affected  by




gravitational effects is made based initially on Richardson number for  which a




table  is  presented.  Model  results  are  given  in two  forms,  a  table  showing




plume  rise,   touchdown  distance,   and  touchdown   concentration   for   each




meteorological   condition  and   a   table   of   concentrations   at   specified




receptors.  In this example both of these tables are  reviewed to  determine the




maximum concentration.  Since  the fenceline is at 100 m, a review of touchdown




distances in excess of  100  m indicates that the maximum concentration  is  8.22




g/m^ and  occurs  at 125  m from the source within stability classes E and F and




2 m/s  winds.   A  review of the  table giving the  post-touchdown concentration




confirms that this concentration exceeds any fenceline value.
                                   6-16

-------
                               TABLE 6-1

                   RVD MODEL RESULTS:  CHLORINE GAS LEAK
   :hlorine  Leak Example
                                   08-05-1988

                                   Input Data
   Pollutant  emission rate (kg/sec)  =  1.261
   Exit  gas velocity (m/sec)=  97.6
   Exit  Temperature (K)=  283
   Stack Height (m)  =  5     Diameter (m) =  .028
   Pollutant  Concentration (volume %) =  100
   Exhaust Gas  Density (kg/m3)  =  3.045529
   Exhaust Gas  Molecular Weight =  70.9
   Exhaust Gas  Mass Flow Rate (kg/sec)  =  1.261
   Pollutant  Molecular Weight =  70.9
   Molar Volume (m3/mole)  =  2.328003E-02
   Release duration (sec)  =  900  Av. Time (sec) =  900
   Wind  Speeds  (m/sec)  =  1.0    2.0    4.0    6.0    8.0
   Distance  (m)  =   100
   Ambient Temperature (K)  =  283  283   283  283  283  283
   Wind  Speed Profile Exponents =  .15   .15  .2  .25  .3  .3
   (Friction  Velocity)  / (Wind Speed at z=10m)
   ".06   0.06   0.06   0.06   0.06   0.06
                                                            10.0
          Dense Gas Behavior

           Stability  Class

         123456
 Wind
Speed
  1.0
  2.0
  4.0
  6.0
  8.0
 10.0
(0=Non-Dense Behavior   l=Dense Gas Behavior
     2=Combinations that cannot occur)

             Release Richardson Numbers
1
1
2
2
2
2
1
1
1
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
2
2
1
1
2
2
2
2
 Wind
Speed
   1.0
   2.0
   4.0
   6.0
   8.0
  10.0
                   Stability Class

                        2         3
64534.0
8066.7
1008.3
298.8
126.0
64.5
64534.0
8066.7
1008.3
298.8
126.0
64.5
66809.8
8351.2
1043.9
309.3
130.5
66.8
69165.8
8645.7
1080.7
320.2
135.1
69.2
71604.9
8950.6
1118.8
331.5
139.9
71.6
71604.9
8950.6
1118.8
331.5
139.9
71.6
                               6-17

-------
                             TABLE 6-1

                 RVD MODEL RESULTS:  CHLORINE GAS LEAK
              Dense Plume Trajectory

Stability  Wind  Plume Touchdown
  Class   Speed   Rise  Distance
          (m/sec)   (m)     (m)
    Touchdown
  Concentration
(ug/m3)       (ppm)
1 1.0
1 2.0
2 1.0
2 2.0
2 4.0
3 1.0
3 2.0
3 4.0
3 6.0
3 8.0
3 10.0
4 1.0
4 2.0
4 4.0
4 6.0
4 8.0
4 10.0
5 2.0
5 4.0
6 1.0
6 2.0
9.2
7.3
9.2
7.3
5.8
9.3
7.4
5.9
5.1
4.7
4.3
9.4
7.5
6.0
5.2
4.7
4.4
7.6
6.0
9.6
7.6
65.85
140.79
65.85
140.79
305.06
63.41
135.50
293.39
464.18
644.90
833.93
61.07
130.41
282.18
446.24
619.76
801.20
125.52
271.40
58.81
125.52
0.11042E+08
0.77962E+07
0.11042E+08
0.77962E+07
0.53969E+07
0.11231E+08
0.79366E+07
0.54998E+07
0.43939E+07
0.37293E+07
0.32747E+07
0.11422E+08
0.80792E+07
0.56044E+07
0.44804E+07
0.38045E+07
0.33420E+07
0.82239E+07
0.57106E+07
0.11616E+08
0.82239E+07
0.38158E+04
0.26940E+04
0.38158E+04
0.26940E+04
0.18649E+04
0.38809E+04
0.27426E+04
0.19005E+04
0.15183E+04
0.12887E+04
0.11316E+04
0.39470E+04
0.27918E+04
0.19366E+04
0.15482E+04
0.13147E+04
0.11548E+04
0.28418E+04
0.19734E+04
0.40141E+04
0.28418E+04
                             6-18

-------
                      TABLE 6-1


          RVD MODEL RESULTS:  CHLORINE GAS LEAK
 Concentrations at Specific Receptor Distances
Stability  Wind Distance  Concentration
  Class   Speed
         (m/sec)    (m)       (ug/m3)
            1.0     100.0   0.54271E+07   0.1875E+04
            1.0    100.0   0.54271E+07   0.1875E+04
            1.0     100.0   0.51773E+07   0.1789E+04
            1.0    100.0   0.49390E+07   0.1707E+04
            1.0    100.0   0.47117E+07   0.1628E+04
                      6-19

-------
6.6 Instantaneous Gas Releases

    Scenario;  An instantaneous chlorine  discharge results from  failure of  a

transfer  line  from  a pump.   An  instantaneous  chlorine  cloud  is formed  for

which an  estimate of  the  peak  concentration at  100 m  from the  release  is

required.



    Discussion:  This  scenario  represents  any instantaneous gas  release  which

may  result  from  a  vent,  stack, pipe  or compressor  failure,  relief valve  or

similar case.   Dispersion  estimates  require characterization  of  the  cloud

primarily to determine if  calculations should be made using a dense or passive

model.  Calculations are guided by Section 4.6 and Flowchart C-7.


    Source Parameters

    - release temperature   -  283 K
    - molecular weight - 70.9 g/g-mole
    - ambient temperature -  283 K
    - transfer line: - diameter - 2 cm
                     - length in release region -3m

    Sample Estimates

    1) Because  the   release  is  single  component, calculations  of  mean
       values  do  not apply.   Begin by calculating the  mass released from
       the known  volume.   In other cases, release mass  may be  known and
       volume  will  be determined.   In this  example,  it is conservatively
       assumed  that all gas  in the  isolated pipe  segment  is  available
       instantaneously as  an emission input  to  a dispersion model.  From
       volume considerations:

          Vi =  3 m  (3.14 (0.02/2)2) = 0.0009 m3

       The release mass is determined using either a  known density or,  in
       this case, the  molar volume and molecular weight.

                (0.0009 m3)(70.9 g/g-mole)  283 K
          Qt = 	  	   =   2.95  g
                     0.0224 m3/g-mole       273 K


    2) Calculate  Richardson  number  as   a check  on  puff  density.   The
       Richardson number is given  by:

                       /70.9 g/g-mole  • 283 K     \ .0009  °-33
          Ri =  2,722  (	1_	  -  1] 	     =  391
                       \28.9 g/g-mole  • 283 K      / (1 m/s)2

                                   6-20

-------
   which indicates dense gas modeling should be used.
3) Steps  in  the  heavy gas  simulation  are  given  in  Flowchart  D-6
   (Section 5.6.2):


              14.7 (1.45 (0.0009 m3))0'5
      Rmax =  	  =  0.53  m
                       1 m/s

   where  1.45  is the  ratio  of the density difference to  the  density
   of air.  This is  identical to the parenthetical expression  in Step
   2.

   Initial conditions for  releases  of  the type  must  be assumed.   In
   this case it  is  assumed that the gas  released  forms  a hemisphere.
   An  initial   radius  for   the  cylindrical   heavy  gas  model   is
   determined  by equating the  hemispherical  and cylindrical  volumes
   and assuming the  radii are equal:


      	 irr3 =
                2
   so that h = 	 r.   Then, for the model
                3

            2      ,          /3 Vn \0.33
      V0 = 	 ir R0J and Ro =|
                               2 TT
   The initial radius is:

            3   (0.0009 m3)0'33"
                   3.14
                                = 0.077 m
   The entrained volume is:

                   , /0.53 m \L2
      V  =0.0009m3/ 	 )  =  0.009m3
                     \ 0.077 m/

      VT = 0.0009 m3 + 0.009 m3 = 0.0099 m3

   and:

              0.0099 m3
      H  =  	  =  0.0112 m
            3.14 (0.53 m)2
                               6-21

-------
4) H is less than 5 cm so Rmax is recalculated:

           = 2.52V0.0099 m3 = 0.25 m
5) The release  is sufficiently  small such  that  the  initial  spread
   radius is trivial  relative  to turbulent spreading and calculations
   can be  completed  using the  simple Gaussian  puff  model  without
   considering virtual distances.
6) Dispersion   parameters  for   this   instantaneous   release   are
   determined using Figures 5-3  and 5-4 (Section 5.3.2)  at a distance
   of 100 m  for a stability of F  (as  defined in Section  2.4).   Peak
   concentration is then estimated as indicated in Section 5.6.5:

           0.127 (2.95 g)
      C =  	  =  0.325 g/m3
           (1.2 m)2 0.8 m
                               6-22

-------
6.7 Continuous Releases of Fugitive Emissions

    Scenario;  The maximum  hourly average  concentration  estimate is  required

for ambient  ethylene dichloride at  a fenceline  receptor  100 meters  downwind

from a production facility.



    Discussion:   Normal  production  of ethylene  dichloride  in vinyl  chloride

plants results in fugitive emissions from storage and vents.   Specific sources

of the  emissions cannot  be  specified.  As  a  result, simulations make  use  of

emission factors  to  provide  average emissions  plantwide.   These  emissions  are

used in a continuous ground level area source dispersion model.


    Source Parameters

    area of emissions at the plant - 100 m x 100 m
    production rate - 204,000 Mg/yr in continuous operation over the  year


    Sample Estimates

    1) Section  4.7  and  Flowchart  C-8  guide  estimates  of  emissions.
       Emissions  are  obtained  from   emission  factors published by  EPA
       (1987b).  Plant-wide  emissions  are  calculated from the production
       rate  and an  emission factor from  various fugitive  sources.   The
       emission factor per production unit is given by:

          chlorination vent                        0.0216 kg/Mg
          column vents                             0.06   kg/Mg
          process storage vents                    0.0003 kg/Mg
          process fugitive                         0.265  kg/Mg
              Total                                0.3469 kg/Mg

       Total emissions for the plant are given by:

          g = 204 x 103 Mg/year (0.3469 kg/Mg)  = 70.77 x 103 kg/year

            = 2.24 g/s


    2) Dispersion estimates are provided by  calculations  using continuous
       area  source  equations   (Flowchart  D-4).   Fugitive  emissions  are
       generally  assumed to be  dilute  and  therefore  not  dense.    The
       virtual  source distance  (Section 5.5.1)  using the  width of  the
       square plant area is:

          oy = W/4.3 = 100 m/4.3 = 23.2 m

                                   6-23

-------
   From Figure 5-1 ,  under stable (F) conditions:

      xv = 620 m

   and:

      Xy = (620 + 100 m) = 720 m

   for receptors at 100 m.


3) Concentrations are calculated using Section 5.6.4:


      C  =  _  exp (-0.5 (H/oz)2)
              TT az (100 m) ay (720 m) u

   Since the  source  is  at ground level, H = 0 and the exponential term
                       2.24 g/s                          ,    .
      C  =  _  =  1.24 x 10~2 g/m
              3.14 (2.3 m) 25 m (1 m/s)
                               6-24

-------
6.8 Continuous Gaseous Emissions from Land Treatment

    Scenario:  Sludge  containing 1000 ppm benzene is  applied to  a  one  acre

land treatment site at  a  rate of 1  lb/ft2  and filled to a depth  of  8 inches.

