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
------- |