EPA-600/2-76-089a
April 1976
Environmental Protection Technology  Series

-------
               RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped  into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
     1.   Environmental Health Effects Research
     2.   Environmental Protection Technology
     3.   Ecological Research
     4.   Environmental Monitoring
     5.   Socioeconomic Environmental Studies

This report has been  assigned  to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new  or improved technology required for the control  and
treatment of pollution sources to meet environmental quality standards.
                    EPA REVIEW NOTICE

This report has been reviewed by  the U.S. Environmental
Protection Agency, and  approved for publication.  Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.

-------
                                       EPA-600/2-76-089a
                                       April 1976
          TECHNICAL  MANUAL

         FOR  MEASUREMENT OF

          FUGITIVE  EMISSIONS:

    UPWIND/DOWNWIND SAMPLING METHOD

          FOR INDUSTRIAL EMISSIONS
                      by

              Henry J. Kolnsberg

TRC--The Research Corporation of New England
            125 Silas Deane Highway
        Wethersfield,  Connecticut 06109
            Contract No. 68-02-2110
             ROAP No.  21AUY-095
         Program Element No.  1AB015
      EPA Project Officer: R.M. Statnick

  Industrial Environmental Research Laboratory
   Office of Energy, Minerals, and Industry
      Research Triangle  Park, NC 27711
                Prepared for

U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Research and Development
            Washington, DC 20460
                      •»-<>*•
          230 SOUCA Ft- 1

-------
                            TABLE OF CONTENTS
 SECTION
                                                                   PAGE
 1.0            OBJECTIVE	     i

 2.0            INTRODUCTION  	     2
    2.1            Categories of  Fugitive  Emissions   	     2
      2.1.1         Quasi-Stack  Sampling  Method  	     2
      2.1.2         Roof  Monitor Sampling Method  	     3
      2.1.3         Upwind-Downwind  Sampling  Method  	     3
    2.2            Sampling Method Selection  	     4
      2.2.1         Selection Criteria  	  	     4
      2.2.2         Application  of Criteria  	     6
    2.3            Sampling Strategies 	     9
      2.3.1         Survey Measurement  Systems   .  	   10
      2.3.2         Detailed Measurement  Systems   	   IQ

 3.0             TEST PROGRAM PROCEDURES	   12
    3.1            Pretest Survey	   12
    3.2            Test Plan    	   13
    3.3            Upwind-Downwind  Sampling Strategies  	   17
    3.4            Survey Upwind-Downwind  Measurement System  ...   17
      3.4.1         Sampling Equipment	   18
      3.4.2         Sampling System Design  	   19
      3.4.3         Sampling Techniques 	   22
      3.4.4         Data Reduction	   31
    3.5            Detailed Upwind-Downwind Measurement System .  .   31
      3.5.1         Sampling Equipment  	   32
      3.5.2         Sampling System Design  	   33
      3.5.3         Sampling  Techniques 	   34
      3.5.4         Data Reduction	   39
   3.6           Atmospheric Tracers 	   39
   3.7           Quality Assurance 	   43
4.0

APPENDIX

   A
ESTIMATED COSTS AND TIME REQUIREMENTS 	   46
TEST PROCEDURES APPLICATION
                                 111

-------
                             LIST OF TABLES
TABLE

 2-1


 3-1


 3-2

 3-3

 4-1

 4-2

 4-3
 FIGURE

  3-1

  3-2

  3-3

  3-4

  3-5

  3-6

  4-1
                                                   PAGE

Upwind-Downwind Sampling Method Application
to Typical Industrial Fugitive Emissions
Sources 	
Pre-test Survey Information to be Obtained
for Application of Fugitive Emissions Sampling
Methods	   14

Matrix of Sampling System Design Parameters  ...   21

Atmospheric  Stability Categories   	   28

Conditions Assumed for  Cost Estimation  of
Upwind-Downwind Sampling Programs  	   47

Estimated Manpower Requirements  for  Upwind-
Downwind  Sampling Programs   	  	   A8

Estimated  Costs for  Upwind-Downwind  Sampling
Programs   	


              LIST OF FIGURES


                                                    PAGE

 Typical Sampler Locations for Selected
 Source Site Configurations  	   23

 Maximum Downwind Sampler Distances  	    27
                                                     9 Q
 Maximum Crosswind Sampler Distances  	   ^

 Pollutant Concentration Ratios  for  Crosswind
 Locations    	
 Pollutant Concentration Ratios  for  Vertical
 Locations    	
 Vertical  Concentration Distribution Factors ...    38

 Elapsed  Time Estimates for Upwind-Downwind
  Sampling Programs  	
                                   IV

-------
                            LIST OF FIGURES
                               (continued)



FIGURE                                                          PAGE

^~2            Cost-effectiveness of Upwind-Downwind Sampling
               Programs	      52


A-1            Portland Cement Plant Site Layout  	  A-3

A~2            Portland Cement Plant Emission Clouds  	  A-9

A~3            Portland Cement Plant Separate Source Clouds .  .  A-13

-------
 1.0  OBJECTIVE




     The objective of this procedures document is to present a guide




 for the utilization of the Upwind-Downwind Sampling Strategy in the




measurement of fugitive emissions.  Criteria for the selection of the




most applicable measurement method and discussions of general informa-




 tion gathering and planning activities are presented.  Upwind-downwind




sampling strategies and equipment are described and sampling system




design, sampling techniques, and data reduction are discussed.




     Manpower requirements and time estimates for typical applications




of the method are presented for programs designed for overall and speci-




fic emissions measurements.




     The application of the outlined procedures to the measurement of




fugitive emissions from a Portland cement manufacturing plant is pre-




sented as an appendix.
                                  -1-

-------
2.0  INTRODUCTION



     Pollutants emitted into the ambient air from an industrial plant




or other site generally fall into one of two types.  The first type is




released into the air through stacks or similar devices designed to




direct and control the flow of the emissions.  These emissions may be




readily measured by universally-recognized standard stack sampling tech-




niques.  The second type is released into the air without control of




flow or direction.  These fugitive emissions usually cannot be measured




using existing standard techniques.



     The development of reliable, generally  applicable measurement pro-




cedures is a necessary prerequisite  to  the development of strategies




for  the control  of fugitive emissions.   This document describes  some




procedures for the measurement  of  fugitive  emissions using  the upwind-




downwind measurement method described  in Section 2.1.3  below.









 2.1   Categories  of  Fugitive Emissions



      Fugitive emissions  emanate from such a wide variety of circumstances




 that it is not particularly meaningful to attempt to categorize them




 either in terms  of  the processes or mechanisms that generate them, or




 the geometry of  the emission points.  A more useful approach is to cate-




 gorize fugitive emissions  in terms of the methods for their measurement.




 Three basic methods exist—quasi-stack  sampling, roof monitor sampling,




 and upwind-downwind sampling.  Each is  described in general terms below.









      2.1.1  Quasi-Stack Sampling Method



      In this method,  the fugitive emissions are captured in a temporarily




  installed hood  or enclosure  and vented to  an  exhaust duct  or  stack  of
                                     -2-

-------
 regular cross-sectional area.   Emissions are then measured In the ex-




 haust duct using standard stack sampling or  similar well recognized




 methods.   This  approach Is necessarily  restricted to those sources of




 emissions  that  are  isolable and physically arranged so  as  to  permit the




 Installation of a temporary hood or enclosure  that will not interfere




 with  plant operations or alter  the character of  the process or  the emis-




 sions .









      2.1.2 Roof Monitor Sampling Method




      This  method is used to measure the fugitive  emissions entering the




 ambient air from building or other enclosure openings such as roof moni-




 tors, doors and windows.  The method is especially applicable to  situa-




 tions in which  enclosed  sources  are too numerous  or physically  configured




 to preclude the application of  the quasi-stack method to each source.




 Sampling is, in general,  limited to a mixture of  all uncontrolled emis-




 sion  sources within the  enclosure and requires the  ability to make  low




 velocity exhaust air measurements and mass balances  of  small  quantities




 of materials entering and leaving the enclosure through  the openings.








     2.1.3  Upwind-Downwind Sampling Method




     This method is utilized to measure the fugitive emissions  from




 sources typically covering large areas that cannot be temporarily hooded




and are not enclosed in a structure allowing the use of the roof moni-




 tor method.  Such sources include material handling and storage opera-




 tions, waste dumps,  and industrial processes  in which the emissions are




spread over large areas.  These features are  embodied in the  typical
                                   -3-

-------
industrial sources and their emitted pollutants listed in Table 2-1.

     The upwind-downwind method quantifies the emissions from such sources

as the difference between the pollutant concentrations measured in the

ambient air approaching (upwind) and leaving (downwind) the source site.

It may also be utilized in combination with mathematical models and

tracer tests to define the contributions to total measured emissions of

specific sources among a group of sources.



2.2  Sampling Method Selection

     The initial step in the measurement of fugitive emissions at an

industrial site is the selection of the most appropriate sampling method

to be employed.  Although it is impossible to enumerate all the combina-

tions of influencing factors that might be encountered in a specific

situation, careful consideration of the following general criteria should

result in the selection of the most effective of the three sampling

methods described above.



     2.2.1  Selection Criteria

     The selection criteria listed below are grouped into three general

classifications common  to all fugitive  emissions measurement methods.

The  criteria are  intended to provide only representative examples and

should not be  considered a complete listing of  influencing factors.



     2.2.1.1   Site Criteria

     Source Isolability.  Can  the emissions be  measured  separately  from
     emissions  from  other sources?  Can the source be  enclosed?