Determine the one-hour  average  concentration of benzene 200 m downwind during

the first one hour  after  application.   The following  input data are  known  or

assumed:


    D     =0.027 cm2/s
    P     = 0.1244 atm
    MWoji = 142 g/g-mole (assumed to be decane)
    M     =1 lb/ft2 = 0.489 g/cm2
    hp    = 8 in = 20.3 cm
    R     =82.06 cm3  • atm/g-mole • k
    T     = 298 K
    t     = 3600 sec
    ppm   = 1000 g benzene/10^ g oil
    A     =1 acre = 40,468,564 cm2
    xr    = 200 m fenceline distance


    Discussion;    Simulations of emissions  from  land  treatment   are  handled

using the procedures  specified  in Section 4.8 and  release Flowchart  C-9.   The

resultant emission  rate,  along  with the area  of the  land treatment operation,

is then  used to  determine  virtual   distance,  dispersion  rates,  and receptor

concentration, as outlined in Flowchart D-4.



    Sample Estimates
          /
    1) Average emissions  over first hour  after application are estimated
       using the eguation of Section 4.8:


              ^(0.027  cm2/s)(0.1244  atm)(142 g/g-mole)(0.489 g/cm2)^
          E = (.
                5(20.3 cm)(82.06 cm3 atm/g-mole •  K)(298 K)(3600 s),

               (1000 ppm)  • (40,468,564 cm2) • 2 x 10~6


          E = 0.414 g/s
                                   6-25

-------
2) As  indicated  in  Flowchart  D-4,  dispersion  from land  treatment
   emissions is  treated as a  continuous, passive  area  source.   The
   virtual  horizontal   distance   (Section   5.5.1),   based  on   the
   equivalent square dimension of  the facility, is used to  determine
   horizontal  dispersion.

   •  estimate ayO

         W = (40,468,564 cm2)0-5 = 6361 cm = 63.6 m

         CTyo(W) =63.6 m/4.3 = 14.8 m

   •  determine  the  horizontal  virtual  distance  xv, using avo  and
      Figure   5-1   (assume   stable   conditions   to   ensure   peak
      concentrations from the ground level source)

         xv = .41 km = 410 m                             (F stability)

   •  calculate receptor plus virtual distance, xv

         xy = 200 m + 410 m = 610 m


3) Calculate  dispersion  parameters   (Section  5.3.1)   and  receptor
   concentration  (Section 5.6.4)

   •  determine av at  the modified horizontal receptor distance,  xv
      (Figure 5-1)

         ay (610 m) = 21 m                               (F stability)

   •  determine az at the receptor distance (Figure 5-2)

         oz (200 m) = 4 m                                (F stability)

   •  determine receptor concentration

                    0.414 g/s
         C = 	!	
              ir (21 m)(4 m)(l m/s)

         C = 0.00157 g/m3
                               6-26

-------
6.9 Municipal Solid Waste Landfill

    Scenario:  Hourly  concentration estimates are  required for  emissions  of

perchloroethylene from a municipal landfill in Ohio.



    Discussion;   Concentration  estimates  from landfills  are determined  using

either  emission  factors,  an  emissions  model  or  site-specific  measurements

(Flowchart C-1Q).  In this example, measurements  are  not available and  the  VOC

emission model in  Section  4.9  is used.   Once VOC emissions are calculated,  VOC

emissions profiles (Appendix A,  item 4)  are used  to determine what fraction of

the  total  is  perchloroethylene.   Dispersion calculations use  a  continuous

ground-level area source.


    Source Parameters

    -  Amount of  Refuse - 3.8 million tons
    -  Landfill area - 3 hectares (30,000 m2)
    -  Distance to nearest offsite receptor (100  m)


    Sample Estimates

    1) From Flowchart C-10 and Section 4.9, emissions are calculated as:

          E(g/s)  = (1 (g/s)/million tons)  3.8 million tons = 3.8 g/s

       From  the   emissions  profiles,  perchloroethylene  constitutes  0.3
       percent of total VOC emissions or 0.011 g/s.


    2) Dispersion  estimates  use the  virtual  point source,  area  approach
       with initial horizontal dispersion (Section 5.5.1) given by:

          ayo = (30,000 m2)°-5/4.3  = 40.3 m

       which,  under  stable  conditions,   results  in  a  virtual  source
       distance of approximately 1,200 m using Figure 5-1.


    3) Dispersion   parameters   (Section   5.3.1)   are   obtained   from
       Figures 5-1 and 5-2:

          Oy (100 m + 1,200 m)  = 44 m

          az (100 m) =  2.3 m
                                   6-27

-------
4) The maximum estimated concentration  (Section  5.6.4) is  calculated
   from:

                  0.011 g/s                       _     ,
      C = 	  =  3.46 x 10~5  g/m3
          3.14 (44 m)  (2.3  m)  1 m/s
                               6-28

-------
6.10 Continuous Emissions from an Herbicide




    Scenario;  2,4-DB butoxy ethanol ester, a restricted herbicide,  is  applied




to  a  farm  field of  four  acres.   Maximum  post-application one-hour  average




concentrations are to  be estimated.   The  property boundary  is  located  100  m




from the edge of the study field.








    Discussion;   Pesticide/herbicide  dispersion  estimates  can   usually  be




obtained  by   using  an  emission  factor  and  a  ground-level,   area   source




dispersion model.  Emission factors are  often difficult to  obtain but may  be




available  from  the  technical  literature  or  state  agricultural  agencies.




Unless emissions estimates are sensitive to  meteorological  conditions,  maximum




short-term   calculations   will  be   controlled   by  dispersion.    The  worst




dispersive  conditions  for  ground  level  area  sources  occur under  low  wind




speed, stable conditions.








    Source Parameters




    field size - 14 acres (approx. 56,660 m^)






    Sample Estimates




    Section  4.10  and Flowchart  C-ll  indicate that  emission  factors  must  be




specified  since  the  evaporation   rates  of  herbicides  vary  widely based  on




composition and sources for emission factors are not  generally  available.  The




following provides an example from a typical data source.   The best emissions




estimates for  2,4-DB butoxy ethanol ester were found  in results  of inventory




and  flux   studies  performed  by  the  California  Department  of  Food  and




Agriculture, a typical  data source.  In one  study,  they determined  that half




of  the  applied pesticide  consisted of the  active ingredient.   Losses  through




evaporation  were  approximately  equal   for  the   first  two  months  after






                                   6-29

-------
application at  a rate of approximately one  third of the applied material  per

month.  Biodegradation and sequestration were not found to be  significant.   In

the   sample   study,   data  were   provided  on   total   regional  application,

application  rate  and  acreage   of  application.   An   emission  factor   of

approximately  1.2  Ib  active  ingredient  per acre  per  month  or 1.87 x  10"^

g/h-m2 was also determined.
    1) Total emissions for this  example  are given by the  emission factor
       times the field area:

          g = 56,660 m2 (1.87 1CT4 g/h-m2}  = 10.6 g/h =  2.9 x 10~3 g/s
    2) Flowchart  D-4  indicates  that  concentration  estimates  for  this
       example are  made with passive  area  source  dispersion  equations.
       Worst case meteorology for  area source estimates is represented by
       low  wind  speeds,  stable  (F)  conditions.    For  this  example  a
       virtual  source  distance  (Section  5.5.1)  for  a  square  field  is
       given from:

          W  = (56,660 m2)0-5  =  238 m

       resulting in:

          cryo = 238 m / 4.3  =  55 m

       From  Figure  5-1,   the  distance  at  which  a  stable  horizontal
       dispersion parameter is 55 m is:

          xv  = 1700 m

       To determine the centerline  concentration  at 100 m  from  the field
       edge:

          Xy — \Xj-* T X^)

             = (219 m +1700 m)  = 1919 m

       where xr  includes  the distance from the field  center  to the edge
       summed with the distance to the property line (100 m)
    3) The horizontal dispersion  parameter  is determined from  Figure 5-1
       (Section 5.3.1).
          oy (1919 m) = 62
m
                                   6-30

-------
   The vertical  dispersion  parameter is  determined at the  receptor
   distance from Figure 5-2.


      0Z  (219 m)  = 4.4 m


4) Concentration (Section 5.6.4) is given by (with H = 0):


                       Q
      C =
            ir oz(219 m) oy{1919 m) u
                 2.9 x 1CT3 g/s                      ,    .
                                        =  3.39 x 1CT6 g/m3
            3.14 (4.4 m)  (62 m)  1 m/s
                               6-31

-------
6.11 Equipment Openings

    Scenario:   A common source of  emissions  due  to  equipment  openings  is  found

in the  production  of  coke  where  opening of  the ovens  at the  completion  of

processing results  in  a near  instantaneous  release.  One  toxic component  of

the emissions  is toluene.   It is  desired  to estimate  a 15  minute  average

concentration at distances  beyond  50  m downwind of the  source.



    Discussion:  Emissions from coke  ovens  result primarily from charging  and

discharging operations and fugitive  losses  which occur on a  continuous basis.

The example  presented  is for  the near  instantaneous  emissions  which  result

from  discharging   the  completed   coke   through   the  oven  doors.    Sample

simulations are  based on  the impact of  a single  furnace  although   in  real

applications  total  emissions  from a  battery of  ovens  over time would  be more

typical.

    Simulations  require  determination of an  emission  factor for the  oven and

the total  emissions based on  oven  capacity.   Dispersion  estimates  are  made

assuming that  the  release is  instantaneous  and  within the wake  cavity formed

by the oven battery.



    Source Parameters

    oven door dimensions:  - height =  5m
                          - width  = 40 m
    coke production per oven = 20  Mg on an 18-hour cycle


    Sample Estimates

    1) Section  4-11  and Flowchart  C-12  guide the  user through  typical
       calculations for  equipment openings.   To begin,  the  emissions  are
       estimated  using  emission  factors  (EPA,   1987b).   Total  toluene
       emissions  from  coke  production  are  0.48   Ib/ton  (0.24  g/kg).
       Emissions from  door openings  must be  approximated.   From  AP-42,
       coke pushing  emissions  account for approximately  three  percent of
       VOC  emissions.   Coke   pushing   emissions   are   then   given   by
       multiplying the emission factor times the  production rate:

                                   6-32

-------
      Q  =  0.24 g/kg coke (0.03)  20,000 kg coke = 144 g toluene


2) Flowchart D-6  shows  the  path  for dispersion  estimates  if  the
   release  amount  is  known.   Because  the  release  is  heated  and
   insufficient data on component  density are available, the  cloud is
   treated  as   passive  for  this  example.   Estimates   of  volume
   dispersion  parameters   require  some  assumptions  on  the   initial
   release  dimensions  (Section 5.5.2).   For this  example,  the  coke
   oven door dimensions are assumed:

      oz =  Hv / 2.15  =  5  m/ 2.15 = 2.3 m
      oy =  W  / 4.3   =  40 m/ 4.3  = 9.3 m

   From Figures 5-3 and 5-4, under stable conditions:

         Xyy = 994 m
         xvz
   and:
         xvz = 542 m
         xv  = (994 + 50 m) =  1044 m
         xz  = (542 + 50 m) =   612 m

   for receptors at 50 m.
3) Concentrations are calculated using Section 5.6.5 (with H = 0):

      Oy  =  9.7 m
      oz  =  2.5 m

               0.127 (144 g)                    _
      C =  	  =  0.078 g/m3
             2.5 m (9.7 m )2


4) From Appendix  E,  the  15 minute  average  (assuming a  1 m/s  wind
   speed) is given by:

          (900) (1 m/s)
      N = 	 = 46 giving A = 1
            2 (9.7 m)
   and
            A - 0.5               _
      F = 	 = 2.73 x 10~2
          46 (0.3989)

   and the average peak concentration is:

      Cavg = 2.73 x ID"2 (0.078 g/m3) = 2.13 x 10~3 g/s
                               6-33

-------
6.12 Continuous Gaseous Emissions  from Surface  Impoundments

    Scenario;    One-hour  concentration estimates  of benzene are  desired at  a

receptor  located  200  km  downwind  of a  surface  impoundment.   The  following

known data are relevant to the simulation:


    Source Parameters

    C0 = 1000  g/m3
    H  = 5.5 x 10~3 atm-m3/g-mole
    A  = 1500  m2
    F  = 43.7  m
    D  = 1.8 m
    Q  = 0.0016 m3/s
    xr = 200 m fenceline


    Discussion;  Estimates of emissions from impoundments are  determined using

the  procedures specified in  Section  4.12 and  Flowchart  C-13.   The  first

equilibrium constant  equation is used because  the Henry's Law Constant  for

benzene  in  water  is  available.   The  resultant  emission rate,  along with the

area of the surface  impoundment,  is then used to  determine  horizontal virtual

distance,  dispersion  rates,  and  receptor  concentrations,  as  outlined  in

Flowchart D-4.


