United States      Environmental Monitoring    EPA-600/4-83-011
           Environmental Protection  Systems Laboratory       May 1982
           Agency         Research Triangle Park NC 27711
           __
«>EFy\     Technical
           Assistance
           Document: Quality
           Assurance
           Guideline for
           Visible Emission
           Training Programs

-------
                                    EPA-600/4-83-011
  Technical  Assistance  Document:
   Quality Assurance Guideline for
Visible Emission Training  Programs
                         by

                  PEDCo Environmental, Inc.
                   505 South Duke Street
                 Durham, North Carolina 27701
                   Contract No 68-02-3431
                  Work Assignment No. 1 82
                      EPA Officers.

                     Thomas J. Logan
                  Quality Assurance Division
            Environmental Monitoring Systems Laboratory

                         and

                      Kirk E. Foster
               Stationary Source Compliance Division
             Office of Air Quality Planning and Standards
            Environmental Monitoring Systems Laboratory
               Office of Research and Development
               U.S. Environmental Protection Agency
            Research Triangle Park, North Carolina 27711
                       Mav 1 982
                                              Protectl0fl *****

-------
                                 DISCLAIMER
     This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion.  Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial  products constitute endorsement or re-
commendation for use.  This report provides suggestions and guidelines for
person's involved in observer certification and visible emission evaluation
and is not a supplement to EPA Reference Method 9.  Valid observations can
still be made even though all recommendations herein are not completely
followed.

-------
                                 'CONTENTS
Figures                                                                  v
Tables                                                                 vii
Acknowledgment                                                        viii

1.   Introduction                                                        1

2.   Organization, Planning, and Training Announcements                  2

          Organization                                                   2
          Planning                                                       2
          Training announcements                                         2

3.   Classroom Training                                                  4

          Example lecture material                                       5

4.   Training Equipment                              .                    7

          Method 9 design and operating specifications                    7
          Smoke generator                                                8
          Transmissometer                                               13
          Setup, operation, and shutdown procedures                     32
          Storage and maintenance of the smoke generator                45
          Common problems, hazards, and corrective actions              47

5.   Certification Requirements                                         53

          Practice sessions                                             54
          Certification testing                                         57
          Grading and documentation procedures                          69

6.   Quality Assurance - Techniques and Procedures                      77

          Quality assurance audits                                      77
          Quality assurance for classroom training                      79
          Quality assurance for certification procedures                80
          Tracking program quality                                      80

7.   Visible Emissions Training Literature                              87

-------
                            CONTENTS (continued)

                                                                      Page

Appendix A     Sample lectures for VE training program                  89

Appendix B     Performance audit and system audit                      122

Appendix C     Analysis and example calculations of VE training
                 program errors                                        142

-------
                                 FIGURES

Number                                                                Page

   1      Example Schematic Layout of a Simple Generator                9

   2      Smoke Generator Component List                               10

   3      White Smoke Generating Equipment                             12

   4      Black Smoke Generating Equipment                             14

   5      Transmissometer                                              15

   6      Example Response Time Check Form                             17

   7      Calibration Wand.                                            18

   8      Calibration Error Check Form                                 20

   9      Neutral-Density Filter Form                                  21

  10      Smoke Generator Performance Evaluation Data Form             23

  11      Example Light Source                                         25

  12      Typical Photocell Schematic                                  26

  13      Examples of Acceptable and Unacceptable Smoke Stability
            Conditions                                                 31

  14      Recommended Calibration Stamp                                32

  15      Part and Supply Checklist                                    33

  16      Example Generator Operation Procedure                        35

  17      Electronic Panel of Smoke Generator Control Console          36

  18      Procedure for Practice Session                               55

  19      Opacity Reading Trainfng Form                                56

  20      Operator's Smoke Generator Checklist                         59

  21      Sample Certification Test Form                               62

  22      Example Completed Certification Test Form                    64

-------
                           FIGURES (continued)

Number                                                                Page

  23      Beaufort Scale of Wind Force                                 65

  24      Certification Run Identification Stamp                       68

  25      Two Methods for Determining Average Deviation for 25
            Readings                                                   71

  26      Certification Stamp                                          73

  27      Sample VE Program Roster                                     74

  28      Certification Letter                                         76

  29      List of Statistics Useful  in Evaluating VE Training
            Schools                                                    83

  30      Example Control Chart for Tracking Training School
            Performance                                                84

  31      Statistics Checklist Form                                    85

-------
                                  TABLES



Number                                                                Page



  1       Transmissometer Design and Performance Specifications         8



  2       Properties of Candidate Fuels                                48



  3       Common Smoke Generator Malfunctions                          50



  4       Average Deviation Chart                                      70

-------
                               ACKNOWLEDGMENT


     This report was prepared for the U.S. Environmental Protection Agency,
Division of Stationary Source Enforcement, and the Environmental Monitoring
Systems Laboratory, Research Triangle Park, North Carolina, by PEDCo Environ-
mental, Inc., Cincinnati, Ohio.   It was based on technical  work prepared by
several contractors.  Accordingly, appreciation for contributions is given to
Mr. Norman Edmisten, formerly of Del Green Associates, Inc., Mr. Robert Missen,
formerly of Pacific Environmental Services, and Messrs. Thomas Rose and Willie
Lee of Eastern Technical  Associates.

     The project was directed at PEDCo by Mr.  Carl Nelson,  and managed by
Ms. Barbara Blagun.  The principal editor, Ms. Barbara Blagun, would like to
express her appreciation for the invaluable guidance and assistance provided
by Dr. John Richards, formerly of PEDCo Environmental, and  Messrs.  Thomas
Logan and Kirk Foster who served as project officers for the U.S. Environmental
Protection Agency.
                                    vm

-------
                                  SECTION 1
                                INTRODUCTION

     Visual observation of plume opacities is currently the most effective
and economical method of determining compliance with opacity-based air pollu-
tion control regulations.  It is one of the primary enforcement tools used in
the United States to determine compliance with particulate emission regula-
tions.  The legality and credibility of this method have repeatedly been up-
held in court.  Recently, however, industry has begun to exercise their rights
by challenging the techniques used to read plume opacity.   Due to the role of
visible emissions (VE) observations in compliance and enforcement of air
pollution control laws, it is imperative that personnel  conducting these ob-
servations make accurate and defensible readings.
     Field inspectors and observers are required to document their plume read-
ing skills by periodic participation in a rigorous smoke training and certifi-
cation program.  It is therefore essential  that VE observers continue to have
the benefit of high-quality training and testing.  Accordingly, EPA's Divi-
sion of Stationary Source Enforcement (DSSE)  and the Environmental  Monitoring
Systems Laboratory (EMSL) have furnished this document to  individuals respon-
sible for the general conduct of the VE training and certification program.

-------
                                  SECTION 2
             ORGANIZATION, PLANNING, AND TRAINING ANNOUNCEMENTS

2.1  ORGANIZATION
     To ensure a coordinated and consistent program one agency training super-
visor should have overall responsibility for the smoke training and certifi-
cation program.  This person will  likely need the support of at least two
other people, the smoke generator operator and the operator's assistant.  The
smoke generator operator will be responsible for the preparation, maintenance,
calibration, and operation of the generator.  His assistant will  be respon-
sible for documenting, reading, grading papers, and monitoring trainees.  The
roles of training supervisor and smoke generator operator could be combined
if the classes are held only two or three times a year.

2.2  PLANNING
     Planning the year's training and certification program is an essential
aspect of any successful program.   Although the number and frequency of
training sessions will be partly determined by past experiences and demands,
the school should be scheduled at least twice each year to accommodate per-
sons needing semiannual recertification.  Certifying previous graduates
while the smoke school is in operation is more efficient and less costly
than scheduling a separate session.   A summer/winter schedule is  generally less
desirable than a spring/fall schedule because of adverse weather  conditions
as well as the need to prepare inspectors for the heavy spring/summer source
compliance activities.

2.3  TRAINING ANNOUNCEMENTS
     Because VE training is an expensive program, it is important to provide
a high-quality training program and to have the greatest number of students

-------
possible.  All local air pollution control agencies within the State or area
should be sent in advance a copy of the training schedule, a brief discussion
of course content, and a breakdown of the cost per trainee.   In addition,
industry, trade associations, and journals generally publish such information
as a service to their membership.  Personnel  at major industries and problem
sources may be extended a special invitation.   Public service notices may be
submitted to several newspapers and should include a request for preregistra-
tion by interested individuals.  Following these steps gives an indication
of student load and establishes a point of contact for any followup notices.

-------
                                   SECTION 3

                              CLASSROOM TRAINING


     Classroom training program is an essential part of any smoke reading

certification program.  This training, which is most effectively completed

with an intensive 1- or 2-day classroom lecture/discussion session, is

beneficial for the following reasons:

     1.   It increases the visible emission observers'  knowledge and con-
          fidence for the day-to-day field practice and application.

     2.   It reduces training time required to achieve certification.

     3.   It trains the smoke reader to properly record and present
          evidence that will withstand the rigors of litigation, and
          greatly strengthens an Agency's compliance and enforcement
          program.

     4.   It provides a forum for the periodic exchange of technical
          ideas and information.  For example, periodic refresher
          courses provide field personnel with updated information
          and developments, and reinforce good practices and techniques.
          It can also point out poor techniques and questionable short-
          cuts that may have become incorporated into routine procedures
          and operations.

     Many States only require classroom training for initial certification.

Because the fields of opacity reading technology, legal development, and
court decisions develop and change rapidly, however, a full-day refresher
course should be given at least once every 3 years as a criterion for certifi-

cation renewal.  Further, the training supervisor should attend one of the

instructor seminars offered periodically by EPA because these serve as a forum

for discussing and distributing new material, techniques, and training aids.

     To assure quality training and optimum student learning, lecturers should

be selected with care.  They should be experienced, knowledgeable, organized,

and have good visual aids and current hand-out materials.

-------
3.1  EXAMPLE LECTURE MATERIAL
     This section,  along with Appendix  A,  provides  only an  overview  of  the
classroom session.   Separate EPA documents cited  in Section 7  should be re-
ferred to for further,  more detailed guidance concerning lecture  materials.
The set of lectures included in Appendix A provides samples of material  that
should be presented as  part of the classroom lecture/discussion session.  Al-
though these example lectures are not intended to be a  model  for  all  schools
to follow, the lectures do address most of the important topics covered in a
visible emission training program.  Thus,  this material can be expanded to
accommodate changes in  the state-of-the-art and can be  tailored to  needs  and
regulations of a specific agency.
     The following describes a typical, six-lecture classroom training  program:
     1.   Lecture 1 - The student is introduced to  the  history, principles,
          and theory of opacity.
     2.   Lecture 2 - The sources of visible emissions  should be  presented
          by experienced enforcement personnel or an engineer thoroughly
          familiar with source conditions and opacity reading procedures  and
          problems.  Note:  The use of quality 35-mm slides illustrating
          common source and plume conditions is recommended during  this
          particular lecture.
     3.   Lecture 3 - The proper procedures for conducting  field  inspections
          are discussed.
     4.   Lecture 4 "- The influence and impact of meteorology on  air quality
          are described.
     5.   Lecture 5 - The legal aspects of visual emission  and opacity
          measurement should be presented by an attorney familiar with  the
          practices and problems of air pollution control  enforcement.
     6.   Lecture 6 - The actual testing procedures are discussed.   This
          lecture relies upon the foundation built by the previous  five
          lectures.
     Appendix A also includes two example quizzes.   A short quiz  should be
given at the conclusion of the classroom series to  indicate the trainee's
comprehension of the material presented in the lecture  and  to indicate  whether
the key points of the lectures have been sufficiently emphasized.  If problems
arise with specific questions, it indicates that the material  has not been
clearly presented.  It may then be possible to clarify  these points  immediately

-------
or at least to make the appropriate adjustments in subsequent sessions.   This
provides a QA check on training effectiveness.   Two sample quizzes are pre-
sented in Appendix A.   Note that the last few questions  allow a brief critique
of the course.  These questions will alert the instructor to parts of the lec-
ture that need improvement and thus allow the instructor to constantly improve
his presentation.

-------
                                  SECTION 4
                             TRAINING EQUIPMENT

     In 1950, the first smoke generator was constructed and used for VE train-
ing by the Los Angeles County Air Pollution Control District.  The VE train-
ing procedures were first outlined by EPA in the Federal Register on
December 23, 1971.  The most recent standards of performances and specifica-
tions for smoke generators were published in the Federal Register, Volume 39,
No. 219 on November 12, 1974, as a part of "Method 9 - Visual Determination of
the Opacity of Emissions from Stationary Sources" (Appendix 1).
     Most Federal, State, and local air pollution control agencies conduct
VE training and certification courses at least every six months.  This fre-
quency is necessary to maintain opacity reading certification.
     This section pres'ents performance specifications and operating procedures
for smoke generators that, if followed under a QA program, will ensure nation-
wide uniformity and consistency with Method 9 criteria.  An integrated QA
program is particularly important, since VE enforcement procedures are fre-
quently challenged in court and even more importantly, to assure that sources
will correctly be observed in either compliance or violation.
     An integral part of this program involves the design and operation of
the smoke generator in accordance with the requirements of Method 9.  In the
following sections, the design and operation of the smoke generator and its
associated transmissometer are explained, and procedures are given for assuring
adequate performance.

4.1  METHOD 9 DESIGN AND OPERATING SPECIFICATIONS
     Method 9, as published in the Federal Register, contains design and
operating specifications for the smoke generator used in the training and
certifying of observers.  Method 9 was developed by EPA in support of NSPS

-------
promulgations.  Many states now make reference to Method 9 and even more are
moving toward the common use of its operational and design requirements.
4.1.1  Smoke Generator Specifications
     The procedures to follow in checking compliance with the design specifi-
cations are in Section 3.3.2 of Method 9.  The specific items to be checked
are listed in Table 1.  The manufacturer should determine the specifications
of the light sources, photocell  spectral response, angle of view, and angle
of projection.  The generator operator is responsible for checking the cali-
bration error, zero and span drift, and response time.   Each of these items
will be discussed in detail.
       TABLE 1.  TRANSMISSOMETER DESIGN AND PERFORMANCE SPECIFICATIONS
        Parameter
           Performance
Light source

Photocell spectral response

Angle of view
Angle of projection
Calibration error
Zero and span drift
Response time
Incandescent lamp operated at +5% of
nominal rated voltage
Photopic (daylight spectral response
of the human eye)
15 degrees maximum total angle
15 degrees maximum total angle
+3% opacity, maximum
+1% opacity, 30 min
5 s, maximum
4.2  SMOKE GENERATOR
     The design and operation of the smoke generator have evolved since the
mid-1960's.  The basic components of a smoke generator include:
     1.  Black and white smoke generating units
     2.  Fan and stack
     3.  Transmissometer system
     4.  Control panel and strip chart recorder
A schematic layout of an example commercial  smoke generator is shown in Fig-
ure 1,  and a detailed component list is provided in Figure 2.   The smoke gener-
ator can either be a stationary or mobile unit depending on the  training
                                      8

-------







^s
/
Q i \(iy
Xs7
^s,^
onoonents
Vaporizer
Mixing section
Wain blower




S
/
/


>-/


\
\

1


/•









s

6 n1




©














=





























(


/
(.







(



1
i
5 A
1




d


D


©



(j



0
)





&


JL
'T^



D










i
c




^_



t
^




D


1*



>






S.(


fio)

1









r




i
 4.   Stack section
 5.   Combustion chamber
 6.   Storage-compartment
 7.   Storage compartment
 0.   Stack support
 9.   Fenders
10.   Interconnect box to control  panel
11.   Transmissometer system
12.   Remote control  panel  and strip chart recorder
     a.   Strip chart recorder
     b.   Digital  readout
     c.   Transmissometer control
     d.   Fuel  controls
     REMOTE
  CONTROL PANEL
       Q
       q
0
0
0
d.
          Figure  1.   Example  schematic  layout  of a  simple  generator.

-------
                    TRAILER-MOUNTED COMPONENTS
Trailer hitch

Storage compartments
Stack support for transporting
White smoke vaporizer
Black smoke combustion chamber
Ambient air/smoke mixing chamber
Induced draft (I.D.) fan
Lower stack section
Upper stack section
Hinged support flange
Transmissometer system  (light
source and photocell assembly)
Stack fan
Main electrical junction box
Remote fuel line hookup
Flexible fuel lines
Electrical interconnect cables
Solid sheet metal trailer bed
Trailer axle and brake assembly
Fuel storage tanks
Hydraulic system for stack lift
                          CONTROL CONSOLE
Digital opacity meter
Digital or strip chart
opacity recorder
Main power on/off switch
I.D. fan control switch
Stack fan control switch
Fuel pump control switch

Light source on/off switch
Transmissometer span control
Transmissometer zero control
Fuel pump selector switch
Power indicator light

Toluene fuel control valve
Fuel oil control valve
Connects for fuel lines
Amphenol connectors for
electric supply and electronics
Fuel pumps
Bell or buzzer
Speaker system to communicate
with trainees
            Figure 2.   Smoke generator component list.
                                10

-------
 location.  Most  units are mounted on  trailers,  therefore  they  can  be  trans-
 ported to  training  locations,  thereby improving  attendance  and lowering  travel
 expenditures.  The  unit should be stored  indoors  between  training  exercises
 in order to  lessen  weather deterioration  and  to  improve security.
     Some  negative  trade-offs  are associated  with a mobile  unit.   It  is  sub-
 ject to increased wear due to  travel  and  usually  requires additional  refur-
 bishment,  calibration, and adjustment upon arrival at  the training  location.
 Due to these circumstances, the QA  issue  is of  increased  importance.
     The design  and purpose of a smoke generator  provide  controlled black
 and white  smoke  plumes and a means  of accurately measuring  and recording
 plume opacity.   A smoke generator must be able to generate  smoke with an
 opacity range of 0  to 100 percent and have sufficient  accuracy to allow  the
 operator to control and stabilize the opacity of  the smoke.  Most generators
 have two basic equipment components:   1)  the mechanism that actually  produces
 and controls the smoke and is  usually mounted on a trailer  for portability
 and 2) the mechanism that monitors  and records the various  smoke opacities
 produced.   The use of a separate console  for housing the control and  record-
 ing functions has proven highly desirable because it may be moved away from
 the generator.   The console table also provides a working surface for making
 notes, reviewing records, etc.  The console and trailer unit must have inter-
 connecting lines for the fuels and  electronics.
 4.2.1  Operating Principles
     The training, testing, and certification of smoke observers require the
 use of a device  that can produce black and white smoke of any  given opacity.
 The device must be able to achieve and hold opacities in 5 percent increments
 from 0 to  100 percent opacity.  The desired testing opacity must be stabilized
 at ±2 percent for a minimum of 5 seconds.   After the plume is  stabilized for
 2 or 3 seconds,  the bell  or buzzer  is sounded for the trainee  to make a read-
 ing.   Another 2 to 3 seconds  are allowed for the reading.   Stability and con-
 trol  are essential  and must be checked as  part of the QA program.
     White smoke is produced  by dispensing No. 2 fuel  oil  into the propane-
heated vaporization chamber (Figure 3).   After it vaporizes, the vapor is
cooled until  it condenses into a white aerosol cloud.   The opacity of white
smoke varies  in proportion to  the volume of oil  vaporized  and is regulated

                                      11

-------
    -tJ
Components
1.
2.
3.
4.
5.
6.
7.
8.
Transmissometer
Vaporizer
Exhaust manifold
I.D. fan
Fuel oil storage
Pump
Fuel oil
Dilution
control valve
air damper
           Figure 3.  White smoke generating equipment.
                                12

-------
by adjusting the flow of fuel oil with  the  fuel oil needle  valve  located  on
the fuel control panel.
     Black smoke is produced by  the  incomplete  combustion of  toluene  in the
double-wall combustion chamber (Figure  4).  The opacity of  the  black  smoke
plume is regulated by adjusting  the  toluene fuel flow  controlled  by the toluene
needle valve located on the console.
     Controlling smoke opacity is more  of an art than  a science because of the
many parameters involved, such as fuel  quality, temperature,  and  generator
condition.  Most operators soon  learn their equipment  characteristics and
become very skilled in regulating plume opacities.  This underscores  the  im-
portance of conducting the smoke school with an experienced team.

4.3  TRANSMISSOMETER
     The transmissometer is perhaps  the most critical  component of the smoke
generator, the very heart of the system.  Appendix 1 of Method 9  sets forth
the specifications for smoke meters  or  transmissometers (Ref. 3.3.2).  An
acceptable smoke generator must  be in full compliance  with  these  design and
operating specifications.   The basic smoke generator transmissometer  (Figure 5)
includes a light source, a photopic  photocell detector, and a readout device
with a calibrated span of 0 to 100 percent opacity.  The light-to-photocell
path is approximately 4 ft in length, but only  1 ft, the stack portion of its
length, is exposed to smoke.   The remaining 3 ft are continually  flushed with
ambient air from the small  fans  to prevent smoke buildup and  soiling of the
lamp or photocell.   The transmissometer is located in  the 3-in.  diameter
crossarms of the upper section of the 12-in. diameter  stack.  The following
subsections discuss in detail  individual components of the transmissometer
and procedures for checking their performance.
4.3.1  Calibration Procedures
     Before conducting the calibration sequences,  the  operator should re-
fresh his memory by reviewing  Method 9.   The transmissometer  is  calibrated
according to this  method.
     Through field use, the term "calibration"  has been expanded  from its
original  meaning in Method 9.   When a generator operator refers  to "calibrat-
ing"  the transmissometer,  he  usually means that a  check has been made on

                                     13

-------
Components
1.  Transmlssometer
2.  Stack
3.  I.D.  fan
4.  Combustion chamber
5.  Pump
6.  Toluene control valve
7.  Nozzle
8.  Dilution air damper
9.  Toluene storage


     Figure 4.   Black  smoke  generating  equipment.
                           14

-------
 b
b
Components

1.  Light source
2.  Smoke stop
3.  Photocell
4.  Fan
5.  Stack
              Figure 5.   Transmissometer.
                          15

-------
calibration error, zero and span drift, and response time of the system, in

addition to simply establishing stable 0 to 100 percent readings.

     Method 9 specifies that generator transmissometers be calibrated every
6 months or after any modifications or repair to the transmissometer or read-
out device.  Experience has shown that it is strongly recommended that the

calibration also be performed before and after each certification course.
These calibration checks determine whether any significant drift or devia-
tion has occurred during the certification training period.   Since the cali-

bration procedure has been refined to a fairly simple and expedient procedure,

these additional checks are not too time consuming and add significantly to

overall quality assurance.  Keeping in mind the expanded definition of cali-

bration, the following step-by-step procedure should be followed:

     1.   Allow a 30-minute warmup time for the transmissometer and readout
          device before starting calibration sequence.

     2.   Make sure that the light source input voltage is within the +5 per-
          cent nominal rated voltage of the light bulb.  This voltage cannot
          be varied for span purposes.

     3.   Turn light source to OFF position and establish 100 percent span
          reference point on .readout device.

     4.   Turn light source to ON position and establish zero span ref-
          erence point on readout device.

     5.   Repeat steps 2 and 3 until the 0 and 100 percent span points
          are established as specified by Method 9.  (Refer to Appendix 1
          of the Method.)

     6.   Check for drift of either 0 or 100 percent span points.  A maxi-
          mum of 1 percent over 30 minutes is allowed.

     7.   Check the response time of the transmissometer readout device
          with a stopwatch to ensure that it does not exceed 5 seconds.  Re-
          sponse time is determined by switching the light to the OFF
          position and measuring the amount of time until the recorder
          reaches full scale.  A total response time that is too lengthy
          can cause the transmissometer operator to misread the final value
          of the opacity reading.  Record all response time checks on an
          appropriate form (Figure 6).

     8.   Insert three neutral-density (ND) filters into the stack using
          the calibration wand  (Figure 7) to intersect the light beam
          across the  transmissometer cross arm.  Take care to prevent
          erroneous readings from stray  light.