     Source Location.   Is the  source indoors or out?   Does location
                                    -4-

-------
                                                                                                   TABLE 2-1


                                                                                        UPWIND-DOWNWIND SAMPLING METHOD
                                                                          APPLICATION TO TYPICAL INDUSTRIAL FUGITIVE EMISSION SOURCES
 I
Ol
 I
Industry
Coke
Making







Primary
Aluminum


Primary
Copper
Sand &
Gravel







Source
Coal Handlin
Coal Storage
Charging Ov-
ens
Coking, Door
& Oven Leak
Coke Pushing
Quenching
(Controlled
but some
leaks)
Coke Handlin
Coke Storage
Bauxite Han-
dling &
Storage
Alumina Cal-
cining &
Preparation
Alumina Stor-
age
Mining
Hauling
Tailings Pond
Slag Dump
Quarrying
Truck Hauling
Delivery &
Storage
Rock Transfer
Crushing &
Screening
Drying (Leak-
age)
Product Stor-
age
Product Load-
ine
Ai*s
'roduct
Delivery



Par-
ticulate
Emission
Coal Dust
Coal Dust
Coal Dust
Tars
Coke Dust
Tars
Coke Dust
Tars
Coke Dust
Tars

Coke Dust
Ore Dust
Alumina
Dust

Alumina
Dust
Dust
Dust
Dust
Dust
ust
ust
ust
ust
ust
ust
ust
ust



Gas & Vapor
Emissions
HC.CO.NO
Pyridine
HC,CO,NOX
H2S,NH3,CS2
Phenol
HC,CO,H2S,NH
HCN, Phenol
HC,CO,H2S,NH
HCN,S02,
Phenol
-
-
-

S02








Industry
Electri
Furnace
Steel






Iron &
Steel
Found-
ries


Coal

Asphalt






Source
Scrap & Sinte
Delivery
Lime & Silica
Delivery
Pollution Con
trol Equip-
ment -
Dust Transfer
& Storage
Storage Ponds
for Slagging
Waste Water
Coke, Silica
Sinter Deliv
ery & Stor-
age
Pollution Con
trol Equip-
ment Transfe
& Storage

Mining
Hauling
Storage &
Transfer
Screening &
Crushing
Transfer
Drying
Storage Piles
Waste Transfer
Gravel Deliv-
ery
Asphalt Stor-
age
torage Piles
sphalt Batch-
ing
rier & Blower
eactor Dis-
charge
eactor Charge

roduct Trans-
fer
Par-
ticulate
Emission
Iron Dus
Steel
Dust
Dust


Fume Dus
-

Dust
Dust


Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Dust
ust
ars
ust
ust,
Tars
ust,
Tars
ust,
Tars
ust,
Tars
ust,
Tars
Gas & Vapor
Emissions
-
-



H2S,S02

-
-


-
Hydrocarbons ,
CO

ydrocarbons ,
Odors
ydrocarbons ,
Odor
ydrocarbons ,
Odor
rdrocarbons ,
Odor
'drocarbons ,
Odor
'drocarbons ,
Odor
Industry
Coal
Gasifica-
tion




Petroleum
Refining
Source
Coal Delivery
& Storage
Coal Transfer
Waste .Trans-
fer
Scrubber Sol-
ids
Settling Pond

Crude Transfe
Crude Storage
Distilate
Storage
: Distilate
Transfer
Gasoline Stor-
age & Transf ei


hosphate
ertili-
er





Leakage in
Drains &
Sewers
Waste Water
Storage &
Transfer
Process Leaks
Gasoline Ter-
minal Loading
Mining
Storage Piles
Rock Trans-
fer (Truck,
Conveyer)
Settling Pond
jypsum Pile
'roduct Stor-
'roduct Deliv-
ery



Par-
tlculate
Emission
Coal Dus
Coal Dus
Dust

Dust


_




Dust
Tars

ust
Dust
ust
'luorides
Dust
Dust
Dust



Gas & Vapor
Emissions

Hydrocarbons ,
H2S,NH3,S02
Phenol
Hydrocarbons
H2S,S02,NH3
Phenol ,
Hydrocarbons
Hydrocarbons ,
RSH,H2S
Hydrocarbons ,
RSH,H2S
tydrocarbons
Hydrocarbons
Hydrocarbons

Hydrocarbons ,
H2S,HF,
Phenols
H2S,S02,NH3,
Phenols, HF,
Hydrocarbons
H2S,SO2,NH3,
Phenols, HF,
Hydrobarbons
Hydrocarbons


02 f Fluorides
02
Luorides
Luorides




-------
    permit access of measuring equipment?

    Meteorological Conditions.  What are the conditions representative
    of typical and critical situations?  Will precipitation interfere
    with measurements?  Will rain or snow on ground effect dust levels?
    2.2.1.2  Process Criteria

    Number and Size of Sources.   Are emissions from a single, well
    defined location or many scattered locations?  Is source small
    enough to hood?

    Homogeneity of Emissions.  Are emissions the same type everywhere
    at the site?  Are reactive effects between different emissions
    involved?

    Continuity of Process.  Will emissions be produced long enough to
    obtain meaningful samples?

    Effects of Measurements.  Are special procedures required to pre-
    vent the making of measurements from altering the process or emis-
    sions or interfering with production?  Are such procedures feasible?
    2.2.1.3  Pollutant Criteria

    Nature of Emissions.  Are measurements of particles, gases, liquids
    required?  Are emissions hazardous?

    Emission Generation Rate.  Are enough emissions produced to provide
    measurable samples in reasonable  sampling time?

    Emission Dilution.  Will transport air reduce  emission concentra-
    tion below measurable levels?
     2.2.2  Application of Criteria

     The application of the selection criteria listed in Section 2.2.1

to each of the fugitive emissions measurement methods defined in Section

2.1 is described in general terms in this section.
                                   -6-

-------
     2.2.2.1  Quasi-Stack Method




     Effective use of the quasi-stack method requires  that  the  source




of emissions be  isolable and that an enclosure can be  installed capable




of capturing emissions without interference with plant operations.  The




location of the  source alone is not normally a factor.  Meteorological




conditions usually need be considered only if they directly affect  the




sampling.




     The quasi-stack method is usually restricted to a single source




and must be limited to two or three small sources that can be effectively




enclosed to duct their total emissions to a single sampling point.




Cyclic processes should provide measurable pollutant quantities during




a single cycle to avoid sample dilution.  The possible effects  of the




measurement on the process or emissions is of special  significance  in




this method.  In many cases, enclosing a portion of a process in order




to capture its emissions can alter that portion of the process  by chang-




ing its temperature profile or affecting flow rates.  Emissions may be




similarly altered by reaction with components of the ambient air drawn




into the sampling ducts.  While these effects are not necessarily limit-




ing in the selection of the method, they must be considered in  designing




the test program and could influence the method selection by increasing




complexity and costs.




     The quasi-stack method is useful for virtually all types of emis-




sions.   It will provide measurable samples in generally short sampling




times since it captures essentially all of the emissions.   Dilution of




the pollutants of concern is of little consequence since it can usually




be controlled in the design of  the sampling system.
                                    -7-

-------
     2.2.2.2  Roof Monitor Method




     Practical utilization of the roof monitor method demands that the




source of emissions be enclosed in a structure with a limited number of




openings to the atmosphere.  Measurements may usually be made only of




the total of all emissions sources within the structure.  Meteorological




conditions normally need not be considered in selecting this method




unless they have a direct effect on the flow of emissions through the




enclosure opening.




     The number of sources and the mixture of emissions is relatively




unimportant since the measurements usually include only the total emis-




sions.  The processes involved may be discontinuous as long as a repre-




sentative combination of the typical or critical groupings may be in-




cluded in a sampling.  Measurements will normally have no effect on the




processes or emissions.




     The roof monitor method, usually dependent on or at least influ-




enced by gravity in the transmission of emissions, may not be useful




for the measurement of larger particulates which may settle within the




enclosure being sampled.  Emission generation rates must be high enough




to provide pollutant concentrations of measurable magnitude after dilu-




tion in the enclosed volume of the structure.









     2.2.2.3  Upwind-Downwind Method




     The upwind-downwind method, generally utilized where neither of




the other methods may be successfully employed, is not influenced by




the number or location of the emission sources except as they influence

-------
 the locating of sampling devices.   In most cases, only the total con-




 tribution to the ambient atmosphere of all sources within a sampling




 area may be measured.   The method  is strongly influenced by meteorolog-




 ical conditions, requiring a wind  consistent in direction and velocity




 throughout the sampling period as  well as conditions of temperature,




 humidity and ground moisture representative of normal ambient condi-



 tions .




      The emissions  measured by the upwind-downwind method may be the




 total contribution  from a single source or from a mixture of many sources




 in  a large area.  Continuity of the emissions is generally of secondary




 importance since the magnitude of  the ambient air volume into which  the




 emissions  are dispersed is large enough to provide a degree of smooth-




 Ing to  cyclic emissions.   The measurements have no effect on the emis-



 sions or processes  involved.




      Most  airborne  pollutants can  be  measured by the upwind-downwind




method.  Generation rates  must be  high enough to  provide measurable




 concentrations  at the sampling locations  after  dilution  with the ambient




air.  Settling  rates of the  larger  particulates  require  that the sampling




 system be  carefully designed  to ensure that representative  particulate



 samples  are  collected.








2.3   Sampling Strategies




     Fugitive emissions measurements may,  in  general, be  separated into




two classes or levels depending upon  the degree of accuracy  desired.




Survey measurement systems are designed to screen emissions  and provide




gross measurements of a number of process influents and effluents at a
                                   -9-

-------
relatively low level of effort in time and cost.  Detailed systems are




designed to isolate, identify, and quantify individual contaminant con-




stituents with increased accuracy and higher investments in time and




cost.









     2.3.1  Survey Measurement Systems




     Survey measurement systems employ recognized standard or state-




of-the-art measurement techniques to screen the total emissions from a




site or source and determine whether any of the emission constituents




should be considered for more detailed investigation.  They generally




utilize the simplest available arrangement of instrumentation and pro-




cedures in a relatively brief sampling program, usually without pro-




visions for sample replication, to provide order-of-magnitude type data,




embodying a factor of two to five in accuracy range with respect to




actual emissions.









     2.3.2  Detailed Measurement Systems




     Detailed measurement systems are used in instances where survey




measurements or equivalent data indicate that a specific emission con-




stituent may be present in a concentration worthy of concern.  Detailed




systems provides more precise identification and quantification of spe-




cific constituents by utilizing the latest state-of-the-art measurement




instrumentation and procedures in carefully designed sampling programs.




These systems are also utilized to provide emission data over a range




of process operating conditions or ambient meteorological influences.




Basic accuracy of detailed measurements is in the order of +10 to + 50
                                   -10-

-------
percent of actual emissions.  Detailed measurement system costs are




generally in the order of three to five times the cost of a survey sys-



tem at a given site.
                                  -11-

-------
3.0  TEST PROGRAM PROCEDURES




     This section describes the procedures required to successfully




complete a testing program utilizing the upwind-downwind sampling method




described in Section 2.1.  It details the information required to plan




the program, describes the organization of the test plan, specifies the




types of sampling equipment to be used, establishes criteria for the




sampling system design, and outlines basic data reduction methods.









3.1  Pretest Survey




     After the measurement method to be utilized in documenting the




fugitive emissions at a particular site has been established using the




criteria of Section 2.2, a pretest survey of the site should be con-




ducted by the program planners.  The pretest survey should result in an




informal, internal report containing all the information necessary for




the preparation of a test plan and the design of the sampling system by




the testing organization.




     This section provides guidelines for conducting a pretest survey




and preparing a pretest survey report.








