VISIBLE EMISSIONS WORKSHOP
REFERENCE MATERIAL
Visible Emissions Concepts
and Special Problems
US Environmental Protection Agency
Office of Air, Noise, and Radiation
Division of Stationary Source Enforcement
Washington DC 20460
-------
FORWORD
The following document is a compilation of selected technical information
and publications on visual evaluation of plume opacity by the trained observer
method and use of this technique in the control of visible emissions from
stationary sources. The reference manual is intended as instructional background
material for persons attending the Region V Seminar for Visible Emissions
Instructors sponsored by the Division of Stationary Source Enforcement, Office
of Air, Noise and Radiation, U.S. Environmental Protection Agency. This document
was not designed or intended to be a self-instructional document, nor to be
reproduced as an agency publication.
vi
-------
TABLE OF CONTENTS
Page No.
SECTION A: BACKGROUND INFORMATION ON RINGELMANN AND EQUIVALENT
OPACITY TECHNIQUES FOR VISUAL EVALUATION OF VISIBLE
EMISSIONS
A-l. "Reading Visible Emissions," Chapter 10, Air Pollution Control
Field Operations Manual, 1962 3
A-2. "Development, Calibration & Use of a Plume Evaluation Training
Unit," J. Coons, APCA Journal, 1965 15
A-3. "Plume Visibility as a Control Basis," J. Coons, Mining
Congress Journal, 1968 21
A-4. EPA Method 9 Description, from EPA Guideline for Development
of a Quality Assurance Program: Visual Determination of
Opacity Emissions, EPA 650/4-74-005i, 1975 27
SECTION B: CHALLENGES TO THE ACCURACY AND VALIDITY OF VISUAL
PLUME READINGS
B-l. "Inadequacy of the Ringelmann Chart," Lionel S. Marks,
Mechanical Engineering, 1937 35
B-2. "Smoke Readings Vary With Observers" Electrical World,
January 15, 1975 41
B-3. "The Ringelmann Number as an Irrebutable Presumption of Gui1t—
An Outdated Concept, Donald J. Henz, Natural Resources Lawyer,
May 1970 45
B-4. "Visual Plume Readings--Too Crude for Clean Air Laws," Raymond
E. Haythorne and James W. Rankin, Natural Resources Lawyer,
Summer 1974 55
SECTION C: OPACITY AND VISUAL EFFECTS OF SMOKE PLUMES
C-l. Measurement of Opacity by Transmissometer and Smoke
Readers, William D. Conner, EPA Memorandum Report, 1974 ... 79
C-2. "Development of an Improved Smoke Inspection Guide," Andrew H.
Rose, Jr., et al., APCA Journal, August 1958 87
C-3. "Opacity and Visual Effects of Smoke Plumes," W. Conner,
Excerpts from EPA Report 650/2-74-128, November 1974 .... 93
i i i
-------
Page No.
SECTION D: RELIABILITY OF OPACITY STANDARDS
D-l. EPA Response to Remand in Portland Cement Association v.
Ruckelshaus, Appendix III, Part A: Opacity Standards, 1973 . 103
0-2. Excerpts from Various Collaborative Test Reports Summarizing
Results of Collaborative Testing of EPA Method 9 . • 145
D-3. Data Graphs 1-5 of Opacity Measurements and Observations
Made During Collaborative Testing of EPA Method 9 and
Summary Tables .' 151
D-4. Summary of Contract Report—Examining the Properties of
Qualified Observer Opacity Readings Averaged Over Intervals
of Less Than Six Minutes, T. Hartwell, 1 976 159
SECTION E: MEASUREMENT OF OPACITY AND PARTICULATE EMISSIONS USING
INSTRUMENTAL METHODS
E-l. "The Correlation of Plume Opacity to Particulate Mass
Concentration," D. S. Ensor and M. J. Pilat, reprinted from
AIChE Symposium Series: Air Pollution Control and Clean
Energy (Vol. 72, No. 156), 1976 165
E-2. "Instrumental Method Substitutes for Visual Estimation of
Equivalent Opacity," Herbert C. McKee, APCA Journal,
August 1971 175
E-3. "Texas Regulation Requires Control of Opacity Using Instru-
mental Measurements," Herbert C. McKee, APCA Journal,
June 1974 179
E-4. "Measurement of Opacity and Particulate Emissions with an
On-Stack Transmissometer," Heinz P. Beutner, APCA Journal,
September 1974 183
E-5. Chart Showing Opacity-Mass Emissions Correlations for
Various Source Categories, W. Conner, 1977 190
SECTION F: REMOTE MEASUREMENT OF PLUME OPACITY
F-i. Remote Measurement of Plume Opacity, W. Conner, Excerpt
from EPA 650/2-74-128, November 1974 193
F-2. "Remote Measurement of Smoke Plume Transmittance Using Lidar,"
Charles S. Cook et al., Applied Optics, August 1972 197
F-3. Amendment to NSPS to Allow Methods Other Than Reference
Method 9 for Measuring Plume Opacity, Federal Reqister,
Vol. 42, No. 99, May 23, 1977 .... . . . . ... 205
v
-------
Page No.
SECTION G: VISIBLE EMISSION TRAINING CERTIFICATION
G-1. Smoke School Evaluation Checklist, R. Missen, 1977 209
G-2. "Factors Affecting Accuracy in Visual Emissions Evaluations,"
R. E. Hague et al. , 1977 226
G-3. Affidavits of Visible Emission Course Attendees Describing
Experiences at Training Courses, 1974 235
SECTION H: ENFORCEMENT OF OPACITY STANDARDS
H-l. "The Validity of Visible Emission & Opacity Standards,"
Mark Mestel, January 1976 243
H-2. "Opacity as a Readily Enforceable Standard," Pamela
Giblin, 1972 255
H-3. "Simplified Visible Emission Standard," Emory J. Crofoot,
June 27, 1971 273
H-4. "The Opacity Witness," Kenneth B. Malmberg, EPA, 1976 . . . 283
SECTION I: MISCELLANEOUS OPACITY INFORMATION
1-1. Table: IGCI Reports Consensus on Industrial Emission
Levels Producing "Clear" or "Near Clear" Stacks, IGCI,
APCA Journal, July 1973 303
1-2. Visible Emissions Observation Form, EPA Division of
Stationary Source Enforcement-Proposed Form, August, 1981 . 305
1-3. "Smoke: A Fable," Burl E. Gilliland and Robert K. Roney,
Journal of Irreproducible Results, 1978 307
v
-------
SECTION A: BACKGROUND INFORMATION ON RINGELMANN AND EQUIVALENT
OPACITY TECHNIQUES FOR VISUAL EVALUATION OF VISIBLE
EMISSIONS
Page No.
A-l. "Reading Visible Emissions," Chapter 10, Air Pollution Control
Field Operations Manual, 1962 3
A-2. "Development, Calibration & Use of a Plume Evaluation Training
Unit," J. Coons, APCA Journal, 1965 15
A-3. "Plume Visibility as a Control Basis," J. Coons, Mining
Congress Journal, 1968 21
A-4. EPA Method 9 Description, from EPA Guideline for Development
of a Quality Assurance Program: Visual Determination of
Opacity Emissions, EPA 650/4-74-0051, 1975 27
-------
A-l
Chapter Ten from Air Pollution Control Field Operations Manual,
M. WeisbHnd, USHEW, Public Health Service, Washington, D.C. 1962
CHAPTER TEN
READING VISIBLE EMISSIONS
Once a plume or effluent ii identified as an air
contaminant, it must be measured by some standard
to determine whether or not a violation of the law has
occurred, or it must be evaluated to determine the size
or severity of a given air pollution problem. Here the
inspector may be required to make evaluations based
primarily on direct observation.
First-hand observation, despite its inherent sub-
jectivity, is not so inexact a means of evaluation if it is
recalled that air pollution is primarily a nuisance
affecting the sense perceptions. Most laws in the his-
tory of air pollution control are based on standards
which express annoyance to the senses by the use of
such qualifying terms as "noxious", "excessive", "re-
pugnant", "injurious". Relative air pollution intensi-
ties thus may be rated according to the sense of sight,
imell, and, in some cases, touch.
While the sense of smell is extremely unreliable
— a.ad of no value at all with odorless contaminants —
and the sense of touch, or, rather, physiological re-
sponse. is only valid in detecting certain types of toxic
substances, sight can be employed with a practical
degree of accuracy to distinguish between shades or
opacities of visible emissions. For this reason, maxi-
mum permissible emission standards based on the
visual determination of the effluent are widely used
around the world.
Compliance with maximum permissible emission
standards is determined by visual evaluation of visible
emissions, and source testing of emissions which are
invisible or near the threshold of vision. This chapter
is concerned with the evaluation of visual emissions by
the use of the Ringelmann Standard, and, specifically.
Section 24242 of the California State Health and Safety
Code.
1 TECHNIQUES OF VISUAL DETERMINATION
The only practical maximum permissible emis-
sion standard which developed for large-Kale use in
the history of air pollution was one which related to
shade or opacity, that is, the darkness or optical density*
of the plume. Standards which limited emissions ac-
cording to grain or dust loading alone were obviously
impractical due to the difficulty of conducting .ource
tests at all of the sources of air pollution.
The benefits of basing smoke statutes on opacity
or density are quite evident, even though equipment
and fuel regulations have increasingly assumed prece-
dence in control legislation. Since the visual standard
is specific with reference to a cut-off point and time
interval, it is simply and directly enforced. AH an
inspector need do is observe an emission of ar opacity
or density beyond that allowed for a specific period of
time in order to cite a violator for excessive smoke.
Also, although the visual standard is limitrd to esti-
mations of particles of pollution which obscure vision,
its application simultaneously tends to reduce grain
loading and gaseous -ontaminants. (As the grain load-
ing in the plume increases, the light transmission de-
creases exponentially.) In order to comply with the
opacity standard, more efficient combustion or equip-
ment operation is necessary. The Rirgelmann stand-
ard, therefore, is most versatile in accomplishing gross
reductions of atmospheric pollutants in a community,
and can 1« applied not only to sr loke, but fumes, dusts
and mists ai ising from a variety of problems as well.
It is perhaps one of '.he most comprehensive types of
rules adopted.
Il should be cautioned, however, that while such
bene:'ts ran be assumed, they ct.nr.ot ai vays be pre-
cisely predicted or evaluated. No useful correlation
exists between the shade ar opacity of an effluent with
an}' quantitative measurement of a plume. Some corre-
lation can be made u t the study of a specific operation,
particularly when grain loadings, operating conditions,
and opacities were previously correlated with great
care. Determination of opacity and shade of any emis-
sion alone, however, gives no specific measurement of
the quantities of contaminants being emitted.
A. Dneription and Use of lb* Ringtlmann Chart
The history, description ~.r.d general use of the
Ringpimann Chart is discussed in the Bureau of Mines
Information Circular *7718 (August, 1955). Sinco
this document has formed the bans of smoke regula-
tion in many cities, and is used as evidence in many
court actions, it is quoted in full here.
Sum-nary
The Ringelmann Smoke Chart fulfills an im-
portant net: I in smoke ibatement work and cer-
Tin pn ' lems in the rornbujtion of fuel. A k ,ow-
ledge of its history aad method of preparation is,
therefore, of interest to many. Since instructions
on its use are not shewn on the recent edition of
the chart, those included in this revision of the
piavious information circular, by Rudolf Kudlich.
now are a necessary complement to the chart.
More detail regarding the use of the chart is in-
cluded than was given in the earlier manuscript.
Introduction
The Ringelmann Smoke Chart, giving shades
of gray by which the density of columns of smoke
rising from stacks may be compared, was devel-
oped by a Professor Ringelmann of Paris. Maxi-
milian Ringelmann, born in 1861, was nrofesscr
of agricultural engineering at l'lnstitute National
Agronomioue and Director de la Station d'Essais
de Machines in Paris in 1888, and held those posi-
tions for many years thereafter.
The chart apparently was introduced into
the United States by William Kent in an article
published in "Enginee.ing News" of November
J
-------
156 Air Pollution Cc
11, 1897, with a comment that he had learned of
it in a private communication from a Bryan Don-
kin of London. It was said to have come into
somewhat extensive use in Europe by that time.
Kent proposed in 1899 that it be accepted as the
standard measure of smoke density in the stand-
ard code for power plant testing that was being
formulated by the American Society of Mechani-
cal Engineers.
The Ringelmann Chart was used by the en-
gineers of the Technologic Branch of the Federal
Geological Survey (which later formed the nu-
cleus of the present Bureau of Mines) in their
studies of smokeless combustion beginning at St.
Louis in 1904, and by 1910 had been recognized
officially in the smoke ordinance for Boston passed
by the Massachusetts Legislature.
The chart is now used as a device for deter-
mining whether emissions of smoke are within
limits or standards of permissibility (statutes and
ordinances) established and expressed with refer-
ence to the chart. It is widely used by law-
enforcement or compliance officers in jurisdictions
that have adopted standards based upon the chart.
In 1906 copies of the chart were prepared by
the Technologic Branch of the Federal Geological
Survey for use by its fuel engineers and for public
distributions. Upon its organization in 19t0, the
Bureau of Mines assumed this service together
with the other fuel-testing activities of the Tech-
nologic Branch.
Additional copies of the chart may be ob-
tained free by applying to the Publications Dis-
tribution Section. Bureau of Mines, 4800 Forbes
Street, Pittsburgh 13, Pa.
Description and Method of Preparing the Chan
The Ringelmann system is virtually a
scheme whereby graduated shades of gray, vary-
ing bv five equal steps between white and black,
may be accurately reproduced by means of a rec-
tangular grill of Slack lines of definite width and
spacing on a white background. The rule given
by Professor Ringelmann by which the charts
may be reproduced is as follows:
Card 0 — All white.
Card 1 — Black lines 1 mm. thick, 10 mm.
apart, leaving white spaces 9 mm.
square.
Card 2 — Lines 2.3 mm. thick, spaces 7.7
mm. square.
Card 3 — Lines 3.7 mm. thick, spaces 6.3
mm. square.
Card 4 — Lines 5.5 mm. thick, spaces 4.5
mm. square.
Card 5 — All black.
The chart, as distributed by the Bureau of
Mines, provides the shades of Cards 1. 2, 3, and 4
on a single sheet, which are known as Ringelmann
No. J, 2, 3, and ~. respectively.
Use of Chart
To learn to use the chart, it is supported on a
level with the eye, at such a distance from the ob-
server that the lines on the chart merge into
shades of gray, and as nearly as possible in line
with the stack. The observer glanMs from the
smoke, as it issues from the stack, to the chart and
notes the number of the chart most nearly corres-
Fitld Operation!
ponding with the shade of the smoke, then records
this number with the time of observation. A clear
stack is recorded as No. 0, tad 100 per cent black
smoke as No. 5.
To determine average smoke emission over a
relatively long period ot time, such as an hour,
observations are usually repeated at one-fourth or
one-half minute intervals. The readings are then
reduced to the total equivalent of No. 1 smoke as
a standard. No. 1 smoke being considered as 20
per cent dense, the percentage "density" of the
smoke for the entire period of observation is ob-
tained by the formula:
Equivalent usuU o( No. 1 »mok« X 0.20 p«rc«aue*
¦¦ — unoka
Number ot obianratioiu daiwty
Figure X-l shows a convenient form for recording
and computing the percentage of smoke density.
This procedure is often used on acceptance tests
of fuel-burning equipment.
The timing and extent of observation* made
for the purpose of determining compliance with
a local smoke abatement ordinance depend upon
the wording and smoke limitations of the ordi-
nance.
There are two general methods ot using the
chart. One is for the observer to make actual ref-
erence to it, as previously described, while judging
the smoke shade. The other method is basaa on
the fact that, with proper experience, it it unnec-
essary for an observer to continue to refer to the
chart. By repeated reference to the chart, during
a suitable training period, the shades of the Ring-
elmann scale become fixed in the observer's mem-
ory. Hence, the chart is used by most cities only
for training and examination of smoke inspectors,
before certification that they are proficient in
judging smoke shade on the Ringelmann scale
without referring to the chart. Since smoke-shade
observations by inspectors, thus trained and certi-
fied, are easily made and are accepted as evidence
in courts, this latter method of using the chart is
preferred by most authorities.
B. Smoke Mtuuring Melbodt
Although virtually all of the control agencies have
employed the Ringelmann Chart as a means of defin-
ing smoke standards, the methods by which smoke
densities and opacities are determined by enforcement
officers in the field vary. Some agencies use special
measuring devices, whereas other agencies train their
personnel to sight-read the effluent emissions within a
prescribed degree of accuracy without making direct
reference to the Ringelmann Chart. Some of the de-
vices used are described as follows:
t. Smoke Tintometer: This i n s t r um ent,
developed prior to 1912, used tinted glasses
graduated to the Ringelmann scale for visual
comparison with the smoke. It contains two
apertures, one for observing the smoke and one
for viewing the clear sky through the opening
or through one of the tinted glasses. Tim in-
strument is probably not significantly mote
effective than a trained sight-reader.
4
-------
Reading Visible Emissions
Location
Hour .SjQQtWsW.^i'#
Rite
Point of observation
Distance to stack
Direction of stack
Direction of wind
Velocity of wind
Zquiv. So. 1 Units
«.T • Qhits No• ^ .«..35......
Units Kb. k
Units So. 1
U3. Chits HO. 0
?V?. Ttalts
6866
Int. - Bu. of Mines, Pah., Pa
Figure X • 1. Rinftlmann Chart reading schema tuggeatad by the Buraau of Mint*.
-------
lit
Air Pollution Control Fitld Operation!
m'm •¦>»
mm ¦
m m,wt. ¦
m- mfmm^m,
¦.: afgn ••?¦': w-d ¦%•»"' ¦£¦*¦¦»,» ¦• »«¦'.
m. -mt«iV,^,''w^4^|,:,,W^
¦ ¦ "» ¦- ¦ ¦'¦»¦ '¦•>¦?¦.».¦•, ¦
a¦ m,a,¦.ai¦
a ¦ a ¦•"» • »,« *,».»,» ¦ a. ». a
m. m. ,»•-•«» ¦;*¦ ¦¦-
a.* •
1
"S
b
i
41
*
1
of
X
6
-------
Reading Visible Emission*
2. Umbrascope: This is a tube using tinted glass
segments which can be adjusted to cover one-
half of the field of view. The smoke can then
be compared visually with the darkness of the
glass. Its main disadvantage is its small range.
3. Smokescope: This instrument uses a film disc
of two shades graduated to *2 and *3 on the
Ringelmann Scale. One aperture is used for
viewing the smoke and one for viewing the
film reference disc against the background. Its
disadvantage is that the quantity of light fall-
ing on the reference disc may be influenced by
objects nearly in line with the smoke. Judg-
ment and skill are required in its use.
4. Photoelectric Cells: Photoelectric smoke me-
tering equipment measures variations in the
intensity cf n beam of light passing through
the effluent in the stack, thus directly measur-
ing opacity or optical density. Because the
equipment is permanently built into the stacks
at the sources of air pollution, these devices
are not portable for transporting by field in-
spectors to the stacks. However, they may be
required either by permit condition or law to
be constructed, and are of particular impor-
tance in training air pollution inspectors to
sight-read effluents.
In these devices, a constant light source is
used to illuminate a standard photoelectric
cell, both of which are diametrically opposed
across the stack. The light source must pass
through any smoke which will rise in the
stack before falling on the cell. Thus the cell
will produce a current of electricity which is
directly proportional to the amount of light
falling upon the cell, i.e., the lumens of light
falling upon the cell are directly related to the
micro*mpei-es of current transmitted by the
cell to th<» smoke measuring meter.
The metering devices vary in construction —
some use a potentiometer, a closed circuit re-
lay, or a direct reading microammeter. Re-
gardless of the type of smoke meter used, nr
the method of calibration, the amount of cur-
rent generated by the cell will be either 4i-
rectly or inversely proportional to the amount
of light falling on the cell, and will exhibit a
straight curve function which may be con-
sidered to be linear.
5. Smoke Comparison Charts-. Several smoke
comparison charts were reported in the
A.S.N1.E. (1936) Power Code. One of these
was a circular chart with radial lines of vari-
ous widths. The operator spins the chart on
an abject inverted through a hole in the center
of the chart. The apparent shades of gray on
the spinning chart are then compared with the
159
smoke. Shades of gray have also been pro-
duced on photographic film for smoke com-
parison charts. Another device consist* of
black lines photographed on celluloid film
which is seen partly by transmitted light and
partly by reflected light.
6. Sight Reading: In Los Angeles County, the
direct sight reading method of determining
violations of visible emissions without the use
of the Ringelmann Chart at the scene of the
violation was required for several reasons.
First, Section 24242 limits emissions of air
contaminants according to opacity, as well as
Ringelmann Smoke Density. It was found
that the general principles of visual determi-
nations apply to both opacity of colored
plumes as well as smoke density. Smoke read-
ing devices are limited to gray smoke. Sec-
ondly, it was found that devices were cumber-
some and were generally not significantly
more accurate in establishing opacity viola-
tions. Inspectors can be trained to read within
V4 Ringelmann or 10% opacity. Thirdly, the
methods by which the inspector is trained and
the methods by which he reads visible emis-
sions have been validated by the court*.
The accuracy and validity of the sight
reading method, without immediate reference
to the Ringelmann Chart, are determined by
intensive initial training, periodic refresher
courses and proficiency testing. The initial
and continuous refresher training an inspector
receives is of prime importance in establishing
his expertness.
The training conducted consists of two
important phases: (1) Training in the prin-
ciples of smoke and opacity readings, and (2)
proficiency training in the reading of smoke.
C. Printipht of Smoke end Opacity Reeding*
The Bureeu of Mines pamphlet, quoted previ-
ously, describes the various cards of the chart. The
terms "density" and "opacity", however, as used in
the literature are only inferentially defined. The
semantics of these has been the subject of much
controversy and have been confused by defense at-
torneys attempting to invalidate the entire smoke
reading procedure.
The definitions of these terms are, however, rather
simple, as they are limited by the law and the nature
of the Ringelmann Chart.
1. "Smoke Density": "Density" means the "quan-
tity of anything per unit of volume or area", as
defined by Webster's Dictionary. An examination
of the Ringelmann Chart discloses the obvious
fact that the shades of gray smoke are reproduced
' Bmd ia part on rtfumca 3.
7
-------
160
Air Pollution Control Field Operations
according to the ratio of the area occupied by the
black grid lines to the total area of each card, and
are expressed at the per cent of each card black-
ened. Since the black grid lines represent opaque
areas, and the white spaces the area through
which light is transmitted, it is implicit in the
design at the RuigeJmann Chart that "smoke
density" can only be defined as a measure of de-
gree of opacity. This definition does not imply
any relationship with the definition which might
be made in terms of "weight per unit volume"
To determine the per cent densities of each
Ringelmann Card we need simply subtract the
square of the distance across a white space from
the distance between the black lines as measured
from the center of each line. Since the first card
of the chart contains black lines 1 mm. thick, 10
mm. apart, leaving white spaces 9 mm. square,
the per cent of the area which is black is calcu-
lated as follows:
(10-) —(9*) -
100 — 81 = 19.0% of total area covered by
the black (opaque) grid lines
By similar calculation. Cards 2, 3, and 4 are
found to be 40.71%, 60.31% and 79.75% black.
Since the accuracy demanded of the chart for all
purposes in which it is to be used will not be 1%
or less, these panels may be considered a* perfect
20% increments. That is, the density of Card 0
is 0 per cent; Card 1, 20 per cent gr»y; Card 2,
40 per cent gray; Card 3, 60 per cent gray; Card
4, 80 per cent gray; Card 5, 100 per cent black.
Thus the Ringelmann Smoke Densities on the
chart may, by the optical phenomenon of the
blending of black lines to form varying shades of
gray when the chart is placed at a distance from
the observer, be measured by equivalent percent*
ages of opacity. This is true because the black
grid lines are seen only as the obscured area, and
the white spaces are seen as the background or the
area through which the light i* transmitted.
Smoke density is therefore nothing more than de-
gree of opacity.
2. "Opacity": The term "opacity" means the degree
to which transmitted light is obscured. The de-
gree of opacity is usually rated directly in p^j..
centage of perfect opacity. 0 per cent opacity be-
ing equivalent to perfect transparency' and 100%
opacity being perfectly opaque. tts a,r Pollution
work opacity is actually judged by to
which an observer's view is obscured- That is, the
expert reader makes his judgment of opacity on
the basis of the amount of backgwd. sky or
light that he cannot actually ««« througj, ^
emission. This manner of observing and record-
ing the opacities of visible emissions is clearly
implicit in the law. For instance. Section 24242
of the State Health and Safety Code prohibits
smoke of periods totaling more than three minutes
in any one hour . . .
(a) ... as dark or darker in shade as that des-
ignated as No. 2 on the Ringelmann
Chart. . .
(b) Of such opacity as to obscure an oburvtr't
view (emphasis supplied—-ed.) to a degree
equal to or greater than does smoke .. .(de-
scribed in "a", above)
It is obvious that the Ringelmann Chart and
the Opacity method are measured in direct equiv-
alents in that they both measure the same thing.
The determination of density is actually the de-
termination of opacity. The difference is that the
Ringelmann Chart is a recognized standard ap-
plied only to shades of gray smoke. The opacity
system it applied only to the shades of all other
colored emissions.
3. Other Principles: In reading tmoke it it not nec-
essary for a trained observer to actually use the
Ringelmann Chart in hit smoke meaturement.
The thought process is the same without the chart
for all color emissions. The U. S. Bureau of Mines
pamphlet states that "observer* with proper ex-
perience find it unnecessary to continue to refer
to the chart". The Superior Court of Lo* Angela*
County, in a recent appeal, ruled that "in proving
a violation, a witness may testify although he did
not have a Ringelmann Chart actually in the field
with him at the time he made his observations.
One does not have to have a color chart in his
hands to recognize a red flower, a blue sky, or a
black bird". * Thus, through training and repeated
examination, enforcement personnel are made
proficient in applying standard Ringelmann ref-
erence readings to field determination of both the
shade and density or opacity of any visible emis-
sion, without regard to its basic color — whether
black, white, yellow, or any other color.
The AS.M.E. Power Test CodeU) states that
tmoke may be read with Ringelmann Charts from
several blocks away with a clear atmosphere and
clear sky. Thus the distance from the smoke is
limited primerily by conditions of visibility and
positive identification of the source. The 1955 re-
vision of the Bureau of Mines Circular doe* not
specify any required distance from the smoke.
The A.SM.E. recommends placing the chart
so that the same light jallt on the chart at on the
tmoke. The observer should not be looking toward
• Pnplt iit. Plywood Mfr't of California (CRA 32S4-S), Shell Oil
Company [CRA-ilM), Union Oil Company (CJM 3303.08).