                                      16

-------
               RESPONSE TIME CHECK (Method 9, Section 3.3.2.7)
Checked by:
Date:
Time:
                Check
      Response time,
            sec
                  1
                  2
                  3
                  4
                  5
                                       Time:
Check
1
2
3
4
5
Response time,
sec





               Figure 6.   Example response  time check  form.
                                     17

-------
POSITIONING
  NOTCHES
                                        SPECTRAL
                                        RESPONSE
                                         FILTER
                                         OPEN
20%
                                          50%
                                          75%
                                         SOLID
       Figure 7.  Calibration wand.
                      18

-------
       9.   Take five random and nonconsecutive readings for each filter.
      10.   Record the nominal  values  of the 20,  50,  and 75 percent
           opacity ND filters  on the  appropriate transmissometer cali-
           bration form (Figure 8).   Repeat this step four times,
           record all  values,  and graph the calibration error.
      11.   If  any reading  exceeds the +3 percent calibration value for that
           specific filter,  take corrective action  before completing the
           certification test.   Check the electrical  and readout systems.
      12.   Have the completed  calibration error  forms signed by the
           individual  conducting these tests.
      13.   File all  signed documents  (Figures  6  and  8)  with the VE certifica-
           tion course records  for the period  in question.   This provides  the
           QA  element  in certifying observer skills  necessary to support
           enforcement litigation.
      To ensure a  properly calibrated transmissometer,  and  thus prevent possible
 legal  challenges  of calibration accuracy,  it  is strongly recommended  that  only
 glass  metallic ND filters that are National Bureau  of  Standards (NBS)-tracea-
 ble should be  used.
      The two most common  types  of ND filters  are composed  of gelatin  and glass
 metallic materials.   Gelatin filters are fairly inexpensive but may deteriorate
 rapidly.  They inherently have  lower quality  and greater variation  than glass
 metallic filters.   Gelatin ND  filters  frequently do  not  meet the  Method 9  re-
 quirements of  +2  percent  accuracy unless .calibrated  by  the  more sophisticated
 spectrophotometers  that generally are  not  available.  The  glass metallic ND
 filters, as ordered from  the manufacturer,  have an NBS  calibration  curve
 accurate to +0.5  percent  of their nominal  rated values.  Calibration  curves
 must  be provided  with each filter to ensure the exact calibration  is  at 540
 nm on  the electromagnetic spectrum.  This  information should be recorded on
 the ND filter  form  shown  in Figure 9.
     Due to the critical  nature of the  photocell alignment,  no  part of the
 transmissometer system should be  removed from the stack  assembly  during cali-
 bration.  A few degrees of misalignment  can shift readings  as  much  as 5 to 10
 percent.  Inserting the ND filters into  the center of the stack at  the trans-
missometer crossarm eliminates  any need  for removing either  the light source
or photocell  assembly in order  to conduct  the calibration.  A  calibration unit
 is now available  from generator manufacturers or can specifically be  built for

                                     19

-------
    CALIBRATION ERROR CHECK  (Method 9, Section 3.3.2.5)
Check performed by: Date:

Filter
1
2
3
Time:

Reading
1

2

3

4

5

Average
opacity, %

Maximum
error, %

Average
reading, %

    100


     8CU
     60J
<_>     ^
<
Q_
O
a:
UJ


-------
Filters calibrated by:
Date calibrated:
Were NBS-traceable filters used?  yes
no
Filter
1
2
3
Nominal
opacity,
20
50
75
Measured
opacity,
01
10



                 Figure  9.   Neutral-density filter form.
                                   21

-------
retrofit on older model smoke generators.  Figure 7 illustrates a typical
filter holder and calibration wand.

4.3.2  Recording Instrument Span

     The next step in the calibration process is to set the span of the  trans-

missometer readout device and to measure any drift.   Since the transmissometer

and recorder are a combination of electronic and mechanical devices, their

resultant outputs will display some drift, which is defined as the amount of

deviation in percent opacity per unit of time from a given setting.

     The following procedure may be followed on a transmissometer and recorder

system to check zero and span drift.

     1.   Warm up the smoke generator/transmissometer system for at least
          30 minutes, but do not start generating smoke.  This step is needed
          to allow the system to stabilize sufficiently before the span
          checks.

     2.   Turn the light switch to the OFF position.

     3.   Adjust the output readings (digital and chart recorder) to read
          100 percent on the chart scale.  This adjustment will be on the
          readout device and will usually be labeled "zero adjustment."

     4.   Turn the light switch to the ON position.

     5.   Adjust the recorder to read 0 percent on the chart scale.  This
          adjustment knob is located on the control  panel  of newer model
          generators.

     6.   Repeat steps 2 through 5 until a stable response is obtained at
          0 and 100 percent on the recording instrument scale.

     7.   Perform the calibration error test.

     8.   Operate the generator for approximately 1 hour.   Repeat steps 2
          through 5.   The difference between the readings at the start
          and end of the hour indicates the resulting instrument drift.

     9.   Record the difference in chart scale for the zero and span on
          the smoke generator performance data sheet (Figure 10).

     This total  procedure, which is commonly referred to as a "zero and span"

check,  must be repeated before and after each test run.   If the drift exceeds

1  percent opacity after a typical 30-minutes test run, the instrument must be

corrected to 0 and 100 percent of scale before testing is resumed.  The drift
readings  should be recorded in order to develop a historical  record of

                                     22

-------
Generator number:
Operator:  	
 Generator Evaluation  Data  Form
	  Manufacturer:
                  Owner:
Zero drift, % chart:
Rise time, sec:  	
                  Span drift,  % chart:
                  Fall time,  sec:   	
Light source lamp voltage, volts:
                       Angle of view (9) calculation
     9 = 2 tan ": d/2L = 	  (Note:  9 must not exceed 15°)
     where:  L = 	mm
             d =        mm
Tested by:
Verified by:

20 percent
% chart
% chart
% chart
% chart
	 % chart

Calibration Error
50 percent
% chart
% chart
% chart
% chart
	 % chart
Date:

75 percent
% chart
% chart
% chart
% chart
	 % chart

y: Date:
       Figure  10.  Smoke generator performance evaluation data form.
                                      23

-------
instrument operation.   As a QA measure, this operation log should be thoroughly
reviewed at least every 6 months,  and it should identify any instrument
deterioration or failure so that preventive maintenance can be initiated.
4.3.3  Light Source
     The smoke generator transmissometer light source in Figure 11 must emit
light in a specified visible light spectrum range between 400 and 700 nm.
Incandescent lamps are normally satisfactory if operated at +5 percent of
their nominal  rated voltage.   If an incandescent lamp is operated at a differ-
ent voltage, though, it will  probably emit additional  light outside the
specified visible spectrum and thereby introduce significant calibration
error.   A typical light source voltage is  determined  by the manufacturer and
is either listed on the bulb or is available in the manufacturer's catalog.
The operating voltage  should be measured with an accurate volt/ohmmeter at
the base connection of the bulb and verified by the operator.   Method 9
requires that all lamps be operated within +5 percent of their nominal  rated
voltage.  Therefore, a lamp rated  at 12 V  should be operated only between
11.4 and 12.6 V.
4.3.4  Photocell Spectral  Response
     The most common photocell found in smoke generator transmissometer
applications is the selenium photovoltaic-type shown  in Figure 12.  The
photovoltaic series of photocells  generate a voltage  or current signal  pro-
portional to the intensity of light detected by the cell.   This voltage is
the signal received by the recording and readout components of the control
console.
     The transmissometer photocell selected for generator applications must
have a photopic response range of 400 to 700 nm as specified in the Federal
Register, Volume 39, Ho. 219, November 12, 1974.
     The photopic response is generally verified through a careful technical
review of the data supplied by the specific generator manufacturer.  This in-
cludes a check to ensure that a tungsten bulb is being used and that the photo-
cell has the appropriate filters necessary to provide a photopic spectral
response output.  If the manufacturer's data are insufficient to make an
accurate assessment, however, the transmissometer unit should be modified to
ensure photopic spectral response.
                                     24

-------
                                                 REFLECTOR
                                          INCANDESCENT
                                              BULB
AMPHENOL
CONNECTOR
ALIGNMENT
  SCREWS
                                                            END CAP
           Figure  11.   Example light source.
                            25

-------
 ACTIVE
CELL AREA
   REMOVABLE
  SUPPORT PLATE
                                                                               CUIFUT
    FRONT  VIEW
REAR  VIEW
                                                      ~|  C~
                                                           -I
                                      234567
                                      1-FRONT CASE
                                      2-SUPPORT PLATE
                                      3-OPTICAL GLASS
                                      4-PhOTOPIC FILTER
                                      S-OPTICAL GLASS
                                      6-SELENIUM ELEMENT
                                      7-INSULATOR
                                      8-8ACK COVER
                               EXPLODED  VIEW
            Figure  12.   Typical  photocell   schematic,
                                       26

-------
4.3.5  Response Time
     The response  time measures  the  time  lapse  required  for the  recorder to
change its reading from 0  to  100 percent  opacity  on  the  recording  instrument
following an  instantaneous  light change in  the  transmissometer.  This  time
interval should be measured by a stopwatch.  Again,  this  determination should
not be attempted until the  instrument  has been  warmed  up  for at  least  30
minutes.  The response time allowed  by Method 9 is five  seconds  or less; an
ideal response time is between 4 and 5 seconds.   Shorter  response  times  can
cause the transmissometer  readout to fluctuate  constantly,  making  it very
difficult to determine precise opacity readings.  All  response time checks
should be recorded on an appropriate form as shown in  Figure 6.
     The response  time is  determined by producing a  series  of five simulated
0 to 100 percent opacity values  by switching the  light source on and off and
observing the time required to move  full  scale.   Note  that  the generator must
not be producing smoke during these  checks.  The  following  steps are suggested
for checking response time:
     1.   Warm up  the transmi ssometer/readout system for  at least  30 minutes.
     2.   Span the instrument as  described  in Subsection  4.3.2.
     3.   Using a stopwatch, measure the  time the recorder  takes to move
          from 0 to 100 percent  on the opacity  readout device.
     4.   Make sure that the recorder reads 0 percent with  the light source
          on.
     5.   Switch the light  off and start  the stopwatch.
     6.   Stop the watch when the recorder  reaches 100 percent opacity
          and record the lapse time.
     7.   Make sure that the recorder stops and stabilizes  at 100  percent
          opacity scale with the  light off.
     8.   Reset the stopwatch, switch light source on, and  start the watch.
          Stop the watch when the recorder  reaches 0 percent  opacity scale,
          and record the lapse time.
         trans!p"ir 'crr.eter system response time adjustment ~.s primarily a  i'u.-sc-
tion of the adjustment of the chart recorder damping.  It is recommended that
either an electronic technician or the manufacturer make these modifications
to the recorder.
                                     27

-------
4.3.6  Angles of Projection  and  View
     The angles of projection and  view refer  to  the  projection  of  light  from
a light source and the view  of this light by  the photocell.   Light from  a  given
source is projected out in nearly  all  directions,  resulting  in  highly  variable
light scattering conditions.  The  ideal  system would have  a  narrow collimated
beam light path.  This optimum condition is approximated by  limiting the angle
of projection to a near-parallel  travel  band  of  light waves.  Similarly, with-
out shielding to restrict the angle of view,  a photocell can respond to  in-
coming light from nearly all angles to its flat  face (approximately 180  de-
grees).  Thus, by limiting the angles  of projection  and viewing to a maximum
of 15 degrees, the light scattering errors are minimized over the  approximately
1-ft stack diameter.
     Limiting the angles of  projection and view  in a generator  transmissometer
serves to restrict the photocell's response to those light waves passing
through the smoke plume approximately  perpendicular  to the long axis of  the
plume.  This results in the  most accurate measurement of plume  opacity.
4.3.7  Estimation of Angle of View
     The construction geometry should  be checked thoroughly  to  ensure  that the
total angle of view of the smoke plume,  as seen  by the photocell,  does not
exceed 15 degrees.  The limiting aperture is  the point in  the path between the
photocell and the smoke plume where the angle of view is most restricted.   In
smoke generator transmissometers,  this is normally an orifice plate.   Since
correct angle of view usually is established  by  the  manufacturer of the  trans-
missometer, the manufacturer's specifications should be checked prior  to per-
forming the measurements and calculations illustrated in this subsection.   The
total angle of view (V) should then be calculated by use of  the following
equation:
                              V =  2  tan"1 |j-
where
          d = sum of photocell diameter and diameter of limiting aperture, mm
          L = distance from photocell  to limiting aperture,  mm
                                     28

-------
4.3.8  Estimation of Angle of Projection
     As is the case with angle of view determination, the general construction
geometry should be checked to ensure that the total angle of projection of
the lamp on the smoke generator does not exceed 15 degrees.  The total angle
can be calculated by use of an equation similar to that for angle of view
determination (subsection 4.3.7) except for a slight difference in the defi-
nition of the parameters.
     The total angle of projection (P) should be calculated by use of the
following equation:
                              P = 2 tan"1 ||-
where
          d = sum of length of lamp filament and diameter of limiting
              aperture, mm
          L = distance from lamp to limiting aperture, mm
     The correct angle of projection is established by the manufacture of the
transmissometer but should be rechecked by use of the equation in this subsec-
tion.  Consistent units must be used in the calculations for both angle of
view and angle of projection.
4.3.9  Stability of Smoke
     Method 9 does not specify a time interval during which the generated smoke
should remain stable or at a reasonably consistent value.  A trained observer
should observe the plume momentarily at 15-seconds intervals when making field
observations, but trainees may need 2 or 3 seconds to make such determinations.
The generator must therefore be capable of holding any set opacity value to
+2 percent for 5 seconds in order to minimize the possibility of incorrect
observations.
     EPA Region IV, for example,^has selected a 2-second observation period
for the reader to make the opacity determination.   The generator operator
signals the beginning and the end of each observation period.  Further,
Region IV has stipulated two criteria for determining whether the opacity is
adequately stable for making a certification reading.  These criteria are:
     1.   The average opacity during the 2-second period is within +_1.5 per-
          cent of the intended opacity.

                                     29

-------
     2.   The variation of the trace does not exceed  1.0 percent
          deviation of opacity from its average position at any
          time during the 2-second period.
Figure  13 illustrates examples of acceptable and unacceptable smoke  stability
conditions as evaluated by the above criteria.
4.3.10  Documentation and Logging for Quality Assurance
     The opacity reading record on field certification activities should be
clearly recorded and maintained and is best provided  by a continuous strip
chart recording showing the actual transmissometer output tracings.  This
method provides full documentation of the spanning effort and adjustments
during the training and certification exercise, and preserves the precise
opacity reading conditions for quality assurance.   The strip chart also facil-
itates the marking of each period and observation point for any future com-
parisons.
     As indicated in examples e and f of Figure 13, the generator readout indi-
cates opacity conditions above and below the intended readings.   At  this point,
the generator was producing a very stable plume and the control  should have
been adjusted slightly to the desired 40 percent opacity.   The strip chart
recording documents a great deal  about the operation of the smoke generator
and should be examined for acceptability in the QA review.
     All zero, span, mid-range linearity, and calibration  filter readings
should also be recorded on the strip chart or digital  recorder in order to
provide a permanent record that these tests were performed and that  the Method
9 criteria were achieved for each individual training school.   The strip chart
or digital recorder should allow the information required  through the use of
a recommended calibration stamp as shown in Figure 14.
     All of the performance verification procedures described in this section
should therefore be documented in writing and dated.   Some agencies prefer to
use a form such as that shown in  Figure 10 to record this  information,  although
a bound logbook is highly recommended because it best assures a  complete rec-
ord of all events that concern the performance of  the smoke generator,  includ-
ing records  of repair and maintenance work,  spectral  response checks, calibra-
tion checks,  response time checks, etc.   These records are then  added to the
permanent files of the VE school.
                                     30

-------
                   OPACITY  LEVELS ACCEPTABLE FOR CERTIFICATION


J'J






























(
\



A



bi
"

,


1






























j



j















3



D



































^

^
j

;


































t5



j















j



u































\
'


H



b



































^



J|

a. variation of <1.5%
     at 40% opacity
b. variation  of <1.5%    c. variation of <1.5%
   above 40%  opacity        below 40% opacity
                  OPACITY LEVELS UNACCEPTABLE FOR CERTIFICATION


3



j




























-
>i

^
1



n


4
1

i>
K,
^






























O



J















,J



J



-































•4



b





i

j

.

\
^






















*



J















j



J


























<
(







*



'J





























1
|
\>
|


^j

d. variations >1.5%
     at 40% opacity
e. constant opacity
     >1.5% above
     40% o pa-city
f. constant opacity
>1.5% below 40% opacity
                         MARKING OPACITY READING PERIODS


J



b

































,
-1
4
— »
•»


M
«1



'ikj


























j



j












j



J




































4
4M
1

y
i
,r
j

ll^o
^i!
I


i ,
1
Bl
~T"7 ~]
; '








i>



J|

               g. Valid readings
               h. Invalid readings
               Figure 13.   Examples of acceptable and unacceptable
                           smoke stability conditions.
                                       31

-------
                         LOCATION
                         DATE 	 RUN 	TIME
                         1-25            26-50
                         GENERATOR
                         OPERATOR
                         VERIFIED BY
                     Figure 14.  Recommended calibration stamp.

     Proper entries and records can form the data base of an excellent QA
program for the school.  The records can be used for statistical studies to
develop confidence intervals for the training sessions and the individual run
forms can be used to develop confidence limits on individual readers.

4.4  SETUP, OPERATION, AND SHUTDOWN PROCEDURES
     Operating an opacity training and certification program is an expensive
but essential endeavor.  It is therefore important to minimize the time and
effort involved.  This is best done by being prepared and thoroughly familiar
with the activity.
     Since opacity training and certification are not a continuing daily opera-
tion, additional preparation is necessary.   The smoke generator operator should
remove the generator from storage before the training course and run through
an operating sequence.  This dry run usually identifies minor repair, preven-
tive maintenance, and missing parts or inventory required for assuring high
quality and efficient certification runs.   Figure 15 provides a checklist of
parts and supplies.   The following step-by-step operating procedure can be
applied to smoke generators to prevent delays and breakdowns during field
certification training sessions.   The smoke generator operator should practice
these steps until a high degree of proficiency is reached.   (NOTE:  An
asterisk (*) indicates that the step must be performed as specified by Method
9.)
      1.   Assemble and inventory the spare parts and accessories by use of a
          checklist similar to that shown in Figure 15.
                                     32

-------
      GENERATOR CHECK
       SPARE PART CHECK
Toluene (2 tanks)
Kerosene (1 tank)
Funnel
Extension/power cord
Ground fault interrupter
Fuel interconnects (2)
Electrical interconnect
Propane torch
Propane tank (full)
Striker
Tip for vaporizer
Vaporizer
Calibration filters
Calibration staff
Calibration stamp
Digital voltmeter
Flag and staff
Control console
Work table
Chart paper
Pen for recorder
Loud speaker system
Fire extinguisher
Standard operating
  procedures manual
Box fuses 15 A main
Box fuses 10 A blower
Box fuses 3 A fans
Box fuses \h A light
Box fuses 3/4 A amplifier
Torch tip (white smoke)
Torch tip igniter
Tank "0" rings
Valves
Bulbs TS67 (12 V) (5)
1 k linear pot 10 turn (1)
Spare tire (mobile units)
Hydraulic jack
Disconnects (2)
Duct tape
Tool kit
Lubricating equipment
      FIELD FILE CHECK
China marking pens
Filter calibration wand
Calibration log
Roster log
Test forms
Grading acetates
Certification stamp
Stamp pad
Felt tip pens
Extra ballpoint pens
Clipboards
Large rubber bands
             Figure 15.  Part and supply checklist.
                             33

-------
 2.   Connect the mobile generator and move it to the area of operation.
     Be sure to select a trailer location that will  provide trainees
     the optimum viewing conditions in respect to background, wind
     direction, and sun position.

 3.   Chock trailer wheels and unhitch smoke generator from towing
     vehicle.   Remove towing vehicle to avoid interference with
     trainee's view of the smoke stack plume.  Adjust trailer dolly
     jack until front of trailer is slightly higher  than the rear
     but level from side-to-side.

 4.   Check for any damage or evidence of tampering with generator
     since the last operation.

 5.   Check fluid level  in hydraulic pump, remove the stack securing
     bracket,  and partially raise the stack to resume operation.

 6.   Lubricate main blower motor.   This must be done prior to each
     certification school by injecting two drops of  lubricating oil
     into each oil cup.

 7.   Check both the toluene (black smoke) and fuel oil  (white smoke)
     storage tanks to ensure an adequate supply of fuel for the en-
     tire certification course.  The generator will  burn 0.1 gal
     toluene or fuel  oil  per certification run; 10 gal  of each fuel
     is adequate for most training and certification sequences.

 8.   Place the generator's control module with detached console on a
     table or  stand approximately 10 ft from the generator trailer.  The
     control console is normally operated on the left side of the trailer,
     facing away from the trainees.  This position prevents trainees
     from viewing the transmissometer readout system during certifi-
     cation runs.  If the control  console is an integral part of the
     model generator being used, proceed with checking  the operability
     of switches, lights, meters,  etc.

 9.   Make electrical  connections by following steps  10  through 15.
     The electrical connection  procedures that apply to the console
     model of  generator are diagrammed in Figure 16.

10.   Connect one end of the 16-ft electrical interconnect cable to
     the main  junction box at the rear of the smoke  generator trailer,
     and the other end to the electrical control console.  (The inter-
     connect cable connectors are keyed and therefore cannot be con-
     nected incorrectly.)

11.   Attach the console AC power (117 VAC) cord to the  read control
     panels (3 pin amphenol; refer to Figure 17).  Do not plug into
     AC source at this time.

12.   Turn all  switches on electrical control console to the OFF position.
                                34

-------
                  Connect the generator
                    console fuel  lines
                   Connect the generator
                 console electrical  lines
                  Connect the power line
                   Turn on propane tank
                   Light toluene igniter
                     Light white smoke
                      vaporizer torch
                      Crank up stack
                    Turn on main power
                       Zero and span
                      transmissometer
                                           Connect speakers and
                                                microphone
Open toluene valve
  for black smoke
 Open oil  valve
for white smoke
    Figure 16.  Example generator operation procedure.
                            35

-------
000
 6    S    <
o  o  o
1 - AC Input
2 - Main Output
3 * Photocell Light Source
4 - Fuse Light Source
S - Cigtul RtjdouC 12V DC (-)
S - Digttil Readout 12V DC
7 - Full Pv*?» 121 DC (-)
3 - Fuel Pu-?s 12V DC (•)
9 - Fuel Pu»ps 12V DC (•)
                              PANEL  IDENTIFICATION
                                                   PIN  WIRE COLOR   VOLTAGE
                                                          Red
                                                         aUck
                                                         Green
              117 V AC (-)
              117 V AC («)
              Eirth-Ground
                            3-PIN  CONSOLE CONNECTOR
                                                              SYSTOt
                                                                          um COLOR
                                                                                      VOLTAGE
Morn
Stick Fin!
Uln 6lo»
-------
     13.  Close both fuel control valves.  This will generally be  in  the
          clockwise direction.

     14.  The control console normally'requires an extension cord  to  reach
          the AC power source.  The extension cord should be at  least 3
          wire, 14 gauge, and weatherproof.  Plug the control console into
          a 20 A, 117 VAC, 60 Hz electrical circuit.

     15.  Set up and test the PA system by connecting the components
          (microphone, speakers, etc.) in their proper places.

     16.  Identify and connect the fuel line.  Two fuel line interconnect
          assemblies are included with the generator.  The shorter set of
          fuel lines (10 ft) is connected between the fuel control panel
          and the fuel shortage tanks.  The longer fuel line assembly (16
          ft) is connected between the fuel control panel and the  fuel in-
          put connectors on the rear of the trailer.

     17.  Connect the fuel lines to the proper input and output  connectors
          on the rear of the fuel control section of the control console.
          On newer generators, the ends of the fuel lines will be  color
          coded.

     18.  Open the vents on both fuel storage tanks to allow proper fuel
          flow through the fuel pumps and to the generator.

     19.  Connect the long set of lines to the rear of the control console.
          Before connecting these lines to the connectors on the trailer,
          bleed each fuel line separately into a container by briefly
          turning on each fuel pump to ensure that most of the air bubbles
          have been removed from the lines.  Clean fuel may be returned to
          its respective reservoir.

     20.  Tighten all fuel line connectors to prevent leaks in the fuel
          delivery system.

4.4.1  Smoke Generator Electrical Console Operation

     Verify that all  electrical switches and fuel  control valves are  in the
OFF position and then proceed with the following steps.

      1.  Turn main power switch to ON position.

      2.  Turn on stack transmissometer fans.   Check fans visually to
          ensure that they are operating.

      3.  Turn the main blower on momentarily to ensure that it  is
          operational.   Then, turn off the main blower until  you begin
          the actual  smoke generation phase.
                                      37

-------
      4.  Activate the transmissometer and readout system  (digital meter
          or strip chart recorder, or both) by turning the electrical point
          switch to ON position.  Allow the electronic components to warm up
          for at least 30 minutes prior to attempting the calibration phase.
          Unless the drive mechanism of the strip chart recorder is driving
          the pen to its extreme positions do not adjust the controls during
          this warmup period.  Proceed with calibration of the transmissometer
          system.  See Section 4.3.1 for additional details.

     *5.  Establish 0 and 100 percent span adjustment of transmissometer.
          This procedure varies slightly depending on whether the readout
          system is a strip chart or digital recorder device.  Refer to
          your generator's operating manual for detailed instructions on
          the specific readout system on your smoke generator.

4.4.2  Recorder Transmissometer Readout System

     Although procedures will vary depending upon the manufacturer, the fol-
lowing steps can generally be followed.

      1.  Turn light source to OFF position.

      2.  Adjust transmissometer zero control to read 100 percent opacity
          on the chart scale.

      3.  Watch recorder for a few minutes to check for significant instru-
          ment drift.

      4.  Turn light source"to ON position.

      5.  Adjust transmissometer span control on the recorder to read 0
          percent opacity on the chart scale.  Check for any significant
          drift or erratic changes in recorder setting.   NOTE:  The recorder
          readout may be in percent transmittance.   This can be changed by
          reversing the recorder input leads.

      6.  Repeat steps 1 through 5 several times until the instrument
          stabilizes.

4.4.3  Digital Transmissometer Readout System

     The following steps present a general procedure for span adjustment for

digital  transmissometer readout systems.