     3.1.1  Information to be Obtained




     In order to design a system effectively and plan for the on-site




sampling of fugitive emissions, a good general knowledge is required of




the plant layout, process chemistry and flow, surrounding environment,




and prevailing meteorological conditions.  Particular characteristics




of the site relative to  the needs of  the owner, the products involved,




the space and manpower skills available, emission  control equipment
                                  -12-

-------
installed, and the safety and health procedures observed, will also




influence the sampling system design and plan.  Work flow patterns and




schedules that may result in periodic changes in the nature or quantity




of emissions or that indicate periods for the most effective and least




disruptive sampling must also be considered.  Most of this information




can only be obtained by a survey at the site.  Table 3-1 outlines some




of the specific information to be obtained.  Additional information will




be suggested by considerations of the particular on-site situation.









     3.1.2  Report Organization




     The informal, internal pretest survey report must contain all the




pertinent information gathered during and prior to the site study.  A




summary of all communications relative to the test program should be




included in the report along with detailed descriptions of the plant




layout, process, and operations as outlined in Table 3-1.  The report




should also incorporate drawings, diagrams, maps, photographs, meteoro-




logical records, and literature references that will be helpful in plan-




ning the test program.









3.2  Test Plan




     3.2.1  Purpose of a Test Plan




     Measurement programs are very demanding in terms of the scheduling




and completion of many preparatory tasks, observations at sometimes




widely separated locations, instrument checks to verify measurement




validity, etc.  It is therefore essential that all of the experiment




design and planning be done prior to the start of the measurement pro-
                                  -13-

-------
                                TABLE 3-1


               PRE-TEST SURVEY INFORMATION TO BE OBTAINED
          FOR APPLICATION OF FUGITIVE EMISSION SAMPLING METHODS
Plant
Layout
Drawings:
  Building Layout and Plan View of Potential Study Areas
  Building Side Elevations to Identify Obstructions and
     Structure Available to Support Test Setup     ;
  Work Flow Diagrams
  Locations of Suitable Sampling Sites
  Physical Layout Measurements to Supplement Drawings
  Work Space Required at Potential Sampling Sites
Process
  Process Flow Diagram with Fugitive Emission Points
     Identified
  General Description of Process Chemistry
  General Description of Process Operations Including
     Initial Estimate of Fugitive Emissions
  Drawings of Equipment or Segments of Processes Where
     Fugitive Emissions are to be Measured
  Photographs (if permitted) of Process Area Where
     Fugitive Emissions are to be Measured
  Names, Extensions, Locations of Process Foremen and
     Supervisors Where Tests are to be Conducted
Operations
  Location of Available Services (Power Outlets, Main-
     tenance and Plant Engineering Personnel, Labora-
     tories, etc.)
  Local Vendors Who Can Fabricate and Supply Test System
     Components
  Shift Schedules
  Location of Operations Records (combine with process
     operation information)
  Health and Safety Considerations
Other
  Access routes to the areas Where Test Equipment/Instru-
     mentation Will Be Located
  Names, Extensions, Locations of Plant Security and
     Safety Supervisors
                                    -14-

-------
gram in the form of a detailed test plan.  The preparation of such a

plan enables the investigator to "pre-think" effectively and cross-check

all of the details of the design and operation of a measurement program

prior to the commitment of manpower and resources.  The plan then also

serves as the guide for the actual performance of the work.  The test

plan provides a formal specification of the equipment and procedures

required to satisfy the objectives of the measurement program.  It is

based on the information collected in the informal pretest survey re-

port and describes the most effective sampling equipment, procedures,

and timetables consistent with the program objectives and site charac-

teristics .



     3.2.2  Test Plan Organization

     The test plan should contain specific information in each of the

topical areas indicated below:


     Background

          The introductory paragraph containing  the pertinent infor-
     mation leading to the need to conduct the measurement program and
     a short description of the information required  to answer that
     need.

     Ob j ective

          A concise statement of  the problem  addressed by  the test
     program and a brief description of  the program's planned method
     for its solution.

     Approach

          A description of the measurement scheme and data reduction
     methodology employed in  the  program with a  discussion of how each
     will answer the needs identified in the  background  statement.
                                   -15-

-------
 Instrumentation/Equipment/Facilities

     A description of  the instrumentation arrays  to be used  to
 collect  the  samples and meteorological data identified in  the
 approach description.  The number and frequency of samples to be
 taken and  the sampling array resolution should be described.

     A detailed description of the equipment  to be employed  and
 its purpose.

     A description of  the facilities required to  operate the
 measurement  program, including work space, electrical power,
 support  from plant personnel, special construction, etc.

 Schedule

     A detailed chronology of a typical set of measurements  or a
 test, and  the overall  schedule of events from the planning stage
 through  the  completion of the test program report.

 Limitations

     A definition of the conditions under which the measurement
 project  is to be conducted.  If, for example, successful tests can
 be conducted only during occurrences of certain wind directions,
 those favorable limits should be stated.

 Analysis Method

     A description of  the methods which will  be used to analyze
 the samples collected  and the resultant data, e.g., statistical or
 case analysis, and critical aspects of that method.

 Report Requirements

     A draft outline of the report on the analysis of the data to
be collected along with definitions indicating the purpose of the
 report and the audience for which it is intended.

 Quality Assurance

     The test plan should address the development of a quality
assurance program as outlined in Section 3.7.  This QA program
 should be an integral part of the measurement program and be in-
 corporated as a portion of the test plan either directly or by
reference.

 Responsibilities

     A list of persons who are responsible for each phase of the
measurement program, as defined in the schedule, both for the
 testing organization and for the plant site.
                             -16-

-------
3.3  Upwind-Downwind Sampling Strategies



     The upwind-downwind sampling method, as described in Section 2.1.3,




is used to quantify the emissions from a source to the ambient atmosphere




by measuring pollutant levels in the atmosphere.  Upwind measurements




are made within the ambient air approaching the site of the source,




using sampling equipment suitable for the specific emissions to be mea-




sured, to determine the baseline concentration of pollutants in the




air.  Downwind measurements are made of the air within the cloud of




pollutants emitted by the source, using sampling equipment similar to




that used for the upwind measurements, to determine the total of the




ambient air and the source's contribution to the concentration of pol-




lutants.  The pollutants contributed by the source to the cloud at the




sampling locations are determined as the difference between the measured




upwind and downwind concentrations.  Measurement of the wind speed and




direction at the site are combined with the pollutant concentrations at




the sampling locations in diffusion equations to back-calculate the




source strength of the emissions.  Section 3.4 and 3.5 describe the




equipment used for sampling, the criteria for sampling system design,




sampling techniques, and data reduction procedures for respectively,




survey and detailed upwind-downwind sampling programs.








3.4  Survey Upwind-Downwind Measurement System



     A survey measurement system, as defined in Section 2.3, is designed




to provide gross measurements of emissions to determine whether any




constituents should be considered for more detailed investigation.  A




survey upwind-downwind measurement system in its simplest form utilizes
                                  -17-

-------
a single upwind sampler for the determination of the concentration of




the pollutants of concern in the ambient air approaching the source of




the emissions and two or three identical downwind samplers for the de-




termination of the pollutant concentration and distribution in the am-




bient air leaving the source.   These data, combined with measurements




of the ambient air wind speed and direction, are used to calculate the




emission rate of the source.









     3.4.1  Sampling Equipment




     Pollutants that may be measured by the upwind-downwind technique




are limited to those that can be airborne for significant distances,




i.e., particulates and gases.   The gross measurement requirements for




survey sampling of particulates are best satisfied by high volume fil-




ter devices to provide data on the average emission rate, particle size




distribution, and particle composition.  Particle charge transfer or




piezoelectric mass monitoring devices may be utilized for continuous or




semi-continuous sampling of intermittent emission sources where peak




levels must be defined.




     Gaseous emissions in survey programs are usually grab-sampled for




laboratory analysis using any of a wide variety of evacuated sampling




vessels or chemical bubblers.  Continuous or semi-continuous sampling




of specific gases may be accomplished using such devices as continuous




monitor flame ionization detectors  (for hydrocarbons) and automated




flame photometric devices  (for sulfur dioxide).
                                    -18-

-------
     3.4.2  Sampling System Design

     The number and location of the devices used to collect samples is

extremely important to the successful completion of a survey upwind-

downwind sampling program, especially since the program is designed for

minimum cost and provides for no replication of samples.  The design of

the sampling system is influenced by such factors as source complexity

and size, site location and topography, and prevailing meteorological

conditions which govern the distribution of the pollutant cloud in the

ambient atmosphere.  Most situations will in general fit into some com-

bination of the following parameters:
     Source - Sources may be either homogeneous, emitting a single type
     or mixture of pollutants from each and every emission location, or
     heterogeneous, emitting different types or mixtures of pollutants
     from different locations.  The resultant cloud of pollutants will,
     for a homogeneous source, be homogeneous.  The pollutant cloud for
     a heterogeneous source may be either heterogeneous or, as a result
     of mixing by suitably directed or turbulent ambient air flow, homo-
     geneous .  The physical size of a source will determine the extent
     of the pollutant cloud and may influence its homogeneity, the prox-
     imity of different emissions to each other largely influencing the
     degree of mixing in the cloud for a given downwind distance.

     Site - Sites in general may be open on level terrain with free
     access of ambient air from all sides, partially obstructed by hills
     or buildings that interfere with or influence the ambient air flow
     either up- or downwind, or located in a valley between hills or
     large buildings that influence the air flow both up- and downwind.
     Each type of topography will influence the extent and homogeneity
     of the pollutant cloud depending on the direction of the wind flow
     relative to the obstructions.

     Meteorology - The direction of the prevailing wind determines the
     basic location of upwind and downwind samplers.  It will influence
     the pollutant cloud in every instance except that of a homogeneous
     cloud at an open level site.  In other instances, the wind may be
     directed generally across or parallel to obstructing hills or
     valleys which may result in channeling, lofting, or swirling of
     the air flow across the site that will distort the pollutant cloud.

     The homogeneity of the ambient air approaching the measurement
     site, while not in the strict sense a meteorological condition,
                                  -19-

-------
     may affect the composition and distribution of different pollutants
     within the pollutant cloud.  Contributions from sources upwind of
     the site may result in variations in the pollutant concentrations
     in the ambient air passing over the site and thus in the pollutant
     cloud as well.

     Wind speed, which can affect the cloud's size and distribution,
     need not be considered as a governing design factor since it is to
     some degree controllable by scheduling to avoid periods of either
     excessive wind velocity or calm conditions.  Wind speeds within
     normal limits are taken into consideration in data reduction cal-
     culations.
     Table 3-2 presents a matrix of 20 possible combinations of these

parameters (cloud homogeneity, site topography, wind direction and am-

bient air homogeneity).  The simplest combination, that of a homogeneous

cloud in an open level site with homogeneous ambient air, would typically

require a single upwind sampler and two downwind samplers located within

the cloud.  The complexity of the sampler system design is, in general,

increased by changes in the parameters as follows:
     Cloud Homogeneity.   A heterogeneous cloud will generally limit the
     placement of the downwind samplers to the portion of the cloud
     that contains the combined emissions from the various sources.  It
     may also require the addition of samplers in the cloud to provide
     data on the extent  of the effects of the heterogeneity and the
     consequent variability of the pollutant distributions.  This param-
     eter will not affect the upwind samplers.