Southern California Eiitan Company (CRA 3327). Novrmbtr
2t. 193$. Memorandum Opinion of Superior Court, County at
Lot Aaatlw.
8
-------
Reading Visible Emissions
tit
the sun while the face of the chart is shaded.
About the same amount of light should be re-
flected from the white portion of the chart as
comes from the background of the smoke. Opaque
smoke charts are seen wholly by reflected light
while light colored smoke is seen mainly by trans-
mitted light. Thus, with a properly placed chart,
60% of the light reflected from a No. 2 Ringel-
mann Chart is equal to 60% of the light trans-
mitted through a No. 2 black smoke plume.
l. Smoke Reading School*
With these principles clearly and legally estab-
lished, it is now possible to train inspectors to become
expert smoke readers. An expert smoke reader ran be
defined as one who can distinguish smoke densities
within a margin of error of Yi Ringelmann or ten per
cent opacity in a significant number of readings dur-
ing both the hours of light and darkness and from any
given view of the emission.
pi(ur« X ¦ 3. A.P.C.D. Smoke School. Inspectors are trained
.j make visual determination! of Ringelmann Smoke Density by
usa of a black (moke generator equipped with a photoelectr c cell.
The stack on the left terves a white (moke generator used to train
inspectors in reeding smoke opacities. To qualify at an expert,
the inspector must not deviate more than plus or minus 10^
from the smoka measuring meter for eny 50 reading*.
During the early period ot sT.oke reading instruc-
tion at the A.P.C.D . technical problems were encoun-
tered in reading and sustaining a desired smoke
density or opacity with the training equipment then
* Based on reference 2.
available. As a result, a specially-designed black smoke
generating system was placed in operation in 1953, and
a comparable white smuke generating system one year
later. (See Figures X-4 and X-5.) A specially designeu
metering system was installed to replace the conven-
tional density recorder since in the latter device the
several stages of amplification were subject to malfunc-
tions which influenced the accuracy of meter readings.
i. Smoke Generating Equipment
In the design of smoke generating equipment, at-
tention was given to a means of regulating air fuel
ratios, to preventing horizontal distortion of the smoka
plume thxoi' jh the action of wind, and to accurate
determination of the actual smoke density or opacity
being observed by trainees.
The systems now utilized consist of two smoke
generating units, each with a vertical .tack and a
density or opacity Jete- ion system.
(1) Black Smoke System: In this unit, the combustion
chamber consists of a 40-cubic foot rectangul. r
steal box, lined with six inches of refractory fire
clay.
The oil burner is a modified mechanical pressure
atomizing type, with the combustion air fan set
to operate continuously at full capacity. Various
degrees of incomplete combustion of the fuel are
obtained by altering air fuel ratios through adjust-
ment of the fuel flow rate. This adjustment is se-
cured through manual operation of a needle valve
located in the fuel supply line, between the pres-
sure pump and the spray nozzles. Combustion
products pass from the chamber through a hori-
zontal duct and cooling chamber into a vertical
stack where they pass through the opacity detec-
tion system and are vented into the atmosphere.
A force drnft fan, discharging approximately 500
cubic feet of dilution air per minute into the base
of the stack, deters wind-caused horizontal distor-
tion of the smoke plume. The cooling chimber
prevents secondary combustion from occurri :ig it
the base of the stack as the combustion products
are diluted with air.
(2) White Smoke System: In this system, thu specific
opacity desin'd is obtained by controlling the rate
at which a distillate type of oii is sprayed into a
vaporizing chamber where it is diluted with air.
Up in leaving this chamber, the generated smoke
enters a vortical stack where it passes through the
opacity detection system before venting to the at-
mosphere. At times, it is ne essary for the operator
to adjust both the oil pressure on the spray nozzle
and the heating chamber temperatures in order to
sustain a specific opacity.
(3) Opacity and Density Detection Systems: The
opacity and density detection systems comprise a
9
-------
Air Pollution Control Fitld Operations
Deem detection system
STACK —»
12" Dtantar * 16' Hiji
AIR SUPPLY TUBE
COOLING
CHAMBER
OIL BURNER
COMBUSTION CHAMBER
REFRACTORY LINING
DILUTION FAN
Figure X - +¦ Dctign of (ray unofc* gtiMralor uaad by >i>* A.P.C.D.
OPACITY DETECTION SYSTEM
OIL PRESSURE GAUGE
MANUAL OIL CONTROL VALVE
OIL SPRAY NOZZLES
OIL SUPPLY LINE
OIL SUPPLY PRESSURE PUMP
tDILUTION AIR
VAPORIZING CHAMBER
VAPORIZATION
" HOT PLATE"
DISTILLATE OIL BURNER | "EATING CHAMBER
I
STACK
12" Diameter x 16* Hifh
AIR DILUTION FAN
1200 CFM
— COMBUSTION GAS VENT
REFRACTORY lining
Fijur* * 'S' Daugn "hit# tmoka (tlMrator luad by tha A.P.C.D.
10
-------
Reading VitibU Emiuiaru
163
light source, a photoelectric cell and a milliam-
meter.
A photoelectric cell and a light source are posi-
tioned at opposite ends of a light tube protruding
horizontally from each side of the smoke stack.
Upon receiving light energy passed through the
tube, the photoelectric cell generates an electrical
current which deflects the pointer of a milliam-
meter. The face of the meter has been modified
so as to indicate, not the amperage reaching the
meter, but rather the per cent of opacity or dens-
ity of the measured plume. Normally, a milliam-
meter reading 100 would be registering free
passage of light from the source of the cell. It
therefore was necessary to reverse the scale of the
meter so that 100 would indicate that no light
energy was reaching the cell and that 0 would
indicate free passage of light. When the operator
adjusts the foot-candle output of the light source
to cause a full scale deflection in the milliammeter,
an opacity of zero is read. Conversely, if the light
is turned out, no light energy reaches the cell and
the nulliammeter, therefore, reads one hundred,
or 100% opacity.
The detection system has been calibrated to per-
mit accurate reading of smoke opacity and density
in terms of the Ringelmann standard. This has
been accomplished by inserting photographic
plates of known opacities and densities between
the light source and the photoelectric cell.
Through this method milliammeter deflections of
20, 40, 60 and 80 are produced.
b. Training Methods
The training objective of the "smoke-reading
school" is a level of individual inspector proficiency
which permits field determination of smoke opacity or
density within a margin of error of V4 Ringelmann, or
10 per cent opacity. The past experience of the Dis-
trict indicates an optimum average deviation of T to
8 per cent with an acceptable and attainable group
deviation of not more than tO per cent. Since any
emission equal to or greater than No. 2 Ringelmann
constitutes a violation in Los Angeles County, it can
be proven that a violation has occurred when an in-
spector trained in the District smoke school determines
that an emission equals or exceeds No. 2tt Ringel-
mann.
During the instructional sessions, trainees are po-
sitioned approximately fifty feet from the smoke gen-
erator stacks.
As the training run begins, inspectors observe the
emission and. in the case of a black or gray emission,
compare it with a Ringelmann Chart posted on the
generator stack. Whenever the generator operator
sounds a bell, the trainees individually record their
reading of opacity or density, and then are advised by
the Instructor as to the actual opacity or density, as
determined by the detection system.
Training runs are conducted under a variety
of conditions simulating actual field situations under
varying conditions of light and darkness, and from
many different possible views of the plume in respect
to the inspector's position and the light source. Back-
ground color and varying lighting conditions may be
critical to accurate opacity determination of white or
other colored emissions, and these varying factors,
therefor*, are introduced during the training runs.
During the initial smoke-reading training given a
new inspector, approximately one hour is devoted to
the training run, and, during refresher sessions, ap-
proximately thirty minutes. In total, each inspector
receives twenty-four hours of smoke-reading instruc-
tion during his initial entry-training, and annually,
thereafter, receives six refresher training sessions, each
of four hours' duration.
c Proficiency Testing
Testing of individual inspector proficiency is con-
ducted both as a diagnostic device to determine the
need for additional training and as a means of assur-
ing a level of expertise that will fully meet the re-
quirements of the District's field operation program
and the possible test of a court action.
Since all new inspectors must serve a six-month
probationary period, failure to qualify fully during
the training program may lead to termination of em-
ployment during the probationary period.
During the test, inspectors are required to record
fifty readings of effluent density or opacity from the
smoke generators. The determinations are recorded
on a "Proficiency Record" (See Figure X - 7) when-
ever the generator-operator sounds a bell.
Since air pollution emissions of 40 per cer. o| ici-
ty or density are illegal in Los Angeles County, ttsting
of inspectors is concentrated on determination of the
inspector's ability to discriminate between emissions ,n
the range of 30 to 50 per cent opacity or density, with
testing also of the highs and the lows.
Scoring of the inspector's "Proficiency Record" is
accomplished by recording opposite the inspector's
reading, the actual opacity or density as determined
by the detection system. The trainee's deviation from
the meter reading is then indicated in an adjacent col-
umn, and for each set of twenty-five readings indicat-
ed on the record the following computation is made:
1. The total number of correct readings.
2. The total number of plus and minus readings.
3. The average plus and minus deviations for each
set of twenty-five readings as well as for the en-
tire fifty.
4. The percentage of readings deviating less than H
Ringelmann from the actual opacity or density.
11
-------
Air Pollution Control Field Operations
AI ft POLLUTION CONTROL OISTRICT •• COUNTV Of LOS ANGCLCS
414 SOUTH SAN PIBHO STMCT, LOS ANSILCS 11. CALIFORNIA. MAO It On ».47l!
NO
NAUC .
_or Z—
* « »"¦"»
SMOKI SCHOOL
_____ CXAMINtR.
OATS i
mm mu».
\. Sntor rooting* lncotu«*« I, 2, 4 on* 9 In AUgtimtfln nti«oor« or frtctlOAO only, Oo Mt «ntoi
tno* dtciMll/* Cntor froctlono no itot tMn 1/4.
2. fntor rooClngo I* 7, 8, 10 on* II in porcoot opocity fro* 0 to. 100 )»i or I0*t.
3. Entor dovUtlono lacolu*no 3 t6 I« portont only, l.o., ono * Aovlotlon • 201, * * «ovlotion •
It eolu««o 9 and 12 «ator s tfifforooco N»t«oon coluono 7ono« of 10 to* h.
4. If /our rooming '• '••• th* «nt«r tM dovlotion oo alnw*. if it |» jrootor
1
2
¦i."
J
...i
-1
7
-J
¦H
_? _
TT
.. If
u
13
*
i
s
I.
1 •
-r.
J?
J
4
X
it
i
<
11
X S
x .
*1
1 *
1
<
»
it
il
i
t
i5
1"
ij
«*
i
4
i
i
j
a
~*i
1 *
!*
1
3
»
1
1
2
j
~to
»
J
•u
1
»
10
,
11
9>
H
•!
a
~20
1*
5
ut
•10
3
1
»
.„r J„
u
,u
J
3
k
~10
17
*
Ji
i
JO ..
22-
17
%
_i
5
1 1A
2
~ *
U
II
Oi_.
k
«
, f 5
u
*
K
4
1
*10
w
2*.
J
i-n
_jn
. .in
It
It -
6
i
~ 5,
30
,J
u
i
<0
..to
.ia
ao
L«
1
1
~10
a
?
1
7
M
w..
..tin,.
21
s
«
*
*
4
~10
a
4
_
4
n
w
.•}
XI
M
H
•
»
J
3
m
n
JA
A. .
,t 5
»
M ..
¦ .10,.
8
-a.
«
* «
IS
U
.«.
%
}
,4 .
10
ao
a
«r
U
k
k
25
s
j:
.
u
AO
W._
~ «
»
A(
TO
~ «
u
k 1/1.
«
~ 5.
no. of CorrMt .
u
«
«_
NO. Of Corroot n
I?
4
1
of Rlu.
13
k«
- «
NO. Of P\it%
u
U_
2
JLA-
no. of ulmi* 1
u>
_a—
_1_S_
NO. Of Mlltvt J
Slock «0O«l0f0
Su« of Plot Dovlotloao
*M«or of PI«o AooOlnfO
Avorofo *tv* Oov. i
JUL.
jjy "jii ftrr.i,
NO. PlttO JUOOlAQl
MAI I to MMlifl
Su« of *'«• Pcvlotlooo
NwMOor of 'I wo floooififo
Avorogo Plus Oov. ,
JUi *
-20t.
Block ftoo0f*«0
Sua of MlflwO Oovlotlooo
mim00r of lllmi* HooClftf*
Avorogo Minwo Oov* ¦
-UL
3w of Mingo 0o».
MO. MlMlO Mooting*
Wfiito *004l0f0
Sum of Mlnwo Oovlotlooo
kgnOof of ttlm» flooding*
Avorofo Minus Oov, a
Sw» Qf MlOIlO OOV.
NO. MlflUi AoO0l*fO
JiL
•lock ftoo4l*«»
¦•reojjj of 0*«din«* I0» Oo.lotlo. oX .M(r
*Mto «oo«Ugt
Poreoot of Aoodlftfo (Of Oo* lot loo ond uMor
_JH »
rnunTATlONS HIOW ARC FOR SLACK ANO WHITI REAOINOS C0WINI0
Numr of Corr*€t RM^l"*1
Nuao«r of Roo4lm» o»*r *
or 10* owl"*1*
NucMar of Ro««lnf»
ri*,.l««. or 30*
Average 0«vlot loo fof 90 fteodlogo
Jbfl I
50
jTtfur* X • i Sraok# School Proficioocy owowinotioo shoot
12
-------
Reading Visible Emissions
i65
\ w /
/1 \
Figure X • 7. Light source should emanate from the rear of observer during daylight hours (reflected light).
Figure X - S. During darkneM, the light source should emanate from behind the plume, opposite tha obaarvtr (transmitted light)-
A
w
Figure X • 9. Readings should ba mada at right angle* to wind direction and from any diatanca necessary to obtain a clear
view of stack and background.
13
-------
166
Air Pollution Control Field Operations
i. The total number of readings deviating more
than Vi Ringelmann from the actual opacity or
density.
6. The total number of reading! deviating more than
1 Ringelmann from the actual opacity or density.
The inspector successfully completes his training
and qualifies as an expert in smoke reading if his pro-
ficiency record meets two requirements:
1. He must not deviate more than 10 per cent from
the actual opacity or density determination of the
detection system for an average of fifty readings,
and
2. He must demonstrate a consistency and reliability
in his determinations.
Even though a trainee achieves an average within
the allowable 10 per cent deviation, he does not qualify
if any reading deviates 20 per cent or more from the
actual, or his readings are inconsistent with one an-
other.
An analysis of the Proficiency Record indicates
whether or not the inspector was guessing, whether his
determinations are consistent, and whether he reads
high or low.
Weakness on any of these factors dictates the need
for further training.
2. Reading Smokt in the Field
On the basis of the training given in the smoke
school, and the accepted methods by which visual de-
terminations are made, the following general rules
apply to sight-reading in the field.
a. Reading Air Contaminants
1. Black smoke is read in densities and recorded in
Ringelmann numbers.
2. All other colored emissions are read in opacities
and recorded in percentages.
3. All opacity readings are related to corresponding
densities on the Ringelmann Chart in the follow-
ing manner:
Ringelmann
#2
*2 Vt
*3
»3J4
*4
«4H
•S
Opacitt
20%
40%
50%
60%
70%
80%
90%
100%
4.
6.
Light source should be from rear of observer dur-
ing daylight hours.
Light source should he behind plume during hours
of darkness (transmitted light).
Readings should be made at right angles to wind
direction, and from any distance necessary to
obtain a clear view of stack and background.
b. Recording Air Contaminants
1. Readings are recorded in the appropnate space on
the report or notice form as taken.
2. Observation times may be noted in terms of min-
utes and quarter-minutes, but not in terms of
seconds except in extraordinary circumstances
3. Record all emissions during observation, showing
consecutive changes in readings.
4. Total only the time in which emissions of 40%
or greater and/or No. 2 Ringelmann or greater are
observed.
5. Color of visible emissions should be recorded as
seen and as it changes.
6. It is advisable to record all or a significant portion
of the periods of excessive smoke observed during
the inspection.
7. A Violation Notice cannot be issued unless the
source emitted excessive smoke for more than
three minutes in any one hour and preferably for
more than 4 minutes. Continuing emission should
be recorded for at least six minutes of violation.
8. Any one hour means any period of sixty consecu-
tive minutes.
9. Photograph should be taken before or after but
not during visual determinations.
c. Smoke from Moving Sources
1. Smoke from tailpipe* and exhausts of vehicles is
generally read in the same way that it is read
from stationaiy sources. The observer following
or pursuing a vehicle, however, should avoid
reading directly into the plume, if possible. Line
of observation should intersect the smoke trail at
as wide an angle as possible. Error of reading
smoke in this fashion should be compensated for
2. Smoke should be read at its point of maximum
density.
3. Stop watch should be used to record accumulated
violation time.
REFERENCES
American Society at Mechanical Enginem. fount Ttn Codti.
Dtitrmiiuiniu. 19)4.
Griswold. S. S.. Ptnnelce, W. H.. McEwn, L. H., Truant 'I
Air Fallmtom t mutton. pm«m«d before 51 it Annual Mmmt
Air Pollution Control Allocution. Philadelphia, Penniylxnu,
May ». 19)8.
Hocker, Arthur J., Opmriiy Prtatipiei report to the
Dimtor ul Enforcement. 6-21-IS.
Kudlich, Rudolf, Rimgtim*** Smei* C/uri. United Sum De-
partment of the Interior. Bureau of Minn Information Grmlir
"7711, reviled by L. R Burdick, August 19)1.
14
-------
A-2
J. D. COONS,
Consulting Engineer,
H. A. JAMES,
Statistician,
H. C. JOHNSON,
Senior Engineer,
M. S. WALKER,
District Counsel,
Bay Area Air Pollution
Control District
Development, Calibration a
Use of a Plume Evaluation
Training Unit
The Bay Area Air Pollution Control District adopted a regulaii n "..•CJnii ;,lume
visibility in terms of the Ringelmann Chart and also in terms of ob-a ra'>;m ./ the ob-
server's view. Since both limitations were adopted in terms requirmi 'he ¦udqmmt of
an observer, preparation for enforcement required development and eaUuntwu ;f a train-
ing unit, so that inspectors could be qualified to testify as expert vriti*esaett under ikiz
regulation.
This paper summarizes the design, calibration and use of the Iraini ri • '".it dsvdciml
for this purpose.
la 1947, the State of California
adopted Health and Safety Code Sec-
tion 24242, which allowed an indirect
evaluation of non-black plumes against
the Ringelmann Chart under the fol-
lowing defining language:
"ia) As dark or darker in shade as
that designated as No. 2 on the Ringel-
mann Chart, as published by the United
States Bureau of Mines, or
"(b) Of such opacity as to obscure
an observer's view to a degree equal to
or greater than does smoke described
in subsection (ai of this section."
This definition was adopted by the
Los Angeles County APCD in its Rule
.50. and much the same form was
adopted by the Bay Area APCD in
ir>60 iu its Regulation 2.
Both the Ringelmann Chart and the
"equivalent obscuration" criterion re-
quire the judgment of an observer.
Their enforcement requires trained
observers whose testimony can have full
acceptance. No training method ex-
isted tor the newer concept of equivalent
obscuration.
The Los Angeles County APCD was
the first to proceed in developing a
training method, and has reported its
prpsent practice based on that work.'
A -ubstantial basis existed for the
calibration of light-transmission instru-
ments in terms of Ringelmann number,
and long experience in Los Angeles
bore out the theoretical relationship.
However, because the practical effects
of the inescapable variations in the
design and operation of a different
training unit wore not thoroughly
established, it was decided that the
Bay Area APCD unit would be cali-
brated against a group of observers
established statistically as a standard,
rather than against any "standard
instrument" or any single observer's
judgment.
This paper is a summary of the design,
calibration and use of the training unit
developed by the Bay Area APCD.
Description of Unit
Genera/
The Plume Evaluation Training Unit
consists of a gray-black plume gen-
erator, a white plume generator, in-
strumentation and accessories.
The unit was built on a trailer, and is
equipped to operate from any adequate
supply of either 110 or 220 volt alternat-
ing current. It carries its own supply
of oil and gas fuels for plume generation.
Gray-black Unit
The gray-black plume generator is
similar to a forced-air oil-fired furnace.
Oil is pumped through a burner into a
refractory-lined combustion chamber,
and the combustion products are
emitted through a twelve-inch stack.
The combustion chamber is surrounded
by a double shell of steel, with an air
space between for cooling. Two blow-
ers provide combustion air to the com-
bustion chamber, and also force air
through the cooling space; some of the
cooling air may oe exhausted into the
stack, with the combustion products, to
provide additional stack velocity. The
damper controlling this air flow has
been locked in a fixed position since
optimum position was determined in
initial operation.
Gray-black smoke is produced by
varying the degree of excess fuel sup-
plied to the burner. A hand-operated
valve permits rapid changg_cj_gjumg.
appearance from clour i.r rf-y dense
smoke.
White Unit
The white plume .s venerated by
feeding diesel oil vapinto an air
stream, where it e.>ud.ii.:es and is
emitted. Compressed a... ;n iUe sup-
ply tank forces, the die^e' ..i! through a
oil. The flow of oil vr.por from the
heater, regulated h" a hand-operated
valve, controls plui. o .usance.
A blower idschar^s or addition.
small spiral-tube heater, which
heated by propane fuel to vaporize the
Instrumentation
Two complete j:iEbmis;di>n
measurement sysc>M.< ha\j l^er. pre-
j vided, one for each plume .wnei-ator.
The systems are ider :icci in design, and
are made up of mat iied .•« nnoi.enw to
minimize differences n ¦spor.se.
The system nae&s. rp- relative
intensity of light mur-ini'.vd through
the stack on a path perpe»-ai. «ii&. :o the
flow of materia; in U\e stac-. A
standard light source nd'ataf through
the stack toward a p'.otor at the other
end of the light path. W -iations in the
amount of particuht- 'i, cr ;n the
stack cause variations in 'he intensity
of the light reaching the photocell.
The electrical signal g«v war-an by the
photocell is a linear ii uctio'i of the
light intensity reaching the photocell.
The strength of the toce!2s
checked twice a year »uyii.si known
density films, and i-e r. iaced an-
nually. No deviations i« ;*• ponse have
been noted.
May ' 965 / Volum* 15, No. 5
J99
Reprinted from APCA Journal, Vol. IS, No. S, May, 1966
-------
Calibration Procedure
The primary objective of the calibra-
tion program was to establish the
Plume Evaluation Training Unit as a
standard for the training of observers
against the legal standards set by
Regulation 2. Broadly, this required
the establishment of a statistically
self-consistent standard group of ob-
servers. and the calibration of the
training unit instruments against ob-
servations made by this group.
The major phases of the calibration
program were:
i,l) Establishment of a group of
observers as a primary standard of
Ringelmann number, by training the
group to evaluate the gray-black plumes
in direct comparison with the Ringel-
mann Chart.
12) Use of the standard group as a
primary measure of Ringelmann num-
ber, to calibrate the meter measuring
light transmission through gray-black
plumes.
(3) Establishment of the group as a
primary standard of obscuration of
view, by training the group to evaluate
obscuration of view, first by gray-black
jilumes and then by white plumes.
(4) Use of the standard group as a
primary judge of obscuration of view, to
calibrate the meter measuring light
transmission through white plumes.
In each phase of the work, there was
thorough statistical analysis of group
consistency, degree of correlation and
other statistical factors.
Selection of Group
In order to prevent any bias due to
previous training, the group was selected
from applicants having no previous
experience in Ringelmann observations.
Personnel were recruited by placing
requests with the student employment
activities of local colleges. Applicants
were required to have valid operator's
certificates from the California Depart-
ment of Motor Vehicles, to set a mini-
mum acceptable level of visual acuity.
No other requirements were set, except
the legal requirements for APCD
employment, such as U. S. citizenship
and Social Security registration.
None of the applicants had had
previous experience in the use of the
Ringelmann Chart or any other plume
evaluation method.
During the course of the work, several
of the original group left to take other
employment. As a result, final calibra-
tion of the gray-black unit was based on
results from nine observers, and final
calibration of the white unit was based
on results from seven observers.
Genera/ Procedure*
Throughout the program, the record-
ing of each observation by each in-
dividual of the group was done by that
individual. Upon the sounding of a
tell signal by the supervisor, each
individual recorded his judgment of the
Ringelmann number or equivalent ob-
scuration at that instant, using a mark-
sense card for ease of statistical work.
Thus, for each bell signal, there was a
recording from each individual. Care-
ful arrangement and supervision mini-
mised the possibility of comparison of
readings by group members.
Prior to the beginning of field work,
the group was given the following
preparation:
(1) A Ringelmann Chart was dis-
played and discussed, to familiarize
the group with its purpose.
(2) The purpose of the program was
discussed briefly in general terms, and
the importance of careful individual
work was stressed.
(3) "Flash cards" carrying various
numbers were shown to the group, and
each member recorded the numbers
on mark-sense cards as he would be
expected to do in field work; this was
done to familiarize them with the
marking practice, and to provide a
check of simple marking accuracy as a
separate factor, when judgment of
plume was not involved. This proce-
dure was carried out indoors with three
series of 20 observations each before
their first experience in plume evalua-
tion. A second group of three series
was conducted, also indoors, after
completion of 100 series (2000 readings)
on the training unit. The results were
used to estimate the amount of error
which might be expected from this
method of recording visual observations.
The flash card values were randomly
distributed by use of a table of random
numbers.
Based on the records of the seven
individuals whose data were used in
plotting both calibration curves,
overall error in this work was I.qq
percent, with 10 errors in 420 trials
(2.38 percent) on the first test series
and six errors in 420 trials (1.43 percent)
on the second test series.
Results of these trials are summarized
iu Table I.
Phase 1: Gray-black Standardization
During this phase, each group mem-
ber compared the plume from the (fray-
black unit to a Ringelmann Chart
mounted on the unit, recording his
evaluation on signal. Statistical analy-
sis of the recorded evaluations de-
termined the degree to which the mem-
bers of the group agreed with one an-
other on each reading as the work
progressed. This phase continued until
group consistency reached a satisfactory
level.
The standard deviation was com-
puted for each of the first 60 trials, and
the mean value of these standard
deviations was determined. A similar
mean standard deviation was obtained
for the last 60 readings of Phase l.
(Data from these latter 60 readings
were included in the plotting of the
curve for the gray-black unit.) These
two mean values were compared by a
statistical test known as the "t" test,
which is designed to indicate any
significant change in grouping of data.