    Sample Estimates

    Emission Calculations - Quiescent Case

    1) Determine  the  equilibrium  constant  with  the  first  form  of the
       equation:

          Keq = (40.9)(5.5 x 10~3) - 0.225


    2) Calculate the gas phase mass transfer coefficient

          kg =  {1.26 x 10-2)(1500)-°-055 = 0.008 m/s


    3) Determine the liquid phase mass transfer coefficient

       •  determine  the  k^ equation  to use based on the  fetch to  depth
          ratio:
                                   6-34

-------
                     43.7 m
                    	 =  24.3
             D       1.8 m


       •  Based on  the  criteria in  Section 4.12, the second  k^  equation
          applies:

                              .    43.7               ,              ,
             ki  =  6.84 x 1CT8  	 + 3.35 x 1CT6  = 5.01  x 1CT6 m/s
                                   1.8
    4) Calculate the overall mass transfer coefficient

                    1                    1        s
          Kg=|
=  5.0 x 10~6  m/s
                5.01 x 10~6       (0.008)(0.225)   )


    5) Determine the bulk concentration in the impoundment:


                        (0.0016 m3/s)(1000 g/m3)                       ,
          CL = 	 = 175.8 g/m3
                (5.0 x 10~6 m/s)(1500 m2) + (0.0016 m3/s)

    6) Calculate the area source emission rate:

          E = (5.0 x 10~6 m/s)(175.8 g/m3)(1500  m2) = 1.32 g/s


    Emission Calculations - Aerated Case

    An  aerated  impoundment  example  is  presented  as  a modification  to  the

quiescent  case.   Assume  the  impoundment  has a  single 75  horsepower aerator

(i.e., POWR =75), aerating half the impoundment.   Repeat steps 1 through 6.


    7) Estimate the turbulent  liquid-phase and  gas-phase  mass  transfer
       coefficients.

          kla = 0.2623 (75/U500 m2(0.5» = 0.0262

          kga = 0.021 (75/1)0-4 = 0.1181


    8) The overall  turbulent and complete  mass transfer coefficients are
       given by:

                /i               i
          Kt  ={ 	  +  	= 0.0132 m/s
                \0,0262     0.225 (0.1181)
                                   6-35

-------
          K   =  0.0132 (0.5)  + (1-0.5)  5.0 x 10~6  = 0.0066  m/s


    9) Emissions are calculated as:

          E = (0.0066 m/s) 175.8 g/m3 (1500 m2)

          E = 1,740 g/s

       Aeration provides a significantly higher emission rate.


    Dispersion Calculations

    As  indicated   in   Flowchart  D-4,   dispersion  from  surface  impoundment

emissions  is  treated   as  a  continuous,  passive   area  source.   The  virtual

horizontal   distance,   based   on  the  equivalent  square  dimension   of  the

impoundment,  is  used  to  determine  horizontal   dispersion   while  vertical

dispersion  is  handled as  with a point  source.   Concentrations  are estimated

with the continuous point source Gaussian equation.

    The following  demonstrates dispersion for the  quiescent example.   For the

aerated impoundment, dispersion estimates will differ only by  the  area source

emission rate.
    1) Determine  the  horizontal  virtual  distance   for  the  quiescent
       impoundment (Section 5.5.1):

       •  estimate avo:

             W = (1500 m2)0-5 = 38.7 m

                    38.7

                    4.3
ayo = 	  =  9 m
       •  Determine  the horizontal  virtual  distance,  xv, using  ovo and
          Figure   5-1   (assume   stable   conditions   to  ensure   peak
          concentrations from the ground level source):

             xv =  .24 km = 240 m                             (F Stability)


       •  calculate receptor plus virtual distance, xv

             X  =  200 m + 240 m = 440 m
                                   6-36

-------
2) Determine  dispersion  parameters  (Section  5.3.1)  and  calculate
   receptor concentration (Section 5.6.4):

   •  determine ay,  at the modified horizontal  receptor  distance,  xy
      (Figure 5-1)

         ay (440 m) = 16 m                               (F Stability)


   •  determine az at the receptor distance (Figure 5-2)

         az (200 m) = 4 m                                (F Stability)


   •  determine receptor concentration (where H = 0)

                     1.32 g/s
         C =
               ir (16 m)(4 m)(l m/s)

         C =  6.6 x 10~3 g/m3
                               6-37

-------
6.13 Relief Valve Discharge (Two-Phase)

    Scenario;  The relief  valve  scenario represents  estimates  of the  maximum

1-hour  concentration  resulting  from  a  two-phase   mixture  of  chlorine  and

suspended chlorine droplets.



    Discussion;   Two-phase   releases  can  result   from  both   relief  valve

discharges and liquid  releases  from pressurized tanks.  Release  estimates  for

relief valves  must  be  specified,  while tank releases can be calculated.   The

example shows the use of the RVD model  for a heavy gas.  Simulation  using  RVD

requires careful  definition  of all  input parameters  to represent  the  density

of the liquid/vapor mixture.   As such, it is possible  to use the  one-phase  RVD

model  to  simulate  aerosol  (droplet)  dispersion by providing  as  input  the

"equivalent" molecular  weight of  the aerosol  mixture.  However, the accuracy

of such a simulation has not been evaluated.


    Source Parameters

    Chlorine gas: - molecular weight =70.9 g/g mole
                  - temperature = 249 deg. K
                  - release rate = 3840 g/s
                  - exit velocity = 30.4 m/s
                  - exit diameter = 20 cm
                  - release height = 10 m
                  - fraction of release in liquid phase = 20 percent
    nearest receptor distance = 50 m
    ambient temperature = 283 deg. K
    R*  =  8.31 x 107 dyne cm /(g-mole K)


    Sample Estimates

    The  methods   for  estimating  concentrations for  a  two-phase  release  are

presented  in Section  4.13 and  Flowchart  C-14.  Due  to  design differences in

chemical plant processes,  a  generic method  of  obtaining emissions  and release

parameters is not available and these parameters must  be supplied by  the user.
                                   6-38

-------
1) Because the flashed  liquid  fraction is specified for this example,
   the  first  step  in  calculations  is  a  determination   of   mean
   parameters of the  released  gas  liquid/stream.   The mean density in
   this example is given by:
      Pm =
            (0.2/0.0034) + (0.8/3.214)

         = 0.01693 g/cm3

   where (from the ideal gas law):

            70.9 g/g-mole (1.01 x 106 dynes/cm2)                 _
      pv = 	  =  0.0034 g/cmj
           (8.31 x 107 dyne cm/(g-mole K))  249 K


   and from references on physical  parameters:

      Pl = 3.214 g/cm3


2) An equivalent molecular weight must also be  calculated:

               Pm R* T
      MWe  =   	
                  P

                            -  /         ., dyne-cm  \
               (0.01693 g/cm3) / 8.31 x 107 	 \ (249 K)
                               y           g-mole K /

                           1.01 x 106 dynes/cm2

           =  346.8 g/g mole

   Volume  release  rate can  be estimated using  the mean density and
   mass rate.

           q       3840 g/s               _   _          _
      V = 	 = 	_	 = 2.27 x 105 cm3 = 0.227 m3
           pm    0.01693 g/cm3


3) Dispersion calculations  begin with  a  check  of  the  importance  of
   buoyant  effects  (Flowchart  D-l).   A  ratio  test   (Section  5.1.2,
   Step  2) indicates  the  plume  is  dense  relative  to  air.   The
   Richardson number test provides  the same result:

                    Ms Ta
      Ri  =  2722 /	-1
                   28.9 Ts
                  V

                  /346.S g/g-mole 283 deg.K   \    0.227 m3
          =  2722 /	,	__ -1\	
                  \ 28.9 g/g-mole 249 deg.K   / (1 m/s)3 0.2 m

          =  39,046


                               6-39

-------
    Table 6-2 provides RVD  results  (Section 5.6.2) for this  example  beginning




with a  listing  of model  inputs.  The  second portion of the  output  identifies




those  cases  in which  the  model  is  applicable.   In  this  section,   a  "0"




indicates  that  the release  is passive  and the model  is  inapplicable,  a  "1"




indicates  that  the  gas   is  influenced  by  gravitational  effects  and  a  "2"




indicates  that   the  meteorological  condition  identified  is  not  likely  to




occur.    The determination  of  whether  the  gas  is  affected by  gravitational




effects  is made  based initially on Richardson number for  which  a table  is




presented.  Model  results are  given in two  forms,  a  table showing plume rise




and touchdown distance and  concentration for each meteorological condition and




a  table  of concentrations  at  specified  receptors.   In this  example both  of




these tables  are  reviewed  to  determine the maximum  concentration.   Since the




fenceline  is  at 50  m, a review of touchdown  distances in excess  of 50  m




indicates that  the maximum  concentration is 4.26  g/m^  and occurs in stability




classes  B through E  with 4 m/s winds at  71 m.   A review of the table giving




post-touchdown  concentrations  confirms  that  this concentration exceeds  any




fenceline value.
                                   6-40

-------
                            TABLE 6-2

             RVD MODEL RESULTS:  CHLORINE TWO-PHASE RELEASE
   Chlorine Relief Valve Example
                                           08-16-1988
                                   Input Data
                                               .2
   Pollutant emission rate (kg/sec) =  3.84
   Exit gas velocity (m/sec)=  30.6
   Exit Temperature (K)=  249
   Stack Height (m)  =  10     Diameter (m) =
   Pollutant Concentration (volume %) =  100
   Exhaust Gas Density (kg/m3) =  17.08235
   Exhaust Gas Molecular Weight =  349.9
   Exhaust Gas Mass Flow Rate (kg/sec) =  3.84
   Pollutant Molecular Weight =  70.9
   Release duration (sec) =  900  Av. Time (sec) =  900
   Release pressure (atm) =  4
   Wind Speeds (m/sec)  =  1.0    2.0    4.0
   Distance (m)  =  50
   Ambient Temperature (K) =  283  283  283  283
   Wind Speed Profile Exponents =  .15  .15  .2
   (Friction Velocity)  /  (Wind Speed at z=10m)  =
   0.06   0.06   0.06   0.06   0.06   0.06
                                                6.0
                                                   283
                                                  .25
8.0

 283
.3  .3
10.0
          Dense Gas Behavior

           Stability  Class

         123456
 Wind
Speed
  1.0
  2.0
  4.0
  6.0
  8.0
 10.0
(0=Non-Dense Behavior   l=Dense Gas  Behavior
     2=Combinations that cannot occur)

             Release Richardson Numbers
1
1
2
2
2
2
1
1
1
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
2
2
1
1
2
2
2
2
 Wind
Speed
   l.C
   2.C
                   Stability Class

                        2         3
   4.0
   6.0
   8.0
  10.0
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
38897
4862
607
180
76
38
.9
.2
.8
.1
.0
.9
                              6-41