      1.  Turn light source to OFF position.

      2.  Adjust the control  setting of the digital readout system to read
          100 on the digital  meter.   This number actually represents 0 light
          transmission across the stack transmissometer, or 100 percent
          opacity.
                                     38

-------
      3.   Turn light source to ON position.
      4.   Adjust transmissometer zero control  to read 0 on the digital
          panel  meter.
      5.   Check for drift in both the 0 and 100 percent readout modes.
          Some drift is normal until the readout system has warmed up for
          about 30 minutes.
      6.   Repeat steps  1 through 5 several times to document instrument stability.,
*4.4.4  Response Time Test
     Determine the response time by producing a series of five simulated 0
and 100 percent opacity values and observing the time required for stabili-
zation.  Opacity values of 0 and 100 percent may be simulated by alternately
switching the light source power off and on while the smoke generator is not
producing a plume.  The actual time can be determined by use of a stopwatch.
Optimum response time should be between 4 and 5 seconds, but no greater than
5 seconds.  All  response time checks should be recorded on a form similar to
the one shown in Figure 8.
*4.4.5  Transmissometer Calibration
     Step-by-step procedures for the calibration of VE generators should be
followed in detail as outlined in Section 4.3 of this report.  The calibration
procedures outlined in Method 9 are essential to the proper operation of a
smoke generator and should be thoroughly understood by the generator operator.
4.4.6  Upper Stack Section Elevator
     The following steps should be followed to properly elevate the generator's
upper stack section.
      1.  Remove the stack cradle hold-down bracket.
      2.  Close hydraulic bleed valve at front of hydraulic pump reservoir.
      3.  Pump handle on hydraulic cylinder to raise stack.
      4.  Place a "C" clamp on flanges between stack sections for additional
          stability.  NOTE:  Do not pump handle after stack reaches the up-
          right position.  The stack can be damaged if additional force is
          applied by the hydraulic cylinder.
                                     39

-------
4.4.7  Procedures for Producing Black Smoke Plumes

     Prior to beginning this subsection, complete the procedures outlined in
Subsections 4.4.1 through 4.4.4.   The following steps outline the procedure
for producing black smoke plumes.

      1.   Be sure the stack transmissometer fans are operating properly.

      2.   Turn on the main blower fan.

      3.   Verify that the combustion chamber pilot light control valve and
          the white smoke generator gas control valve are turned off
          (clockwise).

      4.   Open the main control  valve on the 20-Ib propane bottle; this is
          the fuel for the white  smoke  generating unit.

      5.   Remove the heat register plate on the black smoke combustion
          chamber, move pilot control  valve to the open  position, and use a
          small  propane torch to  ignite the pilot flame  inside the combustion
          chamber.

      6.   Replace the heat register cover plate.   The adjustment arm for
          the register louvers  should be located so that it cannot vibrate
          closed.

      7.   Adjust register louvers  for proper air flow.   This will require
          trial- and error-adjustments  until  the best settings are achieved.

      8.   Turn fuel  pump electrical  switch to toluene position as indicated
          on control  panel.

      9.   Open toluene fuel  control  valve slightly to allow fuel lines to
          fill with  toluene.   This can  be observed by watching the toluene
          flow in the input line  connected at the rear of the generator
          trailer.

     10.   A small  flame now should be evident inside the combustion chamber
          and the recorder should  be indicating a low level  smoke plume
          (less  than 20 percent opacity).

     11.   Check  to be sure that the pilot light is burning.

     12.   Check  for  possible fuel  leaks at all  connecting points in fuel
          lines  and  check for kinks  or  blockage in the fuel  lines.

     13.   Turn off pilot light  and determine whether fuel can flow into the
          fuel tray  mounted inside the  combustion chamber.   Before attempting
          this,  make sure pilot light is off.   Do not attempt to light the
          unit if more than 0.25  oz toluene is  in the fuel  tray.
                                     40

-------
     14.  Allow sufficient time for removal of air from  remote  fuel  lines.
     15.  Repeat the preceding steps if the system appears  to be  functional.
          If the unit fails to function, review all steps and details  care-
          fully until the problem is found and eliminated.
     16.  Allow the combustion chamber to warm up for a  reasonable  time  before
          proceeding with 20 percent opacity.  Allow a 10-  to 20-minute  warmup
          before starting the initial run.
     17.  The desired black smoke opacity readings now can  be set by adjust-
          ing the toluene fuel control valve.  Stable readings  can  be  achieved
          after a few minutes practice.  Best results are achieved  with  slow
          but firm changes in the control knob setting.
     18.,  Readings of 80 percent opacity or higher should be held for  only
          brief periods to protect the generator from excessive heat buildup.
     19.  The pilot light should be left on between black smoke runs and un-
          til the generator is shut down for the day.
     20.  Be sure that all toluene is burned out of the  combustion  chamber
          before proceeding to shut down the generator.
4.4.8  Black Smoke Assembly Shutdown Procedures
      1.  Turn fuel pump at control console to OFF position.
      2.  Turn toluene fuel control valve to OFF position.
      3.  Be sure all fuel is burned in the combustion chamber.   This  will
          be indicated by a few minutes of clear stack emissions.
      4.  Turn off main valve on propane bottle if no other runs will  be
          conducted with either black or white smoke.
      5.  Cool the combustion chamber by running the main fan for at least
          10 minutes after fuel  shutdown.
4.4.9  Procedure for Producing White Smoke Plume
     Prior to generating the white smoke,  be sure that all  the toluene is
burned out of the black smoke combustion chamber before  proceeding.  The
main blower should be turned on  and running, and the front of the smoke  genera-
tor trailer should be slightly elevated to prevent any fuel  oil  spillage  from
pooling in the vaporizer cabinet.   The following steps outline the  procedure
for producing white smoke plumes.
                                     41

-------
      1.   Make  sure  the vaporizer burner valve is  in the closed or OFF posi-
          tion  (fully clockwise).

      2.   Open  main  valve on propane bottle.

      3.   Open  lid on vaporizer assembly cabinet.   Check for any fuel oil
          or liquid  in the cabinet chamber.

      4.   Ignite the vaporizer burner flame  with a small propane torch.

      5.   Adjust the vaporizer burner valve  until  there is a smooth blue
          flame at the input to the fuel vaporization burner chamber.

      6.   Allow the  vaporization chamber to  warm up for at least 5 minutes;
          a little longer time is needed in  extremely cold weather.  The
          intake throat of the vaporization  chamber should appear slightly
          red before any No. 2 fuel oil  is  injected.

      7.   Slightly open No.  2 fuel  oil  valve on console fuel control panel.
          A low opacity white smoke plume now will be indicated on the
          console recorder and it will  be visible  from the stack.

      8.   If no white smoke is visible,  shut off the fuel  oil  valve and check
          that  the vaporizer burner is  working.

      9.   Slowly open the fuel oil  control  valve one more time.  White smoke
          should be  indicated and visible.

     10.   If no smoke is produced,  shut off  the vaporizer burner and allow
          the unit to cool down.

     11.   Check for  blockage or leaks in fuel oil  lines.

     12.   Check to ensure that fuel oil  will  flow into vaporizing chamber.

     13.   If all checks are positive and no  excess oil is in the vaporizer,
          repeat the startup steps.  If the  unit still does not generate,
          shut  the unit down and repeat each step  until the problem is found
          and corrected.

     When the unit is ignited and operating,  the white smoke vaporizer assembly

should produce  any desired opacity.  Be sure to always guard against flooding

of the fuel oil vaporization chamber.  Do not open the fuel oil control  valve
beyond the initial point when 100 percent opacity is registered on the console

opacity recorder.

4.4.10  White Smoke  Assembly Shutdown Procedure

     The following steps outline the proper shutdown procedure for the white
smoke assembly.

                                     42

-------
     1.    Turn off the fuel pump at control console.

     2.    Turn off fuel  oil control valve.

     3.    Allow all fuel to vaporize from the vaporization chamber.  This
          will require only 2 or 3 minutes and will be indicated by a 0 per-
          cent opacity reading on the recorder.

     4.    Turn off vaporization chamber propane torch.

     5.    Turn off valve on propane bottle if no other runs are to be con-
          ducted with either white or black smoke or if the generator will
          be unattended over a lunch break, etc.

     6.    Run main blower for 10 minutes to cool vaporizer heat chamber.

     7.    At the completion of the certification runs, repeat the calibration
          procedures and record calibration data on the applicable forms.

4.4.11  Generator Shutdown

     When the training session is over, the generator must be shutdown and

secured  for storage.  The following steps outline the generator shutdown

procedure.

     1.    Ensure that both fuel control valves are completely off  (turned
          clockwise).

     2.    Be sure that all propane control valves are fully off (starting
          with the main bottle valve, the white smoke torch valve, and the
          toluene igniter valve).

     3.    Disconnect fuel lines at trailer and insert ends into their
          respective fuel storage tanks.  If proper connect procedures were
          followed, the blue connector should go into the toluene tank and
          the green connector should go into the kerosene or fuel  oil storage
          tank.

     4.    Disconnect fuel supply lines (the short set) from fuel storage
          tanks and elevate them to avoid unnecessary spilling of fuel.
          Turn on the fuel pumps to drain the lines of fuel.

     5.    Remove fuel lines from console disconnects.  Elevate one end to
          fully ensure fuel drainage and cap the ends for storage.

     6.    Cap the male connectors on console.

     7.    Turn off the PA system.  Disconnect the PA system components and
          store properly.
                                     43

-------
      8.  Turn off all electrical control switches, starting with the induced
          draft fan (when the generator has cooled sufficiently), then the
          recorder power control switches, and finally the main power switch.

      9.  Disconnect the generator from the AC power source.

     10.  Remove the electrical system interconnects from the trailer junction
          box and the rear of the console control panel.   Coil and store the
          cables in a dry and secure facility.

4.4.12  Procedures for Lowering Upper Stack Section

     The following procedure should be followed to properly lower the upper
stack section.

      1.  Open cradle hold-down bracket.

      2.  Partially open hydraulic bleed valve.

      3.  Slowly tip the stack so it can work with the hydraulic system.

      4.  The stack can be lowered by inserting a large screwdriver between
          upper and lower stack flange sections and applying leverage.

      5.  Be careful  not to  let the stack get out of control and drop too
          suddenly.   This can be controlled by careful  adjustment of the
          bleeder valve.

      6.  Close stack cradle and secure in place for travel.  Bleeder valve
          should be left open to prevent damage to stack  caused by accidental
          activation of the  pump.

4.4.13  Transporting Smoke Generator

     The smoke generator is  designed to be transported to various training
sites to reduce student travel  time and cost.   The following items should be
checked prior to transporting the generator.

      1.  Bolt down stack in the cradle support.

      2.  Protect fuel  storage tanks by padding the inside portion of their
          storage compartments.

      3.  Check trailer tail  lights and turn  signals  for  proper operation.

      4.  Ensure that the tow vehicle is  adequate and  has the proper hitch
          assembly.

      5.  Check tires  for proper inflation and be sure  that the lug nuts  are
          secure.
                                     44

-------
      6.   Be sure that the trailer and two vehicles have sufficient braking
          capabilities to ensure safe transport since the smoke generator
          is heavy.
      7.   Transport the control  console inside the towing vehicle—not on
          the trailer.
      8.   Upon arrival at the designated training site, check the trailer
          and generator component system for possible damage, loose bolts,
          etc.

4.5  STORAGE AND MAINTENANCE OF THE SMOKE GENERATOR
     Proper storage and maintenance procedures are essential for smoke genera-
tors since they not only increase the lifetime of the instrument but also
provide better quality assurance.
4.5.1  Storage of Smoke Generator
     The modern smoke generator is a complex and sensitive electronic instru-
ment that requires care and protection in both handling and storage.  The
electronic console must be stored in either a heated facility or environment
when not in actual operation, and must be covered with a plastic sheet when
exposed to rainfall.  The electrical interconnect cable can be stored in a
generator storage compartment as long as it is reasonably dry.  Although the
basic smoke generator trailer unit can be stored outdoors, storing the unit
in a warehouse or garage will increase its lifetime and result in lower main-
tenance requirements and operating costs.  If the trailer is stored outdoors,
the stack transmissometer section must be protected by a waterproof cover.
A tarpaulin covering the entire generator may be used, but the covering must
place no stress on the stack fans or the transmissometer electrical connec-
tion.  Any unit stored in such a manner should be checked frequently to en-
sure the integrity of the protection.
4.5.2  Maintenance Procedures
     Before and after each smoke school, the following routine maintenance
procedures should be performed.
      1.  Check all electrical cables to verify that there are no frayed
          parts, that all contact points are clean, that the connectors
          are not bent, and that the cables have good integrity.
                                     45

-------
 2.   Check all  fuel  lines  to ensure that they are unkinked;  have no
     weathered  or worn spots,  splits,  or leaks;  and that the connectors
     are in good condition.   Since  ferrules  can  be easily lost inside
     the connectors,  verify  that these are present.

 3.   Use liquid soap  or similar leak-check commercial  products to check
     the integrity of the  fittings  to  the propane tank for leaks.

 4.   Check the  propane gas supply by weighing the cylinder.

 5.   Assemble and check the  PA system  to ensure  proper operation.  The
     PA system  should be stored with the console unit.

 6.   Check that the  small  drain hole in the  bottom of  the stack is open
     so that water will not  accumulate in the stack.

 7.   Check the  transmissometer calibration to ensure that the system
     has not aged or  deteriorated and  that the calibration curve is
     still good.

 8.   Check the  recorder pin  assembly to ensure that it is operating
     correctly.   This can  be performed by turning on the transmissometer
     and turning the  signal  attenuator to the right and left to make sure
     that the recorder pin will  travel  up and down the scale.   An alter-
     nate procedure  is to  turn the  transmissometer on  and off to ensure
     that the pin will drive back and  forth  smoothly from one side to
     the other.

 9.   Inspect the fire box  after it  has cooled by inserting a mirror in
     the side.   Make  sure  that the  torch and deflector plate are in
     good condition and that the fire  box has not been damaged in
     transit.

10.   Place a drop of  oil on  each of the bearing  surfaces and in the oil
     drop slot  on either side  of the electric motor.

11.   Verify that all  storage compartments contain the  appropriate items
     and that these  compartments are locked.

12.   Keep a bound logbook  to record all  events that bear on  the performance
     of the smoke generator.   Such  events include records of all  repair
     and maintenance  work, spectral  response checks, calibration checks,
     response time checks, dates of use,  number  of runs completed, best
     estimate of fuel  consumption,  and any other pertinent information.
     This book  should be maintained by the generator operator and checked
     every six  months by the training  supervisor.   Entries should be
     made whenever the generator is serviced, repaired, and/or operated.
                                46

-------
4.6  COMMON PROBLEMS, HAZARDS, AND CORRECTIVE ACTIONS
     The proper operation and maintenance of a smoke generator requires per-
sonnel with a high degree of skill and instrument knowledge.  The training
supervisor is responsible for ensuring that a qualified person is present
and that proper calibrations, checks, and safety procedures are followed.
     The generator should always be in top condition and ready to operate.
To ensure optimunToperation, the preliminary certification checks should be
performed at least one day prior to the scheduled training and certification
runs.  For this preliminary run, the generator operator should review the
applicable provisions of Method 9 and sections of this publication to ensure
that all steps are performed.
     A number of problems can develop that may interfere with the proper opera-
tion of the smoke generator.  Some of the more common problems and solutions
are discussed in this section.
4.6.1  Fuels
     The use of improper fuels can cause several problems.  The following
fuel selection guidelines should be helpful in avoiding these problems.
4.6.1.1  Black Smoke —
     Originally, most smoke generators produced black smoke by the incomplete
combustion of fuel oil.  In the late 1960's, many operators found better re-
sults using benzene as the fuel.  While benzene was fairly successful, it is
a highly toxic substance and hazardous when improperly used.  In June 1977,
EPA issued a memorandum which stated that an adverse health problem may be
associated with the use of benzene and strongly suggested that other fuels
be used instead.  Possible alternatives include toluene, xylene, kerosene,
and No. 2 fuel oil.  Table 2 lists the various properties of these fuels.
Although all of these fuels are suitable, toluene appears to be the most
suitable fuel for producing black smoke.   Care must always be used when handling
any fuel since they are flammable.  Possible adverse health effects are also
associated with prolonged and unnecessary exposure to benzene, toluene, and
xylene.
                                     47

-------
                   TABLE 2.  PROPERTIES OF  CANDIDATE  FUELS
Fuel
Benzene
Toluene







Xylene
Kerosene


No. 2 oil



Composition
C6H6
C7H8







C8H10
Higher par-
affins and
naphthenes
Paraffins,
aroma tics
and nap-
thenes
Freezing
Point,
°C
55
-95







-25/+13
___


—



Boiling
point,
°C
80
111







138-144
79-227


282



Threshold
limit
value,
ppm
10
100







100
a


a



Remarks
Carcinogenic
Requires in-
creased accuracy
in adjustment of
fuel -to-air
ratio compared
with that re-
quired for
benzene
Burns too hot
Gives brown smoke


Clogs valves



 Threshold limit value will be greater than 100 ppm, although exact value
 depends on concentration of each specific compound.

4.6.1.2  White Smoke —
     White smoke is made by vaporizing kerosene or No. 2 fuel oil.  This
usually is done by injecting the fuel oil either onto the hot exhaust from
a self-contained gasoline engine or onto a heated surface, causing the liquid
droplets to vaporize and then condense into a thick white cloud.  Both fuels
appear to perform equally well.   The No. 2 fuel oil tends to clog the valves
and produces slightly less smoke per gallon of fuel than does kerosene.
Variations in the volatility of some kerosenes result in their tendency to
ignite or flash when the vaporizer overheats.   The operator should be aware
of this and be prepared to react accordingly if kerosene is to be burned.
                                     48

-------
4.6.2  Overheating
     Overheating either the combustion chamber or the vaporizer can permanently
distort the unit, resulting in unstable smoke production or reduced opacity
capability.  Overheating is easily avoided by properly fueling the vaporizer
and limiting the durations of high opacities.  Once a vaporizer is severely
damaged, it must be replaced in order to achieve satisfactory service.
4.6.3  Breakdowns
     The new model generators are designed and fabricated for durable field
use and portability.  Like all electromechanical  devices, however,  the genera-
tor will occasionally break down or malfunction.   The problem must  be diagnosed
and repairs made expeditiously to maintain the proper training and  interest  of
the attendees.  The inventory of spare parts should be based on the more common
needs and malfunctions.  Experience over the years shows most failures to be
in the areas listed in Table 3.
4.6.4  Modifications to Improve Performance
     In recent years, the operation of the smoke  generator has been improved
with a number of important modifications in:
     o    design of the black smoke combustion chamber
     o    switch from the exhaust manifold vaporizer to the hot
          plate vaporizer
     o    photopic transmissometer system
These improvements are adaptable to most basic generator units, if  the frame
and support assembly are sturdy and in good condition.  Cost may be optimized
by simply replacing the troublesome component.  Outlined are some modifica-
tions that can be made, however, to improve the operation of some of the
older model generators.
     1.   Place a baffle plate or refractory assembly in the air inlet opening
          of the combustion chamber to provide better air/fuel mixing conditions
          and to stabilize smoke production.  Some experimentation  may be
          needed to determine optimum location for the baffle.
     2.   Reduce fuel line plugging significantly by adding an inline fuel
          filter, which is standard equipment on  the newer model generators.
                                     49

-------
                TABLE  3.   COMMON SMOKE GENERATOR MALFUNCTIONS
       Malfunction
         Cause
      Problem
Power failure
Overloading the circuit
to blow a fuse or trip
the circuit breaker;
this should not occur
because the generator
should not overload
the standard 110
VAC/15 A service
                           Electrical  connection
                           has  been  broken  or
                           disconnected
Locate service panel;
check fuse or circuit
breaker; replace or re-
set as necessary; if
power fails again, the
circuit is being over-
loaded; locate and
use another electrical
service

Reconnect and secure
Loss of signal  to
the recorder
Light source burned
out or broken

Loose wiring
                           12 V power supply
                           failure
Replace bulb
Locate and repair loose
connections; this con-
dition is frequently
intermittent and may
require some movement
of wiring and tracing
and insulation with a
test of a continuity
meter

Connect to a 12 V DC
battery until power
supply pack is re-
placed
Unstable smoke
Wind influences, e.g.
> 15 mph
Check orientation of
trailer in relation to
wind direction; rotate
trailer 30 to 60 degrees.
Poor air/fuel
mixture
Air flow into combus-
tion chamber or dete-
rioration of combus-
tion chamber; wind
direction may also be
a problem
Determine most probable
cause by inspection and
experimentation;  if
combustion chamber is
deteriorated or warped
by heat, replacement
may be required;
attempt to install a
baffle at air inlet by
placing a brick or metal
baffle in entrance
(continued)
                                     50

-------
TABLE 3 (continued)
       Malfunction
         Cause
      Problem
Fuel flow problems
or line blockage
Dirty fuel filters
                           Fuel pump failure
                           Fuel line kinks or
                           plugging
                           Carbonization of
                           burner tips
Replace or install fuel
line filters

Replace fuel pump;
implement a temporary
measure by switching
the fuel lines to the
operable pump;  Note:
Replacing fuel pump
requires shutting off
generator for 30 to 60
minutes  if possible,
replace at lunch break
or end of day

Inspect fuel lines; if
plugged, disconnect ends
and blow out debris

Run a thin stiff wire
through the orifice
                                    51

-------
     3.   Minimize wind shear by increasing the main blower air flow since
          many of the older model generators have inadequate blower
          capacity.  Wind shear is a common problem in areas that fre-
          quently experience wind speeds in excess of 12 mph.  If the
          training session area has a prevailing wind direction, position
          the generator so that the wind does not blow directly into any
          openings in the smoke flow passages.   If the wind shears off the
          plume at the top of the stack, a 90 degree elbow with laminar flow
          vanes may be used to allow the smoke to exit the stack parallel
          to the wind.  This provides the trainees a better opportunity for
          certification by reducing the eddy effect.  The elbow has been
          proven in field use, with some individuals reading successfully
          during winds exceeding 20 mph.
4.6.5  Safety Requirements
     The normal safety requirements for trailer towing and handling apply
to transporting the generator.  Brakes, lights, etc. should be checked each
time the trailer is moved.  Wheel chocks must be used whenever the trailer
is disconnected from a vehicle, and additional  care must be exercised if the
generator is located on an inclined surface.
     Since smoke generation fuels are flammable, proper and sensible
storage and handling precautions must be followed.  Possible health hazards
of handling the fuels should be recognized and the equipment should be
operated only in open ventilated areas.  Proper explosion-proof fuel should
be used and any fuel transfer operations should be conducted with care.   A
fully charged and functional fire extinguisher must always be readily
available whenever the generator is operating.
     Because unit surfaces become hot while the generator is in operation,
persons not involved and familiar with the equipment should remain at a
safe distance.   A pair of protective gloves should be available to handle
any hot items,  and operators should always wear safety glasses when working
around the generator.  Carelessness and shortcuts can result in serious
injuries.
     Some internal components of the console have a 110-V electrical charge
that poses a potential hazard especially during damp field conditions.  A
voltmeter connected to a grounding rod can be used to monitor electrical
shorting conditions.
                                     52

-------
                                  SECTION 5
                         CERTIFICATION REQUIREMENTS

     This section provides procedures for conducting the certification part
of the training program.  To ensure proper quality assurance, the certifica-
tion must be attended by at least two people—one to operate the generator
and the other to instruct the students, monitor student activity, ensure that
the smoke is readable, answer student questions, grade records, etc.   A third
person is also recommended to assist with the above activities as well as to
provide an additional element of quality assurance, especially for classes
with 25 or more trainees.
     Prior to firing up the generator, the instructor should identify the
components, explain the operation of the generator to the trainees, and allow
them to examine the equipment.  Safety requirements should be discussed, and
the trainees should be reminded to stay away from the generator during train-
ing and test runs.  The instructor should also emphasize that the generator
has hot surfaces that can cause serious burns and that there are numerous
electrical cables and fuel lines which, if accidentally disconnected, could
shut down the generator and delay the entire program.
     The first part of the certification test will be a "test run."  Both
black and white smoke will be emitted, and the opacity announced in order to
familiarize the trainees with the procedures and help "calibrate the eye"
to  the announced readings.
     After the initial test run, certification runs will be made in blocks of
50  readings (25 black and 25 white).  After the certification criteria have
been achieved, the trainee must check that the form has been completed cor-
rectly and signed.   It also is recommended that the trainee recheck the
arithmetic on the form to avoid any  errors in transcription or arithmetic
that might preclude  certification.   Each form will then be checked by at
least two other people to further assure that all criteria and requirements
have been satisfied.  Those trainees who successfully meet the criteria will
receive a letter of  certification and a copy of the qualification  form.
                                     53

-------
      The original qualification form is retained  by  the  school  for  at  least 3
years  for  possible presentation in any future  legal  proceedings  or  challenge
of  certification.  Note:  Certification is valid  for a period of only  6  months
according  to Method 9.
     Recertification procedures are identical  to  testing procedures, except
that the lecture series is usually omitted.  Though presently not required,
it  is  recommended that the trainee repeat the  entire lecture portion of  the
school every third year to reemphasize techniques and to familiarize the
trainee with new material  and procedures.

