EPA-600/2-76-284
December 1976
Protection T^ctmology Series
            DEVELOPMENT  OF  PROCEDURES FOR THE
              MEASUREMENT OF FUGITIVE  EMISSIONS
                                   Industrial Environmental Research Laboratory
                                        Office of Research and Development
                                       U.S. Environmental Protection Agency
                                 Research Tri"r?$8 ParK, North Carolina  27711

-------
               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-600/2-76-284

                                   December 1976
  DEVELOPMENT OF PROCEDURES

     FOR THE MEASUREMENT OF

         FUGITIVE  EMISSIONS
                     by

 P.W. Kalika, R.E. Kenson, and P. T.  Bartlett

 TRC-The Research Corporation of New England
           125 Silas Deane Highway
       Wethersfield, Connecticut  06109
           Contract No. 68-02-1815
            ROAP No.  21AUZ-004 '
         Program Element No. 1AB015
   EPA Project Officer: Robert 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

-------
                                FOREWORD






      This document constitutes  the final report for the work accomplished




 on EPA Contract No.  68-02-1815, Methods of Measurement of Fugitive




 Emissions.




      The Environmental Technology Division of TRC - The Research Corp-




 oration of New  England was responsible for the work performed in this




 program.  The work was conducted under the technical and administrative




 direction of Dr. R. M. Statnick of the Industrial Environmental Research




 Laboratory, Process Measurements Branch of the Office of Research and




 Development of  the Environmental Protection Agency, which is located at




 Research  Triangle Park, North Carolina.  Mr. John E. Yocom, Chief Engi-




 neer  of TRC, was Program Director and Mr.  Peter W. Kalika, Engineering




 Manager, was project manager.  Dr. Robert E. Kenson and Mr. Paul T. Bart-




 lett  served as  task managers.  Major technical contributions were pro-




vided by Mr.  Don L. Shearer,  Mr. William A. Marrone, Mrs. Roberta Huston,




Mr. Kevin C.  Tower, Miss Nicola F. Grappone, and Miss Jo Anne Marchese.
                                  ii

-------
                           TABLE OF CONTENTS


SECTION                                                           PAGE

1.0              SUMMARY	1

2.0              INTRODUCTION  	   4
   2.1             The Nature of the Fugitive Emission	4
                    Problem
   2.2              Measurement Techniques 	   6
   2.3              Scope of Work Undertaken	7

3.0              CONCLUSIONS 	   9
   3.1             Industrial Sources  	   9
   3.2             Sampling Strategies 	   9
   3.3             New Methods	9
   3.4             Cost-Effectiveness  	  12
   3.5             Quasi-Stack Strategy  	  12
   3.6             Alternative Strategy  	  13
   3.7             Method Limitations  	  13

4.0              RECOMMENDATIONS	14
   4.1             Industry Studies  	  14
   4.2             Strategy Tests	14
   4.3             New Methods	15
   4.4             Documentation	15
   4.5             Package Development 	  15
   4.6             Impact Determination  	  15
   4.7             Refinement of Quasi-Stack 	  16

5.0              DISCUSSION	17
   5.1             Task I - Identify the Sources of	17
                    Fugitive Emissions
      5.1.1          Literature Search for Sources 	  17
      5.1.2          Trade Association/Industry/Govern-  	  18
                      ment Agency Contacts
      5.1.3          Matrix Charts for Identification of   ....  21
                      Selected Industries
      5.1.4          Fugitive Emission Sampling Techniques ....  23
                      and Strategies
   5.2             Task II - Evaluating Various Sampling	27
                    Strategies
      5.2.1          Evaluating 18 Selected Industries 	  27
      5.2.2          Suitability of Sampling Strategies  	  28
      5.2.3          Evaluate Sampling/Analysis Techniques ....  28
      5.2.4          Selection of Sampling/Industry Com-   ....  32
                      binations for Further Evaluation
      5.2.5          Cost-Effectiveness of Sampling  	  34
                      Strategies
   5.3             Task III - Prepare a Technical Manual for the .  36
                    Measurement of Fugitive Emissions
                                   iii

-------
   5.4              Task IV - Field Test of a Fugitive Emis- ... 36
                     sion Technical Manual
      5.4.1           Summary of Procedures and Analysis   .... 40
           5.4.1.1      Description of the Mold Pouring    .... 40
                         Hood Arrangement
           5.4.1.2      Description of Sampling Port Loca- .... 44
                         tions
           5.4.1.3      Field-operated Instrument and Chem-  ... 46
                         ical Laboratory
           5.4.1.4      EPA Method 5 Particulate and Conden- ... 48
                         sible Train
           5.4.1.5      G.I. Cascade Impactor and Condensi-  ... 51
                         ble Train for Particulates
           5.4.1.6      IKOR Model 206 Continuous Particu-   ... 54
                         late Monitor
           5.4.1.7      Beckman Model 109 Flame lonization   ... 56
                         Detector for Total Hydrocarbons
           5.4.1.8      Intertech Model Infra-2 Non-disper-  ... 56
                         sive Infrared Analyzer for Carbon
                         Monoxide
           5.4.1.9      High-volume Filter Sampler for Am- .... 58
                         bient Particulates
      5.4.2           Analytical Procedures  	 60
           5.4.2.1      Sample Recovery and Preparation in .... 60
                         the Field
           5.4.2.2      Analytical Laboratory Procedures 	 62
      5.4.3           Evaluation of Particulate Concentra- .... 65
                       tion Data
           5.4.3.1      Background Tests 	 70
           5.4.3.2      EPA Method 5 Test Comparison	.70
           5.4.3.3      Cascade Impactor Test Comparison 	 74
           5.4.3.4      Continuous Monitor Test Comparison .... 76
           5.4.3.5      Test Setup Limitations 	 80
           5.4.3.6      Critique of Sampling Methods 	 80
           5.4.3.7      Presentation of Particle Sizing Data ... 81
           5.4.3.8      Particulate Composition  	 84
      5.4.4           Evaluation of Hydrocarbon Concentra- .... 85
                       tion Data
      5.4.5           Evaluation of Carbon Monoxide Concen-  ... 89
                       tration Data
   5.5              Task V - Issue a Technical Manual	91
   5.6              Units of Measure	93

APPENDIX

A                 LITERATURE REFERENCES TO FUGITIVE EMISSIONS

B                 SUMMARY OF TRADE ASSOCIATION/INDUSTRY/
                    POLLUTION CONTROL AGENCY CONTACTS ON
                    FUGITIVE EMISSIONS

C                 MATRIX CHARTS USED TO IDENTIFY AND CHARACTER-
                    IZE FUGITIVE EMISSION SOURCES
                                   iv

-------
D                 MEASUREMENT OF FUGITIVE EMISSIONS - LITERATURE
                    REFERENCES TO METHODS

E                 SAMPLE CRITIQUE OF SAMPLING METHODS

F                 EPA METHOD 5 DATA SHEET

G                 CASCADE IMPACTOR DATA SHEET

H                 IKOR CONTINUOUS MONITOR DATA SHEET

I                 HI-VOL FILTER TEST DATA SHEET

J                 LABORATORY WORKSHEET FOR FIELD SAMPLES

K                 CONVERTING UNITS OF MEASURE

-------
                            LIST OF FIGURES


FIGURE                                                        PAGE

5-1      Simplified Flow Sheet for Gray Iron Foundry	31

5-2a     Cost-effectiveness of Four Fugitive Emission   ....  35
           Sources - Sampling Strategies Applied to Different
           Sources

5-2b     Cost-effectiveness of Four Fugitive Emission   ....  35
           Sources - Sampling Strategies Applied to the Same
           Source

5-3      Illustration of Test Set-up for Measuring Fugitive  .  39
           Emissions from Mold Pouring in a Gray Iron Foundry

5-4      Schematic Diagram of Sample Port Locations  	  45

5-5      EPA Method 5 Particulate Sampling Train Including . .  49
         t  Condensibles

5-6      G.I. Cascade Impactor Assembly  	  52

5-7      IKOR Continuous Particulate Monitoring System  ....  55

5-8      Schematic Diagram of Laboratory Trailer Setup  ....  57
           Showing Hydrocarbon and Carbon Monoxide Analyzers

5-9      High Volume Sampling Station for Ambient Particu- . .  59
           lates

5-10     Test 17:  Continuous Monitor Trace	77

5-11     Test 21:  Continuous Monitor Trace	78

5-12     Layout of Molds	86

5-13     Plot of Mean Half-Hour Hydrocarbon Concentrations . .  88
           for Period April 7 - 21, 1975

5-14     Plot of Mean Half-Hour Carbon Monoxide Concentra- . .  92
           for Period April 7-21, 1975
                                 vi

-------
                            LIST OF TABLES
TABLE                                                          PAGE

3-1      Classification of Industries for Possible Future ... 10
           Studies of Fugitive Emissions

3-2      Evaluation of Candidiate Strategies for Measuring  .  . 11
           Fugitive Emissions

5-1      Summary of TRC Studies Relevant to Fugitive Emis-  .  . 19
           sions Measurement

5-2      Matrix Chart of Fugitive Emission Sources in Copper  . 20
           Smelting

5-3      Classification of Industries for Possible Future ... 22
           Studies of Fugitive Emissions

5-4      Feasible Sampling Strategies for Industries/ 	 25
           Processes Selected for Possible Fugitive
           Emissions Studies

5-5      Method of Rating Hazard Potential of Fugitive Emis-  . 29
           sions of 18 Selected Industries

5-6      Matrix of Iron and Steel Foundries Fugitive Emission . 30
           Source/Sampling Strategy Combinations

5-7      Matrix of Final Selections of Fugitive Emissions ... 33
           Sampling Strategy/Industry Source Combinations

5-8      Summary of Tests - Mold Pouring	  . 41

5-9      Foundry Activity During Testing  	  .  . 42

5-10     Particulate Concentration Data 	 67

5-11     Summary of Statistical Analysis-Total of All Iso-  .  . 68
           kinetic Tests for Each Method

5-12     Summary of Statistical Analysis-Total of 10 Tests:  .  . 69
           All Methods Iso-kinetic and All Methods were Run
           Concurrently

5-13a    Differences in Test Method Performances-Total  .... 71
           Particulate Concentration, GR/DSCF

5-13b    Differences in Test Method Performances-Total  .... 72
           Condensibles Concentration, GR/DSCF
                                vii

-------
5-13c    Differences In Test Method Performances; Partlculate . 73
           on Filter(s) Concentration, GR/DSCF

5-14     Particle Size Distribution (% by Count):  EPA  .... 82
           Method 5

5-15     Particle Size Distribution (% by Count):  Mean .... 83
           Values and Standard Deviation

5-16     Particle Size Distribution (% by Count):  Statis-  . . 83
           tical Differences Between Process Emissions
           and Background Condition

5-17     Total Hydrocarbon Concentration  	 87

5-18     Carbon Monoxide Concentration Data 	 90
                                viii

-------
1.0  SUMMARY

     This document is a final report  of EPA Contract No. 68-021815

"Methods of Measurement of Fugitive Emissions."  TRC - The Research

Corporation of New England conducted the studies undertaken in this

program under the supervision and sponsorship of the Industrial Environ-

mental Research Laboratory, Process Measurements Branch of the Office of

Research and Development of the Environmental Protection Agency, which

is located at Research Triangle Park, North Carolina.

     It has been recognized that fugitive (non-stack) emissions from

industrial processes are a potentially significant portion of the total

air emissions in many industries.  The purpose of the program undertaken

was to develop basic procedures which could be used to sample and analyze

these fugitive emissions.  The TRC studies were conducted over a 12-month

period between July, 1974, and June, 1975, and divided into the following

tasks:
     1.  Task I - Identify the Sources of Fugitive Emissions.  This
         task consisted of a literature search, a review of TRC exper-
         ience in fugitive emission measurements, and contacts with
         industries, regulatory agencies, and industry trade associa-
         tions.  The results are presented in the form of a matrix
         showing industries, processes, fugitive emission sources, and
         type of pollutants emitted.

     2.  Task II - Evaluate Various Sampling Strategies.  In this task,
         TRC considered the application of three broad measurement ap-
         proaches to the sources listed in the matrix from Task I.  Cost
         and accuracy estimates were made for each of the methods.  These
         approaches are:

         A.  Quasi-stack.  Here the fugitive emission is temporarily
             hooded or enclosed and a temporary duct or stack and fans
             are installed to permit sampling by means of standard stack
             sampling methods.

         B.  Roof Monitor.  The term "roof monitor" encompasses a var-
             iety of similar situations in which fugitive emissions
                                   -1-

-------
            leave a building or enclosure through roof monitors, roof
            ventilators, and windows and doors.

        C.   Upwind-downwind.  A carefully designed upwind-downwind
            sampling network is used under selected meteorological
            conditions to show the contribution of fugitive sources
            to ambient air quality.  Three dimensional arrays may be
            needed to permit measurement of pollutant flux past a
            sampling grid.  Mathematical models and tracer studies
            may be required to describe complex situations.

    3.   Task III - Prepare a Technical Manual for the Measurement of
        Fugitive Emissions.  Based on Task I and Task II and discussions
        with EPA, the quasi-stack method was selected for preparation of a
        test procedures document.  A draft was prepared which used as an
        example case the measurement of fugitive emissions from mold
        pouring at a gray iron foundry.

    4.   Task IV - Field Test of a Fugitive Emission Technical Manual.
        In  this task, TRC conducted a field test of the quasi-stack
        method for mold pouring at a gray iron foundry in southern
        Connecticut.

    5.   Task V - Issuance of a Technical Manual.  At the completion
        of  the field test program and after all data had been analyzed,
        the Field Test Procedures Document was thoroughly evaluated by
        TRC and EPA and modified as appropriate.  The approved document
        was issued as Volume 2 of the final report.
   The basic conclusions developed by TRC in this study were:
7

   1.  Eighteen industries/processes were believed to have potentially
       significant fugitive emissions.  These were selected from all
       the industries/processes studied based on the potential of their
       fugitive emissions to be hazardous and to have high emission
       rates.

   2.  The industry or process step/sampling strategy combinations
       best able to define and measure the fugitive emissions of these
       18 industries/processes were identified and studied.  This re-
       sulted  in the specification of three basic sampling strategies
       (quasi-stack,  roof monitor, and upwind-downwind) which could
       be used  to  define  and measure most of the potentially signifi-
       cant fugitive emissions identified in this program.

  3.   For  the  field test portion of this program (Task IV), a quasi-
       stack sampling  strategy was identified as capable of measuring
       a  significant  number  of fugitive emission sources and as being
      amenable  for  implementation within reasonable time and cost
                                 -2-

-------
         constraints.  Mold pouring in a gray iron foundry in Southern
         Connecticut was chosen as the fugitive emission source for ap-
         plication of the quasi-stack strategy.

     4.  A quasi-stack sampling strategy was found to be a valid method
         for studying mold pouring fugitive emissions provided that con-
         tinuous analysis methods are used and that background pollutant
         concentrations are separable from those from the operation
         studied.
     As a result of this study, TRC makes the following basic recommen-

dations:
     1.  EPA should consider further studies of the fugitive emissions
         of six specific industries.

     2.  EPA should fund the study of roof monitor and upwind-downwind
         sampling strategies.

     3.  EPA should develop specific instrument packages for fugitive
         emissions.
     This report consists of two volumes.  One, which is titled, "Methods

of Measurement of Fugitive Emissions — Industrial Fugitive Emissions

Sources and Sampling Strategies," documents Tasks I, II, and IV of this

study.  A second, which is titled, "Technical Manual for the Measurement

of Fugitive Emissions — Quasi-Stack Sampling Method for Industrial Fugitive

Emissions," documents Tasks III and V.
                                   -3-

-------
 2.0  INTRODUCTION

      The program was  undertaken over  a  12-month period between  July,

 1974, and June,  1975,  for  the  purpose of development of sampling  proce-

 dures to determine the emission rate  and chemical and physical  charac-

 teristics of fugitive  emissions.

      It has been recognized  that  as the emissions from industrial pro-

 cesses through stacks  or vents are controlled, by 90 percent, for exam-

 ple, overall process  emissions may not be reduced 90 percent.   Non-stack

 or fugitive emissions, which can be produced by almost every industrial

 process, could become  significant portions of the total process emis-

 sions.  For the purpose of this program, fugitive emissions were  defined

 as those emissions of  air  pollutants  not directed through ducts or stacks

 and not amenable to measurement by established source sampling  methods.




 2.1  The Nature of the Fugitive Emission Problem

      Almost every industrial operation  capable of emitting air  pollu-

 tants exhibits some fugitive emissions.  These emissions vary with the

 basic process  and the  manner in which a given process is carried  out,

 e.g., a new cement plant with  enclosed  raw materials handling and good

 housekeeping will have far less fugitive dust emissions than an old,

 poorly  operated  plant  with open storage.

      The mechanism for release  also varies widely.  Fugitive emissions

of gaseous pollutants   result from:
     1.   Leaks in piping, packing glands, and other components  of  pro-
         cess equipment.

     2.   Exposure of liquids containing a component of appreciable va-
         por  pressure.
                                   -4-

-------
     3.  Exposure of unstable solids, one of whose breakdown products
         is a gaseous pollutant.

     4.  Generalized emissions from a building roof monitor or roof
         fan where many smaller fugitive emissions accumulate within
         and are vented from the building.

     5.  Operations carried out outdoors or partially outdoors (e.g.,
         coke charging and pushing).
     Gaseous emissions, once they have left the confines of a process

and when they are not exhausted through a duct or stack where emission

rates can be measured, are fugitive emissions.  On the other hand, par-

ticulate fugitive emissions, because of gravity and inertial effects,

may or may not become fugitive emissions once they have left a process.

     Fugitive particulate emissions behave differently from gaseous fu-

gitive emissions.  This difference is principally a function of parti-

cle size.  Particles below about five to 10 microns behave similarly

to gases because of their slow settling rate.  Above this range of par-

ticle size, the settling rate will influence how far a particle will

travel and how easily it will become reentrained.  Thus, measurements

of fugitive particulate emissions must include particle size determina-

tions.  In addition, meteorological conditions and the geometry of the

release point produce complications in describing the behavior of par-

ticulate as compared to gaseous fugitive emissions.

     Typical mechanisms for the release of particulate fugitive emis-

sions include:
     1.  Operations carried on outdoors (e.g. , material handling, coke
         charging and pushing).

     2.  Participates formed from the reaction of gaseous fugitive emis-
         sions with the atmosphere, moisture, or other gases (e.g.,
         ?2®5 from phosphorous emissions, and NH^Cl from adjacent
         NH3 and Cl£ sources).
                                   -5-

-------
      3.  Leaks from confined process streams carrying particulates or
          condensible vapors (e.g., coke oven door leakages).

      4.  Outdoor storage of solid materials (e.g., raw material piles,
          dried sludge beds).

      5.  Accumulations of dust from inadequate housekeeping which are
          subject to reentrainment.

      6.  Traffic on dusty plant roads.
      Another factor which complicates the assessment of fugitive parti-

  culate matter from an industrial process is the ubiquity of airborne

  particulate matter as a background component of generalized air qual-

  ity.  Unless the particulate fugitive emission is unique chemically

  and/or physically and has not already influenced the characteristics

  of  surrounding background particulate matter, it may be difficult to

  assess the impact of the fugitive emission in question.



  2.2  Measurement Techniques

      The techniques presently available for measurement of fugitive

  emissions are:


      1.   Material Balances
      2.   Quasi-Stack Sampling
      3.   Roof Monitor Sampling
      4.   Upwind-downwind Sampling


      Material balance techniques involve no actual tests, but simply

 determine what enters and exits  the process.  The material unaccounted

 for  by product or measured  emissions is assumed to be the fugitive

emission.

     The quasi-stack method  involves the temporary enclosing, hooding,

and ducting of the emission  source  such that well-established source

sampling and analysis techniques  can be applied.
                                   -6-

-------
     The roof monitor method refers to the specialized sampling of fugi-

tive emissions from well-defined building openings such as roof monitors,

skylights, ceiling vents, windows, etc.

     The upwind-downwind method is generally applied to large outdoor

sources and involves carefully designed arrays of samplers upwind and

downwind of the suspected fugitive emission source.  These samplers can

be located both at ground level and at elevated positions.  The method

involves meteorological measurements which must be related to the sam-

ples taken.  Diffusion modeling and tracer techniques may also be in-

volved to assist in identifying individual sources of fugitive emissions.



2.3  Scope of Work Undertaken

     The program to develop procedures for the measurement of fugitive

emissions was divided into the following tasks:


     Task I   - Identify the sources of fugitive emissions (15 percent
                of total program effort)

     Task II  - Evaluate various sampling strategies (15 percent)

     Task III - Prepare field test procedures document (10 percent)

     Task IV  - Field test of fugitive emission field test procedures
                document (45 percent)

     Task V   - Issuance of final field test procedures document (15
                percent)


     This report (Volume I) summarizes the results of Tasks I, II, and

IV, as follows:
     Section 3.0 - Conclusions based upon the program undertaken
     Section 4.0 - Recommendations for further studies
     Section 5.0 - Discussion of the procedures and results of the work
                   performed
     Appendices  - Detailed descriptions of specific methods, refer-
                   ences, and detailed matrix charts.
                                   -7-

-------
     The studies of the "Methods of Measurement  of  Fugitive Emissions --




Quasi-Stack Sampling Method for Industrial  Fugitive Emissions"  (Tasks




III and V), comprise a separate volume  (Volume 2) of  this  report.