(In this application, it would indicate
any significant change in the "scatter"
of readings from individuals from a
given trial.) A "t" value of 2.00 or
more is considered significant, and the
value found in this comparison was
9.52. The training period of Phase i
was thus shown to have significantly
improved the group's consistency jn
estimating Ringelmann value for any
single trial.
As a measure of group consistency,
Table I—Co-ordination Trial Remits
Man
No.
• First 60 Trials .
No.
Read Errors %
.—Second 60 Trials—•
No.
Read Errors %
•—Total
No.
Read
—120 Trials—„
Errors err
1
60
14
23.3
60
6
10.0
120
20
16.7
2
60
o
3 3
0
—
—
60
2
3.3
3a
60
0
0
60
0
0
120
0
0
4
SO
0
0
0
—
—
60
0
0
51'
60
0
0
60
0
0
120
0
0
6*
60
0
0
60
0
0
120
0
0
7
60
0
0
60
0
0
120
0
0
8h
60
0
0
60
3
5.0
120
3
2.)
!>"
60
4
6.7
60
2
3.3
120
*6
5.0
10"
60
3
5.0
60
0
0
120
3
2.5
11
60
0
0
60
1
1.7
120
1
0.8
12"
60
0
0
60
0
0
120
0
0
13
60
(1
0
60
1
1.7
120
1
OS
14b
60
2
3.3
60
1
1.7
120
3
2.5
15b
60
1
1.7
60
0
0
120
1
0.8
Total
900
26
2.9
780
14
1.8
1680
40
2.38
•Indicates men from whom data wis used on gray-black unit calibration only.
b Indicates men from whom data was used on calibration of both grav-hlack and white
units.
200
16
Journal of th* Air PoOwlton Control A*»ociortc«i
-------
Fig. 1. Plum* •valuation training unit.
the mean standard deviations from the
mean Ringelmann estimates were de-
rived from the individual trial standard
deviations.
This method was used with data
from 15 men for the first three series
(60 trials) of gray-black smoke training.
A similar study of data from 13 men
for the three series immediately preced-
ing the calibration trials (over 3000
trials later) showed an improvement
from 0.60 to 0.32 Ringelmann units.
The above standard deviations in-
clude observations of all participants.
Using the data from only the nine men
participating iti the actual gray-black
meter calibration gives a beginning
standard deviation of 0.50, which
improved to 0.28 immediately preceding
the calibrations.
1*1 >on completion of this phase, the
group was considered to be established
as a primary standard group, capable of
stating the shade of a gray-black plume
on the basis of the Ringelmann Chart.
This primary standard group now
became the basis for determining, in
Phase 2, the calibration curve of the
meter installed on the gray-black unit.
During this entire phase, the optical
measurement instruments were discon-
nected to avoid any possibility of the
introduction of bias by the operator or
group supervisor.
Phase 2: Calibration of Gray-black
Unit Instruments
The field operation during this
phase was exactly the same as in Phase
1, with one exception: the measuring
instruments were connected and in-
strument readings were recorded for
each signal. The instrument face, and
the values recorded, were concealed
from the standard group to avoid bias.
The statistical operation differed
from Phase 1 in that the data now
allowed two operations: first, the check
of group consistency on each reading
could be continued as in Phase 1; the
mean standard deviation of the group
of 13 men calculated from eight ran-
domly selected series was 0.34 Ringel-
mann units. Second, the calibration
curve for each of the various Ringlemann
values (as indicated by group reading)
could be evaluated statistically.
For a number o.' observations in
which the means of the observers'
readings were identical, there was
some degree of variation among the
corresponding meter readings. These
variations in meter reading were then
taken to be deviations about a mean
meter value related to a Ringelmann
number estimate. By this method, the
relationship between the observers'
estimate and the instrument reading
was established. The meter then be-
came a secondary standard suitable for
the training of new observers. This
resulted in production of a calibration
curve for the gray-black unit, as shown
in Fig. 2.
The calibration curve was based on
the estimates by nine observers over
80 series of 20 readings (a total of
14,400 points). This resulted in a
third degree curve of best fit with the
equation of M » 1.60R' — 10.60R* —
4.84 R + 94.5 where M = meter reading
and R - Ringelmann reading. These
data had a standard deviation of
10.363 meter divisions.
Results from several men were
not used in the plotting of the calibra-
tion curves for both the gray-black and
the white unit. One man was too
poorly co-ordinated to mark the card
properly, as shown by his response on
the flash card trials. Others filed
incomplete cards or were absent from
an excessive number of trials.
The standard deviation from the
mean was computed by machine for
each of the individual curves. The
mean value of all these standard devia-
tions was taken as a measure of the
expected meter reading corresponding
to the visual Ringelmann estimate by a
standard population.
Figure 2 shows the relationship
found between Ringelmann number
estimates and meter readings for gray-
black plumes.
This plot shows an essentially straight
line from Ringelmann 1 to 3 with a
slope of approximately —25 meter
divisions per Ringelmann unit.
The curvatures near Ringelmann
0 and 5 are attributed to the bias
brought about by the impossibilities of
meter readings less than 0 or exceeding
100. Although at Ringelmann 2, 3,
and 4 there are readings somewhat
plus and somewhat minus, there can be
no meter readings or observer judg-
ments below Ringelmann 0 or above
Ringelmann 5. The curve of statistical
summary therefore hendg at these
extremities.
-I 90
40
»o
20
MNMLMANN UNTO
Fig. 2. Calibration curv* far gray-black unit.
Phase 3: Group Standardization:
Obscuration
The preliminary stage of this phase
consisted of familiarizing the standard
group with the degree of obscuration
caused by gray-black plumes. This
was done by raising a target behind the
stack of the gray-black unit and in-
structing the group to note the obscura-
tion, in addition to evaluating the
Ringelmann number.
After the familiarization the group
members were asked to observe a
similar target behind the white plume,
and to record on signal the Ringelmann
number of a gray-black plume which,
in their judgment, would cause an
obscuration equivalent to that ob-
served. During this initial training
period on the white plume, the group
consistency of 11 men (two men vol-
untarily left the program to accept
other jobs), as determined by the
mean standard deviation, was 0.58
Ringelmann numbers. This phase con-
tinued until group consistency in
individual readings reached a satis-
factory level, as indicated by a mean
standard deviation value of 0.41 Ringel-
mann numbers for the 9 men continuing
the calibration test. At that point
the group was considered to be estab-
lished as a primary standard group,
capable of stating equivalent obscura-
tion in terms of Ringelmann number.
This standard group was used to de-
termine the calibration curve of the
white unit instruments, as described
in the next section.
Phase 4: Calibration of White Unit
Instruments
The field operation during this phase
was the same as that in the second
stage of Phase 3, except that the
measuring instruments were connected
and readings were recorded for each
signal. As in calibration of the gray-
black unit, the meter face and the
May 1965 / Volume 15, No. 5
17
301
-------
i i i *
RINGELMANN UNITS
Fig. 3. Calibration curv* for whit* unif.
recorded values were kept from the
standard group.
The data from this phase allowed
itatistieai determination of the proper
instrument reading to correspond to
each croup-determined value, establish-
ing a calibration curve for the white
unit, as shown in Fig. 3.
The calibration curve was baaed on
readings recorded by seven observers
over 58 series of 20 readings {a total of
>120 points). These gave a third
dearee curve of best fit having the
equation M =• 0.79R1 — 6.07R* —
•5.35 R + S5.84. where M is the meter
reading and R is the Ringelmann read-
ing. The data had a standard devia-
tion of 12.3 meter divisions.
Figure 3 shows an essentially straight
line from Ringelmann 1 to Ringelmann
4, with a sloj>e of approximately —20
meter divisions per Ringelmann unit.
The data also permitted a further
check of the Phase 3 objective of group
consistency; during this work, the
mean standard deviation of the seven
men remaining, based upon six ran-
domly selected series, was 0.45 Ringel-
mann numbers.
Use of th*i Training Unit
Training in Gray-Black Plumes
The training L'nit is the basis of
training of new inspectors, and refresher
work for experienced personnel, in the
evaluation of visible plumes.
Xew personnel are shown the Ringel-
mann Chart, and are given a period of
practice in evaluating gray-black plumes
against the Chart. They are then
shown a scries of plumes, identified by
the instructor as to Ringelmann num-
ber, on the basis of the reading of the
calibrated meter.
After this introduction, they record
their individual judgments of Ringel-
mann number of the plume each time
the signal bell is rung, while the in-
structor records the corresponding
meter value. The procedure is gen-
erally similar to that used during the
calibration process. A special form
has been developed for the training
work, on which the observer records his
readings. After each series of 25
readings, the instructor reads the meter
values for the series, and each observer
records these in a separate column next
to his own readings. This i>ermits the
observer to note his progress, and also
permits calculation of qualification
criteria, as explained below.
This training continues until the
observer qualifies in the evaluation of
gray-black plumes. Normally, an in-
experienced person will qualify after
10 to 20 series of 25 readings each.
Training in Equivalent Obscuration
When the observer has attained
proficiency in evaluating gray-black
plumes, his attention is called to the
effect of these plumes in obscuring the
view of background objects. When
necessary for this purpose, the target
behind the stack is used. The similar
effect of the white oil-mist plume is then
demonstrated, and the instructor shows
a series of white plumes, calling out the
equivalent Ringelmann values of the
plumes.
When a degree of familiarity with
the effects of the white plume has been
achieved, the observers record their
judgments of the obscuring power of the
white plume at each bell signal, in
terms of the Ringelmann number of a
gray-black plume which would give simi-
lar obscuration. The recording and cal-
culation procedures and qualification re-
quirements are identical to those for the
gray-black plume.
Qualification Criteria
The Bay Area APCD has established
certain criteria which its personnel must
meet in order to be considered qualified
observers for the evaluation of visible
plumes in terms of its regulations. The
criteria are identical for evaluation of
both gray-black and non-black plumes;
they are applied separately to the
oliserver's performance on each type of
plume.
The criteria are based on the devia-
tion of the observer's readings from the
meter values for each of the 75 reading
of a set of three consecutive series of 25
readings. For ease of calculation, each
unit of the Ringelmann scale is given
a value of 20, or five for each quarter
unit. Any deviation of 20 or more
(one full Ringelmann number or more)
disqualifies the observer. At least
two of the three series must contain no
more than two deviations as great as 15.
The average deviation of each series,
obtained by adding the absolute values
of all deviations and dividing by 25,
must be 10 or less.
After initial qualification, each quali-
fied observer must spend a minimum
of four hours in any two-month period
in refresher training, and during this
18
work must qualify on both tyjjes of
plumes. Experience has indicated
that longer grinds betwepn refresher
courses might Im>. permissible without
sacrifice of proficiency; however, be-
cause of the need for a high degree of
reliability in the testimony of these
observers, any reduction in total train-
ing must be considered carefully.
Training History
Approximately 25 enforcement per-
sonnel of the Bay Area APCD have
been trained on this unit to date;
industrial personnel, and personnel of
other agencies, have also participated
in some degree of training on the unit.
Among APCD inspectors, who are
required to maintain their qualification
as a condition of employment, there
has never been a failure to reach the
required proficiency, indicating that
there is probably no special physical
aptitude required. It has been noted
that personnel who are training pri-
marily for purposes of familiarization
frequently have somewhat greater dif-
ficulty; this may indicate that a fair
degree of motivation is required to
evoke the concentration which is es-
sential to good performance.
Among experienced APCD personnel,
the average deviation is usually in the
range of three to five, rather than the
allowable 10; and averages between
one and two for a single series occur
with some frequency.
Testimony of observers trained on
this unit has been used in many cases
before the Hearing Board of the Bay
Area APCD. Both the calibration
and the training methods have been
accepted by the Board after rather
extensive and searching cross-examina-
tion.
The unit has thus served successfully
all of its major purposes: To provide
affected industries with reliable advice
by trained personnel as to their com-
pliance or noncompliance with regy.
latory requirements concerning visible
plumes; to allow familiarization of
industrial personnel with these visible
plume standards; and to provide a
basis for acceptance of expert testimony
by APCD personnel as to violations of
the standards.
Acknowledgments
The authors wish to acknowledge
with thanks the cooperation and as-
sistance of virtually the entire staff of
the Bay Area APCD in the work re-
ported, under the direction of Mr.
Benjamin Linaky, then Control Officer.
Particular thanks are due Mr. John E.
Yocom, then Director of Technical
Services, for his advice and assistance
in the general planning of the work:
Mr. Hulan Brinkley, Senior Inspector,
for field supervision of the standard
group; Mr. Val Vitols, then Assistant
Journal of rtw Air Pollution Control Auodation
-------
Engineer, for detailed mechanical de-
sign; and Mr. Thomas Brennan,
Senior Enforcement Engineer, for de-
velopment of training criteria for
inspectors.
The various locations of the unit
during the calibration were arranged
through the courtesy of Mr. Benn
Martin, Senior Civil Engineer, City and
County of San Francisco; Mr. Marvin
Scott, Airport Manager, County of
Contra Costa; and Colonel Matsinger,
U. S. Marine Corps.
REFERENCES
1. S. Smith Griswold, W. H. Parmelee,
and L. H. McEwan, "Training of Air
Pollution Inspectors," Paper presented
at the diet Annual Meeting of the Air
Pollution Control Assoc., Philadelphia,
May, 1958.
2. Stan report, "The Construction and
Calibration of the Plume Evaluation
Training Unit" (unpublished). On
file with the Bay Area APCD, 1480
Mission Street, San Francisco, Cali-
fornia.
May 1965 / Volum* 15, No. 5
19
203
-------
A-3
Regulatory methods should aim at
acceptable air quality at acceptable
cost
PLUME VISIBILITY AS A
CONTROL BASIS
By JOSEPH D. COONS
Consulting Engineer
San Rafael, Calif.
IT is almost a tradition that any discussion of plume
visibility use as its first reference the work of Maxi-
milian Ringelmann, published in 1898.' The article is
in French and hard to locate; its chief value, perhaps,
it just as a reference to the length of time that the
Ringelmann chart has been with us.
More readily available and widely used is U.S.
Bureau of Mines Information Circular 7718, which in-
cludes a copy of the Ringelmann chart, with instruc-
tions for its use.2 This chart was developed as a method
for estimating smoke emissions from combustion op-
erations. It provides a standard visual reference
against which an observer can evaluate the darkness
of shade of a plume of smoke. It allows designation of
plume appearance at any instant as corresponding to
one of six main values, ranging from Ringelmann O
for a clear stack to Ringelmann 5 for totally black
plumes, with intermediate shades of gray numbered
from 1 through 4. Common usage today includes half-
tod quarter-values, to give a total of 21 possible
values.
Iqulvalmnt obscuration eonempt
part of California law
In 1947, the California legislature adopted a pro-
hibition3, which was based on the Ringelmann chart,
but added a new concept derived from it In addition
to prohibiting smoke of the shade of Ringelmann 2 or
darker, the law also prohibited emissions . . of such
opacity as to obscure an observer's view . . ." to the
nine degree as the prohibited smoke.
This required that observers experienced in esti-
'A preuntation at thi National IPtsltrn Mining Con/trencf, Dtnrer,
ftit—ry 1968
irBIL IMS
mating smoke by the Ringelmann chart also become
familiar with the degree of obscuration of view re-
sulting from each shade of smoke, so as to evaluate
plumes of other materials—and other colors—in terms
of equivalent obscuration. In effect, the observer, on
viewing a non-smoke plume, must be able to say, "That
plume obscures my view to the same degree as would
a smoke plume having a Ringelmann shade of 2" (or
1, or 3, etc.). The terms "Ringelmann (equivalent)"
and "equivalent opacity" have been used to indicate
this judgment.
Los Angeles County, having at that time one of the
most pronounced air pollution visibility problems in
the country, was the first to act under this legislation.
Its Air Pollution Control District pioneered in the
development of training equipment and methods for
the qualification of its inspectors as observers of
equivalent obscuration.4 Other nearby counties moved
in the same direction, some of them using the Los
Angeles training facilities.
In I960, the (San Francisco) Bay Area Air Pollu-
tion Control District adopted similar provisions. Recog-
nizing that both the Ringelmann and equivalent ob-
scuration concepts rested legally on the existence of a
"qualified observer," this District undertook a cali-
bration program* using a group of observers as a sta-
tistical "primary standard."
Experience with this new concept in Los Angeles
and the Bay Area indicates that inspectors can be
trained to evaluate non-black plumes, at or near the
legal dividing line of Ringelmann 2 (equivalent), with
a high degree of reproducibility, if they are careful
of the many variables involved in such evaluations.
The approach in both instances was aimed at the satis-
faction of legal needs, rather than development of
optical theory. Since the courts have upheld the re-
st
-------
for nonblack plumes than for black plumes and depend
on plume illuminating and viewing conditions. Be-
cause of this, nonblack plumes could be evaluated
differently when viewed on different days or vtwid
from different directions, though their aerosol content
had not changed."
suits (under rather heavy challenge in a few in-
stances), the aim seems to have been met.
Both of these areas have set the legal limits at
Ringelmann 2 and equivalent. To my knowledge, there
is no experience in equipment, enforcement or legal
precedent at any lower limit.
Validity of Rlngolmann method questioned
From its inception, the meaning and physical validity
of the phrase, "of such opacity as to obscure an ob-
server's view . . has been perplexing workers in
optics and aerosol research. As early as 1937, Marks*
raised several questions as to the validity of the Ringel-
mann method, even for smoke plumes, with optical
factors as one of his considerations.
The U.S. Public Health Service recently published a
report on an extensive study of this question, done
in cooperation with the Edison Electric Institute.T
This report is recommended for detailed study, and it
is heavily quoted here as the "PHS report" In the
authors' Abstract, the report states, "Results indicate
that visual effects are not intrinsic properties of the
plumes but vary with the background of the plumes
and with illuminating and viewing conditions. Varia-
tion was much greater with white plumes than with
black. Tests conducted with trained smoke inspectors
showed that their evaluations of nonblack smoke
plumes were significantly influenced by these varia-
tions." (Emphasis added).
Table 2 of the PHS report shows the results of evalu-
ations of white oil-fog plumes by six trained observers.
Each of these men was allowed to choose the viewing
position he thought best, and then recorded 60 evalu-
ations of the plume as its density was varied. The
calibrated plume was varied to include ten separate
occurrences of each of six values, ranging from 0.0
to 4.2 in Ringelmann (equivalent). Three of the six
observers had individual averages, on the ten occur-
rences of Ringelmann 1.00 (equivalent), of Ringel-
mann 1.80, 1.90 and 1.98 (equivalent). These were
nearly double the calibrated value. Errors at all values
tended to be larger for the white plume than for the
black, although both black and white showed serious
error as the values dropped below 2.00.
In the conclusions, the report notes that . . evalu-
ations of plumes by a visual effect are more stringent
Bsmfif of obscuration standards unclear
Thus it appears that legal requirements are satisfied
by obscuration regulations and practices In Lot
Angeles and the Bay Area, but there is considerable
uncertainty as to just what the visual observations
mean physically. Such regulations are easy to admini-
ster, since they do not require plant entry and com*
plex sampling. It is possible to adopt and enforce
standards based on obscuration by non-black plums,
but it is not clear what such regulations will accom-
plish in terms of benefit to a community.
These factors have led some other jurisdictions to
different directions. The Air Pollution Control Board
of the State of Illinois9 in 1967 adopted, a clear state-
ment that "The Ringelmann chart shall be used far
grading the light obscuring power of smoke. It shall
not be used for determining metallurgical fume enh*
sions or measuring the opacity of non-combuattaa
process emission." It defines smoke as "Small gas-
borne particles resulting from incomplete combustion
. . .". New Texas regulations contain a similar pro-
vision.9
It is of interest to note, also, that an ASME Com-
mittee recommended in 1966 that the Ringelmann
number be restricted to "the use for which it was
originally intended;" that is, to smoke.10
Contamination of air may not
constitute pollution
As a return to first principles, it may be useful hare
to review what is meant by the phrase, "control of air
pollution." The Engineers Joint Council has defined
air pollution as "the presence in the outdoor atmos-
phere of one or more contaminants ... in quantitis^
of characteristics and of duration such as to be injur-
ious to human, plant or animal life or to property, or
. . . (to) . . . unreasonably interfere with the comfort-
able enjoyment of life or property."11
In brief, air pollution consists of contaminat m of
the atmosphere to an extent which causes certain
effects; contamination below that level does not con-
stitute air pollution.
Tliis is not a new concept; it is contained, explktty
or implicitly, in most air pollution laws. It is used in
Colorado law" and many others and is the clear impli-
cation of the air provisions (in Section 107) of tilt
Clean Air Act of 1967. In effect it says, that use of tiht
atmosphere for waste disposal is acceptable, providad
the manner of disposal is such as to avoid the stated
effects. The underlying basis for this concept is worth
some brief comment.
First, as a matter of equity, the use a person my
make of that portion of the environment to which
has legal access should not be arbitrarily limited. 8
22
MIMXP. CONG It KM JODUIAt
HI
-------
his use is to be limited, it ought to be because that
use, alone or in conjunction with other surrounding
uses, causes some evident or reasonably predictable
effect beyond that which is reasonable in the circum-
sUnces.
Second, it must be recognized that all our activities,
including simple existence, do and must generate
wastes. To demand, even as a goal, that there be no
disposal of any waste to the environment of land, air
and water is to demand, in essence, the cessation of all
activity. What is needed is not the complete prohibi-
tion of all disposal of wastes to the environment, but
the management of such disposal toward the goal of
proper environmental maintenance.
Thus, the consideration of any air pollution control
reguls .ion must begin with the consideration of the
effect: which the proposal is intended to reduce or
elimin ite.
Viewing of plumes may arouse mmotions
The most obvious effect of a visible plume is that we
can see it. This is, of course, not a physical effect in
the sense of direct contact between the affected party
and the emitted material; it is an esthetic effect and
perhaps, some times, an emotional one.
The sight of steam or smoke is not harmful in and
of itself. Most complaints appear to arise from the
complainant's belief—right or wrong—that the plume
represents a threat to his health, or is an unreasonable
or deliberate act against the community.
Hie primary question then is whether the plume
constitutes air pollution. Is it an unreasonable inter-
ference with the enjoyment of life and property? In
older to answer this, some consideration must be given
to what makes the plume visible, and what it would
take to make it less visible.
Plumes are visible if they contain a sufficient num-
ber of small solid or liquid particles, with diameters
of the order of 0.2 to 0.8 microns—approximating the
wave length of visible light. There are a number of
other factors which affect the degree of visibility; as
noted in the PHS report, some of these factors are not
inherent in the plume itself.
In combustion operations, the existence of dark
plumes for extended periods usually indicates faulty
equipment or misoperation, which can be corrected by
whatever means will lead to improved, combustion. For
this reason, the Ringelmann 2 limitation on smoke,
with some flexibility for short periods of upset, has
traditionally been applied simply as a means of re-
quiring good equipment and practice. While it is easy
to generalize about the effect of such plumes on air
quality, there has been very little work toward defin-
ing these effects quantitatively; because of the relative
ease of correction, there simply hasn't been the moti-
vation to undertake such work.
When the equivalent obscuration concept is applied
to plumes of materials other than smoke, however, the
nature of the plume, its significance in air quality, and
the feasibility of correction are all much more varied.
As has been noted, the variability in visibility evalua-
tion is also much greater.
"Wet plum•" can vary greatly in appearance
One very general problem in the application of the
equivalent obscuration concept is the "wet-plume"
situation. Water droplets, in the appropriate size range,
can be about as effective as anything else in producing
visible effects, as any cloudy or foggy day well illus-
trates.
Many operations produce gas streams which are
warmer than the outside air and contain nearly their
full capacity of water vapor at the elevated tempera-
ture. Under certain rather broad ranges of outside
temperature and humidity, the emitted water vapor
will condense, producing huge masses of visible water
droplets. Thus a highly visible plume results, even
though the gas stream may contain no particles in the
light-scattering size ranges, as emitted. Such an emis-
sion can have extreme variations in appearance from
day to day, or even hour to hour, with no change in
actual emissions.
In the usual case, of course, there are some other
particles present. It then becomes a problem to deter-
mine whether the plume appearance would be accept-
able if the water were not present. Yocom has devel-
oped a prototype device for such situations,13 but it has
not, to my knowledge, been used officially anywhere.
Most agencies appear to handle such problems "ad-
ministratively." I know of no agency which in practice
prohibits steam emissions from cooling towers; on the
other hand, I know of no agency which has regulations
drawing a clear and practical line between acceptable
"wet-plumes." There have recently been some inter-
esting and informative papers on the dilemma faced by
the operator—and the control officer—in such cases.14'14
Passing on from the visibility of the plume itself, we
come to a consideration of what happens to the visible
particles as they leave the vicinity of the stack.
Al'KIL ions
23
-------
terials for these reactions, is also a source of particles
emitted directly in its exhaust. Some types of vegeta-
tion also emit sufficient raw materials to cause forma-
tion of photochemical particles and haze as a natural
j^enomenon.
These sources do not have a constant or fixed part in
visibility reduction, even in a given area. The windy
days which sweep out an area, preventing photo-
chemical particle foi-mation and accumulation of other
particles, may at the same time significantly increase
the total particle load by picking up large amounts of
surface dust. Widespread hazes observed in many areas
appear to develop through natural phenomena, al-
though very little is known of their causes and the
nature of the particles involved.
plum* vis bility poor measure of
parti
-------
ing the source, and become a part of the total load oI
visibility reducing particles in that area. This is tlx
principal air quality effect which is ascribed to these
particles.
For specific materials, there may be specific toxic
effects of emitted particles. Control of these toxk
agents to maintain safe levels is ordinarily accom-
plished by limiting the amounts or concentrations with-
out regard to particle size, since there is little pres-
ent basis for distinguishing between the various sixes
as to toxic effect.
If water droplets are present in the plume, the dis-
tance over which they will persist is dependent largely
on the atmospheric humidity and temperature, al-
though evaporation may also be significantly influenced
by the nature of the other materials present. The ulti-
mate evaporation of the water normally produces no
important atmospheric effects, either desirable or un-
desirable.
'I particles behave mora likm gat
1 effects of other small particles are not so easily
evaluated.
Because volume and weight vary with the cube of
the diameter, the weight of particles diminishes rapidly
as the diameter decreases. Thus, for a given material,
it takes one million particles with a diameter of 0.2
micron to equal the weight of one particle of 20 micron
diameter. In air quality sampling, or source sampling,
these small particles will usually make up a negligible
raction of the weight of material collected.
Further, these small particles are approaching
molecular size, being perhaps 200 to 500 times the
diameter of the larger gas molecules, and only
slightly larger than the average distance between gas
molecules. Thus, they behave more nearly like gas
molecules than do the larger particles; they settle
very slowly in air, and their travel is much more
affected by collisions with gas molecules.