     Site Topography.  Depending on the relationship of the topography
     obstructions and the wind direction, this parameter may affect
     both upwind and downwind samplers.  Hills and valleys may cause
     lofting or depression of the pollutant cloud, requiring sampler
     elevation on towers or limiting the downwind distance of samplers
     within the cloud.  They may also provide funnelling effects that
     limit the dispersion of the cloud and restrict the lateral position-
     ing of the downwind samplers.  Upwind sampler locations may be
     restricted by lofting or depression of the ambient air approaching
     the site.

     Wind Direction.  Changes in this parameter alone are not generally
     a major factor in the sampling system design.  They will dictate
                                    -20-

-------
                                                                                   TABLE 3-2


                                                                  MATRIX OF SAMPLING SYSTEM DESIGN PARAMETERS
I
N>
Cloud
Homogeneity
Homogeneous
CD
Heterogeneous
(2)
Site
Topography
Open
(1)
Hill
(2)
Valley
(3)
Open
(i)
Hill
(2)
Valley
(3)
Wind
Direction
Not a
Factor
(0)
Parallel to
Hill (1)
Over Hill
(2)
Down Valley
(1)
Across Valley
(2)
Not a
Factor
(0)
Parallel to
Hill (1)
Over Hill
(2)
Down Valley
(1)
Across Valley
(2)
Ambient Air
Homogeneity
Homogeneous
(1)
Heterogeneous
(2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homogeneous
(1)
Heterogeneous
(2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Cloud Homogeneity:
Homogeneous - Sources at site all emitting same pollutants.
- Sources of different pollutants grouped so that
emissions are mixed before sampling.
Heterogeneous - Sources at site emitting identifiably different
pollutants - no mixing before sampling.
Site Topography:
Open - Site on flat terrain ambient air access from any
direction unhindered.
Hill - Site close enough to rise in terrain or large
buildings to cause channeling or lofting Of am-
bient air.
Valley - Site between rises in terrain or large buildings.
Uind Direction:
Parallel to - Wind across site channeled against side of hill.
Hill Usually changes shape of pollutant cloud and
distribution of pollutants within cloud.
Across Hill - Wind across site from or to hill top. Can cause
lofting of depression of pollutant cloud.
Down Valley - Wind across site channeled against sides of hills.
Across Valley - Wind across site from hill to hill.


Homogeneous - Pollutants in approaching air evenly distributed.
Heterogeneous - Pollutants in approaching air measurably different
at points over site. Usually caused by emissions
from nearby upwind external source.

-------
     changes in the design in combination with other factors such as
     site topography, described above, or the presence of external
     sources, which may influence the homogeneity of the approaching
     ambient air, described below.

     Ambient Air Homogeneity.  The presence of external emission sources
     that may result in variations in the pollutant concentrations and
     distributions in the air approaching a site may require the addi-
     tion of samplers both upwind and downwind to ensure that the mea-
     surements of the pollutants of interest are not unduly influenced
     or masked.  Samplers typically are required within and outside of
     the external source cloud both upwind and downwind.
     Typical sampler locations for selected source site configurations

illustrating some of these effects are sketched in Figure 3-1.  The

configurations are identified by a four-digit number referring, in left-

to-right order, to the numbers assigned to 'the parameters identified in

the matrix of Table 3-2.  A configuration with a homogeneous cloud emitted

at a valley site with cross-valley wind direction and homogeneous am-

bient air is thus identified as 1321.



     3.4.3  Sampling Techniques

     Sampling must be scheduled and carefully designed to ensure that

data representative of the emission conditions of concern are obtained.

Effective scheduling demands that sufficient knowledge of operations

and process conditions be obtained to determine proper starting times

and durations for samplings.  The primary concern of the sampling design

is that sufficient amounts of the various pollutants are collected to

provide meaningful measurements.

     Each of the various sample collection and analysis methods has an

associated lower limit of detection, typically expressed in terms of

micrograms of captured solid material and either micrograms or parts

per million in air of gases.  Samples taken must provide at least these
                                    -22-

-------
Wind
            1101
2102
           1221
2311
                                 Legend
                               Homogeous cloud source
                          ItsfD Heterogeneous cloud source
                            0 External source
                           A \ Sampler
                           --•:•:::: Source cloud
                           ^^ External source cloud
     Fig. 3-1.  Typical sampler legations for selected source site configurations.
                                    -23-

-------
minimum amounts of the pollutants to be quantified.  The amount (M) of

a pollutant collected is the product of the concentration of the pollu-

tant in the air (x) and the volume of air sampled  (V), thus,


     M (micrograms) = x (micrograms/cubic meter) x V (cubic meters).


To ensure that a sufficient amount of pollutant is collected, an ade-

quately large volume of air must be passed through such samplers as

particle filters or gas absorbing trains for a specific but uncontrolla-

ble concentration.  The volume of air (V) is the product of its flow

rate (F) and the sampling time (T),


      V (cubic meters) = F (cubic meters/minute) x T (minutes).


Since the sampling time is most often dictated by  the test conditions,

the only control available to an experimenter is the sampling flow

rate.  A preliminary estimate of the required flow rate for any sam-

pling location may be made if an estimate or rough measurement of the

concentration expected is available.  The subsitution and rearrangement

of terras in the above equations yields Equation 3-1:
  F (cubic meters/minute) = M (micrograms)/x (micrograms/cubic meter)
     x T (minutes).                                                (3-1)
This equation permits the calculation of the minimum acceptable flow

rate for a required sample size.  Flow rates should generally be adjusted

upward by a factor of at least 1.5 to compensate for likely inaccuracies

in estimates of concentration.

     Grab-samples of gaseous pollutants provide for no means of pollu-

tant sample quantity control except in terms of the volume of the sample.
                                   -24-

-------
Care should be taken, therefore, to correlate the sample size with the

requirements of the selected analysis method.

     Sampler location is also important in obtaining representative

data.  Downwind sampler location is especially critical to ensure that

samples are taken at points known to be within the pollutant cloud at

measurable concentrations.  A rough estimate of acceptable downwind

sampler locations may be made utilizing the basic equation    for the

diffusion of gases and particulates in the atmosphere from a ground-

level source:  x = Q/ifKu, where


         X = pollutant concentrations at receptor point, gm/m3
         Q = source emission rate, gm/sec
         K = product of standard deviations of vertical and
             horizontal pollutant distribution, m2
         u = wind speed, m/sec


This equation assumes a Gaussian distribution of pollutants in both the

vertical and horizontal directions and no deposition or reaction of

pollutants at the earth's surface.

     By rearranging terms, the product of the standard deviations (K),

which are functions of the downwind distance (x) of the receptor from

the source, may be determined as a function of easily estimated or

measured parameters in Equation 3-2:


                              K - Q/TTXU,                         (3-2)
        Turner, D. Bruce, "Workbook of Atmospheric Dispersion Estimates,"
U.S. Department of Health, Education and Welfare, Public Health Service
Publication No. 999-AP-26, Revised 1969.
                                  -25-

-------
where
         Q is estimated from published emission factors,
         X is set equal to a selected value related to
           the sampling method detection limit and
         u is measured at the site.
The maximum downwind sampler distance from the source along the axis of

the wind direction (x) may then be determined from the curves of Figure

3-2, which relate K and x for various atmospheric stability categories.

These categories are listed and explained in Table 3-3.

     When suitable x-distances, which may be any distance less than the

maximum determined from Figure 3-2, have been selected, cross-wind dis-

tances (y) perpendicular to the x-axis that will ensure that samples

are taken within the limits of the cloud must be determined.  Maximum

cross-wind distances, which are a function of the distribution of the

pollutant concentrations within the cloud, are plotted as a function of

x in the curves of Figure 3-3 for the same atmospheric stability cate-

gories used in determining x.  Downwind samplers should in general be

located at two different x-distances within the limits of the maximum

as determined above and at cross-wind y distances less than the maximum

indicated in Figure 3-3 on opposite sides of the wind direction axis.

     Upwind samplers should ideally be located on the wind direction

axis just far enough upwind to prevent sampling the backwash of the

pollutant cloud.  A minimum upwind distance of x   /10, where x    is
r                                               max            max

determined using x equal to the sampling method's lower detection limit,

will usually be sufficient.
                                   -26-

-------
                 Atmospheric
                 stability
                 category
100
200          400          600          800
  Maximum downwind sampler distance from
  source along wind direction axis (x) - meters

Fig. 3-2. Maximum downwind sampler distances.
1000
                        -27-

-------
                            TABLE 3-3

                ATMOSPHERIC STABILITY CATEGORIES
Wind Speed
m/sec
< 2
2-3
3-5
5-6
> 6
Dav*
Solar Altitude*
> 60°
A
A-B
B
C
C
35°-60°
A-B
B
B-C
C-D
D
15°-35°
B
C
C
D
D
Nig
Overcast or
> 50% Clouds
-
E
D
D
D
ht
< 50% Clouds
_
F
E
D
D
*Day is one hour after sunrise to one hour before sunset.

+Solar altitude may be determined form Table 170, Solar
 Altitude and Azimuth, Smithsonian Meteorological Tables.
 Use neutral class D for overcast conditions at any wind
 speed. - Parital cloud cover (60 percent to 85 percent)
 will reduce effective solar altitude one division (e.g.,
 from > 60°. to 35°-60°) for middle clouds and two divi-
 sions (e.g., from >60° to 15°-35°) for low clouds.
                                  -28-

-------
   300
   200
   100
 E


I
 I  50
 a.
•a
 o
 u
 E
 3
 E
'x
 co
    10
Atmospheric
stability
category
                                  I
       0    100    200          400           600          800

                    Downwind sampler distance (x) - meters
                                            1000
              Fig. 3-3.  Maximum crosswind sampler distances.
                                   -29-

-------
      To illustrate the application of the equations and curves presented

 in this section,  assume a source emitting particulates into a four  meter

 per second wind at an estimated  rate of  10 grams  per  second, and a  sam-

 pler with a lower detection limit of .001 gram and  flow rate of 0.67

 cubic meter per minute.   For  a sampling  time  of 10  minutes, the required

 pollutant concentration,  x» at the sampler is x = M/FT,  where


          M = .001 gram
          F = 0.67 cubic meter/minute x 1.5 adjustment  factor =
              1 cubic  meter/minute
          T = 10 minutes,  and
          X = .001/10  = lO"1* grams/cubic  meter


 The  product of the pollutant  cloud's standard deviations, K,  is found

 in  Equation 3-2,  K =  Q/irxu, where


          Q = 10_grams/second
          X = 10 ^ grams/cubic meter
          u = 4 meters/second, and
         K =  10/ir x!0ltx4 = 8x 103 meters  squared


      To measure the emissions during midday with  clear skies,  Table  3-3

 indicates  an atmospheric  stability category B for the  four  meter/second

wind.  Figure 3-2  for  K = 8 x 103  and category B  indicates  a maximum

 sampler downwind  distance of 680 meters.   Figure  3-3 for x  =  680 meters

 and  category  B indicates a maximum cross-wind  distance of 145  meters.