-------
                           TABLE 6-2

           RVD MODEL RESULTS:  CHLORINE TWO-PHASE RELEASE
              Dense Plume Trajectory

Stability  Wind Plume Touchdown
  Class   Speed  Rise  Distance
          (m/sec)   (m)      (m)
    Touchdown
  Concentration
(ug/m3)      (ppm)
1
1
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
5
5
6
6
1.0
2.0
1.0
2.0
4.0
1.0
2.0
4.0
6.0
8.0
10.0
1.0
2.0
4.0
6.0
8.0
10.0
2.0
4.0
1.0
2.0
13.4
12.6
13.4
12.6
10.0
13.4
12.6
10.0
8.7
7.9
7.4
13.4
12.6
10.0
8.7
7.9
7.4
12.6
10.0
13.4
12.6
13.27
32.83
13.27
32.83
71.84
13.27
32.83
71.84
114.43
159.78
207.45
13.27
32.83
71.84
114.43
159.78
207.45
32.83
71.84
13.27
32.83
0.11362E+08
0.62434E+07
0.11362E+08
0.62434E+07
0.42608E+07
0.11362E+08
0.62434E+07
0.42608E+07
0.33715E+07
0.28414E+07
0.24811E+07
0.11362E+08
0.62434E+07
0.42608E+07
0.33715E+07
0.28414E+07
0.24811E+07
0.62434E+07
0.42608E+07
0. 11362E+08
0.62434E+07
0.39263E+04
0.21575E+04
0.39263E+04
0.21575E+04
0.14723E+04
0.39263E+04
0.21575E+04
0.14723E+04
0.11651E+04
0.98188E+03
0.85737E+03
0.39263E+04
0.21575E+04
0.14723E+04
0.11651E+04
0.98188E+03
0.85737E+03
0.21575E+04
0.14723E+04
0.39263E+04
0.21575E+04
                            5-42

-------
                     TABLE 6-2

      RVD MODEL RESULTS:  CHLORINE TWO-PHASE RELEASE
 Concentrations at Specific Receptor Distances

Stability  Wind Distance Concentration
  Class   Speed
         (m/sec)    (m)       (ug/m3)       (ppra)


    1       1.0      50.0  0.11922E+07   0.4120E+03
    1       2.0      50.0  0.30538E+07   0.1055E+04


    2       1.0      50.0  0.11922E+07   0.4120E+03
    2       2.0      50.0  0.30538E+07   0.1055E+04


    3       1.0      50.0  0.11922E+07   0.4120E+03
    3       2.0      50.0  0.30538E+07   0.1055E+04


    4       1.0      50.0  0.11922E+07   0.4120E+03
    4       2.0      50.0  0.30538E+07   0.1055E+04


    5       2.0      50.0  0.30538E+07   0.1055E+04
    6
    6
1.0     50.0  0.11922E+07   0.4120E+03
2.0     50.0  0.30538E+07   0.1055E+04
                      6-43

-------
6.14 Two-Phase Instantaneous Release

    Scenario;  An  instantaneous  pressurized chlorine  discharge  results  in  a

flash vaporization  forming an instantaneous cloud  of  65 percent vapor and  35

percent suspended droplets.   An  estimate of the peak  concentration  at 100  m

from the release is required.



    Discussion:  Release  estimates for  two-phase  instantaneous  releases  are

beyond the scope of this  workbook and data must be  specified (Section 4.14 and

Flowchart  C-15).   Techniques   followed  for   dispersion   estimates   require

characterization of the cloud dilution and density prior to estimates  using a

dense  or passive  model.    Evaporation  of  droplets  in  the cloud  cannot  be

simulated using the workbook.  As a result, cloud density is assumed to change

only by dilution.


    Source Parameters

    - released mass - 50 kg
    - molecular weight - 70.9 g/g-mole
    - ambient temperature - 283 K
    - chlorine boiling point - 239 K

    Sample Estimates

    1) Calculate  the  density of the  initial  cloud  assuming the  cloud
       volume  is  attributable only  to vapor  and the  temperature  is the
       boiling temperature  of chlorine.   The vapor mass  is  65  percent of
       total  mass   (50 kg)  or 32.5 kg.   The  cloud  volume  is  calculated
       from the mass of vapor using the molar volume:


                32,500 g              ,          239 K
          V = 	   0.0224 m3/g-mole  	
              70.9  g/g-mole                      273 K

            = 9.0 m3

       and the  mean density can  be  calculated  as  in  Section  4.13  or by
       simple volume relationships:

          pm = mass/cloud volume

          pm = 50,000 g/9.0  m3 =   5556 g/m3


                                   6-44

-------
2) Mean molecular weight  for  an instantaneous release (Section 5.1.1)
   is  a parameter  which  would  normally  be  specified.    A  simple
   calculation  for  this  example  can be  performed  by relating  th«
   total cloud density  (vapor  + droplets)  and the cloud molar volume
   to obtain the mean molecular weight;

                   ,          ,         239 K
      Ms = 5556 g/mj (0.0224 nv-Vg-mole)  	  =  109 g/g-mole        8
                                        273 K
   Richardson number is given by:


                 /109 g/g-mole 283 K
      Ri =  2,722|
                                    - 1
                                          (9.0 m3)

                 .9 g/g-mole 239 K      / (1 m/s)2

which indicates dense gas modeling should be used.
                                                  0.33
=  19,481
3) Steps in  the heavy  gas  simulation are given in  Section 5.6.2 and
   Flowchart D-6:
              14.7

              1 m/s
                   (3.46)(9.0 m3)
                                 0.5
                                     = 82 m
   where 3.46 is  the  ratio of the density  difference to  the  density
   in air.  This  is  identical to the parenthetical  expression in the
   Richardson number (Step 2).

   Initial conditions for  releases  of this type must  be  assumed.  In
   this  case,  it   is   assumed  that  the   gas  released  forms  a
   hemisphere.  An initial  radius  for the cylindrical heavy gas model
   is  determined  by  equating  the  hemispherical   and  cylindrical
   volumes:
          irr3 = irr2h
   so that h =
                r.  Then for the model
      vo =
            ir R03 and Ro =i
                              '3 Vn\0'33
                               2 ir
The initial radius is:


         3   (9 m3)10-33

         2    3.14
                           = 1.62 m
                               6-45

-------
   The entrained volume is:

                       r
                  1.62 ml
      VT = 9.0 m3 + 998.6 m3 = 1,007.6 m3
         _ / 82 m i             _
V = 9.0 m3 /	\  =  998.6 m3
   and:
              1,007.6 m3
      H  =  	  =  0.05 m
             3.14 (82 m)2
4) Instantaneous  dispersion  estimates  are  calculated  for  using  a
   horizontal virtual point  source  approach for an area  source since
   the cloud depth is small (i.e., xz is insignificant):

      Xyy = (23.26 • 82 m)1-12 = 4721 m

      Xy.  = (100 m + 4721 m) = 4821 m


5) Dispersion parameters (Section 5.3.2) are  obtained for F stability
   and used to determine concentration (Section 5.6.5):

      0y (4821 m) = 38 m

      crz (100 m) = 0.8 m

           0.127   50,000 g           _
      C =  	  =5.5 g/m3
            (38 m)2 0.8
                  m
                               6-46

-------
6.15 Liquid Release from a Pipe

    Scenario;   Concentrations are  estimated  from  a  transfer  pipe  failure

between  unsymmetrical  dimethyIhydrazine  tanks  resulting  in  an  unconfined

liquid pool.



    Discussion;  This  scenario represents  cases where liquid  is  released from

pipes  on  the  ground  and,  due  to  its  low  volatility,   pooling  occurs.

Evaporation  of  the liquid  results  in formation of vapor  for  which continuous

area source techniques can  be  used  to simulate dispersion.  It  is  assumed for

screening  that  flow  in the  pipe  is continuous  and frictionless.  The  pool

evaporation rate is assumed to reach  a steady state after spreading  such that

the evaporation rate equals the pipe flow rate.


    Source Parameters

    - hydrazine (UDMH):  density - 0.786 g/cm3
                         molecular weight - 60.1 g/g-mole
                         vapor pressure - 19.88 x 104 dynes/cm2
    - flow rate  - 0.0001 m3/s
    - ambient temperature - 283 K
    - minimum receptor distance - 100 m


    Sample Estimates

    1) The  path through calculations  is presented  in  Section 4.15  as
       shown in Flowchart C-16.   Estimates are calculated separately for
       pool  spreading  and  evaporation  and  subsequent dispersion.   The
       liquid release rate is given by:

          qi = 0.786 g/cm3 (100 cm3/s) = 78.6 g/s


    2) Evaporation is  calculated using an  intermediate parameter  of the
       evaporation model  to determine  the  area at which  the  evaporation
       rate equals the release rate:

           1.54 x 10~4 (1 m/s)°'78 (60.1 g/g-mole)0'67 (19.88xl04 dyne/cm2)
       B = 	
                                        283 K

         =  1.683
                                   6-47

-------
           78 6
         i /o.b            _
      A =(	I  = 58.8 m2
           1.683
          ^


   Since the spill  is continuous and unconfined, no check  is  required
   for determining  maximum area (Section 4.15).
3) The areaoof the evaporating pool is used in area  source  dispersion
   calculations  using Flowchart  D-4.   The  virtual source  distance
   (Section 5.5.1) is obtained assuming a square  pool:
      W  = J58.8 m2 = 7.67 m
      ay =  W/4.3  =  7.67 m/4.3  =  1.78 m


   From Figure 5-1, under stable conditions,:



   and:
xv = 70 m
         xy = (100 + 70 m) =  170 m


   for receptors at 100 m.



4) Dispersion parameters  from  Section 5.3.1 are used  for  F stability
   to determine concentration in Section 5.6.4 (with H = 0).


      cry = (170 m) = 7 m


      oz = (100 m) = 2.3 m


                    78.6 g/s                      _
      C =  	  =  1.55 g/m3
            3.14 (7 m) 2.3 m (1 m/s)
                               6-48

-------
6.16 Low Volatility Liquid Releases from Tanks

    Scenario;   Maximum  concentrations  are   estimated   from  a  leak  in  an

unpressurized tank containing unsymmetrical dimethylhydrazine (UDMH) .



    Discussion;  This  example represents  a  scenario where  liquid is  released

from  a  storage  tank and  pools on the ground.   Evaporation  of  the  liquid

results in a plume  of  vapor for which continuous area source techniques can be

used to simulate dispersion.


    Source Parameters

    UDMH: - density - 0.786 g/cm3
          - molecular weight - 60.1 g/g-mole
          - vapor pressure - 19.88 x 10^ dynes/cm^
          - tank pressure = atmospheric pressure

    leak: - 100 cm below liquid level
          - 8 cm2 area
    meteorology - wind speed - 2 m/s
                - temperature - 283 deg. K
                - stable (F)
    receptor distance - 100 m (fenceline)
    impoundment area - 2500 m2


    Sample Estimates

    The   path   through  calculations   is  presented   in   Section   4.16   and

Flowchart C-17.   Estimates  are  calculated separately  for  release  from  the

tank, pool spreading and evaporation and subsequent dispersion.