-------
3.0  CONCLUSIONS




     The study of industrial sources of fugitive emissions and fugitive




emission measurement methods, as well as the field test of a quasi-stack




sampling strategy, has led to the following conclusions:









3.1  Industrial Sources




     There were 18 industries/processes believed to have potentially




significant fugitive emissions.  These 18 (Table 3-1) were selected




from all those studied based on the potential for their fugitive emis-




sions to be hazardous and to have high emission rates.  Other criteria




were also used to rank these 18.









3.2  Sampling Strategies




     There are three basic sampling strategies suitable for measurement




of the fugitive emissions from most potentially significant sources in




all the industries studied.  These three, the quasi-stack, the roof mon-




itor/enclosure vent, and the upwind-downwind sampling strategies, are




described briefly in Table 3-2, including their limitations and general




accuracy ranges.









3.3  New Methods




     For some significant fugitive emission sources, present sampling




strategies may be inadequate.  An example was the copper smelting indus-




try, where openness of building structures limits present sampling strat-




egy application.  New developments in remote sensing were seen as po-




tentially useful tools in this application.
                                   -9-

-------
                   TABLE 3-1

       CLASSIFICATION OF INDUSTRIES  FOR
 POSSIBLE FUTURE STUDIES OF FUGITIVE EMISSIONS
   Category

 Metallurgical
 Energy/Fuels
Chemical Products
Rock Products
Agricultural Opera-
 tions
   Industry/Process

Primary Aluminum
Primary Copper
Electric Furnace Steel
Iron & Steel Foundries
Coke Making

Coal Mining
Coal Gasification
COG (Char-Oil-Gas)
Shale Oil
Petroleum Refining
Oil Production

Plastics
Tire & Rubber

Phosphate Fertilizer
Lime
Sand & Gravel
Asphalt Batching

Cattle Feedlots
Soil Tilling
Grain Harvesting
                   -10-

-------
                                                              TABLE 3-2

                                 EVALUATION OF CANDIDATE STRATEGIES FOR MEASURING FUGITIVE EMISSIONS
    Candidate .Strategy.

1.  Roof-Monitor/Enclosure
    Vent Strategy (Use of
    Available Enclosures
    such as Buildings
     Quasi-Stack Sampling
     Strategy (Construc-
     tion of special en-
     closures with off
     take duct for
     sampling.)
 3a
     Upwind-Downwind
     Strategy (Simul
     taneous Upwind-
     Downwind sampling.)
 3b
     Upwind-Downwind
     Strategy (Upwind™
     Downwind sampling
     together with dif-
     fusion modeling and
     tracer studies.)
             Requirements

    Openings  must be regular and
    amenable  to sampling for
    concentration and air flow.
    Impractical if too many
    openings.

    Ventilation (natural or
    power)  must be such that a
    material  blance can be
    established.
3.  If natural ventilation             3.
    exhausts the fugitive
    emissions, enough data must
    be taken to establish emissions
    under a variety of
    meteorological conditions.

1.  Enclosure must capture essentially 1.
    100% of fugitive emissions.
2.  Enclosure must not affect emission 2.
    rate or physical and chemical
    characteristics of emissions.

3.  The scale of the fugitive emissions 3.
    may be too great for construction
    of a reasonably sized enclosure.

1.  Fugitive emissions must be unique  1.
    and identifiable in relation to,
    or not influenced by, other
    plant emissions.

2.  Sampling array must encompass the  2.
    fugitive emissions cloud and
    provide enough detail to establish
    concentrations across the entire
    cloud.

3.  Wind flux measurements are needed
    at several points both laterally
    and longitudinally in the cloud
    path.

1.  Model must be applicable to        i.
    fugitive emission points.
       Applicability

Buildings housing multiple
units which have fugitive
emissions - Cl  caustic cells,
aluminum pots, electric
furnaces, etc.

Buildings containing materials
storage which emits pollutants
while stored, e.g., synthetic
fertilizer, odors from render-
ing.

Buildings containing a variety
of materials handling
and processing operations.
Can be applied to a wide range
of relatively small operations
such as welding, solvent use.
Can be used for measuring
leaks from piping, packing
glands, etc.

This method where applicable
can provide quite accurate
results.

Must be applied when two
previous strategies cannot be
used.
                                                                        Applicable where fugitive
                                                                        emissions of the same type
                                                                        are  emitted from several areas
                                                                        of a plant.
- 25%-50%
                                                                                                             - 25%
Applicable where the preceeding
three  strategies cannot be
used alone.
                                                                                                           -  50%
                                                                                                                           - 200%
                              2.  Detailed micro-meteorological
                                 data are required.
                             3.  Field measurements must be carried
                                 out under specified meteorological
                                 conditions.
                                            Applicable where various  types
                                            of fugitive emissions  are
                                            present and where background
                                            pollutants influence
                                            meas ur emen t s
                                                          -11-

-------
3.4  Cost-Effectiveness

     The cost-effectiveness of each  sampling  strategy when  applied  to  a

specific source varied considerably.   For  different  sources,  different

optimum sampling strategies could be found.   It was  found that  cost-

benefit tradeoffs could be examined  for  each  of the  three fugitive  emis-

sion sampling strategies by developing error  limits  vs.  cost  curves for

selected fugitive emission sources.



3.5  Quasi-Stack Strategy

     The quasi-stack sampling strategy was identified  as being  widely

applicable to many significant industrial  fugitive emission sources,

and as an excellent candidate for field  tests.  It has  generally accept-

able accuracy limits and in many cases can be made cost-effective.   It

was found to be a valid method for the study  of mold pouring  emissions

provided that continuous analysis methods  are used and  certain  other

criteria are met.   These other criteria  include:


     1.   Background pollutant concentrations  are  separable  from the
         operation studied.

     2.   Physical  location of sampling allows proper positioning of re-
         quired  equipment and isolation  of the  test  area from other
        plant operations.

    3.  Sampling  times  where significant  fugitive emissions  will occur
        are  compatible  with instrument  response  times.

    4.   Sample  concentrations are within  the analysis capability of
        the  instruments.

    5.   The quasi-stack configuration can be designed to capture all
        of  the  fugitive  emissions to be studied  without serving as a
        general ventilation hood for all  in-plant pollutants,  or other-
        wise biasing pollutant  concentrations  or characteristics.
                                  -12-

-------
3.6  Alternative Strategy

     As an alternative to the quasi-stack sampling strategy in an iron

foundry or similar situation, a roof monitor sampling approach would be

applicable under certain conditions, which include:
     1.  High background pollutant concentrations prevent separation
         of any single pouring or other operation.

     2.  Physical limitations prevent setup or isolation of a quasi-
         stack sampling area.

     3.  Response times of instruments prevent capture or display of
         single event fugitive emissions.

     4.  Sample concentrations from single events are below the instru-
         ment threshold.

     5.  The quasi-stack configuration cannot be used for emission cap-
         ture and the building structure is well enclosed and has well
         defined vents.
3.7  Method Limitations

     Standard particulate measurement techniques (time-averaging meth-

ods such as filter collection) may be unsuitable for fugitive emission

studies, especially where emissions are transient in nature and release

is over a short time duration.

-------
 4.0  RECOMMENDATIONS

      TRC,  on the  basis  of  its  studies of fugitive emissions under EPA

 Contract 68-02-1815,  makes  the following recommendations:



 4.1  Industry Studies

      EPA should consider further studies of the fugitive emissions from

 some of the industries  listed  in Table 3-1.  In particular, the indus-

 tries most in need  of study would be:
      1.   Primary  copper
      2.   Coke making
      3.   Electric furnace steel
      4.   Petroleum refining
      5.   Coal gasification
      6.   Iron and steel  foundries (continue and expand work begun in
          this study)
      This  listing is based upon both the mass rate and the hazardous

 nature  of  their  fugitive emissions.  In particular, the fugitive emis-

 sion  potential of a significant new energy process, coal gasification,

 should  be  examined in detail.  The impact of this industry on air  qual-

 ity in  the future could be significant, and an important part of this

 impact  could be  the fugitive emissions.



4.2  Strategy Tests

     EPA should fund the study of the roof monitor and upwind-downwind

sampling strategies for fugitive emission measurement.  A comparison

of all three major strategies could then be made after field  test  of

these fugitive emission sampling methods.
                                   -14-

-------
 4.3  New Methods




      EPA should consider funding of research in the application of re-




 mote sensors for the study of fugitive emission sources.   For some po-




 tentially significant sources, the three major strategies  are inade-




 quate.   This research might be conducted in parallel with  field tests




 of either the roof-monitor or upwind-downwind strategies to  attempt



 correlation.









 4.4  Documentation




      EPA should develop a fugitive emissions measurement handbook.   This




 would outline sampling strategies available to the  investigator,  includ-




 ing a basic methodology and procedures.   By use of  a multi-volume  for-




 mat,  annual updates  would be possible  as technology advances.








 4.5  Package Development




      EPA should develop instrumentation  packages designed  to measure




 fugitive emissions.   Although they may not  be  adopted as standard EPA




 methods, they could  be presented  as EPA-recommended methods.  This would




 help  standardize  fugitive emission measurements and aid in the  develop-




 ment  of  correlations between investigators,  including better estimates




 of  relative  and absolute  errors.   These may  include  remote sensing methods.








4.6   Impact  Determination




     EPA should consider  the  special needs of  fugitive emission impact




estimation procedures.  Since present  stack  plume models may need modification
                                  -15-

-------
 to be  used for the simulation of the impact on ambient air of fugitive




 emissions, a comprehensive study program may be required in this area.




 TRC  studies of the impact of fugitive sources on ambient air quality




 has  shown that estimation of the source strength, source height, and




 diffusion rate of the fugitive pollutants from open sources, open vent/




 roof monitor sources, or even short stack sources can be considerably




 in error unless examined in some detail and appropriate modifications




 made in the diffusion models.  This would aid EPA efforts in many areas




 of study, where fugitive emission impacts up to now have not been con-




 sidered.








 4.7  Refinement of Quasi-Stack




     EPA should sponsor research to refine quasi-stack fugitive emis-




 sion measurement methods.  The limitations found by TRC in its field




 studies, especially with regard to time intervals of fugitive emis-




 sions and their variability and isolability, point out areas for fur-




 ther development.   Extensive testing of various continuous monitors,




especially for particulate matter mass rate and particle size, may be



required.
                                  -16-

-------
5.0  DISCUSSION




     The overall objective of the program was to develop methods for the




measurement of fugitive emissions of air pollutants from several indus-




trial processes.  Fugitive emissions, for the purposes of this study,-




were defined as those emissions of air pollutants which are not directed




through ducts or stacks and which are not amenable to measurement by es-




tablished source sampling methods.  The methods developed under this pro-




gram were to be adequate for quantifying fugitive emissions within rea-




sonable and calculable error limits and capable of use for assess-




ing measures designed to control fugitive emissions.




     A five task program to accomplish this was conducted as follows:









5.1  Task I - Identify the Sources of Fugitive Emissions




     In order to identify possible industrial sources of fugitive emis-




sions, a four part program was conducted with the following results:









     5.1.1  Literature Search for Sources




     A bibliography of references specifically directed to fugitive emis-




sions was developed and is presented in Appendix A.  The key documents




used to develop the fugitive emission data for industries believed to




have significant fugitive emission potential are included.  A substan-




tial number of the key documents were EPA research project reports




wherein overall process emissions and in some cases fugitive emissions




were quantified.  Only recently have any significant number of fugitive




emission studies been reported in the open literature.
                                  -17-

-------
     Another  source  of  fugitive emission  information was TRC in-house in-




formation developed  during studies  for  industrial and governmental clients.




Table 5-1 summarizes these.  Because  of their  confidential nature (with




a few exceptions), it was not possible  to cite these in the bibliography.




They are a substantial body of information,  especially in regard to fu-




gitive emission measurement methods.








     5.1.2  Trade Association/Industry/Government Agency Contacts




     Six trade associations and two companies  were contacted to obtain




information concerning fugitive emissions in industries of potential in-




terest.  Four of  these contacts have supplied  data which will be useful




in  identifying and quantifying fugitive emissions.  Six air pollution




control agencies were contacted but no  "hard"  data were obtained from




them.  Appendix B summarizes all the contacts  which were made.  The con-




tacts made did have an appreciation for the fugitive emission potential




of  the industrial processes with which  they were familiar.  However, in




common with many others, they had for the most part not conducted studies of




fugitive emissions because of the priorities which have had to be given




to  the characterization and control of  stack emissions.  One exception




to  this was the contact with Dr. Fred Templeton, Assistant to the Presi-




dent of the Metal Mining Division of Kennecott Copper.  Dr. Templeton sup-




plied EPA and TRC with fugitive emission  data  developed by Dr. John




Heaney of the Utah Copper Division of Kennecott Copper.  These data, in




conjunction with those available from other sources, were used to develop




a matrix chart which describes most of  the fugitive emission factors




for the copper smelting industry.  Table  5-2 summarizes these emission
                                   -18-

-------
                                      TABLE 5-1

                         SUMMARY OF TRC STUDIES RELEVANT TO
                           FUGITIVE EMISSIONS MEASUREMENT
     Client Category

 1.   Iron "and Steel Works


 2.   Inorganic Chemical
 3.   Steel Company
     Coke Works

 4.   Municipal Health Department
     (Report on Proposed Iron
      & Steel Works)
 5.  Inorganic Chemical Company
 6.  ASHRAE
     (Research Project TRP-93)

 7.  USPHS, DHEW, NAPCA
     Contract CPA-22-69-14
 8.  Assembly Plant
     Electric Components

 9.  EPA
     Contract CPE-69-5

10.  Cement Company
11.  Cement Company
12.  Sinter Plant in
     Steel Mill
13.  Iron Foundry


14.  Rubber Manufacturing


15.  Turbine Manufacturing

16.  Rubber Manufacturing


17.  Parking Garage


18.  Gravel Pit


19.  Coke Works
20.  EPA
     Contract 68-02-0047
     Fugitive Emission Studies Done

Welding operations gas and particulate
emissions

Emissions of dust from ground, materials
handling and storage piles in plant.

Smoke emissions from leaking doors and coke
pushing and charging

Contribution of dust emissions from proposed
plant on local air quality.  Similar plant
emission data and diffusion model used
for prediction

Dust emissions from uncontrolled vents and
from open windows of dry bleach plant

Use of air conditioning equipment to reduce
outdoor air pollutants drawn into buildings

Leakage of emissions inand out of structures
such as houses.  Measurement of fugitive
emissions from stoves, etc.

Solvent emissions from coating operations
where concentrations v ary.

Study of emissions from disposal of solid
wastes

Upwind and downwind particulate emission
measurement to determine plant contribution

Upwind and downwind particulate emission
measurement plus diffusion modeling

Analysis of local air quality and contribu-
tion of sinter plant emissions to this.
Used up and downwind measurements.

Contribution to particulate emissions of
unpaved parking lot

Uncontrolled curing press emissions
measured

Fuel and Solvent transfer system emissions

Emissions inventory of uncontrolled sources
in tire plant

Dispersion modeling to determine CO con-
centrations at peak times

Air quality measurements of particulates
from fugitive  and other sources

Air quality measurements in plant  including
those near fugitive sources

Evaluation of  particulate  control  strategies
which include  fugitive emissions  from
demolition
                                    -19-

-------
                                     TABLE  5-2

                 MATRIX CHART OF  FUGITIVE  EMISSION  SOURCES
                               IN COPPER  SMELTING
Process Step
1. Ore Concentrate
Storage and
Handling
2. Limestone Storage
and Handling
3. Slag Handling
4. Dust Collection
and Transfer
5. Roaster toading
and Operation
6. Calcine Transfer
7. Revcrberatory
Furnace Loading
and Operations
8. Matte Transfer
9. Converter
Loading and
Blowing
10. Blister Copper
Transfer
Particulate Fugitive Lmissions
Hnss Kate
1



10 lbs/ton(3)
Copper
1
i
1
1
2.0 Tons/
DayU)

1
*~5 ft
Dayu
,


T>

Composition
(Annroximate)
Cu 28%(3) Si02 117.
Te 24% Other 5%
S 32%
CaO ~ 60%
Cu 0.5%^ Si02 38%
Fe 40% S 1.5%
Unknown
Cu 5%(4) Pb 18%
Zn 16% As 60%
Cd 0.5X 70%<2um
Unknown
Cu 5%(A) Pb 18%
Zn 16% As 602
Cd 0.5% 70%<2um
Cu 42%^ Si02 1%
Fe 32% S 25%
Cu 1%(A) Pb 50%
Zn 8% As 37%
Cd 4% 10%<2vra

Gaseous Fugitive Kralssions
Mass Rate
Unknown
Unknown
0.4 Ibs.S^
/Ton Slag
Unknown
i ,
5.2 Tons/
Day S02 W
)
/8.3 lbs.S(2)
1 /Ton of
| Matte
* (1)
35-40 Tons/
Day S02
/"Olbs.S^
1 /Ton Cu
Compos i tion
^Approximate)
Unknown
Unknown
Unknown
Unknown
SOx-5-6*/^ ,
SOx-5-15%k ' ;
Unknown
SOx-5-6%^)
SOx-1-3%^ ' '
Unknown
SOx-l-5-4%^J
SOx-14-19%l'J'
Unknown
     Utah Copper Division Data Supplied by Kennecott Copper Corporation.
(2)

(3)


(4)



(5)
Noranda Mines Ltd.  Data Supplied by Kennecott  Copper Corporation.

"Systems Study for  Control of Emissions -  Primary Non-Ferrous Smelting Industry"
Arthur C.  HcKce and Company  NTIS Report No. PB 184 884.

Measurement of Sulfur Dioxide, Particulate and Trace Elements in Copper Smelter
Converter  and Ronster Rcverberatory Gas Streams, KPA Report No. 650/2-74-11 by
Couljol Sy^tcnis Laboratory.

Air Pollution Control Technology and  Costs in Nine Selected Areas, IGCI Report on
KPA Com rnrt  68-0?  0101
(6)   Emission includes  all  these operations.
                                       -20-

-------
factors.




     Mr. Tom Keller of Bethlehem Steel contributed information on roof




monitor emission measurement techniques.  Although not used as a samp-




ling strategy in the field test program, this information offered val-




uable guidelines for conducting such tests in the future.








     5.1.3  Matrix Charts for Identification of Selected Industries




     A list of possible industries and processes which might have fugi-




tive emissions was developed.  To identify where these occur, a matrix




chart was evolved which lists vertically the industry or process studied




and horizontally the operations which make up these industries or pro-




cesses.  These were provisions on the charts for symbols to characterize




the type of fugitive emission encountered.  Appendix C shows these charts.




     The completed matrix charts were used as a basis for selection of




industries which might involve significant fugitive emissions.  A list




of 18 such industries was developed and the matrix charts of these indus-




tries were consolidated.  These industries were reviewed in more depth




so that the list would be refined to identify the most serious candidate




industries for fugitive emissions testing.  Tentative sampling strategies




concepts were developed for these 18.  Table 5-3 identifies the indus-




tries and their major potential fugitive emission problems.




     In the absence of "hard" data, mass rates of fugitive emissions




were estimated not in terms of absolute numbers but in terms of their




potential fraction of total air emissions.  Where the potential frac-




tion was high (> 10 percent), the industry or process might be worthy




of further study.  Industries with potential for release of low TLV and/
                                   -21-

-------
                                TABLE 5-3
            CLASSIFICATION OF  INDUSTRIES FOR  POSSIBLE  FUTURE
                     STUDIES OF FUGITIVE EMISSIONS
Category
Metallurgical
Energy/Fuels
Chemical Products
Rock Products
Industry/Process
Primary Aluminum
Primary Copper
Electric Furnace Steel
Iron & Steel Foundries
Coke Making

Coal
Coal Gasification
COG  (Char-nil-Gas)

Shale Oil

Petroleum Refining

Oil Production
Plastics
Tire & Rubber
Phosphate Fertilizer
Lime
Sand & Gravel
Asphalt Batching
Agricultural Opera-
tions
Major Fugitive Emissions
Fume, Fluorides, PNA's
Fume, S02» Dust
Fume, Kish, CO, Odors
Fume, Odors
Hydrocarbon Fumes, Odors,
CO, PNA's
Dust
Hydrocarbon Fumes, CO, PNA's
Hydrocarbon Fumes & Vapors,
CO, PNA's
Dust, Hydrocarbon Fumes
& Vapors, CO, PNA's
Hydrocarbon Vapors, Odors,
PNA's
Hydrocarbon Vapors, Odors
Hydrocarbon Vapors, Odors
Hydrocarbon Vapors, Odors
Dust, Fluorides, S02
Dust
Dust
Dust, Odors, PNA's
Dust, Odors
                                    -22-

-------
or known carcinogenic materials were considered for further study.  Highly




reactive or toxic fugitive emissions also help to place an industry on




this list.




     The 18 industries selected for possible study could be reduced to




four categories  (see Table 5-8):  metallurgical, fuels/energy, chemical




products, and rock products.  Only one industry escaped classification




in Table 5-3:  agricultural operations.  It was important, however, that




this category be studied, since its fugitive dust problems had signifi-




cant impact on measurements of particulate emissions in other industries




such as rock products.  The state of Colorado, for example, cannot meet




particulate matter primary air quality standards because of agricultural




contributions.