      Downwind samplers must then  be  located within  the limits  of a  tri-

 angle with an apex at  the source, an altitude  of  680 meters  along the

wind  direction axis and a base 145 meters wide  on each side  of  the axis.

      The upwind sampler should be located along the wind direction axis

at a minimum distance  of x   /10 = 68 meters  from the source.
                          ma v
                          max
                                   -30-

-------
     A more detailed description of the application of this method is




presented in the appendix.









     3.4.4  Data Reduction




     When the sampling program has been completed and the samples have




been analyzed to yield pollutant concentrations in such terms as micro-




grams per cubic meter in the ambient air at each downwind sampling site,




the measured upwind concentrations are subtracted to yield the concen-




tration provided by the source at each sampler.  These values are then




back-calculated through known diffusion equations that take into account




the variables of topography and meteorology to produce statistical dis-




tributions of the concentrations within a pollutant cloud generated by




a given source.  These calculations yield source strengths of the emis-




sions in such terms as grams per unit time.  A library of computer pro-




grams to assist in the performance of the calculations is maintained in




the User's Network for Applied Models of Air Pollution (UNAMAP) at the




Environmental Protection Agency's Research Triangle Computer Center.




Additional programs may be obtained through many environmental consul-




tants.








3.5  Detailed Upwind-Downwind Measurement System




     A detailed measurement system is designed to more precisely iden-




tify and quantify specific pollutants that a survey measurement or




equivalent data indicate as a possible problem area.  A detailed system
        Bulletin American Meteorological Society, Vol. 56, No. 12,
December, 1975.
                                  -31-

-------
 is necessarily more  complex  than  a  survey  system  in  terms of  equipment,




 system  design, sampling  techniques, and data  reduction.  It requires a




 much  larger  investment in  terms of  equipment, time,  and manpower and




 yields  data  detailed and dependable enough for direct action  toward




 achieving emissions  control.  Detailed systems in general employ sam-




 pling arrays or networks to measure the concentration and distribution




 of specific  pollutants in  the ambient air  approaching and leaving a




 source.  These actual measurements  of the  pollutant  distribution within




 a cloud and  the variations in meteorological  conditions during the sam-




 pling period replace the assumptions utilized in survey sampling sys-




 tems.  Detailed systems are frequently employed to compare emissions at




 different process or operating conditions  to  determine which conditions




 dictate the  need for emission control.




     The data provided by the sampling arrays are processed in conjunc-




 tion with more detailed meteorological data which are taken simultan-




 eously to determine source emission rates and ambient distributions in




much the same manner as the simpler survey systems.








     3.5.1   Sampling Equipment




     The pcLltants to be characterized by  a detailed upwind-downwind




 sampling system fall into the same two basic classes—airborne partic-




ulates and gases—as those measured by survey systems.   Detailed sam-




pling and analysis equipment is generally selected to obtain continuous




or semi-continuous measurements of specific pollutants rather than sim-




ple grab-sampled  measurements.




     Particulate  samples are collected using filter impaction, piezo-
                                  -32-

-------
 electric, particle change transfer, light or radiation scattering, elec-




 trostatic, and size selective or adhesive impaction techniques.  Gases




 are sampled and analyzed using flame ionization detectors, bubbler/im-




 pinger trains, non-dispersive infrared or ultraviolet monitors, flame




 photometry, and other techniques specific to individual gaseous pollu-




 tants .




     The selection of suitable sampling equipment should be influenced




 by such considerations as portability,  power requirements, detection




 limits and ease of control.









     3.5.2  Sampling System Design




     The basic  criteria reviewed  in  Section 3.4.2 for  the design of a




 survey sampling system are generally applicable to  the  design of a de-




 tailed system.   The need for  replacement  of survey  assumptions as  to




 pollutant distribution with actual  measured values, however,  most  fre-




 quently requires  the  design of a  sampling array or  network that will




 provide samples  of  a  distribution at various distances  downwind of the




 source  in both  the  horizontal and vertical  directions.   Sampler loca-




 tions may generally be  determined in the  same manner as  those for  a




 survey  system.   For detailed measurements,  each  location must provide




 for sampling across a  section of  the  pollutant  cloud horizontally  and/




or vertically.  Horizontal distributions may be measured by adding a




number  of  samplers  (usually at least  two) at either side of the  survey




sampler location at distances estimated to yield significantly differ-




ent pollutant concentrations.   Vertical distributions may be measured




by placing a tower of suitable height at each survey sampler  location
                                   -33-

-------
and adding  samplers over a range of heights on each  tower.   Combina-


tions of horizontal and vertical distributions may be measured by plac-


ing a grid  of horizontally and vertically  spaced samplers at each


survey sampler location.  Actual numbers of samplers, their  spacing,


and heights of towers required must be determined for each location.  A


rough guide for estimating the required spacing is presented in Section


3.5.3.




    3.5.3.  Sampling Techniques


    The guidelines presented in Section 3.4.3 for the design and loca-


tion of samplers for a survey system are applicable  to detailed systems.


The assumption of a Gaussian distribution  of pollutants in the cloud,


sufficient  for data reduction in survey systems is reasonable as a rough


guide to locating samplers within the pollutant cloud as in  Section


3.4.3, and for the spacing of sampling arrays as outlined below.


    The approximate concentration of a specific pollutant within a


cloud in which concentrations vary in accordance with a Gaussian dis-


tribution at a given downwind distance from the source is greatest at


ground level on the wind direction axis of the cloud.  Assigning this


concentration the unit value x , the concentration (x) at any cross-


wind distance (y) from die axis is expressed as
                           X = XQ e



The term F(y) may be expressed as y/y , where y  is the maximum cross-


wind sampler distance determined from Figure 3-3.  The relationship x/X
                                                                       0

is plotted .as a function of y/y  in Figure 3-4.  This may be used to
                                   -34-

-------
    1.0
    0.8
2
o
'5  0.6
8
4-1

I
2  0.4
    0.2
                         0.5                1.0                1.5

                           Y/Y_ - crosswind distance ratio
2.0
           Fig. 3-4.  Pollutant concentration ratios for crosswind locations.
                                      -35-

-------
 determine the probable concentration at a sampler location relative to



 the concentration at the axis and the concentrations at lateral dis-



 tances from that location to assist in the horizontal spacing of sam-


 plers in an array.



      The concentration in the vertical direction from any ground level



 point will decrease as the height,  Z, increases in a similar relation-



 ship.  The ratio of the concentration at the elevated point to that at



 ground level,  Xi/X» is plotted in Figure 3-5 as a function of Z/Z , where
                n                                                 ni


 Z  is a function of the downwind distance from the source and the atmo-



 spheric stability as plotted in Figure 3-6.   Figure 3-5 may be used to



 determine the  relative concentrations at elevated points to assist in



 the design of  sampling towers and the vertical spacing of samplers in


 an  array or grid.



     In general, arrays should be designed to provide data at concentra-



tions approximately two to four times greater or less than the concen-



tration at a selected ground level sampling point.  Physical limitations



at the site or very unstable atmospheric conditions will often preclude



the compliance with this design guideline by limiting the available



horizontal positions or by requiring an impractical tower height.  In



such situations, the need to adjust the requirements of the guideline



must be recognized and the array designed to compensate for the limita-



tions.



     Upwind sampling arrays will generally be less complex than downwind



arrays unless a nearby pollutant source results in a heterogeneous am-



bient air mix.  In this case, the guidelines for downwind array design



presented in this section may have to be applied to the upwind array
                                   -36-

-------
I    I    I   I        I    I    I    I
              0.5                1.0



               Z/Z_  - vertical distance ratio
1.5
2.0
 Fig. 3-5.  Pollutant concentration ratios for vertical locations.
                            -37-

-------
   500
   100 —
N
s
t
                               Atmospheric
                               stability
                               category
            100    200
800
                 400           600
          Downwind distance (x) - meters
Fig. 3-6.  Vertical concentration distribution factors.
1000
                                    -38-

-------
design.




     Wind speed and direction should be measured at each sampler or




array location.  Pretest survey observations should indicate whether




stratification will occur to a degree which will require wind data at




more than one level.




     An example of the application of these guidelines to the design of




survey and detailed systems for the measurement of pollutants at a. Port-




land cement plant is presented as an appendix to this document.









     3.5.4.  Data Reduction




     Samples are analyzed to yield concentrations of specific pollutants




in such terms as micrograms per cubic meter at each sampling site.




Measured upwind concentrations are substituted into appropriate diffu-




sion equations to provide ambient air background concentrations at each




downwind site and the background concentration subtracted from the mea-




sured downwind concentration at each site to yield the source contri-




bution.   These values are then substituted into diffusion equations to




back-calculate source strengths in terms of grams per unit time, util-




izing UNAMAP or other available computer programs.








3.6  Atmospheric Tracers




     In some instances, prevailing process or meteorological conditions




prohibit the collection of samples containing measurable, clearly de-




fined amounts of a specific pollutant for the back-calculation of source




strengths.   In many such cases, the atmospheric tracer method may be




employed to determine a typical distribution of a general class of pollu-
                                   -39-

-------
 tant analogous to the pollutant of concern.

      The use of tracers should be considered under any of the following

 circumstances:


      When the pollutant background concentration is either excessively
      high or inhomogeneous.   This can be caused by significant emis-
      sions from external upwind sources.

      When the fugitive emissions are  of  such a  complex nature that an
      excessive number of downwind vertical profiles are required  to
      characterize the emissions.

      When physical limitations  prohibit  the  installation of  adequate
      instrumentation  for specific pollutant  concentration measurement.

      When the nature  of the  specific  pollutant  prohibits its measure-
      ment with acceptable instrumentation or indicates large probable
      errors  in measurement.

      When estimates of fugitive emissions are being made for non-oper-
      ating processes  or planned operations.