5.1  PRACTICE SESSIONS
     Prior to the actual  certification runs,  the trainee should  be exposed
to a series of practice readings over the  full  range of opacity  levels in
order  to allow an adjustment or "eye calibration" period.  Calibration with
the standards is an  important element of the  6-month recertification
requirement of Method 9.   Trainee calibration can be achieved in several
ways, although experience  has shown that one  of the most effective methods
is to focus on the 25,  50, and 75 percent  opacity levels.
     The practice procedures  consist of the following steps outlined in the
flow chart  diagrammed in  Figure 18.
      1.   Set up the  generator and perform the  zero and span checks.
      2.   Set the chart speed to zero.
      3.   Distribute  practice test forms  (Figure 19)  and  explain their use.
         The first blank  on  the form is for  the trainee's  observed  opacity
         estimate, the second for the  announced transmissometer readings,
         and the third for  the difference in the two values.
      4.  Generate a  25 percent opacity white smoke plume for about  3 minutes.
      5.  Encourage the observers  to  walk  around and  view this  opacity
         level  at different  angles  to  the sun.
      6.  Repeat steps  4 and  5 for 50 and  75  percent  opacity.   Note:  Stan-
         dards  of 20,  40,  60, and 80 percent opacity can also  be used.
      7.  Begin  the practice  certification when the smoke becomes steady;
         indicate the  beginning of  the reading with  the  word "Ready."
                                     54

-------
Zero and span check transmissometer and set chart speed to zero
                   Distribute practice forms
   Generate 25, 50, and 75 percent opacities in white smoke
               Generate four white smoke plumes
                      Read actual values
                   No
                                Yes
   Generate 25, 50, and 75 percent opacities in black smoke
               Generate four black smoke plumes
                      Read actual values
                   No
                                Yes
                     End Practice Session
          Figure  18.  Procedure for practice session.
                             55

-------
en
CTl


Group
Color:
W B


Group
Color:
W B


Group
Color:
W B


Group
Color:




P 1 ume
No.
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Ru
Obs
















n No. 1
SG
















i
Dev
















Rur
Obs
















i No. 2
SG
















i
Dev
















Ru
Obs
















n No. 3
SG
















i
Dev
















Ru
Obs
















i No. 4
SG
















i
Dev
















           stands for  the  observer
       difference between  the  two.
's  opacity  reading,  SG,  the opacity from the smoke generator, and Dev, the


   Figure  19.   Opacity reading training form.

-------
      8.   Wait at least 1 second and say "Mark" to indicate that the observer
          should record his observation, which he estimates at the moment
          "Mark" is announced.
      9.   On a duplicate practice form, record the transmissometer value at
          the moment "Mark" is announced.
     10.   Repeat steps 7 through 9 four times for values around the
          standards, e.g., 25, 50, and 75 percent opacity.
     11.   After four different opacities are generated, read the actual
          values recorded for the transmissometer and have the trainees
          check their answers.  Repeat steps 7 through 11 with smaller
          shifts in opacity until the trainees become proficient at judging
          opacity in increments of 5 percent.
     12.   Shutdown the white smoke generator.
     13.   Start up the black smoke generator.
     14.   Repeat steps 4 through 11 for black smoke plumes.
     15.   Shutdown the black smoke generator.
After the practice sessions, begin the actual certification testing.

5.2  CERTIFICATION TESTING
     The trainees should now be prepared for the certification runs.  They
have been exposed to the theory and background on opacity, proper techniques
of reading, parameters of variables that influence accuracy of readings,
legalities of sound documentation, and plume opacities simulated by the smoke
generator.
5.2.1  Preparing for Certification
     The training supervisor and assistant must pay careful attention to the
details of the certification operatives, and the staff must be able to state
with certainty that all aspects of Method 9 have been fulfilled.  This requires
meticulous care of all records and information and stipulated QA checks of
these procedures, thereby ensuring that any passing trainee is technically
and legally certified.
                                     57

-------
     An important element of the QA program is a checklist that itemizes the
important steps in the certification of opacity readers.  This checklist ful-
fills two important QA needs:
     1.   It assures that an orderly preparation and implementation
          procedure has been completed.
     2.   It provides documentation and evidence in support of
          quality certification.
     Figure 20 presents an example format for the checklist, which should be
marked off or initialed as appropriate after each item is completed.  The
training supervisor should check and sign the sheet at the end of the cer-
tification exercise.  This form then becomes a part of the official file or
record of the course.
5.2.2  Test Forms and Recording Procedures
     Test forms vary greatly with the specific needs and experiences of each
agency.  A commonly used form that is both low in error and easy to grade is
shown in Figure 21.  The test form should be printed on two-copy paper so that
the original can be turned in for the official file and the carbon copy can
be graded by the trainee.   Two test forms printed on regular bond paper with
carbon paper between them is also satisfactory, although this arrangement
is more cumbersome to use and subject to greater error and misunderstanding
in reconciling the original with the carbon copy.  In all cases, the agency
should retain the original test page for the official file and certification
record.
     The trainee is to circle one answer per line that is judged to be the
generator opacity at the indicated signal for reading and may change any
answer by simply Xing out the wrong answer and circling the new choice as
indicated in the following example:
          20       25      (5&    (35)      40       45       50
     It should be noted that the most common error with this form is placing
the answer on the wrong line.  Again, the procedure is to X the incorrect
answer and circle the correct one.
          20       25       ©      35       40       45       50
          20       25      ¥j&?     {&)      40       45       50
                                     58

-------
Date:	  Agency:  .
Operators:	  Location:
Note:  Initial  each item as completed.   Write  NA  (Not  Applicable) where  pro-
       cedure does not apply.
                            Generator Preparation
     Check for damage and vandalism                             	
     Verify that a complete inventory of parts
     and supplies (refer to inventory list)
     exists
     Check inflation of tires
     Check operation of brakes, hitch, safety
     chain, etc.
     Lubricate fans and motors
     Check that stack is secure for traveling
     Perform preliminary check of console
                            Adequate Fuel Supply
     Toluene
     Diesel or kerosene oil
     Propane
     Other (specify kind and use)  	
                               Generator Setup
     Check background, sky, and wind conditions
     for best generator orientation
              Figure 20.  Operator's smoke generator checklist.
 (continued)
                                     59

-------
Figure 20 (continued)

     Check electrical service and availability
     (110 VAC/20 A, 3 wire grounded connection)

     Ensure that generator is leveled and wheels
     are chocked

     Connect fuel and electrical  lines

     Check for fuel leaks (lines, connections,
     and tanks)

     Check fans for smooth,  normal  noise level  operation

          Main or induced draft fan

          Transmissometer fans

     Be sure air volumes appear adequate


                           Operation of Generator

     Be sure generator logbook is available

     Be sure that both fuel  pumps are operable

     Check chart recorder operation and paper  supply

     Check that all fuel valves are in full  OFF position

     Check both combustion systems  for operating
     condition

     Check operation of safety ignitors

     Turn on console for 30-minute  electrical warmup

     Verify that all  readout systems are working


                     Transmissometer System  Calibration

     Zero and span check

     Drift check

     Response time

(continued)
                                     60

-------
Figure 20  (continued)

     Light source voltage

     Calibration error (use NBS  filters)

     Calibration record completed


                             Generation  of  Smoke

     Raise and secure  stack

     Run black smoke stability and  range  test

     Run white smoke stability and  range  test

     Check for excessive leaks and  proper draft conditions


                            Public  Address  System

     Check that Total  Smoke System  is  calibrated and
     operating properly for training and  certification
     purposes

     Comments -- please list any problems or conditions
     encountered in the setup or startup  of the smoke
     generator.
                                     61

-------
AFFILIATION
Course L
Date
Distance
ocation
NAME
RUN
1


Sunglasses
Sky
and
Direction
Wind



to Stack
READING
NUMBER
1
2
3
4
S
6
7
3
9
10
11
12
13
U
15
16
17
18
19
20
21
22
23
24
25

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

5
5
5
5
5
5
5
5
5
S
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10

IS
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15

20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20

25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25

30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30

35
35
35
75
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35

40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40

45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45

50
50
50
50
50
50
50
50
SO
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55

60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60

65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
55
65
65
65
65
65
65

70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70

75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 30 85
75 30 85
75 80 35
75 80 35
75 80 85
75 80 85
75 80 85
75 30 85
75 80 85
75 80 85
75 30 85
DEVIATION
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95

READING
ERROR

100 1
100 2
100 3
100 4
100 5
100 6
100 7
100 8
100 9
100 10
100 11
100 12
100 13
100 14
100 15
100 16
100 17
100 18
100 19
100 20
100 21
100 22
100 23
100 24
100 25

ERROR
NUMBER
26
27
23
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
4 7
43
49
50

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

5
5
5
5
5
5
S
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10

15
15
15
15
15
15
15
15
15
15
IS
15
15
15
15
15
15
15
15
15
15
15
15
15
15

20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20

25
25
25
25
25
25
25
25
25
25
25
25
25
2S
25
25
25
25
25
25
25
25
25
25
25

30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30

35
35
•35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35
35

40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40

45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45
45

50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55

60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
50
60

65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65

70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70

75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
DEVIATION
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95

100 26
100 27
100 28
100 29
100 30
100 31
100 32
100 33
100 34
100 35
100 36
100 ?7
100 38
100 39
100 40
100 41
100 42
100 43
100 44
100 45
100 46
100 47
100 43
100 49
100 50

Figure  21.  Sample certification test form.
                     62

-------
     The form should be completed with one answer per line and all the
pertinent information provided (see Figure 22).  All  entries must be made in
ink.  The observer's name, affiliation, run number, course location, and date
should be recorded; if sunglasses are used, the type and lense color should
be noted.
     The windspeed should be estimated to within a 3 to 5 mph range.  If an
anemometer is not available, the Beaufort wind scale, Figure 23, may be used.
     Smoke school certification should not be attempted in winds above 25 mph
which interfere with high quality, stable smoke simulation.  A personal dis-
comfort factor under these conditions tends to introduce an additional un-
acceptable error.
     The wind direction can be estimated to within eight points of the compass
(N, NE, etc.) by observing which way a flag is blowing, or by observing the
direction a few blades of grass are blown when thrown into the air.  The north
direction can be obtained by referring to a map.
     The sky condition should be filled in as:
     1.   clear - less than 10 percent of the sky covered with clouds
     2.   scattered - 10 to 50 percent of the sky covered
     3.   broken - 50 to 90 percent of the sky covered
     4.   overcast - more than 90 percent of the sky covered
     Certification can be achieved even under total overcast conditions pro-
vided the trainees have a contrasting background available.  Readings should
not be attempted during conditions of precipitation.
     The observer's orientation with respect to both the plume and the sun
should also be indicated on the test form.
5.2.3  The Test
     The trainees should now be familiar with the testing procedures and ready
to  begin actual certification runs.  The test forms should be distributed and
the top  portion filled in by the trainee for identification as described in
Subsection 5.2.2.
     Method 9 stipulates that a valid test must have 25 white and 25 black
smoke plume readings.  The candidate must demonstrate the ability to assign
opacity  readings in 5 percent increments, within the following criteria:
                                     63

-------
AFFILIATION JrM TVf
Course Location K~T
-USqt
*./
P l\
•n/f 1 1
•J
NAME rfa K ri J)ne
Sunglasses
Date .3-.3O- ?3.
Distance
and
Direction
to Stack
READING
rioo co /.
Sky CL /£/)/
.TO
\ r '
f/

/.

Ho
9
•ffJA



( *



rf

RUN t 3

Wind JL /a /
w


Hi- (



'=*/



}

ERROR
NUMBER
1
2
3
4
5
5
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

QLr
\'^
26
27
23
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
i4
45
46
4 7
43
49
30

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

10 ING
1BER
(2)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5 {



1 5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
CD

10
10 '
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
12
s>



10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10

£2)
ClSJ
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15



15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
ii

20

25
JS*
Is
25
25 ,
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25



25
A
dt
25
25
25
25
25
25
25
25
25
25
25
 30
30
dE
30
30
30
30
30
30
30
Jl
 Jfl.

30
30
) 30
©
30
30
30
30
30

35
35
35
35
35
35 <
35
35
35
(j5y
Q5J
Q5J
35 (
35
35
35
35
35
35
35
35
dE>
35
35
35



35
35
35

1 35
35
35
35
35
35
35
35
35
) 35
35
1 35
35
35
35
> 35
35



35

40
40
40
40
40
35")
W
40
40
40
40
40
2D
40
40
40
40
40
40
40
40
40
40
40
40



40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
40
(S3
40
40
40
40

45
45
45
45
45
45
45
45
Q[^
45
45
45
45
45 (
45
45
45
45
45
45
45 '
45
45
45
45



45
45
45
45
45
(3^
45
45
45
45
/3S
Qy
45
45
45
45
45
45
45
45
1 45
45
45
45
45

50
50
50
50
50
50
50
^^
50
50
50
50
ii
GS>
50 '
50
50
50
50
5fl
Cs(Q
50
50
50
50



50
50
50
50
50
50
50
50
50
50
50
50
Q&
50
50
50
50
50
50
50
50
(Sfl)
50
50
50

55
55
55
55
55
55
55
55
55
55
55
55
55
55
Sy
« 5
55
55
55
55
55
55
55
55
55



55
55
55
55
55
55
^P
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55

60
60
60
60
60
60
60
60
60
60
60
60
60
60
60

60
60
60
60
60
60
60
60
60



60
60
60
60
60
60
60
60
60
02^
60
60
60
60
60
60
60
60
60
60
50
60
60
60
60

65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
GiS^
65
65
65
65
65
65
65
65



65
65
65
65
65
65
65
(S^
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65
65

70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70



70
70
70
70
70
70



70
70
70
70
70
70
70
70
OD
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70
70

75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 30 85
75 80 85
75 80 85
75 80 85
75 80 85
75 30 85
75 80 85
75 80 85
75 80 35
75 £0_ 85
75 ©> 85
75 80 35
75 30 35
75 80 85
75 80 85
75 80 85
DEVIATION


75 80 85
75 80 35
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 30 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 85
75 30 85
75 80 85
75 30 85
75 80 85
75 80 85
75 80 85
75 80 85
75 80 35
75 80 85
DEVIATION
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90



90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95



95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95

100 1
100 2
100 3
100 4
100 5
100 6
100 7
100 8
100 9
100 10
100 11
100 12
100 13
100 14
100 15
100 16
100 17
100 18
100 19
100 20
100 21
100 22
100 23
100 24
100 25

ERROR

100 26
100 27
100 28
100 29
100 30
100 31
100 32
100 33
100 34
100 35
100 36
100 37
100 38
100 39
100 40
100 41
100 42
100 43
100 44
100 45
100 46
100 47
100 48
100 49
100 50

Figure 22.  Example completed  certification  test form.
                          64

-------
      Beaufort Scale of Wind Force
 (Compiled by U.S.  Weather Bureau, 1955)
Beau-
fort
number
0
1
2
3
4
5
6
7
8
9
10
11
12 or
more
Miles
per
hour
Less
than 1
1-3
4-7
8-12
13-18
19-24
25-31
32-38
39-46
47-54
55-63
64-72
73 or
more
Knots
Less
than 1
1-3
4-6
7-10
11-16
17-21
22-27
28-33
34-40
41-47
48-55
56-63
64 or
more
Wind effects observed on land
Calm; smoke rises vertically
Direction of wind shown by smoke drift;
but not by wind vanes
Wind felt on face; leaves rustle;
ordinary vane moved by wind
Leaves and small twigs in constant
motion; wind extends light flag
Raises dust, loose paper; small
branches are moved
Small leaves in trees begin to sway;
crested wavelets form on inland
waters
Large branches in motion; whistling
heard in telegraph wires; umbrellas
used with difficulty
Whole trees in motion; inconvenience
felt walking against wind
Breaks twigs off trees; generally
impedes progress
Slight structural damage occurs;
(chimney pots, slates removed)
Seldom experienced inland; trees
uprooted; considerable structural
damage occurs
Very rarely experienced; accompanied
by widespread damage
Very rarely experienced; accompanied
by widespread damage
Terms used
in USWB
forecasts

Light

Gentle
Moderate
Fresh
Strong

Gale
Whole Gale

Hurricane
Figure 23.   Beaufort Scale of Wind Force.
                  65

-------
     1.   No reading may be in error by more than 15 percent
          opacity.
     2.   The average error must not exceed 7.5 percent for
          either set of 25 white or 25 black smoke readings.

Failure to meet either of these criteria is considered to be an unacceptable
demonstration of reading accuracy, and the observer has failed that particular
run of 50 readings.
     The runs can be repeated as many times as needed to meet the criteria
for certification.   A minimum of 10 runs should be made available during a
given certification session.  Generally, students pass certification in three

or less runs and more than 95 percent pass within the 10 runs.  The steps
below are to be followed in the testing session:

      1.  Set up the generator, allow warmup period, and perform the zero
          and span checks.

      2.  Distribute the certification forms, and have the trainees fill
          out the top (Figure 22).

      3.  Open the white smoke valve.

      4.  Generate the first standard opacity smoke plume.

      5.  Turn off all valves.  Wait until emissions are not visible and
          the transmissometer trace is flat.

      6.  Repeat zero and span checks.

      7.  Announce the start of the test, give the run number and time.

      8.  Start the strip chart recorder.

      9.  Open the fuel valve and allow the smoke to stabilize within  1.5
          percent opacity of a 5 percent scale line (Figure 13).

     10.  When the smoke has stabilized, say "Number one" and begin the
          trace on the strip chart.

     11.  After 1 to 2 seconds, say "Mark" if this reading  stayed within the
          1.5 percent opacity limits.  Stop the recorder and number the
          reading.  Restart the strip recorder.   If unacceptable, say
          "Scratch" and repeat the recorder.  Students do not mark their
          papers unless "Mark" is announced.  Mark the unacceptable reading
          "Void" on the chart and initial.
                                     66

-------
     12.   Repeat steps 8 through 10 for the remaining 24 white smoke
          readings.   The opacity of these readings must be selected in a
          random order.
     13.   Shut off the white smoke valve.
     14.   Check the zero and span transmissometer drift.  If the values are
          within 1 percent of 0 and 100, go on to step  15.  Otherwise, wait
          15 seconds and repeat the check.
     15.   Open the black smoke valve and proceed with steps 4 through 11
          for readings 26 to 50.  The opacities must again be selected at
          random.
     16.   Collect the original or top sheet of the two-part form.  After
          all trainee forms have been collected, give the correct readings.
          This allows the trainees to check their own reading so they can
          adjust and calibrate their readings as necessary.
     In general, people frequently tend to read either  slightly higher or
slightly lower than the actual value.  If that is happening, the observer can
make the mental adjustment and at that point can proceed refining and rein-
forcing his skill.  It is suggested that the trainees continue their readings
for all runs even though they may have, certified in the first few runs.  This
additional practice serves to refine the newly developed skills.  This is not
as important during recertification tests, however, especially if the person
is routinely engaged in making official opacity readings.
5.2.4  Ensuring a Valid Certification Run
     The instructor must make sure that the trainees read the plume opacity
at the same time the transmissometer records the opacity.  Many schools rely
on audible signals for informing the trainees when to read the plume.  A
major problem with this is that some trainees tend not  to read instantaneously,
as suggested, but rather give the plume a lingering stare.  This is a common
reason for failure to certify, since the transmissometer reading may have
already started moving to the next opacity reading.  Readings must be made
within a second of notification to mark.
     This potential problem can be corrected easily by  using a public address
system to cue the readers and by modifying the chart recorder so that the
pen gives an instantaneous rather than a continuous trace.  A rubber band
should be wrapped around the tracing bar and anchored to a heavy paper clamp
                                     67

-------
on top of the console.  When the generator has sufficiently stabilized for  the
first reading, the operator should place a finger on the trace bar.  While
pressing down on the trace bar to allow the pen to contact the paper, he should
simultaneously announce "Number one."  The pen should trace for 2 to 3 seconds.
The operator should then release the bar simultaneously and announce "Mark."
The trainees should be instructed to read between the two announcements of
"Number one" and "Mark."  This results in an exact trace of the transmissometer
reading during the precise period the students are to read.  Should the smoke
destabilize and cause the opacity to jump greater than +2 percent, the in-
structor should void that number by announcing "Scratch" and indicate it on
the chart record.
     The generator operator should circle and number each acceptable trace
for full  and accurate identification.  He should indicate with an X any un-
acceptable trace, write VOID adjacent to it,  and initial it to avoid any
confusion on the validity of readings.   The traces on the strip chart recorder
discussed above are illustrated in Figure 13.   The operator should clearly
mark the strip chart at the beginning and end of each run and identify each
run with the date,  time, run number,  and color of smoke.  This is easily
accomplished by use of a rubber stamp similar to the one illustrated in
Figure 24.   The operator should note  and initial  any event that could impact
the validity of a run on the strip chart.   In  addition,  any lengthy or de-
tailed explanation  should be noted in the bound logbook.

                         METHOD 9 CERTIFICATION RUN
                         DATE	
                         RUN #	
                         TYPE
                         CHART SPEED_
                         + STOP TIME_
                         OPERATOR
             Figure 24.   Certification run identification stamp.
                                     68

-------
5.3  GRADING AND DOCUMENTATION PROCEDURES
     If a test form similar to the one shown  in Figure  24  has  been  used,
accurate grading is relatively simple with  the aid of a grading  key.  The
moderator can produce this key by marking the correct value for  each  reading
with a china marking pencil on an acetate copy of the form.  The moderator
should then place the acetate key over each form to be  graded  and identify  the
difference between the correct answers and  the trainee's answers.   With
practice, an error exceeding 15 percent  (three or more  5 percent increments)
is quickly spotted.
     The moderator should count the number of increments of 5  percent error
for both sets of 25 readings and compare the average deviation (D)  for each
run of 25 readings using the following equation:

     fi     (Sum of Positive Deviations) - (Sum of Negative Deviations)
     u '                               25
An alternate system is to total the number of increments of 5  percent error
(both positive and negative) and read the average deviation from the chart
presented in Table 4.   If a chart is unavailable, the total number  of percent
increments on 25 readings can be multiplied by 0.20 to get the average devia-
tion.  These two methods are illustrated in Figure 25.
     Several different grading methods are used.  Some training supervisors
prefer that all grading be done by the course moderators without trainee
participation.  Involving the trainees, however, enhances learning.   The
moderator collects the original sheet of the test form at the  end of each
run.  The trainee retains the second copy for initial  grading.  After all
original copies for a  run have been collected, the transmissometer  opacity
values are announced.   The trainees grade their copy of the run.   The generator
operator can prepare a master copy of the opacity readings as  the values are
announced to be used as the acetate grading key for the official  grading.
     After the readings have been announced, the trainees are given a few
minutes to check their paper and to recognize any needed adjustments in their
reading skills.  The training staff, however, must make the official deter-
mination of certification.
     In sessions involving a large number of students, another run may be
started while one staff member continues grading.   The graded forms must be
                                     69

-------
                     TABLE 4.   AVERAGE DEVIATION  CHART
Increment of
error
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Average
deviation, %
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
2.20
2.40
2.60
2.80
3.00
3.20
3.40
3.60
3.80
4.00
4.20
4.40
4.60
4.80
5.00
Increment of
error
26
27
28
29
30
31
32
33
34
35
36
i- 37
38
39
40
41
42
43
44
45
46
47
48
49
50
Average
deviation, %
5.20
5.40
5.60
5.80
6.00
6.20
6.40
6.60
6.80
7.00
7.20
7.40
7.60
7.80
8.00
8.20
8.40
8.60
8.80
9.00
9.20
9.40
9.60
9.80
10.00
Each error of 5 percent opacity,  either positive or negative,  is  counted as
one increment.
                                    70

-------
Reading
Deviation,
Increment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
+ 5
0
-10
0
- 5
-10
0
0
0
-15
0
+ 5
-10
-10
+ 5
0
0
0
+ 5
0
+10
0
- 5
- 5
- 5
1
0
2
0
1
2
0
0
0
3
0
1
2
2
1
0
0
0
1
0
2
0
1
1
1
                                                Method 1:
                                                =   (Z pos read)  -  (I neg  read)
                                                n =             25
                                 D
                                                Method 2:

                                                D = Number of 5% increments x 0.2

                                                D = 21 x 0.2 = 4.2%
TOTAL
    105
   21
  Figure 25.  Two methods for determining average deviation for 25 readings.
                                      71

-------
crosschecked for accuracy before a trainee is officially notified of certifi-

cation.  A copy is sent to the individual along with a letter of certification
and preferably a wallet card stating the period of certification.  Details of
this technique are listed in the following step-by-step procedures.

      1.   Complete an official run of 25 white and 25 black readings.

      2.   Collect the original (top white) copy of each trainee's certi-
          fication form (Figure 22).   Place them in the field file and
          retain in the possession of a training staff member.

      3.   Take a blank certification form, remove the carbon copy, and
          place the test form on a clipboard.

      4.   Write "Master" in the space for the observer's name.   Fill  in the
          run number, date, location, and time of the test run.

      5.   Correctly record the values on the  form as they are read from the
          strip chart.   Be sure to ensure accuracy.   Make five  random
          rechecks with the chart record.  Allow the trainees to grade
          their (carbon) copy as this is done.

      6.   Place a grading acetate over the master,  watching alignment
          carefully.

      7.   Using a marking pencil  make a diagonal  slash from upper left to
          lower right (\) over each marked number on the grading master.
          File the acetate with the originals.

      8.   Turn the grading master over and record the names of  the trainees
          who think they have qualified.   Theirs  are the first  papers to be
          graded.

      9.   Remove those  papers from the file.   Select one and align the
          acetate over  it.

     10.   Count from  the mark on the  acetate  to  the  mark on the  trainee's
          paper.   Each  5 percent is counted as  one.   For example:

                   (3)        (2)        (1)
          30        35        40        45       50           Error = 3
                  Value                       Correct
                                               value

     11.   Add all  of  the errors for the 25  white  smoke readings  and record
          the sum on  the line marked  "Deviation."