With their low weight and slow settling properties,
these particles do not cause dustfall problems near
the stack. Settling and impaction appear to be negli-
gible factors in their removal from the atmosphere,,a
so that they are presumably much less significant in
the soiling of surfaces than are particles of perhaps 2
to 5 micron diameter.
These particles travel away from the source as a
plume, gradually dispersing until they are no longer
visible as a plume. The rates of travel and dispersal
depend chiefly on wind speed and atmospheric turbu-
lence. Over a variable and uncertain period, which
has been estimated"1 as on the order of a few days on
the average, they are removed from the atmosphere.
The chief mechanisms of removal for particles of this
size appear to be either incorporation in raindrop
nuclei or growth by agglomeration to a settleable size.
If ventilation is sufficiently poor, however, they may
accumulate in the atmosphere of the area surround-
Need tor fine particle control herd to assess
The final questions are those of the necessity and
feasibility of control of these small particles. In the
present consideration, these are two-pronged ques-
tions: We must first consider the necessity or feasi-
bility of fine particle control itself and then whether
the use of plume visibility as a criterion is likely to
meet the needs.
The necessity for control of fine particles emitted
in visible plumes is difficult to assess accurately. The
extreme goal of air purity would permit no emissions;
the extreme goal of waste disposal would permit un-
limited emissions. Somewhere between these two ex-
tremes lies the best goal of limited disposal with limited
effects.
As has been noted, the simple fact that the plums
can be seen may, in some instances, be regarded as un-
desirable and indicative of the need for control Where
the control is not too difficult or costly, a community
may decide that no further justification is required.
This has been the history of use of the Ringelmarm 2
limit on smoke from combustion operations. In cases
where the control is more difficult or costly, a much
more thorough appraisal may well be required.
Where general reduction in visibility in an area la
the concern, the importance of the visible plume de-
pends chiefly on the relative importance of the particles
in such plumes as compared to the total of all par-
ticles, from all sources, on the days when undue re-
striction of visibility is noted.
In evaluating this total particle load, consideration
must be given to natural sources—fog, windblown
dust, pollen, sea-salt nuclei, and others—and to
human activities which may intensify these by char {-
ing the nature of the surface, or of the vegetation, e i-
posed to action of the wind.
Man-made particle loads will include, of course, the
particles emitted as visible plumes. Other sources,
which will have vai-ying significance in varying areas,
include dust from vehicular traffic on unpaved or dusty
roads, and from agricultural or earth-moving opera-
tions. Demolition work and open fires would also con-
tribute to this load.
Photochemical reactions major cause of hates
A major source of visibility interference in many
areas is the formation of particles in the atmosphere,
generated as by-products of photochemical smog re-
actions. The automobile, a chief source of raw raa-
Z 5
MINING CONQBKIIS JOURNAL
-------
A* a consulting engineer tor the past
four year*, Joseph D. Coons has been
particularly active in the development
of air pollution legislation and regula-
tions and in their administration and
enforcement. He has worked with local,
state and federal agencies and legis-
lative bodies. Before starting his own
consulting practice in 1964, Coons was
director of enforcement for the Bay
Area Air Pollution Control District in
San Francisco for four years. He is currently associated
with the district as the professional engineer member of
its advisory council.
These questions are not new; they have nearly all
been raised, in one forum or another, since the first
use of the equivalent obscuration concept However,
such recent work as the Public Health Service report
cited appears to give them substantially greater weight.
Yet some jurisdictions have used both the Ringel-
mann and the equivalent obscuration limits, at the level
of 2, for a number of years, with success in the courts.
Most control agencies recognize that these difficulties
exist. However, many of them are understandably re-
luctant to give up a control tool, which is administra-
tively easy and legally enforceable, until some satis-
factory substitute can be found.
Lmgitlativ* bodfat should preM«( with cor*
In these circumstances, it would seem prudent for
legislative bodies considering such questions to pro-
ceed with care. The severe errors in judgment by
trained observers at values below Ringelmann 2
(equivalent), and the uncertainties in both necessity
and feasibility of stringent controls, argue strongly
against any use of lower limits until some of these
questions can be resolved.
Two basic questions must be weighed carefully in
any deliberations as to the need for or effectiveness of
this type of emission limitation.
First, what is the actual need for control of fine par-
ticles from non-combustion sources? This requires
evaluation of the severity of visibility reduction in the
specific area, and of the relative importance of all the
various sources of particles contributing to such visi-
bility reduction.
Second, what are the practical capabilities for con-
trol of fine particle emissions from the affected opera-
tions in the area? Fundamentally, of course, no regula-
tion should be adopted which cannot be met. Beyond
this, however, it is an accepted principle that where
regulations impose burdensome and disruptive re-
quirements, the need for the regulation should be
sufficient in both magnitude and certainty to justify
the action.
Answers to these questions will not be firm. They
will be judgments based on estimates which are them-
selves difficult to develop. But their consideration will
26
lead to evaluation of specific problem situations in the
area and to a better understanding of those situations*
Such understanding should limit the application of
this control tool to those cases where considered judg-
ment indicates it is needed and feasible.
It is to be hoped that it will also lead to development
of other regulatory methods which are more certain hi
their application and effect and which achieve ac-
ceptable air quality at acceptable cost
RsmncNCES
1 Ringelmann, M. "Method of Estimating Smoke _ ....
by Industrial Installations," Rmvuo Technique, 268, Ji
1898*
2 Kudlich, R. (Rev. by L. R. Burdick). "Ringefan^^
Smoke Chart," E/.S. Bureau of Minat Information Ciro^^
7718, August 19SS.
3 Stat* of California, Health and Safety Code Chanter i.
Division 20, Section 24242 (1947).
* Oriswold, S. S.. W. H. Parmelee, L. H. McEwan. "Ttal*.
ing of Air Pollution Inspectors." Paper presented at Aom
Meeting, Air Poll. Control Assoc., Philadelphia, May 19SS.
5 Corns, J. D., H. A. James, H. C. Johnson, M. S. Walker.
"Development, Calibration and Use of a Plume Evaluate*
Training Unit," /. Air Poll. Control Amoc., May 1965.
'Marks, L. S. "Inadequacy of the Ringelmann
Mecfi; Eng., September 1937.
T Conner, W. D., J. R. Hodkinson. "Optical Propertia* aatf
Visual Effects of Smoke-Stack Plumes," Publication Nol 999.
AP-30, U. S. Public Health Service. Cincinnati, 1967.
I Stat* of Illinois, Air Pollution Control Board, Regoletfaa
3-3.121 and 1-1.43 (1967).
'State of Tens, Air Control Board, Regulation No, 1
(1967).
10 "Recommended Quide for the Control Dust Emisaie»_
Combustion for Indirect Heat Exchangers," ASK E SM>
ardNo. APS-1 (1966).
n Bishop, C. A., "EJC Policy Statement on Air Pothrtfaa
and its Control," CAem. Eng. Progr. 53:11 (195<); or em
Heating, Piping & Air Conditioning, May 1958.
" State of Colorado, Air Pollution Control Act Sectlea
3(6).
II Yocom, J. E., "Problems in Judging Plume Opedty—A
Simple Device for Measuring Opacity of Wet
J. Air Poll. Control Atao
-------
A-4
Excerpt from EPA Report No. 650/4-74-005i, Guidelines for Development
of a Quality Assurance Program: Visual Determination of Opacity Emissions
from Stationary Sources, November 1975
METHOD 9—VISUAL DETERMINATION OF THE
OPACITY OF EMISSIONS FROM
STATIONARY SOURCES
Reproduced from Appendix A, "Reference Methods,"
Federal Register, Vol 39, No. 219; Tuesday, November 12, 1974
27
-------
5.8.1.1 Variables Accuracy-
Many ststlonary sources discharge visible
emissions lota the atmosphere; these emis-
sions are usually In the shape of a pluffie.
This method Involve* the determination of
plume opacity by qualified observers. The
method Includes procedures for the training
and certification of observers, and procedures
to be used In the Held ror determination of
plume opacity. The appearance of a plume as
viewed by an observer depends upon a num-
ber of variables, aome of which may be con-
trollable and some of which may not be
controllable In the Held. Variables which can
be controlled to an eitent to which they no
longer eurt a significant Influence upon
plume appearance Include: Angle of the ob-
server with respect to the plume; angle of the
observer with respect to the sun: point of
observation of attached and detached steam
plume; and angle of the observer with re-
spect to a plume emitted from • rectangular
stack with a large length to width ratio. The
method Includes specific criteria applicable
to than vvUblit.
Other variables which may not he control-
lable In the field are luminescence and color
contrast bstween thi plume and the back-
ground against which the plume Is viewed.
These variables exert an Influence upon the
r.ppearance of a plume as viewed by an ob-
server, and can affect the ability of the ob-
server to accurately assign opacity values
to the observed plume. Studies of the theory
of plume opacity and field studies have dem-
onstrated that a plume Is most visible and
presents the greatest apparent opacity when
viewed against a contrasting background. It
follows from this and Is confirmed by field
trials, that the opacity of a plume, viewed
under condltlone when a contrasting back-
ground I* present can he assigned with the
greatest degree of accuracy. However, the po-
tential for a positive error Is also the greatest
when a plume la viewed under such contrast-
5.8.1.2 Principle and Applicability-
1. Principle and applicability.
1.1 Principle. The opacity of rmlMtooe
from stationary sourcee Is determined vis-
ually by a qualified observer.
1.3 Applicability. This method 19 appli-
cable for the determination of the opacity
of emissions from stationary sources pur-
suant to I 60 11(b) and for qualifying ob-
servers for visually determining opacity of
emission*.
lng conditions. Under conditions presenting
a less contrasting background, the apparent
opacity of a plume Is less and approaches
«ero as the color and luminescence contrast
decrease toward zero. As a result, significant
negative bias and negative errors can ha
made when a plume la viewed under less
oontrsstlng conditions. A negative bias de-
creases rather than Increases the poeslblllty
that a plant operator will be cited far a vio-
lation of opacity standards due to observer
error.
Studlee have been undertaken to determine
the magnitude of positive errors which can
be made by qualified observer* while read-
ing plumes under contrasting conditions and
using the procedures set forth In this
method. The results of these studies (field
trials) which Involve a total of 709 sets of
28 readings each are as follows;
(1) For black plumss (133 sets at a smoke
generator), 100 percent of the sets were
read with a positive error' of less than TJ
percent opacity; M percent were read with
a positive error of .less than 5 percent opacity.
(3) For white plumes (170 sets at a smoke
generator, 1M sets at a coal-fired power plant,
391 sets at a sulfuric acid plant), 89 percent
of the seta were resd with a positive error of
less than 7J percent opacity; 93 percent were
read with a positive error of less than S per-
cent opacity.
The positive observational error associated
with on average of twenty-five readings Is
therefor* established. The accuracy of the
method must be taken Into account when
determining possible violations, of appll-
cable opacity standards.
'For a set, positive error—average opacity
determined by observers' 33 observations-
average opacity determined from transmla-
someter's 38 recordings.
5.8.1.3 Procedures for Opacity Observations-
3. Procedures. The observe* qualified la
accordance with paragraph S of this method
shall use the following procedures for vis-
ually determining the opacity of emissions:
3.1 Position. The qualified observer shall
stand at a distance sufflcleot to provide a
clear view of the emissions with ths sun
oriented In the 140* sector to his hack. Con-
sistent with maintaining the above require-
ment, the observer shall, as b»u«* " Possible,
make hu observations from a position such
that bu line. Of vision ts 2
perpendicular to the plunie dl'eeiton, and
when observing opacity , 1^2
rectangular outleU (e.g. root monitor*.
baghouses. nonclrcular stsejs). approal-
ntstely perpendicular to the
the outlet. The observer's »»• SSj?
not Include mole than one
when multiple stacks ars 'J**"'*?! £?. „L
any esse the observer shoi>J*?>
serrations with hb line of
lar to the longer axis of such a setof mulU
pis stacks (e.g. stub stacks 00 baghonees).
3.3 Field records. The observer shall re-
cord the name of the plant, emission loca-
tion, type facility, observer's name and
affiliation, and the date on a field data sheet
(Figure 9-1). The time, estimated distance
to the emission location, approximate wind
direction, estimated wind speed, description
of the sky condition (presence and color oi
clouds), and plume background arc recorded
on a field dst« sh*et at the time opacity read-
ings are initiated and completed.
3.3 Observations. Opacity observations
shall be made at the point of greatest opacity
in that portion of the plume where con-
densed water vapor la not present. The ob-
server shall not look continuously at the
plume, but Instead shall observe the plume
momentarily at 18-second Intervals.
3.3.1 Attached steam plumes. When con-
denssd water vapor Is present within ths
plume ss it emerges from the emission out'
let, opacity observations shall be mad* be-
yond the point In the plume at which con-
densed water vapor la no longer visible. The
CO
-------
5.8.1.3 Procedures (cont.)-
obaeTver thill record th* ipproiliMl* dla-
tance from the emission outlet to the point
to the^pluma at which th* observation* tn
mad*.
2 3-2 Detached ilwn plum*. When water
vapor la the plum* conctens** and become*
visible at a distinct dhtua from tt* emis-
sion outlet, the opacity of mbatonl should
be evaluated at the emlMloa outlet prior to
the condensation of water hw* >wt Ibt for-
mation of the atearn plum*.
2.4 Recording observation*. Opacity on-
servatlon* ibtll be recorded to the nearest •
percent at lS-second Intervals on an ob-
eervatlonal record aheet. (fie* Figure 0-3 for
an example.) A minimum of 3* observation*
shall be recordsd. Bach momentary observa-
tion recorded shall be deemed ,to represent
tli* aveng* opacity of cmlsrtooe tor a 15-
secood period.
M Data Reduction. Opacity usll be d*»
tarmlaad a* *a aver*#* at it eoneeeutlv*
ob—f ration* reoorded at ll-eeoond interval*.
Divide it* obaerratlona recorded on Ut* rec-
ord itmt Into aet* at M consecutive obser-
vations. A aet le oompoaad of any 94 eon-
•ecutlfe observation*. Set* need not be eon*
secuttve la time and tn no cue (ball two
set* overlap. Far each aet of 34 observation*,
calculate the average by summing the opacity
ot the 34 observation* and dividing Uila sum
by 34. If an applicable standard specifics an
averaging time requiring mora than 34 ob-
servation*. calculate the avrrage tor all ob-
servations mad* during tb* apaeifled time
period. Record the average opacity an a record
aheet. (See Plgur* 9-1 for an example.)
5.8.1.4 Qualifications and Testing-
3. Quallflcation$ and testing.
S I Certification requirements. To rteelv*
ccrtifUntloa as a qualified observer, a ran'
dldate must be tested and demonstrate liie
ability to assign opacity readings In B percent
Increment* to 25 different black plume* and
35 different wblta plume*, with an error
not to exceed IS percent opacity on any one
reading and an average error not to exceed
7 5 percent opacity In each category. Candi-
date* ehalt ba tasted according to the pro-
cedures described in paragraph 3 2. Sunk*
generator* used pursuant to paragraph 9.3
•hall b« equipped with a amok* mater whleh
meets the requirements of paragraph t J.
The certification shall be valid for a period
of 4 months, at which time the qualification
procedure must h* repeated by any observer
In order to retain certification!
5.8.1.5 Certification Procedure-
3.2 Certification procedure. Tlie certifica-
tion te»t corulsta of allowing the candidate a
complete run of 50 plumes—28 black plume*
and 20 will to plumes—generated by a smoke
generator. Plumes within each aet of 25 black
and IS white ruua shall be presented In ran-
dom order. The candidate nmlgn* an opacity
vaIu* to each plume and record* Ills obser-
vation on a suitable form. At th* completion
of each run of 50 reading*, th* score of th*
candidate Is determined If a candidate fall*
to qualify, the complete run of 50 reading*
inu«t be repeated In any rctest. The «mok*
test may be admlntstcred as part of a smoke
school or training progrnm, and may be pre-
ceded by training or familiarization run* of
th* smoke geuorator during which candidates
are rhown black and white Dlumas of known
opacity.
-------
5.8.1.6 Smoke Generator Specitications-
IJ tank* generator ipec motions. Any
¦aofce puntor uaad for Uu puifom 0 display la*
stack opacity btitd upon t pathlength equal
to the stack exit diameter, on a full 0 to 100
percent chart recorder real*. The smoke
outer optical design and performance shall
meat the specifications shown la Table >-1.
TIm smoke meter iluUI be calibrated aa pre-
aenbed la paragraph M.I prior to tht oon-
duot of each amok* reading teat. At the
oomplotlon of each test, the aero and spea
drift eball be checked and If the drift ex-
eaeda ±1 pert*nt opacity, the coodltloo shall
be corrected prior to conducting any subse-
quent teat ruas. The smoke meter shall ba
demonstrated, at the time of Installation, to
meet the specification* listed la Table 0-1.
Title demonstration shall h* repeated fol-
lowing any subsequent repair or replacement
at the photocell or associated electronic cir-
cuitry Including the chart recorder or o«tp«3
meter, or every • imbMih vhlekmr wm
fin*
ruu e-i—won Mem Desjcw urm
mmiuxcf erscmcsnoiis
Para outer: Speci/loadon
a. Light eouroe Incandescent lamp
operated at nominal
b. Spectral response
of photooelL
«. Angle of rleir
d. Angle of projec-
e. Calibration error.
5.8.1.7 Calibration of
f. Zero and spaa
drift
g. Reapome time...
Smoke Meter-
rated voltage.
Photoplo (daylight
spectral response of
the human eye—
reference 41).
II* maximum total
angle.
II' maximum total
angle.
±3% opacity, maxi-
mum.
±1 * opacity, 30
mlnutea.
£S seoonda.
3JJ Calibration. The smoke meter Is
calibrated after allowing a minimum of "SO
minuses wannup by alternately producing
aUnulated op&cltyof 0 percent and 100 per-
cent. When eLable response at 0 percent or
100 percent-ft- noted, the (moke meter la ad-
Justed to produce aa output of 0 percent or
100 percent, aa appropriate. This calibration
shall be repeated until stable 0 percent and
100 percent readings are produced without
adjustment. Simulated 0 percent and 100
percent opacity values may be produced by
alternately switching the power to the light
source on and off while the smoke generator
is not producing smoke.
5.8.1.8 Smoke Meter Evaluation-
3.3.2 Smoke meter evaluation. The imoke
meter design and performance are to be
evaluated as follows:
3.3.3.1 Light source. Verify from manu-
facturer's data and from voltage msasuie-
mciits nude at the lamp, as Installed, that
the lamp Is operated within ±t percent of
the nominal rated voltage.
3.1.2.3 Spectral response ot photocell.
Vertfy from manufacturer's date that the
photocell has a photoplc response; l.e„ the
spectral sensitivity ot the cell shall closely
approximate the standard spectral-luminos-
ity curve for photoplo vision which Is refer-
enced la (b) of Table 3-1.
3.S 21 Angle of view. Check construction
geometry to ensure that the total angle of
view of the smoke plume, as seen by the
photocell, dots not exceed !•*. The total
angle of view may be calculated from: f=*
t*a-> d/2U where f—total angle ot view;
d^the sum of the photocell dlametor+the
diameter ot the limiting aperture; and
L = the distance from the photocell to the
limiting aperture. The limiting aperture la
the point la the path between the photocell
and the smoke plume whe<* the angle ot
view is most roslrlcted. Tn smote generator
smoke meters this Is normally aa ortflcs
plate.
3.3.3.4 Angle of projection. Check con-
struction geometry to ensure that the total
angle of projection of the lamp on the
smoke plume does not exoeed IS*. The total
angle of projection may be calculated front:
1-3 tan ' d/2L, where #= total angle of pro-
jection; d=> the sum of the length of the
lamp filament + the diameter of the limiting
aperture; and L== the distance from the lamp
to the limiting aperture.
3.3.3.3 Calibration error. Using neutral*
density Alters of known opacity, check the
error betwooa the actual response and the
theoretical linear response of the smoke
meter. This check Is accomplished by first
calibrating the smoke meter according to
3.3.1 and then Inserting a series of three
neutral-density Alters of nominal opacity of
30, 80, and 75 percent la tlie smoke meter
pathlength. Filters cnllbarted within ±3 per*
cent shall be used. Care should be taken
when Inserting the niters to jnevent stray
light from affecting the meter. Make a total
ot Ave noaconacciitlre readings' for each
filter. The maximum error on any oae read*
lag ah all be 3 percent opacity.
3.3 J.I Zero and spaa drift. Determine
the aero and span drift by calibrating and
operating the smoke generator la a norma
manner over a l-hour period. The drift Is
measured by checking the aero and spaa at
the end of thli period.
33.3.7 Response time. Determine the ie*
spouse time by producng the series of five
simulated 4 percent and 100 percent opacity
value* and observing the tUne required to
reach stable response. Opacity values ot o
percent and 100 percent may he simulated
by alternately switching ths powr_- to the
light souros off 'and on "Mle the smoke
generator Is not operating.
o «
-
-------
FIGURE 9-2 OBSERVATION RECORD PAGE OF
COMPANY OBSERVER
LOCATION TYPE FACinTT
TEST NUMBER POINT OF EMSSTOT
PATE
Hr.
Mln.
Seconds
JTEAm PluhE
(check 1f applicable)
COMMENTS
0
15
30
lb
Attached
Detached
0
1
2
3
4
5
6
7
8
9
!?
1
12
13
1}
J
16
J7
i$
fa
Jo
St
a
a
24
?*
V
gft
29
IMUl HWIR, VOL M,
FIGURE 9-2 OBSERVATION RECORD PAGE OF
(Continued)
COMPANY OBSERVED
LOCATION TYPE FA. ICTTy
TEST NUNBEK POINT OF EMISSTCST" . |
PATE ~
Hr.
Kin.
Seconds
" STEAM MUMf
(check If applicable)
COWCNTS
0
IS
JO
45
Attached
Detached
30
3)
32
33
34
15
36
37
is
'40
4)
42
45
44
45
46
4)
46
44
SO
SI
52
S3
54
55
Si
S>
Ss
|>B DOC.T4-M1M FUad UBl
21*—TUCJDAV, NOVlMMi IS, lff4
-------
FIGURE 9-1
RECORD OF VISUAL DETERMINATION OF OPACITY
PAGE Of
COMPANY
LOCATION
TEST NUMBER,
DATE
TYPE FACILITY
CONTROL DEVICE_
HOURS OF OBSERVATION,
OBSERVER
OBSERVER CERTIFICATION DATE_
OBSERVER AFFILIATION
POINT OF EMISSIONS ¦
HEIGHT OF DISCHARGE POINT
CLOCK TIME
H OBSERVER LOCATION
Distance to Olscharge
Direction froa Discharge
Height of Observation Point
8ACKGR0UN0 DESCRIPTION
WEATHER CONDITIONS
Wind Direction
Wind Speed
Ambient Temperature
SKY CONDITIONS (clear,
overcast, i clouds, etc.)
PLUK DESCRIPTION
Color
Distance Visible
OMR i.»jfor:iatiom
Initial
Final
SUMMARY OF AVERAGE OPACITY
Set
Number
Tim.
ODJCltS
Start—End
Sum
'verage
Readings ranged from to i opacity
The source was/was not In compliance with .at
the time evaluation was made.
-------
SECTION B: CHALLENGES TO THE ACCURACY AND VALIDITY OF VISUAL
PLUME READINGS
Page No.
B-l. "Inadequacy of the Ringelmann Chart," Lionel S. Marks,
Mechanical Engineering, 1937 . 35
B-2. "Smoke Readings Vary With Observers" Electrical World,
January 15, 1975 . . 41
B-3. "The Ringelmann Number as an Irrebutable Presumption of Guilt—
An Outdated Concept, Donald J. Henz, Natural Resources Lawyer.
May 1970 45
B-4. "Visual Plume Readings—Too Crude for Clean A1r Laws," Raymond
E. Haythorne and James W. Rankin, Natural Resources Lawyer,
Summer 1974 55
-------
B-l
INADEQUACY
RINGELMANN CHART
By LIONEL S. MARKS
GRADUATE SCHOOL OF ENGINEERING, HARVARD UNIVERSITY. CAMBRIDGE, MASS
DETERMINATION of smoke density by comparing the
apparent density of smoke as it issues from a stack with
a Ringelmann chart located 50 ft from the observer has
become almost standard in the United States. The object of
this article is to demonstrate that such a comparison does not
give any reliable information about smoke density and that
the incorporation of this procedure in smoke ordinances either
should be abandoned or should be qualified as indicated later.
What is written here about the inadequacy of the Ringelmann
chart applies in part or in whole to other visual methods for
the determination of smoke density.
The first definition of smoke in Webster's New International
Dictionary will presumably be acceptable both to laymen and
to engineers. It is as follows:
Smoke—The gaseous products of burning organic materials, as wood,
coal, peat, tobacco, etc., when rendered visible by the presence of
small particles of carbon which finally settle as ton. Smoke is thus
always the result of imperfect combustion.
"Imperfect combustion" can be accepted as including such
phenomena as the thermal decomposition or "cracking" of
hydrocarbons which undoubtedly accompanies combustion in
many cases and may be responsible for the liberation of soot.
Density of smoke is the weight or volume of soot in each
cubic foot of the gases issuing from a smokestack. The com-
mon belief that the appearance of smoke is an indication of its
density is entirely fallacious. Smoke of a definite density may
have any color from light gray to black, depending on the
conditions under which it is observed.
FACTORS AFFECTING VISUAL DENSITY OF SMOKE
Darkness or visual density of smoke depends on the amount
of light coming through it. If no light comes through it, it is
black. Light which penetrates the smoke comes from the
background of sky and will vary with the brightness of the
sky. The fraction of the light of thesky which penetrates will
depend primarily on two factors: true density of the smoke
and thickness of the smoke screen. Some other factors that
influence the appearance of smoke are the effect of (») wind,
(4) the presence of steam, and (c) reflected sunlight when the
sun is behind the observer. In general, darkness or visual
density of a column of smoke depends primarily on the first
three factors mentioned, brightness of the sky, true density of
the smoke, and thickness of the smoke column.
To demonstrate the influence of thickness of the smoke
column on visual density, the equipment shown in Fig. 1 was
set up in the laboratory of che Brooklyn Idison Company.
Smoke was generated by spraying a mist of turpentine over an
active tire and the products of combustion were stirred vigor-
ously by a tan in the base of the large cylinder shown in the
figure, and were then discharged through the three stacks
shown. The actual density of the smoke must have been the
same in all the three stacks. A comparison of the visual densi-
ties close to the stacks, before dilution with the air into which
FIG. 1 INFLUENCE OF STACK DIAMETER ON APPARENT SMOKE
DENSITY
(Actual composition of the smoke is the same for all three stacks.)