      The  atmospheric  tracer  method, which may be considered  as a  special

 detailed  system,  consists of the  introduction into  the atmosphere,  at

 the  source site under consideration,  of  a readily identifiable material

 similar  in the character  of  its  diffusion in the atmosphere  to the pollu-

 tant  of  concern.   The quantity  released  may  be  controlled  to provide

 readily measurable concentrations.  A detailed  downwind  measurement

 system, designed  using  the guidelines  of Section 3.5,  is used to  col-

 lect  samples  of  the tracer and  to  determine  its  dispersion for the known

 and  controllable  source strength.  This  dispersion  will  be analogous to

the dispersion of the  pollutant of concern and will permit the predic-

tion of pollutant concentrations for a range  of  source strengths.
                                   -40-

-------
    3.6.1  Tracers and Samplers




    Both particulate and gaseous atmospheric tracers are in general




use.  The most commonly used particulate tracers are zinc-cadmium sul-




fide and sodium fluorescein (urinine dye).  The primary gaseous tracer




is sulfur hexafluoride (SF6).




    Zinc-cadmium sulfide is a particulate material which can be ob-




tained in narrow size ranges to closely match the size of the pollutant




of concern.  The material is best introduced into the atmosphere in dry




form by a blower type disseminator although it can also be accomplished




by spraying from an aqueous or solvent slurry.  The zinc-cadmium sul-




fide fluoresces a distinctive color under ultraviolet light which pro-




vides a specific and rapid means of identification and quantification




of the tracer in the samples.




    Sodium fluorescein is a soluble fluorescing particulate material.




It is normally spray disseminated from an aqueous slurry solution to




produce a particulate airborne plume, the size distribution of which




can be predetermined by the spraying apparatus.  Sodium fluorescein can




be uniquely identified by colorimeter assessment.




    Sulfur hexafluoride is a gas which can be readily obtained in ordi-




nary gas cylinders.  Sulfur hexafluoride can be disseminated by meter-




ing directly from the gas cylinder through a flow meter to the atmosphere.




The amount disseminated can be determined by careful flow metering and/or




weight differentiation of the gas cylinder.




    Particulate tracers are usually sampled with filter impaction de-




vices or, for particles over 10 microns in diameter, the more easily




used and somewhat less accurate Rotorod sampler which collects particles
                                   -41-

-------
on an adhesive-coated U- or H-shaped rod which is rotated in the am-




bient air by a battery-driven electric motor.




     Sulfur hexafluoride gaseous samples are collected for laboratory




gas chromatograph analysis in non-reactive bags of such materials as




Mylar.









     3.6.2  Tracer Sampling System Design




     All of the design guidelines presented in 3.4.2 and 3.5.2 may be




applied to the design of a tracer sampling system as site conditions




dictate.  Their application is, in general, simplified since the source




strength may be controlled to provide measurable tracer concentrations




at readily accessible sampling locations.




     A single upwind sampler will usually be sufficient to establish




that no significant amount of the tracer material is present in the




ambient atmosphere approaching the source.









     3.6.3  Tracer Sampling and Data Analysis




     The methods introduced in Section 3.4.3 and 3.5.3 for determining




sampler design and location are fully applicable to tracer sampling.




Like the design guidelines, they may be more easily applied because the




source strength is easily controlled.




     The analysis of the data is also simplified since the source strength




is known and no back-calculation is required.
                                   -42-

-------
3.7  Quality Assurance

     The basic reason for quality assurance on a measurement program is

to insure that the validity of the data collected can be verified.

This requires that a quality assurance program be an integral part  of

the measurement program from beginning to end.  This section outlines

the quality assurance requirements of a sampling program in terms of

several basic criteria points.  The criteria are listed below with  a

brief explanation of the requirements in each area.   Not all of the

criteria will be applicable in all fugitive emission measurement cases.


     1.  Introduction

             Describe the project organization, giving details of the
         lines of management and quality assurance responsibility.

     2.  Quality Assurance Program

             Describe the objective and scope of the quality assurance
         program.

     3.  Design Control

             Document design requirements and standards applicable  to
         the measurement program as procedures and specifications.

     4.  Procurement Document Control
             Verify that all design specification accompany procurement
         documents such as purchase orders.

     5.   Instructions,  Procedures,  Drawings

             Prescribe  all activities that affect the quality of the
         work performed by written  procedures.   These procedures must
         include acceptance criteria for determining that these activ-
         ities are accomplished.

     6.   Document Control

             Ensure that the writing, issuance,  and revision of proce-
         dures which prescribe measurement program activities affecting
         quality are documented and that these  procedures are distrib-
                                   -43-

-------
      uted  to  and used at  the  location where the measurement  program
      is  carried out.

  7.   Control  of Purchase  Material, Equipment, and Services

         Establish procedures to ensure  that purchased material
      conforms to the procurement specifications and provide  veri-
      fication of conformance.

  8-   Identification and Control of Materials, Parts, and Components

         Uniquely identify all materials, parts, and components
      that  significantly contribute to program quality for trace-
      ability  and to prevent the use of incorrect or defective ma-
      terials, parts, or components.

  9.   Control of Special Processes

         Ensure that special processes are controlled and accom-
      plished by qualified personnel using qualified procedures.

10.   Inspection

          Perform periodic inspections where necessary on activities
      affecting the quality of work.  These inspections must  be or-
      ganized and conducted to assure detailed acceptability  of
      program components.

11.   Test Control

         Specify all testing required to demonstrate that applicable
      systems and components perform satisfactorily.   Specify that
      the testing be done and documented according to written proce-
      dures, by qualified personnel, with adequate test equipment
      according to acceptance criteria.

12.   Control of Measuring and Test Equipment

         Ensure that all testing equipment is controlled to  avoid
     unauthorized use and that test equipment is calibrated  and
     adjusted at stated frequencies.   An inventory of all test equip-
     ment must be maintained and each piece of test equipment labeled
     with the date of calibration and date of next calibration.

13.  Handling, Storage,  and Shipping

         Ensure that equipment and material receiving,  handling,
     storage, and shipping follow manufacturer's recommendations to
     prevent damage and deterioration.   Verification and documentation
     that established procedures are followed is required.
                               -44-

-------
14.  Inspection, Test, and Operating Status

         Label all equipment subject to required inspections and
     tests so that the status of inspection and test is readily
     apparent.  Maintain an inventory of such inspections and oper-
     ating status.

15.  Non-Conforming Parts and Materials

         Establish a system that will prevent the inadvertent use
     of equipment or materials that do not conform to requirements,

16.  Corrective Action

         Establish a system to ensure that conditions adversely
     affecting the quality of program operations are identified,
     corrected, and commented on; and that preventive actions are
     taken to preclude recurrence.

17.  Quality Assurance Records

         Maintain program records necessary to provide proof of
     accomplishment of quality affecting activities of the measure-
     ment program.  Records include operating logs, test and in-
     spection results, and personnel qualifications.

18.  Audits
         Conduct audits to evaluate the effectiveness of the mea-
     surement program and quality assurance program to assure that
     performance criteria are being met.        . ••'
                              -45-

-------
4.0  ESTIMATED COSTS AND TIME REQUIREMENTS




     Table 4-1 presents a listing of the conditions assumed for esti-




mating the costs and time requirements of upwind-downwind fugitive emis-




sions sampling programs using the methodology described in this document.




Four programs are listed, representing minimum and more typical levels




of effort for each of the survey and detailed programs defined in Sec-




tion 3.3.  The combinations of conditions for each program are generally




representative of ideal and more realistic cases for each level and will




seldom be encountered in actual practice.  They do, however, illustrate




the range of effort and costs that may be expected in the application of




the upwind-downwind technique except in very special instances.









4.1  Manpower




     Table 4-2 presents estimates of manpower requirements for each of




the sampling programs listed in Table 4-1.  Man-hours for each of the three




general levels of Senior Engineer/Scientist, Engineer/Scientist, and




Junior Engineer/Scientist are estimated for the general task areas outlined




in this document and for additional separable tasks.  Clerical man-hours




are estimated as a total for each program.  Total man-hour requirements




and approximately 500 man-hours for a simple survey program and 1500




man-hours for a more complex r.urvey program; and 2800 man-hours for a




simple detailed program and 4500 man-hours for a more complex detailed




program.
                                   -46-

-------
               TABLE 4-1

CONDITIONS ASSUMED FOR COST ESTIMATION
OF UPWIND - DOWNWIND SAMPLING PROGRAMS
Parameter
Site
Location
Emmission
Source
Emission
Character
Wind
Measurement
Sample
Sites
Samplers
Towers
Experiments
Estimated
Basic
Accuracy
Survey Program
Simple
Open Area-
Accessible
Well Defined
Steady
External
Source
One Upwind
Two Downwind
3
0
1
± 500%
Complex
Congested-
Limited Access
Complex
Steady
Measured
On Site
One Upwind
Three Downwind
8
4 Low
1
± 150%
Detailed Program
Simple
Open Area-
Accessible
Well Defined
Cyclic
Measured
On Site
Vertical
Arrays-
One Upwind,
Two Downwind
16
4 High
2
± 125%
Complex
Congested-
Limited Access
Complex
Cyclic-
Measured
at Two Levels
Two Measure-
ments On Site
Grid Arrays-
One Upwind,
Two Downwind
30
4 High -
Grids
4
± 75%
                   -47-

-------
                                                                  TABLE 4-2


                                           ESTIMATED MANPOWER REQUIREMENTS FOR UPWIND - DOWNWIND

                                                             SAMPLING  PROGRAMS
00
i
— 	 	 	 	 Estimates in Man-Hemrs

Task
Pretest Survey
Test Plan Preparation
Equipment Acquisition
Field Set-Up
Field Study
Sample Analysis
Data Analysis
Report Preparation
Totals
Engineer/Scientist Tota]
Clerical
Grand Total
Survey Programs
Senior
Engr/Sci
4
8
4
8
24
20
20
16
104


Engr/
Sci
12
12
4
12
48
20
20
16
144
472
40
512
Junior
Engr/
Tech
0
0
12
12
120
36
36
8
224



Senior
Engr/Sci
8
12
4
16
60
60 ;
60 j
64
284
!
|
1

	 • 	 1 	 1 	 i 	 	

Engr/
Sci
24
16
8
48
200
60
60
64
480
1376
120
1496
Junior
Engr/
Tech
0
4
28
64
268
108
108
32
612





Senior
Engr/Sci
Q
12
8
80
120
120
120 |
160 i
628 i
	 Detailed Programs 	
Engr/
Sci
24
24
24
140
300
240
240
80
972
J2628
i
200
2828

Junior
Engr/
Tech
0
12
48
100
536 :
96
96
40 ;
1028
Senior
Engr/Sc
12
16
12
120
180
180
180
200
900
i





uomplex
Engr/
Sci
36
32
36
280
560
320
320
200
1784
4240
Junior
Engr/
Tech
16
12
52
240
796
180
180
80
1556

280
4520 !


-------
4.2  Other Direct Costs




     Table 4-3 presents estimates for equipment purchases, rentals,




calibration, and repairs; on-site construction of towers and platforms;




shipping and on-site communications for each of the listed programs.