     12.   If  no single  error  is more  than 3 and  the  total  number is 37 or
          less, repeat  steps  9 and 10 for the  25  black smoke readings.
                                    72

-------
     13.  If no single error  among the black smoke readings  is more than 3
         and if the total  number is 37 or less, stamp the  paper  "Quali-
         fied" (Figure 26).
     14.  Sign the line labeled  "Graded by," and have the  supervisor verify
         the grade.
     15.  Enter the trainee's name, address, and qualifying  run number in
         the VE program roster  in a bound logbook (Figure  27).
     16.  Compute the average deviation for each color of  smoke and record
         this value on the test form under "Deviation" in  the roster.
     17.  Inform the trainees who have qualified.
     18.  Send a certification letter and wallet card to each trainee who
         qualified.
                          UALIFIE
                      Graded fay_
                              M
                      Verified by.
                    Figure 26.   Certification stamp.
5.3.1  Maintenance of Records
     The  top sheet of the test form collected from each trainee  at  the end
of each run is part of the documentation of an individual's certification as
a qualified observer.  A training staff member must collect these before
announcing the correct values, and these forms must remain in  staff custody
thereafter.  This control procedure prevents cheating or manipulation.
     A bound logbook should be maintained  for recording all events  that might
affect the performance of the smoke generator.  This logbook should include
records of repairs, maintenance work,  spectral response checks,  calibration
checks, response time checks, etc.   In another logbook, records  should be
kept of the number of attendees receiving  training; the number of trainees
certifying; and their name, address,  scores and average deviation,  dates of
training, etc.  The original of each individual's certifying run, the
                                    73

-------
3
CO
                       o
                       -5
                       GO

                      CO
                       3
                       ft)
                       r-t-

                       -5
                       fD
                               CD
                               -S
                               CU
                               Q.
                               00

                              IO
       fD
 CO
 O)
 3
T3

 fD
T3
 -S
 O
CQ

 Q)
 3

 -s
 o
 CO
 r+
 tt>
 -5
                       CU
       CU
                              O
                              CU
                              n>


























































































































































z
CU
rt>
CU
3
Q.
J»
Q-
Q.
-s
fD
in
co
Fee paid
Training
Recertifying
QR Number
Date
Dev. White
Dev. Black
Bias White
Bias Black
Certification
CD
fD
3
fD
-5
CU
rfr
O
-5

=tfe
CU
r+
ID
                                                                                                                                                      O
                                                                                                                                                      O
                                                                                                                                                                                     O
                                                                                                                                                                                     n
fD     c-t-
3     -"•
n-     o
CO     3

-------
checklist, the recorder strip charts,  the VE program roster,  and any other
pertinent information should be maintained in the agency's official  file.
     This information may be needed for presentation at legal  proceedings
as evidence that the inspector or person in question has been  certified as a
qualified VE evaluator by a recognized smoke training and certification group.
     These files should be arranged by training session and maintained for
at least 5 years to be available for use in any future legal  proceedings that
may occur.
5.3.2  Certification Letters
     Within 2 weeks of the training session, each trainee who  successfully
meets the Method 9 criteria should be mailed a letter of certification or
verification and a copy of their qualification form.  An example letter is
shown in Figure 28.
     Some agencies provide wall certificates and/or wallet cards to each
successful participant.  At least one of these should be provided and it
should contain the following information:
     1.   Participant's name
     2.   State where accomplishment took place
     3.   Date of certification
     4.   Date of expiration or statement that certification expires 6 months
          from date of certification
     5.   Location of course
     6.   Signature of course moderator or other selected official
Certificates should be numbered in sequence and a record maintained in order
to account for each certificate.
                                     75

-------
                 ENVIRONMENTAL PROTECTION AGENCY
          RESEARCH TRIANGLE PARK, NORTH CAROLINA   27711
                                     August  12,  1981
Mr. John Doe
U.S. EPA
Research Triangle Park, North Carolina  27711

Dear Mr. Doe:

Please be advised that you successfully completed  our  recent
Visible Emissions Evaluation course.  Having attended  the
lectures (March 8, 1982) and participated  in the smoke evalua-
tion sessions, you met the following certification criteria:

     1.   The average deviation for the sets of 25 black
          and 25 white smoke emissions was  less than  7.5%.

     2.   The deviation of each reading was 15% or less.

This certification is valid until September 7,  1982.

                                     Sincerely yours,
                                     John  L.  Forrest
                                     Physical Science Technician
                                     State Air Quality Training
                                       Division
                 Figure 28.  Certification letter.
                               76

-------
                                  SECTION 6
                QUALITY ASSURANCE - TECHNIQUES AND PROCEDURES

     Several recent court decisions have favored industrial efforts to resist,
question, and discredit the application of opacity readings in enforcement/
compliance documentation.  Problems can be most readily avoided through a
structured and consistently applied QA program that includes the elements
stressed throughout the previous sections.  Although this section will
describe in concise detail a recommended program, it is not intended to re-
present the ultimate or ideal program.  Each school should review its opera-
tions and needs, and then design an integrated QA program accordingly.
     A QA program, in concept and implementation, is a management tool or
operational mechanism for assuring credible training and certification.  To
ensure success, this program must receive the commitment, support, and follow-
up of management.   This additional effort will help to readily identify smoke
school weaknesses  and shortcomings that must be resolved.  If properly
implemented, this  program will provide the extra element of documentation and
credibility that will withstand the investigation and scrutiny of special
interest challenges.

6.1  QUALITY ASSURANCE AUDITS
     The QA program consists of two distinct program operations that are com-
plementary and mutually supportive.  One program is operated integrally with
the presentation of the smoke school  and focuses on the operation of the
smoke generator.  This program provides the extra effort needed to assure
routinely accurate and reliable training and certification.  Such an effort
includes the planning of the program; preparation, calibration, and operation
of the generator;  recordkeeping and documentation; and trainee control.  This
program has been suggested and emphasized throughout the previous sections.
                                     77

-------
     Another  important element of QA is auditing.  A VE program audit  is  an
external review of the program by use of structured evaluation forms.  The
audit is usually conducted by supervisory personnel to assure objectivity.
Audits are designed to identify program weaknesses and deficiencies that  must
be addressed  and corrected in order to maintain an effective and high-quality
opacity training and certification program.
     A performance audit is a structured and routine review of the various
steps in the  training and certification process.  It generally consists of a
number of specified checks to ensure that proper and accurate procedural  ac-
tivities have been completed.  The performance audit is conducted by the
training supervisor and/or designated staff, and is most applicable to fairly
specific and  routine procedures.   Appendix B includes example formats for in-
spection and operation of the smoke generator/transmissometer unit and the
handling of certification test records.  Although not addressed in this manual,
additional performance audit formats could be developed for other specific
procedures important to overall  program quality assurance.   In this process,
the staff would systematically review the content and operation of the program
as outlined in this manual.   In  any audit situation,  checklists should be de-
vised for each specific activity.
     Records pertinent to the audit must.be completed and maintained as part
of the agency's official  documentation file.   The generator operating log,
smoke school file,  and other records are also components  of the documentation
and recordkeeping process.   The  importance of the performance audit as an
integral  part of the training activity cannot be overemphasized because it is
vital to training and certification efforts.   This  audit  provides  assurance
that the requirements of  Method  9  have been fulfilled and that a sound and
high-quality program has  been provided.
     Complementary  to the performance audit is  the  system audit.   The system
audit is conducted  periodically,  perhaps on an  annual  basis,  through an on-
site comprehensive  review of the  total  smoke school  program.   This  audit should
be conducted to assess all  aspects  of the training  activity and should readily
identify problems and weaknesses  that should and can  be corrected.   The system
audit preferably should be  conducted by upper management  such as the enforce-
ment program manager and/or  EPA  Regional  Office personnel.   Participation by
the enforcement program manager  is  preferable because opacity certification

                                     78

-------
directly impacts and supports the day-to-day enforcement/compliance operation.
This is particularly important since more than 90 percent of participate
emissions compliance is determined and documented by opacity readings.
Participation by enforcement management also provides a needed element of
management involvement and concern.   The smoke school is typically a service
provided by the Technical Support Division.   It also demonstrates interest by
the Enforcement Division in requiring the most credible operation possible.
     A system audit should have a positive impact.  Management attention, con-
cern, and opacity training and certification should increase with increased
court challenges and litigation.   The system audit provides a mechanism for
management to gain needed program familiarization.  It can also be used as a
planning document to assist agencies in obtaining the support to update
training materials and equipment.
     The system audit is a review of the entire training program and can be
used to assist in a systematic check.  An itemized audit form similar to the
one included in Appendix B should be followed.   The checklist approach assures
that all aspects of the program are evaluated.   It also lends itself to the
expedient preparation of an audit report.  The form can be attached to a brief
narrative that summarizes the findings, conclusions, and recommendations.  It
also provides the mechanism for followup on  the important recommendations to
assure that proper corrective measures have  been taken.

6.2  QUALITY ASSURANCE FOR CLASSROOM TRAINING
     Quality assurance activities for the classroom portion of a training and
certification program ensure that adequate facilities are available and that
the lectures adequately cover important subject matter.   A portion of the
example system audit checklist included in Appendix B addresses basic require-
ments of classroom training.  In addition to the system audit, the classroom
examination can be used to evaluate lecture  content and delivery.  Evaluation
questions can also be used to assess the adequacy of the classroom facilities
(see example quiz in Appendix A).
                                     79

-------
 6.3  QUALITY ASSURANCE FOR CERTIFICATION PROCEDURES
     Because field certification activities are time consuming and  expensive,
 this part of the training process must be completed accurately and  expedi-
 tiously.  Field activities must also be well documented.  Recordkeeping
 functions for certification are amenable to a performance audit and are
 addressed in Appendix B, whereas the general QA review of the field certifi-
 cation program can be effected through a system audit.
     Other factors that are important to the field certification program can
 be addressed as part of this system audit.   These include the selection of an
 appropriate site, instructions to participants, operation of the generator,
 certification of generator outputs, and grading and documentation of reader
 results.  Each of these components is addressed in the system audit checklist
 included in Appendix B.
     The system audit checklist also presents additional factors to be con-
 sidered in planning and performing the overall  certification testing.   Over-
 emphasis of one area of certification,  such as  the field test, may result in
 understaffing of classroom or QA efforts.   In addressing the criteria necessary
 to produce certified readers,  the system audit  checklist provides a method
 of assuring that all  components are adequately  evaluated.

 6.4  TRACKING PROGRAM QUALITY
     A principal  objective of  a QA program  for  VE  training is to ensure that
observers meet established performance  standards.   In  addition,  Reference
Method 9 or other more stringent standards  prescribe certain minimum levels
of equipment quality and  performance;  therefore it is  imperative that a
mechanism exist to  ensure compliance with  these standards.  Thus,  a QA program
must be able to track both qualitatively and quantitatively  the  overall pro-
gram performance as well  as  the performance of  individual  program elements.
6.4.1  Analysis of  Opacity Error
     Some factors related to  program performance are easily  quantified because
they can be directly measured,  e.g., transmissometer accuracy and an observer's
mean deviation.   Other factors  rely upon somewhat  more subjective means for
quantification, such as  the  use of examinations to determine the effectiveness
                                     80

-------
of classroom training.  In either case, the techniques used to make such quan-
tification must maintain consistency since the performance tracking provided
by a QA program relies upon comparable standards.
     To a great extent, overall program performance can be judged by the
opacity reading abilities of the trainees.  Since the reading ability is
really the combined result of the observer's ability, his classroom train-
ing, and the accuracy of the smoke generation equipment, however, potential
and measured errors associated with each factor must be estimated.  Appendix
C details the procedure for calculating the overall bias and variability due
to equipment and operational  procedures.  The general procedure requires the
summation of variances or errors associated with each step in determining an
opacity value and comparing that value and its associated error with the value
observed by the attendees.   Accordingly, error values must be determined for:
     1.   Standard opacity (calibration) filters
     2.   Transmissometer calibration
     3.   Reporting of opacity value (rounding-off)
     The reported opacity value, which, incorporates these errors, is then
compared with the opacity values reported by the attendees.   The bias  and
variance associated with differences in these two values describe the  combined
error of the attendee.
     The error analysis outlined in Appendix C establishes the basis for
determining the adequacy of an important element in a VE training school, i.e.,
to what accuracy can certified trainees read opacity.  Secondarily, but of
equal importance, the analysis provides a method for evaluating equipment
status and the effectiveness  of other elements of the overall program.
6.4.2  Evaluating Quality-Related Data
     Accurate recording and analysis of program  information are fundamental
to a good QA program.  In particular, the ability to compare current data
with data from historical or contemporary programs is useful in assessing
the achievement and maintenance of program quality.  Close tracking of
quality-related statistics can identify problems in classroom training and
field certification and possibly prevent the unnecessary failure of smoke
school participants.
                                    81

-------
     Several  statistics  important  to evaluating  the  training program's  per-
formance have already  been  discussed.  These  include assessment of operational
error,  which  is  an  accumulation of several errors caused  by calibration filters,
transmissometer  calibration,  and opacity value reporting.  The error of the
individual  reader  is the main criterion for certification.  For an overall
training program,  however,  statistics  representing group  performance are of
greater value.  In  particular, average observer  bias and  variance provide
insight into  overall participant performance  and can identify training  and
equipment biases.   Although the percentage of trainees certified is indica-
tive of opacity  reading  ability, it is also a reflection  of the effectiveness
of classroom  training.   Figure 29  lists key statistics that can be useful  in
evaluating training school  performances.
     These statistics  are of particular use for  making comparisons with
previous data or similar data from other VE training facilities.  In this
way, trends and  significant deviations from a trend  can be quickly
observed.  A  useful tool for displaying quality  control trends is the
control chart.  Values of significant  parameters are plotted for succeeding
periods of time  or events,  thus developing a  trend line.  Typically, lines
indicating upper and lower  acceptable  values  are also charted to clearly
identify adverse trends. The major advantage of the control chart, or
similar tracking systems,  is that  it clearly  illustrates  changes in factors
indicative of program  quality.  It can therefore serve as a warning to
supervisors that procedures and equipment may need review or may signify to
management that  a  system audit is  needed.  A  typical control chart is shown
in Figure 30. A basic QA  program  would include  the  tracking of the
statistics calculated  in Appendix  C and those of Figure 29.
     Figure 31 lists  some of these statistics in a checklist form suitable
for auditing  purposes.  As  may be  noted, quality control  criteria specified
by Method 9 are  included in the  figures for ease of  reference.  Other
criteria values  in the figure represent those levels that reflect the
experience of VE training  schools  with good QA  programs.
                                     82

-------
Date:  	 Location:

Instructors:
Number of participants

Average of participants' mean
  deviations

Variance of participants' deviations
Average of participants' variance or
  standard deviation

Average of maximum positive and nega-
  tive biases

Percentage of participants that
  certified

Percentage of participants that re-
  certified
Percentage of participants failing
  due to mean deviation >7.5%, due
  to single deviation >15%
  both

Percentage of participants that
  passed the classroom examination

Average examination score
Average years experience of partici-
  pants
Percentage of participants within a
  given employment group (government,
  business,  consultants, others)
Smoke generator calibration error

Opacity reporting error
 Figure 29.   List of statistics useful  in evaluating VE training schools,
                                     83

-------
 MEAN OF
TRAINEE'S
 AVERAGE °
DEVIATIONS
         1978
                      CLASS MEANS,
                              ANNUAL  MEANS'1
                                                UPPER CONTROL LIMIT FOR CLASS MEANS
                                                      UPPER CONTROL LIMIT
                                                       FOR ANNUAL MEANS
                                               LOWER CONTROL LIMIT FOR ANNUAL MEANS
                                               LOWER CONTROL LIMIT FOR CLASS MEANS
1979
1980
                    Figure 30.   Example  control  chart for tracking  training
                                      school  performance.
                                       84

-------
Statistics
Calibration filter
error
Bias3X9
so, a 2
1
Total error
Transmissometer
calibration error
Bias )L
SD2 a 2
Total error
(X2 ±2SD2)
Opacity reporting
error
BiasaX3
Total error
(X3 +2SD3)
Overall SG/TC error
Bias I
__ ov
SD
ov
Overall error
(X +2SD )
v ov — ov
Student
certification
Percentage of
students that
certify
Percentage of
students that
recertify
Calculated
value

































Criteria
value
-




2%b




l%b




1.5%






2.5%






90%


95%
Acceptable
yes

































no

































Comments

































(continued)
                   Figure 31.   Statistics checklist form.
                                     85

-------
Figure 31 (continued)
Statistics
Percentage of
students that fail
due to:
mean deviation >7.5%
single deviation
>15%
Either mean devia-
tion >7.5% or
single deviation
>15%
Calculated
value




Criteria
value

2%
8%
10%
Acceptable
yes




no




Comments




 Specified in Reference Method 9.
 Calculated pursuant to Reference  Method 9,  i.e., the absolute value of actual
 deviations is used to determine mean deviations.
GSmoke generator/trainer.
                                    86

-------
                                 SECTION 7

                   VISIBLE EMISSIONS TRAINING LITERATURE


     The following is a list of selected references pertaining to visible

emissions determinations and training programs.

Conner, W.  D.   Measurement of Opacity by Transmissometer and Smoke Readers,
EPA Memorandum Report, 1974.

Conner, W.  D., and J. R. Hodkinson.  Optical Properties and Visual Effects
of Smoke Plumes.  Environmental Protection Agency.  Office of Air Programs,
Edison Electric Institute and Public Health Service, 1967.

Coons, J. D. ,  et al.   Development, Calibration, and Use of a Plume Evaluation
Training Unit.  JAPCA 15:199-203, May 1965.

Crider, W.  L., and J. A. Tash.  Status Report:  Study of Vision Obscuration
by Nonblack Plumes.  JAPCA 14:161-165, May 1964.

Hamil, H. F.,  R. E. Thomas, and N. F. Swynnerton.  Evaluation and Collaborative
Study of Method for Visual Determination of Opacity of Emissions from
Stationary Sources.  EPA Contract 68-02-0626, U.S. Environmental Protection
Agency, Research Triangle Park, N.C., January 1975.

Malmberg, K.  B.  EPA Visible Emission Inspection Procedures.  U.S. Environ-
mental Protection Agency, Washington, D.C., August 1975.

Osborne, M. C., and M. R. Midgett.  Survey of Transmissometers Used in
Conducting Visible Emissions Training Courses.  Environmental Monitoring
and Support Laboratory, U.S. Environmental Protection Agency, March 1978.

Ringelmann, M.  Method of Estimating Smoke Produced by Undustrial Installa-
tions.  Rev.  Technique, 268, June  1898.

U.S. Environmental Protection Agency.  Office of Air Quality Planning and
Standards, Emission Standards and Engineering Division, Evaluation of EPA
Smoke School  Results.  October 9,  1974.

Weir, A., Jr., D. G.  Jones, and L. T. Paypay.  Measurement of Particle Size
and Other Factors Influencing Plume Opacity.  Paper presented at  the  Inter-
national Conference on Environmental Sensing and Assessment, Los  Vegas,
Nev., September 14-19,  1975.
                                     87

-------
Wohlschlegel,  P.,  and D.  E.  Wagoner.   Visual  Determination of Opacity Emis-
sions from Stationary Sources.   Guidelines for development of a quality
assurance program, Vol.  9.   EPA-650/4-74-005-i,  U.S.  Environmental Protection
Agency, Washington, D.C., November 1975.

Yocom, J. E.  Problems in Judging Plume Opacity:   A Simple Device for Measur-
ing Opacity of Wet Plumes.   JAPCA 13:36,  January 1963.
                                     88

-------
                                 APPENDIX A
                   SAMPLE LECTURES FOR VE TRAINING PROGRAM

LECTURE 1:   HISTORY, THEORY, AND METHODS FOR EVALUATING VISIBLE EMISSIONS
     This first lecture is designed to introduce the student to the history,
principles, and theory of opacity.  The instructor should reemphasize the
purpose of the course and expand upon and clarify the introductory remarks
during the orientation.  He should explain that certification is necessary to
assure VE evaluations and that it will be discussed in more detail later in
the day.  The lecture should cover each of the following points.

History of the Method
     The entire VE evaluation system is derived from a technical concept
developed by Maximilian Ringelmann in the late 1800's to measure black smoke
emissions from coal-fired boilers.  The Ringelmann Chart, which was adopted
by the U.S. Bureau of Mines in the early 1900's, has found extensive use in
efforts to assess and control smoke emissions in this country.  Since the
early 1950's when the Ringelmann concept was expanded by the introduction of
the term "equivalent opacity," the chart has become a very reliable and
useful VE compliance/enforcement tool.
     The Federal Government has discontinued the use of Ringelmann numbers in
Method 9 procedures for Standards of Performance for New Sources  (NSPS).  The
procedures are now based solely on opacity.  Many states, however, still
refer to the Ringelmann Chart to evaluate black and gray plumes in their
regulations.  The general trend, however, is to read all smoke as percent
opacity.  Certified evaluators must therefore be familiar with both systems.
Opacity Theory
     As the use of the Ringelmann Chart for assessing smoke emissions increased,
considerable curiosity developed as to the theory and scientific  foundation
for this effective tool.  This curiosity grew with the transition to use of
opacity readings as the basis for evaluating visible particulate  emissions of
                                     89

-------
both white and black  smoke.   In  practice,  the evaluation of opacity by  the
human eye is a very complex  phenomenon,  and  for  the most part  the  theory
behind it is not completely  understood.   It  is well documented, however, that
visible particulate emissions  can  be measured with good accuracy and  repro-
ducibility by properly trained/certified observers.
     Previous study results  indicate that plume  opacity readings are  influ-
enced by many factors such as:   particle density and  particle  refractive
index, particle size  distribution,  plume background,  pathlength, distance and
relative elevation to stack  exit,  time  of day, and lighting conditions.
Of particular significance is  particle  size.  Particles decrease light  trans-
mission by both scattering and direct obscurations.   Particles  in  the diameter
size range of visible light, 0.4 to 0.7 urn,  have the  greatest  light-scattering
effect.
     By literal definition;  opacity is  the reduction  in visibility of an
object or background  as viewed through  the diameter of a plume; in terms of
physical optics, opacity is  dependent upon transmittance  (I/T  ) where I
                                                            i0          o
is incident light flux and  I,  the  light flux leaving  the plume.  Percent
opacity is therefore  defined as:
                         Opacity = (1 - I/j  ) x  100
                                           o
The relationships between light transmittance, plume  opacity,  and  Ringelmann
number are presented  in Table A-l.
              TABLE A-l.  COMPARISON OF LIGHT EXTINCTION TERMS
Light transmission,
%
0
20
40
60
80
100
Plume opacity,
°/
/o
100
80
60
40
20
0
Ringelmann
number
5
4
3
2
1
0
                                      90

-------
 Opacity  Reading  Methods
      Over  the years,  the  procedures and guidelines for  reading opacity have
 been  refined and more rigidly  defined.  The  three basic  techniques currently
 used  in  reading  visible emissions  are:
      1)    Time exemption  (frequency distribution)
      2)    Time averaging  (Method 9)
      3)    Stopwatch  (time  accumulation)
 Many  states are  using a combination of these procedures  in opacity readings
 depending  on the exact wording of  the applicable emission regulation.  The
 different  techniques  and  provisions must be  thoroughly understood, since
 legally  sound opacity documentation must be  consistent and in accordance with
 the applicable regulation.  The three procedures are summarized below.
 Time  Exemption --
      Nearly all  State Implementation Plan control strategies contain opacity
 regulations that are  based on  the  time exemption procedure.  This procedure,
 which historically was devised to  control coal  combustion sources, allows the
 source a stipulated number of minutes per hour to be in violation of the
 allowable  emission level.  The observer typically makes readings on 15-second
 intervals  and reads for several minutes longer than the stipulated exemption
 period.  The individual readings in excels of the allowable standard are then
 counted  to determine  the status of source compliance.
 Time Averaging --
     Commonly referred to as EPA Method 9, this procedure was developed by
 EPA in support of NSPS and has been adopted widely by State and local  air
 control agencies.  This procedure also requires reading on 15-second intervals
 over a period of 6 minutes (24 consecutive readings).  The sum of the readings
 is then mathematically averaged,  and that value determines source compliance
 status.
 Stopwatch Procedure --
     This procedure has recently  been  used by several  air control  agencies to
more effectively address  intermittent  or  highly variable emission levels,
e.g.,  coke oven emissions.  The observer  uses two stopwatches.   One  watch  is

                                     91

-------
activated at the start of the  observation  period  and  stopped  at  the  end.   The
second watch is activated in an  accumulating  time mode  each time  the emissions
exceed the stipulated emission level  and  stopped  each time the emissions  fall
below that level.   The total accumulated  time is  then read and recorded from
the second stopwatch over the  total  observation  time  of the first watch.
Variables that Influence the Accuracy of  Opacity  Measurements
     The human eye is a unique instrument  for making  opacity  readings  due to
its capability to  discern very narrow color bands of  the electromagnetic
spectrum and thus  identify light wave frequency.  The purpose of  the VE
training and certification program is to  refine  or  calibrate  the  human eye to
a scale that distinguishes opacity in 5 percent  increments.
     As previously mentioned,  several  factors influence the accuracy of
opacity reading.  The reader should  understand these  factors  and  take  the
necessary precautions to minimize errors.  The standard observer  form  requires
the observer to note and describe the conditions  of reading that  are needed
to substantiate the validity of  observations. This QA  measure in the  field
reading and in the review of field reading by supervising personnel  better
assures sound documentation for  enforcement purposes.
Sun Angle --
     Small particles in the plume tend to  scatter light in the forward direc-
tion at small angles with reference  to the direction  of the sun's rays.
Thus, a plume would appear to  be much more opaque if  it is not viewed  with
the sun behind the observer.  The most accurate  readings are  taken when  the
sun is within a 140 degree sector behind  the  observer.
Wind Direction --
     It is important that the  plume  be read through a plume diameter approxi-
mately the size of the stack exit.  If the plume is blown toward  or  away  from
the observer, it is likely to  be read through a  longer  pathlength than if
readings are taken with the wind blowing  perpendicular  to the line of sight.
The longer the pathlength through the plume,  the greater the  plume opacity
with diameters of constant loading and particle  size  distribution.
     The error involved in observations made  nonperpendicular to  the plume is
illustrated in the following example.  Assume a  plume rising  from a  rectangular