31
35
-------
682
0
Mechanical Engineering
PIO. 2 EfPECT Of THICKNESS of a smoke column on the adsorption op a beam op light
(Intensity of the light beam and »«uil smoke density ire the same in both views, but, by routing the rectangular smokestack through an
angle of 90 deg, thickness of the smoke column was reduced from 12 in. in view at left to less than ) in. in view at right.)
the smoke is discharged, shows a Ringelmann density of 5 for
the largest and between 1 and 2 for the smallest stack.
Another demonstration of the influence of the thickness of
the smoke column on the absorption of light passing through it
is shown in Fig. 2. This demonstration was also carried out in
the laboratory of the Brooklyn Edi$on Company In this case
the density of the smoke was determined by a General Electric
photoelectric cell. Light from a source to the right of the
smoke column passes in a parallel beam through the smoke
to a photoelectric cell at the left of the picture. The intensity
of the beam of light arriving at the cell is shown by the indi-
cating instrument located above the clock. The smoke issues
through a rectangular stack of dimensions )', j X 12 in. which
is rotatable through an angle of 90 deg. The two photographs
of Fig. 2 were taken 19 sec apart and show the instrument indi-
cations (a) when the light traversed 12 in. thickness of smoke
and (4) after rotation of the stack through 90 deg so that the
light traversed less than 3 in. thickness of smoke. The con-
36
-------
September, 1937
683
traction of the issuing smoke stream is evident in the photo-
graph and reduces its thickness below the stack dimensions.
The corresponding instrument readings are seen to be about 4
and 1. These can be considered as giving the relative fractions
of the light absorbed in the two cases or as measures of the
apparent smoke densities.
Another demonstration of the influence of stack diameter on
apparent density of smoke uses a "liquid smoke" made from
India ink and water. India ink consists of soot, obtained from
the thermal decomposition of natural gas, suspended in a liquid
I/a few drops of India ink are added to a large quantity of water
and the latter is then stirred, the soot will remain uniformly dis-
tributed (or a long period. Individual soot particles are ex-
ceedingly fine—beyond the range of the ordinary powers of a
microscope From the point of view of its constitution this
liquid smoke is the equivalent of ordinary smoke but with the
substitution ot a colorless liquid, water, for colorless gases. As
this liquid smoke is used here only for comparative demon-
strations, us use is believed to be |ustitied.
Fig. 3 shows four glass cylinders, approximately 1, 2, 4, and 8
in. in diameter The largest was filled with about 20 lb of
water and 15 drops ot India ink were added and well stirred.
Some of the resulting liquid was then transferred to the other
cylinders as shown in Fig. 3. The white board behind the
cylinders was brightly illuminated. The photograph indicates
apparent densities ot this liquid smoke varying from No. 5
Ringelmanr: chart lor the largest diameters down to about No.
1 for the smallest diameters. The liquid sinoke, however, had
the same actual density in all four cylinders.
Translated into corresponding chimneys, this indicates that
smoke of No 1 Ringelmann-chart density issuing from a 2l/r
ft stack miehr appear as dark as No. 5 when issuing from a 20-
ft stack. With the same gas velocities, a 20-ft stack handles
64 times as much gas as a 2l/rft stack. Consequently a 20-ft
stack discharging smoke of No. 5 density could be replaced by
sixty-four 2l/rft stacks, and in that case the smoke might ap-
pear to have a density not greater than No. 1. These demon-
strations indicate clearly that a visual determination of smoke
density has no significance in terms of actual smoke density un-
less proper allowance is made for the influence of stack diameter
on the apparent density.
Liquid smoke can also be used to demonstrate the influence of
the brightness of the sky background on the apparent density
of smoke. In Fig. 4, the three smallest cylinders are shown as
in Fig. 3 with unchanged contents. The only change is that
a gray background is substituted for the upper half of the white
background. The influence of this change on the apparent
density of smoke is evident. In these cylinders the upper halves
ot the liquid have apparent densities that are much greater than
in the lower halves; in the largest cylinder, which is not shown,
no change appears, since in this case all the light is intercepted.
This demonstration indicates clearly that a visual determination
of smoke density has no significance in terms of actual smoke
density unless the condition of the background is determined,
as, for example, by a light-intensity meter such as is used by
photographers, and proper allowance is made for any departure
from a standard brightness of the background.
The effect of wind velocity on apparent density of smoke
may sometimes be considerable. On an absolutely still day,
smoke ascends vertically with a velocity that diminishes as it
rises, in consequence both of cooling and of spreading. Fur-
thermore, cooling diminishes the gas volume and therefore in-
creases the soot content per cubic foot or the actual smoke den-
sity. With a brisk wind, smoke is immediately diluted with
FIG. 3 INFLUENCE OF DIAMETER OF CONTAINING VESSEL ON APPARENT DENSITY
(Composition ot liquid smoke," i mixture of India ink and water, in all four vessels is identical.)
?7
-------
684
MechaniCal Engineering
'I'-##*w$8i **%
3§ptSpp3
M?#ll
ImAs
^&UBUW<
mmm^
FIG. 4 INFLUENCE OF THE BACKGROUND ON APPARENT DENSITY Or "LIQUID SMOKb"
air on leaving the stack and its actual and apparent densities
arc diminished. It is evident that smoke issuing from a stack
with a given actual density may appear to be very much denser
on a calm day than on a windy day.
VISUAL DENSITY THE ONLY CONCERN OP SMOKE ORDINANCES
Smoke ordinances of American cities concern themselves only
with the apparent or visual density of smoke. The purpose of
the ordinances is to control an assumed menace to health and
property resulting from the pollution of the atmosphere by
soot. These ordinances do not attempt to control the chemical
composition of the gases that carry the soot. Objection can be
raised on aesthetic grounds to smoke issuing from a chimney.
It can hardly be maintained seriously that the basis of objections
by boards of health and other civic bodies to smoke is aesthetic;
the basis is always the menace to health and property. Ap-
pearance of smoke is not in itself the concern of the smoke ordi-
nances but is assumed to indicate the degree of menace of smoke
to health and property.
Existing smoke ordinances do not concern themselves with
the total quantity of smoke or soot discharged into the atmos-
phere in any communitv or by any plant. What they endeavor
tocontrol is the completeness of combustion of fuel and the limi-
tation of the permissible content of soot per cubic foot of chim-
ney gases. As previously pointed out and as demonstrated
in Fig. 3, a given total volume of smoke of a given actual den-
sity may appear to be dense smoke if it issues from a single large
chimney of may appear to be unobjectionable if it issues from
a large number of smaller chimneys, although the same total
quantity of soot is discharged in both cases. Actually, the
smoke from a multiplicity of small stacks would be a greater
menace to property, since such stacks cannot practically be
made as high as large stacks and consequently discharge their
gases at a lower level where the deposit of soot may reach ob-
jectionable concentrations as compared with the much wider
distribution of soot from smoke discharged at a high level.
Limitation of the total permissible quantity of soot discharged
into the atmosphere would amount to a limitation on the total
quantity of fuel burned in a community or a plant, and no city
is likely to limit the industrial activities of its population in
that manner.
The immediate objective of smoke ordinances is to prevent
the emission of smoke exceeding some definite actual density.
This objective is not realized by existing ordinances. Devices,
such as the photoelectric type of smoke indicator, will permit a
fairly accurate determination of actual smoke densities, but
these devices arc complicated installations attached to the
smokestack or tlues and cannot be used by a smoke inspector.
A smoke inspector can only make visual observation* and this
method, it has been shown, is entirely unsatisfactory unless
the observations are corrected to compensate for chimney di-
ameter, condition of sky, wind, and other conditions.
THREE OBJECTIONS TO CSE OF Ris-CELM*1*"'' CHART
The Ringelmann chart is unsatisfactory, cvcn for the deter-
mination of visual Jcnsitv, because
CO It is inconvenient for the smoke inspector, since the
chart should be set up at a distance of 50 ft from him, a°d this
38
-------
September, 1937
685
ij often impracticable; (2) it is likely to be misleading since
it involves a comparison of the light reflected from the chart
with that transmitted through the smoke, and the chart's ap-
pearance will depend on its position relative to the sun and on
other factors; and (3) it demands judgment on the part of
the observer of the comparative darknesses of two widely sepa-
rated surfaces.
These objections to the Ringelmann chart as a device for
ascertaining visual density can be largely eliminated by substi-
tuting a device such as the umbrascope,1 in which the smoke
is looked at through an eyepiece while half the field of vision
is occupied by a clouded glass or glasses. If the light-absorbing
capacity of the clouded glass is equal to that of the permissi-
ble smoke, inspection will show ac once when the permissible
visual smoke density is exceeded. This instrument eliminates
the three stated objections to the Ringelmann chart and cor-
rects also for the effect of background on apparent smoke density
but does not compensate for stack diameter.
Compensation of observed visual densities for stack diameter
must be based on an experimental determination of the absorp-
tion of light when passing through a column of smoke of con-
stant density but of various thicknesses. As insuring constant
smoke density throughout a series of tests is difficult, these ob-
servations can be made on liquid smoke. Such observations
have been made for the author in the laboratory of the Brook-
lyn Edison Company and indicate that, with light from an in-
candescent-lamp bulb passing through smoke of constant den-
sity, the percentage of light absorbed by a layer of unit thick-
ness is a constant fraction of the light arriving at that layer.
For example, if light of intensity 10 is reduced 20 per cent, or
to intensity 8, in passing through a layer of smoke of unit thick-
ness, then in the next layer of the same thickness the arriving
intensity is 8, the reduction is 20 per cent, or 1.6, so that the
light will leave that layer with intensity 8 — 1.6 ¦¦ 6.4.
The curves of Fig. 5, each of which corresponds to a fixed
actual smoke density, show how apparent smoke density
changes with stack diameter. Smoke which when discharged
from a 4-ft stack has the apparent density 1 on the Ringelmann
chart will appear to be of density greater than 3 if discharged
from a 20-ft stack; smoke showing density 2 from a 4-ft stack
would have an apparent density between 4 and 5 from a 20-ft
stack.
A smoke ordinance that prohibits emission of smoke exceed-
ing some stated visual density on the Ringelmann scale will, in
effect, make severe demands on plants with large stacks but may
be altogether too lenient with domestic or other small stacks.
Insistence on a lower actual smoke density for large stacks is
justified, since it is practicable in large plants to have high-
grade combustion equipment, scientific selection of fuel that is
proper for that equipment, and continuous supervision and auto-
matic controls of combustion. But improvement in combus-
tion, and consequent reduction in actual smoke density, cannot
reasonably be expected co proceed so far, with increase in size
of plant, as to result in no increase in apparent smoke density.
For example, if the quantity of fuel burned is increased four-
fold, the stack diameter would be doubled for the same stack
velocities. Wich unchanged visual density the actual smoke
density would have to be nearly halved.
Complying with very stringent regulations in normal opera-
tion of a plant is possible. During operations such as the
starting up of banked or of pulverized-coal furnaces, the possi-
bility of smoke is always present, but this need not exceed,
either in duration or in actual density, such magnitudes as can
' A.S.M.E. Power Test Codes, Instruments and Apparatus, Part 20,
Smoke-Density Determinations, p. 10.
4 ft 12 16 20
Inside Stack Diam«t«r,Ft
no. 3 CU*VM OF CONSTANT ACTUAL SMOKX DXNSITY, SHOWING
VAX! ATI ON OF ATP AUNT SMOKX DENSITY WITH CKANGB IN STACK
DIAMETER
be readily tolerated. Occasionally, unforeseeable derange-
ments may occur in an otherwise adequate plant and may give
rise to dense smoke; such occurrences should not be penalized,
however, if it can be demonstrated that they were truly un-
foreseeable.
A satisfactory smoke ordinance should have characteristics
somewhat as follows:
(1) It should insist on a high standard of smokelessness for
about fifty-five minutes out of each hour.
(2) It should permit smoke of certain stated visual densities
during the remaining five minutes, these permissible densities
varying with stack diameter and condition of the sky so as to
correspond to attainable actual densities.
(3) The owner of the plant should not be held culpable for
exceeding the permissible smoke density if he is able to demon-
strate that the smoke resulted from an unforeseeable derange-
ment of a good modern plant which is under continuous expert
supervision.
Carbonization of a paste composed of 55 per cent of finely
crushed coal and 45 per cent of petroleum, according to a report
that was recently issued by the Fuel Research Director, results
in the production of a soiid smokeless fuel together with oil
and gas by-products. In the process, which has been developed
by Coal and Associated Industries, Ltd., the coal is completely
converted into coke and the quantity of gas produced, 12,900
cu ft per ton of the mixture, is more than sufficient to supply
the heat required for carbonization. Each ton of the coal and
petroleum mixture produced 55.8 gal of clean crude liquid.
The total quantity of refined spirit was 15-7 gal per ton of the
mixture.—Tht Engitutr, May 28,1937, p. 617.
39
-------
B-2
Reprinted from Electrical World, January 1975
Smoke readings vary with observers
When regulatory stipulations are strict, readings made by a single
smoke observer using a Ringelmann chart are not precise enough
to provide good emission control or reliable enforcement
Recent stack-emission tests conducted
under the auspices of the American
Society for Testing & Materials
(ASTM) indicate that the traditional
Ringelmann method fur reading smoke
density can be misleading when local
air-quality-control regulations are
strictly administered. These tests were
conducted as a part of ASTM's Project
Threshold, which was designed to deter-
mine ihe reliability of the standard
methods for measuring Ihe major pollu-
tants in ambient air, and in emissions
from principal sources.
ASTM does not take sides in such
controversial issues. Its role is to vali-
date standard test methods, using the
best techniques available, and then to
put its findings on the record. Standard
icst reliability comprises accuracy, re-
producibility. and repeatability of the
data collected.
Project Threshold has as its objective
the validation of some 35 standard
methods for sampling and analyzing
ambient air and emissions from station-
ary sources. Phase 1, which dealt with
six ambient-air methods, was concluded
in 1973.
Jn January. 1971. the first on-site
interlaboratory tests under Project
Threshold were conducted to determine
ihe reproducibility of stack-emission
measurement methods. Teams from
four cooperating laboratories carried
out a week-long testing program at an
oil-lired electric utility plant. The meth-
ods studied were those for measuring
particulate matter, residue collection,
sulfur and nitrogen oxides, and velocity
of stack gases.
Similar field tests were mode later at
a coal-fired electric generating station, a
coal-fired industrial boiler, a foundry,
and a cement plum. In June. 1973. on-
site cooperative testing of the Ringel-
m:mn method was conducted in Detroit
and Wayne County. Mich.
Ringelmann history
Jn )Rg». Professor Maximilian Ring-
elmann presented his "smoke-reading"
chart, which has become the standard
device for evaluating smoke-emission
density. H employs a set of white sheets
overprinted with grid lines of predeter-
mined thickness, which produce gradu-
ated shades of gray, varying by five
equal steps between white and black. It
was first used in the US in 1904 by the
Geological Society to study "smokeless"
combustion in St Louis, and Boston of-
ficially recognized the chart in 1910.
when it was incorporated into the city's
smoke-control ordinance.
The US Bureau of Mines also
adopted the chart in 1910, and th[s
paved the way for iis acceptance nation-
ally. It is still the only accepted standard
in practically every city in the country
for determining regulatory compliance
on smoke densities and opacities of vis-
ible emissions.
Air-pollution-contro] laws are essen-
tially ihe adaptation of legal concepts to
empirical evidence and standards which
have been designed to achieve gross re-
ductions of air pollution. Those that
have been adopted by the Los Angeles
Pollution Control District are com-
monly recognized as the model in the
field of air-quality control. This is the
country's largest pollution-control dis-
trict. and its techniques and procedures
have been widely adopted by other
community enforcement authorities.
The California Control Act of 1947
contained Rule 50. which accepted the
Ringelmann chart as a standard for
reading the densities and opacities of
slack emissions. Rule 50 has been tested
several times in court, where questions
have been raised over the validity of the
chart us a standard and over the practice
of reading densities without actual ref-
erence to ihe chart in the field.
As early as 1951. the California Su-
preme Court ruled tha.1 the Ringelmann
chart does not lack certainty, in later
cases, the court determined that Rule 50
was both constitutional and enforce-
able. and that a witness could testify to
the opacity or density of visible emis-
sions without using a Ringelmann chart
on the scene—provided he had been
properly trained.
In general, these and other decisions
have upheld the practical use of the
Ringelmann chart in the field. But they
are now being challenged anew in the
US Court of Appeals for the District of
Columbia, in the case of Portland Ce-
ment Assn vs. Ruckelshaus. The thrust
of Portland Cement's argument is that
the test is arbitrary, and that inspectors
are unable, with any reasonable degree
of accuracy, to determine if the smoke
density is less than the regulatory stipu-
lation of 0 5 RN (Ringelmann Num-
ber). to which the cement company is
obliged to adhere.
The critical question being tested in
court, and which lias been examined by
ASTM. is this: How accurately can den-
sity and opacity observations be made?
Leaders of the utility industry, along
with those who manufacture cement,
are following these court proceedings
closely, since the outcome could throw
all air-quality-control programs into a
dilemma.
Standards
The Environmental Protection
Agency, in accordance with the Clean
Air Act of 1970. has set standards for
visible emissions at 1.0 RN for electric
uliltiv facilities, and 0.5 RN for cement
plants. The regulations established by
state and kical governments may be
even stricter, and they usually stipulate
that visible emissions may not exceed 1
or 2 RN for more than a specified
amount of lime wiihin specific periods-
e.g.. for no more than three minutes in
one hour, as in Los Angeles County.
The smoke observer's qualification is
very important in the use c>f the Ringel-
mann method. His observations and tes-
timony are admissible in court, whether
he uses the chart during observation in
the licld. or only his trained eye. To be
ccrtilicd. the observer must complete a
smoke-readinc course with a content
approved by EPA. Upon completing the
course, he must be able to make nu-
merical assignments to the nearest 0.25
RN. with an error not to exceed 15% on
any one reading, and xn avenge error
not to exceed 7.5'i. All qualified obscrv-
R«^rint*d from ltnuuy IS. 197S Jimm ol ClKtriCfl WM4
4i
-------
ers must pass (his test every year to re-
main certified.
ASTM evaluation
The ASTM Test Method D 3211-73T.
which incorporates the Ringelmann
chart, is a subjective test that covers the
determination of the relative density of
black smoke by visual observation by a
certified observer. Each smoke reader in
the test assigns a number-correspond-
ing in his opinion to a shade of gray on
the Ringelmann chart—to the density of
an observed plume. The observer may
make an "unaided-eye" reading, using a
Ringelmann chart for direct compari-
son, or he may use an instrument called
a "smoke scope." His readings are taken
at constant time intervals, at a specified
distance from the stack, and at a specific
location with respect to the plume and
Ihe sun.
ASTM Project Threshold evaluation
Test Method D 3211-73T was designed
and conducted to study the following
characteristics:
¦ The observer-to-observer vari-
ability in estimating the relative density
of black smoke.
¦ The effect of smoke density on the
variability of Ringelmann Number esti-
mates.
• The presence of bias in the test
method.
In this evaluation, seven certified ob-
servers participated in the smoke-read-
ing tests. The reading experience of
those participating averaged seven
years-ranging from one to twenty.
Four observers had performed smoke
observations on a more-or-less regular
basis as part.of their duties with state
and county air-pollution-control organi-
zations. The other three had not per-
Between-obsarver standard deviation
s» - 0.1 + 0.3 M - 0.07 M*
T 2
Mean tmoke dan.rty (M). RingUm«nn number
Plot of measurement deviation (Fig 1) among observers in ASTM test of R.ngelmann method
formed smoke observations regularly,
but had made smoke-density estimates
within two months prior to the test date.
All seven were fully qualified to partici-
pate in the test.
The Ringelmann method tests were
performed in the early afternoon, under
clear skies, and with a west wind blow-
ing at 2 to 25 knots. All observers, using
the "unaided-eye" method, made con-
current smoke-density estimates on sig-
nal at 30-second intervals, for a total of
60 reading periods. During the test,
each observer estimated the density of
black smoke plumes from the stacks on
an incinerator, an industrial power-gen-
eration facility, and a coke plant, and
from a smoke generator.
The observers worked shoulder-to-
shoulder, but independently. The objec-
tive was to determine how well (or how
poorly) each agreed with the ratings as-
signed by the others, as well as how ac-
curately he could read a plume with a
known value.
To evaluate the test, the differences
Table I: Citation probabilities (%)
Actual tmoka
density (RN)
Maximum density allowed
(RN)
0.5
1.0
1.5
2.C
2.5
0.00
0%
0%
0%
0%
0%
0.25
10
0
0
0
0
0.50
50
2
0
0
0
0.75
75
20
1
0
0
l.OO
92
50
8
1
0
1.25
97
73
27
3
0
1.50
99
87
50
13
1
1.75
100
93
70
30
6
2.00
100
99
85
50
15
2.50
100
100
98
84
50
3.00
3.50
8 8
8 8
100
100
98
100
85
99
between the density estimates repre-
sented a measure of variability, and the
standard deviation of all estimates pro-
vided a measure of the precision of the
test method. The standard deviation be-
tween observers (Sb). called "reproduc-
ibility," was calculated by the equation:
Sb - 0.1 + 0.3 M - 0.07M*
where M is the mean smoke density
reading determined by averaging all
ratings made at a given moment.
The observers also estimated the
Ringelmann Number for plumes of
known density that were produced by a
smoke generator. The difference be-
tween observer estimates and known
values provided a measure of accuracy
of the test method, and the average of
the differences over the tests by all ob-
servers provided a measure of bias.
The Columbus Laboratories of Bat-
telle Memorial Institute coordinated the
test program, evaluated the measure-
ment techniques, and made a statistical
analysis of the data.
The standard deviation, or reproduc-
ibility, established by the tes(s is plotted
in Fig I. As might be expected-since it
is easier to discern with certainty near
white or near black (as opposed to vary,
ing degrees of gray)-the greatest dis-
agreement among observers was in the
middle range of the chart.
As to accuracy with plumes of known
densities, the observers were able, on
the average, to make accurate reading*
although some observers regularly as-
signed darker, and others lighter, values
to the same smoke plume. A significant
bias was found for readings darker than
3.25 RN.
The practical significance of these
findings is that smoke readings by a
single ohscrver are not precise enough
to provide very good control of visible
emissions; therefore, single readings are
not very reliable for enforcement pur-
poses when regulatory controls ure
strict. Biased readings arc considered
likely, unless the mean value of several
observers is used.
The heart of the issue is expressed by
these questions:
42
-------
¦ How certain can a plant operator
be that he will not be falsely charged
with a violation?
¦ How certain can an enforcement
officer be that he will not mistakenly al-
low a violation to go undetected?
Table [, compiled from the analysis
of the ASTM test data, attempts to an-
swer these questions. In the table, actual
smoke-density values are arranged
against regulatory density limits, and
the probability of a violation citation is
presented in each instance.
For example, the table indicates that,
when the regulation calls for I RN max-
imum and a plant has an actual plume
density of 0.75 RN. the operator has a
20% probability of being falsely cited
for a violation. At the same time, an en-
forcement official has a 27% probability
(100% -73%) that, when the regulation
calls for a I RN maximum and a plant
has an actual plume density of 1.25 RN,
a citation will not be issued.
If an operator in a regulated zone
that allows I RN maximum wants best
assurance that he will not be cited, he
must keep his actual plume value under
0.S RN, where his chances of false cita-
tion are only 2%. From this it follows
that, if the regulator wants assurance
that plume values will not be greater
than 1 RN. the regulated limit should
be set at 0.5 RN.
The picture is considerably better
when the regulation limit is 2 RN. In
this case, the operator has essentially
the same protection (13%) against false
citation when he is 0.5 RN to the good,
as the enforcement officer (16%) has of
not catching a plume that is 0.5 RN
above the limit.
The most important observation from
these probability data is that, when the
true smoke density corresponds exactly
with the regulatory limits, the chances
of a citation are 50%. no matter how
good or how bad the reading of the
smoke.
The obvious conclusion from these
tests is that the Ringelmann method can
be misleading in judging compliance
under strict regulatory control. The
ASTM test data findings have been
published by the Battelle Memorial In-
stitute, and the information cited here
must now be taken into account on the
national level.
Meanwhile, better measurement
methods must be devised if stack-emis-
sion standards are to be effective. How-
ever, the ASTM Committee on Sam-
pling & Analysis of Atmospheres
(Committee D-22) has no plan to de-
velop a new method for reading smoke
densities and opacities.
For the immediate future, it seems
apparent that regulatory agencies, when
considering violation citations, must use
cumulative sets of readings, and should
not rely on single evaluations made by
one person. ¦
43
-------
B-3
Reprinted from Natural Resources Lawyer, May 1970
DONALD J. HENZ •
The Ringelmann Number as an
Irrebuttable Presumption of
Guilt — an Outdated Concept
DEVELOPMENT OF THE RINGELMANN CHART
What is the Ringelmann Chart? What prompted its development? The
chart is actually a creature of necessity designed to solve the air pollu-
tion problems of a century past. Looking at the history of man we see
.that air pollution was a mark of progress —for while he learned to control
fire, he could not control its by-products. Until fairly recent times, man
could only comprehend the most primitive of pollution problems. These
were evidenced by smoke and odors, they being perceived by the
senses. Thus concern centered itself in these two areas.
Common law as it developed provided relief from smoke and odors in
the form of a nuisance action. However, it became imperative that
courts balance the equities before granting injunctive relief and, as a
result, it was both difficult and costly for the plaintiff to prove his case.
A smoke belching mill employing the entire community could not be
closed for the sake of a single resident who happened to live in the
fallout area. That this was a pragmatic approach cannot be denied.
However, the court couJd see smoke and, if all the lights light up in the
plaintiffs argument, he was granted some relief. While the degree of
smoke nuisance could possibly be determined with some little degree of
objectivity, not so with odors which to this day are amenable to only the
most subjective of tests. (For example, I was appalled when working in
a rather large dairy, to find that an inspector would sniff each can of milk
as it came off the trucks to determine whether the cows had been in the
onion patch. Quality Control could not predict the results when the
inspector had a cold. Presumably the tainted milk went into pre-flavored
cottage cheese.) So it is, that the early pollution control laws concerned
themselves with that which everyone could see —smoke.
* Mr. Henz is Senior Engineer and Legal Counsel to Pedco Environmental. Inc. of
Cincinnati. Ohio. The opinions expressed herein are his and do not necessarily reflect
those of his employer. He holds an M.E. degree from the University of Cincinnati and a
1 D. from Salmon P. Chase College.
:32
45
-------
DONALD J. HENZ 2.13
But then even (he sense of sight can take on a subjective air. It was
not until nearly six hundred years after the first known smoke abatement
law. that a means was devised to measure the pollutant concentration of
factory plumes with some degree of objectivity. In 1897, the Ringel-
mann Chart, prepared by Maximilien Ringelmann of Paris, was brought
to this country.