    1) Using the release model, the liquid release rate is calculated as:
             = 0.8 (8 cm2) 0.786 g/cm3  /1960 (100 cm) + 0

             = 2227 g/s
    2) Evaporation is  calculated using  an intermediate parameter  of the
       evaporation model  to determine  the area  at which the evaporation
       rate equals the release rate (as described in Section 4.15):
                                   6-49

-------
   n   1.54 x 10~4 (1 m/s)°-78  (60.1  g/g-mole)0-67  (1.988  x  105  dyne/cm2)
   a —
                                     283  K

     = 1.683


         /2227\1'°*       ,     •
      A =/  	 \    =  2037 m2
         I  1.683 I

   The area of the evaporating  pool  is the smaller of  the impoundment
   area (2500  m2)  and  the  area  at  which evaporation  across  the  pool
   equals   flow  into  the  pool   (2037  m2).    Therefore,  the   vapor
   emission rate equals the  liquid release rate:

      qv =  q1 - 2227 g/s

   Alternatively, qv can be  calculated as in Section 4.15.
3) Dispersion calculations  follow  using Flowchart  D-4.   The  virtual
   source  distance  using  the  width  of   a  square  impoundment  is
   (Section 5.5.1):

      W    = /2037   = 45.1 m
      oyo  =  W/4.3  = 45.1 m/4.3  =  10.5 m


   From Figure 5-1, under stable (F) conditions,:

        xv = 280 m

   and: xv = (280 + 100 m) = 380 m


4} Dispersion   parameters  are   determined   for   stability  F   in
   Section 5.3.1.  Concentrations are  calculated  using  Section 5.6.4
   (with H = 0):

      ay (380 m) = 14 m

      az (100 m) = 2.3 m

                    2227 g/s                       -
      C =  	  =  22.01 g/m3
            TT (14 m) (2.3 m) (1 m/s)
                               6-50

-------
6.17 High Volatility Liquid Release  from a  Pipe

    Scenario:    Aminomethane liquid  is  released  from a  small  pipe,  and  a

concentration estimate is  required  for  a site with a minimum receptor distance

of 100 m downwind.



    Discussion;   The  high volatility liquid  release is intended  to  represent

calculations for materials  which when  released  immediately volatilize and  are

airborne.  Estimates of the liquid  release rate are required,  but  the material

can be considered  gaseous at the  source.   A continuous  dispersion  model  can

then  be  used  to  estimate  downwind   dispersion.    For   pipe  releases,   the

screening technique is to assume  that   the  volume rate of release equals  the

emission to the  atmosphere and is  determined by  the flow volume in  the pipe.


    Source Parameters

    - liquid characteristics - density   - 0.69 g/cm^
                             - molecular weight  - 31 g/g-mole
                             - boiling  point - -6.3 deg. C
    - liquid mass flow rate - 9,8  g/s
    - wind speed - 1 m/s
    - ambient temperature  - 283 K


    Sample Estimates

    1) A technique  for estimating emissions  is provided  in  Section  4.17
       and Flowchart C-19.  The release rate is  assumed to equal the  pipe
       flow rate:

          qi =  9.8 g/s

       In  this   example,  it  is   assumed  that   aminomethane   has   a
       sufficiently  high  vapor pressure  that  evaporation  from  a  small
       leak   is   instantaneous  and  no   pooling   results.    Such   an
       instantaneous  vaporization  results  in   a vapor  cloud  which  is
       roughly at the boiling point  of  the  liquid.   This  assumption could
       be checked using techniques  presented in  Section 4.15.

    2) Simulations  of  dispersion   follow  Flowchart D-l  for   passive
       continuous  releases.  A  check  of  plume  density  shows  that  the
       release at the boiling temperature is slightly more dense than air
       (Section  5.1.2).   Calculation  of  the   Richardson  number  as  a
                                   6-51

-------
   density check requires a  determination  of the volume release  rate
   using the  molar  volume  and molecular weight:

           9.8  g/s    (0.0224 m3)   (267 K)
      V = 	!	
          31  g/g-mole    g-mole      (273 K)

        = 6.9 x 10~3  m3/s


   and effective diameter:
         /4 (6.9 x 10~3)
      d =/	.	   = 0.094 m
        \  3.14 (1 m/s)


   The Richardson number  is given by:

                 (31 g/g-mole)  (283 K)    \     0.0069  m3/s
      Ri =  2722 / 	  - 1
                 28.9 g/g-mole (267 K)     /  (1 m/s)3 (0.094 m)

         =  27

   A value of less  than 30 confirms that  the  release is  essentially
   passive.
3) Since,  in  this   scenario,   no   pooling   occurs,  point   source
   dispersion calculations  (Flowchart D-2) can  be used  to  calculate
   concentration.   No information  is  available  on the setting  of  the
   leak  relative  to  the tank  dimensions or other structures  so that
   the  estimate  cannot  evaluate   the  effects of  downwash.   In  the
   absence of  this  information, the  most conservative assumption is
   to consider the source a continuous ground level point  source.   In
   addition,  because  the  release  is  not heated,  no buoyancy plume
   rise occurs (Section 5.2.3).
4) Point  source  dispersion  calculations  (Flowchart  D-3)  begin with
   specification  of  dispersion  parameters  (Section  5.3.1)  for  F
   stability:

      Oy (100 m) = 4 m

      oz (100 m) = 2.3 m
5) Because  there  is  no  plume  rise,  buoyancy  induced  dispersion
   (Section  5.4)  does  not  apply.    Concentration  is  estimated  by
   (Section 5.6.4 with H = 0):


                  9.8  g/s                     .
      C = 	_	  = 0.339 g/m3
          3.14 (4 m) (2.3 m) 1 m/s
                               6-52

-------
6.18 High Volatility Liquid Releases from Tanks

    Scenario:  Aminomethane  liquid is  released  from a  minor tank  leak' and a

maximum  concentration  estimate  is required  for a  fenceline  receptor  100  m

downwind.



    Discussion;  The high  volatility  liquid release  is  intended  to represent

calculations  for materials  which, when released,  immediately evaporate  (no

pooling  results).   In  this  example,  it  is assumed  that  aminomethane  has  a

sufficiently  high  vapor  pressure  so  that evaporation  from  a small leak  is

instantaneous  and  no pooling  results.   Estimates of  the  liquid  release  rate

are required,  but  the  material can be  considered gaseous  at the source.  This

scenario  shares  many  of  the  calculations  found  in   example  Section  6.17.

Differences  lie  only  in  the  release  and emission  calculations.   In  both,

steady-state release estimates  are used in a  dispersion model.   In  a  refined

analysis,  the  examples would  show different  release rates  as  pressure  in the

transfer line  or tank  is  reduced.  A continuous  dispersion model  can then  be

used to estimate  downwind dispersion.


    Source Parameters

    liquid characteristics - density  - 0.69 g/cm^
                           - molecular weight - 31 g/g-mole
                           - boiling point 	6.3 deg. C (267 deg.  K)
    tank pressure - 2.03 x 10^ dynes/cm^
    atmospheric pressure - 1.01 x 10^ dynes/cm^
    ambient temperature - 283 deg. K
    release depth - 100 cm below liquid level
    release area  - 0.01 cm^
    wind speed -  2 m/s
    stability - neutral


    Sample Estimates

    1) Techniques to estimate  emissions are provided in Section 4.18 and
       Flowchart  C-19.   The release rate is calculated by:
                                   6-53

-------
         = 0.8(0.01 cm2) 0.69 g/cm3 •
         /  1960 (100 cm) + 2(2.03 x 106 - 1.01 x 106)  dynes/cm2

        >                          0.69 g/cm3
         = 9.8 g/s

2) Simulations  of  dispersion  follow Flowchart  D-l  for  continuous
   releases.  No information is available on the location of  the  leak
   relative to  the tank  dimensions  or other  structures  so  that  the
   estimate cannot evaluate  the  effects  of downwash.   In the absence
   of  this  information,  the  most  conservative   assumption  is  to
   consider the source a continuous ground level point source.

   A check  of plume  density shows that the release  (31  g/g-mole)  at
   the  boiling  temperature  is slightly  more  dense  than  air  (28.9
   g/g-mole).   Thus,  a  check of the Richardson number ('Section 5.1.2)
   must be made.

   The  volume rate  is  calculated  at ambient  temperature using  the
   molar weight and gas molar volume at ambient temperature:

            9.8 g/s   /         m3  \ /267 K
      V = 	 I 0.0224 	 1 (  	
          31 g/g-mole  \       g-mole / V 273 K

        = 6.9 x 10~3 m3/s


   An equivalent diameter is calculated:

           4V
      d =]
           •iru
           4 (6.9 x lO-3)
        = / 	  =0.094
         \   (3.14) 1 m/s

   Then,

                  (31 g/g-mole)(283 deg. K)    \   0.0069 m3/s
      Ri = 2722 /	 -1] 	 = 27
                  28.9 g/g-mole (267 deg. K)    / (1 m/s)3 0.094 m


   Since Ri < 30, the release is essentially passive.


3) As with Example  6.17,  no downwash,  plume  rise,,  or  buoyancy induced
   dispersion   apply.    Therefore,   once  dispersion   parameters  are
   determined  (Flowchart  D-3,  Section  5.3.1),  receptor  concentration
   can be calculated (Section 5.6.4):

      <7y (100 m) = 4 m

                               6-54

-------
o~ (100 m) = 2.3 m
C =
              9.8  g/s
      3.14 (2.3 m) (4 m) 1 m/s
=  0.339 g/m3
                          6-55

-------
REFERENCES

Beilstein, 1987:  Handbook of Organic Chemistry.   Springer-Verlag, New York.

Beychok, M.,  1979:  Fundamentals of Stack Gas Dispersion, Irvine, CA

Cox,  A.  and  R.  Carpenter,  1980:   Further  Development  of  a  Dense  Vapour
    Dispersion  Model  for Hazard  Analysis.   Heavy Gas and Risk  Assessment,  S.
    Hartwig (ed.) D. Reidel Publishing, Dordrecht, Holland.

Environmental   Protection   Service,   1985:    Introduction   Manual,   Technical
    Information for Problem Spills (TIPS), Technical  Services  Branch.   Ottawa,
    Canada

Fingas, M. , I. Buist, R. Belore,  D.  Mackay,  and P. Kawamura,  1986:   The Input
    of  Spilled  Chemicals  into  the  Environment.   Hazardous  Materials  Spills
    Conference, St. Louis.

Green, D., 1984:  Perry's Chemical Engineers Handbook.  McGraw-Hill, New York.

Havens,  J.  and  T.  Spicer,  1985:   Development  of  an  Atmospheric  Dispersion
    Model  for  Heavier-than-Air  Gas  Mixtures.   U.S.  Dept.  of  Transportation
    CG-D-23-85.

Hoot,  T.,  R.  Meroney,  and  J.  Peterka, 1973:  Wind Tunnel Tests of Negatively
    Buoyant Plumes.  EPA 650/3-74-003.

Hunsaker,  J.   and  B.  Rightmire,  1947:    Engineering  Applications  of  Fluid
    Mechanics.  McGraw-Hill, New York.

Leahey,  D.  and  M.  Davies,  1984:   Observations of  Plume  Rise   from  Sour Gas
    Flares, Atm.  Envir., 18:917-922

List,  R. ,  1968:   Smithsonian Meteorological  Tables.   Smithsonian Institute,
    Washington, D.C.

National  Oceanographic  and  Atmospheric Administration,  1988:   ALOHA  - Areal
    Locations   of  Hazardous   Atmospheres,    Technical   Appendix,   Hazardous
    Materials Response Branch, Seattle, WA.

Pasguill, F.,  1976:   Atmospheric  Diffusion (2nd ed.).   John Wiley  & Sons, New
    York.

Petersen, W. ,  1982:   Estimating  Concentrations Downwind from  an Instantaneous
    Puff Release.  U.S.  EPA, ESRL.

Slade, D., 1968:  Meteorology and Atomic Energy,  U.S.  Atomic Energy Commission
    (T10-24190).

Thibodeaux, L.  and S.  Hwang, 1982:   Landfarming  of Petroleum Wastes  - The
    Modeling Problem, Environmental Progress, 1(46).

Turner, D., 1970:  Workbook of Atmospheric Dispersion Estimates, Office of Air
    Programs  Publication AP-26.  U.S. Environmental Protection Agency.
                                      R-l

-------
U.S.   Environmental  Protection  Agency,  1986:   Compiling Air  Toxic  Emission
    Inventories.   EPA-450/4-86-010.