                                      92

-------
stack with a width of approximately 2 ft has an opacity of 20 percent.
Based on Beer's Law, the relationship between transmittance and pathlength is
as follows:
     where b = the extinction coefficient of the plume; an intrinsic
               property of the plume due to particle characteristics
           L = pathlength
For our example:
                    opacity = (I - I/j. ) x 100 = 20 percent
                                      o
                       I/j  = 0.80
                          o
                          L = 2 ft
Solving for b:
                              Ln (0.8) = -b(2)
                                -0.223 = -2b
                                 0.112 = b
On a 45 degree angle the actual  pathlength (L'),  as presented  in  Figure  A-l,
would be as follows:
                         L'  = cos  ^45 '  ~  0.707  =  2'84  ft
The apparent opacity is  obtained  by use  of the extinction  coefficient  pre-
viously calculated (0.112)  and substituting b and  L'  into  Beer's  law.
                         I,    _  -0.112  (2.84)
                          Ve
                         J/T   =*-°-318
                         VT   =  0-73
                            o

     Thus,             opacity  =  1  - ^ =  0.27  or  27%
                                     93

-------
          n
        Q
Figure A-1.   Top view of discharge.
                 94

-------
In this case, there was a 35 percent error between  the  two  measurements.
Proper positioning for plume observations  is  therefore  influential  to  the
accuracy of opacity measurements.
     Variable wind conditions can  cause the plume  to  shift,  putting the ob-
server in a poor viewing location.   The observer should change  positions if
possible, maintain the proper sun  angle,  and  remain perpendicular  to the
plume travel.  Such changes must be noted  on  the recording  form.   If a suitable
position cannot be found, the observer should note  the  conditions  and  discon-
tinue reading until viewing conditions improve.
Effect of Observer Elevation Angle on Observed Opacity  —
     As the observer moves closer  to the base of the  stack,  the angle  of sight
and the pathlength through the plume both  increase, causing  the observed
opacity to increase even though the cross-plume opacity remains constant.
Table A-2 illustrates how observed opacity decreases  with distance from the
base of a stack emitting a plume of 20 percent opacity.   Figure A-2 presents
the variation of observed opacity  with distance from  any elevated  source.
The observed opacity is 27 percent, as evaluated from a distance of one stack
height (H), drops to 22 percent from a distance of  2H,  and  to 21 percent from
a distance of 3H.
                    TABLE A-2.   EFFECT OF ELEVATION  ANGLE

Observer
distance
(Y)
H
2H
3H
Observed
elevation
angle,0
(e)
45°
27°
18°

Observed
pathlength
(po)
1.41 D
1.21 D
1.05 D^

Actual
opacity, %
<°a>
20
20
20

Observed
opacity, %
<°o>
27.2
22.2
21.0

Deviation, %
+7.2
+2.2
+1.0
                                     95

-------
      	 PATHLENGTH THROUGH PLUME
2K
                         2H
              Figure A-2.   Variation of pathlength through plume with distance
                                  from an elevated source.
                                             96

-------
The observed pathlength (P )  is  calculated  as  follows
                               D    D VH^  +  Y2
                                o ~       Y
where D = actual width of plume
      f = pathlength factor
                                   f.!a
                                        D
                                   T =  1 - 0 (decimal  form)
where T = transmittance
      0 = opacity
                                          rf
                                 0Q  =  1  -  T'
     Sample calculation:   Y = 2H
                                           yra  lm
                         f _ 1.12D _  .  1?
                         f	|j1.12

                         T = 1  - 0.2  =  0.8
                  Thus,  0Q = 1  - 0.81'12 =  22.2%

     The observer must be close enough  to the emission point to have reasonable
visual range.  Earlier versions of Method 9 recommended that the observer
stand a minimum of two stack heights  but not further than 0.25 mile from the
source.  While this is no longer required,  it does indicate approximately
what distances between the source and the observer are considered optimum.
Background --
     Maximum accuracy is obtained when  the  plume is read against a contrast-
ing background.   The reader should choose the background of maximum color
contrast with the plume, i.e.,  a green  tree is considered best for a light
colored plume and a blue sky for a black plume.  Many observers, however,

                                     97

-------
prefer a contrasting object such as a tree or building to be included in the
background for all  plume readings.
Viewing Point --
     Method 9 procedures expressly  require that the observation point be that
of maximum opacity.  This normally  means as close to the point of emission as
possible since the plume tends to dissipate with travel  distance.  Exceptions,
though, must be made in situations  when uncombined water or secondary partic-
ulate formation occurs.  The observation must be made through a portion of
the plume where uncombined water is not present since the water vapor can
condense when exposed to cold ambient temperatures and form highly opaque
plumes.  Always indicate on the observation form where the viewing point is
relative to the stack exit or source point.
Time Interval between Readings --
     Staring continuously at a reasonably steady state plume while making 15-
second interval readings causes eye fatigue,  which can result in reduced
visual  acuity.  To prevent this, the observer should only glance at the plume
momentarily and make the opacity determination at the regular 15-second in-
tervals.  Any deviation from this reading procedure must be noted and explained
on the observation form.
Atmospheric Haze --
     Both natural  or manmade atmospheric haze will  generally reduce the
contrast between the plume and its  background, thereby reducing the opacity
reading.  Hazy conditions should not significantly affect the opacity reading
if the visibility is at least 3 miles.
Wind Speed and Atmospheric Stability --
     A strong wind or unstable atmospheric conditions cause rapid dispersion
of the plume, thus reducing opacity.  As such interferences generally favor
the source, this would not hinder official  opacity readings.  Accurate read-
ings can be made under these conditions as long as the plume remains reasonably
intact at the stack exit and as long as the opacity is read as close to the
stack exit as possible.
                                      98

-------
Night Viewing --
     Questions sometimes have been  raised about the  feasibility  of  conducting
visual  opacity observations at night.   Plumes  can  be read  accurately  at  night
if special  training has been conducted and certain reading conditions  exist.
The opacity is judged by use of a light as a  background  target.   A  few agencies
regularly conduct night observations,  e.g., Los Angeles  Air Quality Mangagement
District.  Procedures for certification and field  observations under  nighttime
conditions, however, are not addressed in Method 9.

LECTURE 2:   SOURCES OF VISIBLE EMISSIONS
     This lecture should be given either by an experienced enforcement person
or by an engineer thoroughly familiar  with source  conditions and opacity
reading procedures and problems.  The  lecture  should be  illustrated by quality
35-mm slides showing sources and plume conditions  commonly seen  in  field
operations.
     Opacity readings are generally made in the course of  routine source
inspections or after a casual observation of apparent visible emissions.
Readings can be made either on or off  the property of the  alleged violator.
Observations must, however, be made in conformance with  location and  viewing
conditions conducive to accurate readings, because any field readings  and
documentation may be used as evidence  in a court of  law.   A source  represen-
tative may request to be present during the reading  period.  If  that  is  not
possible or desirable, the source should be contacted immediately after  the
readings are taken to determine the operating  conditions during  the period of
reading.  The source may be in a legitimate condition of upset or malfunction,
and thus any enforcement action would  be unwarranted. The source should be
provided the opportunity to explain or defend  its  operating condition  during
the period of alleged violation.
     The camera is very effective for  illustrating the visual  appearance of
heavy visible emissions.  It does not, however, replace  accurate opacity ob-
servations and can only be used as  secondary evidence in court.   The  inspec-
tor can testify that the photograph accurately represents  the plume location
and geometry that he observed.  Care must be  exercised in  taking photographs
so as not to divulge details of processes or  operations  designated  confidential

                                      99

-------
or as proprietary information by the source owner/operator.   A telephoto lens
should not be used; and if the inspector is on company property,  prior per-
mission is required to use a camera.
     The observer must also be cognizant of responsibilities  and  obligations
not to inadvertently divulge secret processes  or operations  in the course of
day-to-day inspection or surveillance activities.   A source  employee might
discuss a secret operation or process in order to  explain excessive emissions
or successful controls, but this privileged information must  not  be divulged
to others.
     It is highly recommended that the observer make a perimeter  survey prior
to and following the observation to determine  plant configuration and confirm
that multiple plume interaction did not cause  inaccurate opacity  readings.
This initial  survey also identifies the most appropriate location with respect
to Method 9.   Numerous EPA publications and workshops are available on the
techniques and procedures for conducting onsite inspection.
Combustion Sources
     Combustion and incineration historically  have been the major VE sources.
These sources include fuel (primarily coal  and oil) combustion for space
heating and power generation, incineration  for waste disposal  or  reclamation,
mobile sources, and process furnaces or operations.
     The principles of combustion (time, temperature, turbulence, and oxygen)
should be discussed along with combustion practices that have  proven to be
successful.  Unique conditions and applications for reducing  combustion emis-
sions are continually reported by field personnel.   An explanation should be
presented detailing when emissions are most likely to occur and what can be
done to reduce them.
Noncombustion Sources
     Industrial process losses, such as fumes, dusts, mists,  gases, and vapors,
are classified as noncombustion sources of  emissions.  Such emissions cannot
truly be called "smoke" because this term refers only to the  visible effluent
resulting from combustion, and consists mostly of  soot and fly ash.  Opera-
tions that emit noncombustion pollutants include grinding, melting, cooking,
materials handling, reaction processes, drying, and calcining.
                                     100

-------
     Because a wide variety of industries produce process  visible emissions,
this lecture should be tailored to the industrial activity where the majority
of the attendees will be working.  Limit this part of the  lecture to a few
examples of the problems and techniques involved in identifying, evaluating,
and controlling visible emissions from noncombustion sources.
     Many VE sources do not fall into the above categories.  Some sources such
as area and fugitive emissions, may be associated with the source operation.
Nonurban sources include demolition, road dust, farming, stockpiles, blasting,
quarrying, roof monitors, and perhaps open burning.  In urban problem areas,
the remaining uncontrolled particulate is emitted from a slightly different
set of major sources.  Despite the tendency of many agencies not to rely on
opacity observations in regulating these sources, the emissions can be ac-
curately read and effectively controlled by opacity regulations.  The basic
principles of opacity reading apply.  The major differences are that opacity
is usually read at or near ground level, and the pollutants are not emitted
from a confined stack-.   The biggest problem in reading fugitive emissions is
usually defining a single emission point and reading that plume.  Some reason-
able judgment in that regard will be necessary.  Such judgments and conditions
of observation must be thoroughly described on the emission form.  The stop-
watch technique may be preferable in reading some of these source conditions,
especially if the plume is intermittent and highly variable.   The stopwatch
procedure, however, should be stipulated in the applicable regulation.
Contaminated Water Aerosol Plumes
     Plumes containing  condensed droplets or water have been  variously de-
scribed as "moist," "wet," "steam," "contaminated water aerosol," or "con-
densed water vapor" plumes;  however, they are usually referred to as "steam
plumes."  Although this term is not technically correct since steam is a gas,
it has become the generally accepted description of a plume containing drop-
lets of condensed water.   In a "dry" plume,  the temperature remains  above the
dewpoint and therefore  the water remains in the gaseous state with no
effect on opacity.   As  the plume cools  and falls below the dewpoint, conden-
sation occurs and water droplets are formed.   At this point the visible
"steam" occurs.   These  condensed droplets scatter light,  generally causing
the plume to appear completely opaque.   Because nearly all  opacity regulations
                                    101

-------
exempt "steam" or water droplets,  care must be exercised to exclude the
effect of "steam" in any official  plume opacity readings.
     The moisture contained in the plume can condense within the stack itself.
This plume is generally referred to as an "attached"  plume unless the billow-
ing white plume forms downwind of  the stack, in which case it is referred to
as a "detached" plume.  In both cases, the contaminated water aerosol  will
usually revaporize eventually and  then disappear.   The formation and disap-
pearance of "steam" are affected by relative temperature and ambient humidity.
     Because of the exemption provisions, it is important  that the observer
readily recognize and discern "steam" contamination.   The  presence of "steam"
in the exit stack gases is determined by the billowy  white appearance and the
rather rapid dissipation of a dense plume.  A dry  plume generally does not
have the billowy appearance, dissipates very slowly,  and diffuses as it travels
downwind.
     Contaminated water aerosol plumes can be expected from high water con-
sumption processes, wet control systems (e.g., scrubbers and wet electro-
static precipitators), and cooling processes.  Therefore,  the inspector or ob-
server must have some familiarity  with the process and control  system in
order to expertly assess stack opacity conditions.
Condensible and Secondary Plume Formations
     Both condensible and secondary plume formations  are phenomenon of in-
creasing significance and concern, particularly at industrial  processes and
combustion sources using high-sulfur fuels.   Condensible plumes  result from
condensation of vaporized particulate as plume temperatures decrease.   Secondary
       -\
plumes'vflccur as particulate is formed from reactions  of species  within the
plume.  At many sources that have  applied high efficiency  particulate and
sulfur dioxide control systems, a  high opacity plume  persists.
     The major sources of these plumes are fossil  fuel-fired power plants,
coal-fired cement kilns, wood products drying operations,  and Kraft pulp mill
recovery furnaces.  Secondary formation of condensation products results under
certain atmospheric conditions. Under conditions  of  high  humidity, fine
sulfuric acid mist particles bind  with water molecules from the  atmosphere
to form large light scattering particles.  Such problems are particularly
perplexing since industry has spent large sums of  money to substantially

                                    102

-------
reduce mass loadings but is still faced with serious opacity control  prob-
lems.
     Another important formation occurs as a result of ambient cooling of the
flue gas.  Particles of this type (generally organics) are in the vapor state
at stack gas temperatures and upon cooling in the ambient air, change to a
liquid or solid state.  These emissions are often further complicated by
so-called steam plumes, and extreme care and keen observation are required
to accurately document both condensible and secondary plume formations.
The opacity must be read at some distance from the stack, and in this regard,
the guidance in Method 9, Section 2.3 is:
          "Opacity observations shall be made at the point of
          greatest opacity in that portion of the plume where
          condensed water vapor is not present."

Other Factors Affecting Plume Opacity
     Plume opacity can be significantly affected by a number of factors.   The
degree of influence of each factor varies widely from source to source.  Some
of the factors are presented in the following subsections.
Control Hardware—
     The type and performance of particulate control systems can have a sig-
nificant impact on opacity.  Most control devices are highly efficient in
collecting large particulates (i.e.,  > 20 urn),  but collection efficiency
generally drops with decreasing particle size.
     Growing evidence indicates that  electrostatic precipitators (ESP's)  and
wet scrubbers have fractional  efficiency curves of the type shown in  Figure
A-3.
     The particles in the 0.2- to 0.5-ym range  are particularly difficult to
collect due to the limitations of basic physical  mechanisms such as impaction
and field dependent charging.   Consequently, nonideal  performance probably
leads to the rapid increase in the quantity of  0.2- to 0.5-range emissions.
This corresponds to the range that scatters visible light most effectively
since the particle diameters are approximately  equal  to the wavelength  of
visible light.
                                     103

-------
i
I
u
z
LLI
IU
z
o
99.98

 99.9
 99.8
 99.5
  99
  98

  95

  90
O
O
     60
     30
                                               MEASUREMENT METHOD:
                                                 ^CASCADE IMPACTORS
                                                 O OPTICAL PARTICLE COUNTERS
                                                 • DIFFUS10NAL
                                               PRECIPITATOR CHARACTERISTICS:
                                                 TEMPERATURE - 335°C
                                                 SCA • 85 M2/(M3/$ec)
                                                 CURRENT DENSITY - 35 nA/CM2
      0.05
             0.1
   0.5        1.0
PARTICLE DIAMETER,
                                                                       5.0
                                                                             10.0
         Figure A-3.   Measured fractional  efficiency of a hot-side ESP
                    installed on a pulverized coal  boiler.
      Familiarity should be gained with the various control  devices and with
 the variations in system characteristics from source to source.  Variations
 are generally observed and defined over a period of several plant inspections
 and opacity readings.   Therefore, it is important to document and meticulously
 record one's observations and review the source file and records prior to
 subsequent inspections and opacity readings.
 Process Operation --
      Many processes have a cyclic operation (e.g., metallurgical furnace melts)
 Since visible emissions tend to be highly variable during the cycle, compli-
 ance documentation readings should be made during the period when highest
 opacity emissions are expected.  Many source operations (e.g., aluminum
 plants) have primary and secondary collection systems—collection hoods
 connected to the primary system and roof monitors connected to the secondary
 system.  The system configurations and the operating conditions during opacity
 readings should be observed and noted.
                                      104

-------
Raw Materials --
     The quality and content of raw materials  can  significantly  influence
particulate emissions.   The source of influence can  be as  simple as  the ash
content of fuels or as  complex as the chemistry of materials  and resultant
secondary reactions that might increase visible emissions.  A well-designed
control system may not  be compatible with a significant raw material  change.
This could explain a sudden VE increase and subsequent compliance efforts.
Circumvention --
     Most agency regulations address circumvention efforts  that  are  practiced
to escape control requirements.  Circumvention can be accomplished through
multiple stacks, bypasses, diffusers, air dilution,  etc.  The inspector must
be on guard against circumvention efforts or the alternation  of  operations  to
produce higher emissions at night or on overcast days, weekends, and holidays.
Atmospheric Conditions  --
     Wind direction, wind speed, atmospheric stability, turbulence,  relative
humidity,.and several other meteorological parameters influence  the  appear-
ance of the plume.  The wind direction affects the orientation of the observer
and the plume.  Plumes  blown directly toward or away from'the observer gener-
ally result in readings through longer pathlengths through  the plume than
those observed when the wind blows perpendicular to  the observer's line of
sight.  The longer the  pathlength through the plume, the higher  the  apparent
plume opacity.  As is the case with unstable atmospheric conditions, stronger
winds increase dilution and thus reduce opacity.
Miscellaneous Factors --
     Other factors that can influence opacity reading include background,  sun
angle, observer distance, and visual aids such as sunglasses. Therefore,  it
is important that these conditions and their respective influences be extensively
documented before, during, and after the readings on the observation report.

LECTURE 3:  FIELD OPERATIONS
     This section discusses the proper procedures for conducting field inspec-
tions.  This presentation should be made by an experienced field inspector
or engineer.
                                     105

-------
     Making technically sound opacity readings  requires  adequate  preparation.
In addition to being currently certified,  the reader must know the  source.
This preparation requires review of the official  source  file  to determine
process operating conditions, type and location of control  equipment,  history
of any VE problems, possible observation sites, applicable regulations,  pres-
ence of condensed water vapor plumes, pertinent operating data, and names  of
contact, etc.
     An opacity reader should have adequate supplies and equipment, including
inspection forms, writing pad or surface,  field logs, stopwatch,  sling psy-
chrometer, range finder, compass, binoculars, and camera.  It is  recommended
that field personnel always be equipped with a  hard hat, safety boots, safety
goggles, and in some instances, coveralls  and a respirator.  Other  equipment
that may be necessary depending on the type of source and the observation
conditions includes topographic maps and a hand-held anemometer.   In some
instances, agency field personnel have conducted observations with  little, if
any, of this equipment, and data frequently have been recorded on a standard
note pad, without the aid of even a watch.  The field personnel must be
adequately equipped and familiar with the use of all equipment.  Fully acceptable
readings., however, can be made pursuant to Method 9 with only a watch and  proper
forms.  The remainder of the equipment simply makes the  inspector's job more
convenient, and does not necessarily render the observation more  accurate.
     An agency-approved standard opacity form is essential to the quality
assurance of an enforcement program.  This form assures  that all  proper data
will be obtained and recorded to document accurate readings and it assures
better consistency of technique and information reported by different readers.
The consistency and adequacy of documentation should be  reviewed  and verified
by the supervisor or legal personnel prior to initiating enforcement actions.
     The reading data must be reduced or summarized in accordance with appli-
cable regulations.  Thus, if an NSPS source is involved, at least 6 minutes of
observation must be recorded to obtain an average of 24  individual  consecu-
tive readings.  In some instances where more than one regulatory  provision
might apply, the readings could be reduced by frequency  count, averaging,
and/or time accumulation, with the more stringent provision taking precedent.
     The need for thorough and accurate field notes should be  reemphasized.
Any field observation is fresh and clear in one's mind shortly after the

                                     106

-------
reading; most enforcement actions,  however,  happen  after a  considerable lapse
of time, perhaps years.   Hurried or sketchy  notes can  jeopardize  the  quality
and credibility of such  enforcement actions.
     Agency staff should make a copy of the  observations available  to source
personnel by leaving a carbon copy  or by later mailing a copy  from  the office,
depending on agency policy.   This should be  done with  minimum  delay and
should be officially noted in the record of  action.   Preferably,  a  represen-
tative of the source should  sign the front of the observation  form  to acknowl-
edge receipt.  Under no  circumstances, though, should  the inspector attempt
to summarize the results as  a finding of noncompliance or compliance. This
is a conclusion of law and the inspector's domain  is  limited  to conclusions
of fact.  In addition, such  action  would preempt the  inspector's  supervisor
and agency legal staff.

LECTURE 4:  AIR POLLUTION METEOROLOGY
     Meteorology has a large influence on air quality.  This  impact can be
from the microscale of building turbulence to the macroscale of pressure
systems.  Several meteorological parameters  directly  influence plume  opacity
(wind speed and direction, humidity, cloud cover, atmospheric  haze, etc.).
The field inspector must be  aware of these factors  and their  potential impact
on opacity readings.
     Although air pollution  meteorology can  be quite  complex,  especially  the
mechanics and mathematics that pertain to diffusion or dispersion modeling,
the opacity reader should be familiar with meteorological  concepts.  This
lecture should be given  by an air pollution  meteorologist,  a  specialist who
can present the material with a practical application.  The lecture content
should describe the basic plume behavior (Figure A-3)  and its  formation and
associated meteorological conditions (Table  A-3).
     Plume behavior and  transport (vertical  and horizontal) are largely a
function of atmospheric  stability and lapse  rate and  should be reasonably
understood by the opacity observer.  Of more direct influence  in  making
quality opacity readings are relative humidity, wind  speed, cloud cover,  sky
contrast, etc.  Several  techniques  can be used to measure or estimate these
conditions.
                                     107

-------
t
                                             NEUTRAL BELOW, STABLE ABOVE (FUMIGATION).
                                             UNSTABLE (LOOPING)
  TEMPERATURE-
                                             STABLE BELOW, NEUTRAL ALOFT (LOFTING)
                      Figure A-3.   Basic plume behavior.
                                        108

-------
                            TABLE  A-3.   PLUME BEHAVIOR AND RELATED WEATHER
   Description of
   visible plume
    Typical
   occurrence
  Me teo rological
    conditions
  Dispersion and
  ground contact
Fanning

Narrow horizontal fan;
1ittle or no vertical
spreading; if stack is
high, resembles a mean-
dering river, widening
but not thickening as
it moves along; may be
seen miles downwind; if
effluent is warm, plume
rises slowly, then drifts
horizontally
Fumigation

Fan or cone with well  de-
fined top and ragged or
diffuse bottom
At night and in early morn-
ing, any season; usually as-
sociated with inversion
layer(s); favored by light
winds, clear skies,  and snow
cover
During change from inversion
to lapse condition;  usually
nocturnal inversion  is
broken up through warming of
ground and surface layers by
morning sun;  breakup com-
monly begins  near ground and
works upward, less rapidly
in winter than in summer;
may also occur with  sea
breeze in late morning or
early afternoon
Inverted or isothermal
lapse rate; very sta-
ble; light winds; very
little turbulence
Adiabatic or super-
adiabatic lapse rate
at stack top and below;
isothermal  or inverted
lapse rate above;
lower layer, unstable
or neutral, upper layer
stable; winds light to
moderate aloft, and
light below; thermal
turbulence in lower
layer, little turbu-
lence in upper layer
Disperses slowly; con-
centration aloft high
at relatively great
distance downwind;
small probability of
ground contact, though
increase in turbulence
can result in ground
contact; high ground
level concentrations
may occur if stack is
short or if pjlume moves
to more irregular ter-
rain
Large probability of
ground contact in rel-
atively high concen-
tration, especially
after plume has stag-
nated aloft
(continued)

-------
TABLE A-3 (continued)
    Description of
    visible plume
     Typical
    occurrence
   Meteorological
     conditions
   Dispersion and
   ground contact
Looping

Irregular loops and waves
with random sinuous move-
ments; dissipates in
patches and relatively
rapidly with distance


Coni ng

Roughly cone-shaped with
horizontal axis; dissi-
pates farther downwind
than looping plume
Lofting

Loops or cone with well
defined bottom and poorly
defined, diffuse top
During daytime  with  clear  or
partly cloudy skies  and  in-
tense solar heating;  not
favored by layer-type cloudi-
ness, snow cover  or  strong
winds
During windy conditions,  day
or night;  layer-type  cloudi-
ness favored in day;  may
also occur briefly  in a
gust during looping
During change from lapse
to inversion condition;
usually near sunset on  fair
days; lasts about an hour
but may persist through
night
Adiabatic or super-
adiabatic lapse rate;
unstable; light winds
with intense thermal
turbulence
Lapse rate between dry
adiabatic and isother-
mal; neutral  or stable;
moderate to strong
winds; turbulence
largely mechanical
rather than thermal
Adiabatic lapse rate
at stack top and above;
inverted below stack;
lower layer stable,
upper layer neutral or
unstable; moderate
winds and considerable
turbulence aloft; very
light winds and 1ittle
or no turbulence in
layer below
Disperses rapidly with
distance; large proba-
bility of high concentra-
tions sporadically at
ground relatively close
to stack
Disperses less rapidly
with distance than loop-
ing plus large probabil-
ity of ground contact
some distance downwind;
concentration less but
persisting longer than
that of looping
Probability of ground
contact small  unless
inversion layer is shal-
low and stack  is short;
concentration  high with
contact but contact
usually prevented by
stability of inversion
layer; considered best
condition for  dispersion
since pollutants are
dispersed in upper air
with small  probability
of ground contact

-------
     Relative humidity is measured by a psychrometer,  following  a  simple pro-
cedure.  If this instrument is not available,  an adequate reading  can,  in
most instances, be obtained from airports,  weather stations,  or  perhaps a
local TV station.  High humidity days (>_ 70 percent)  are usually associated
with atmospheric turbidity that can interfere  with accurate opacity readings.
     Wind direction is determined readily by the direction of plume travel  in
relation to the orientation of the compass.  Wind speed can be measured by  a
small rotameter or hand-held anemometer.  Area wind information  can be  obtained
for the period in question from the local weather station.  A reasonable
estimate of wind speed can also be made by use of the Beaufort scale.   Accurate
determination of wind speed is not as important as other measurements.   It  is
basically important to support that reasonable readings were, in fact,  possible
and that unusual turbulence and plume shearing or separations were not  occur-
ring.