The Ringelmann Scale consists of six sections numbered from 0 to 5,
each section being approximately 5-3/4 x 8-^4 inches in size. Section 0 is
completely white while section 5 is completely black. These two sec-
tions are usually not included in the chart. Sections 1 to 4 consist of a
white background with intersecting heavy black lines imprinted thereon,
the lines growing progressively wider from section I to section 4. No. I
Ringelmann is twenty percent black; No. 2 is forty percent; No. 3 is
sixty percent; and No. 4 is eighty percent. The chart was originally
designed to be posted fifty feet from the observer. Estimates of the
density of the plume was then made by comparing it with the intervening
chart, selecting the section which most nearly resembles the smoke. The
Ringelmann Chart has become a standard tool for estimating plume
particulate density.
HISTORY OF THE CHART'S ACCEPTANCE
The Chart was first used in St. Louis in 1904 and.incorporated into the
Boston ordinance in 1910.1 It has had a strange history of acceptance
within this country. While being commonly used in ordinances and
elsewhere since the turn of the century, few cases referred to it and none
actually approved its use until 1951. Despite judicial silence, however,
the Chart became recognized as the single most important factor in air
around our major industrial complexes. It gave to the courts the most
practical and, for a long time, the only means of objectively measuring
the particulate density of factory plumes.
Since 1947, the courts have been more generous in their notice of the
Ringelmann Chart2 and the court approved its use in People v. In-
ternational Steel Corp., 102 Cai. App. 2d Supp. 935, 226 P. 2d 587,
(1951) stating:
. . .Estimates of the density of smoke may be made by glancing from this
chart so displayed to smoke, and picking out the section on the chart
which most nearly resembles the smoke. This mode of measuring the
density of smoke has been in use. it appears, for over fifty years. This
1 Kennedy & Porter, Air Pollution: Its Control and Abatement, 8 Vand. L. Rev. 854.
866 (1955V
J Kennedy, The Legal Aspects of Air Pollution Control With Particular Reference to
the County of Los Angeles. 27 So. Cal. L. Rev. 373. 380 (1954).
46
-------
234 THE RINGELMANN NUMBER
affords a reasonabfy certain mode of determining and stating the density
and opacity of smoke, and we think that the statute adopting it is not
lacking in certainty.
Indeed the courts now take judicial notice of the Chart and it is used
as a standard in current air poHution control legislation in most jurisdic-
tions.3 Jean J. Schueneman has observed. . . More recently, regu-
lations have been adopted which: prohibit emission oF more than a
certain number of pounds of particulate matter per hour. .. ; prohibit
emission of particulate matter having an opacity equivalent to or greater
than smoke of a certain Ringelmann Number;.. .*
Since it is awkward and time consuming to set up a chart for each
observation, the charts are used only for training, after which the ob-
server depends on memory to judge plumes and classify them according
to Ringelmann Number. This is a widely accepted practice and has been
repeatedly upheld by the courts, as has the practice of assigning an
"equivalent opacity" to plumes other than black or gray.® The court, in
People f. Plywood Manufacturers of California, 137 Cal. App. 2d Supp.
859. 291 P. 2d 587,(1955), stated:
In proving a violation, a witness may testify although he did not have
Ringelmann Chart actually in the field with him at the time he made his
observations. One does not have to have a color chart in his hands to
recognize a red flower, a blue sky. or a black bird. The question is one of
credibility, not competency. Nor do we see any difficulty arising from the
fact that a plume of smoke, for example, may appear less dark than
Ringelmann No. 2 from one position, but darker than Rirgefmann No. 2
from another viewpoint. If the contaminant has the substance that, fairly
viewed from any position, gives it a shade as dark or darker than Ringel-
mann No. 2. it is condemned, no matter how light in color it may look to
someone situated at another vantage point.
PRESUMPTIVE GUILT BASED SOLELY ON RINGELMANN -
THE CONTEMPORARY PROBLEM
While judicial acceptance of the Ringelmann Chart has been based on its
long use rather than scientific considerations, many law making bodies
have failed to take note of that fact. Instead they see Ringelmann as a
handy tool for measuring guilt. The consensus is that smoke control
3 All hue two of (he 79 municipal or regional regulations checked by the author,
provide that il is unlawful lo exceed a certain Ringelmann Number. The two jurisdictions
which Jo not prosecute on the basis of a Ringelmann Number are St, Cloud. Minn., and
Springfield Township. Pa.
* Paper hy Jean J. Schueneman. Air Pollution Control Association 57th Annual
Meetinu. 1964.
4 Purdom. Swtrcc Monitoring, in 2 Aih PoiLUTlOM 543 [ 1967).
47
-------
DON ALU J. HENZ 235
regulations referencing the Ringelmann Chart are not so indefinite as to
render the regulation invalid.6 A typical ordinance reads as follows:
A person shall not discharge into the atmosphere from any single source of
emission whatsoever any air contaminant for a period or periods aggregat-
ing more than three minutes in any one hour which is:
(a) As dark or darker in shade as that designated as No. 2 on the
Ringelmann Chart, as published by tte United States Bureau of Mines, or
(b) Of such opacity as to obscure an observer's view to a degree equai to or
greater than does smoke described in subsection (a) of this section.
Notice that no mention is made of the amount of pollutants one puts into
the atmosphere.
If the regulation is to do a proper job it must limit the tonnage of each
pollutant with some degree of specificity. The Ringelmann Chart does
not do this. It is, therefore, not appropriate for the accomplishment of
the object of the legislation —the protection of clean air. This was the
real issue in People v. International Steel Corp. The closest the court
came to this was when it rejected one of the defenses thus:
It is a so urged that the statute is unreasonable and discriminatory because
under it one who discharges an air contaminant only slightly below the
prescribed limit of color or opacity is exempt from the prohibition even
though if he continues his operation long enough he will discharge more
contaminant into the air than one who continues for only a short time
beyond the three-minute minimum. This is only another way of saying that
the line between permission and prohibition is drawn in the wrong place or
that no such line can be drawn. But the drawing of such a line is very
largely a matter of legislative discretion, the exercise of which will not be
reversed by the courts unless abused.
The determination of guilt on the basis of a Ringelmann reading flies
in the face of logic. Fourteen years prior to the case, Lionel S. Marks,
the engineer's engineer, disclosed the results of his research into the
adequacy of the Ringelmann Chart.7 He found that the stack diameter
greatly influenced the apparent smoke density. One of his experiments
consisted of a smoke generating box with three stacks of varying diame-
ters. While each stack vented the same atmosphere, the largest regis-
tered Ringelmann No. 5 while the smallest registered Ringelmann No. 1.
This is, of course, the same effect every boy has observed when looking
for minnows or colorful stones in the shallow pools of ponds and creeks.
The water is clear and the bottom is seen in great detail. But look
4 Annot.. 78 A.L.R. 2d 1305 (1961).
7 Marks. Inadequacy of the Ringelmann Chart, 59 Mechanical Engineering 681
1937).
48
-------
236
THE R1NGELM ANN NUMBER
beyond the shallow area and the water appears muddy. Marks hit upon
the real issue of International Steel fourteen years before the courts
dealt with it, when he wrote:8
Smoke ordinances of American cities concern themselves only with the
apparent or visual density of smoke. The purpose of the ordinances is to
control an assumed menace to health and property resulting from the
pollution of the atmosphere by soot. These ordinances do not attempt to
control the chemical composition of the gases that carry the soot. Objec-
tion can be raised on aesthetic grounds to smoke issuing from a chimney.
It can hardly be maintained seriously that the basis of objections by boards
of health and other civic bodies to smoke is aesthetic; the basis is always
the menace to health and property. Appearance of smoke is not in itself
the concern of the smoke ordinances but is assumed to indicate the degree
of menace of smoke to health and property.
Existing smoke ordinances do not concern themselves with the total quan-
tity of smoke or soot discharged into the atmosphere in any community or
by any plant. What they endeavor to control is the completeness of
combustion of fuel and the limitation of the permissible content of soot per
cubic foot of chimney gases. As previously pointed out and as demonstra-
ted m Fig. 3. a given total volume of smoke of a given actual density may
appear to be dense smoke if it issues from a single large chimney or may
appear to be unobjectionabfe if it issues from a large number of smaller
chimneys, although the same total quantity of soot is discharged in both
cases. Actually, the smoke from a multiplicity of small stacks would be a
greater menace to property. Since such stacks cannot practically be made
as high as large stacks and consequently discharge their gases at a lower
level where the deposit of soot may reach objectionable concentrations as
compared with the much wider distribution of soot from smoke discharged
at a high level. Limitation of the total permissible quantity of soot dis-
charged into the atmosphere would amount to a limitation on the total
quantity of fuel burned in a community or a plant, and no city is likely to
limit the industrial activities of its population in that manner.
The immediate objective of smoke ordinances is to prevent the emission of
smoke exceeding some definite actual density. T/tis objective is not real-
ized by existing ordinances. Devices, such as the photoelectric type of
smoke indicator, will permit a fairly accurate determination of actual
smoke densities, but these devices are complicated installations attached
to the smokestack or flues and cannot be used by a smoke inspector. A
smoke inspector can only make visual observations and this method, it has
been shown, is entirely unsatisfactory unless the observations are cor-
rected to compensate for chimney diameter, condition of sky, wind, and
other conditions. (Emphasis supplied).
Marks was able to plot his empirical data and thus devise the chart
shown in FIGURE NO. I.9
" Id. at 684.
9 Id. at 685.
49
-------
DONALD J. HENZ 237
0
4 8 12 16 20
Inside Stack Diameter, Ft.
FIGURE NO. 1
Four years later in 1941. the Smoke Prevention Association of
America adopted a model Smoke Abatement Ordinance, the pertinent
part of which reads:10
4. It shall be unlawful for any person to use any new or reconstructed
plant for the production of heat and power, or either of them, until he shall
have first procured a certificate from the Smoke. . . certifying that the
plant is so constructed that it will do the work required, and that it can be
so managed that no dense smoke shall be emitted from the chinmey
connected with the furnace or firebox.
To use an old cliche, the Ringelmann Chart is conspicuous by its
absence. Was the California Court aware of the Association's reason-
10Smoke Prevention Association of America, Inc., Manual of Ordinances and
Requirements (1941).
50
-------
THF klN( i KI MANN NIJMHKR
iny.' Ch;mces are. the Association didn't think Ringelmann was appro-
priate for the accomplishment of the object of the legislation either—and
ihtii presents a constitutional problem.
Proponents of the Ringelmann Chart like to cite the 54 year old case
of Northwestern Laundry r. Des Moines. 239 U.S. 486. 36 S. Ct. 206,
60 L. Ed. 396. (1916) which upheld an ordinance utilizing the Ringel-
mann Chart, stating:
. . There are no constitutional restraints upon state actions against the
emission of dense smoke injurious to the common welfare; the only re-
quirement is that the regulation be free from arbitrariness.
The Chart may not have been arbitrary when smoke signals were still
an accepted mode of communication, but it certainly is arbitrary today.
There have been more objective methods of particulate sampling devel-
oped since 19 16. These include the following methods: 1) sedimentation.
2) filtration. 3) impingement, 4) photoelectric, 5) electric precipitation,
and 6) thermal precipitation. All of these methods, albeit some are
crude, determine particulate emission rates more accurately than does
Ringelmann.
Statutes and ordinances which provide for prosecution solely on the
basis of a Ringelmann reading, without regard to the lighting conditions,
wind velocity, and smoke stack diameter, may be held lo be uncon-
stitutional because such a visual test is an unreasonable method of
determining emission pollution content, in light of the present sate of the
art. Times have changed. The U.S. Supreme Court has recognized that
statutes, valid when enacted, may become invalid because of changed
conditions. Nashville, C. and St. L. Ry. v. Wallers, 55S. Ct. 486, 294
U.S. 405. '79 L. Ed. 949 (1935). Such legislation lacks reasonableness
unless the means are reasonably necessary and appropriate for the
accomplishment of the legitimate objects falling within the scope of the
power. 11 Am. Jur.. Constitutional Law § 303. Ringelmann. without
modification, is scientifically discredited as an appropriate means of
determining pollution rates.
Even the U.S. Public Health Service recognizes the inadequacy of
Ringelmann:11
Estimation of smoke density must lake into account variations in lighting
and background against which the smoke is viewed. Different readings
may he obtained on dull days and bright days or when the position of the
observer is shifted, Readings are also influenced by variations in the
illumination of the chart. The chart in all cases is illustrated by the ambient
light near the position of the observer which may be different from that at
"United States Dept. of Health. Education and Welfare, Robert A. Taft Sanitary
Engineering Center. Fine Particle Techniques in Air Pollution 7 (Training Course Manual
1957).
51
-------
DONALD J. HENZ 239
the stack. These limitations of the Ringelniann Chart method have been
recognised by many people engaged in smoke control work.
And C. A. Lindstrom writes:12
The Ringelmann Chart is a method of judging the shade of gray of a
particulate laden plume emitted from a source of combustion. It is as-
sumed that the darker the shade of gray, the greater the concentration of
black-colored particulate matter in the plume. The particulate matter im-
parting the blackness consists mostly of soot, flyash, and other solid and
liquid particulates less than I-micron in diameter.
It must be cautioned that application of the Ringelmann Chart, alone, for
the determination of the shade of gray of any plume, provides no specific
measurement of the concentration of pollutant in the effluent. (Emphasis
supplied.)
The Mechanical Engineer's "bible" states, "While the Ringelmann
Chart is frequently used to evaluate stack emissions, it is a crude and
inaccurate method."13
Another expert in the field states:14
Many investigators have criticized the use of the Ringelmann Chart for the
evaluation of smoke density. It is admittedly subjective. The orientation of
the sun and the intensity of illumination have a decided effect upon the
amount of light reflected. Wind and stack exit velocity will influence the
mixing of the stack gases with ambient air. with a result that greater mixing
will decrease the apparent density. The thicker the smoke column, the
greater will be its apparent density. These objections are partially over-
come by "reading" the plume as close to the stack discharge point as
possible. The stack gases may be diluted by the addition of air prior to
discharge, (hereby decreasing their apparent blackness. (Emphasis sup-
plied).
Of visual methods in general, Purdon writes:"
One of. the earliest concerns about air pollution related to the smoke
originated from combustion processes. Lacking current technology, a form
of emission control evolved which was based upon the visual observance
of the smoke plume. The earliest efforts to standardize observations were
based upon plume reflectance, but more recently plume transmittance has
been utilized to a greater extent.
Methods for the visual evaluation of emissions have the fault that the
optical density may not be related to the weight of contaminant discharged
to the air. (Emphasis supplied).
It is not the purpose of this paper to besmirch the record of the
"United States Dept. of Health. Education and Welfare. Robert A. Taft Sanitary
Engineering Center. Elements of Air Quality Management 9-7 (Training Course Manual
1967',.
13 Tiggs & karlsson. Stack Emissions, in Standard Handbook for Mechanic ai.
Enginekrs 7-S6 (7th ed. 19671.
11 Purdom. Source Monitoring, in 2 Air Pollution 541 (2d ed.) (1967).
" Id. at 54 I.
52
-------
240 THE RINGEl.MANN NUMBER
Ringelmann Chart. Where would we be today had not such a chart been
developed and adequately utilized in the past regulations? It has served
us well. As seen above, however, the chart covers only a small portion
of the problem and, in view of current knowledge, fails to do this
objectively. Today's problems are different than those of the past. Not-
withstanding this difference, the chart can still be a valuable tool for
establishing a prima facia case. It should be used as no more than a
rebuttable presumption of guilt, thereby placing the burden of going
forward on the defendant.
As a result of the Clean Air Act of 1967, much legislation has been
recently passed. For example, 66 of 90 state and local jurisdictions have
adopted or amended their regulations since January I, 1967. All but one
of these regulations provide for prosecution solely on the basis of a
Ringelmann number reading.
Notwithstanding high, and very often irrational feelings regarding air
contaminants, it does not seem possible that the court's view in People
v. International Steel, can long endure. In this sense. International Steel
may well become a pyrrhic victory to the foes of air pollution, for sooner
or later, some defense attorney will show the court Lionel Mark's
thirty-three year old experiment —and the court will submit to logic.
53
-------
B-4
Reprinted from Natural Resources Lawyer, Vol. 7, Mo. 3, Summer 1974
ROBERT E. HAYTHORNE
JAMES W. RANKIN*
Visual Plume Readings—
Too Crude for Clean Air Lawst
Any traffic policeman can distinguish between an auto travelling at 10
miles an hour and one going twice as fast. He won't need a speedometer
or radar to make this evaluation, yet no one would suggest that a 10 mph
speed limit should be enforced by the policeman's eyeball judgment
Smoke plume inspectors cannot judge plume density as accurately as
the policeman can judge a 10 mph speed. One third of the trained in-
spectors at an EPA smoke reading school rated a 10 percent plume as
20 percent or more opaque. Nevertheless, EPA regulations prescribe a
10 percent opacity limit for some plumes despite the fact that enforce-
ment of that limit must depend completely on the eyeball judgment of the
inspector.
The implications of such regulations are starkly illustrated by a recent
case which reached the U.S. Supreme Court.1 In that case an inspector,
without a warrant, entered a company's premises. Without notice to the
operator he evaluated the opacity of the plume emitted from the plant's
stack. His reading indicated that a violation was talcing place. The opera-
tor, not having been aware of the inspection, did not secure any evidence
concerning emissions at that time, and therefore had no defense. His evi-
dence of compliance at other times was excluded. The Colorado appeals
court stated that fundamental due process rights were violated because the
defendant was foreclosed from presenting any rebuttal evidence and held
the unreasonable search provisions of the Fourth Amendment were vio-
* Messrs, Haythorne and Rankin are partners in the firm of Kifkland & Ellis,
Chicago, Illinois. Holding J.D. degrees from the University of Chicago, they are mem-
bers of the Illinois Bar.
fAdapred from a presentation made at the Southwestern Legal Foundation
Second Annual Short Course on Environmental Law and Litigation, Dallas, Texas on
Tune 18, 1974. The authors acknowledge extensive reference to in excellent article by
Donald I. Henz published under the title The Ringelmann Number As An IrrebutabU
Presumption of Guili—An Outdated Concept, 3 NUturm, Res..LaW, 292 (1970^.
lAir Poll. Board v. Western Alfalfa Corp. 42 U.S.LW. 4756, U.S. 7
40 L. Ed. Id 607 (1974).
457
55
-------
458 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
lated.2 In April 1974,3 the U.S. Supreme Court reversed, per Mr. Justice
Douglas, on the basis that sights seen "in the open fields" are not so pro-
tected. It declined to rule whether the failure to give notice was a due
process violation-
Legislative bodies continue to prescribe plume opacity limits in anti-
pollution laws—exceeding the limit violates the law. Administrative agen-
cies vigorously enforce this concept. And the courts have upheld this pro-
cess. Enigmatically, these governmental branches apparently overlook the
underlying fact that, even if constitutional guarantees are respected, eye-
ball opacity readings are exceedingly inaccurate at present levels and that,
even if such readings were accurate, opacity is not a reliable, accurate
measure of the amount of pollution in the plume.
This paradox may be better understood by reference to history rather
than logic or reason. Following the historical perspective presented in part
I below, parts II and III of this article will discuss, respectively, the in-
adequacy of visual plume evaluation, and opacity as a legal standard.
Where appropriate, some comments concerning highway safety will be in-
cluded to furnish a reference for comparison.
I. HISTORY
At the turn of the century dense black smoke and automobile traffic were
considered nuisances. Local ordinances prohibiting emission of Mdense
smoke"4 were upheld by the courts as a means to abate nuisances.3 Auto-
mobiles also created nuisances. When Peter J. Rodgers got himself "good
and drunk" and drove his auto through a saloon window he was guilty of
committing a "public nuisance."6 Automobiles frightened horses and pro-
tective laws were passed.7
2Western Alfalfa Corp. v. Board, 510 P.2d 907 (Colo. App. 1973).
3Air Poll. Board v. Western Alfalfa Corp., U.S. , 40 L. Ed. 2d 607
(1974).
~This was not a new development. In the reign of Edward II (1307-1327) "a
man was put to torture ostensibly for filling the air with 'a pestilential odor* through
the use of coal." I Air Pollution 5 (2d ed. 1968).
sHarmon v. City of Chicago, 110 111. 400 (1884); Moses v. U.S., 16 App, D.C.
428. 50 A.L.R. 532 (1900). See also A.L.R.2d 1328; Kennedy and Porter, Air
Pollution: Its Control ar.d A batement, 8 Vand. L. Rev. 854 (1955),
tfState v. Rodgers. 90 N.J.L. 60, 99 A. 931 (1917).
T"Perhaps the wildest of these laws, according to the National Automobile Club,
were drawn up by the Anti-Automobile Society that was formed back in Pennsylvania
when the problem was first coming to the fore. There the farmers decided that anyone
driving a horseless carriage along the road at night should come to a stop every mile"
and send up a signal rocket, then wait 10 minutes for the road to clear. If a team of
horses should approach along the road, the motorist was obliged tb pull off the road
and cover his vehicle with a large canvas or painted cloth that would blend with the
surrounding landscape. If the horses refused to pass even then, the motorist had to
56
-------
VISUAL PLUME READINGS 459
"Dense smoke" is not a precise concept. Accordingly, when Professor
Ringelman's charts8 became known, they were incorporated as the stan-
dard for the permissible density of smoke emissions beginning with a 1910
Boston smoke ordinance.9 Thereafter, Ringelmann No. 2 (40 percent
opacity) became the accepted standard for smoke ordinances.10
With the passage of time, two varieties in plume evaluation developed:
(i) It became unnecessary to use the chart in making an evaluation.
The courts concluded that a trained observer could remember11 shades
of smoke.12 As stated by one court "One does not need to have a color
chart in his hand to recognize a red flower, a blue sky or a black bird."1*
That court would have been right if the problem bad been to distinguish
between a black plume and a white plume. When, however, it comes to
differentiating between various shades of gray, white or beige, anyone who
has ever dealt with an interior decorator knows a color chart-in-hand is
essential.
(ii) Light-colored plumes became subject to evaluation even though a
black (or gray) chart cannot be compared to a white plume. Typically,
regulations came to prohibit not only plumes "darker in shade than that
designated as No. 2 on the Ringelmann Chart" but also plumes "of such
opacity as to obscure an observer's view to the same degree." The pro-
visions have been upheld by the courts.14
Until the end of World War II smoke abatement was largely a matter
of local concern. Thereafter, public awareness was directed to protection
of the environment rather than the abatement of nuisances and the states
undertook, or were imposed with, the duty to regulate and control pollu-
tion.15 The 1947 California Health and Safety Code stated its purpose
take his vehicle apart piece by piece and hide the pieces under the nearest bush." As
quoted in Fishex * a »«>««, Vehicle Traffic Laws 21 (1974) [hereinafter cited as
Fisher * Reedex].
*The Ringelmann charts consist of white paper marked with black cross-hatched
lines covering 20%, 40%, 60% and 80% of the surface. These are referred to as
Ringelmann Numbers 1, 2, 3 and 4 respectively. When held at a distance they appear
to be different shades of gray and are intended to be used for visual comparison with
dark smoke plumes. Their use has been described in a series of information circular*
issued by the U.S. Bureau of Mines, the latest of which is IC 8333 [hereinafter cited
as IC 8333].
9Mass. Acts 1910, ch. 651, S§ et seq.
10Henz, supra notet, at 233-35.
nThat observers may forget is indicated by the current EPA requirement that
they be recertified at least every six months, 36 Fed. Reg. 24,895, Test Method No. 9
(12-23-71).
"People v. Int'l. Steel Corp., 226 P.2d 587, 592 (Sup. Ct. LA. Cty. 1951).
,3People v. Plywood Mfgrs., 291 P.2d 587, 591 (Sup. Ct. L.A. Cty. 1955).
14People v. Plywood Mffrs., supra note 13; State v. Fry, 49S P.2d 751 (Or.
App. 1972).
"Comparable developments in the field of highway safety had taken place 50
years earlier. Before 1910 New York and Pennsylvania had passed statewide laws
57
-------
460 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
was "to safeguard life, health, property and the public welfare and tc
make possible the comfortable enjoyment of life and property."1' The
Clean Air Act of 1955, 42 U.S.C. § 1857, indicated the responsibility ol
the states to prevent air pollution and, referring to this legislation, th<
U.S. Supreme Court stated that the "sole aim" of a Detroit ordinance "is
the elimination of air pollution to protect the health and enhance the
cleanliness of the local community."17
The decade of the 1970s has brought another major evolutionary
change in environmental control. The extensive 1970 amendments to the
Clean Air Act18 put the federal government squarely into the direct con-
trol of air pollution.19 Under the amended statute the Environmental Pro-
tection Agency (EPA) has been directed to prescribe standards for
stationary source emissions and to approve state pollution control plans
or, lacking an acceptable state plan, to impose its own plan for the state.
As soon as the 1970 Clean Air Act Amendments were passed EPA
moved promptly to comply with its statutory directive, and proceeded to
cut existing opacity standards in half. On August 17, 1971, EPA pro-
posed standards which limited opacity of power plant emissions to 20
percent and acid and cement plant emissions to 10 percent.20 Then, on
November 15, 1971, the Agency published regulations requiring that
state implementation plans assure that national emission standards will
not be exceeded.21 Thus alerted, the states submitted their plans for EPA
approval. Thirty-eight plans adopted the 20 percent opacity limit for
industrial sources, one plan prescribes a 10 percent limit, and three pro-
hibit any visible discharges whatsoever.2*
At the time EPA cut opacity limits in half and wrote them into federal
law, two authoritative sources questioned the use of eyeball opacity read-
ings as a valid measurement of air pollution and indicated that the accur-
acy of such readings was not adequate for the purpose of enforcement
which superceded local ordinances and highway safety became recognized not as a
matter of nuisance abatement but as a specific independent public interest goal. Set
Fisher 4 Reeder, supra note 7, at 22 et seq.
»«CaJif. Stats, 1947, ch. 632 £ 2419S-9; see People v. Int"l. Steel Corp., supra
note 12.
i'Huron Portland Cement Co. v. Detroit, 362 U.S. 440, 445 (1960).
ls42 U.S.C. I 1S47, et seq.
1DCotn parable developments in highway safety occurred in 1966 with the passage
of the National Traffic and Motor Vehicle Safety Act of 1966 and the Highway
Safety Act of 1966. See Fisher & Reeder, supra note 7, 25 at et seq.