U.S.   Environmental   Protection  Agency,   1987a:   "Hazardous  Waste  Treatment,
    Storage  and  Disposal  Facilities  (TSDF)   - Air  Emission  Models",  Draft
    Report, U.S.  EPA OAQPS, April,  1987.

U.S.  Environmental  Protection  Agency, 1987b:   Emission Factors for  Equipment
    Leaks of VOC and HAP.  EPA-450/3-86-002.

U.S.  Environmental  Protection Agency,  1987c:   Industrial Source Complex  (ISC)
    Dispersion Model  User's Guide,  Second  Edition (Revised), Volume  1  and 2.
    EPA-450/4-88-002a and 002b.

U.S.   Environmental  Protection Agency,  1987d:  On-Site  Meteorological Program
    Guidance for Regulatory Modeling Applications,  1987.  EPA-450/4-87-013.

U.S.   Environmental  Protection  Agency,  1988a:   Air  Emissions  from  Municipal
    Solid Waste Landfills  - Background Information for  Proposed Standards  and
    Guidelines,  Office  of Air  Quality  Planning  and  Standards  (Preliminary
    Draft).

U.S.  Environmental  Protection Agency,  1988b:   Procedures for Evaluating Impact
    of Stationary Sources, Braft Report, September 1988.
                           f  -   ,  •   . - -• •*•",' *
Van Ulden,  A.  1974:  On Spreading of a Heavy Gas Released Near the Ground,  1st
    International Loss Prevention Symposium,  The Hague/Delft.

Verschueren, K.,  1983:  Handbook  of Environmental Data on  Organic Chemicals.
    Van Mostrand Reinhold Company, New York.

Wallis, G., 1969:   One Dimensional Two-Phase Flow,  McGraw-Hill, New York.

World Bank,  1985:   Manual of Industrial Hazard  Assessment Techniques.  Office
    of Environmental and Scientific Affairs, Washington.
                                      R-2

-------
   APPENDIX A




EMISSION FACTORS

-------
                                   APPENDIX A

                                EMISSION FACTORS


    One alternative  for  estimating air toxic emissions from sources is through

the  use of  emission  factors.   Emission  factors have  been developed  for  a

number  of  processes and  pollutants.   Emission factors provide  an estimate of

emissions as a function of  source  activity such as process  rate  or some other

operating  parameter.  Emission  factors are intended  to be  used  for  making

preliminary estimates of toxic air emissions.  As  sach,  they represent generic

factors whose  applicability to  a  specific  source may  be questionable.   These

factors  will  not   likely  provide  exact  estimates  of  emissions  from  any

particular  source.   The  source  of  an  emission  factor  must  be  carefully

evaluated  to determine  whether it  is  applicable to  a  particular facility.

Emission factors  are available  for both area and point sources.   Some sources

of emission factors applicable to air toxics emissions are presented below.


    1) U.S.  Environmental  Protection  Agency.    Locating and  Estimating  Air
       Emissions from Sources of (Substance).  EPA 450/4-84-007a-q.

    EPA has  underway  a  program to  compile and  publish  emission  factors for
various air  toxics.   To  date,  sixteen  reports have  been published  as part of
this  program.   The substances  covered by  this  series include:   acrylonitrile,
carbon  tetrachloride,  chloroform,  ethylene  dichloride,  formaldehyde,  nickel,
chromium,  manganese,  phosgene,  epichlorohydrin,  vinylidene  chloride,  ethylene
oxide, chlorobenzenes, PCBs, POM, and benzene.

    2) U.S.  Environmental  Protection Agency.   Survey of  (Substance)  Emission
       Sources.

    A second series of  reports on specific air  toxics  has been developed by
EPA as  part  of  the National Emissions  Standards  For Hazardous  Air Pollutants
(NESHAPS)    program.    The   substances  covered   by   this   series  include:
trichloroethylene  (EPA  450/3-85-021),   perchloroethylene (EPA  450/3-85-017),
ethylene  oxide   (EPA  450/3-85-014),  chloroform  (EPA  450/3-85-026),  ethylene
dichloride  (EPA  450/3-84-018),  methylene  chloride  (EPA  450/3-85-015),  and
carbon tetrachloride (EPA 450/3-85-018).
                                      A-l

-------
    3) U.S.   Environmental  Protection Agency.   Preliminary  Compilation of  Air
       Pollutant  Emission   Factors   for    Selected   Air   Toxic   Compounds.
       EPA-450/4-86-010a,  1987.

    This preliminary report presents  emission  factors  of air toxic  pollutants
for  a variety  of sources with  varying activity  levels.   This  listing  gives
little technical  detail concerning  the derivations or applicability of any of
the  factors  therein.   This preliminary  report is currently being updated  and
expanded and a data management  system is  being  developed to  allow for  easy
access  of  the  factors.   The  updated factors  are associated  with  pollutant
names and CAS numbers, process  descriptions and  SIC  codes,  emission  source
descriptions  and  SCC  codes,  notes  on the  derivation of  the  factors and on
control measures  associated  with the  factors,  and references.   The  emission
factors can  be used to obtain guick,  rough estimates of air toxic  emissions.
More detailed data on the emission  sources  can be obtained from  the  Notes  and
References  Sections  listed   in   the  emission  factor  tables.    The  primary
limitation of using  just  the emission factors listed  in this  compilation is
that  their  accuracy  in application  to a  given  source is  not   shown.   More
accurate emissions  estimates may  reguire  evaluation of  the  application  of
available test data to specific  source characteristics.

    4) U.S.   Environmental  Protection  Agency.   Compilation  of  Air  Pollutant
       Emission   Factors,   Fourth  Edition.    AP-42,   September  1985.    Air
       Emissions  Species Manual,  Volume I,  Volatile  Organic Compound  Species
       Profiles.   EPA  450/2-88-003a,  1988.   Air Emissions  Species  Manual,
       Volume  II,  Particulate  Matter  Species  Profiles.   EPA-450/2-88-003b,
       1988.

    Another tool for estimating air toxic emissions involves the  use  of VOC/PM
factors presented in AP-42 and  species profiles presented  in Volumes I and II
of  the  Air  Emissions  Species  Manual.  AP-42  contains  emission factors  for
total VOC  and PM rather  than for  a single chemical  compound.   These factors
can be used with profiles contained in the Species Manual to  estimate releases
of  specific  toxic compounds  based  on the  total  amount of VOC or PM released
from a source.  The Species Manual  shows  the percent by weight and  percent by
volume of specific  chemicals  in  emissions from specific chemicals in emissions
from  specific processes.   The VOC  profiles were  obtained from  the 1980  VOC
Data Manual,  new  VOC  profiles developed from  readily  available  existing data,
and  new VOC profiles  developed  from  original  data  as  part  of  the  VOC
speciation  field  sampling  program.   The  PM profiles  were obtained  from the
1984 Source  Composition Library  and from the  literature.   In addition  to the
VOC  and  PM  profiles,  profile  assignments  linking   the profiles  to  source
categories are presented in the Species Manual.  Species  profiles for VOCs and
PM  are  developed  from generic  sources  and  may not  be  representative of
emissions from an individual facility.

    5) U.S.   Environmental Protection Agency.   Fugitive  Emission  Sources of
       Organic  Compounds  -  Additional  Information  on  Emissions,  Emission
       Reductions, and Costs.   EPA-450/3-82-010, 1982.

    This document contains the  data and methodologies  which  EPA  believes  most
accurately   characterize  average  synthetic  organic  chemical  manufacturing
industry  eguipment  leak  emission  rates  of  VOC,  effectiveness  of  control
techniques,  and control  costs for  selected  equipment used in the processing of
organic chemicals.  The emission factors (on Page  1-4) can  be used to estimate
VOC  emissions from any  industrial  plant which has the  selected  equipment and
handles organic chemicals.
                                      A-2

-------
APPENDIX B




 GLOSSARY

-------
                                  APPENDIX B

                                   GLOSSARY

Accident - In this workbook,  an unplanned release event not  considered in the
design  of  a facility;  a  major or  catastrophic release (e.g.,  release from a
tank rupture disk).

Boiling  -  Vaporization resulting from  heat transfer  to liquids  with boiling
temperatures below ambient or spill  surface temperatures.

Cavity  - An  aerodynamic recirculation zone formed in the lee of an obstruction
(e.g., building) in a wind field.

Continuous  Release   -  A  release for which  the  discharge   rate  is  not  time
varying or instantaneous; a  dispersing  plume  for which  sampling  at  a receptor
observes  little or  no variability  due  to  a  time change  of  emissions  (see
instantaneous).

Dense gas  - Gas  mixtures  that are  denser than the surrounding  air.   A dense
gas may  rise  at first  due  to its  initial  upward momentum  from a  stack,  but
then  sinks due  to   its excess density.   In  the  context  of this workbook,  a
release can  result  in dense gas  effects  if its  release Richardson  number  is
> 30.

Depositing  - Removal  from  a plume  of  gases  or  particles by chemical  or
physical interactions with surfaces  or by precipitation.

Emissions  -  A  release  of  gas,  aerosols  or   particulate matter  to  the
atmosphere;  a  gaseous release  is  an  emission  while a  liquid  release  must
evaporate before being emitted.

Evaporation  -  Vaporization  resulting  from   mass  transfer;  i.e.,  turbulent
interactions between the atmosphere  and the liquid surface,

Fugitive emissions  - Emissions  resulting from a  source for which quantifying
release  parameters,  emission  rates,  and locations is difficult.   Emissions
estimates  may  result from  emission factors or  mass balance  calculations  and
typically represent a number of small disperse emission sources at a facility.

Instantaneous release  - 1)  For a liquid, a release  which  occurs within a very
short time.   An instantaneous  liquid release may result  in formation of  an
instantaneous gaseous  or two-phase  cloud if the liquid is  highly volatile or a
pool from which emissions may  be  continuous  or highly  time  dependent.  2)  For
a gas, a release is  considered instantaneous for modeling if:

    -  the travel time to a  receptor is  long relative to  the  length of  the
       release
    -  the averaging or sampling time is long relative to the release time

An instantaneous model  is  used if the concentration estimate at a receptor is
significantly affected  by the upwind and or downwind  edge of the cloud.   For
this  workbook,  releases  of  less  than five  minutes duration  are arbitrarily
considered instantaneous.
                                      B-l

-------
Neutral -  1)  A stability  class in which  mixing is  controlled by  mechanical
turbulence; 2) A passive or tracer release,  see passive release.

Passive release  - Emissions  to the  atmosphere which disperse as  a  tracer,
i.e., as a massless  point  in space which in  no way influences the environment
in which it disperses;  sometimes  called a  neutral buoyancy release  as  opposed
to releases for which buoyancy is important  in dispersion.

Pool, pooling  - Accumulation of  released  liquid  in  a  puddle  on a  surface;
pools  may  be  unconfined  or  spreading,   confined  by  a  berm  or  dike  or
steady-state for which  evaporation  equals  input to the  pool  and  spreading  is
halted.

Reactive -  Dispersing pollutants which chemically interact  with  surfaces  or
other chemicals  resulting  in a transformation  to another  chemical.   Reactions
are  important  in  estimating pollutant  losses  or  formation  as  well  as  in
estimating energy balances in plumes.

Release - Chemicals or pollutants leaving containment, stacks or vents.

Slumping - Initial  spread  of a dense gas characterized by vertical sinking and
horizontal spreading resulting from (negative)  buoyant or  gravitational forces
acting on the cloud.

Two-phase  release  - Releases  consisting  of  a vapor  and  suspended  liquid
droplets resulting  from the violent flash vaporization  of superheated liquids
as they rapidly depressurize.

Volatile  - A  liquid  subject  to  high vaporization due to  its high  vapor
pressure.   In this workbook, the term high volatility  is  intended  to represent
liquids that  pool  after  release  requiring  an evaporation model  to simulate
emissions.    If  a  doubt  exists,  the  liquid  can   be   simulated  using  an
evaporation model.   An assumption  of  rapid  evaporation of a  high volatility
liquid is a conservative assumption.