LECTURE 5:  LEGAL ASPECTS
     This lecture should be presented by an attorney thoroughly  familiar with
the practice and problems of air pollution control enforcement as  well  as
those recent court decisions that have affected the field of opacity reading.
It is becoming increasingly important that all field opacity readings  be
conducted with the approach and attitude that the documentation  must with-
stand the rigors of court scrutiny and interrogation.  - Because the evidence
is only as strong as the weakest element, it is important that all aspects  of
the program (from classroom training and certification to field  readings and
documentation) be conducted within the requirements and structure of a  good
quality assurance program.  This requires the  field observer to  have thorough
training and to be familiar with the legal  requirements and conditions.
     As the frequency and amount of fines and  penalties increase,  industry  is
growing more concerned about the significance  of a technically and legally
sound opacity enforcement program.  Opacity readings by field observers ac-
count for more than 90 percent of agency enforcement actions.  In response,
industry is spending a great deal of money to  better understand  the theory
and practice of opacity reading.  Industry has also initiated several
precedent-setting legal actions to weaken this important tool.  Therefore,  an
                                     111

-------
understanding of the legal  requirements and cases  is  essential  to avoid
needless jeopardy in important enforcement actions.
Applicable Regulatory Provisions
     Field readings made by the opacity observer will  usually document compli-
ance or noncompliance with  several  levels or layers of regulatory provisions.
To help identify and promote the required understanding of the applicable
provisions, the more common regulatory requirements are listed below:
Clean Air Act, as Amended 1977 --
     Section 110 - State Implementation Plan Requirements
     Section 111 - New Source Performance Standards
     Section 112 - Emission Standard for Hazardous Air Pollutants
     Section 113 - Federal  Enforcement Authority
     Section 114 - Inspection, Monitoring, and  Entry
     Section 119 - Nonferrous Smelter Orders
     Section 120 - Noncompliance Penalty
     Section 160 - Prevention of Significant Deterioration
     Section 169 - Prevention of Significant Deterioration (Section A)
     Section 303 - Emergency Powers
State Air Pollution Control  Laws and Regulations --
     Nearly all  states have opacity emission regulations that must be  at
least as stringent as those of the  applicable Federal  regulations.   States
are generally adopting Federal regulations (i.e., NSPS) but omissions  and
even inconsistencies may be discovered.   Therefore, it should be  remembered
that several opacity emission regulations may apply to VE  control  for  any
pollutant source.
Local Laws, Ordinances,  and Regulations  --
     Familiarity should  be  gained with these regulations which, in some cases,
are more stringent than  State or Federal  regulations.
Important Legal  Cases
     Opacity observations as a viable enforcement tool  have been  repeatedly
upheld by several state  courts as well  as the U.S. Supreme Court.   Significant
                                    112

-------
court cases concerning opacity as a viable enforcement tool are reviewed in
several references listed below.  Opacity standards have withstood the rigors
of many serious challenges in the courts of this country.  Recent rulings
include:
     1.   Air Pollution Variance Board of the State of Colorado v. Western
          Alfalfa Corporation, 419 U.S. 815, 94 S. Ct. 2114 (1974), at
          footnote 1.
     2.   State of New Jersey v. Fry Roofing Co., Docket No.  C-3682-72 (N.  J.
          Superior Court, 1974, attached as Appendix A).
     3.   St. v. Fry Roofing Co., 495 F. 2d 751, 4 ERC 1116 (Ore.  Ct.  of App.
          1972).
     4.   People v. Plywood Manufacturers, 291 P. 2d 587 (Sup.  Ct. of  Los
          Angeles, Ca. 1955).  California enacted a statutory opacity  require-
          ment as early as 1947.  Cal. Health and Safety Code 24242.
     5.   Essex Chemical  Corp. v. Rickelshaus, 158 U.S.  App.  D.C.  360, 486  P.
          2d 427 (1973).
     6.   Portland Cement Association v. Rickelshaus,  486 F 2d  375, U.S.
          Court of Appeals, District Columbia Circuit  (June 29, 1973).
     The following court cases established important legal  precedents  that
are significant to the conduct of a sound enforcement  program.
     The case of Air Pollution Variance Board (Colorado) v. Western Alfalfa
Corporation, No. 73-690,  U.S. Supreme Court, May 20, 1974,  pointed out the
necessity of immediately  notifying the source where opacity readings had
documented a violation.  The Court ruled that the source management must  have
opportunity to defend itself, and that it is unreasonable to  expect the
source representation to  reconstruct in court the operating conditions on the
day of the opacity observation several  days or months  later as  an  explana-
tion of the observed excessive emissions.
     The case of Donner-Hanna Coke Corporation v. Administrator U.S. EPA,
Civil  Action No. 77-232,  in the U.S.  District Court for  New York,  January 1978
pointed out the necessity of having clearly defined procedures  in  opacity
documentation.  The court ruled that EPA was remiss in not  having  documented
the applicability and accuracy of the "stopwatch" reading method.
                                     113

-------
Legal Procedures
     The inspection/enforcement staff must work  closely with  the legal  staff.
Before referring the case to the attorney, the field  observer and supervisor
should review the evidence and soundness  of the  opacity readings.  The  attorney
should review the evidence again and thoroughly  discuss any evidence of pos-
sible case weakness.  Corrective actions  (such as  additional  field readings)
should be taken if possible.  These tier  level  reviews  provide a final  layer
of QA review to the enforcement action.
     The attorney and enforcement staff should again  review the applicable
regulatory provisions and be cognizant of any duplicate opacity regulations
(i.e., State Implementation Plan, NSPS and Prevention of Significant Deterio-
ration provisions).  The legal staff will  probably decide if  the case is to
be prosecuted as a civil or criminal action.   The  inspector should have a
reasonable understanding of each procedure and his/her  role as an expert
witness.  The film "Role of the Witness"  illustrates  proper presentation and
behavior in this role.
Legal Restraints
     Different types of sources and different areas of  the country require
different VE restrictions.  Certain operations may be exempt  from the regu-
lations, and exceptions may be allowed for others  (such as agricultural
burning) during certain periods of time.   The specific  variances applicable
to the course location should be discussed.
Authority to Enter Facilities
     Section 114 (a) 2 of the Clean Air Act duly authorizes the control  of-
ficial to enter any facility to make inspections,  take  samples or readings,
and gather information and records.  Similar authorization is contained in
the enabling authority of most or all  states.   This authority, however,  is
being subjected to frequent challenges.   Any inspector  denied entry should
seek assistance from his/her supervisor and agency attorney to obtain a search
warrant.
Signing Waivers
     It is a fairly common practice for companies  to  request  nonemployees or
visitors to sign liability waivers as a condition  of  entry onto their industrial

                                    114

-------
facilities.  EPA employees cannot be denied entry for refusal  to sign such
waivers, and are specifically instructed not to sign such statements.  To
sign such waiver statements could jeopardize the rights of the individual and
his/her employer in cases of unforeseen injury or damage.
Confidential Information
     In order to explain excessive emissions, the company may  divulge or
share confidential  information.  It is important to understand the liability
associated with such information and the precautions that must be taken to
protect it.  Further decisions and guidance are presented in 40 CFR Part 2,
Public Information.

LECTURE 6:  TESTING PROCEDURES
     The preceeding classroom lectures have built the foundation for this
lecture.  The student should now be familiar with the theory,  history, source
conditions, meteorology, and reading techniques.  The next step is to work
with the smoke generator to "calibrate" the eyes and develop a proficiency
and confidence for making field readings.
     The trainee must be fully informed and aware of the events and procedures
in field training.   The training supervisor or assistant will  be in charge of
this portion of the training.  Most instructors have found 35-mm slides
helpful in presenting the material and in displaying the generator equipment
and components.  High-quality slides can effectively simulate  smoke in incre-
ments of 10 or 20 percent opacities.
     Practice forms may be distributed to the class, and practice recordings
may be referenced to assure the skillful completion of the forms.  Needed
field equipment should be listed:  clipboard, ballpoint pens,  comfortable
clothing, and folding chair.  The class should be told that any trainee who
intends to wear sunglasses in training and certification must  wear the same
type of glasses when making field opacity readings.  Glasses that change
intensity with changing sunlight and those with nonstandard colors are not
recommended.  The best sunglasses for opacity reading are those with gray
and green tints.
     The trainees should be reminded of the time and location  for the field
testing and certification.

                                     115

-------
     A short quiz should be  given  at  the  conclusion  of  the  classroom series
for two reasons.   First, the quiz  will  indicate  the  trainee's  comprehension
of the material  offered  in  the  lectures.   The  test need not pose  an  obstacle
to the student,  but simply  indicate whether  the  student understands  the  key
points of the lectures.   Problems  with  specific  questions would indicate
that this particular material has  not been clearly presented.   It may be pos-
sible to clarify  important  points  or  at least  make adjustments  for subsequent
sessions and provide a QA check on the  training  effectiveness.
     Second, the  last few questions on  the quiz  would allow a  brief  critique
of the course.   These questions would alert  the  instructor  to  any parts  of
the lecture that  need improvement.  Both  the quiz and critique  will  provide
information to allow the instructor to  continually improve  the  training
program.  Two sample quizzes are provided in Figures A-5 and A-6.
                                    116

-------
                                      NAME	
                                      DATE	
                                      LOCATION	

                             SMOKE SCHOOL QUIZ

 1.  A certificate is valid for:
     A.  6 months   B.  1 month    C.  1 year
 2.  The observer should stand,
     A.  facing the sun  B.  with the sun at his back  C.  With the sun at
                                                           his side
 3.  The opacity scale is used for
     A.  Black smoke     B.  White smoke     C.  Red smoke  D.  Blue smoke
     E.  A & B           F.  All of the above
 4.  A North wind blows from
     A.  South      B.  North      C.  West       D.  East
 5.  The smoke reading shall be taken with no allowance for deviation
     A.  At the top of the stack   B.  At the top of the plume
     C.  At the widest part of the plume     D.  At the narrowest portion
                                                 of the plume
     E.  At the densest portion of the plume
 6.  Smoke reading is a/an 	measurement.
     A.  Objective       B.  Subjective
 7.  The following is not a part of white smoke generation equipment:
     A.  Toluene    B.  Fuel pump       C.   Gasoline engine      D.   Blower
 8.  The following is not a part of black smoke generation equipment:
     A.  Gasoline engine      B.   Toluene    C.  Fuel  pump       D.   Blower
 9.  White smoke emitted from the generator consists of:
     A.  Talcum powder   B.   Oil  soot   C.   Oil vapor       D.  Steam

                 Figure A-5.  Sample Smoke School  Quiz I.
(continued)

                                     117

-------
Figure A-5 (continued)

Short answer or listing.

10.   Define opacity.



11.   What are the advantages of the visual  method of plume evaluation?




12.   What are the limitations of visual  plume evaluation?


13.   Sketch the proper  relationships of  (A) Smoke Reader, (B) Stack, (C) Wind
     direction (use plume),  and (D) Sun's  position (indicate North).
14.   What color light is most visible to human eyes?
     Why is this important?
15.   What are light obscuring mechanics?
16.   What is a hold harmless agreement?
17.   What should be added to this course?
18.   What should be removed?
19.   What should receive extra emphasis?
                                     118

-------
                                        Name 	
                                                         1 point

                              SMOKE SCHOOL QUIZ

Multiple Choice - 3 points each
 1.  A certificate is valid for:
     A.  6 months        B.  1 month         C.  1 year
 2.  The observer should stand:
     A.  Facing the sun       B.  With the sun on his back    C.  With sun at
                                                                  his side
 3.  The opacity scale is used for:
     A.  Black smoke     B.  White smoke     C.  Red smoke    D.  Blue smoke
     E.  A & B           F.  All of the above
 4.  A North wind blows from the:
     A.  South      B.  North      C.   West       D.  East
 5.  The smoke reading shall be taken:
     A.  One stack diameter above the top of the stack
     B.  At the top of the plume
     C.  At the widest part of the plume
     D.  At the narrowest part of the plume
 6.  The plume axis should be:
     A.  Along the reader's line of sight    B.  At right angle to the reader's
                                                 line of sight
 7.  Smoke reading is a/an 	measurement.
     A.  Objective       B.  Subjective
 8.  White smoke emitted from the generator consists of:
     A.  Talcum powder   B.  Oil soot   C.  Oil vapor    D.  Steam

                 Figure A-6.  Example Smoke School  Quiz II.
(continued)
                                     119

-------
Figure A-6 (continued)
Matching - 1 point each
 9.  	20% opacity
10.  	  0% opacity
11.  	Water vapor
12.  	 50% opacity
13.  	Gaseous pollutants
14.  	Wind speed
15.  	80% opacity
16.  	Light scattering
17.  	Photocell
18.  	Neutral density filter
19.  	 100% opacity
A.  Transmissometer
B.  Not measurable
C.  0 Ringelmann
D.  2% Ringelmann
E.  Steam Ringelmann
F.  5 Ringelmann
G.  4 Ringelmann
H.  Calibration
I.  1 Ringelmann
J.  Beaufort scale
K.  Opacity
Short answer or listing - 8 points each
20.  What is opacity (25 words or less)?


21.  What are the advantages of the visual method of plume evaluation?


22.  What are the limitations of visual plume evaluation?


23.  Distinguish combustion versus process sources.


24.  How can photographs be useful in  smoke  reading?


25.  List typical sources of BLACK smoke.
                                      120

-------
Figure A-6 (continued)

26.   List typical  sources of WHITE smoke.




27.   What is a sling psychrometer?




28.   What changes  would you recommend to improve the lecture content?




29.   Were speakers adequately prepared?
30.   Were classroom facilities generally adequate?  If not, what changes would
     you recommend?
                                     121

-------
                                 APPENDIX B
                     PERFORMANCE AUDIT AND SYSTEM AUDIT

     This appendix provides an example grading and recordkeeping checklist
for field certification test performance audits (Table B-l),  a checklist for
smoke generator performance audits (Table B-2), and an extensive information
checklist for a smoke school training and certification system audit
(Table B-3).   These forms are designed to check and document  QA for visible
emissions training programs.
                                     122

-------
       TABLE B-l.   GRADING AND RECORDKEEPING CHECKLIST FOR PERFORMANCE
                     AUDIT OF FIELD CERTIFICATION TESTS
Agency
    Date
Operator

Auditor
    Location
     Activity check
                                    Adequate
yes
no
Comments/corrective action
Field Operations

Certification forms (carbon/NRC
paper)

Field certification forms dis-
tributed individually to par-
ticipants by field certifica-
tion test supervisor

Participants instructed to sign
completed form before separating
copies

Participants instructed to sub-
mit original, completed, signed
form to test supervisor

Forms reviewed by supervisor for
name, signature, illegible char-
acters, completeness

Completed forms checked against
list of participants; discrep-
ancies noted

Both forms filed in a secure
area

Strip chart reviewed for cor-
rectness by test supervisor and
smoke generator operator

Operator certified that test
completed in accordance with
provisions of Method 9;
noted on chart

(continued)
                                     123

-------
TABLE B-l (continued)
Activity check
Chart with test forms and par-
ticipant list secured by test
supervisor
Office Operations
Master opacity sheet from strip
chart audited for correctness
Forms checked against partici-
pant list
Certification forms graded; mean
and maximum deviations noted
against master opacity sheet;
name, statistics, and pass/
no pass recorded on standard-
ized grading form
Opacity readers with mean devia-
tion <_ 7.5% and no deviation
> 15% passed by grader
Grading of certification form,
and pass/no pass status re-
viewed by checker/auditor
Test forms, grading form,
class list, strip chart and
master opacity chart for-
warded to training coordi-
nator; accuracy of grades
and pass/no pass status
certified
Certificates to passing
students issued by training
coordinator in accordance
with school requirements
and 6-month certification
period
Test forms, grading forms,
class list, strip chart,
master opacity list, and
list of certified readers
filed in secure area by
training coordinator
Adequate
yes









no









Comments/corrective action

                                     124

-------
         TABLE B-2.  SMOKE GENERATOR FOR CHECKLIST PERFORMANCE AUDIT
Agency
    Date
Operator

Auditor
    Location
     Activity check
                                    Adequate
yes
no
Comments/corrective action
Pre-Setup Inspection

General damage or damaged com-
ponents checked	
Inventory of parts and supplies
completed	
Fans and motors lubricated
Console received preliminary
check
If transmissometer not in-
spected within last 6 months:

  Transmissometer disassembled
  and inspected

  Photocell  detector inspected
  for alignment, damage, and
  proper electrical connection

  Phototopic response required
  by photocell

  Angles of projection and view
  £ 15 degrees
Standard Fuels Used

Toluene
Diesel or fuel oil
Propane
Other (specify kind and use)

(continued)
                                    125

-------
TABLE B-2 (continued)
     Activity check
                                    Adequate
yes   no
Comments/corrective action
Generator Setup

Background, sky, and wind condi-
tions checked for best generator
orientation
Generator leveled and wheels
chocked
Separate control console set up
Checks conducted for fuel leaks
(lines, connections and tanks)
Fans

All fans checked for smooth,
normal noise level
Main or induced draft fan
checked for normal operation
and flow rate
Transmissometer fans checked
for normal operation	
Operation of the Generator

Generator logbook available
and used
Both fuel pumps operable and
not leaking	
Chart recorder paper supply
adequate for proposed runs
Chart recorder visible to
operator or assistant
Chart recorder equipped with
event marker or otherwise
easily marked by operator or
assistant	

(continued)
                                     126

-------
TABLE B-2 (continued)
     Activity check
                                    Adequate
yes
no
Comments/corrective action
Console allowed to warm up for
30 minutes before making opacity
measurements
All opacity readout systems
operational, stable, and con-
sistent
Transmissometer System Calibra-
tion

Initial  zero and span drift
check completed; repeated
after each run
Drift greater than 1% corrected
Response time checked prior to
testing; 0 to 100% opacity
swi ng	
Response time 0 to 100%, <_ 5
seconds
Light source voltage checked
prior to testing	
Voltage - 5% of nominal
Calibration check

  Filters within +_ 2% of
  nominal value

  Five (nonconsecutive)
  checks for each filter
  completed

  Maximum error, 3%
  opacity

  Calibration record
  completed	

(continued)
                                     127

-------
TABLE B-2 (continued)
     Activity check
                                    Adequate
yes   no
Comments/corrective action
Generation of Smoke

Black smoke stability and
range test complete	
Opacity range 0 to 100% for
black smoke
White smoke stability and range
test complete	
Opacity range 0 to 100% for
white smoke
Public Address System

Setup and operation checked
Voice and cuing horn audible
and clear throughout reader
area
Total smoke system calibrated
and operating properly for
training and certification
purposes	
Comments (please list any problems or conditions encountered in the setup or
startup of the smoke generator):
                              Auditor's signature

                              Date
                                     128

-------
      TABLE B-3.   SMOKE SCHOOL TRAINING AND CERTIFICATION SYSTEM AUDIT
I.    General  Information
     Agency_
     Address
     Smoke School  Personnel
     Training Supervisor	
     Smoke Generator Operator_
     Training Assistants	
     Organization Schematic
Date
         Phone_
         Phone_
         Phone_
         Phone
     Schedule of Training/Certification (year
     Classroom              ,   	,
     Field opacity 	, 	, 	, 	;
     Average Number of Students Trained (year
     Agency 	        Industry 	
     Other (specify)
(continued)
                                     129

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
yes
no
Comments/corrective action
II.   Classroom Training

     A.   Registration and
         Orientation

           Forms and records
           Procedure
     B.   Theory and Principles
         of Opacity Reading

           Visual aids
           Materials and Prepa-
           ration
           Definition of opacity
           Opacity vs. trans-
           mi ssometer
           Contracts and back-
           ground effect	
           Particle size
           Chemical and physical
           properties	
           Light scattering
     C.  Sources of Visible
         Emissions

           Visual aids
           Material and  Prepa-
           ration
           Combustion sources
           Noncombustion sources
           Fugitive sources
 (continued)
                                    130

-------
TABLE B-3 (continued)
Activity check

D.
E.
Plume characteristics
Identifying water
vapor
Effect of control
equipment
Opacity Reading and Field
Documentation
Visual aids
Materials and prepa-
ration
Techniques o.f reading
opacity
Standardized forms and
field log
Important parameters
Field equipment
Regulations
Source review
Reading location
Sun angle
Number or time of
readings
Data reduction
Contacting the source
Meteorology
Visual aids
Adequate
yes

















no

















Comments/corrective action



(continued)
                                    131

-------
TABLE B-3 (continued)
Activity check
Materials and prepa-
ration
Atmospheric stability
Relative humidity
Plume characteristics
Wind speed and direc-
tion determinations
Building and obstruc-
tion effects
Cloud cover
F. Legal Aspects
Visual aids
Material and prepa-
ration
Legal history
Legal precedents
Right of entry
Hold harmless agree-
ment
Free and open field
concepts
Evidence and source
information disclo-
sure
Freedom of information
Case preparation
"Adequate
yes

















no

















Comments/corrective action


(continued)
                                   132

-------
TABLE B-3 (continued)
Activity check

G.
H.
I.
Requirements of certi-
fication
Expert witness
Preparing for Testing
and Certification
Visual aids
Materials and prepa-
ration
Requirements for cer-
tification
Review techniques for
reading opacity
Forms
Calibrating the eye
When to read
Use of glasses or
other aids
Closing
Course review
Quiz
Discussion
Refresher Courses
Requirements
Frequency
Adequate
yes















no















Comments/corrective action




(continued)
                                     133

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
                           yes   no
Comments/corrective action
     J.   Presentation and Prepa-
         ration

           Background speeches
           Speaker experience
           Material  organization
           Preprinted notes and
           illustrations
           Visual  aids
     K.
Facilities and Training
Equipment

  Classroom

  a.   Size and suita-
      bility	
           b.   Comfort and con-
               venience
           Visual  aid equipment
           Chalkboard,  easel,
           etc.
III.  Field Training and Certi-
     fication

     A.   Site Selection
Smoke generator power
supply
Suitable backgrounds
for reading smoke
Sufficient area for
expected number of
readers






(continued)
                                     134

-------
TABLE B-3  (continued)
Activity check
Reader and generator
properly oriented
with sun
B. Smoke Generator
General condition and
operability
Operating manual
Transmissometer sys-
tem
Angle of projection
(15° max)
Angle of view (15°
max)
Photopic response
Incandescent light
source
Nominal rated
voltage (- 5%)
Calibration fil-
ters (+. 2%)
Calibration error
(+ 3%)
Ease of calibra-
tion
Zero and span
drift (+. 1%)
Response time (max
5 sec)
Transmissometer fans
Controls (operating
and identified)
Readout devices
(specify type)

Adequate
yes

















no


















Comments/corrective action


(continued)
                                    135

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
yes   no
Comments/corrective action
           Control  Panel

             Operating condition
             Qperability
             Controls identified
             Electrical/fuel
             connections
           Main Fan

             Capacity (cfm)
             Air flow control
           Stack

             Diameter (12"  mini-
             mum)	
             General  condition
             Raising/lowering
             mechanism
           Black Smoke  Generation

             Fuel  supply and  type
             Stability  of plume
             opacity	
             Capable  of  generat-
             ing  0  to 100%
             opacity	
             Color  of  smoke over
             opacity range	
             Safety  of  operation
(continued)
                                     136

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
yes   no
Comments/corrective action
           White Smoke Generation

             Fuel supply and type
             Stability of plume
             opacity	
             Capable of generat-
             ing 0 to 100%
             opacity	
             Safety of operation
           General

             Spare parts inven-
             tory   	
             Maintenance tools
             Maintenance log
             Operations proce-
             dural manual
             Operator protected
             during inclement
             weather
           Public Address System

             Audible in reader
             area
             Bell or buzzer for
             cuing
IV.  Operation of Smoke Generator