     5.1.4  Fugitive Emission Sampling Techniques and Strategies




     A scanning  of the literature on sampling techniques showed that few




new techniques have been developed for measurement of fugitive-type emis-




sions.  Appendix D summarizes specific references found.  One advance




was the development of a procedure for sampling emissions from roof moni-




tors.  This procedure is more of a refinement of a technique previously




used rather than a new sampling technology.  Much the same can be said




about the other  measurements mentioned in the literature.




     Our new development indicated that lidar (light detection and rang-




ing/laser radar) might be applicable to measurement of fugitive dust emis-




sions.  It has two possible advantages:  readings are instantaneous and a




large number of  accurate mappings of fugitive dust cloud might be obtain-




able in the time needed for one such set of data from a standard particulate




sampling test array.  There are two possible disadvantages:   commercial
                                   -23-

-------
 models may not be readily available and we are not  aware of  any  present




 studies which seek to confirm direct  applicability  to  the  type or  problems




 encountered in fugitive dust  measurements  (most work has been done with




 lidar used in stack plume mapping).



      The 18 industries/processes of interest for possible  fugitive emis-




 sions studies were closely examined to see which sampling  strategy might




 be applied to each of the suspected fugitive emission  sources.   If more




 than one strategy or a combination of strategies might be  appropriate,




 this was noted.  An attempt was also  made  to rate the  strategy as  to  its




 applicability.




      The sampling strategy/industry or process combinations  considered




 to be feasible are given in Table  5-4.




      In all cases, a feasible sampling strategy was found  for each in-




 dustry/process of possible future  interest for fugitive emission studies.




 In some cases, however,  there were serious limitations to  the applicabil-




 ity of the proposed sampling  strategy and  it is apparent that alternate




 new or new combinations of older sampling  strategies should  be investi-




 gated.




      The most universally applicable  sampling strategy for outdoor sources




 was found to  be the upwind-downwind technique.  In  16  of the 18  indus-




 tries/processes of interest,  this  strategy would adequately  quantify  most




 of  the major  outdoor fugitive  emissions encountered.   There  are  limita-




 tions  in  that  tracers  and/or  diffusion modeling might  have to be done in




 conjunction with sampling  in order to obtain accurate  results.   Elaborate




sampling systems may be required and  considerable total sampling time may




be required for the definition of each fugitive source, which can  be  a
                                   -24-

-------
                               TABLE  5-4

         FEASIBLE SAMPLING STRATEGIES FOR INDUSTRIES/PROCESSES
            SELECTED FOR POSSIBLE FUGITIVE EMISSIONS STUDIES
Industry/Process


Primary Aluminum


Primary Copper
 Electric Furnace
  Steel
 Iron  A  steel
  Foundries
 Coke  Making
 Coal
 Coal Gasification
 COG
 (Char-Oil-Gas)
 Shale Oil
  Petroleum Refining
 Oil Production
 Plastics
 Tires & Rubber
 Phosphate Fertilizer
 Lime
 Sand & Gravel
 Asphalt Batching
 Agricultural
  Operations
Critique of Sampling Indoor
_Fu_gi tiye Emissigns	

Roof'Monitor/Enclosure  Vent
is ideal.

Roof Monitor/Enclosure Vent
limited in accuracy because
buildings too open.  Quasi-
Stack may he used.

Roof Monitor/Enclosure Vent
plus Ouasi-^tack are ideal.
Roof Monitor/Enclosure "ent
plus Quasi-Stack are ideal.
                         None
                         None
Maybe indoors to some
extent, then Roof Monitor/
Enclosure Vent or Ouasi-
Stack would be applicable.

Maybe indoors to some
extent, then Roof Monitor/
Enclosure Vent or Quasi-
Stack would be applicable.

Maybe indoors to some
extent, then Roof Monitor/
Enclosure Vent or Quasi-
Stack would be apolicable.

None
                          None
Strategy For:  Outdoor
Fugitive Emissions

 Upwind-Downwind Acceptable
 Upwind-Downwind  requires
 tracers  and  complicated
 sampling.
 Slaa  expansion might  require
 nuasi-Stack.  Upwind-Down-
 wind  acceptable.

 upwind-Downwind  has problems
 but acceptable for open
 buildings.

 Dpwind-Pownwind  requires
 tracers  and  complicated
 sampling. Quasi-Stack
 may isolate  individual
 sources  but  there are
 many  of  them.

 Upwind-Downwind  useful,
 but fires  in waste piles  may
 defy  analysis.   Quasi-Stack
 may also be  used to point
 sources.

 Same  problems as Coke
 Making.
 Same problems as Coke
 Making.
                                                        Same problems as Coke
                                                        Making.
                               Upwind-Downwind would aopear
                               to be feasible.  Would be
                               difficult to isolate
                               individual sources, however
                               Emissions Inventory can be
                               used also.  Ouasi-Stack can
                               apply to point sources.

                               Upwind-Downwind would
                               appear to be feasible.
                               Would be difficult to
                               isolate individual sources*
                               however. Emissions Inventory
                               can be used also.
 Roof Monitor/Enclosure Vent   Upwind-Downwind acceptable.
 plus Emission Inventory would
 be ideal combination.

 Roof Monitor/Enclosure Vent   Upwind-Downwind acceptable.
 plus Emission Inventory would
 be ideal combination.
 Quasi-stack may be used
 inside.
                          None
                          None
                          None
                          None
 Upwind-Downwind limited
 because of number of
 sources involved. Consider
 remote sensing.

 Upwind-Downwind limited
 because of number of
 sources involved.
 Consider remote sensing.

 Upwind-Downwind limited
 because of number of
 sources involved. Consider
 remote sensing.

 Upwind-Downwind limited
 because of number of sources
 involved.  Consider
 remote sensing. Quasi-
 stack for hydrocarbons.

 Only methods are Upwind-
 Downwind and Quasi-Stack.
                              -25-

-------
 problem if many are present in one industry/process, and if operating





 cycles are short.



      The possibility of  remote sensing which would allow short  time  in-




 terval mapping of  fugitive dust clouds or gas emissions should  be  con-




 sidered as an alternate  sample strategy.  One example where this might




 be a good choice would be at a copper smelter.  Lidar could be  used  to




 map the particulate matter fugitive dust clouds from storage piles and




 building openings  and a  correlation spectrometer could be used  to  map




 fugitive S(>2 emissions.




      The other two strategies which have been considered have about  equi-




 valent applicability to  fugitive emission sources.  The quasi-stack  and




 roof monitor/enclosure vent strategies measure fugitive emissions  which




 can be directed into and vented through specific openings or enclosed




 areas.  This makes them  ineffective in determining fugitive emissions from




 outdoor sources, such as storage piles or process steps such as crushing




 and grinding of gravel,  lime, etc., where the operation is accomplished




 in an open-sided structure or entirely in the absence of an enclosing




 structure.   However,  for many metallurgical processes which occur  inside




 buildings,  such as aluminum electrolysis or electric furnace steel mak-




 ing,  these  techniques  are ideal for measurement of fugitive emissions.




      It can be  concluded that, although adequate concepts of sampling




 systems  and analytical methods for fugitive emissions measurements can




 be specified  for all the industries currently of potential interest  on




 this project, there is obviously room for improvement.  It appeared  that




it would be advantageous to continue through Task II some additional lit-




erature search effort directed toward finding new fugitive emission  sam-
                                  -26-

-------
pling methods.  It also appeared at this juncture that any of the three

strategies mentioned above would be a good choice for most significant

cases of fugitive emissions.



5.2  Task II - Evaluating Various Sampling Strategies

     Task II was divided into  a series of subtasks.  These subtasks,

which followed in logical sequence from Task I, can be described as fol-

lows:


     5.2.1  Evaluating 18 Selected Industries

     The criteria used for development of the original 18 industry list

were:
     1.  Large number of fugitive emission sources
     2.  Large mass  rate of fugitive emissions
     3.  Large number of plants extant in the United States
     4.  Industry  found near urban areas.
     The list was  reviewed for the purpose of rating the industries for

overall fugitive emissions potential and also the steps within them which

might exhibit major  fugitive emissions so that the list could be further

culled.  The criteria applied were the above four plus the following:


     5.  Hazardous nature of gaseous fugitive emissions
     6.  Particle  size of particulate fugitive emissions (respirability)
     7.  Hazardous nature of particulate fugitive emissions
     8.  Probability of control of fugitive emissions by new technology
         in the immediate future.


     A rating system was therefore used as a guide to aid in the selec-

tion of the final  candidate industries, but this was not the only process

used to select these industries.  It was not possible to rank the indus-
                                    -27-

-------
 tries/processes in direct numerical order of  the  seriousness  of  their




 fugitive emissions.  This was  dictated by the nature of  the fugitive emis-




 sion sources, which makes it difficult to quantify  the emission  rate




 and identify the components of that emission.  Table 5-5 shows how the




 ratings were arrived at for the 18 industries.









      5.2.2  Suitability of Sampling Strategies




      Another subtask involved  evaluation of the suitability of the pro-




 posed basic sampling strategies to measure fugitive emissions of the pro-




 cess steps of the 18 industries where fugitive emissions were considered




 significant.  This was done by developing a matrix  (Table 5-6, for exam-




 ple), listing for each of the  industries/processes  the process steps be-




 lieved to have significant fugitive emissions, the  nature of  the gaseous




 and particulate fugitive emissions, and the basic sampling strategy most




 applicable to the fugitive emissions of the step.  A process  flowsheet




 was also developed or extracted from the literature.  Figure  5-1 is an




 example.









      5.2.3  Evaluate Sampling  Analysis Techniques




      The procedures of subtasks  1 and 2 pointed out there existed sub-




 stantial commonality in sampling and analysis requirements from  industry




 to  industry.   There was an obvious need to specify  the best sampling/




 analysis  techniques which  might  be applied to the measurements of fugi-




 tive emissions, but  these  techniques would also need to be compatible




with the  technical  requirements  and the financial resources of each sam-




pling strategy to which they would be applied.
                                   -28-

-------
                                TABLE 5-5

                   METHOD OF RATING HAZARD POTENTIAL OF
              FUGITIVE EMISSIONS OF 18 SELECTED INDUSTRIES
Industry or
Process
1.
2.
3.

4.

5.
6.
7.
8.
9.
10.
11.
12.
13.
14.

15.
16.
17.
18.

Primary aluminum
Primary copper
Electric furnace
steel
Iron and steel
foundries
Coke making
Coal
Coal gasification
COG
Shale oil
Petroleum refining
Oil production
Plastics
Tire and rubber
Phosphate ferti-
lizer
Lime
Sand and gravel
Asphalt batching
Agricultural op-
erations
A
4
5
3

4

5
5
5
5
5
5
3
4
4
4

2
2
2
5

B
4
5
4

4

5
4
4
4
4
4
3
3
3
4

5
5
3
2

C
4
3
5

5

3
4
(4)
(2)
(2)
3
3
5
3
3

5
5
5
5

Criterion
D E
3
3
5

5

5
1
2
2
2
4
2
5
5
3

3
3
3
1

4
4
3

4

5
1
5
5
5
3
3
4
4
3

2
2
4
1

F
4
4
4

4

5
3
5
5
5
4
4
3
3
4

2
2
3
2

G
4
5
4

3

5
3
5
5
5
4
4
3
3
3

2
2
5
2

H
2
4
3

3

3
3
3
3
3
4
4
3
3
2

4
4
3
5

Average
3.6
4.1
3.9

4.0

4.5
3.0
4.1
3.9
3.9
3.9
3.3
3.8
3.5
3.3

3.1
3.1
3.5
2.9

(  )  = Future Estimates

   A   Large number of fugitive emission sources
   B   Large mass rate of fugitive emissions
   C   Large number of plants extant in the United States
   D   Industry found near urban areas
   E    Hazardous nature of gaseous fugitive emissions
   F   Particle size (respirability) of particulate fugitive emission
   G    Hazardous nature of particulate fugitive emission
   H    Probability of control of fugitive emissions by new technology
       in the immediate future

                      (Highest number rates worst)
                                  -29-

-------
                                TABLE  5-6

                   MATRIX OF IRON AND STEEL FOUNDRIES
                    FUGUTIVE EMISSION  SOURCE/SAMPLING
                          STRATEGY COMBINATIONS
Operation
Coke, silica, sin-
ter, scrap deliv-
ery and storage
Furnace or cupola*
charging

felting

Tap and pour
slagging


fold pouring*


'roduct finishing
'ollution control
equipment dust
transfer and
storage
Particulate
Dust**


Fume**
Carbon-dust**
Smoke (oil)**
Fume**
Dust**
Fume**



Dust**


Dus t**
Dust**



Gas and Vapor
Emissions



CO**
Hydrocarbons

CO**

Hydrocarbons**
(odors)
CO,** S02,
H2S
Hydrocarbons**
(odors)
CO,** PNA**
S02, CO




Sampling
Strategy
Upwind-downwind
or Quasi-stack

Quasi-stack or
Roof Monitor

Quasi-stack or
Roof Monitor
Quasi-stack



Quasi-stack


Quasi-stack
Upwind-downwind
or Quasi-stack


 *Most critical steps for fugitive emissions
**Most significant fugitive emissions
                                  -30-

-------
Binder
                                                                          Dust and gases
                                                                  Return
                                                                   sand
                                                 participate
                                                 emissions
                                                        Y?
                                                    Cores W    i
                                                                                            Dust
                                                                                                      FINISHING
                                                                                                         AND
                                                                                                      SHIPPING
                                                                             CASTING
                                                                            SHAKEOUT
                                                                                           COOLING AND
                                                                                             CLEANING
                                                                                        Core sand
                                                                                        and binder
    SAND
PREPARATION
                                                           CORE
                                                         MAKING
       Figure 5-1:   Simplified Flow Sheet for
                    Gray Iron Foundry

-------
      An  in-depth evaluation or critique of certain sampling/analysis tech-

  niques was  therefore undertaken in the areas of:


      1.  Hydrocarbon analysis (gaseous and particulate)
      2.  Fluorides analysis (gaseous and particulate)
      3.  Total suspended particulates analysis (size distribution,
          chemical nature, continuous analysis)
      4.  Sulfur compounds analysis (identification, continuous analysis)
      5.  Remote sensing (particulates or gas, continuous analysis)


      One of  these critiques is included as Appendix E.  These critiques

  aided in evaluation of sampling strategies and they were also very im-

  portant  in  the preparation of a suitable field test procedure in Task

  III.
      5.2.4  Selection of Sampling Strategy/Industry Combinations for
             Further Evaluation

      An evaluation was made to establish the most common basic sampling

 strategy/industry source combinations in the 18 industry list.  This

 was accomplished by first establishing a matrix of 23 strategy/source

 combinations which would cover all or most of the significant fugitive

 emission sources found on the 18 industry list.  From this a determina-

 tion was made of how many of these were related well enough to list

 common  combinations.   This final matrix of four general combinations

 which also  provides the  first  step of Task III, is shown in Table 5-7.

 It  includes  some  additional  elements  which relate to Task III.

     This list of four was jointly reviewed by TRC and EPA, and led to

the choice for preparation of  a  test  procedure and field demonstration

of a quasi-stack  sampling strategy.
                                   -32-

-------
                                                                                                                TABLE 5-7
                                                                                              MATRIX OF FINAL SELECTIONS OF FUGITIVE EMISSIONS
                                                                                               SAMPLING STRATEGY/INDUSTRY SOURCE COMBINATIONS
 I
u>
Fugitive Source
Sampling Strategy

1 , Upwind-Downwind










2. Roof Monitor







Industry Process Step Other Industries
Applied To Involved Where Applicable
Specific
Coal Reactors Charge Coal Gasification
Gasification & Discharge Petroleum Refining
& Storage Oil Production
Of Coal Shale Oil Product
Char-Oil-Gas Prodn.
Asphalt Batching
General
Coal Preparation
Phosphate Fert.
Lime
Sand & Gravel
Primary Potroom " „ sPeciflc
/luminum Primary Aluminum
Primary Copper
Elec. Furnace Steel
Iron & Steel Found.
General
Phosphate Fert .
Plastics
Tire & Rubber

Has:

Med.
Med.
Med.
Med.
Med.
High





Med.
High
Med.
Med.




Fugitive
s Rate

to High
to High
to High
to High
to High






to High

to High
to High




Emissions
Hazardous Nature

PNA's. Reactive Hydrocarbons
PNA's "
"
1
i ii
PNA's only





Fluorides, PNA' st Small Metallic Particulate
SO , Small Metallic Particulate
Fluorides, Small Metallic Particulate
PNA.s, Small Metallic Particulate




                           3.  Quasi-Stack
Iron & Steel
 Foundries
                                                                        Cupola
                                                                                                       „   ,
                                                                                                       Found'
                                                                                          EUc_  Furnace gteel
                                                                                          Primary Copper
                                                                                          Primary Aluminum

                                                                                             General
                                                                                      Med. to High  ReactiveLjHydrocarbons^ Metal Parji^cuJ.ate
                                                                                      Med. to High    "           "         "        "
                                                                                      Very High     S0?, Metal Particulate of Small Size
                                                                                      Med. to High  PNA's, Metal Particulate of Small Size
4.  Quasi-Stack
                                                     Iron &  Steel
                                                      Foundries
  Mold
Pouring
Asphalt Batching
Phosphate Fert.
Plastics
Tire & Rubber
Lime
Coal Gasification
Coke Making
Shale Oil Prodn.

Specific
Iron & Steel Found.
Elec. Furnace Steel
Primary Copper
Primary Aluminum

  General
Asphalt Batching
Phosphate Fert.
Plastics
Tire & Rubber
Lime
Coal Gasification
Coke Making
Shale Oil Prodn.
Medium   PNA' s,  Hydrocarbons L Metal Parjticula^tes^ o£ Small Size
Medium   Metal Particulates of Small Size
Medium   S0_ Metal Particulates of Small Size
Medium   PNS's,  Metal Particulates of Small Size
                                                                                                                                                                                       Nature  of  Fugitive
                                                                                                                                                                                       Em is sions  Measured
                                                                                                                                                                                       Particulates:
                                                                                                                                                                                        A.  Size Dist.
                                                                                                                                                                                        B.  Mass
                                                                                                                                                                                        C .  Compos it ion
                                                                                                                                                                                        D.  Organics  (PNA)

                                                                                                                                                                                       Hydrocarbons
                                                                                                                                                                                       CO
                                                                                                                                                                                       Sulfur Gases
                                                                                                                                                                                       Phenol s
                                                                                                                                                                                       Pyridines
                                                                                                                 Particulates:
                                                                                                                   A.   Size  Dist.
                                                                                                                   B.   Mass
                                                                                                                   C.   Composition
                                                                                                                   E.   Organics  (PNA)
                                                                                                                 Hydrocarbons
                                                                                                                 CO
                                                                                                                 Sulfur Gases
                                                                                                                 Fluorides
                                                                                                                   A.   Gas
                                                                                                                   B. Particulate

                                                                                                                 Particulates:
                                                                                                                   A.   Size  Dist.
                                                                                                                   B.   Mass
                                                                                                                   C.   Composition
                                                                                                                   D.   Organics  (PNA)
                                                                                                                 Hydrocarbons
                                                                                                                 CO
                                                                                                                 Sulfur Gases
                                                                                                                 Fluorides
                                                                                                                   A.   Gas
                                                                                                                   B.   Particulate
                                                                                                                                  Particulates:
                                                                                                                                   A.   Size  Dist.
                                                                                                                                   B.   Mass
                                                                                                                                   C.   Composition
                                                                                                                                   D.   Organics  (PNA)
                                                                                                                                  Hydrocarbons
                                                                                                                                  CO
                                                                                                                                  Sulfur Gases
                                                                                                                                  Fluorides
                                                                                                                                   A.   Gas
                                                                                                                                   B. Particulate
                                                                                                                                  Odors

-------
       5.2.5  Cost-Effectiveness of Sampling Strategies

       The four final matrix sampling strategy/industry source combinations

  were examined to estimate their cost-effectiveness.  The basic procedure

  was:


       1.   Determine accuracy and reproducibility of the techniques.

       2.   Determine minimum experiments required to achieve the above.

       3.   Compare the four techniques under the best conditions for which
           they can be employed.

       4.   Establish matrix of strategy, industry in which employed, fugi-
           tive emissions measured,  costs,  accuracy, and reproducibility.

       5.   Develop cost curves for  each  strategy and plot against the rela-
           tive error limits (error  limit depends on the number of experi-
           ments run) .


       Figure 5-2a shows  such  a  cost curve.   It  was  developed from the ba-

 sic costs of experiments plotted against  the relative error.   It was as-

 sumed that the two  quasi-stack sampling strategies were equivalent in

 costs.

      This procedure results in relative cost-effectiveness curves for

 different sources of fugitive  emissions, whereas to compare each of the

 strategies to'each other, either a weighting factor for the importance

 of  the fugitive emissions measured needs to be applied, or another pro-

 cedure must be used to determine the cost-effectiveness for the same fu-

gitive emission source.