Total costs are approximately $4500 for a simple survey program and




$17,000 for a more complex survey program; and $34,000 for a simple




detailed program and $64,000 for a more complex detailed program.









4.3  Elapsed-Time Requirements




     Figure 4-1 presents elapsed-time estimates for each of the listed




programs broken down into the task areas indicated in the manpower es-




timates of Table 4-2.  Total program durations are approximately 12




weeks for a simple survey program and 17 weeks for a more complex sur-




vey program; and 21 weeks for a simple detailed program and 41 weeks




for a more complex detailed program.









4.4  Cost Effectiveness




     Figure 4-2 presents curves of the estimated cost effectiveness of




the upwind-downwind technique, drawn through points calculated for the




four listed programs.  Costs for each program were calculated at $30




per labor hour, $40 per man day subsistence for field work for the man-




power estimates of Table 4-2, plus the other direct costs estimated in




Table 4-3.
                                   -49-

-------
-os-
H P < H CO O W
O P fD i-t p* o ua
H 1 O* (» H- W^TJHP SIPOOSJCOCOC
>• COH-H-XJ M M O CO H-  P> H- (D P> H-
f1 H- O M "O (D pJ t! rt f( "d M 3 .a B "d
rt M (D H- O rt (D i-i H- » H- Q- C "d B
TO (D i-l 3 rf Mi i-l C 3H-CT (!) M (D
OQ i-iown OQ i-i i-< 3; 3 re 3
OJOJO H-i-t ft MftoftO^n-
orem os H- na rt ta m
333 P)tn o 0) H-COI-(ITJ
grtrt I-1- 3 H OC C
3MM tc (o n> o>co
M-05 Ort en BMP'
n O o en fH o CD
Co ?r • (D 3 p1 os
rt 1 05 rt P) fD
o T3 ^d re
3 • CD C
en i-t
O
CD
X
re
o
o

-c/>
QT\
00
h~*
o

*»
o
. 0

-0>
u>
vO
o
o

MMNJ Ln Ml— 'VOOO
OOOO O *-JUiOO
OOOO OOO OOOOOO


1— ' NJ I-1 LO NJ J>
OJOi *» (-"OO ONiLnJ^OO
OON O OOO OUiOOO
OOOO OOO OOOOOO

I-1 t-1 00 M U> *- VO
OOOO OOO OOOOOO
OOOO OOO OOOOOO

l-> I-I
MM MVO MOivOOO
vOOOLnM U) O O OOOiOOO
OOOO OOO OOOOOO
OOOO OOO OOOOOO

o
o
CO
rt
H
rt

a














CO
1
nT

n
o
9
*w
re*
X
CO
L J,
|
re
0
o
B
T3
M
re
X





















co
^
re
o
OQ
i-i
Co

en




0
re
CD
H-
M
o
OQ
P)
B
CO


                                         M

-------
                                                                     Weeks
             Task        15          10         15         20         25          30         35         40         45
                       "  I  I  I  I   I  I  i  I  I  I  i   i  I  llJJIilllllllllillllllllllllil J	I
I
01
            Pre-test
            survey
           Test plan
           preparation
Equipment
acquisition
             Field
             set-up
             Field
             study
            Sample
            analysis
            Data
            analysis
           Report
           preparation
                                         s^
k
                                      ^^
                                        mi
                                    •=fe
                                          VTTTi
                                                                                   Simple survey program
                                                                                   Complex survey program
                                                                                   Simple detailed program
                                                                                            detailed program

                                                       15
                                                                  25
                                                           35
                                                                    Weeks

                                     Fig. 4-1. Elapsed-time estimates for upwind-downwind sampling programs.
45

-------
500^
400
300
200
100
               Survey program
                  Detailed program
               50
100         150          200
 Cost in thousands of dollars
250
300
         Fig. 4-2. Cost-effectiveness of upwind-downwind sampling programs.
                                     -52-

-------
        APPENDIX A




TEST PROCEDURES APPLICATION
           -53-

-------
A.1.0  INTRODUCTION




       This appendix presents an application of the upwind-downwind




fugitive emissions measurement system selection and design criteria to a




Portland cement manufacturing plant.  The criteria for the selection




of the method and the design procedures for both survey and detailed




sampling systems as presented in Sections 3.4 and 3.5 of this document



are discussed.
                                  -54-

-------
A.2.0  BACKGROUND INFORMATION

       The following information relative to the operation of  the  sub-

ject Portland cement manufacturing plant would ordinarily be gathered

from interviews and observations during a visit to the plant for a pre-

test survey.

     Portland cement is made from a mixture of finely ground calcareous

(lime component) and argillaceous (alumina component) materials.   The

four major steps for producing Portland cement are:


     (1)  Obtaining raw materials and reducing their size,

     (2)  Grinding, blending and homogenization of these materials
          to obtain desired composition and uniformity,

     (3)  Heating to liberate carbon dioxide and burning to form
          clinker,

     (4)  Grinding or fine pulverization of the clinker with
          addition of gypsum.


     At this location, shown in Figure A-l, limestone is quarried  at

the site by dragline buckets, pulverized in a hammer mill, mixed with

water and pumped to raw material storage.  Other raw materials are de-

livered to storage by rail.  Ball mills reduce first the limestone and

then a limestone-clay mixture to a fine slurry which is stored in  con-

crete tanks prior to its introduction to the rotary kilns.  The slurry

is dried and burned at about 2700°F to form clinker, which is cooled

and stored in bins until needed for the finish grinding operation  where

it is pulverized and mixed with gypsum to produce the finished product.

The cement is stored in silos prior to bagging or transfer to bulk con-

tainer  trucks and railroad cars for shipping.
                                  -55-

-------
                                                        State road
                                                         (3 - lane)
                                                            III
                                            Private roads
                                              (2 - lane)
l I I l  l l I l I l I  I »
Railroad spur line
               Fig. A-1.  Portland cement plant site layout.
                              -56-

-------
     The plant operates on a three-shift, round-the-clock production




schedule including all operations except shipping and unloading of rail-




delivered raw materials, which are normally carried out only on the




8:00 a.m. to 4:00 p.m. shift.  The plant produces about 300 barrels of




finished product per hour, consuming about 600 pounds of raw materials




for each 376 pound barrel produced.




     The raw materials and the finished product are essentially dust;




the principal emissions are also dust.  The largest contributor is the




kiln used to produce the clinker, where the dried mixture becomes sus-




pended in the combustion gases as dust and is delivered through the




stack to the atmosphere.  A multi-cyclone/electrostatic precipitator




combination removes about 95 percent of the dust before it is vented to




the stack.  Other sources of dust are the ball mills, materials trans-




fer operations and packaging operations.  Hoods at the ball mills and




packing house are utilized to capture and transmit about 85 percent of




their emissions to a bag house.  The quarry operation at this plant




contributes little or no dust since the entire process is conducted




with the material in a wet condition.




     The EPA estimates for uncontrolled emissions, as published in the




Office of Air Programs Publication AP-42, Compilation of Air Pollutant




Factors, are 15 to 55 pounds from the kiln and 2 to 10 pounds from all




other sources for each barrel of cement produced.  If the assumed 95




percent effectiveness of the stack controls is correct, 0.75 to 2.75




pounds per barrel could be transmitted to the atmosphere from the stack.




Assuming that 80 percent of all other emissions are hooded, 0.5 to three




pounds per barrel could be transmitted to the atmosphere as fugitive
                                 -57-

-------
emissions.




     The prevailing daytime wind at the plant is from a general easter-




ly direction and averages 10 miles per hour over open, flat, partly




swampy terrain.
                                  -58-

-------
A.3.0  METHOD SELECTION

       Selecting the most practical method to measure the amount of

emitted pollutants reaching the ambient atmosphere involves evaluating

the site, processes and pollutants concerned in terms of the criteria of

Section 2.2 as follows:


     Site Criteria - the various sources at the site are remote from
     one another, both indoors and outdoors, and are not small enough
     to be hooded or otherwise enclosed.

     Process Criteria - emissions are essentially the same from all
     sources at the site with no interfering reactions between emis-
     sions or with other constituents in the ambient atmosphere.
     The process is continuous and does not entail any limitations
     as to the timing of sampling.

     Pollutant Criteria - emissions to be measured are particulates
     whose generation rate and dilution in  the ambient air will pro-
     vide measurable concentrations within  reasonable distances of
     the source.


     The site criteria are, in this case, the determining factors in

selecting the measurement method.  Since the emissions cannot be con-

tained or directed in any manner, only  the  upwind-downwind measurement

method may be successfully utilized to  determine  the plant's contribu-

tion to  the particulate  concentration  in the local atmosphere.

     The basic question  to be  answered by the measurement program  is

"Does  the rate of particulate  emissions from  the  plant exceed  the

accepted regulatory agency standard?"   This question can be  answered

by  a survey program average measurement of  the  total particulate emis-

sions  from the plant,  including  emissions from  the kiln  stack.   If

the survey program indicates that  the  plant's emissions  do exceed  stan-

dards, the question to be answered will then be  "What actions  are  neces-
                                   -59-

-------
sary to reduce emissions to an acceptable rate?"  The answer to this




question requires that the rates of the specific sources of the emis-




sions be separately quantified.  This will require the increased accur-




acy and extent of measurements of a detailed program.  The design of




both systems is described in the following sections.
                                 -60-

-------
A.4.0  SURVEY MEASUREMENT SYSTEM



       To determine the total plant contribution of particulates to the




atmosphere, measurement must be made of the approaching ambient air




containing upwind and all background emissions.  In this case, a single




upwind sampler located between the kiln building and the road to the




east will include the general ambient background particulates plus the




particulates contributed by traffic on the road.  A ground level sampler,




located about 200 meters from the kiln, should provide an accurate mea-




surement.  The downwind measurement must include the contributions from




all the sources at the site, which may be considered as emanating from




a line source at ground level with an overlay of emission from the ele-




vated stack, as illustrated in Figure A-2.  To ensure that the stack




emission contribution to the cloud is being measured, one downwind sam-




pler is located within the estimated confines of the stack plume and




others outside the stack plume as shown on Figure A-2.