Wake  -  The entire zone of wind field disturbance caused  by  an obstruction in
the flow.
                                      B-2

-------
            APPENDIX C




FLOWCHARTS FOR WORKBOOK SCENARIOS

-------
                        FLOWCHRRT C-l

         CONTINUOUS PRRTICULRTE EMISSIONS  FROM  STRCKS
                     NO
YES
      USE
REPRESENTATIVE
EMISSION FRCTOR
 (RPPENDIX fl)
            USE SOURCE
             SPECIFIC
           EMISSION RRTE
                                            c-i

-------
                        FLOXCHRRT C-2

              CONTINUOUS FUGITIVE DUST EMISSIONS
                     NO
      USE
REPRESENTRTIVE
EMISSION FFICTOR
 (flPPENDIX fl)
 USE SOURCE
  SPECIFIC
EMISSION RfiTE
                                            C-2

-------
                     FLOWCHRRT  C-3

INSTRNTRNEOUS PRRTICULRTE EMISSIONS FROM DUCT FRILURES
                     USE SOURCE
                      SPECIFIC
                    EMISSION RRTE
                                     C-3

-------
       FLOWCHflRT C-4

CONTINUOUS FLflRE EMISSIONS
         USE SOURCE
          SPECIFIC
        EMISSION RfiTE
           FLOWCHflRT
              0-7
                          C-4

-------
                        FLOWCHRRT C-5

           CONTINUOUS GRSEOUS EMISSIONS  FkOM STRCKS
                     NO
YES
      USE
REPRESENTflTIVE
EMISSION FRCTOR
 (RPPENDIX R)
            USE SOURCE
             SPECIFIC
           EMISSION RRTE
                                            C-5

-------
                       FLOWCHRRT C-6

         CONTINUOUS GRSEOUS LEflKS FROM TRNKS/PIPES
  MULT I
COMPONENT
 RELERSE
                                         YES
                                                   DETERMINE MERN
                                                     DENSITY RND
                                                    MEflN SPECIFIC
                                                     HERT RRTIO
                                O*-
                    YES
  CRLCULRTE
CRITICRL FLOW
EMISSION RRTE
                          CRLCULRTE
                         SUBCRITICRL
                        FLOW EMISSION
                            RRTE
                          [ FLOWCHRRT )
                                           C-6

-------
                  FL01ICHRRT C-7

INSTRNTRNEOUS GRSEOUS EMISSIONS FROM STflCKS/VENTS
                    USE SOURCE
                     SPECIFIC
                   EMISSION RflTE
                     FLOWCHflRT
                                    C-7

-------
                          FLOWCHRRT C-8

         MULTIPLE FUGITIVE CONTINUOUS GflSEOUS EMISSIONS
      USE
REPRESENTRTIVE
EMISSION FRCTOR
 (flPPENDIX fl)
 USE SOURCE
  SPECIFIC
EMISSION RRTE
                               RRE
                            EMISSIONS
                          FROM FIN RRER
                            OR VOLUME
                                9
                                                C-8

-------
                  FLOWCHRRT C-9

CONTINUOUS 6RSEOUS EMISSIONS FROM LRNO  TRERTMENT
                    DETERMINE
                  EMISSION RRTE
                    FLOWCHRRT
                                   C-9

-------
                    FLOWCHRRT C-10

      CONTINUOUS GRSEOUS EMISSIONS FROM LRNDFILLS
    SITE
  SPECIFIC
  EMISSION
    RflTES
    KNOWN
 USE SOURCE
  SPECIFIC
EMISSION RRTE
LRNDFILL
 IN DRY
 CLIMRTE
                       CflLCULRTE
                     MOIST CLIMRTE
                     EMISSION RRTE
                    CflLCULRTE DRY
                       CLIMflTE
                    EMISSION RRTE

                                              C-10

-------
          FLOWCHRRT C-ll

PESTICIDE/HERBICIDE VOLflTILIZRTIQN
           USE SOURCE
            SPECIFIC
          EMISSION RflTE
            FLOWCHRRT
               0-4
                          C-ll

-------
                     FLOWCHRRT C-12

INSTRNTRNEOUS GflSEOUS EMISSIONS DUE TO EQUIPMENT OPENINGS
      USE
REPRESENTATIVE
EMISSION FflCTOR
 (flPPENDIX fi)
 USE SOURCE
  SPECIFIC
EMISSION RfiTE
                                                C-12

-------
                      FLOWCHRRT C-13

              SURFflCE IMPOUNDMENT EMISSIONS
 CRLCULRTE
EQUILIBRIUM
 CONSTflNT
DETERMINE GRS
 flND LIQUID
MfiSS TRflNSFER
COEFFICIENTS
                           DETERMINE
                         OVERflLL MfiSS
                           TRflNSFER
                          COEFFICIENT
                           CflLCULflTE
                          EQUILIBRIUM
                         CQNCENTRRTION
                              IS
                              THE
                          IMPOUNDMENT
                            RERRTED
                               9
                            CRLCULRTE
                          TURBULENT GRS
                           flND LIQUID
                          MflSS TRRNSFER
                          COEFFICIENTS
                                 NO
                                                     CRLCULRTE
                                                      OVERRLL
                                                     TURBULENT
                                                   MflSS  TRflNSFER
                                                    COEFFICIENT
                           DETERMINE
                         EMISSION RRTE
                            CRLCULRTE
                            COMBINED
                         QUIESCENT/TURB.
                          MRSS TRRNSFER
                           COEFFICIENT
                                          C-13

-------
                         FLOWCHRRT C-14

          CONTINOUS RELIEF VflLVE DISCHRRGE (TWO-PHRSE)
          RELIEF VRLVE
PRESSURIZED TflNK
 USE SOURCE
  SPECIFIC
EMISSION RRTE
           CflLCULRTE
             FLflSH
         FRRCTION flND
           TWO-PHflSE
            OUTFLOW
                            CRLCULRTE
                          MEflN DENSITY
                           PRRRMETERS
                            FLOWCHRRT
                               0-1
                                              C-14

-------
                 FLOWCHRRT  C-15

INSTflNTRNEOUS  RELIEF  VRLVE  DISCHRRGE  (TWO-PHRSE)
                  USE SOURCE
                   SPECIFIC
                   EMISSIONS
                   CflLCULflTE
                 MEfiN DENSITY
                                  C-15

-------
                                 FLOWCHRRT C-16

             EMISSIONS DUE TO L0« VOLflTILITY LIQUID LERKS FROM PIPES
                                              DETERMINE
                                            MflSS TRRNSFER
                                             COEFFICIENT
                                             CONTINUOUS/
                                            INSTRNTRNEOUS
                                              RELERSE
                         LIQUID
                      RELERSE RRTE
                         KNOWN ?
                                                     INSTRNTflNEOUS
                                     YES
   flSSUME
RELERSE RflTE
 EQUflLS PIPE
  FLOW RRTE
 USE SOURCE
  SPECIFIC
RELERSE RflTE
   RSSUME
RELERSE RflTE
 EQUflLS PIPE
SECTOR VOLUME
                     DETERMINE SPILL
                      RREfl BflSED ON
                     LIQUID RELERSE
                          RflTE
USE SOURCE
 SPECIFIC
  RELERSE
  RMOUNT
                     DETERMINE SPILL
                      RRER BRSED ON
                     LIQUID RELERSE
                         RMOUNT
                        DETERMINE
                       EVRPORRTION
                          RflTE
                        FLOWCHRRT
                           D-4
                                               C-16

-------
                       FLOKCHRRT C-17

EMISSIONS DUE TO LEflKS OF LOW VOLflTILITY  LIQUIDS  FROM  TRNKS
                         DETERMINE
                          LIQUID
                       RELERSE RRTE
                         DETERMINE
                        EVflPORflTION
                         RflTE RND
                        SPILL RREfl
                         FLOWCHflRT
                            D-4
                                        C-17

-------
                       FLOWCHRRT  C-18

EMISSIONS DUE TO LEflKS OF HIGH VOLRTILITY  LIQUIDS FROM PIPES
                         fiSSUME
                      EMISSION RRTE
                       EQUflLS PIPE
                        FLOW RflTE
                        FLOWCHRRT
                           0-2
                                       C-18

-------
                       FLOWCHRRT C-19

EMISSIONS DUE TO LEftKS OF HIGH VOLRTILITY LIQUIDS  FROM TRNKS
                         CflLCULRTE
                          LIQUID
                       RELERSE RRTE
                          RSSUME
                       EMISION RRTE
                       EQUflLS LIQUID
                       RELERSE RRTE
                                       C-19

-------
              APPENDIX D




FLOWCHARTS FOR DISPERSION CALCULATIONS

-------
                                FLOHCHRRT D-l

          DETERMINflTION OF DISPERSION CLflSS FOR CONTINUOUS SOURCES


DETERMINE
DENSITY OF
CONTINUOUS
RELERSE
(5.]
.2)






2-STEP
DENSITY/
RICHRRDSON
NUMBER CHECK

   CRLCULflTE
CONCENTRflTIONS
    (5.6.3)
POINT, RRER
 OR VOLUME
 SOURCE ?
                                             D-l

-------
                                      FLOKCHRRT D-2

                          CONTINUOUS POINT SOURCE CRLCULRTIONS
                            FLOWCHRRT
                               D-2
                                                      RECEPTOR
                                                     IN CflVITY ?
                            PLUME IN
                            CflVITY ?
                             (5.6.1)
  DETERMINE
WRKE IMPRCTS
   (5.5.3)
                          ESTIMflTE NOT
                            RVRILRBLE
                                   CflVITY (x < 3ha)
                                         NEflR (3ha < x < IDha)


                                                         K)
                            RECEPTOR
                           LOCflTION
PLUME IN
 *flKE
                                   FflR (x > lOha)
                            DETERMINE
                             VIRTUflL
                            DISTRNCES
                             (5.5.3)
  DETERMINE
BUOYRNT PLUME
RISE (5.2.3)
  FLOWCHRRT
     D-3
                                                                           CRLCULRTE
                                                                            CRVITY
                                                                            IMPRCTS
                                                                            (5.6-1)
                                                                               END
 ESTIMRTE
DISPERSION
PRRRMETERS
  (5.3-3)
                                                                           CRLCULRTE
                                                                        CONCENTRRTIONS
                                                                            (5.6.4)
                                                                                END
                                               D-2

-------
                  FLOWCHRRT D-3

CONTINUOUS POINT SOURCE DISPERSION CRLCULflTIONS
                   ( FLOWCHRRT ]
                   V    D-3    I
                     DETERMINE
                    DISPERSION
                    PRRRMETERS
                      (5.3-1)
                     DETERMINE
                     BUOYRNCY
                  INDUCED INITIRL
                  DILUTION (5.4)
                     CRLCULRTE
                  CONCENTRRTIONS
                      (5.6.4)
                        END
                                    D-3

-------
                 FLOWCHflRT D-4

CONTINUOUS flREfl SOURCE DISPERSION CRLCULRTIONS
                    FLOWCHflRT
                       D-4
                    DETERMINE
                   HORIZONTflL
                     VIRTURL
                    DISTRNCES
                     (5.5.1)
                    DETERMINE
                   DISPERSION
                   PRRRMETERS
                     (5.3.1)
                    CRLCULRTE
                 CONCENTRflTIONS
                     (5.6.4)
                       END
                                   D-4

-------
                  FLOWCHRRT D-5

CONTINUOUS VOLUME SOURCE DISPERSION CRLCULRTIONS
                    DETERMINE
                  VOLUME SOURCE
                     VIRTURL
                    DISTRNCES
                     (5.5.2)
                    DETERMINE
                   DISPERSION
                   PRRRMETERS
                     (5.3.1)
                    CRLCULRTE
                 CONCENTRRTIONS
                     (5.6.4)
                       END
                                  D-5