     The following alternate procedure  for evaluating all strategies on

the same basis was used:


     1.  Determine accuracy and reproducibility of the techniques.

     2.  Determine minimum experiments  required to achieve  this.
                                  -34-

-------
                 300
i
u>
Ui
              O
              O
              (O
             JO
              2
             '5
             •o
              S
              a
              0)
             cc
                 200
           I
           I
           I
           O
            I
            I
            I
    I       I
     I      I
    "\      I
      \      I
     I
     I
     A

      1
      I
                                                          Best effort
                                                      — Least cost
      \
      O
          Upwind - downwind
          (Coal gasification)
                              I
       ^T— Quasi - stack
        •     Q   (Iron and
        •     I  ctaal frtunrli*
100
L  I
     \
             I
                          \
                          o
                           \
                           i
steel foundry)
Pouring or charging
                    Roof - monitor
                (Aluminum pot room)
                                                          I
                                                  I
                                                                       	Best effort
                                                                       — — Least cost
                                                      Roof
                                                     Monitor
                                                           Quasi - stack
                                                       (Pouring and charging)
                                                           I
                                                                                         I
                           50,000  100,000  150,000  200,000  250,000

                                         Dollar cost
                         Cost-effectiveness of four fugitive emission source
                      sampling strategies applied to different sources.
                                                              50,000  100,000  150,000  200,000  250,000

                                                                              Dollar cost

                                                                                  B
                                                             Cost-effectiveness of four fugitive emission source
                                                           sampling strategies applied to the same source.
                                                           (2 converters in one copper smelter building).
                                                                   Figure 5-2

-------
      3.  Compare results of use of all four strategies in same industry
         or process (e.g., copper smelter converter house).

      4.  Determine cost vs. accuracy and reproducibility for each.

      5.  Plot cost vs. reproducibility for each strategy (Figure 5-2b)
         by varying the number of experiments and thereby changing the
         relative error limits of tests and the costs.


      This alternative procedure clearly shows that the cost-effectiveness

of  a  fugitive emission source sampling strategy is dependent on the source

to  be tested.  It can, for instance, be shown that a quasi-stack strategy

is  most cost effective when a single source such as iron cupola charging

is  studied, whereas in a copper smelter converter building (multiple en-

closed sources) a roof monitor strategy might be preferred.  In a coal

gasification plant, the upwind-downwind strategy would probably be the

choice.


5.3  Task III - Prepare a Technical Manual for the Measurement of Fugitive
                Emissions

      A draft of a field test procedure for mold pouring in an iron foun-

dry was prepared.


5.4  Task IV - Field Test of a Fugitive Emission Technical Manual

      Task IV consisted of the field evaluation of a procedure for the

quasi-stack sampling method for industrial fugitive emissions, which

was developed during Task III of the program.  The procedures document

presents general guidelines for the design and conduct of a field test

program utilizing the quasi-stack method.  The field evaluation was in-

tended to gather relevant information for refinement of the procedures

document based on actual field operating experience.
                                   -36-

-------
     As a result of an extensive literature study executed during Tasks




I and II which dealt with the identification of sources of fugitive




emissions and the evaluation of sampling strategies, and further re-




view by TRC and EPA, it was decided to evaluate the quasi-stack samp-




ling method on fugitive emissions.  As an example which was considered




feasible, mold pouring in an iron foundry was chosen for the field study.




Five foundries which were cooperative with TRC upon initial contact




were inspected for suitability for these tests.  A foundry located in




southern Connecticut was selected.




     This foundry operates one cupola furnace which has a melting rate




of about 12 tons  (109 kgs) per hour.  Most of the production occurs in a




large main bay area, where molds are assembled in place and remain sta-




tionary during pouring and cooling.  Most of these molds contain sand




cores.  An overhead crane is used to transport the pouring ladle around




the bay and to deliver the molds to the sand shakeout operation.




     It was possible for the field test setup to be located at one end




of the main bay where it was isolated from much of the foundry activ-




ity.  Operation of the cupola, manufacture of the sand cores, mold




shakeout, product cleaning, and a large part of the mold assembly ac-




tivity occurred either at the other end of the bay or in another part




of the foundry.  A ventilation system was designed and installed to




capture emissions during the initial pouring of hot metal in the mold




and the subsequent cooling period.  The system consisted of an enclo-




sure or hood for a test mold, a fan, and associated ductwork.




     Pollutant emission measurements were made in ports located in a
                                   -37-

-------
 straight run of the ductwork.   The following  listing  summarizes the

 specific equipment employed in the test  program  along with  the types

 of pollutants measured:


      1.  IKOR Model 206 Continuous Particulate Monitor;   dry  filterable
          particulate.

      2.  EPA Method 5 Particulate Train  per Federal Register,  August
          17, 1971:  dry filterable particulate,  organic  and inorganic
          condensibles particle size,  and particulate  composition
          (i.e., iron, silica,  nitrate, and sulfate).

      3.  Glass Innovation (G.I.)  Cascade Impactor with EPA  Method 5
          condensible train;  dry  filterable particulate,  organic and
          inorganic condensibles,  and  particle size distribution (aero-
          dynamic diameter).

      4.  Beckman Model 109  Flame  lonization Detector;  total  hydrocar-
          bons (methane basis).

      5.  Intertech Model Infra-2  Non-dispersive  Infra-red Analyzer;
          carbon monoxide.


      A trailer was located  on  the foundry bay next to the test setup.

 This trailer served as an instrument  and chemical field  laboratory (ref-

 erence Figure 5-3).  In addition,  high volume air sampling  stations were

 used to measure ambient air particulate  concentrations in the foundry

 main bay.

      The test program for Task IV was conducted  during the  period from

 March 7 to  March 21,  1975.   Twenty-three (23) individual mold pouring

 tests were  run as  well as four background tests.  While  a wide variety

 of mold  types were encountered, the mix  was considered representative

 of a normal  production schedule for this  foundry.

      This report presents the  results of  the  measurements and discusses

 their significance, both with  respect to  the  foundry  operation and the

quasi-stack measurement method employed.   The measurement and analysis

procedures are also presented  in detail.

     The purpose of the field  test program was to evaluate  the quasi-
                                    -38-

-------
vo
 I
                                                             Particulate Measurement Devices
                                                               IKOR    EPA  CASCADE
                                                                              IMPACTOR
                   Capture
                    hood
                                                                                                       HC and CO line
                                                                                                            Instruments
                             Figure 5-3:    Illustration of test set-up for measuring fugitive emissions from
                                            mold pouring in a gray iron foundry. (EPA Contract 68-02-1815)

-------
 stack sampling strategy  for measuring fugitive emissions  (in  this  case,

 mold pouring in an iron  foundry) and to provide field experience  for im-

 proving the strategy's test procedures.  Emission concentrations were

 obtained during the evaluation and they were related to the process  con-

 ditions at the time of the test.

      Table 5-8 presents  a summary listing of the tests that were con-

 ducted, the associated casting/core weights of the molds  that were poured,

 and the test methods that were used to measure the mold pouring emis-

 sions.  Table 5-9  summarizes  foundry activity during tests.



      5.4.1  Summary of Procedures and Analysis

      5.4.1.1  Description of  the Mold Pouring Hood Arrangement
               (Refer to  Figure 5-3)

      A hood was designed to enclose a single mold and capture most of

 the emissions given off  from  the mold and ladle during the initial pour

 and after the pour while the  mold was cooling.  After witnessing several

 mold pouring operations  on preliminary inspections, it was observed

 that there was a significant  evolution of hot exhaust gas from the

 ladle alone.   The  enclosure design, therefore, had to take into account

 both the ladle and mold  configurations.  Guidelines established by the

 American Conference of Governmental Industrial Hygienists were used

 to  develop initial  plans.  Modifications to the design were made  to

 incorporate  features dictated by specific plant operating practices.

 Since mold and  ladle transfer to and from the hood would be by over-

head crane, it was  necessary to provide hinged side and top panels for

better access.  This also provided a greater degree of operator safety

while handling  the  ladle during the pour.  The final hood design,  there-
                                    -40-

-------
                                                   TABLE 5-8
                                         SUMMARY OF  TESTS - MOLD POURING
Test
No.
1
2
3
4
5
6
Bl
7
B2
8
9
10
11
12
13
14
15
16
17
B3
18
19
20
B4
21
22
23
Test
Date
4/7/75
4/7
4/7
4/8
4/8
4/8
4/9
4/9
4/9
4/10
4/10
4/10
4/14
4/14
4/14
4/15
4/15
4/16
4/16
4/17
4/17
4/17
4/17
4/21
4/21
4/21
4/21
Process Conditions
Casting Core
Wt. , LBS. ¥t., LBS.
800 250
300 275
1000 900
600 500
600 500
300 150
(Background Test-A.M.)
600 500
(Background Test-P.M.)
600 500
600 500
600 500
300 150
300 50
650 0
600 500
300 125
300 175
250 25
(Background Test-A.M.)
250 30
250 30
250 30
(Background Test-A.M.)
600 300
250 125
250 125
Test Methods
Particulate
EPA Cascade
Method 5 Impactor
j
/
/
T/
,/
/

/
/
/
y
/
T/
/
/
/
/
/
/
/
/
/
/
/
/
'
,
/
/
/
/
/

/







/
/
/
/

/
/

J
/
17
Continuous
Monitor


/
/
/
/
/
•/
/
/
/
/
/
/
/
/
/
/
/
/
/
/

J
/
/
Other
)
> Hi-Vol1
j

J Ditto
)
)
> Ditto
j

Ditto


Ditto

T^-T *- t-n
t Ditto
-
-

Hi-Vol1


Tli ^t*n
Is ± L> i~D

Total HC
(CH4 Basis)
7
/
/
,/
/
/
/
/
'/
/
/
/
/
/
/
y
^
/
/
/
/
/
/
/
/
/
CO
y
/
J
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/

-------
         TABLE 5-9




FOUNDRY ACTIVITY DURING TESTING
Test
No.
1
2,3
4

5
6


Bl


7

B2

8
9
10
11



12


13

14



15


16


17

B3


18




19



20


B4


21


22


23


Test
Pate
4/7/75
A/7/75
4/8/75

4/8/75
4/8/75


4/9/75


ft/9/75

4/9/75

4/10/75
4/10/75
4/10/75
4/14/75



4/14/75


4/14/75

4/15/75



4/15/75


4/16/75


4/16/75

4/17/75


4/17/75




4/17/75



4/17/75


4/21/75


4/21/75


It/21175


4/21/75


Test
Start
Time
11:58
1:12,2:18
12:07

1:29
2:49


10:25


12:06

2:11

12:05
12:49
1:45
11:57



1:05


2:01

11:59



1:25


12:08


2:06

10:08


12:44




1:49



2:37


12:12


12:05


1:09


2:10


Description of Foundry Activity During Testing
•
i 15 molds brought into booth area during tests.
Pouring and positioning molds near booth during
test.
Positioning molds near booth, no pouring.
Hold from test f!5 dropped 10 ft in front of
booth upon removal; sat smoking during test
#6, caught somewhat by draft of booth.
Booth background: no mold in booth, no produc-
tion or molds at trailer end of main bay. Ac-
tivity at Herman press.
Molds transferred and assembled in bay area in
front of booth during last half of test.
Pouring completed ^ 2:30. 16 molds poured near
booth .
Mold pouring on bay in front of booth 12:10-12:20.
Mold pouring on bay cupola Herman.
Flask transfer to Herman during run.
Pouring of molds near cupola during test. Opera-
tors and mold assemblers at lunch during test per-
iod - emissions from mold alone. Booth top in full
raised position during pour.
Mold assembly in front of booth during test. Both
roof ventilators over booth on during test. Pour-
ing at other end of bay.
Pouring molds in front of booth 2:00-2:10. No
other activity. Ventilators on, room clear.
Pouring on main bay from Herman to cupola, 12:00-
1:00. No mold pouring near booth 12:00-1:00. Roof
fans closed during first 15 minutes of test, then
opened. Mold assembly in front of booth 12:35-1:00.
Ventilators on over booth. Flasks next to booth
being transferred to Herman 1:20-1:45. Mold assem-
bly in front of booth 1:20-1:50.
Roof monitors on, mold assembler on break 12:00-
12:30. Pouring on main bay cupola to'Herman 12:10-
12:30, 12:30-1:00.
Molds in front of booth poured. No assembly occur-
ring. Roof fans on.
Mold preparation between Herman and cupola. Prepa-
ration near booth at 10:30. Roof fan on during last
15 minutes of test.
Roof monitors on. Mold assembly near booth during
test. Pouring at cupola end of bay 12:30-1:40.
Molds in front of Herman poured 1/2 hour before
test. Room very hazy ^ 1:15. 1:20 - cleaning Her-
man with air gun.
Mold pouring near booth started near end of test.
Roof vents on. No activity in this end of bay 1:45-
2:15. Pouring at other end of bay 1:45-. Molds
from this afternoon still smoking near booth.
Mold pouring in front of booth fron end of Test 19,
continuing until 2:45. Molds noticeably smoking
during latter part of test (on bay floor).
Mold assembly between cupola and Herman. Flask
transfer to Herman. Core cleaning with air hose in
front of booth ^ 10:25.
Roof vents on. Mold assembly break. Mold pouring
in front of Herman and booth 12:05-12:25. Pouring
at other end of bay, 12:25-test end.
Roof vents on. Mold assembly in front of booth.
One row of molds smoking from pour 12:05-12:25.
Flask transfer to Herman next to booth during test.
Atmosphere very hazy. 2:10 mold blowout next to
trailer. Mold pouring of second rwo in front of
booth 2:10-.
             -42-

-------
fore, was functional from the standpoint of plant operating practices,




provided ventilation characteristics sufficient to capture the mold




pouring emissions, and incorporated physical dimensions dictated by




the sizes and configurations of the mold and ladle.




     A two-foot  (0.65 meter) diameter duct served as the exhaust duct-




work for the hood and as the sampling test section.  This duct discharged




from the top of  the hood and, by the use of three 90° elbows, was ar-




ranged to run horizontally, three feet above the floor.  The horizontal




section which incorporated a continuous run of about 22 feet (7 meters),




contained the sample ports and test equipment probes.




     The front opening of the hood was designed for a face velocity of




150  feet (46 meters) per minute (fpm).  This ensured capture of most of




the  exhaust gases from mold pouring, particularly from the ladle when




in the pour position.  Based on an open face area of 57 square feet (5.4




square meters) for the hood and a design capture velocity of 150 fpm




(46 meters/minute), the required exhaust volume flow is 8500 cubic feet




(240 cubic meters) per minute (cfm).  This exhaust flow in the two-foot




(0.65 meter) diamter duct represents a velocity of 2500-3000 fpm (770 925




meters/minute) which is considered adequate to insure that deposition




of particulate by settling will not occur in the duct.  The fan for this




system was located at the end of the horizontal duct run.  The exhaust




flow was returned directly to the bay.  This was considered necessary




to prevent the hood enclosure from causing room air changeover and




thereby affecting the background concentrations of contaminants which




may have resulted if discharge was to the outside of the foundry struc-




ture.  (The location of this hood in the foundry and the quantity of
                                     -43-

-------
 air moved would have made this system one of the largest exhaust  sys-




 tems  in  the bay area.)  Figure 5-3 schematically depicts the mold pour-




 ing hood arrangement, including associated ductwork, sample equipment




 locations, and field laboratory trailer.  More detailed descriptions of




 the trailer and sampling equipment are presented in other sections.









      5.4.1.2  Description of Sample Port Locations




      The location of four sample test ports in the horizontal duct is




 shown in Figure 5-4.  Pertinent dimensions and the arrangement of the




 sample equipment are also depicted in this figure.




      The placement of the three side ports was chosen to minimize tur-




 bulence  effects.  Disturbances to the normal flow pattern were expected




 on the downstream side of each probe; therefore the actual port loca-




 tions had to be arranged so that any one probe was not in the turbu-




 lent  wake of the preceding probe.




      The IKOR continuous particulate monitor probe was located in the




 furthest upstream port.  This probe was positioned at a point nine




 inches (23 cm) from the far side, which was determined from a velo-




 city  traverse to represent a point of average velocity.  The IKOR probe




was not  traversed during the test program.




      The probe for an EPA Method 5 particulate train including a  con-




densible collection section was located in the middle port.  A six




point horizontal traverse was conducted for each test.  Figure 5-4




gives  the position of the traverse points with respect to the duct




diameter.




    A Glass  Innovation (G.I.)  Cascade Impactor System coupled to a
                                    -44-

-------
                  12"
                  «-~|  A
 Cascade Impactor
       train
  Gas sample probe
                                             TOP VIEW
       IKOR
       train
                          EPA
                          train
Fan
                       3'
2'-.
12'
                                                                        Elbow
                                    FRONT VIEW
                          EPA traverse
                          point location
                    123   4 5 6


                    SECTION A-A
                  Distance from
           Point back wall (inches)
             1          1
            2          3.5
            3          7
            4         17
            5         20.5
            6         23
                    Figure 5-4:   Schematic Diagram of
                                  Sample Port Locations
                                  -45-

-------
 condensible  train  Identical to the EPA Method 5 train was located  in




 the third  port.  As  in  the case of the IKOR, the impactor probe was




 positioned at  a  point nine inches (23 cm) from the rear wall and was




 not traversed  during the tests.  Although downstream of other particu-




 late measurement devices, no significant effect on the mass emissions




 as measured  by the impactor was seen from its placement in the wake  of




 other instrument probes.




      A gas sampling  probe was placed in the vertical port situated at




 the same location  as the G.I. cascade impactor and inserted a distance




 of about nine  inches (23 cm).  Twenty-five feet (7.7 meters) of tygon




 sample line  carried  a sample flow from this probe to a fan-manifold




 system contained in  the laboratory trailer.  This gas flow provided  gas




 samples for  the  Beckman hydrocarbon analyzer and the Intertech carbon




 monoxide analyzer  which were operated in the trailer.




      The placement of the three sampling train probes and the one  gas-




 eous sampling  probe  in  the same duct section was designed so that  all




 equipment  could  be operated simultaneously during each test.  This al-




 lowed for  comparisons among sampling techniques.









      5.4.1.3  Field-operated Instrument and Chemical Laboratory




      A utility trailer was located inside the foundry close to the




hood  ducting near  the fan discharge.  It was employed to provide a




clean  environment within the foundry for sample preparation and re-




covery  and for the operation of the hydrocarbon and carbon monoxide  in-




strumentation and  chart recorders.




     A manifold system consisting of two-inch (5 cm) diameter glass




piping and an exhaust blower was used to collect the gaseous sample
                                     -46-

-------
from the vertical sampling probe  and  tygon  tubing.   This  sytem ex-




hausted excess  flow out of the  trailer  through  flexible ducting on the




blower discharge, while the hydrocarbon and carbon monoxide  analyzers




extracted independent  samples on  a  continuous basis  from  the glass pip-




ing.  Each  instrument  extracted a small portion of the total sample




flow and only as much  as  it needed  for  analysis.




     A strip chart recorder with  a  three channel output was  set up in




the trailer to  present continuous traces for the hydrocarbon analyzer,




carbon monoxide analyzer, and for the IKOR  continuous particulate  moni-




tor.  Although  the IKOR output  was  continuously monitored on the strip




chart recorder  in  the  trailer,  the  instrument itself was operated  at




the test port location, as mentioned  previously.




     The  trailer,  in  addition to  serving as  an  instrument laboratory,




was important in  functioning as a field chemical laboratory.  Having




the availability of a  relatively  clean  environment located close to




the test setup  increased  assurance  that the  integrity of  the samples




would be maintained.   A procedure was adopted and followed (discussed




in more detail  in  Section 5.4.2)  for  handling samples in the field.




All sample  mediums, i.e., filters,  were handled and assembled in the




trailer prior to each  test.  At the end of  each test all collected




samples were again handled and  recovered inside the trailer.  Samples




Were placed in  suitable containers  and  logged properly.  Some sample




preparation steps were done in  the  laboratory trailer that would nor-




mally have  been done at TRC's permanent facilities.  Filters were  stored




in a dessicator after  removal from  their holders and chemical extractions




of the liquid medium in the condensible trains  were performed,  yield-




ing separate organic and  inorganic  fractions.   These steps aided in




shortening  the  overall turnaround time  for  sample analysis and  there-

-------
 by permitting the  test  team  to evaluate data at least on a preliminary




 basis within only  a few days.  This procedure would indicate or  suggest




 the need for modification  to planned procedures as the program pro-




 gressed.



      At the end of each test day, all samples were assembled in  the




 trailer, prepared  for transport to TRC's Wethersfield, Connecticut,




 permanent laboratory, and  logged.  As will be discussed later, samples




 were delivered to  the office laboratory on the day of their respective




 tests and were immediately prepared for subsequent analysis.