S040 C.F.R. § 466.22 et seq.. 36 Fed. Reg. No. 247 (12-23-71). Subsequent regiu
lations prescribe a 20% limit for lead smelters and asphalt, concrete, and sewage
treatment plants. 39 Fed. Reg. No. 47 (3-8-74).
2*40 C.F.R. 5 5118, 36 Fed. Res. 22,398 (11-15-71).
—Analysis of Final State Implementation Plans—Rules and Regulations. U.S.
EPA (Office of Air Programs) Publication APTD1334 (1972), at 51.
58
-------
VISUAL PLUME READINGS 461
The basic Bureau of Mines circular cautioned that (i) visual opacity data
is empirical and depends on a number of variables including particulate
size, depth of plume and lighting conditions, and (ii) the judgment of
observers should not be pressed to too fine distinctions.88 A cooperative
study of the U.S. Public Health Service and the Edison Electric Institute
published by the Department of Commerce in 1967 stated, in effect, that
the amount of pollution in a plume should be measured by some means
other than an eyeball evaluation."
EPA's motives in cutting permissible opacity limits were unquestion-
ably meritorious. But clean air cannot be achieved by governmental fiat
which cannot be reasonably enforced. Even the Ringelmaim No. 2 ordi-
nances did not result in a significant number of plant closings. By com-
parison, no one would suggest the strict enforcement of a 55 mile an hour
speed limit if it depended on the unaided subjective judgment of the
policeman, no matter how trained and experienced he might be in visual
speed estimation.25
Air pollution enforcement officials, however, do not always recognize
the limitations inherent in eyeball opacity readings. Recently, the Texas
Attorney General stated that his office regards prescribed levels of pollu-
tion to be the same as speed limits; that the validity of such regulations is
not subject to question in enforcement proceedings; and that a polluter
will be prosecuted just as a driver "going 61 miles an hour in a 60 mile an
hour zone."8*
Meanwhile the accuracy and validity of opacity readings continues to
be questioned. Following EPA's 50 percent reduction of permissible opac-
ity limits, two authoritative studies of opacity observations have been
published indicating substantial inaccuracy inherent in visual plume opac-
ity readings. First, EPA itself published a report entitled "Average Ob-
servational Error Associated with Smoke Plumes at Known Levels of
Opacity."27 Second, the prestigious American Society for Testing and
Methods published in January 1974 its "Final Report on Interlaboratory
"IC 8333 at 2, 3.
2*"Optical Properties and Visual Effects of Smoke Stack Plumes," U.S. Pub.
Health Serv. Publication 999AP30 (1967 Nafl. Tech. Info. Serv. PB 174705), at 58,
59 [hereinafter cited as the HEW Study].
asSpeedometers and radars used by police are accurate to within one mile per
hour. Fisher, Legal Aspects of Speed Measurement Devices (1967).
20Remarks of Hon. John L. Hill at a luncheon meeting sponsored by the South-
western Legal Foundation in Dallas, Texas on June 19, 1974. An attorney listening to
Attorney General Hill's statement was heaid to remark, "But I don't want to be
charged with speeding for going 55 mph in a 60 mph zone." Shortly after this
luncheon, suit was brought in the U.S. District Court to set aside the Texas Opacity
Regulations. General Portland Cement, Inc., Case No. CA 3-74-621C, N.D. Texas.
-Hereinafter, the EPA Study.
59
-------
462 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
Cooperative Study of the Precision and Accuracy of the Determination
of the Relative Density of Black Smoke (Ringelmann Method) Using
ASTM Method D 3211-73T."28 These reports and other pertinent mate-
rials form the basis for the following analysis.
H. INADEQUACY OF VISUAL PLUME EVALUATION
Simply stated, visual plume evaluations are not valid to determine whether
anti-pollution regulations have been violated because:
a. Eyeball plume readings cannot be made with sufficient precision;
and
b. Opacity, even if it could be measured precisely is nor an adequate
measure of the amount of pollution the plume contains.
A. Precision—The Inaccuracy of the Observer's Evaluation
The concept of precision simply does not apply to visual plume evaluation.
The human eye cannot reliably distinguish opacity differences less than
five percent.29 An EPA observer is permitted an error of 15 percent (opac-
ity) on single readings and an average error of 7.5 percent (opacity) and,
despite that degree of error, he may still be certified.30 Thus, in the evalu-
ation of a 10 percent plume bis error could be ±75 percent of the read-
ing. Such an error would, of course, be a mockery of "precision." By
comparison, violation of a 10 mph speed limit certainly could not be sup-
ported by a policeman's speedometer which, in a calibration run, regis-
tered 2.5 mph when his car was traveling at 10 mph.
The EPA, ASTM, and HEW studies of observer accuracy speak in
terms of average deviations.31 This permits errors of one observer to can-
cel errors of another, a process not applicable to routine field evaluations
where only one observer rates the plume. Realistically, in enforcement
activity, the agency, source operator and courts must deal with the read-
ings of a single observer. In this context the studies show errors radically
greater than the averages.31
The 0 Percent Non-Plume. Observers in the HEW and EPA
studies were shown "plumes" of 0 percent opacity. It is not knows
^Hereinafter, the ASTM Study.
MHEW Study, supra note 24, at 10.
3"36 Fed. Reg. 24895, Test Method No. 9 (12-31-71).
3i.S>£ notes 23, 26, and 27 supra. On this basis the EPA and ASTM studies indi»
cate the average error was *3% and ±10%, respectively in terms of opacity. Applied
to a 10% opaque plume the error in the reading would be ±30% and ±100%,
respectively obviously a travesty on the concept of precision.
^Appendices A-E to this article reproduce, respectively, the tabulated,results of
observations reported in the HEW Study (page 28), the EPA Study (App. B), and
the ASTM Study (Tables 1,1 and 3).
60
-------
VISUAL PLUME READINGS 463
how complete transparency can indicate a "plume." Nevertheless, all
six HEW observers saw a plume and seven of the 12 EPA observers
did also. Thus thirteen of eighteen observers gave an opacity rating
to a non-existent "plume." Of the thirteen, ten rated the "plume"
at five percent or more, four at 10 percent or more and three rated
it at more than 20 percent
10 Percent Opacity. In all three studies, observers were shown 10
percent plumes. These plumes would not have violated a regulation
prohibiting opacity in excess of 10 percent. Yet over two-thirds of
the observers would have found a violation:
(i) In the HEW study all six observers would have rated the
plume in violation and five of them rated it over 20 percent
opaque;
(ii) In the EPA study 60 of 93 observers (almost two-thirds)
would have rated the plume in violation and 29 rated it over
20 percent opaque;
(iii) In the ASTM study 35 percent of the observers would have
rated the plume in violation and IS percent of the observers
rated it 20 percent or more opaque.**
20 Percent Opacity. Currently this is the level which may not be
exceeded under most State and Federal regulations. Yet more than
half of the observers would have found 20 percent plumes to be in
violation;
(i) In the HEW study five of six inspectors would have found
the plume to be in violation and two rated it as almost 40
percent opaque;
(ii) In the EPA study 165 of 331 observers would have rated
the plume in violation and 72 rated it as 30 percent or more
opaque;
(in) In the ASTM study 50 percent of the inspectors would have
rated the plume in violation and over 25 percent of the ob-
servers rated it as 40 percent or more opaque.
40 Percent Opacity. Even at the formerly accepted Ringelmann
No. 2 level more than one-third of the inspectors would have found
a violation where none existed:
(i) In the HEW study three of five inspectors rated a plume
known to be slightly less than 40 percent opaque to be more
dense than it really was;
33See Appendices A-E and note 32 supra.
61
-------
464 NATURAL RESOURCES LAWYER VOL. VH, NO. 3, SUMMER 1974
(ii) In the EPA study 179 of 507 observers would have rated the
plume in violation and 76 rated it at 50 percent or more
opaque;
(iii) In the ASTM study four of five observers would have found
a violation.
Confronted with the foregoing study results—which demonstrate that
a substantial number of observers will find violations where none exist—
proponents respond, "We wouldn't prosecute a violation unless the read-
ing were far in excess of the limit—such as a 40 percent reading to prove
violation of a 10 percent limit." A more sophisticated, but identical,
apology is articulated by EPA as follows:
For example, the agency would not take, and the courts would not sus*
tain, enforcement actions based upon opacity observations that do not
exceed the standard by more than the average opacity error associated
with the particular opacity standard.34
The simple answer to these comments is—if an opacity level is not
going to be enforced it should not be enacted into law Furthermore,
consider the EPA statement! The average error at 10 percent opacity was
found to be -^5 percent (opacity). Thus EPA would not enforce a 10
percent limit unless the inspector's reading were 15 percent. But, the
agency study shows that almost a third of the 15 percent readings related
to five percent or 10 percent plumes which would not have violated the
regulation.
The average error concept furnishes no help in a case where one in-
spector's reading showing a plume violation must be weighed against
another expert's reading showing compliance. To resolve this conflict, a
court would be required to ascertain the point at which all qualified ob-
servers would agree there was a violation. Why shouldn't the regulation
set the limit at this point? Some increased emissions might be permitted
but this is the only method which would insure individual rights against
overly zealous enforcement efforts.3®
A 10 percent opacity limit presents special problems because the aver-
age error may be unacceptably large when expressed as a percentage of
the 10 percent limit. Ten percent opacity is Ringelmann No. 1/2. At this
level the U.S. Bureau of Mines circular dealing with the chart states:
... it is questionable whether results should be expressed in fractional units
because of variations in physical conditions and in the judgment of the
observers."
"EPA Draft Response, P.C.A. v. Ruckelshaus, No. 72-1073 (D C. Cir.), at 26
(emphasis supplied).
"S?? note 26 supra.
3«IC 8333 at 2, 3.
62
-------
VISUAL PLUME READINGS 465
This conclusion is supported by the HEW Study's indication that the
human eye cannot always discriminate between five percent differences in
opacity.37 The EPA Study comments that at the 10 percent level the aver-
age error "was noted to be 5 percent" (opacity)38—i.e., ± 50 percent of
the reading. The ASTM report concludes that "observer readings across
the entire Ringelmann Range were within 10 percentage points of genera-
tor readings"38—i.e., an error of ± 100 percent at the 10 percent level.
Furthermore, all three studies were conducted under controlled "school"
conditions. Qualification under these conditions does not mean the ob-
server can read a plume accurately in the field, for at least three reasons.
First, before a student is tested at school he is typically permitted a
"calibration run" in which plumes of known opacity are emitted. He
then takes the test while the impression is still fresh in his mind. No "cali-~
bration run" is possible in the field. If a transmissometer calibration were
possible, there would be no need for an eyeball rating. Second, the train-
ing course is typically limited to less than three days and does and cannot
deal with demonstration of variations attributable to wind effects, stack
diameter, particle size or moisture in the plume. The smoke machine can-
not generate a "wet" plume, the particle size remains the same, and so
does the diameter of the small stack. Wind conditions are subject to na-
ture only during the short time the student takes his training and test.
Third, at some schools, students who fail the test are permitted to retake
it until they pass or give up.
Thus are the abilities, limitations and training of the observer who has
the power to judge whether or not a manufacturer's plume violates the
law. His task and its implications are a far cry from the identification of
a "red flower," "blue sky" or a "black bird."
B. Opacity Does Not Equal Pollution
Even if an inspector could read a plume with complete accuracy his read-
ing would not measure the amount of pollution it contains. Variations in
stack diameter, gas velocity, particle size, moisture content or light condi-
tions will cause plumes to have different opacities even though they con-
tain the same weight of pollutants. Inspectors are not taught how to
evaluate these variables. Perhaps adequate evaluation is not possible.
1. STACK DIAMETER AND OAS VELOCITY
The most obvious variable is the thickness of the plume. Anyone who has
dropped a penny into the still, clear water of a deep well or spring knows
3THEW Study, note 23 supra, at 10.
siEPA Study, at 3.
30ASTM Study, at 2.
63
-------
466 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
that the coin can be seen clearly for a while and then gradually disappears.
This does not mean that the water at the bottom of the well is less pure
than that at the top. The observer merely has to look through more and
more water, each unit of which contains the same amount of impurity
until the total amount of impurity obscures the coin.
Applying this concept to an industrial stack, Engineer Louis S. Marks
in 1937 constructed a smoke generating box with three stacks of varying
diameter. The same emission registered Ringelmann No. 5 from the largest
stack and Ringelmann No. 1 from the smallest.40 One commentator face-
tiously suggested "many small stacks so that the top of a smoke stack
might look like a candelabra" thus producing a less opaque plume for
each stack.11
Under the same principle, it can be understood that an observation
made parallel to the flow of a plume will produce a higher opacity rating
than an observation made perpendicular to its flow. Fortunately, this
possibility has been eliminated. Most regulations require the observation
point to be "perpendicular to the plume."*8
The velocity of the exhaust gas and the external wind conditions also
will have obvious effects. In a still atmosphere the particulates may build
up, thus increasing opacity. In a high wind panicles may be dispersed so
that there may be no opacity at all even though the same volume of par-
ticulates is emitted.
2. PARTICLE SIZE
A plume containing very small particles may be more than four times as
opaque as one containing an equal weight of larger particles.43 Thus,
plumes of 20 percent and 80 percent opacity could contain identical
weights of pollutants. The explanation of this phenomenon in terms of
optical science can be very complicated.4* Simply stated, however, as a
solid bulk is broken into smaller pieces the total surface area increases
although the weight stays the same. It is the surface area which interferes
with the passage of light and with visibility. Thus opacity increases as
particle size decreases.
3. MOISTURE
Some industrial plumes contain substantial amounts of moisture and a
moist plume can be 100 percent opaque even though it contains no pollu-
"Marks. Inadequacy of Ringelmann Chart, 59 Mechanical Engineering
681 (1937).
"Phelps, The Equivalent Opacity Problem 5, IS (1967).
«36 Fed. Reg. 24.S95, Test Method No. 9 (12-31-71).
¦^Phelps, supra note 40, at 4.
"HEW Study, supra note 24, at 29 et seq.
64
-------
VISUAL PLUME READINGS 467
tant whatsoever. Anyone who has observed a boiling teapot on a kitchen
stove knows this. Pure water vapor can be totally opaque.
Ail Arthur D. Little study describes this characteristic of industrial
plumes as follows:
However, there are two important situations where an observer cannot
adequately evaluate a plume:
Wet Plumes- Many industrial effluents are saturated with moisture at
temperatures above the ambient, and under many situations the plume
from such operations will be highly opaque. The effluent from most
scrubbers fits in this category, and the apparent opacity of the plume
bears more relationship to ambient temperature and humidity than the
amount of pollution being emitted to the atmosphere. The inspectors of
the LACAPCD try to observe such plumes on a day of low relative hu-
midity when the water droplets evaporate at a minimum distance from
the emission point and make a judgment of whether or not the plume
near the stack: outlet would have complied with the regulation had the
water not been present. Such a method is adequate in cases where there
is no residual plume or where the residual plume is clearly in excess of
40% opacity, but is inadequate for intermediate situation*.44
Recognizing this problem, the federal standards provide that "where
the presence of uncombined water* is the only reason for failure to meet
the [opacity standard^1 the source will not be in violation-4*
This attempted solution raises more problems than it solves. Obviously
the eye cannot separate the water from the dust where both appear in
combination. It is said, however, that at a short distance from the stack,
steam will evaporate and the moisture will drop oat of the plume, after
which the opacity of the dust can be measured. An EPA study states:
Most air pollution control regulations including New Source Perform-
ance Standards specifically exempt opacity caused by condensed water
vapor. Visible emissions caused by particulate matter or other air pol-
lutants and which can be distinguished from water vapor are not ex-
empted from the standards. Therefore, a source releasing a mixture of
condensed water vapor and air pollutants to the atmosphere may be cited
for violation if it is shown that opacity attributable to air pollutants
exceeds allowable limits.
The premise inherent in this enforcement approach is that a trained
smoke observer can distinguish between opacity caused by water vapor-
and opacity caused by air pollution. Persons concerned with the enforce*
«Yocotn, Problems in Judging Plume Opacity, 13 I. Am Poll. Ctrl. Assn. 37
(1963).
««40 C.F.R. § 60.52 (d); Fed. Reg, 20793 (6-14-74). Paradoxically, EPA now
proposes to eliminate this exception from the standard, and proposes to add to the test
method a requirement that the observer should rate opacity at a point where no mois-
ture is present. 39 Fed. Reg. 32,857-58 (9-11-74). There is no'scientific foundation
for the assumption :has that point can be identified by visual observ ation.
65
-------
468 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
ment and development of air pollution emission standards have had ex-
perience in observing visible water droplets. It has been their experience
that in most instances condensed steam or water vapor can readily be
distinguished from such pollutants as rock dust, cement dust, fly ash from
coal burning, etc.47
To support its assumption that, at a certain point, the plume will con-
tain no more water, EPA presents a series of photographs which show a
white plume which becomes darker as the distance of the stack increases.
However, the report does not identify the point at which all moisture has
left the plume. Experts disagree, and no scientific foundation is available
to demonstrate the point at which all moisture is dissipated so that only
the opacity attributable to the particulates can be evaluated.48 Besides, at
a distance from the stack the plume size will have changed and wind con-
ditions will have had effect. Both of these conditions can change opacity.
4. LIGHT CONDITIONS AND BACKGROUND
The appearance of a plume will vary according to light conditions and
background. A technical explanation of these variations is given in the
HEW study.49 Another report states:
With white plumes and guides, where appearance is due primarily to the
amounts of light they scatter toward the observer, contrasting backgrounds
may be provided due to color (blue sky) of luminace (dark background).
An overcast sky does not present a good background for evaluating white
plumes since it is white and fails to provide luminace contrast with either
a white smoke or the white guide. On clear days the sky does not present
a good background when viewing toward the sun because of scattering
by atmospheric particulates, which lightens and obscures the color of the
sky. In these conditions the observer should view the guide and the smoke
against a dark background instead of the sky if possible.30
But, in Plywood Manufacturers the court which referred to red flowers
and black birds said:
If the contaminant has the substance that, fairly viewed from any posi-
tion, gives it a shade as dark or darker than Ringelmann No. 2, it is
condemned, no matter how light in color it may look to someone situated
at another vantage point.31
¦•"A Report on Contaminated Water Vapor Plumes (EPA, Office of Control
Technology. December, 1973) at 1.
**See State v. Fry, 495 P.2d 751 (Or. App. 1972).
¦•"HEW Study, supra note 24, ai 10 et seq.
5uConnor. Smith & Nader, Development of a Smoke Guide for the Evaluation of
White Plumes, 18 J. Air Poll. Ctrl. Assn. 750 (1968).
"People v. Plywood Mfgrs., 271 P.Id 587, 591 (Sup. Ct. L.A. Cty. 1955).
66
-------
VISUAL PLUME READINGS 469
Thus, while the problem of observation position may be eliminated by
ihe requirement that the observer place the sun at his back, an inadequate
background may mean that no reading can be made at all.
m. OPACITY AS A LEGAL STANDARD
Constitutional guaranties prohibit the use of opacity limits as legal stan-
dards. Unless and until plume opacity can be measured accurately and
can be reliably correlated with pollution, regulations which establish it
as the criterion for compliance or violation should be held invalid for at
least two reasons.
First, it is axiomatic that a statute or regulation will violate due process
rights if reasonable men can differ as to whether the act of violation was
committed.33 Qualified plume inspectors can differ as much as 100 per-
cent in rating a 10 percent plume. The possibility of such differences
between experts should invalidate most opacity regulations. By way of
comparison, blood alcohol content which can be accurately measured
within limits of ^ 1 percent may become a conclusive presumption of a
traffic law violation only at a level where all qualified observers would
agree the driver is too impaired to be allowed on the highway.5*
A second legal principle invalidates regulations which do not conform
to the governing statute.5* Since the degree of opacity does not equal the
degree of pollution, and since current statutes are directed to the control
of pollution, opacity regulations issued under those statutes should be
invalid. An anti-pollution regulation which is violated by exceeding an
opacity limit creates a conclusive presumption that the degree of opacity
is irrebuttable proof of the degree of pollution. Similar legislative action
has been struck down by the U.S. Supreme Court in three recent cases.
52Lanzetta v. New Jersey, 306 U.S. 451 (1939); Connolly v. Genl. Const. Co.,
269 U.S. 385 (1936).
53In many states a blood alcohol concentration of more than 1/10 of 1% creates
a presumption that a driver is a hazard. But this is only a presumption. In Delaware
the presumption is conclusive (Coxe v. State, 281 A .2d 606 (Del. 1971)). and the
New York statute sets a higher level for conclusiveness (1971 N.Y. Laws, ch. 495).
See also, Ross, The British Law on Drinking and Driving, 60 A.B.AJ. 694 (1974).
There is, however, a world of difference between an opacity reading and blood analy-
sis. A properly conducted blood test is accurate within 10% (at 1/10 of 1% the error
would be 1/100 of 1%), and all qualified observers would agree that a person with
so much alcohol in his blood is impaired. See Standard for Devices to Measure Breath
Alcohol, 38 Fed. Reg. No. 212 (11-5-73); Alcohol and the Impaired Driver, ch.
J, 1973); Lucia, Alcohol and Civilization (1963). By comparison, as discussed
above, if all observers would have to i:;ree that a plume's opacity is 10% or more,
they would have to be shown a 30% opaque plume.
''•Addison v. Holly Hill, 322 U.S. 607 (1944); Grace Line v. Nat. Maritime Bd.,
263 F.2d 709 (2nd Cir. 1959).
67
-------
470 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
In Vlandis v. Kline" a state statute established different tuition rates
for resident and non-resident students and adopted the student's legal
address at the time of application as the sole criterion for determining
his residency throughout the entire period of attendance. Students who
had applied from out of state but had acquired all of the attributes of
bona fide residency brought suit. Sustaining an injunction against enforce-
ment of the statute, the court said:
In sum, since Connecticut purports to be concerned with residency in
allocating the rates for tuition and fees at its university system, it is for-
bidden by the Due Process Clause to deny an individual the resident rates
on the basis of a permanent and irrebuttable presumption of nonresidence,
when that presumption is not necessarily or universally true in fact, and
when the State has reasonable alternative means of making the crucial
determination. Rather, standards of due process require that the State
allow such an individual the opportunity to present evidence showing that
he is a bona fide resident entitled to the in-state rates.
In U.S.D.A. v. Murry-'* a federal statute denied food stamps to house-
holds in which a person over 17 years old was claimed as a dependent
for tax purposes. This was based on the assumption that such a household
would not contain persons who were really needy. Enforcement was en-
joined when a really poor family who were denied stamps brought a class
action. Upholding the injunction the Supreme Court said:
the Amendment wholly missed its target. By creating an irrebuttable pre-
sumption contrary to fact, the Amendment classifies households arbitrarily
along lines that have no rational relationship to the statutory scheme or
the Amendment's apparent purpose. It creates a classification which de-
nies similar treatment to all persons similarly situated and is, on its face
and by its operation as established in this record, grossly unfair. Thus,
there is both a denial of due process and of equal protection.
In Cleveland Bd. of Education v. La Fleur37 a rule was adopted re-
quiring pregnant teachers to take a leave of absence for the three month
period preceding the birth. The stated purpose was "to maintain an ade-
quate continuity of able-bodied classroom teachers." Finding the rule
unconstitutional, the Supreme Court said:
Even assuming arguendo that there are some women who would be physi-
cally unable to work past the particular cut-off dates embodied in the
challenged rules, it is evident that there are large numbers of teachers who
are fully capable of continuing work for longer than the . . . regulations
•'•!412 U.S. 441 (1973): See also The Conclusive Presumption Doctrine: Equal
Process or Due Protection? 72 Mich. L. Rev. 800 (1974) and The Irrebuttable Pre-
sumption Doctrine in the Supreme Court, 87 Harv. L. Rev. 1534 (1974),
"413 U.S. 508 (1973).
"414 U.S. 632 (1974).
68
-------
VISUAL PLUME READINGS. 471
will allow. Thus, the conclusive presumption embodied in these rules, like
that in Vlandis, is neither "necessarily nor universally true," and is viola-
tive of the Due Process Gause.
Apart from constitutional considerations, a rational anomaly is created
by a regulation (based on opacity) which can be violated only during
daylight hours when the wind direction is perpendicular to the rays of the
sun and the sky is not overcast38 and which would permit unacceptable
pollution levels at all other times.
Since plume opacity as a statutory criterion creates an irrebuttable
presumption which violates due process guarantees does this destroy
visual plume reading as an antipollution enforcement tool? Not neces-
sarily. Flume readings could be used as evidence provided their limitations
are taken into account Thus, a very opaque dry plume from a small
diameter stack could be persuasive evidence of pollution despite the size
of the particles or the wind conditions. A court could weigh such evidence
against the opinions of pollution experts and reach a meaningful conclu-
sion responsive to the purpose of the regulation.
But federal EPA and various state regulatory agencies are unwilling
to limit opacity readings to an evidentiary role. More specifically, EPA's
regulatory scheme provides concentration/mass standards to assure that
an adequate control system is designed and installed, and opacity stan-
dards to require proper maintenance and operation of the system after
installation. The agency explains this pattern as follows:
Opacity standards are a necessary supplement to concentration/mass stan-
dards. Opacity standards help ensure that sources and emission control
systems continue to be properly maintained and operated so as to comply
with concentration/mass standards. Particulate testing by EPA method 5
(Stack Tests) and most other techniques requires an expenditure of
$3,000 to $10,000 per test including about 300 manhours of technical
and semitechnical personnel. Furthermore, scheduling and preparation are
required such that it is seldom possible to conduct a test with less than
2 weeks notice. Therefore, method 5 particulate tests can be conducted
only on an infrequent basis.8*
Obviously EPA must police maintenance and operation; its concern
over the means of enforcement is reasonable. Unfortunately its rationale
3SIn the days when an opaque cloud was a nuisance simply because dense smoke
was objectionable it might have been supportable to say that the plume was objection
able in daylight when people could see it but not at night when they could not see it
This approach is not valid when the object is the prevention of pollution. Unfoitun-
ately the Superior Court for Los Angeles disagreed when it upheld ar three minute pei
hour opacity provision incorporated in an anti-pollution statute. See People v. Interna
tional Steel, supra note 12, at 590.
i939 Fed. Reg. No. 47, at 9309 (3-8-74).
69
-------
472 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
becomes unreasonable when it limits the criterion of violation to an unre-
liable measurement and denies an accused operator the opportunity to
submit other evidence to prove that the equipment was, in fact; adequately
maintained and operated.
EPA's position is:
Opacity of emissions is indicative of whether control equipment is prop*
erly maintained and operated. However, it is established as an inde-
pendent enforceable standard, rather than an indicator of maintenance
and operating conditions because information concerning the latter is
peculiarly within the control of the plant operator. Furthermore, the time
and expense required to prove that proper procedures have not been
followed are so great that the provisions of 40 CFR 60.11(d) by them-
selves (without opacity standards) would not provide an economically
sensible means of ensuring on a day-to-day basis that emissions of pol-
lutants are within allowable limits. Opacity standards require nothing
more than a trained observer and can be performed with no prior notice.