-------
                         FLOWCHRRT D-6

              INSTRNTRNEOUS DISPERSION CRLCULRTIONS
                              IS \ PRRTICULRTE
                             ELERSE\ MRTTER
                          "GRSEOUS OR
                         PRRTICULflTE
                           DETERMINE
                          DENSITY OF
                         INSTRNTRNEOUS
                            RELERSE
                            (5.1.1)
                 YES
DETERMINE CLOUD
VOLUME, HEIGHT,
SPRERD RRDIUS,
 MRSS (5.6.2)
POINT OR
 VOLUME
 SOURCE
    9
             VOLUME
                           DETERMINE
                          DISPERSION
                          PflRflMETERS
                            (5.3.2)
                          DETERMINE
                           VIRTURL
                          DISTRNCES
                           (5.5.2)
                           CRLCULRTE
                        CONCENTRRTIONS
                            (5-6.5)
                          DETERMINE
                         DISPERSION
                         PRRRMETERS
                           (5.3.2)
                              END
                                                D-6

-------
        FLOWCHRRT D-7

CONTINUOUS FLRRE CRLCULRTIONS
          FLOWCHRRT
             D-7


CRLCULRTE
FLRRE FLfiME
TIP HEIGHT
(5.2.2)


CRLCULRTE
BUOYRNCY
PLUME RISE
(5.2.3)


        ( FLOWCHRRT  )
                          D-7

-------
                 APPENDIX E




AVERAGING PERIOD OF CONCENTRATION ESTIMATES

-------
                                  APPENDIX E
                  AVERAGING PERIOD OF CONCENTRATION ESTIMATES
    The purpose of  this  appendix is to provide  some  simplified techniques for
concentration averaging  from instantaneous and  continuous equations  provided
in  the  workbook.   Methods  presented  are  applicable  to  ground-level  and
elevated emissions of passive gases and particulate matter.
•   Instantaneous Estimates

    Methods  provided  for  instantaneous  concentration  estimates  (i.e.,  puff
releases with a  duration  of 5 minutes or less) in Section 5.6.5 represent peak
concentrations at the centroid of an expanding puff  transported in the  wind.
Petersen  (1982)  provides  equations for estimating peak average concentrations
over time periods of up to one hour.   The method is as follows:

             C(mean) =  C(instantaneous) x F

    where the correction factor (F)  represents the mean height of the  area of
the Gaussian puff which traverses a  receptor  in  the  sampling time (T).  It is
given by:

             F =  (A - 0.5)/(0.3989 N)

    where:

       A = the   area   under  the   Gaussian  distribution  within  N  standard
           deviations, as  found in Figure E-l

       N = the number of standard  deviations from the peak defined as:

           N =  TU
                2ar
           T  - averaging  time in  seconds
           u  - transport  wind speed  (m/s)
           ar - instantaneous  horizontal  plume   dispersion  parameter  at  the
                receptor distance  (m) (from Figure 5-4)


•   Continuous Estimates

    To obtain the estimate of the  maximum concentration for a longer averaging
time,  multiply the 1-hour  maximum  concentration by the given factor:

              Averaging Time                  Multiplying  Factor

                  3  hours                          0.9 (±0.1)
                  8  hours                          0.7 (±0.2)
                 24  hours                          0.4 (±0.2)
                                      E-l

-------
                            FIGURE E-l


                    AREA UNDER NORMAL CURVE
   4.0
C/5

O
UJ
C
O
CC
<
O
z
<

C/J
tx
O
cr
UJ
CD
2

Z
3.0
2.0
    1.0
      50  60   70   80
                        90    95
98   99
99.8 99.9
99.99
                                E-2

-------
The  numbers  in  parentheses  are recommended  limits to  which one may  diverge
from the multiplying factors  representing  the general  case.   For example,  if
aerodynamic  downwash or  terrain  is  a  problem  at the  facility,  or  if  the
release height is very  low,  it may be appropriate  to  increase the  factors  up
to  the  limits  specified   in  parentheses.    Conversely,   if  the   stack  is
relatively tall  and there  are  no  terrain  or downwash  problems,  it  may  be
appropriate  to  decrease  the  factors.   For  averaging  times  in between  the
values  listed  above, use  the  multiplying factor  for  the  shorter  averaging
time.   For  example,  if a  4-hour  average concentration  is  needed,  use  the
multiplying factor  for  the  3-hour averaging time (0.9).

    T6~db±ain  the estimated  maximum concentration for  a shorten averaging time
than 1-hour,  use thex'l-hour\ maximum concentration  for any  desired averaging^
time between  30  and7 60 minutes.  For/ averaging \times  less than\30  minutes,  a
specific procedure/can  not  be recommended.         •— "


                               "~
                                              t
                                                                   C
                                                       /  }°'2
                                                   *-> i ^//%/.  i
                                  <<3,'^^  JZJ^vT/
                                      E-3

-------
         APPENDIX  F




SELECTED CONVERSION FACTORS

-------
                                  APPENDIX F

                          SELECTED CONVERSION FACTORS
Pressure

1 ATM = 1.013 x 106 dynes/cm2
1 millibar = 1000 dynes/cm2
1 mm Hg = 1333.224 dynes/cm2
1 lb/in2 = 68,947.6 dynes/cm2
1 in. Hg = 33,863.9 dynes/cm2
1 Pascal = 10 dynes/cm2

Volume
   ,3 = 106 cm3
   i3 = 103 liters
1
1
1 cu ft = 28.317 liters
1 liter = 103 cm3
1 m3 = 35.315 cu feet
1 gal = 3,785 cm3

Mass Release Rate

1 g/s = 7.9367 Ib/hr
1 t/yr = 2.8766 x 10~2 g/s
It/day = 10.500 g/s

Concentration
                                       1 cal (g) = 3.9685 x 10~3 BTU
                                       1 BTU = 251.634 cal
                                       1 BTU = 1.0543 x 1010 ergs
                                       1 BTU = 1054 Joules - (N - m)
                                       Heat Rate
                                                           -6
                                       1 cal/s = 1.102 x 10~D BTU/h
Flow
                                       1 m3/h = 3600 m3/s
                                       Area
                                       1 m2 = 104 cm2
                                       1 ft2 = 0.0929 m2
                                       1 hectare = 104 m2
                                       1 acre = 4046.86 m2
Conversions with parts  per million by
volume
ug/m3 = (ppm) 40.87 MW
ppm = (ug/m3)
              0.02447
                MW
                         P  /To
                         Po \ T
                       Po / T

                       P  \ To
                                      F-l

-------

-------
                   APPENDIX  G




CALCULATIONS METHODS FOR DISPERSION PARAMETERS

-------
                                  APPENDIX G

                CALCULATIONAL METHODS FOR DISPERSION PARAMETERS

    Dispersion parameters presented graphically in Section 5.3.1 and  5.3.2  are
statistical fits  to  observed experimental data.  Dispersion  parameters  may be
derived from these figures or from equations presented in this appendix.

    Instantaneous  dispersion parameters  are  derived  from  quasi-instantaneous
releases (Slade, 1968).   The parameters are of the form:

       a =  a  xb

where x is the downwind distance in meters and the coefficients are given by:

           STABILITY               HORIZONTAL             VERTICAL
                                   a        b            a        b

           Unstable (A-C)         0.14     0.92         0.53     0.73
           Neutral (D)            0.06     0.92         0.15     0.70
           Stable (E-F)            0.02     0.89         0.05     0.61

    Pasquill-Gifford  dispersion  parameters  for  continuous  sources  can  be
calculated using downwind distance  in kilometers using Figures 5-1 and  5-2 or
techniques from ISC (EPA, 1987c) given in Tables G-l and G-2:
                                      G-l

-------
                                  TABLE G-l




               PARAMETERS USED TO CALCULATE PASQUILL-GIFFORD oy
oy (meters) = 465.12 (x) tan (TH)
Pasguill
Stability
Category
A
B
C
D
E
P
TH = 0.01745
c
24.17
18.33
12.50
8.33
6.25
4.17

-------
                                     TABLE G-2




                PARAMETERS  USED TO CALCULATE PASQUILL-GIFFORD oz
Pasguill
Stability
Category x (km)
A* <.10
0.10 - 0.15
0.16 - 0.20
0.21 - 0.25
0.26 - 0.30
0.31 - 0.40
0.41 - 0.50
0.51 - 3.11
>3.11
B* <.20
0.21 - 0.40
>0.40
C* All
D* <.30
0.31 - 1.00
1.01 - 3.00
3.01 - 10.00
10.01 - 30.00
>30.00
CTz
a
122.800
158.080
170.220
179.520
217.410
258.890
346.750
453.850
**
90.673
98.483
109.300
61.141
34.459
32.093
32.093
33.504
36.650
44.053
(meters) = a x"
b
0.94470
1.05420
1.09320
1.12620
1.26440
1.40940
1.72830
2.11660
**
0.93198
0.98332
1.09710
0.91465
0.86974
0.81066
0.64403
0.60486
0.56589
0.51179
 * If the calculated value of a_  exceeds 5000 m, a-  is set to 5000 m.
** i
   is equal  to  5000 m.
                                        G-3
                                                        U S. GOVERNMENT PRINTING OFT ICE I')-

-------
                   TABLE G-2




                  (Continued)




PARAMETERS USED TO CALCULATE PASQUILL-GIFFORD az
Pasguill
Stability
Category x (km)
E <.10
0.10 - 0.30
0.31 - 1.00
1.01 - 2.00
2.01 - 4.00
4.01 - 10.00
10.01 - 20.00
20.01 - 40.00
>40.00
F <.20
0.21 - 0.70
0.71 - 1.00
1.01 - 2.00
2.01 - 3.00
3.01 - 7.00
7.01 - 15.00
15.01 - 30.00
30.01 - 60.00
>60.00
°z
a
24.260
23.331
21.628
21.628
22.534
24.703
26.970
35.420
47.618
15.209
14.457
13.953
13.953
14.823
16.187
17.836
22.651
27.074
34.219
(meters) = a x13
b
0.83660
0.81956
0.75660
0.63077
0.57154
0.50527
0.46713
0.37615
0.29592
0.81558
0.78407
0.68465
0.63227
0.54503
0.46490
0.41507
0.32681
0.27436
0.21716
                      G-4

-------
                            ADDENDUM
"A Workbook of  Screening Techniques for Assessing Impacts of Toxic
                Air Pollutants" — September 1988


     Page  l-l:  The  title  and  reference  of  the  publication
     "Procedures for  Evaluating  Impact of  Stationary  Sources"
     (1988b) is now "Screening Procedures  for  Estimating the Air
     Quality Impact of Stationary Sources,"  Erode, 1988. EPA-450/4-
     88-010.

     Page 2-9:  Table 2-2  is taken from Appendix A of Erode (1988)
     referenced above.

     Page  6-17:  Model output  shown  is  based  on a  preliminary
     version of RVD run on  8/16/88.  Results may differ from those
     obtained when  using  the final version  of RVD.

     Page 6-41:  Same comment about RVD model output as above.

     Page  R-2:   The reference   "U.S.  Environmental  Protection
     Agency, 1988b: Procedures of Evaluating Impact of Stationary
     Sources,"  Draft Report, September 1988 is  now "Erode, R. W.,
     1988:  Screening  Procedures  for  Estimating the Air  Quality
     Impact of  Stationary  Sources,"  Draft  for public  comment,
     August 1988. EFA-450/4-86-010.

     Page  E-3:  Delete the  last  paragraph.    Replace  with  the
     following  paragraph:

          To obtain the  estimated maximum concentration,  Ct,
          for a shorter averaging time than 1 hour adjust the
          1-hour maximum concentration by the following ratio:

            Ct = €„ (60 min/t)°-2

            where:

               t -  averaging period of interest (less
                   than 60  minutes)
                        December 8,  1988

-------