         Operators familiar with
         the operation of the
         equipment	
         Generator service and
         maintenance conducted
(continued)
                                     137

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
yes   no
Comments/corrective action
         Generator checked for
         operation prior to day
         of training and cer-
         tification
         Warmup before operation
         prior to day of train-
         ing and certification
         Calibration require-
         ments completed
         Data and information
         properly recorded and
         maintained        	
         Students provided
         reasonable training
         at selected opacity
         levels
V.   Certification Testing

     A.  Meeting Method 9
         Requirements

         Both 25 white and 25
         black readings on one
         sheet          	
         No more than 7.5% average
         opacity deviation ac-
         cepted on either the
         white run of 25 or the
         black run of 25
         No error exceeding 15»
         ooacity allowed	
         Points of test given in
         random order	
         Applicants required to
         certify from the correct
         viewing angle	
 (continued)
                                     138

-------
TABLE B-3 (continued)
     Activity check
                                    Adequate
yes
no
Comments/corrective action
         Applicants properly in-
         structed before testing
         Adequate communication
         between tester and
         teacher
         Standards provided prior
         to tests
         Background provided
         Smoke held at one value
         long enough for evalua-
         tion
         Weather conditions con-
         sidered
         Elbow used during windy
         periods	
     B.   Testing Forms

         Carbon or NCR
         Name
         Loqation
         Date
         Run and color identifi-
         cation
         Distance from stack
         Sun/stack  reader  orienta-
         tion
         Time of day
         Cloud cover
         Wind  speed
         Wind  direction
(continued^
                                    139

-------
TABLE B-3 (continued)
Acti

C.
vity check
Sunglasses (type)
Corrections noted
Student signature
Graded by
Validated by
Issuance of Cards and/or
Certification
Dated properly
Accountable forms
Adequate
yes







no







Comments/corrective action

VI.   Quality Assurance
A. Records Control (Re-
tained for 5 yrs by
agency)
Student lists
Transmissometer cali-
bration data
Transmissometer opacity
data
Student test forms
Issued certificates
B. Auditing Functions
Traceability of cali-
bration standards
Calibration of genera-
tor transmissometer
Auditing of generator
transmissometer data
















(continued)
                                     140

-------
TABLE B-3 (continued)
Activity check

C.
D.
Verification and control
of documentation
Auditing of grading sys-
tem
Performance audits con-
ducted
Maintenance of Control
Statistics
Standards
Transmissometer response
Reported opacity
Applicants final test
score
Percent certification
Other
Calculated according to
accepted statistical
methods
General Information
(specify numbers)
Number of schools per
year
Number of persons per
school
Number of recertifiers
Percentage of new appli-
cants certifying
Percentage of recerti-
fiers certifying
Adequate
yes















no















Comments/corrective action

 The transmissometer  system consists  of  the  light  source,  photocell,  fans, and
 readout instruments.   The  system  must meet  applicable  40  CFR  60, Appendix A,
 Method 9 specifications.
                                     141

-------
                                 APPENDIX  C
              ANALYSIS AND EXAMPLE CALCULATIONS  OF VE TRAINING
                                 PROGRAM ERRORS

     Several elements of the VE training program performance can be esti-
mated, within statistical  limits, through  error  analysis of individual  pro-
gram activities.  These error estimates may be combined to produce an aggre-
gate error estimate.  The probability limits resulting from this error esti-
mate serve as a quantitative measure of program  effectiveness and should be
included as part of its documentation.   Individual errors associated with
the standardized calibration filters, the  transmissometer as compared with
these standardized filters, and the announced opacity as read from the trans-
missometer are determined along with their respective standard deviations (SD).
Combined precision and accuracy can then be estimated for the transmissometer
and smoke generator operator as well as the total precision and accuracy for
the VE training program including trainee precision.  The following sections
provide example calculations and an error analysis for a VE training program.
The descriptions of the statistics are not included since they are available
in several  standard statistical textbooks.
Filter Quality  Error
      In order to determine  the precision and accuracy of the transmissometer,
the error associated with  the standardized filters used to calibrate the trans-
missometer  must be  determined.  The following assumption should be used to
calculate filter quality  error directly from the  specifications of the stan-
dardized calibration  filters.
      1.   Each  NBS-traceable filter used to calibrate the transmissometer
          is accurate  within +0.5 percent opacity of  its nominal  value.
      2.   The area  within  3 SD of the mean of a  standard normal curve  is
          99.73 percent of the total area or approximately  the  total area.
                                     142

-------
     3.   Therefore, under the assumption of normal  distribution of
          errors, the calibration filter standard error (SD,)  will  be
          less than one-third of the nominal error or
                              SD
                              SD
                                1    3
                              SD1 < 0.17% .
          The standard deviation of error associated with the calibration
          filters is therefore less than 0.17 percent opacity.
Linearity - Calibration Error
     To determine how accurately the transmissometer compares with the stan-
dardized filters, the standard deviation (SD2) of the transmissometer must be
obtained by use of the following procedure.   Note:   At least 10 data points
from each filter must be used.
     1.   Use the data obtained from filter  readings taken during the VE pro-
          gram certification runs.  Fifteen  filter  readings must be taken at
          the beginning of the training course, and 15 additional readings
          must be taken at the end.  Record  these data points on a form
          similar to that illustrated in Figure C-l.
     2.   Determine the difference (d..) between each of the 30  values reported
          by the transmissometer and the actual opacities of the corresponding
          filters as shown in Figure C-l.
     3.   Sum the differences obtained in step 2 and determine  the mean (JL)
          and the standard deviation (SD2) of the 30 differences, where d,  is
          the first difference, d  is the second, and so on.
x2 =
SD2-
dl + d2 +

'(Zd^) - (
d3+ '•• +d30
30
Id.)2/n
L n - 1 J
%
Operational  Error
     The following procedure should be used to  determine  the  standard  deviation
(SD3) of the reporting error from the transmissometer,  the  difference  between
                                    143

-------


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
INITIAL
Measured
17.5
17.5
17.5
17.5
17.0
49.0
49.5
49.5
51
49.5
76
74
74.5
74
74
FILTER READINGS
Actual
18
18
18
18
18
50
50
50
50
50
74
74
74 +
74
74
a
di
.5
.5
.5
.5
-1
-1
.5
.5
+ 1
.5
+2
0
.5
0
0


16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
FINAL
Measured
17
20
17.5
19
18
50
49.5
51
49
50
75
74.5
74
73.5
74
FILTER READINGS3
Actual
18
18
18
18
18
50
50
50
50
50
74
74
74
74
74

di
-1
+2
-.5
+1
0
0
-.5
+ 1
-1
0
+ 1
+ .5
0
-.5
0
aMeasured denotes values read by the transmissometer; actual denotes filter
 values; d. denotes the difference (Measured - Actual).

                 n = 30
                   = 19
                           30
                           n - 1
                      19 - (CT/30)
                           29
                                        = 0.80%
                 Figure C-l.  Linearity - calibration error
                           data and calculations.
                                      144

-------
the actual  opacity value as recorded on the transmissometer strip  chart and
the opacity as announced to the nearest 5 percent value  by  the  generator
operator.
     1.   Randomly select one certification run from the school's  field
          training program.  Note:   Avoid selecting the  first run—it usually
          will have the highest error,  since the generator  operator is un-
          familiar with the generator's operation at this point.   On a form
          similar to the one illustrated in Figure C-2,  record  the actual
          transmissometer readings  for  each of the 25 white smoke  readings,
          the 25 readings as announced  by the generator  operator,  and the
          calculated difference (d.) for each of the readings.
     2.   Sum the differences obtained  in step 1 and calculate  the mean (X^)
          and standard deviation (SD3)  of the 25 readings,  where  d1 is the
          first difference, d~ is the second, etc.:
     3.
                                              J25
                 SD.
                                   25
                         Zd.2 - (Zd.)2/n
                                              25
                              n - 1
Repeat steps 1 and 2 for the 25 black smoke readings.   Refer  to
Figure C-2 and the strip chart record trace illustrated in  Fig-
ure C-3.
Probability Limits
     The following procedure is used to calculate the 95 percent probability
limits (PL) for the actual readings.  This calculation establishes a reasonable
estimate of the combined transmissometer and generator operator error in report-
ing individual smoke readings (see Figure C-4).   Figures C-l and C-2 show the
standard errors computed from the previous discussions of filter quality,
linearity-calibration error, and operational error.   The following steps
outline the procedure for determining these probability limits.
                                     145

-------
        Run  #1
WHITE SMOKE
Reported Actual d^
20 21 -1
15 15.5 -.5
10 9 +1
10 10 0
5 5.5 -.5
10 10.5 -.5
15 14 +1
20 19 +1
35 35 0
30 29.5 +.5
45 46 -1
50 50 0
50 49 +1
70 70.5 -.5
60 59.5 +.5
30 30 0
30 29 +1
25 25 0
20 19 +1
20 21 -1
10 10 0
30 30 0
50 49 +1
65 64.5 +.5
56-1
BLACK SMOKE
Reported Actual
35
35
25
15
10
15
20
40
40
45
45
55
65
70
75
80
50
55
40
25
15
10
20
30
35
For White Smoke: For Bl
n = 25 n =
Zdi = 2.5 zd. =
Id.
V-TT = Ir =0-1% *3 •
d.2 = 12.75 Zd.2 =
i i
s = 12.75 - (2.52/25)~h

*>3-
35
34
26
15.5
10
15.5
20
41
39
45
45.5
54
65
71
74.5
80
49
54.5
40
24
16
9.5
20
29
35
ack Smoke:
25
2.0
VH
LQi 2.0
n 25
11.50

Til. 50 - (2.
24
di
0
+1
-1
-.5
0
-.5
0
-1
+1
0
-.5
+1
0
-1
+ .5
0
+ 1
+ .5
0
+ 1
-1
+ .5
0
+1
0




= 0.08%


02/25)T2

                                                                 = 0.69%
Figure C-2.   Example of operational  error data  and calculation.
                              146

-------
          :(-flli:'
                               13'
TOO
i 190
            i70 if—]
                         150
01 I
; no;
•20: i
:30
•
i i  ;50
                       i i
                    I   !
             Figure C-3.  Example of a strip chart recorder
                  trace showing recorded opacity  value.
                                    147

-------
     For White Smoke:
       0.1 = 0 + 0.1
      1.00 =


       PL4 =

       PL4 =

0.1 + 2.20 =

0.1 - 2.20 =
                       (SD2)
                   (SD/]1*

 (0.17)2 + (0.81)2 + (0.72)2]

T4 + 2 (SD4) or

0.1 + 2 (1.10) or

2.3

-2.1
                                           h
                                                  For Black Smoke:
                                                     + x
                                                        3
                                           0.08 =  0 + 0.08

                                                   (SD,)2  + (SD,
                                           1.08 =


                                            PL4 =

                                            PL4 =

                                    0.08 + 2.16 =

                                    0.08 - 2.16. =
                                                   (0.17)2 + (0.81)2 + (0.69)2]

                                                  X"4 + 2 (SD4)  or

                                                  0.08 + 2 (1.08) or

                                                  2.2

                                                  -2.1
Therefore, 95% of the time it is expected that:

1.   An individual  white smoke reading will  not  exceed the true value by more
     than 2.3% opacity.

2.   An individual  white smoke reading will  not  be less than  2.1% opacity
     below the true value.

3.   An individual  black smoke reading will  not  exceed the true value by more
     than 2.2% opacity.

4.   An individual  black smoke reading will  not  be less than 2.1% opacity
     below the true value.
Calculated by
Calculations Audited by
                                                      Date

                                                      Date
     Figure C-4.
                  Example calculations and reporting of probability limits
                        for individual smoke readings.
                                      148

-------
1.   Determine the mean error (X») for white smoke.
                         x4 = x2 + x3

where     )L = mean of the 30 differences (d.) from initial  and
               final calibration error checks using NBS traceable
               filters
          X-, = mean of the differences (d.) from one-certification
           *J                             I
               run for 25 smoke readings for the operational  error
2.   Determine the cumulative standard deviation (SDj for white smoke.
                          )2 + (SD2)2 + (SD3)2

where     SD^ = standard deviation of error of the standardized
                calibration filters, i.e., 0.17
          SD2 = standard deviation of the 30 differences (d.) from
                initial and final calibration error checks using NBS
                traceable filters
          SD., = standard deviation of the 25 differences (d.) from
            •3                                              1
                one certification run for smoke readings for the opera-
                tional error
3.   Using the results from Steps 1 and 2, compute the 95 percent probability
     limits for individual white smoke readings.
                         PL4 = X4 ± 2(SD4)

     This results in a 95 percent probability that:
     o    An individual smoke reading will not exceed the true
          value by more than 	% opacity.
     o    An individual smoke reading will not be less than 	%
          opacity.
4.   Repeat Steps 1, 2, and 3 for black smoke.  Be sure to clearly mark
     whether the calculations are for white  or black smoke.  Sign and
     date the calculation form.   It is recommended that the calculations
     be audited and that the auditor also sign and date the form as shown
     in Figure C-4.

                                149

-------
Aggregate Data Analysis
     Even though the section on probability limits  provides  statistical  calcu-
lations for generator operator's biases along with  other sources  of error,
additional  calculations are still  required to derive  a  "total"  cumulative error
for the VE training program.  The standard deviation  (SDr) of the difference be-
tween the actual opacity values and those recorded  by the trainees measures the
precision of the individual trainees.
     Figure C-5 illustrates a qualifying test form.   The procedures for  calcu-
lating the mean (Xr) and standard deviation (SDr) for white  smoke follow.
Example calculations are illustrated in Figure C-6.
     1.   For each of the 25 white smoke readings on  the qualifying test
          form record the opacity announced by the  generator operator to
          the nearest 5 percent, the reported opacity from the  trainee's
          test form, and the difference (d.)  between  the announced and the
          trainee's values.
     2.   Sum all  the differences  obtained in step  1  and calculate the mean
          (Xr) and the standard deviation (SDr)  of  the  25 differences,
            0                                0
          where d,  is the first difference,  d~ is the second, etc.

                   X  =  dl + d2 + d3  + '••  + d25   =  ^1
                   X5              25                    n

                          (Id.2) - (Zd.)2/n
                  SD5 =      1         1

     3.   Repeat steps 1 and 2 for the 25 black  smoke readings.   Again,
          be sure  to indicate which set of calculations is for  white smoke
          and which is for black.
     The following calculations establish a  95 percent  confidence interval  (CI)
for the "true average" opacity based on 25 readings each of  both  white and black
smoke by an individual.
     1.   Combine  the three means  for  white  smoke (Xr)  to calculate the  mean
          error.  See Figures C-6  and  C-7.
                                     150

-------
AFFILIATION O~t &i"C. Hae.rJC.L4
Course Location R.T p
Sate V3-
Oistance
and
JO

5k
H(L


Direction to Stack f-



NAME 3~~0 K VS TT) 0 C_
RUN
, 2
Sunglasses Na
Sky

o*. 50 -I
" e

A K. Wind 1 i Q p-V (& \ \
f*n
n >=>
-f
b*r
/,
^-
READING

ERROR
NUMBER
1
2
3
4
5
6
7
3
9
10
a
12
13
14
1:
16
17
13
19
20
21
22
23
24
25
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
5
5
5
S
5
5
5
5
5
5
5
5
5
5
S
5
3
5
5
5
5
5
5
5
10
10
10
10
10
10 '
*tfl»
10 -
10
10
10
10
10
10
to
10
10
10
10
10
10
10
10
s^
15
X
15
15
IS
15
2^
OS?
©.
IS
LS
IS
IS
IS
IS
15
15
IS
15
15
IS
15
15
-«.
IS
IS


20
20



20
-«fl-
20
20
20
20
20
20
20
20
20
ji
JHK
3P4
S&
(3?
20
-tt.



25
25
25
25
25
Gy
~*s^
25
25
25
25
25
25
25
25
4|p

25
25
25
25
<0>
30
30
(55)
®
-10s.
30
30
30
M.
<3I>
30
30
30
30
30
30
30
"^38^
30
30
30
30
30
30
30
35
35
35
15
(jf)
35
35
35
35
35
— M-.
35
35
35
35
35
35
(^&
jS
35
35
35
35
35
35
40
40
40
40
40
40
40
40
40
40
($ft)
40
40
40
40
40
40
40
40
40
40
40
40
40
40
45
45
45
45
45
45
45
45
45
45
45
-4*- <
45 -
45
45
45
45
45
45
45
45
45
45
45
45
50
50
50
50
50
50
50
50
50
50
11
42>
S£
50
50
50
50
50
50
50
50
50
50
50
50
55
55
55
55
55
55
55
55
55
55
55



55
55
55
55
55
55
55
55
55
55
55
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
65
65
65
65
65
65
65
65
65
65
65
65
65

65
65
65
65
65
65
65
65
65
65
65
70
70
70
70
70
70
70
70
70
70
70
70


•

70
70
70
70
70
70
70
70
70
75
75
75
75
75
75
75
75
75
75
75
75
75
75
~"f5»
75 -
75
75
75
75
75
75
75
75
75
30
30
30
30
30
30
30
30
30
80
30
SO
30
30
30
W-
oo>
30
30
30
80
30
30
30
30
35
85
35
35
35
35
35
35
35
35
35
35
as
85
35
35
-85-
25
35
35
35
35
35
35
35
90 95 100 1
90 95 100 2
90 95 100 3
90 95 100 4
90 95 100 5
90 95 100 6
90 95 100 7
90 95 100 8
90 95 100 9
90 95 100 10
90 95 100 11
90 95 100 12
90 95 100 13
90 95 100 14
90 95 100 15
90 95 100 16
90 95 100 17
90 95 100 13
90 95 100 19
90 95 100 20
90 95 100 21
90 95 100 22
90 95 100 23
90 95 100 24
9q 95 100 25
DEVIATION c?-V-
READING
ERROR
NUMBER
26
27
23
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
43
49
50
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
5
5
"*V.
5
5
5
5
5
5
S
5.
^®
5
5
S
5
S
5
5
5
S
5
5
5
10
10
'vifl
• o5/
T&L
10
10
10
10
10
.10

50
50
SO
50
50
50
50
50
50
50
50
50
50
50
IS*
02)
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
55
- 55
55
60
60
60
60
60
60
60
60
60

-------
AFFILIATION	

COURSE LOCATION

DATE
                   NAME
                    RUN
                           SUNGLASSES
                SKY
                         WIND
DISTANCE AND DIRECTION TO  STACK
Reading
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
WHITE SMOKE
Reported
20
15
15
10
20
25
30
35
45
60
75
10
5
25
35
35
45
25
20
20
20
20
40
50
50
Actual
25
15
10
5
10
15
25
35
50
70
75
10
5
20
25
35
45
25
15
15
20
10
25
50
40
di
-5
0
+5
+5
+10
+10
+5
0
-5 .
-10
0
0
0
+5
+10
0
0
0
+5
+5
0
+10
+15
0
+10
BLACK SMOKE
Readi ng
1
2
3
4
5
6
7
8
o
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Reported
20
25
30
30
35
20
15
15
25
30
40
50
55
70
70
80
80
35
25
20
20
20
20
10
25
Actual
15
20
25
30
30
15
10
15
20
25
35
45
50
65
75
80
85
30
25
25
20
20
15
10
20
di
+5
+5
+5
0
+5
+5
+5
0
+5
+5
+5
+5
+5
+5
-5
0
-5
+5
0
-5
,0
0
+5
0
+5
     For White Smoke:
          n = 25
                                      For  Black  Smoke:



                                           n  = 25
     x  -   l"  - 75 - 3  n
     X5- —  - 25~ 3"°


     Id.2 = 1025
SD5 =
1025 - (752/25)


      24
= 5.77
                                      X   =1 =  60  -  2  4
                                      A5    n     25    ^4
Id.- = 450



  450 - (602/25)


        24
= 3.57
 Figure  C-6.   Example data and calculations  for  determining  student

                               certification.
                                     152

-------
For White Smoke:
  X6 = (X2)  + (X3)  + (X5)

 3.1 = 0 + 0.1 + 3.0
 SD6 =

5.87 =
 CI    j
   6    6 ~
CI  .
  6
(SD2)
(SD3)

(.722)
                                  (SD5
                                   (5.77)2]
                                           For  Black Smoke
             \/K
                                    3.72 =
 3.1 + 2.35 = 5.5

 3.1 - 2.35 = 0.8
                   x6  =  (x2)  +  (x3)  +  -(x5)

                  2.48 = 0  +  0.08  +  2.4

                  '(SO^2 +  (SD2)2  +  (SD3)2 +  (SD5)

                  (.17)2 +  (.81)2  +  (.69)2 +  (3.57
                              2(SD,)
                   CI,  = X,  + -—5- or
                     6     6  - V2S
                                             CI, = 2.48
                                               6
                                                          T
                                                          ^2
                                                                 V25
                                             2.48 + 1.49 = 4.0

                                             2.48 - 1.49 = 1.0

Therefore, 95% of the time it is expected that:

White Smoke

1.   (Qualified Individual's Name) will  not report an average of 25 white smoke
     readings in excess of 5.5% opacity above the true opacity.

2.   (Qualified Individual's Name) will  not report an average of 25 white smoke
     readings below 0.8% opacity above the true opacity.

Black Smoke

1.   (Qualified Individual's Name) will  not report an average of 25 black smoke
     readings in excess of 4.0% opacity above the true opacity.

2.   (Qualified Individual 's Name) will  not report an average of 25 black smoke
     readings below 1 . 0% opacity above the true opacity.
Calculated by
Calculations Checked by
                                                       Date
                                                       Date
   Figure C-7.  Example calculations for determining the bias and confidence
    intervals for a true average opacity reading by an individual observer.
                                     153

-------
                         X,- = X0 + X0 + Xr-
                          b    c.    6    D
where     5L = mean of the 30 differences (d.)  from initial  and
               final  calibration error checks using NBS traceble
               filters
          X"3 = mean of the 25 differences (d.)  from one certification
               run for white smoke readings for the operational error
          Xg = mean of the 25 differences (d.)  from the trainee's
               test form for white smoke readings
2.   Combine the four standard deviations for white smoke to obtain
     the "total" cumulative standard deviation (SDC) as illustrated in
                                                  o
     Figure C-7.
               SDg = ^(SDj)2 + (SD2)2 + (SD3)2 + (SD5)2

where     SD, = standard deviation of error in standardized calibration
                filters
          SD2 = standard deviation of the 30 differences (d.) from
                initial and final calibration error checks using NBS
                traceable filters
          SO, = standard deviation of the 25 differences (d.) from
             sj                                              I
                one certification run for white smoke readings for the
                operational error
          SD5 = standard deviation of the 25 differences (d.) for white
                smoke  readings recorded on the trainee test form and the
                master sheet for the operator's announced transmissometer
                readings
3.   Compute an approximate 95 percent confidence  interval for "true average"
     opacity based on  white smoke readings.
 where      CIg  =  95  percent confidence  interval  for white  smoke
           Xg  =  average of three means for  the  error  for  white  smoke

                                 154

-------
     SDfi = pooled or  combined  average of four standard deviations
           for white  smoke
     /FT  = square root of the  number of white smoke readings
Using the results computed  from Step 3, record the 95 percent confi-
dence interval on a form similar to  the one illustrated in Figure C-7.
Repeat steps 1 through 4 for the black smoke readings.  Be sure
to clearly indicate whether the calculations are for white
or black smoke.
Sign and date the calculation  form.   It is  recommended that the
calculations be audited and that the auditor also sign and
date the form.
   •U.S. GOVF.MMOT FMHTniC OFFICE: 19«5-537-002/21,50»
                          155

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
                              2.
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 TECHNICAL ASSISTANCE DOCUMENT:   Quality Assurance
 Guideline for Visible Emission  Training Programs
             5. REPORT DATE
             6. PERFORMING ORGANIZATION COOE
7. AUTHOR(S)

  Barbara E.  Blagun
             S. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  PEDCo  Environmental Inc.
  505  South Duke Street
  Durham,  North Carolina   27701
              10. PROGRAM ELEMENT NO.
              1 irCONTRACT/GRANT NO.
                                                              68-02-3431
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental Monitoring Systems Laboratory
 Office  of Research and Development
 U.S.  Environmental Protection Agency
 Research Triangle Park, NC   27711
              13. TYPE OF REPORT AND PERIOD COVERED
                Final
              14. SPONSORING AGENCY COOE


                EPA 600/08
15. SUPPLEMENTARY NOTES
 To be  published as an ORD  Project Report
 This document describes procedures and outlines  technical guidance for assistance in
 establishing effective visible emission observer training programs.   Smoke  generator
 operating  information is presented and appendices gives examples of sample  classroom
 training  lectures as well  as  audit forms and  statistical techniques used to  access
 the quality of a training  program.  The adherence of visible  emissions training
 programs to the materials  in  this document  should result in improved and consistent
 levels  of  observer training.   This documentation will aid enforcement programs  in
 demonstrating that the observers were trained without biases  and in a professional
 manner.
7.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lOENTIFIERS/OP6N ENDED TERMS  C. COSATI Field/Group
8. DISTRIBUTION STATEMENT
 RELEASE  TO PUBLIC
                                               19. SECURITY CLASS (Tint Jteporri

                                                UNCLASSIFIED
                                                                           21. NO. OP PAGES
20. SECURITY CLASS (Tim page/
  UNCLASSIFIED
                                                                          22. PRICE
EPA Form 2220-1 (R«v. 4-771   P«evtou» EDITION is

-------