      5.4.1.4  EPA  Method 5 Particulate and Condensible Train




      Particulate sampling  was accomplished by means of the EPA collec-




 tion train, Method 5, as described in the August 17, 1971, edition of




 the Federal Register.   This  method permits the measurement of  both dry




 filterable particulate  and condensible particulate material.   It is




 shown schematically in  Figure 5-5 and consists of a nozzle, glass-lined




 probe,  filter,  four impingers, vacuum pump, dry gas meter, and an ori-




 fice flow meter.   Following  the probe, the gas stream impacts  directly




 on  a 2-1/2 inch  Millipore  membrane filter supported on a coarse  fritter-




 glass  disc in  a  glass filter holder.  A membrane type filter was used




 to  permit  microscopic particle size analysis and particulate composi-




 tion determinations  to be  performed.  An ice bath containing four im-




 pingers was  attached to  the  back end of the filter via a short section




of  glass connectors.  All  condensible material in the sample stream was




collected here.  The stack velocity pressure was measured using  a cali-




brated type-S pitot tube and inclined manometer, while the stack temper-




ature was monitored by a thermocouple-potentiometer arrangement.  A  nom-
                                    -48-

-------
                   Potentiometer
      Duct wall
  Thermocouple
Nozzle
                Calibrated
                orifice
oo
      Main
Bypass  valve
1

o
/
                                                Leakless
                                                pump
           Figure 5-5:  EPA Method 5 Particulate Sampling
                      Train Including Condensibles
                            -49-

-------
ograph rapidly determined the orifice pressure drop which is representa-

tive of the sampling flow rate required for any pitot velocity pressure

in order  to maintain isokinetic sampling conditions.  It is believed that

particle  pick-up would be weight-biased when either sub or super-isokin-

etic sampling rates are employed.  Generally, a range of within + 10

percent of isokinetic was an acceptable test.  Sampling flow was adjusted

by means  of the needle valve and bypass valve arrangement on the leakless

pump.

     Test data recorded included test time, sample point location,

sample duration at each traverse point, velocity pitot pressure, stack

temperature and pressure, meter volume and temperature, and orifice

pressure  drop (see Appendix F for an example data sheet) .

     At the end of each test, the probe, filter, and impinger assembly

were disconnected and transferred intact to the field laboratory.

Samples were recovered and placed in the following containers:
     Container #1  (plastic petri dish) - 2-1/2" (6.3 cm) membrane fil-
     ter  from glass holder.

     Container //2  (glass jar) - acetone washings of probe and front
     half of filter holder.

     Container #3  (glass jar) - total impinger solution (after volume
     recorded) and water wash of impinger.

     Container /M  (plastic bottle) - silica gel from the fourth impin-
     ger.

     Container #5  (glass jar) - acetone prewash of the probe before
     start of next test.
     A more detailed description of sample recovery, handling,  and

analysis is presented in Section 5.4.2.  A computer program  developed

by TRC was used to calculate the duct emission levels of particulate

material in grains per actual and standard cubic foot,  and the  duct


                                   -50-

-------
volume flow.  Reference is made  to the  Federal Register, August  17,




1971, for the formulas used in performing  the required  computations.









     5.4.1.5  G.I. Cascade Impactor  and Condensible Train  for  Particulates




     A Glass Innovation Model 226 multi-stage cascade impactor was op-




erated in its in-stack mode and  attached directly  to the front end of a




standard EPA Method  5 condensible train with only  slight modifications.




A specially designed type-S pitot was used which allowed for a velocity




pressure reading  to  be taken at  the  impactor nozzle location.  Sampling




was  performed isokinetically and data was  recorded in much  the same




manner as  for the EPA Method 5 train with  the noteable  exception  that




the  cascade train was not  traversed  during the course of the test.  Ap-




pendix G presents an example data sheet for the cascade impactor.




     The impactor unit, which is illustrated in Figure  5-6, consists of




a cone-shaped nozzle and six impactor stages followed by a  built-in




backup filter stage.  All  parts  are  constructed of stainless steel.




The  unit is designed to hold seven radially slotted glass  fiber  fil-




ters and one glass fiber backup  filter  as  follows:






           Filter    Type       Location
1
2
3
4
5
6
7
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Slotted
Top of Stage 0
Top of Stage 1
Top of Stage 2
Top of Stage 3
Top of Stage 4
Top of Stage 5
Bottom of Stage 5
               8      Whole      On backup  filter  support






     The use of an impactor with  filters for  particle  collection on
                                   -51-

-------
                                               EPA probe
       Elbow adapter
            \
Back-up
filter support
            Stage
                                               Nozzle
Type S pitot
                  Figure 5-6:  GI Cascade Impactor Assembly
                                      -52-

-------
each stage was chosen since  it offered  a  lower  tare weight  (as opposed

to steel plate collection) and enhanced particle  collection by minimiz-

ing reentrainment.  The  cascade  impactor  determines the aerodynamic

particle size distribution with  particle  size cutoffs ranging from .5

to 15 microns.  The actual cutoffs obtained depend on the operating

conditions during  a test as  compared  to the unit's calibration data.

Since this impactor is designed  to capture all  particulates, it is also

used to determine  total  filterable particulate  emissions.  As mentioned

previously,  this train contains  a condensible section identical to the

EPA Method 5 train so that condensible  particulates are also measured.

     At the  end of each  test the sampling train was partially discon-

nected with  the impactor and impingers  transferred intact to the field

laboratory.  Samples were assigned containers as  follows:
      Containers  #1-8  (plastic petri dishes) - eight impactor stage fil-
      ters  from the  cascade device.

      Container #9  (glass  jar) - total impinger solution (after volume
      recorded) and  water  wash of  impingers.

      Container #10  (plastic bottle) - silica gel from the fourth impin-
      ger.
     Probe wash and prewash procedures were not followed on the cascade

impactor train since  the impactor  (filter) was located at the front

of the sampling ahead of the probe.  Section 5.4.2 presents additional

detail on the sample  recovery, handling, and analysis procedures.  The

duct emission of particulate in  terms of grains per actual and

standard cubic foot was determined by TRC's computer program as

discussed in Section  5.4.1.4.  The aerodynamic particle size distri-
                                    -53-

-------
 bution was  obtained by  applying the weight percentage of material  cap-




 tured on each  stage to  the mean particle diameter representative of




 that  stage.








      5.4.1.6   IKOR Model  206 Continuous Particulate Monitor




      An IKOR continuous particulate monitor, which was supplied by EPA's




 Industrial  Environmental  Research Laboratory, was used to provide  a quanti-




 tative measurement and  a  real-time continuous display of mold pouring par-




 ticulate emissions.   The  system, shown schematically in Figure 5-7, con-




 sists of a  probe,  sensing unit, gravimetric filter unit, and a control




 unit  with signal integrator.




      Particulate matter in the gas stream flowing through the probe col-




 lides with  an  inline  sensing element resulting in a transfer of charge be-




 tween the particle and sensor.  An electrical signal is generated  which




 is  believed to be  in  response to the mass rate of particulate emissions




 when  the characteristics  of the particles do not vary with time.   A cali-




 bration factor is  obtained by developing a ratio between a simultaneous




 gravimetric sample collected in the filter unit (142 mm glass fiber fil-




 ter)  and an integrated signal of the instantaneous output response.




 This  factor may then be used to convert subsequent integrated signals




 into  mass particulate emissions.




      Sampling  is conducted isokinetically through the one inch I.D.




 probe.  A specially designed venturi section along the probe length




 provides a means for measuring the sample flow rate.  Isokinetic sam-




 pling conditions are attained by matching the pressure drops across the




venturi and a type-S pitot monitoring the duct velocity.  The sample flow




is adjusted by a knob  on the control unit which regulates a variable
                                   -54-

-------
                                      Signal integrator unit
                                                           Blower
                                                           discharge
                                          Sensor unit
                                          (contains sensor element,
                                          gravimetric filter, and
                                          blower unit)
Fig. 5-7 :IKOR continuous paniculate monitoring system.
                   -55-

-------
 speed blower located downstream of  the sensor element and  gravimetric

 filter unit.

      Recorded data included  the test time, sample velocity pressure,

 stack velocity pressure,  stack temperature, sensor temperature,  inte-

 grator parameters,  and  the integrated signal.  An example  data sheet

 can be found in Appendix  H.  For  those tests which include a filter

 run (for calibration purposes) , the filter cartridge was removed and

 transferred to the field  laboratory.  The recovered filter was placed

 in a plastic petri dish.  Handling and analysis of the filter is dis-

 cussed in Section  5.4.2.  Duct particulate emission levels in terms of

 grains per actual  and standard cubic foot were obtained by converting

 integrated signal  data  through application of the calibration factor.



      5.4.1.7   Beckman Model  109 Flame lonization Detector  for Total
               Hydrocarbons

      Figure 5-8 shows the setup of the hydrocarbon flame ionization de-

 tector which  was located  in  the laboratory trailer.  A continuous dis-

 play of the output was  provided by means of a chart recorder.  The hy-

 drocarbon analyzer abstracted a small portion of the total sample flow

 from the gas  sampling manifold system.  Calibration of the hydrocarbon

 analyzer was  performed with  a known concentration of methane gas in

 nitrogen.   The  data  output is therefore presented in terms of the cali-

 brating methane  gas  (as parts per million
     5.4.1.8  Intertech Model Infra-2 Non-dispersive Infrared
              Analyzer for Carbon Monoxide

     The carbon monoxide analyzer is shown schematically  in  Figure 5-8.

The measurement of carbon monoxide by non-dispersive infrared  is the
                                   -56-

-------
          Blower
                                   Sample manifold
                                          (glass)
                                                                      Supply air

                                                                      Fuel supply
Intertech NDIR
carbon monoxide analyzer
Beckman Flame
lonization Detector
                                   3 channel
                                   chart recorder
       Fig. 5-8 :Schematic diagram of laboratory trailer set-up showing hydro-
               carbon and carbon monoxide analyzers.
                                  -57-

-------
 reference method acknowledged by EPA.  Calibration of the carbon monox-




 ide  analyzer was made with a gas of a known carbon monoxide concentra-




 tion.  As in the case of the hydrocarbon analyzer, a sample was obtained




 continuously from a glass sampling manifold and the analyzer output is




 presented on a strip chart recorder.








     5.4.1.9  High-volume Filter Sampler for Ambient Particulate




     The measurement of ambient particulate concentrations at two loca-




 tions  on the foundry main bay was accomplished with high volume sampling




 stations as illustrated in Figure 5-9-  One sampling location was near




 the  booth enclosure while the other was in a production area at the op-




 posite end of the bay, near the cupola.




     A metal shelter encloses a filter-blower assembly in such a manner




 that a measurement of suspended particulates is obtained.  A high vol-




 ume  blower with a flow range of 20 to 60 cubic feet (0.5-1.7 cubic




 meters) per minute draws sample into the shelter hood through slots at




 the  top.  Particulates are captured on an 8 inch by 10 inch (20.3 by




 25.4 cm) glass fiber filter attached to the front of the blower.




     Sample flow through the unit is monitored by a calibrated rotameter




 attached to the blower.  Data collected during each test included sam-




 ple  time, duration of test and the initial and final sample flows.  Ap-




pendix I contains a copy of a data sheet for the hi-vol ambient sampling




procedure.   Filters were removed from the shelter at the end of each




test, folded,  and placed in a manila envelope.  Section 5.4.2 discusses




the procedures  for handling and analyzing the filter samples.
                                 -58-

-------
                      Filter
          Hi-vol blower
             Calibrated.
             flow meter

Fig. 5-9: High volume sampling station for ambient particulates.
                        -59-

-------
      5.4.2  Analytical  Procedures




      A quality assurance procedure was developed which was primarily




 concerned with the  proper methods for sample handling and analysis.  A




 prescribed chain  of custody was assigned for samples taken in  the  field




 and delivered to  TRC's  permanent laboratory facility in Wethersfield,




 Connecticut.   The following sections describe all aspects of the quality




 assurance program including sample recovery and preparations in the




 field,  transport  of samples form the field to the permanent laboratory,




 analytical laboratory procedures, and data reporting.









      5.4.2.1   Sample Recovery and Preparation in the Field




      All  sample media were handled in such a way that initial  loading




 and final recovery  were conducted inside the field trailer, except for




 the high-volume filters, which were loaded and recovered at the sampling




 site.   A  chemical technician operating inside the trailer was  responsi-




 ble for receiving the samples from the test team, for placing  the  recov-




 ered samples  in assigned containers, for labeling and logging  all  sam-




 ples, for  conducting in-field analytical procedures, and for assembling




 samples in a  suitable manner for transport to TRC's laboratory.  It was




 considered essential that these activities be conducted in an  enclosed




 structure  such as the trailer.  During the handling procedure, the test




 team remained outside the field laboratory, thereby minimizing undesir-




able interferences with the collected samples.




     The filters from the EPA, cascade impactor, and IKOR trains were




removed from their respective holders and placed in plastic petri  dishes.




The hi-vol filters were removed at the sampling station, folded, and
                                  -60-

-------
transferred to a manila envelope.  All  filter  samples were  stored in s.

desiccator in the trailer during  the  day  until preparations were  made

for transport to the perinament  laboratory.   Desiccation  in  the  field

shortened the time  required  in  the laboratory  for  the filters to  achieve

a. constant weight.

     Silica gel samples were removed  from the  fourth impinger of  both

the EPA and cascade trains and  placed in  their respective PVC plastic

containers.  The silica gel  samples from  each  test were  transferred  to

their original containers.   All sample  containers were tightly  sealed

and put aside until they were transported.

     The liquid fraction from the first and  second impingers of the  con-

densible train  (EPA and cascade) was  handled as follows:
      1.   Total  liquid  volume was measured at  the end of the test in a
          graduated  cylinder.   This volume was compared to the original
          200 milliliters  of distilled water placed in the impingers and
          the net  gain  in  volume recorded on the field test data sheet.

      2.   The liquid was transferred  to a glass sample container.  Rins-
          ings of  the impingers and graduated  cylinder with distilled
          water  were added to the container.
     Extractions were performed  in  the  field on the liquid samples af-

ter recovery was complete.   The  samples were first extracted three times

with 25 milliliters of  chloroform and then  three times with 25 milli-

liters of ethyl ether.  The  chloroform  and  ether layers were combined

and transferred to a clean glass container.  The container was then

stored on ice.  The remaining water layer from the extraction was trans-

ferred to another glass container and set aside until assembled  for

transport.

     The EPA Method 5 probe was washed  with acetone to rinse out any
                                 -61-

-------
 particulate matter adhering to the glass lining of the probe.  One wash




 (labeled  a pre-wash) was conducted before each test and recorded on  the




 previous  test's  data sheet.  A second probe wash was completed after




 each test.  The  probe wash samples were collected in glass containers




 and stored on  ice.



      Portions  of the distilled water used in the condensible trains  and




 the acetone for  the probe washes were taken daily and transferred to




 glass containers.  These samples served as blanks in the chemical analy-




 sis.



      At the end  of each test day all samples were gathered together  by




 the chemical technician and prepared for transport to TRC permanent  lab-




 atory facility.   The filters were removed from the desiccator and




 sealed in the  plastic petri dishes.  Sample containers from the acetone




 probe washings and chloroform-ether extractions were placed in an ice




 chest.  A log  was prepared which identified the batch of samples being




 transported.   An example of this log appears in Appendix J.  Samples




 were delivered to the permanent laboratory by a member of the test  team




 on  the day of  their respective tests.






      5.4.2.2   Analytical Laboratory Procedures




      Immediately upon delivery to the analytical laboratory the samples




were  unpacked  and prepared by the test team technician for subsequent




analysis.   The lids from the petri dishes were removed and the  filters




 (in  the bottom half of the petri dish) were placed in a desiccator.




The high-volume filters were placed on trays in the desiccator  after




being removed from the manila envelopes.  Sample probe washes and  con-
                                  -62-

-------
densible extractions were  transferred  from their  respective  glass  con-

tainers to tared dishes.   These  dishes were placed  in  an  evaporator hood

to promote drying.  The beaker containing  the  condensible water layer

was placed on top of a hot plate (temperature ^ 100°F [ 38°C] to induce

evaporation and shorten the drying time.   A log was maintained to identify

samples for the analytical chemist performing  the sample  analysis.

     The following is an outline of the procedures used by the analytical

laboratory:


     Filter weight gain -  filters were weighed on an analytical balance
     twice a day and until a constant weight was attained.  Weights were
     reported to the nearest .01 milligrams.

     Silica gel weight gain - silica gel was removed from the plastic
     container and weighed on an analytical balance.  The amount of
     moisture collected was represented by  the difference between this
     value and an original silica gel weight of 200 grams.  The results
     were reported to the  nearest .5 grams.

     Particulate in probe  wash - acetone was evaporated from the tared
     beaker until a completely dry residue remained.  The weight gain
     of particulate in the probe wash, reported to the nearest .01 mil-
     ligrams, was obtained from  the difference in the final and tare
     weights of the beaker as measured on an analytical balance.

     Organic and inorganic condensible - the chloroform-ether and water
     extraction beakers were evaporated until a dry residue remained.
     The organic condensible particulate is represented by the residue
     from the chloroform-ether extraction while the inorganic condensi-
     ble particulate is residue  from the water layer.   The results,  ob-
     tained by a measured weight gain in the tared beakers,  are reported
     to the nearest .01 milligrams for both organic and inorganic par-
     ticulate.

     Microscopic particle sizing of EPA membrane filters - by assuming
     an even distribution of particulate on the membrane filter,  a pie-
     shaped section was cut away.  This piece was  mounted on a viewing
     slide  and  made transparent by a mounting solution.  Dark phase mi-
     croscopy was  used to count the particles.   Results were reported
     in terms of the percentage of particulates by count in the ranges
     of 3y.

     Iron content  of  particulate on EPA membrane filter - a portion of
     the membrane  filter  was dissolved in nitric acid to dissolve the
                                 -63-

-------
      iron.  The distribution of iron and total particulate was assumed
      the same for this portion as for the whole filter.   The iron con-
      centration of the known solution volume was determined by atomic
      absorption.  The results were presented in terms  of  the percent
      iron content of the total particulate.

      Quartz content of particulate on EPA membrane  filter - nitric acid
      was used,  as for iron,  to dissolve a portion of the  filter.   Hydro-
      chloric acid was added  to eliminate any metallic  interferences.
      Silicates  are removed by a hot phosphoric acid digestion.  The re-
      maining quartz was dissolved with hydrofluoric acid.   A color change
      produced when ammonium  molybdate was added was compared with stan-
      dards of known concentration on a spectrophotometer.   Results were
      reported in terms of the percent quartz in the total particulate
      collected.

      Nitrate and sulfate content of particulate on  EPA membrane filter -
      A portion  of the filter was refluxed for 90 minutes  with water.  The
      sulfate concentration of an aliquot was determined by the turbidi-
      metric method.  Treatment of the sample with barium  chloride froms
      a barium sulfate suspension.   The brucine method  was  used for nitrate
      determination.  A yellow color change occurs when brucine reacts
      with nitrate while applying controlled  heat.   A spectrophotometer
      was used to relate the  color change to  a nitrate  concentration as
      compared to standards.


      After analysis was completed,  all filters  (except those destroyed

 in iron,  quartz,  sulfate,  and nitrate determinations), extraction resi-

 dues,  and probe wash residues were  assembled together  and prepared for

 storage.   Filters  were  returned  to  their respective petri  dishes  and

 sealed.   Sample residues were redissolved in the original  solvent (i.e.,

 water  acetone,  chloroform-ether)  and  transferred to clean  glass contain-

 ers.   The  solvent was permitted  to  evaporate again  leaving the sample

 residue.   These containers were  then  sealed.   Samples  were assembled in

a large box and assigned a storage  area.   These  samples will be retained

for possible future analyses  until  such  time as  their  disposition is spe-

cified by EPA.
                                  -64-

-------
     5.4.3  Evaluation of Particulate Concentration Data

     The particulate concentration data obtained in the gray iron

foundry tests were analyzed in detail.  The low concentrations and the

similarity of the results obtained using the three (EPA Method 5, Cas-

cade Impactor, IKOR Continuous Monitor) different sampling methods made

it imperative that a complete statistical analysis of the data be done

to determine if there was:
     A.  Any significant difference between the results using the three
         sampling methods.

     B.  Any significant difference between these results for mold pour-
         ing fugitive emissions and background tests in the gray iron
         foundry.
     The data comparison can be summarized as follows:
     A.  None of the three particulate sampling methods gave statis-
         tically significant difference in results when compared to in-
         plant particulate background as measured by hi-vol filters.

     B.  The instantaneous particulate fugitive emissions can be in-
         ferred from the sample traces from the IKOR online contin-
         uous sampling.  This is approximately an order of magnitude
         higher than the average over the entire test period and the
         general background concentration.

     C.  The period of high fugitive particulate concentration comes
         at the beginning of the test and is of short duration.  This
         results in a low contribution of these actual fugitive emis-
         sions to the total particulate concentration over the test
         period (total = fugitive from pour + background).

     D.  Based upon C, it was considered that only gross estimates of
         fugitive particulate concentrations could be obtained from the
         data taken, since concentrations were low and near error limits
         of measurement.

     E.  Based upon the quasi-stack configuration used, only the IKOR
         continuous monitor was capable of sensing and measuring par-
         ticulate fugitive emissions from the mold pouring operation.
         This could be accomplished without a filter on the IKOR after
         an emission factor has been established, or by accumulating
         on a single filter a large number of short interval mold pour-
                                  -65-

-------
      ing tests (run IKOR only  during peak particulate periods)  in
      the gray iron foundry  and using the accumulated sample  as  an
      average result.


      Table 5-10 presents the particulate concentration data  obtained  us-

 ing the three comparative methods:


      1.  EPA Method 5,  including  condensibles.
      2.  Cascade Impactor,  including condensibles.
      3.  IKOR continuous mass  monitor.


      The duct background test  results are included in the table and am-

 bient background data are presented from hi-vol tests that were conducted

 near the booth during mold  pouring.