A.4.1  Sampler Location




       For a high volume sampler sampling 18 cubic feet per minute, a




desired sample weight of 100 micrograms and a 60 minute sampling time,




the particle concentration required at the sampling point is:   (per




Equation 3-1)






         X = M/FT = lO"4 (gm)/0.5  (m3/min) x 60  (min)




         X = 3.3 x 10"6  (gm/m3)
                                  -61-

-------
*D1
                              *D3
                                                         Stack plume
                                                               100 meters
               Fig. A-2.  Portland cement plant emissions clouds.
                                  -62-

-------
      Local emission limitations,  promulgated on a process weight basis,




 permit 30 pounds per hour of particulate to be transmitted to the at-




 mosphere from all sources.   In order to measure this total emission rate




 in a 10 mile per hour (4.47 m/sec)  wind with the proposed samplers, the




 product of the standard deviations  used to determine the maximum dis-




 tance from the source that  samplers may be located is found using:






                                K  =  Q/irxu                (Equation 3-2)




  K - 30 x 454 x ^L (fS^/u x 3.3  x lO'6  (g)  x 4.47 (^)  - 8 x 10*






      Table 3-3 indicates  the use  of an  atmospheric stability category B




 for clear midday conditions and the wind speed  of 4.47 meters per sec-




 ond.   Figure  3-2 indicates  a maximum sampler downwind distance well in




 excess  of one kilometer for K = 8 x 10^  and category B,  so that any




 sampler location within one kilometer downwind  of the plant  will provide




 satisfactory  measurements.   To  ensure that  the  stack emissions are  also




 adequately measured,  one  sampler  is  located along the wind direction




 axis  through  the  stack  at a distance  of  800 meters  from  the  stack,  at




 point Dl  on Figure A-2.   Two  additional  samplers  are located within the




 fugitive  emissions cloud  outside  the  stack  plume  at  points D2,  300




meters  from the kiln  structure  on its wind  direction centerline,  and D3,




 500 meters from  the kiln  at  100 meters to the south  (cross-wind)  of  its




centerline.




     Samples  are taken  simultaneously at the upwind  and  three downwind




locations for a one hour period chosen to include activities  in  all




phases of the process- kiln operation, grinding, packaging and all phases
                                 -63-

-------
of material transfer including bulk product loading and raw material




unloading.  The samples are analyzed to determine particulate concen-




trations at the sampler locations, which are then used in computer pro-




grammed diffusion equations to determine the source strengths of the



fugitive and stack emissions.
                               -64-

-------
A.5.0  DETAILED MEASUREMENT SYSTEM

       Assuming that the survey measurements indicate an emission rate

in excess of the local regulations, say 40 pounds per hour, a detailed

system must be designed to more accurately quantify the emissions from

the separate sources at the plant.

     The separate sources are identified as individual particulate

clouds on Figure A-3.  Their characteristics and schedules are as fol-

lows:
     (1)  Flotation Tanks - continuous low level emissions.  Cloud
          usually isolated.

     (2)  Ball Mill and Slurry Tanks - continuous emissions.  Cloud
          usually mixed with (3).

     (3)  Raw Materials Storage - continuous low level emissions,
          higher emissions during day shift material unloading opera-
          tions.  Cloud always mixed with (2).

     (4)  Packing and Shipping - emission level variable with activity
          on day shift only.  Cloud always partially mixed with  (5).

     (5)  Finish Grinding Mill - continuous emissions.  Cloud partially
          mixed with  (4).

     (6)  Stack - continuous emissions.

     (7)  Materials Transfer - continuous low level emissions as  back-
          ground to all except (1).
     Assuming that the prevailing wind direction  remains  from the  east:
     Cloud  (1) may be individually measured at  any  time.

     Cloud  (2) may be individually measured only when material  un-
     loading operations are shut down - measurement would  be  improved
     by wetting down raw materials.

     Cloud  (3) may not be individually measured.  Fjnissions may be
     quantified by measuring total of clouds  (2) and  (3) and  subtracting
     individual measurement of  (2).
                                 -65-

-------
                                                        Flotation
                                                   Y"^\ tanks
                                          Finish
                                          grinding
                                          mill
                                                100 meters
Fig. A-3. Portland cement plant separate source clouds.
                        -66-

-------
      Cloud (4)  may not be individually measured.  Missions may be
      quantified by measuring total of clouds (4) and (5) and sub-
      tracting individual measurement of (2).

      Cloud (5)  may be individually measured when packing and shipping
      operations are shut down.

      Cloud (6)  emissions may be measured by stack sampling at any time.

      Cloud (7)  emissions may not be individually measured.   Their low
      level background contribution is present in all clouds measured
      except (1).


      To measure the source strengths associated with clouds (1) through

 (5),  a network  of  individual arrays may be set  up as follows:


                Array [1]  - in  cloud (1)

                Array [2]  - in  combined clouds  (2)  and  (3)

                Array [3]  - in  combined clouds  (5)  and  (6)


      Samples  taken during  first shift operations using  all  three arrays

will  provide measurments  of  the particulate concentrations  in  cloud

 (1),  the combined  concentrations  of  clouds  (2)  and  (3)  and  the combined

concentrations  of  clouds  (4)  and  (5).   Samples  should be taken during

materials  unloading operations  to  provide measurements  of the  maximum

concentrations  of  cloud  (3)  and during  maximum  activity level  in the

shipping area to provide measurements of the maximum concentrations  of

cloud (4).

      Samples taken during  second or  third shift  operations  using arrays

[2] and [3] will provide measurements of the particulate concentrations

of clouds  (2) and  (5).

     Stack samples may be  taken at any  convenient time.

     Analyses of the samples will provide particulate concentrations  at
                                -67-

-------
 the array locations for each source,  which may then be back-calculated




 to  provide equivalent source strengths,  which,  with appropriate sub-




 tractions described above,  will  give  individual source strengths.




      The  flotation tanks are located  very nearly at ground level and




may reasonably  be  considered a ground level source.   Array [1]  may




 therefore be  composed of only ground  level samplers located across the




 cloud (1)  generated by these tanks.




      Raw  material  storage generates a ground level  source  cloud (3),




while the ball mills  and the slurry tanks generate  an elevated  cloud




 (2).   The  array  [2]  used to  sample these  clouds must then  employ both




ground level  and elevated samplers located across the portion of the




cloud  combining the emissions  of both sources.




      Packing  and shipping operations  generate  a cloud (4)  of both  ground




level  and  elevated  emissions,  as do the  finish  grinding mill and clinker




storage in cloud (5).  Array  [3] must then be  composed of  ground level




and elevated  samplers  located  across  the  portion of  the cloud combining




the emissions of both  clouds.




     Assumptions as  to the approximate source  strengths for each of  the




sources are made to provide  the starting  points  for  determining array




locations and spacing.   Based  on the  40 pound per hour rate for the




total emissions indicated by  the survey measurements,  a source  strength




of eight pounds per hour  is assigned  to each of  the  sources of  clouds




(2)  through (5), and four pounds per  hour  for the sources  of clouds  (1)




and (7).   The conditions of the survey example  of Section  A.4.1; with a




10 mile per hour wind from the east,   an atmospheric  stability category




B, a desired sample weight of  100 micrograms and a 60 minute sampling
                                -68-

-------
 time are assumed to apply to the detailed system.

      The best  locations for the arrays, each within the clouds they are

 designed to measure and away from the influences of other clouds, are

 shown at Al, A2 and A3 on Figure A-3.

      For array [1], located about 250 meters downwind of the source of

 cloud (1) in order to avoid the influence of neighboring clouds, the

 approximate particle concentration on the wind direction axis at ground

 level is determined from Equation 3-2, rearranged as


                            X = Q/irKu, where


          X  = concentration (gm/m3)
          Q  = source strength = 4 (Ibs/hr) = 0.5  (gm/sec)
          K  = 110 (m2) - from Figure 3-7
          u  = 4.47  (m/sec),  and


                            X = 3.2 x 10~5(gm/m3)


      The required  sampler flow rate is determined from Equation 3-1,

 rearranged  as


                            F = M/xT,  where


          F  = flow rate (m3/min)
          M  = sample weight  = 10"1*  (gin)
          X  = 3.2 x  ID"5  (gm/m3)
          T  = sampling  time  = 60 (min),  and

                       F  =  5.2  x 10-2  (m3/min)


     The  cross-wind spacing  of the samplers in  the array is  deter-

mined by assigning  a particle  concentration desired to be measured at

a sampler location  of about  1/2 the concentration at the wind direc-
                                 -69-

-------
tion axis, or, x = 1.6 x 105 (gm/m3), and calculating its ratio to the




calculated concentration on the wind direction axis, x •  In this case,




X/X  = 0.5.  Figure 3-4 indicates a value of 0.91 for the cross-wind




distance ratio y/y , in which y is the desired cross-wind sampler dis-
                  m                                          r



tance and y  is the maximum cross-wind distance determined from Figure




3-3.  In this case, y  for x = 250 meters and category B is 68 meters,



and y = 62 meters.




     Array [1], then, would consist of three ground level samplers lo-




cated 250 meters downwind of the flotation tanks with the central sam-




pler on the wind axis and two samplers 62 meters away, one in each of




the two cross-wind directions.   This array will provide measurement




of at least two particle concentrations within the cloud for use in




the back calculation of the source strength at the flotation tanks.




     Similar computations may be made for each of the other arrays,




with the addition of a vertical spacing determination using Figures




3-5 and 3-6 in the same manner as Figures 3-3 and 3-4 for the deter-



mination of cross-wind spacing.
                                 -70-

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-089a
2.
4. TITLE AND SUBTITLE Technical Manual for Meas
Fugitive Emissions: Upwind/Downwind Sam
for Industrial Emissions
7. AUTHOR(S)
Henry J. Kolmsberg
3. RECIPIENT'S ACCESSION-NO.
urement of •• "EPO.R,TDATE
Dlinc Method Apri1 1976

r~ 0 6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION. REPORT NO.
9. PERFORMING OR3ANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
TRC--The Research Corporation of New England 1AB015; ROAP 21AUY-095
125 Silas Deane Highway
Wethersfield, Connecticut 06109
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Labora
Research Triangle Park, NC 27711
15. SUPPLEMENTARY NOTES Proiert
Ext 2557.
11. CONTRACT/GRANT NO.
68-02-2110
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; V75-1/76
14. SPONSORING AGENCY CODE
lory
EPA-ORD
officer for this report is R. M. Statnick, Mail Drop 62,
i6. ABSTRACT ^^Q manuai provides Si guide for the implementation of the Upwind/Down-
wind Sampling Strategy in the measurement of fugitive emissions. Criteria for
the selection of the most applicable measurement method and discussions of general
information gathering and planning activities are presented. Upwind/downwind
sampling strategies and equipment are described. The design of the sampling system,
sampling techniques, and data reduction procedures are discussed. Manpower
requirements and time estimates for typical applications of the method are
presented for programs designed for overall and specific emissions measurements.
The application of the outlined procedures to the measurement of fugitive emissions
from a Portland cement manufacturing plant is presented as an appendix.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Industrial Engineering
Measurement
Sampling
Portland Cements
I8. DISTRIBUTION STATEMENT
Unlimited
b.lDENTIF!ERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Fugitive Emissions
Upwind/Downwind Sam-
pling
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13B
13H
14B
11B
21. NO. O= =AGES
75
22. PRICE
EPA Form 2220-t (9-73)
                                                     -71-

-------