      Mean values and standard  deviations were calculated for all  the  tests

 which were within 90 to 110 percent of true isokinetic, including back-

 ground tests.   Table 5-11 summarizes these calculations.  These results

 are for all isokinetic  tests (all methods not necessarily run concurrent-

 ly) .   The same calculation  was  repeated for isokinetic tests and  back-

 ground tests when all of the tests methods were run concurrently.  Table

 5-12  summarizes the  latter  calculation.

      One aspect of reviewing the sampling strategy was to compare the var-

 ious  test  procedures  for accurately monitoring fugitive emissions.  This

 was done statistically  by conducting "t" and "f" tests on the mean val-

 ues and  standard  deviations (Table 5-112) of the particulate results.

 The "t"  tests  are  presented as  a confidence interval for the true dif-

 ference between the averages of the various test methods for both process

emissions  and background  conditions.  The "f" tests were conducted  to con-

firm the results of the  "t" tests.  Intervals are reported  for  95 or  99

percent confidence.   If  a 95 percent confidence interval is  reported, no

comment regarding a significant difference can be made with  a confidence
                                   -66-

-------
          TABLE 5-10

PARTICULATE CONCENTRATION DATA
            GR/DSCF
Test
No.
1
2
3
4
5
6
Bl
7
B2
8
9
10
11
12
13
14
15
16
17
B3
18
19
20
B4
21
22
23
Test
Date
4/7/74
4/7
4/7
4/8
4/8
4/8
4/9
4/9
4/9
4/10
4/10
4/10
4/14
4/14
4/14
4/15
4/15
4/16
4/16
4/17
4/17
4/17
4/17
4/21
4/21
it/21
it/21
Process conditions
Casting Core wt.,
Wt., Ibs Ibs
800 250
300 275
1000 900
600 500 :
600 500
300 150
(Background Test-a.m.)
600 500
(Background Test-p.m.)
600 500
600 500
600 500
300 150 i
300 50
650 0 '
600 500
300 125
300 175
250 25
(Background Test-a.m.)
250 30
250 30
250 30
(Background Test-a.m.)
600 300
250 125
250 125
Test Method
EPA Method 5
Total
Part.
.020
.014
.011
.009
.016
.008
—
.014
.018
.006
.005
.006
.006
.017
.005
.007
.008
.008
.006
.008
.008
.005
.002
.007
.009
.003
.008
Condensibles
Total
.009
.005
.002
.006
.004
.007
—
.007
.013
.003
.004
.004
.004
.015
.004
.004
.007
.006
.004
.004
.002
.001
.002
.003
.005
.002
.003
Inorg
.003
.001
.001
.002
0
.002
—
.004
.005
.001
.002
.002
.002
.012
.002
.002
.005
.004
.002
.002
.001
.001
.002
0
0
0
0
Org
.006
.004
.001
.004
.004
.005
—
.003
.008
.002
.002
.002
.002
.003
.002
.002
.002
.002
.002
.002
.001
0
0
.003
.005
.002
.003
Probe
Hash
.006
.004
.004
.003
.009
.001
—
.004
.002
.003
0
.001
0
0
0
.002
.001
.001
0
.002
.005
.002
0
.001
.003
.001
.004
Filter
.005
.004
.005
0
.003
0
—
.003
.003
0
0
0
.002
.002
.001
.001
0
.001
.001
.001
.001
.002
0
.003
.001
0
.001
% Iso-
kinetic
68
60
62
61
62
61
—
58
61
101
99
98
100
101
102
104
100
102
100
100
100
103
104
99
103
102
103
Cascade Impactor
Total
Part.
.003
.006
.005
.012
.004
.010
—
.010
—
—
—
—
—
—
—
.002
.003
.003
.003
—
.001
.002
.001
—
.004
.007
.005
Condensibles
Total
.003
.006
.004
.008
.004
.009
—
.010
—
—
—
—
—
—
—
.002
.002
.003
.002
—
.001
.002
.001
—
.004
.005
.003
Inorg
0
.001
.001
.003
0
.005
—
.004
—
—
—
—
—
—
—
.001
.001
.002
.001
—
.001
.001
.001
—
.002
.002
0
Org
.003
.005
.003
.005
.004
.004
—
.006
—
—
—
—
—
—
—
.001
.001
.001
.001
—
0
.001
0
—
.002
.003
.003
Filter
0
0
.001
.004
0
.001
—
0
—
—
—
—
—
—
—
0
.001
0
.001
—
0
0
0
—
0
.001
.002
% Iso-
kinetic
102
98
102
67
58
67
—
33
—
—
—
—
—
—
—
95
97
97
103 ,
—
98
106
101
—
96
99
99
IKOR
Monitor:
Filter
—
—
.001
.001
.001

.0006
.002
.003
.0007
.0015
.002
.0007
.0012
.0013
.003
.002
.002
.0023
.0034
.0033
.0012
.0011
—
.0011
.0023
.0035
Hi-vol
Filter
(@ booth)
\
[ .003
)

.001

.0001
.0002
.0002

.0021

)
} .0008
j
.0024
—
—
—
\
> .0025
)

.0035

.0023


-------
                                                          TABLE- 5-U


                                             SUMMARY OF STATISTICAL ANALYSIS


                                        TOTAL OF ALL ISO-KINETIC TESTS FOR EACH METHOD
Test Method
EPA Method 5
(16 Tests)
Cascade Impactor
(13 Tests)
IKOR Monitor
(20 Tests)
Duct Background
EPA Method 5
(2 Tests only)
Duct Background
IKOR Monitor
(4 Tests)
Room Background
Hi-Vol Near Booth
(11 Tests)
Room Background
Hi-Vol in Production
Area. (7 Tests)
Particulate Concentration, Gr/DSCF
Total Part., Incl. Cond.
Mean, y
.0068
.0035
—
.0075
^^
— —

Std . Dev . , o
.0033
.0019
—
.0007
^_
~~

Total Condensibles
Mean . y
.0044
.0029
—
.0035
~~^
•••

Std. Dev. , 0
.0032
.0015
—
.0007
"
"

Part, on Filter (s)
Mean, y
.00231
.0005
.0018
.00351
.0025
.0019
.0036
Std. Dev.j^ a
.0018
.0007
.0009
.0007
.0013
.0011
.0025
CO
I
           1 Includes Probe Wash.  (Probe wash does not apply to other  test methods.)

-------
                                                        TABLE 5-X2

                                              SUMMARY OF STATISTICAL ANALYSIS
                                      TOTAL OF 10 TESTS: ALL METHODS ISO-KINETIC AND
                                             ALL METHODS WERE RUN CONCURRENTLY
Test Method
EPA Method 5
Cascade Impactor
IKOR Monitor
Duct Background
EPA Method 5 (2 tests
only)
Duct Background
IKOR Monitor (4 tests)
Room Background
Hi-Vol near booth (4
tests)
Room Background
Hi-Vol in production
area (3 tests only)
Particulate Concentration, Gr/DSCF
Total Part., Incl. Cond.
Mean , p
.0064
.0031

.0075
—
"™^
"
Std. Dev. , a
.0024
.0019

.0007
—
^_»
"
Total Condensibles
Mean , y
.0036
.0025

.0035
—
"•"""
Jim"1^
Std. Dev. , cf
.0020
.0013

.0007
—
__
-» —
Part, on Filters
Mean , y
.00271
.0005
.0023
.00351
.0025
.0027
.0052
Std. Dev. , a
.0020
.0007
.0010
.0007
.0013
.0006
.0029
vo
           Includes probe wash.  (Probe wash does not apply to other test methods.)

-------
 greater  than 95 percent.  This statistical analysis is presented in Ta-

 ble  5-13a for  total particulate concentration, in Table 5-13b for total

 condensible concentration, and in Table 5-13c for the solid particulate

 concentration  collected on the filter(s).




      5.4.3.1   Background Tests

      The three comparative test methods were run concurrently at isokin-

 etic conditions for 10 tests.  These 10 tests were conducted over a four

 day  period.  During those four days a total of four hi-vol tests near the

 booth for background testing spanned the production duration for mold

 pouring.   Several mold pouring emission tests were often conducted dur-

 ing  one  hi-vol test.  Duct background data was provided by two isokin-

 etlc tests with the EPA Method 5 train.  These tests occurred in the

 morning  prior  to mold pouring.




      5.4.3.2   EPA Method 5 Test Comparison

      From Table 5-12 it can be observed that there was no statistically

 significant difference between the duct tests by SPA Method 5 and the:



      1.   Duct  background "reference" that was determined by EPA Method
          5 and the continuous monitor in the duct.

      2.   Continuous monitor tests ("particulate on filter").

      3.   Room  background "reference" that was established by the hi-vol
          tests near the booth.



     The EPA Method 5 duct results were higher than the cascade im-

pactor results.  Compared to the cascade impactor, the following confi-

dence intervals were obtained for the difference in results:
                                   -70-

-------
             TABLE 5-13 (a)

 DIFFERENCES IN TEST METHOD PERFORMANCES
TOTAL PARTICULATE CONCENTRATION, GR/DSCF
Comparison of Methods
Cascade Impact or

EPA
Method 5
to-
Duct Backgrounds
by EPA Method 5


Cascade Impactor
to-
Duct Background
by EPA Method 5

Statistical Analysis
"t" Tests
Interval
0.003310.0029



None




0.004410.0026


Confidence
99%



99%




99%


"f" Tests
"f" Ratio
12.0



0.6




10.1


Critical "f"
& % Confidence
8.3; 99%



10.0; 99%




10.0; 99%


Remarks
Significant dif-
ference confirmed
by "f" test.

No statistical
difference con-
firmed by "f"
test.

Significant dif-
ference of lower
Cascade results
confirmed by "f"
test.

-------
                                                        TABLE  5-13  (b)



                                          DIFFERENCES  IN  TEST METHOD  PERFORMANCES

                                         TOTAL CONDENSIBLES CONCENTRATION,  GR/DSCF
i
-o
N>
Comparison of Methods
Cascade Impactor
EPA Method 5
to -

Duct Background
by EPA Method 5


Cascade Impactor
to -

Duct Background
by EPA Method 5
Statistical Analysis
"t" Tests
Interval
None


None




None



Confidence
997.


99%




99%



"f" Tests
"f" Ratio
2.3


0.3




1.3



Critical "f"
& % Confidence
8.3; 99%


10.0; 99%




10.0; 99%



Remarks
No statistical
difference con-
firmed by "f"
test.
No statistical
difference con-
firmed by "f"
test.

No statistical
difference con-
firmed by test.


-------
                 TABLE 5-13 (c)

   DIFFERENCES IN TEST METHOD PERFORMANCES;
PARTICIPATE ON FILTER(S)  CONCENTRATION, GR/DSCF



Comparison of Methods
EPA ° Cascade Impactor
Method 5 (10 Tests)
(10 Tests)
to -
° Continuous Monitor
(10 Tests)

° Duct Background,
By EPA Method 5
(2 Tests Only)
0 Duct Background,
By Continuous Moni-
tor (4 Tests only)
0 Room Background
At Booth, By Hi-Vol.
(4 Tests only)
0 Continuous Monitor
Cascade
Impactor
to -
* Duct Background
By Method 5
" Duct Background
By Continuous Moni-
tor
° Room Background
Statistical Analysis
"t" Tests

Interval
0.0022+0.0020


None


None


None


None


0.0018+0.0011



0.0030+0.0023

0.0020+0.0019


Confidence
99%


99%


99%


99%


99%


99%



99%

95%

(
0.0022+0.0012
At Booth, By Hi-
Vol.
0 Duct Background
By EPA Method 5
Continuous
Monitor
to - ° Duct Background,
By Continuous Moni-
tor
° Room Background
At Booth, By Hi-
Vol.
Duct ° Duct Background,
Back- By Continuous Moni-
ground , tor .
By EPA
Method 5
to - ° Room Background
At Booth, By Hi-
Vol.
Duct
Back-
ground ,
By Con-
tinuous. ^om Background
Monitor Afc Booth> By Hi_
to * Vol.


None
None


None


None


None
99%


99%
99%


99%


99%


99%





None






99%


"{" Tests

"f" ratio
10.8


0.3


0.6


0.03


0.01


2.16



21.6

14.3


29.8


2.5
0.1


0.6


0.5


0.6





0.2


Critical "f"
5. % confidence
8.3; 99%


8.3; 99%


10.0; 99%


9.3; 99%


9.3; 99%


8.3; 99%



8.3; 99%

4.8; 95%


9.3; 99%


10.0; 99%
9.3: 99%


9.3; 99%


21.2; 99%


21.2; 99%





13.8; 99%


— 	


Remarks
Significant
difference c ^--
firmed by "f" tes-. .

No statistical
difference con-
firmed by "f" test.
Ditto


Ditto


Ditto

i
Significant
difference of lower
Cascade results
confirmed by "f" test.
Ditto

Ditto


Ditto


No statistical
difference con-
firmed by "f" test.
Ditto


Ditto


No statistical
difference con-
firmed by "f" test.


Ditto





No statistical
difference con-
firmed by "f" test.
,
!
                  -73-

-------
      1.   Total particulate  (99  percent  confidence  interval)  -  0.0033 +
          0.0029 gr/dscf

      2.   Total condensibles  (99 percent confidence interval) - no  sta-
          tistically significant difference.

      3.   Particulate on filter(s)  (99 percent confidence interval)  -
          0.0022 + 0.0020 gr/dscf


      Thus,  the most similar  results between the two methods were the

 total condensibles.  From the mean values of Table 5-15, total condensi-
                                                                  1
 bles represented 80 percent  of  the total sample collected by the cascade

 impactor train and 56 percent of the total sample  collected by the  EPA

 Method 5 train.   The higher  percentage  of condensibles in the  cascade

 train can be attributed to collecting insufficient dry particulate  on

 each filter stage (as will be discussed in Section 5.4.3.3).

      In  summary,  process emissions in excess of the background concentra-

 tion were not detected by the EPA Method 5 train.  Statistically,  there

 was  no significant  difference between the EPA Method 5 train results and

 the  continuous monitor results.  The EPA Method 5 results were statis-

 tically  higher than the cascade  impactor results, with the exception of

 total  condensibles.



     5.4.3.3   Cascade  Impactor Test Comparison

     Particulate  results  from the continuous monitor are limited to a

dry particulate  (listed  as "particulate on filter" in Table  5-13c)  since

there is no condensibles  train presently designed  for the unit.  The re-

sults obtained showed no  statistically  significant difference  when com-

pared to  the:
                                   -74-

-------
     1.  EPA Method 5 dry  particulate duct test results..


     2.  Background results obtained in the duct by:

         A.  EPA Method 5
         B.  Continuous monitor


     3.  Background results obtained in the room by hi-vol tests near
         the booth.



     The total particulate 99 percent confidence interval for the true


difference between the averages of  the methods was:



     1.  0.0033 + 0.0029 Gr/DSCF when compared to EPA Method 5
     2.  0.0018 + 0.0011 Gr/DSCF when compared to the continuous
         monitor ("Particulate on Filter").



     Several conditions must be discussed pertaining to the cascade im-


pactor results.  First, no traversing was done.  This may result in lower


concentrations than EPA Method 5.


     Secondly, the particulate catch on individual filters was frequently


not detectable or lower than the amount that is recommended by the im-


pactor's manufacturer (minimum of 0.2 to 0.5 mg per filter) for reliable


results.  A longer test time to collect more particulate on the filters


may be required, but this  was prevented by the constraints of test tim-


ing.  The range of total dry particulate collected for the impactor tests


was only 0.5 to 2.75 mg, whereas the total tare weight of cascade fil-


ters was approximately 850 mg.  Thus the greatest collection rate only


represented about 0.3 percent of the total tare weight.  The amount col-


lected and the accuracy of detecting that amount leads to the conclusion


that no meaningful particle size distribution results were obtained and


that the total dry particulate results may be erroneous.


     Thus, in applying the cascade  impactor procedure to the quasi-stack
                                   -75-

-------
 strategy,  a process emission concentration and duration must be substan-

 tially  greater  than this particular mold pouring operation to provide

 meaningful results.



      5.4.3.4 Continuous Monitor Test Comparison

      Particulate  results from the continuous monitor are limited to a

 dry particulate (listed as "Particulate on Filter" in Table 5-13c) since

 there is no cohdensibles train presently designed for the unit.

      The results  obtained showed no statistically significant difference

 when compared to  the:


      1.  EPA Method 5 dry particulate duct test results

      2.  Background results obtained in the duct by:

         A.  EPA  Method 5
         B.  Cascade Impactor

      3.  Background results obtained in the room by Hi-vol tests
         near the booth.


      The continuous monitor results were higher than the cascade impac-

 tor  results  (a  99 percent confidence interval of 0.0018 + 0.0011 gr/dscf

was  obtained).

      From  the continuous monitor trace it was found that substantial mold

pouring particulate emissions occurred at the time of pouring, and, in

a matter of one or two minutes, the emissions returned to the genera

background levels for the remainder of the test (30 minutes total  for

participates).  This is clearly illustrated by Figures 5-10 and 5-11.  Of

all the particulate emission testing procedures that were used during

this program, the continuous monitor was the only procedure that was ca-
                                   -76-

-------
 3
 CO
 H
8
O


CO
       I

      5

-------
 I

?
CO

H
n
o
      M
CO    Ln

-------
pable of defining  (either quantitatively or  qualitatively) this brief,




large emission at  the  time of mold  pouring.  The EPA Method 5 and cas-




cade impactor procedures averaged that  short emission excursion over a




30 minute sampling period, in this  case and  thus reduced the detectabil-




ity of this emission.




     From Table 5-11,  the mean  concentration that was determined by con-




tinuous monitor was 0.0018 gr/dscf  with a  standard deviation of 0.0009




(20 tests).  Figures 5-10 and 5-11  show that the concentration due to the




emissions at the time  of pouring may be an order of magnitude greater




than the general background level,  or approximately 0.02 gr/dscf (at




8500 scfm).  This  brief emission excursion for one mold is average by




EPA Method 5 and cascade impactor procedures over the 30 minute test per-




iod.  This emission excursion has greater significance when it is con-




sidered as an accumulative emission for the large number of molds that




are typically poured per day in this foundry.




     Thus, for a mold  pouring operation such as this where the molds are




stationary at all  times and the particulate emissions only occur briefly,




the EPA Method 5 and cascade impactor procedures average the emissions




over a long period of  time and  essentially monitor the background con-




centration.  A continuous monitor is required to define the emission ex-




cursion.



     The EPA Method 5  and cascade impactor procedures may be suitable to




mold pouring operations where numerous  molds are conveyed to a station-




ary pouring station in a short  period of time.  Monitoring particulate




concentrations in  excess of the background condition would be more likely




to occur than in the case tested for this program, and there would be
                                   -79-

-------
 more  of  a particulate  catch for particle size distribution and composi-




 tion  analysis.









      5.4.3.5  Test  Setup Limitations




      As  discussed earlier, the nature of the mold pouring operation was




 such  that a large booth and large flow volume was required to capture




 the  fugitive  emissions.  The large flow volume probably induced a con-




 vective  air current at a considerable distance from the booth.  Any




 activity adjacent to the booth such as core cleaning, sand compacting,




 or vehicle traffic  (all of which occurred regularly during the testing),




 could generate  local fugitive emissions that might enter the booth.




 Thus, a  general activity background level would be monitored by the sam-




 pling apparatus in  the duct.  That activity emission level may be as sig-




 nificant as the 30  minute averaged emissions from the mold pouring and




 cooling  operation itself.  This is supported by the statistical compari-




 sons  discussed  earlier.









      5.4.3.6  Critique of Sampling Methods




      In  Section 5.4.3.4 a discussion was given of the significance of




 the emission  excursion above general background concentrations at the




 time  of  mold  pouring.  The continuous monitor was the only method that




 defined  that  excursion.  Unfortunately, the continuous monitor did not




 have  the  provision  for monitoring total condensibles, which comprised the




 major portion of the particulate fugitive emissions  (Section 5.4.3.2).




A condensibles  train designed specifically for the continuous monitor




would have  to be larger than the EPA Method 5 train since  the sample  flow
                                   -80-

-------
volume is much greater.  This would  constitute a hybrid monitor that gave




continuous dry particulate emissions and integrated condensible material



emissions.




     For this application, the  cascade impactor results were question-




able.  For more concentrated emissions and longer sampling times, this




method may be a suitable approach  for providing particle size distribu-




tion and total particulates, including condensibles.  The short sampling




times and generally  low emissions  encountered during this program led




to the conclusion  that the EPA  Method 5 train is probably the best all-




purpose method for particle size distribution (by microscopy, which will




be discussed in Section 5.4.3.7),  total particulates including condensi-




bles, and particle composition  (which will be discussed in Section 5.4.3.8.




The major limitation of the EPA Method 5 for this application was the in-




ability to determine brief excursion emissions, as described above.









     5.4.3.7  Presentation of Particle Sizing Data




     The cascade impactor test  results were eliminated from considera-




tion for particle  size analysis for  the reasons outlined in Section




5.4.3.3.  As a backup for the cascade impactor, a membrane filter was




used in the EPA Method 5 train  setup.  Particle size distribution by count




was determined by  microscopy.   Table 5-14 presents the results of that




analysis.



     The statistical analyses of the 16 isokinetic emission tests and




the two EPA Method 5 duct background tests (also isokinetic) are presented




in Tables 5-15 and 5-16.  Mean  values and standard deviations of the par-




ticle size distribution by percent count are presented in Table 5-14.
                                    -81-

-------
               TABLE 5-14

PARTICLE SIZE DISTRIBUTION (% BY COUNT):
              EPA METHOD 5
Test No.
1
2
3
4
5
6
Bl
7
B2
8
9
10
11
12
13
14
15
16
17
B3
18
19
20
B4
21
22
23
% Isokinetic
68
60
62
61
62
61
—
58
61
101
99
98
100
101
102
104
100
102
100
100
100
103
104
99
103
102
103
Distribution
% 3p
6.2
3.4
4.6
—
3.9
2.5
—
1.9
2.3
2.4
2.2
3.2
1.8
3.0
2.0
3.0
3.0
3.0
1.0
5.0
3.0
2.0
1.0
6.0
3.0
2.0
2.0
                  -82-

-------
                   TJSBLE 5-15

    PARTICLE SIZE DISTRIBUTION  (% BY COUNT):
       MEAN VALUES AND STANDARD DEVIATION

Tests
16 Isokinetic Emission
Tests (p.m.)

2 Isokinetic Duct Back-
ground Tests (a.m.)


1 Distribution
3y
3y
y,
Mean
67.03%
30.00
2.35
52.00%
42.50
5.50
o,
Std. Deviation
6.91%
5.53
0.72
2.83%
2.12
0.71
                   TABLE 5-16

    PARTICLE SIZE DISTRIBUTION  (% BY COUNT):
STATISTICAL DIFFERENCES BETWEEN PROCESS EMISSIONS
            AND BACKGROUND CONDITION
Distri-
bution
3y
Statistical Analysis
"t" Tests
interval
15.03% +9.81%


12.50% + 7.14%
3.15% + 2.30%
Confi-
dence
99%


99%
95%
"f" Tests
u.pii
ratio
8.91


9.60
34.13
Critical "f"
& % confidence
8.53; 99%


8.53; 99%
4.49; 95%
Remarks
Statistical
difference con-
firmed by "f"
test
Ditto
Ditto
                      -83-

-------
 Table 5-15  presents  the "t" test results as a confidence interval, as




 discussed in  Section 5.4.3, and the "f" test results to confirm that a




 statistical difference between the sets of data existed.




      From the analysis it can be concluded that there is a statistical




 difference  between the process emissions (monitored in the afternoon)




 and  the background condition (monitored in the morning).  Approximately




 two  thirds  of the process emission particles were less than one micron




 in size, and  approximately one third were in the one to three micron size




 range.  This  compares to the duct background condition of approximately




 half  of the particles at less than one micron and about 40 percent between




 one and three microns in size.




      As discussed above, there is a statistical difference between the




 process and background data.  It can be concluded that a process emis-




 sion  particle characteristic, not a general background particle charac-




 teristic, was monitored; but, as discussed in Section 5.4.3.1, the concen-




 tration that was monitored by EPA Method 5 did not exceed the general




 background  concentration.









      5.4.3.8  Particulate Composition




      The average particulate catch on the EPA Method 5 membrane filters




was only about 1 mg for the isokinetic tests.  For most laboratories and




 standard analytical methods this is an insufficient particulate sample




 for analysis of composition by percent weight.  Longer sampling time to




obtain more of a particulate catch was ruled out since most of the emis-




sions occurred in a one to two minute period at the time of pouring, as




discussed in Section 5.4.3.4.
                                   -84-

-------
     Analytical methods  described  in  Section 5.4.2 were employed despite




the insufficient  sample.   The  accuracy of  the  results prevents reporting




meaningful data for  composition by percent weight, but it can be reported




that traces of Si02,  Fe,  SO^"  and  N0~3 were detected on most of the fil-



ters.




     To improve particulate composition  results  for operations such as




the one tested, entire filters rather than filter fractions must.be




used for each analysis.   For operations  with molds conveyed to a station-




ary pouring station,  both entire filters and longer sampling periods




should be used to improve particulate composition results.









     5.4.4  Evaluation of Hydrocarbon Concentration Data




     The layout of molds on the foundry  floor  included double rows of




molds along the length of the  bay.  Pouring began, as shown in Figure




5-12, at approximately noon each day at  the cupola area opposite the test




booth.  Pouring progressed along one row towards the booth, then started




on the second row, and again progressed  toward the booth.  Pouring was




usually completed by 2:30 or 3:00 p.m. each day. Table 5-9 presents a




detailed description of foundry activities during each test.




     Table 5-17 presents the total hydrocarbon (methane basis) concentra-




tion data that was obtained as outlined  in Section 5.4.4.  Concentrations




are reported from the daily plots at 30  minute intervals between noon




and 3:00 p.m. for each of the  nine test  days.   The mean concentration




and standard deviation was calculated for  each time,  tabulated in Table




5-17, and presented  graphically in Figure  5-13.




     Normally three  tests (mold pourings)  were conducted  at  the  test booth
                                    -85-

-------
 Pouring
 direction
   Holding
   furnace  Cupo|a
     DO
n
nn
an
a
      Molds
an
a
nn
n
    n
nn
1st  2nd
row  row
nnnn
       Herman
       press
        fJ
Test
booth

Test
duct

Fan
    Laboratory
    trailer
                       Foundry
                       bay
  Fig. 5-12. Layout of molds.
      -86-

-------
                                                               TABLE 5-17
                                                     TOTAL HYDROCARBON CONCENTRATION
                                                              (ppm as
Day



Time



12:00
12:30
1:00
1:30
2:00
2:30
3:00
Total Hydrocarbons, ppm (as CH^)
4/7/75
48
46
34
30
28
58
> 64
4/8/75
5
14
25
29
29
62
47
4/9/75
7
10
38
20
16
30
50
4/10/75
4
17
231
70
48
46
86
4/14/75
22
96
62
26
70
39
96
4/15/75
20
22
c.200
56
41
54
—
4/16/75
12
12
c.200
50
20
72
50
4/17/75
—
—
48
56
48
70
72
4/21/72
82
60
46
66
76
72
68
Statistical Analysis
Mean HC Cone . ,
ppm (as CH^.)
25
34.6
98.2
44.8
41.8
55.9
66.6
Std. Dev., HC
ppm (as CHtt)
27
30.6
85.2
18.7
21.0
15.1
17.8
I
00
-4
I

-------
zoo
180
160
140

120
O
|ioo
80

60
40



20
«••




—

/
/
/
i
••
8
_


<


;
*
f ***
Startof
1st test
i
I I
/
i


4
•
9 -No. of test days



/T/ Standard deviation


^
\
\ 4
V/v Plot of mean values 1 1
* T i^
\ *• • ^***^l
^™ 1
-*--••
y^-Duct
X^ background
X test
Startof * startof
2nd test 3rd test
^ t
12--00 1:00 2:00 3=00
Time of day
Fig. 5-13.  Plot of mean half-hour hydrocarbon
          concentrations for period April 7-21,1975.
               -88-

-------
each day.  The average start time  for  the  three tests is shown in Figure




5-13.  One background test was conducted in the duct when no mold wss in




the booth.  This  test is  also presented in Figure 5-13, and It generally




follows the pattern  of the mold pouring tests.




     It can be observed  from Figure 5-13 that two peaks occurred during




the day.  Daily observations determined that these peaks occurred about




50 minutes after  pouring activity  in the vicinity of the booth or the




adjacent Herman press—not necessarily after a mold was poured in the




test booth.   Entries in  Table 5-17 show that pouring occurred near the




booth  (first  row  pouring)  and Herman press area between 12:10 and 12:30




on April 8, 10, and  21  (booth)  and April 15 and 16 (Herman press).  Fig-




ure 5-13 illustrates the first  peak occurring from this pouring activity




about  1:00 p.m.   The second  row near the booth was poured between 2:00




and 2:45 on April 9, 14, 16, 17,  and 21, as recorded in Table 5-17.  The




second peak of Figure 5-13 appears at  3:00 p.m. or later, after pouring




had been completed.




     It can thus  be  concluded  that HC  monitoring essentially defined a




background or general foundry  activity condition and isolation of a spe-




cific  mold or the emissions  contribution of one mold was not distinguished




from general  background  concentrations.









     5.4.5  Evaluation of Carbon  Monoxide  Concentration Data




     Table 5-18 presents the  carbon monoxide  concentration data  that was




obtained as outlined in  Section 5.4.1.8.   Concentrations are presented  from




the daily plots at 30 minute  intervals between noon  and 3:00 p.m.  The




mean concentration and standard deviation  was calculated for each  time,
                                    -89-

-------
                                                           TABLE :5-18

                                               CARBON MONOXIDE CONCENTRATION DATA,
                                                               PPM
Day
12:00
12:30
1:00
Time 1 : 30
2:00
2:30
3:00

4-7-75
27
28
28
30
32
66
66
Carbon Monoxide Concentration, ppm
4-8-75
5
31
24
25
25
85
60
4-9-75
5
26
28
20
20
55
36
4-10-75
4
25
44
35
35
46
48
4-14-75
9
50
31
19
34
56
43
4-16-75
5
21
27
20
15
65
38
4-17-75
—
—
31
24
23
45
70
4-21-75
17
14
12
12
18
20
22
Statistical Analysis
Mean CO
Concentration, ppm
10.3
27.9
28.1
23.1
25.3
54.8
47.9
Standard
Deviation
8.7
11.2
8.8
7.1
7.6
19.0
16.6
O

-------
tabulated In Table 5-18, and presented graphically in Figure 5-14.  A back-



ground test was also plotted on Figure 5-14.




     From Figure 5-14  it can be observed  that as in the case of total hy-




drocarbons, there are  two peaks again appearing.  From the pouring ac-




tivity description given in Table  5-18 and Table 6-9, it was noted




that CO concentrations increased almost immediately after pouring occurred




near the booth and in  the adjacent Herman press area.  The one duct back-




ground test also closely followed  the pattern of Figure 5-14.




     Thus, it can be concluded that  a general background or activity con-




dition was monitored during  the CO testing.  Isolation of a specific mold




or  the emissions contribution  of one mold was not distinguishable from




general background concentrations.




     As in the  case  of the HC  measurements,  this procedure for CO moni-




toring may be more suited  to the  foundry  with a central pouring station(s),




If  quasi-stack  procedures  are  to  be  applied  to  the  type of foundry opera-




tion that was  tested,  the  test mold  will  have to be isolated better from




general room activity.  If  this is not  possible, a  broader approach for




defining emission  concentrations  will have to be employed, such as roof




monitoring or  traditional  industrial hygiene measurement  techniques.









5.5 Task V -  Issue  a Technical Manual



     After revision  and thorough  testing in the field,  a  final draft  of




the field  test  procedure was issued.
                                    -91-

-------
80
70

60
uw
50
8
| 40

30

20
10

0
^hM
8
•
M»


4
\
I
7- No. of test days jl
T T //
Standard deviation a / f "
1
i

8
1
A
'~"V
XV 1
\N-
\
\

" - f T/7 A
Plot of mean values j ,-'*" ^ '-.^ / ,' Duct
\^X ^^^ XJ ' M background
/ M 	 '" I test
-7l 1
Average Average Average
. start of start of start of
A 1st test A.2ndtest 1 3rd test
.. T T
i • i i
1 11
12:00 1:00 2:00 3:00
                  Time of day

Fig. 5-14.  Plot of mean half-hour carbon monoxide
          concentrations for period April 7-21,1975.
             -92-

-------
5.6  Units of Measure




     Although it  is  the policy of EPA to use the  metric  system  for qual-




itative descriptions,  the British system is generally  used  in this re-




port.  Appendix K contains a conversion table to  aid readers who wish




to  use metric  units where only British units are  cited.
                                      -93-

-------
                APPENDIX A




LITERATURE REFERENCES TO FUGITIVE EMISSIONS
                 -94-

-------
 1.   "^ocessing Emissions and Occupational Health  in  the Ferrous Foundry
     Industry,  C. E. Bates, L. D. Scheel,  AIHA Journal, 452, August,


 2.   "Systems Analysis of Emissions and Emission Control in the Iron
     Foundry Industry," A. T. Kearney & Co., Chicago,  Illinois, Report
     on EPA Contract CPA-22-69-106, February, 1971.

 3.   "Environmental Surveys of Aluminum Reduction Plants," P. J. Shuler,
     P. J. Bierman, HEW Publs., (HIOSH) 74-101, April,  1974.

 4.   "Air Pollution Control in the Primary  Aluminum Industry," Singmas-
     ter & Breyer, New York, New York, Report on EPA Contract CPA 70-71,
     March, 1972.

 5.   "Investigation of Fugitive Dust - Sources, Emissions, and Control,"
     PEDCO Environmental Specialists, Cincinatti, Ohio, Report on EPA
     Contract 68-02-0044, June, 1974.

 6.   "Soil Conditions that Influence Wind Erosion," W.  S. Chepil, U. S.
     Dept. of Agr. Tech. Bull. 1185, 1958.

 7.   "Asphalt Hot-Mix Emission Study," The  Asphalt  Institute, College
     Park, Maryland, Pub. RR-75-1, 1975.

 8.   "Atmospheric Emissions from  the Asphalt Industry," L. Laster, EPA
     Publ. EPA-6509-73-046.

 9.   "Air Pollution  from Coal Refuse Disposal Areas,"  V. M. Sussman and
     J. J. Mulhern,  JAPCA 14, 279, July, 1974.

10.   "Control of Coke Oven Emissions," Battelle Columbus Laboratories
     Report to AISI, December  31,  1971.

11.   "Control of Coke Oven Emissions," T. Dancy, AISI  Yearbook, p. 181,
     1970.

12.   "Systems Study  for Control of Emissions - Primary Non-ferrous Smelt-
     ing Industry,"  A. G. McKee & Co., Cleveland, Ohio.  Report on NAPCA
     Contract PM86-6585, June, 1969.

13   "Retrofit S02 Techniques  for Existing Equipment at Arizona Smelters,"
     I. J. Weisenberg, P. Wondra,  68th Annual APCA Meeting, Boston, Mas-
     sachusetts, Paper 75-23.1, June,  1975.

14.   "Compilation of Air Pollutant Emission  Factors,"  EPA Pub.  AP-42,
     second edition, April,  1973.

15.   "Solutions for  Feedlot Odor  Control Problems:  A Critical Review,"
     R. M. Bethea, JAPCA 22,  765, October,  1972.

16.   "The Problem of Fugitive  Dust in  the Highway Construction Industry,"
                                 -95-

-------
      F. A. Rennlnger, 46th Annual Meeting Highway Research Board, Wash-
      ington D. C., January, 1967.

 17.   "Hydrocarbon Pollutant Systems Study," MSA Research Corporation,
      Homer City, Pennsylvania, Report to EPA, October 20, 1972.

 18.   "The Identification of Effluents from Rubber Vulcanization," S. M.
      Rappaport, presented at the EPA Conference on Environmental Aspects
      of Chemical Use in the Rubber Processing Operations, Akron, Ohio,
      March 13, 1975.

 19.   "The Contribution of Open Burning of Land Clearing Debris to Air
      Pollution," M. Feldstein, JAPCA 1JJ, 542, November, 1963.

 20.   "Control of Particulate Emissions from Lime Plants - A survey,"
      L. J. Minnick, JAPCA, 195, April, 1971.

 21.   "Particulate Pollutant System Study," Midwest Research Institute,
      Kansas City, Missouri, Report on EPA Contract CPA 22-69-104, May,
      1971.

 22.   "Source Control of Air Emissions," H. F. Elkin, R. A. Constable,
      Hydrocarbon Processing, 51, 113, October, 1972.

 23.   "Fluorine Emissions from Wet Process Phosphoric Acid Plant Process
      Water Ponds," W. R. King and J. K. Ferrell, 68th Annual APCA meet-
      ing, Boston, Massachusetts, paper 75-25.7, June, 1975.

 24.   "Air Pollution Control Technology and Costs in Seven Selected
      Areas," IGCI, Stamford, Connecticut, Report to EPA on Contract 68-02-
      0208, December, 1973.

 25.   "Odors and Air Pollution from the Treatment of Municipal Waste
     Water," G. P. Sutton, 64th Annual Meeting of the Air Pollution
      Control Assn., Atlantic City, N. J., June, 1971.

 26.   "An Evaluation of Charging and Tapping Emissions for the Basic
     Oxygen Process," R. P. Mattis, 68th Annual APCA Meeting, Boston,
     Massachusetts, Paper 75-75.1, June, 1975.

 27.  "Control of H2S Emissions During Slag Quenching," F. H. Rehmus,
     D. P. Manka, E. A.  Upton, JAPCA 23_, 864, October, 1973.

28.  "A Systems Analysis Study of the Integrated Iron and Steel Indus-
     try," H.  W.  Lownie, J. Varga, Battelle Memorial Institute, Colum-
     bus,  Ohio, Report to NAPCA on Contract PH-22-68-65, May, 1969.

29.  "Exposure to Coal Tar Pitch Volatiles at Coke Ovens," N. Fannick,
     L. T. Gonskor, J. Shockley Jr., AIHA Journal, J33, 461, July, 1972.
                                -96-

-------
                  APPENDIX B

SUMMARY OF TRADE ASSOCIATION/INDUSTRY/POLLUTION
 CONTROL AGENCY CONTACTS ON FUGITIVE EMISSIONS
                      -97-

-------
                   SUMMARY OF TRADE ASSOCIATION/INDUSTRY/POLLUTION CONTROL AGENCY

                                  CONTACTS ON FUGITIVE EMISSIONS
          Organization   Contact

  1.   National  Crushed    Mr. Rick Renninger
      Stone Assn.         V.P. Operations
  2.   American  Iron  &
      Steel  Inst.

  3.   Aluminum  Assn.
Mr. Ed Lally
Dr. William Balgord
Manager Env. Services
  4.   Asphalt  Institute   Mr. V. Piuzinauskas
                         Chem Engineer
  5.  American  Foundry-
     men's Assoc.


  6.  Kennecott Copper
     Corp.


  7.  Bethlehem Steel
     Co.

  8.  Rubber Manufac-
     turers


  9.  EPA Region III
     (Philadelphia,
     Pa.)

10.  Arizona Div. of
     Air Pollution
     Control

11.  Texas Air Control
     Board - El Paso
llr. Hulsen
Dr. Fred Templeton
Mr. Tom Keller


Mr. Frank T. Ryan
Director, Env. Health
Safety Affairs

Mr. Lou Fallison
Mr. Mead Sterling
Mr. Jim Shoults
12.  Pa. Dept. of Env.   Mr. Feigenbaum
     Resources Alle-
     gheny Cty.
13.  Bay Area Air Pol-   Mr. Warren Grouse
     lution Control
     District-Californ-
     ia
14.  Los Angeles Air     Mr. William Krenz
     Pollution Control
     District-Californ-
     ia
Results

Contributed reports, reprints, EPA proposal,
all on fugitive dust in the stone industry.
Discussed subject in person at office.

Referred TRC to AISI environmental committee.
Contributed coke oven emission article, steel
making flowsheets on office visit.

Requested letter to environmental committee.
Suggested NIOSH data as being useful to us.
Visited office for discussion.

Discussed report on hydrocarbon emissions done
for them by Esso.  Visit was made and data was
discussed before report issued.  We will get
copy.

Discussed problems of cupola charging and pour-
ing.  Suggested other contacts.  Also dis-
cussed emissions from mold making and pouring.

Will contribute their data on fugitive emis-
sions at Utah smelter.  Would like meeting
with TRC/EPA.

Discussed Roof Monitor Sampling Methods.
TRC may make presentations to environmental
committee.  Other contacts suggested who may
supply data.
Suggested other contacts.  Thinks dust blowing
off open hearth roofs is fugitive.  EOF Tap
and charge emissions not well controlled.
Smelters have 10-25% of S02 as fugitive emis-
sion.  Furnace leakage a problem.  Storage
pile fugitive dust.

Fugitive S02 at smelters 5-10% of total.
Storage piles a problem — need to build
structures over them.
Most coke oven emissions are  fugitive.  They
have no data, but air quality near coke plants
known.

Uses API data as hydrocarbon  emission guide-
lines.  No new data known.
                         Uses API data as hydrocarbon  emission guide-
                         lines.  No new data known.
                                               -98-

-------
                  APPENDIX C

MATRIX CHARTS USED TO IDENTIFY AND CHARACTERIZE
           FUGITIVE EMISSION SOURCES
                      -99-

-------
             KEY TO PROCESS STEP HEADINGS ON MATRIX CHARTS


  List  process or industry, whichever provides best description.  For each

  process  step (described below), enter symbol (from attached list) de-

  scribing character of emissions.


  1.  Material Delivery - Emissions from transfer of raw material from
     transport  mode to plant storage area (e.g., carbon black delivery
     by railroad car  to tire plant).

  2.  Material Storage - Raw material emission from process storage areas
      (e.g.,  sand blown from cement plant pile, evaporation of hydrocarbons
     from oil storage tanks).

  3.  Material Transfer - Emissions in getting materials from storage to
     process use point (e.g., coal dust from conveyors in coal washing
     plant).

  4.  Mixing  and Grinding - Emission from putting ingredients together.  Also
     sizing  and classifying emissions in this category (e.g., mixing of lime
     and  silica for cement manufacturing).

  5.  Ingredient Preparation - Emissions from special steps needed to make a
     prime ingredient of the process (e.g., drying of stone for use in
     asphalt paving,  mold making for iron casting).

  6.  Reactor Charging - Emissions during opening of reactor to add raw
     materials  (e.g., blast furnace loading with iron, limestone and coke).

  7.  Process Leakage  - Leaks of main process system where emissions are involved
     (e.g.,  coke oven door seal leakage).

  8.  Enclosure  Leakage - Open building emissions not treated by pollution
     control air handling systems (e.g., dust blown out of windows or
     doors in foundry due to insufficient hooding).

  9.  Reactor Discharge - Emissions when product is removed from main reactor
     (e.g.,  tapping of a blast furnace).

10.  Final Product Preparation - Any emissions from size reduction or
     final forming needed to make usable product (e.g., cement clinker grinding
     to a powder)

11.  Solvent Evaporation - Emissions from uncontrolled solvent removal
     operations  (e.g., drying of ink in printing).

12.  Ground Dust - Dust  emissions not primarily from the process but made
     airborne by wind or traffic near or in the plant (e.g., dust from
     truck deliveries on dirt roads).
                                 -100-

-------
13.  Waste Transfer - Emissions from transfer of waste from process  to
     storage point.  This includes waste captured by all forms of pollution
     control equipment  (e.g., conveying of slag to slag pit from blast
     furnace).

14.  Waste Storage - Emission from waste in a disposal area (e.g., vapors
     from wastewater settling pond).

15.  Product Transfer  - Emission  during transfer of product from process
     line  to storage area or transport  (e.g., conveying of grain to
     storage silos).

16.  Product Storage - Emissions  from product during storage before delivery
     to user  (e.g.,  vapors  from creosoted  telephone pole piles).

17.  Product Packaging - Emissions  in going  from process or storage to
     shipping  or final retailing or wholesaling containers  (e.g., fertilizer
     bagging).

18.  Mining and Blasting - Emissions from excavation,  material breaking
      and explosive blasting within mining type  operations  (e.g.,  dust plumes
      from blasting).
                                   -101-

-------
                 EXPLANATION OF SYMBOLS USED ON MATRIX CHARTS
                        TO DESCRIBE EMISSIONS CHARACTER
Symbol                       Meaning                       Particle Size

  M           Mining                                            >2y
  m           Mining Dust                                       <2y
  S           Stockpile Dust                                    >2y
  s           Stockpile Dust                                    <2y
  T           Truckload and Road Dust*                          >2y
  t           Truckload and Road Dust*                          <2y
  C           Crushing and Grinding Dust                         >2y
  c           Crushing and Grinding Dust                         <2y
  D           Transfer Material Dust                            >2y
  d           Transfer Material Dust                            <2y
  0           Organic Particulate (Solid)                        >2y
  o           Organic Particulate (Solid)                        <2y
  P           Metallic Particulate                              >2y
  p           Metallic Particulate                              <2y
  W           Aqueous Mist                                      >2y
  w           Aqueous Mist                                      <2y
  L           Organic Particulate (Liquid)                       >2y
  1           Organic Particulate (Liquid)                       <2y
  G           Odoriferous Gas
  g           Gas
  V           Odoriferous Vapor
  v           Vapo r
  R           Reactive Gas Emission
  r           Reactive Vapor Emission
  F           Agricultural Field Dust                            >2y
  f           Agricultural Field Dust                            <2y
  B           Smoke from Burning
  E           Dust  from Pollution Control Equipment              >2y
  e           Dust  from Pollution Control Equipment              <2y
  Y           Fume  or Process Dust from Reactors                 >2y
  y           Fume  or Process Dust from Reactors                 <2y
  K           Livestock Generated Soil Erosion                   >2y
  k           Livestock Generated Soil Erosion                   <2y
  M           Mining & Blasting Dust                            >2y
  m           Mining & Blasting Dust                            <2y
 * Within Plant Confines
                                   -102-

-------
                                                           MATRIX CHART - FUGITIVE EMISSIONS  SOURCES  IDENTIFICATION AND CHARACTERIZATION
 I
H1
O

 I
                                                                                                                                                   COMMENTS ON  COMMENTS ON
                                                                                                                                                    FUGITIVE     CANDIDATE
                                                                                                                                                   EMISSIONS    MEASUREMENT

-------
            APPENDIX D

MEASUREMENT OF FUGITIVE EMISSIONS
 LITERATURE REFERENCES TO METHODS
              -104-

-------
 1.   "Fugitive Dust Measures by  the Rotary Aabient Adhesive Impac-

                            G" HarmS "                      '
 2-
 3.  "A Microscopal  Classification of Settled Particles Found in the
     ™nl^ ^V Coke-^king Unit," R. A. Herrick and L. G. Benedict,
     JAPCA 19 (5), 325, May,  1969.

 4.  "Device for the Continuous Determination of the Dust Flow in Flow-
     ing Gases," J.  Bosch,  Staub  32_, 8, November, 1972.

 5.  "Roof Monitor Emissions  - Test Methodology," T. E. Kreichelt and
     T. G. Keller, JAPCA  22_,  640, August, 1972.

 6.  "The Collection and  Analysis of Inorganic Dust Downwind of Source
     Effluents, "M.  L.  Feldstein, B. Potter, A. E. Alcocer, and H. Moore,
     APCA 61st Annual Meeting, St. Paul, Minnesota, June, 1968.

 7.  "Automation of  the Conventional Gravimetric Dust Measuring Method
     for Quasi Continuous Following of the Courses of Processes," D.
     Eickelpasch, Proc. Second Intl. Clean Air Congress, Washington,
     D. C. , December, 1970, p. 446, Academic Press, New York (1971).

 8.  "The Use of Multiple Tracers in Determining Source Contributions
     from a Smelter," S.  L. Kiel, R. E. Stephenson, L. D. Smoot, and
     D. H. Barker, AIChE  78th National Meeting, Salt Lake City, August,
     1974.

 9.  "A Monitoring System for the Detection and Control of Airborne
     Dust," F. A. Renninger,  Dust Topics Magazine, October, 1966.

10.  "A Comparative  Study of  Particulate Loading in Plumes Using Multi-
     ple Sampling Devices," B. R. Meland, JAPCA, 1J3 (8), 589, August,
     1968.

11.  "Personal and High-Volume Air Sampling Correlation Particulates,"
     P. M. Duvall and R.  C. Bourke, Environmental Science and Technology
     jj"(8), 765, August,  1974.

12   "Development of Methods  for  Sampling and Analysis of Particulate
     and Gaseous Fluoride from Stationary Sources,1  A. D. Little, Inc.,
     Report to EPA on Contract R2-72-126.

13   "Field Evaluation  of the High Volume Particle Fractionating Cas-
     cade Impactor," R. M.  Burton, J. N. Howard  R. L. Penly, P. A. Ram-
     say, T. A. Clark,  JAPCA  23,  277, April, 1973.
                                  -105-

-------
 15.   "Comparison of  'Light Scattering Diameter1 Based on Forward Scat-
      tering Measurements and Aerodynamic Diameter of Aerosol Particles,"
      C.  J. P. van Buitjenen, F. Oeseburn, Atmospheric Environment J3,
      855,  1974.

 16.   "Developments in Sampling and Analysis Instrumentation for Station-
      ary Sources," J. S. Nader, JAPCA 23, 587, July, 1973.

 17.   "Identification and Quantitative Analysis of Particulate Air Con-
      taminants by X-Ray Diffraction Spectrometry," P. 0. Warner, L.
      Sand, and J. 0. Jackson, JAPCA 22_, 887, November, 1972.

 18.   "A  Portable Battery-Operated Immediate-Readout Dust Particle Analy-
      zer," W. Withstandley, W. J. Moroz, G. W. Anderson, JAPCA 21_, 571,
      September, 1971.

 19.   "Size Distribution Measurement of Airborne Coal Dust by Optical
      Particle Counters," B. Y. H. Lin, V. A. Marple, K. T. Whitley,
      N.  J. Barsie, Am. Ind. Hyg. Assn. Jour. 35, 443, August, 1974.

 20.   "A  New Approach to Roof Monitor Particulate Sampling," A. Sinka,
      R.  Marek, L. Gnan, JAPCA, 25, 397, April, 1975.

 21.   "Lidar Effort," E. W. Burgess, Lt. W. M. Pekny, R. E. Meyers, Des-
      eret  Test Center, Fort Douglas, Utah, July 21, 1971.

 22.   "Lidar-Traces Atmospheric Diffusion Measurement System," Ross, Des-
      eret  Test Center, Fort Douglas, Utah, August, 1971.

 23.   "Laser Radar Technology," Lt. W. M. Pekney, Deseret Test Center,
      Fort  Douglas, Utah, October, 1971.

 24.   "Composition and Concentration Measurements of Atmospheric Pollu-
      tants by Remote Probing (Lidar)," Lt. W. M. Pekney, December, 1971.

 25.   "Lidar Study of the Keystone Stack Plume," W. B. Johnson, E. E.
      Uthe, Stanford Research Institute, Menlo Park, California, April 17,
      1971.

 26.   "Lidar," R. T. Collis, Applied Optics, August, 1970.

 27.   "Remote Measurement of Smoke Plume Transmittance Using Lidar," C.
      S. Cook, G. W. Bethke, W. D. Conner, Applied Optics, August, 1972.

 28.   "Long-Path Monitoring of Atmospheric Carbon Monoxide with a Tunable
     Diode Laser System," R. T. Ku, E. D. Hinkley, J. S. Sample, Lin-
     coln Laboratory, Massachusetts Institute of Technology, Lexington,
     Massachusetts, March, 1974.

29.  "Bistatic Monitoring of Gaseous Pollutants with Tunable Semiconduc-
     tor Lasers," E. D. Hinkely, Lincoln Laboratory, Massachusetts  Insti-
     tute of Technology, Lexington, Massachusetts, March, 1974.
                                 -106-

-------
            APPENDIX E




SAMPLE CRITIQUE OF SAMPLING METHODS
               -107-

-------
     Two  important parameters which are of prime interest in particulate

emission  measurements are particle size and material identification.

Both are  obtainable with a reasonable degree of accuracy under certain

test conditions.  Another parameter not readily obtained, however, is

the instantaneous particulate emission rate from a fugitive source,

Some progress has been made toward this goal in present techniques.

     The  methods of particulate fugitive emission measurement available

involve seven basic techniques:
     1.  Standard filter impaction, collection, and weighing.
     2.  Piezoelectric mass detection.
     3.  Light scattering (Nephelometry of Lidar).
     4.  Radiation scattering (B gauges).
     5.  Electrostatic precipitation with controlled voltage.
     6.  Size selective impaction and weighing.
     7.  Adhesive impaction.
     Filter Impaction.  This technique collects most suspended particu-

late matter with the exception of fine aerosols and particles (< 0.1 ym)
                                                                      1
unless special filters are used which collect fine particles well at

lower sample flow rates.  The total particulate can be weighed with some

degree of accuracy and analysis by instrumental or chemical techniques

can readily be done.  Some size distributions can be determined by use

of a cyclone before the filter or two different porosity filters in a

series.  The filter catch can be examined under microscopes (optical and

electronic) for determination of particle size.  As a routine measure,

however, this would be prohibitively expensive.  Another problem of this

technique is that it also gives only the average particulate concentra-

tion over the sampling period.  If fugitive emission rates are  time de-

pendent, a lot of detail of the emission characteristics can be lost.
                                -108-

-------
     Piezoelectric.  This  technique accurately determines total parti-




culate loadings of samples in  very short time periods (< 1 second).  It




is, therefore, essentially on  online  continuous particulate matter moni-




tor.  There are two  problems in  that  size distribution and material an-




alysis are not possible.   These  are major disadvantages, since these are




the critical  criteria  for  fugitive particulate emissions.  Some size dis-




crimination could be done  by suitable air sample prefiltration/separation,




but this has  limited applicability unless multiple instruments are used.




     Light Scattering.   Use of a Nephelometer allows the direct measure-




ment of total particulate  concentration and  also the size distribution




in  a continuous online monitoring system.  There would be no question




that it met two important  criteria except that these conclusions must




be  qualified. Experience  has  shown  that this instrument must be precal-




ibrated with  the  particle  size and  type which is to be examined.  There




are serious problems unless this is  done.  Also, dilution is required




for measurements  of concentrated particulate bearing gas streams, which




introduces  a  possible  error.   The use of the Nephelometer is a question




of  how much pretesting can be  tolerated  to get good field data.  Also,




a material  analysis is not obtained  by this  technique and indeed must be




obtained before measurements can be  made with the Nephelometer.




     Another  light scattering based  system is Lidar, which remotely




scans a sector of the  atmosphere, and the amount of scattered beam de-




tects the  particulate  concentration in that  sector.  The analysis of




dust from  fugitive sources can be made instantly, and entire areas can




be  "particulate  mapped" over short  time periods.  Correlations with




other methods have not been well-proven at  the  present,  and  commercial
                               -109-

-------
 availability of equipment in in doubt.




     Radiation scattering.  The primary technique is to measure scatter-




 ing of  3  radiation  from a radioactive source by collected participates.




 Since only small amounts of sample need to be collected, data acquisi-




 tion is semi-continueus.  Two different sample collection techniques




 have been employed.  One uses a moving paper tape upon which the sample




 impacts and is then read with the g-scatter.  The other uses a condensa-




 tion chamber (water droplets trap the particulate) to concentrate the




 particulate on a glass surface, which is then read with the 3-scatter.




 Both are  readily adaptable for automatic sampling/readout.  Hand port-




 able instruments can be used so that areas of fugitive emission could




 be mapped by an operator walking through a specified sector.  The method




 does require calibration with dust of the same type as being measured




 and also  calibration of the individual instruments themselves.




     Electrostatic  precipitator.  This is basically a small-scale ver-




 sion of an electrostatic precipitator and can be used to collect a re-




 presentative sample on a suitable medium.  Particle size and mass weight




 can be determined as easily as with a filter.  Efficiency for fine par-




 ticulate  is better  than a filter; a semi-continuous version is avail-




 able for  submicron  particle sizing.  A change in collector rod voltage




 is used to discriminate size.  The techniques seem ideal for inorganic




 dust and  aerosols,  but no data on organics such as tars has been found.




These techniques may be applicable to specific situations where size




distribution of inorganic particulate needs to be measured.




     Size selective  impaction.   A cascade impactor such as the Andersen




Sampler can be  used.  It is reproducible and has been correlated to
                               -110-

-------
other techniques  and  also with known standards.  It offers a reasonable

range of size  detection  for studies of hazardous particulates.  The

problem is that it  requires a large sample  (plates are weighed) for ac-

curate results and  that  this is  only an average over the sample time

period.  It  is, however, one of  the easiest size discriminators to use

under field  test  conditions.

     Adhesive  impaction.  For size detection with low sample concentra-

tions and/or short  sampling times, these  are ideal.  They can be acti-

vated whenever important sampling is needed and many samples can be

taken during a day.  The problem is the associated laboratory time spent

counting particles  is large, and the analysis  still is an average of

the  time interval (although short, results  are not instantaneous).  It

would probably be advantageous  to have  some available for special tests

of localized fugitive emissions.

     The most  viable combination depends  upon  the strategy needed for

measurement  of the fugitive emissions.   The best  recommendations are as

follows:


      1.  Pseudo duct strategy

         A.   Piezoelectric mass monitor
         B.   Andersen impactor
          C.   EPA train  as backup and confirmation of  mass monitor
               (also used for organic particulates).

      2.   Roof monitor strategy

         A.   Multiple hi-vols for TRC 4" filters
         B.   Multiple Andersen  impactors
          C.   Piezoelectric mass monitor  for special runs

      3.   Upwind-downwind strategy

         A.   Multiple hi-vols or TRC 4"  filters
         B.   Multiple Andersen  impactors
                                 -111-

-------
C.  Piezoelectric mass monitor for special runs
D.  Rotorod Adhesive Impactor for FP tracer tests and
    other special situations
E.  Consider experimental Lidar test
                          -112-

-------
      APPENDIX F




EPA METHOD 5 DATA SHEET
         -113-

-------
Firm Name 	
Plant Location
Test No.
                                               FIELD DATA SHEET
                                       (EPA Method 5 Sampling Procedure)
                                      	  Pump No.
            ^O	  Ambient Temp.
Sampling Location   Ho\dL  Penxviv-»Q.     Bar Press.
Date       1-IQ05	  Probe Id en
Teste*     W.M. ?  c».p.	  Filter Ide
2-3 Probe Diameter VlG>
in.
OH- £>H = \.2
t. No. 40 CP -871 Probe Heater Setting —
nt. No. IO7 Box Temp. Setting 	
min
min
amp
F
Port
i-j
/M
4>
, 1















Point
1
Z
3
H
5
fc









3
4
5;
Time
Min.















5
5
5
5
S
5,
























6J7
8J9 1
.0
Velocity
P
in H20






























^

£>









26
27 .
28 2
9 bo
PSTACK






























|
1
1
1
O
0









•*>
7
2
0
°l
1









O
5
O
0
1
<


























2
33 :
)4 :
15
TSTACK






























°l
t>
d
fe
(t>
t









1
8
7
«
H
5









ko
f















41
42
43 '
w
43
.. ..Initial
tfeter volume
cu. ft.















^
H
4
H
M
4
4








1
1
1
3
8
$
?








\
4
7
0
3
<4
&








4
H
7
7
s
0
z.








1
5
7
5
fe>
P
?








50
\
|
)
1
1
1
\
1








Pump
Vac.
in Hg.















Remarks






F'^?\ .
voiwiwe









-------
        APPENDIX G




CASCADE IMPACTOR DATA SHEET
           -115-

-------
                                                    CASCADE IMPACTOR DATA SHEET
Test Location 	
Test No.     \S
Date        4 - I
Test Time
Operators
Test Point Location    Hot A
Start of Test (Time) 	17.
Initial  Meter Volume (Cu.  Ft.)
Initial  Average Meter Temp (°F)
Filters:
  Collection Paper 1
  Collection Paper 2
  Collection Paper 3
  Collection Paper 4
  Collection Paper 5
  Collection Paper 6
  Collection Paper 7
  Back-up Filter
                                  1OS.QS
                               Filter No.
                               IO3
                               lOl
                                I OO
                                 T '
Velocity Pressure iP Cin H20)       O-M-O
Orifice 4H C
-------
            APPENDIX H




IKOR CONTINUOUS MONITOR DATA SHEET
              -117-

-------
Test Location   HO\A  p€>U.
«. t

Time Constant
(Sec.)
IOO
••
ii
t •
ti
Sensitivity
JO
«•
'•
M

Instantaneous
Mass Flow MS
2.
1
.5
-

Integrated
Mass. Read.
Mi
O
M
to
m-s

Back-Up
Filter
No.
^
•%
»*
II
H

-------
         APPENDIX I




HI-VOL FILTER TEST DATA SHEET
            -119-

-------
HIGH VOLUME FILTER TEST
     DATA SHEET
; EQUIPMENT
IDENTIFICATION

V4V-2.
UM-3

i
i
i
1
i
;
	 I
LOCATION

ooorK
P Y"O(JLutA •
Area-






FILTER
NUMBER

UIS3
...»






TIME
ST*.r.T
i
1
U-.13
•r«






INITIAL
VOLUME FLOW
•40
Bo-U fcosz
C-oto SSV*





*
i
TIME
END

y.*i
V.M





j
i
FINAL
VOLUME FLO™

i4
Mt.







-------
              APPENDIX J




LABORATORY WORKSHEET FOR FIELD SAMPLES
                 -121-

-------
                            LABORATORY WORKSHEET
                                    FOR
                          EPA FUGITIVE EMISSION  STUDY
                              PROJECT 32397-04
Test Location

Test Date  	

Sent By  	
Reported To
              KJ.
               \\) .
    Identification
Test No.
            Sample No
Sample Description
Ana lysis Required
Special Comments
                5

                (p
                         Ikor
                                   -122-
                     vot,
                     lot.
                     iot.

                     Ujt.

                     iot.
                     tot

                                                                  GPA

-------
        APPENDIX K




CONVERTING UNITS OF MEASURE
          -123-

-------
     EPA policy is to express all measurements in metric units.   When




implementing this practice will result in undue cost or lack of clarity,




conversion factors are provided for the non-metric units.  Generally,




this report uses British units of measure.  For conversion to the metric




system, use the following conversions:
       To convert from:
to:
Multiply by;
acfm
°F

ft
ft/sec
gal /me f
gpm
gpm/ft2
gr/scf
in
in H20
Ib moles
Ib moles/hr
Ib moles/min
tons
nm3/hr
°C

m
m/sec
1/m3
1/m
1/min/m2
gm/m3
cm
mm Hg
gm moles
gm moles/min
gm moles/sec
kg
1.70
Subtract
divide
0.305
0.305
0.134
3.79
40.8
2.29
2.54
1.87
454
7.56
7.56
907.2

32, then
by 1.8












                               -124-

-------
TECHNICAL REPORT DATA 1
(Please read Instruction s on the reverse before completing!
1 REPORT NO. 2.
EPA-600/2-76-284
4. TITLE AND SUBTITLE
Development of Procedures for the Measurement of
Fugitive Emissions
7.AUTHOB
-------