Normally, it is not even necessary for the observer to be admitted to the
plant to determine properly the opacity of stack emissions. Where ob-
served opacities are within allowable limits, it is not normally necessary
for enforcement personnel to enter the plant or contact plant personnel.
However, in some cases, including u'mes when opacity standards may not
be violated, a full investigation of operating and maintenance conditions
will be desirable. Accordingly, EPA has requirements for both opacity
limits and proper operating and maintenance procedures.64
Several conclusions are revealed by this statement:
1. Investigation without an opacity reading can be used to determine
proper maintenance and operation.
2. Evidence other than plume readings may be used to prove by the
Agency a source is in violation even if its plume complies with
the allowable opacity limits.
3. When the plume is read to exceed the opacity limit, the operator
will not be permitted to use the same type of evidence to prove
compliance.
This is unreasonable. If the Agency can find a source to be in viola-
tion even though it does not exceed allowable limits, then the operator
whose source exceeds the limit should be permitted the opi. aunity to
prove it is not in violation.81 This is not possible when opacity limits are
«o n,id.
61 Pessimism on this point may not be warranted. The U.S. Supreme Court de-
clined to pass upon the holding by the lower court (the Colorado Court of Appeals)
that the failure to give notice deprived the defendant from securing rebuttal evidence
and therefore the fundamental elements of due process were lacking. See 510 P.2d
907, 909-910 (Colo. 1973).
70
-------
VISUAL PLUME READINGS 473
written into the regulations or are treated as irrebuttable presumptions of
violation.
EPA's position on opacity readings has been questioned by the U.S.
Court of Appeals for the District of Columbia Circuit in Portland Cement
Association v. Ruckelshaus where the court stated:
It may be, as EPA argues, that the opacity test is an important enforce-
ment tooi, and that the results of an opacity test, which is normally per-
formed at some distance from the plant by trained observers, offers a
cheaper and faster method of determining compliance than enforcement
of the particulate concentration standard. However, it is one thing to use
a method of testing to observe possible violations of a standard; it is
another to constitute that method as the standard itself. If the opacity
test is to be a standard, and if violations can result in enforcement actions
without further testing, the standard must be consistent with the statute
and congressional intent63
This comment is not a final ruling and the case remains pending. In
that case and companion cases,** minimum opacity levels of 10 percent
and 20 percent are questioned. This contrasts with existing decisions
upholding plume evaluations at much higher levels of opacity.94 Perhaps
the minimum opacity levels at issue in the current cases will dramatize
the limitations of visual plume evaluations sufficiently that the final deci-
sions in these cases will invalidate plume opacity as a regulatory standard.
Today, plumes which are only barely discemable may carry pollutants
exceeding allowable limits under current anti-pollution laws. Old methods
of measuring opacity cannot evaluate these faint fumes with sufficiently,
reliable accuracy to justify continued reliance on this eyeball method to
enforce present day requirements. Pressing visual evaluations beyond their
effective limits is futile. No matter how much expertise is applied or how
many readings are made, it is impossible to measure the size of an atom
with a yardstick.
"Portland Cement Assn. v. Ruckelshaus, 486 F.2d 373, 400 (D.C. Cir. 1973).
"Essex Chemical Co. v, Ruckelshaus and Appalachian Power Co. 4S6 F.2d 427
(D.C Cir. 1973).
84In Huron Portland Cement Co. v, Detroit {supra note 17), the opacity level
was not in issue. In the International Steel. Plywood Mfgrs, and Fry cases (notes 12,
13, 14 supra), the opacity readings ranged from 40% to 80%.
71
-------
474 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
Appendix A
Table 1. Evaluation of Black Training Plumes
By Trained Smoke Inspectors
In-stock trans, %
15
30
45
60
75
90
Equiv. Ring. No.
3.50
3.00
2.40
1.90
1.30
0.50
Insp. No. 1
3.68
3.08
2.65
2.14
1.65
120
Std. Dev.
0.34
0.34
0.37
0.24
0.33
0.38
Error
+0.18
+0.08
+ 0.25
+0.24
+0.35
+0.70
Insp. No. 2
3.58
2.95
2.53
2.03
1.55
1.10
Std. Dev.
0.32
022
0.31
0.34
0.24
0.37
Error
+0.08
-0.05
+0.13
+0.13
+0.25
+0.60
Insp. No. 3
3.58
3.00
2.67
2.30
1.68
1.-23
Std. Dev.
0.24
0.35
0.33
0.29
0.34
0.39
Error
+0.08
0
+0.27
+0.40
+0.38
+0.73
Insp. No. 4
3.38
2.70
2.10
1.75
1.45
0.84
Std. Dev.
0.42
0.33
0.30
0.19
0.31
0.24
Error
-0.12
-0.30
-0.30
-0.15
+0.15
+0.34
Insp. No. 5
3.70
2.94
2.43
1.78
1.35
1.03
Std. Dev.
0.50
0.28
0.32
0.26
0.20
0.21
Error
+0.20
-0.06
+0.03
-0.12
+0.05
+0.53
Insp. No. 6
3.50
3.00
2.35
1.98
1.43
1.10
Std. Dev.
0.34
0.25
0.33
0.31
0.24
0.12
Error
0
0
-0.05
+0.08
+0.13
+0.60
Table 2. Evaluation of White Training Plumes
By Trained Smoke Inspectors
In-stock trans, %
Equiv. Ring. No.
15
30
45
60
75
90
4.20
3.30
2.60
1.80
1.00
0
3.90
2.94
2.42
1.70
1.03
0.03
0.28
0.10
0.40
0.33
0.34
0.09
-0.30
-0.36
-0.18
-0.10
+0.30
+0.03
3.92
2.83
2.59
1.78
1.13
0.16
0.31
0.30
0.41
0.20
0.41
0.21
-0.28
-0.47
-0.01
-0.02
+0.13
+0.16
3.73
2.97
2.47
1.70
0.97
0.17
0.24
0.38
0.46
0.42
0.56
0.17
-0.47
-0.33
-0.13
-0.10
-0.03
+0.17
4.28
3.55
3.15
2.55
1.98
1.18
0.31
0.44
0.28
0.37
0.21
0.23
+0.08
+0.25
+0.55
+0.75
+0.98
+ 1.18
4.18
3.58
2.78
2.38
1.80
1.06
0.23
0.35
0.24
0.32
0.29
0.11
-0.02
+0.28
+0.18
+0.52
+0.80
+ 1.06
4.08
3.58
3.08
2.53
1.90
1.10
0.37
0.46
0.30
0.51
0.30
0.17
-0.12
+ 0.28
+0.48
+ 0.73
+0.90
+ 1.10
Insp. No. 1
Std. Dev.
Error
Insp. No. 2
Std. Dev.
Error
Insp. No. 3
Std. Dev.
Error
Insp. No. 4
Std. Dev.
Error
Insp. No. 5
Std. Dev.
Error
Insp. No. 6
Std. Dev.
Error
72
-------
Appendix B
Tabulation of Smoke School Results
Transmissomctcr Reading
0
5
10
15
20-25
30-35
40-45
50-55
60-65
70-75
80-85
90-95
100
-20
-15
2
7
16
22
17
10
9
3
2
-10
5
12
28
67
48
27
34
25
17
4
-5
2
15
52
93
104
70
48
37
24
26
0
5
7
31
24
100
114
141
109
60
42
32
35
6
5
6
4
31
31
93
100
103
70
44
36
19
12
10
1
1
23
15
58
51
58
51
22
22
9
2
15
6
6
14
16
18
23
17
4
2
1
20
73
-------
476 NATURAL RESOURCES LAWYER VOL. VII, NO. 3, SUMMER 1974
Appendix C
Table 1. Observer Estimates and Smoke Generator Readings of
Smoke Density for Test No. 1
Reading Period'*'
Observer I 2 3 4 5 67 8 9 10 11 12
A 2*4 2 li/4 4 2% 4*4 1*4 4 1 1% 3*4 %
B 1% 1*4 'A 3*4 2*4 4*4 1*4 4*4 % 1% 2% 1*4
C 2% 2*4 1 3% 3 4*4 2% 4*4 34 2 4 1V4
D 2 1*4 1 3 2*4 4 1*4 4 1*4 2 3V4 2
E 2*4 2 1 3*4 2*4 4*4 1*4 4*4 *4 1 3V4 1%
Smoke Generator"" 2*4 1% % 3>/i 2*4 4*4 1% 4Ya Ya 1*4 3*4 1*4
(a) Entries are Ringelmann number estimates.
(b) Based on smoke generator transmissometer reading.
Appendix D
Table 2. Observer Estimates and Smoke Generator Readings of
Smoke Density for Test No. 2
Reading Period'*'
Observer 1 2 3 4 5 6 7 8 9 10 11 12
A 334 4 *4 4 1*4 4 3% 4 3% 1*4 *4 4
B 3% 3% *4 3*4 2*4 4*4 3*4 3 2*4 1% *4 4
C 4*4 4% 1 4V4 2*4 4*4 4 4 3*4 2 Ya 4*4
D 4 4*4 1*4 4Ya 2 4*4 4 3*4 3 2*4 1 4V4
E 4*4 4*4 *4 4*4 1 Ya 4*4 4% 4 3 1*4 *4 4V4
Smoke Generator"" 4*4 4*4 Ya 4*4 1 Ya 4V4 4*4 4*4 3Ya 1 Ya *4 4*4
(a) Entries are Ringelmann number estimates.
(b) Based on smoke generator transmissometer readings.
Appendix E
Table 3. Observer Estimates and Smoke Generator Readings of
Smoke Decsity for Test No. 3
Reading Period'"
Observer
I 2 3
4
5
6
7 8 9 10 11
12
A
4 3% 3*4
1
5
*4
*4 3*4 3*4 3% 2*4
0
B
3Ya 3*4 2Ya
1*4
5
%
*4 244 3*4 4*4 2%
0
C
4*4 3Ya 3*4
3
5
*4
% 3 3 3 3*4
0
D
4 3*4 2*4
m
5
*4
*4 2 2% 3Ya 2*4
V4
E
4 4 3
1
5
*4
% 134 2*4 3 Ya 1
*4
F
4*4 4 3
2 Ya
5
%
*4 2*4 3*4 3Ya IYa
*4
Smoke Generator""
4 3*4 3
1*4
5
*4
Vi IYa 214 4 134
*4
(a) Entries are Ringelmann number estimates.
(b) Based on smoke generator transmissometer readings.
74
-------
VISUAL PLUME READINGS 477
Appendix C (continued)
Reading Period'*'
;; 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
3ia y4 4Vi 3% 2% % 3Vi 2V4 1% % 4 3 V* 1 3V4 4V4 Vi 3 2V4
v* 2?i 1 3% 3 2V4 V* 3Vi 1% IVi Vi 4V4 2Vi V4 Vi 2Vi 4Vi W 2V4 1%
"i -'i rn 4V4 3Vi 3% 2 4 2M 2V4 V& 4M 3>/i Vi 1 3Vi 4% Vi 3*4 3
'•••. 34 2 3% 3V4 3 2 4 2Vi 2V* IVi 4>A 3V4 1 V/t 2V4 4V4 Vi 2% 2V4
-«? 34 IVi 4 3Vi 2Vi IVi 3% IVi 1 Vi 4V4 2 % 2Vi 4% Vi 2*4 1*4
} 3% H4 41/4 3H 3 m 4 2 IVi % 4V4 3 Vi 1 3*4 4% *4 3V4 1%
Appendix D (continued)
Reading Period <•>
•3 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
34 3 34 2Vi 2 2*4 4 4% V4 3V4 4 3 1*4
24 2»/4 1% 1H IVi 2 2% 5 Vi 1% 2*4 1% IV4
5'u 3W 3Vi 3Vi 2 2Vi 4 4% % 2V4 3% 2% 2
3U 3 214 2 IV* 1% 4 4%% 1% VA 214 1%
3:i 3 2*4 2*4 IVi 2 4*4 5 Vi 2% 3*4 2 IV*
>4 34 2% 2Vi IVi 2*4 4% 5 *4 3*4 3% 2Vi 1*4
Appendix E (continued)
Reading Period'*'
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
2 414 1 2 % 3 3 3 14 2% Vi 4% IVi 3*4 Vi Vi *4 2 1*4 3*4
24 4*4 1% 234 2*4 414 3Vi 4 2 3 1*4 4*4 2*4 3% 114 *4 IVi 3Vi 214 4
3*4 4 2W 2M 2 3W 3 3 1 3 % 4*4 2 3*4 *4 *4 1% 3 2*4 3V4
2 4 2 2*4 IVi 3*4 2% 3 1*4 214 % 4 1?4 3V4 1V4 *4 % 2*4 2 4
Hi 4 % 1% 3A 314 2Vi 3*4 4 1*4 14 4Vi 114 3Vi Vi 0 Vi 2*4 1 3%
24 4 2 2?4 134 3*4 3*4 3 1*4 2*4 Vi 4Vi 2% 3 % *4 1 2*4 2 394
1% 4 1 l?-4 1 31/4 2Vi 3V4 *4 1?4 Vi 4Vi IVi 3*4 % V* Vi 2*4 1% 3*4
75
-------
478 NATURAL RESOURCES LAWYER VOL. vn, NO. 3, SUMMER 1974
Editor's Note : The subject matter of the preceding article is, at the pub-
lication date, before governmental agencies and courts. After the type was
set on the preceding article, two material developments occurred. Rather
than revising the text, those developments are noted here:
1. The U.S. Department of Commerce has questioned the propriety of
opacity evaluations as standards under anti-pollution legislation and has
urged this position upon the U.S. Environmental Protection Agency, which
does not agree. The Department's position is stated as follows:
The Department of Commerce believes that opacity limits have not been
satisfactorily correlated to give rates of particulate concentration emissions
or mass emissions to establish capacity as a standard. Further, Commerce
has questioned whether such standards would be subject to accurate visual
determination. Commerce, therefore, recommended that opacity limits not
be adopted as a standard where a particulate concentration or mass emis-
sion standard is established. Commerce believes such opacity limits should
only be used in those cases to create a rebuttable presumption of a viola-
tion of the particulate or mass emissions standards. 39 Fed. Reg. 37,466
(1974); see also Fed. Reg. 37,922 (1974).
2. The Environmental Protection Agency, in connection with the re-
mand of Portland Cement Association v. Ruckelshaus, 486 F.2d 427
(1973), increased the opacity standard for cement plants from 10 per-
cent, as originally promulgated, to 20 percent, asserting that it would be
extremely rare that a plant which was meeting the mass emission stan-
dards would generate a plume of greater opacity. 39 Fed. Reg. 39,872
(1974).
76
-------
SECTION C: OPACITY AND VISUAL EFFECTS OF SMOKE PLUMES
Page No.
C-1. Measurement of Opacity by Transmissometer and Smoke
Readers, William D. Conner, EPA Memorandum Report, 1974 .. . 79
C-2. "Development of an Improved Smoke Inspection Guide," Andrew H.
Rose, Jr., et al., APCA Journal. August 1958 87
C-3. "Opacity and Visual Effects of Smoke Plumes," W. Conner,
Excerpts from EPA Report 650/2-74-128, November 1974 .... 93
-------
C-l
UNiTED STATES ENVIRONMENTAL PROTECTION AGENCY
subject: Measurement of Opacity by Transmissometer date: October 16, 1974
and Smoke Readers
from: William D. Conner, CPL
to-. R. L. Ajax
Attached are some observations on opacity and (1) its relationship
with the mass of particulates in an effluent and (2) its evaluation by
smoke readers. You will note that I have assumed opacity to be an
intrinsic optical property of an emission that can be accurately measured
by a light transmission instrument with the proper design specifications.
This assumption is consistent with the definition of opacity which has
been stated to ba "the degree to which emissions reduce the transmission
of light and obscures the view of an object in the background" (F.R. 36:
247, December 23, 1971). It is also consistent with the smoke reader
(Method 9) certification procedure where the smoke readers attend smoke
schools and are certified when they learn to read opacities of training
plumes with acceptable accuracies relative to transmissometer measurements
of their opacities.
79
-------
Opacity-mass relationship
The relationship between the light transmittance and mass concen-
tration of the particulate in an effluent has been described (See
Conner and HodkinsonC) and Pi 1 at and Ensor(^). It is expressed by
the equation:
were C = the mass concentration
m
T = the light transmittance (opacity = 1-T)
L = stack or plume diameter (path length of transmittance
measurement)
K = constant that depends on the particulate characteristics of
the emission
Due to the dependence of K on the characteristics of the particulates
in the emissions (average size, size distribution, refractive index,
density, shape), the quality of the opacity-mass concentration relationshio
for an emission source is usually related to the stability of the particu-
late characteristics of the source. The average size and density of the
particulates in the emission are generally the most important particulate
characteristics affecting the relationship; except, particle size is of
little importance if the particles are small (less than one micron) and
absorb light (e.g. carbon). Refractive index is an important characteris-
tic when the average size of the particulates is below three or four
microns, but is of little importance when the average size exceeds three
or four microns. The size distribution (polydispersity) and shape
(irregularity) of the particles are important because they reduce the
effects of the other particulate characteristics on the opacity-mass
concentration relationship.
Although general observations on the expected effects of different
particle characteristics on the opacity-mass concentration relationship
can be made, accurate calculations of the relationship from the particulate
characteristics of sources are generally not possible due to the inter-
dependent effects of the characteristics, the limited understanding of
particle shape effects, and the difficulty of obtaining the necessary
data on the particulate characteristics of the sources. Consequently,
emperical determinations of the relationship are usually preferred. An
emperical determination is made by installing a light transmission
instrument in the stack to monitor the opacity of the source being
studied while making concurrent particulate mass measurements. To
80
-------
2
establish a functional opacity-mass concentration relationship for a
source, the tests are best conducted on a source with good control
equipment to permit measurements over a range of emission levels. The
range of emission levels is obtained by reducing the efficiency of the
control equipment. Some variation in the emission levels may also be
obtained by changing plant operations but plant operators are usually
less receptive to this method. Emperical opacity-mass concentration
data are being obtained for sources in most categories for which new
source performance standard opacity standards have been promulgated.
Data on the opacity-mass concentration relationship of a coal-
fired power plant have been published by Schneider.Data for kraft
recovery furnaces have been reported by the National Council for Air
and Stream improvement,and data for a cement plant have been pub-
lished by Buhne.^5' These data are shown in the attached figures. The
data are for the stack size (effluent depth) indicated on the opacity
scales. To convert to a different stack size, it is necessary to divide
the mass concentration scale by the ratio of desired size to the indi-
cated size.
The data reported by Buhne were obtained at a rotary kiln cement
plant in Germany. These data were collected over a range of plant
operating conditions. Opacity-mass concentration data are also shown
for a rotary kiln cement plant in the U.S.A. These data were obtained
by Cotrell Environmental Sciences and are reported by Kelly.I6' The
agreement between the data from these two plants is very good. Additional
tests at a third cement plant are planned in November.
The data shows that the opacities of the sources were correctable
with the mass concentrations of the particulates in the emissions. This
means that the transmissometer may be calibrated to monitor the mass
concentration. To monitor the mass emission rate, it would be necessary
to also monitor the stack gas velocity.
Mt = Mc VA
where ¦ emission rate
M„ = mass concentration
c
V ® gas velocity
A = area of stack cross section
The data shown in the attached figures were obtained at one or two
plants; consequently, application of the data to all plants within a
source category implies that the mass-opacity relationship is not
significantly different. This conclusion is not fully justified without
81
-------
3
further investigation although limited data available looks very promlsin
A complete compilation of these and additional data will be published in *
early 1975.
Visual opacity determination
The assessment of the opacity of white plumes by smoke readers has
been observed to be dependent upon the plume's background, the plume's
environmental liobting conditions and the observer viewing position.
Crider and Tash(7' noted the variability in the contrast between white
plumes and their background when viewed under different environmental
lighting and background conditions. Conner and Hodkinson(') observed
that smoke readers viewing white plumes on clear days with the sun in
front of them evaluated the plumes higher than observers viewing the
same plumes with the sun behind them. Halow and ZeekW observed that
smoke readers evaluate the amount of light a white plume scatters from
the sun and its surroundings toward them relative to the brightness of
the plume's background, not the reduction of transmitted light (opacity
of the plumes).
These observations do not apply to a black plume that scatters
little light. The appearance of the black plume is due primarily to its
reduction of transmitted light from its background. Nevertheless, even
black plumes scatter some light from the sun and surroundings toward the
observer and should not be evaluated against a dark background where the
scattered light relative to the background would be high. In practice,
however, this is not a problem since a sufficiently contrasting background
will normally be available except at night.
Due to the effects of viewing conditions on the opacity evaluation
of a plume by a smoke reader, and the many types of plumes and the
variability of the viewing conditions confronting the smoke reader in
the field, it is necessary to consider the intent of the smoke school
and certified smoke reader (Method 9).
The objective of the school should be to train smoke readers to
evaluate the opacities of good quality light scattering (white) plumes
and good quality light absorbing (black) plumes under conditions where
they are most visible and exhibit the highest opacities to observers.
Then any bias the training would produce in the smoke reader would
result in an underestimation of the opacity of plumes when evaluated
under field conditions different than training conditions. This
training is substantially accomplished with the present smoke schools.
The training smoke generators produce good quality black and white plumes
and the students generally view the plumes under conditions where they
appear to the observer to have the highest opacities.
82
-------
4
Viewing conditions for best observation of the training plumes
are generally obvious to the students. The highest opacity of the
black plume is observed when viewed against a bright overcast or blue
sky background or other light backgrounds. Dark overcast skies, hills,
or dark structures should be avoided. The highest opacity of the
white training plume is observed when viewed against dark hills, trees,
structures, or blue skies. Overcast sky backgrounds should not be used
and distant hills and structures should be avoided when the atmosphere
is hazy. A close dark structure such as a telephone pole, tree, or
building should be used as a training background for the white plumes
when the atmosphere is hazy and the sky overcast.
The observed opacity of white plumes is actually greatest on clear
days when viewed with the sun illuminating them from behind but is too
sensitive to the position of the sun for evaluation and training the
smoke reader (see scattering patterns of plumes, AP-30, p. 41).For
this reason the training and plume evaluations are made with the sun to
the observer's back where the amount of sunlight scattered by the plume
toward the observer is not sensitive to the position of the sun.
Although the training plumes represent high quality black and white
emissions that generally exceed the scatter or absorption efficiencies
of emissions from stationary sources and can be expected to have higher
observed opacities, some exceptions for white plumes can exist. The white
training plume consists of spherical (oil) particles with a refractive
index of 1.44 and a size near 0.4 micron diameter (see analysis of plume
AP-30, p. 42). 0) A plume composed of spherical particles with a higher
refractive index around 1.5 (e.g. glass spheres) and a narrow size range
around 0.5 micron diameter could produce a plume observed to have a higher
opacity than the oil plumes. The ppacity of such a plume could be over-
estimated by the observer. In practice there are probably few sources
with particulate emissions in tin's category. One such source may be a
glass furnace. Light transmission and observer evaluation field data should
be obtained for such sources. Most sources (including cement plants) are
not affected because their particulates are too large, polydisperse,
opaque or irregular.
The training of smoke readers on overcast days to properly evaluate
the white plumes on clear days may present a problem. White plumes on
clear days when illuminated by direct sunlight scatter more light toward
an observer than on overcast days when illuminated by diffuse sunlight
(see "air light" scattering by white plume on overcast day p. 22 and clear
day viewed from west p. 21, AP-30).t'' Consequently, the observer trained
on an overcast day tnay evaluate the opacities of white plunes too high
on clear sunny days. Similarly, an observer trained on a clear day may
evaluate the opacities of white plumes low on overcast days. Data
83
-------
5
available to date has not shown this to be a problem. However, further
data would be desirable. In any event, if the data indicates that this
is a problem, several possible solutions are available. One solution
would be to limit the training to times when the sun is out; however,
this may be unnecessarily restrictive. A second solution would be to
positive bias the white smoke generator opacity when traininq is con-
ducted on overcast days. The magnitude for such a compensation would
be provided by the test data. A third solution is offered by the use
of white smoke guides by the observer (see Conner, Nader, and Smith), (9)
The report describes a white guide which has four different light scatter-
ing opacities (20, 40, 60, and 80 percent) that appear like the white
training plume of equivalent opacities when viewed under the same lighting
and background conditions. Use of smoke guides may also improve the
accuracy of smoke readers and reduce the need for recertification.
A determination of the error in the evaluation of the opacity of the
black and white training plumes by trained smoke readers under training
conditions is important. It represents the maximum accuracy that can be
expected of the method. Published data to determine this is limited, a
study of the evaluations of training plumes by trained smoke readers is
reported in AP-30.* ' This data is not well suited for determining the
accuracy of the smoke readers of today. During the middle nineteen-
sixties when the data was obtained, there was little interest in reading
low opacity plumes since visible emissions standards were around forty
percent opacity. The tests were conducted with smoke readers at a smoke
school where the Ringelmann Chart was considered the primary standard and
the school's smoke generator transmissometers were calibrated in terms of
Ringelmann numbers (a procedure not used today). This nonlinear calibra-
tion resulted in white plumes below thirteen percent opacity being assigned
0 Ringelmann number. Black plumes below five percent opacity were assigned
0 Ringelmann number. In addition, the smoke readers were not trained to
read smoke only with the sun to their back (a procedure now taught at the
schools).
Extensive tests at a smoke school should be conducted with certified
smoke readers to determine the accuracy of the readers and verify that
the smoke reader training procedures are under conditions of highest
observed opacities. The tests should be conducted with the black and
white training plumes under the four different major plume lighting and
background viewing conditions confronting the reader (1) overcast sky,
hill background (dark object), (2) overcast sky, overcast background,'
(3) clear sky, hill background (dark object), (4) clear sky, sky background
On the clear days the sun should be at least twenty degrees to the observer's
back. Similar comparative in-stack transmissometer-observer field tests
should be conducted at plants in the field.
84
-------
6
Much of this data is being developed by the Methods Standardization
and Performance Evaluation Branch-QAEKL. They have contracted Southwest
Research Institute to evaluate the performance of Method 9. SIJR has
conducted Method 9 tests with trained smoke readers using the white and
black smoke school generators and coal-fired power plant emissions at a
field site. Preliminary results of the test with the smoke school
generators indicated that the 7.5 percent opacity accuracy requirement
for certification was maintained by the readers.
85
-------
7
References
1. Conner, W. D. and J. R. Hodkinson, Optical Properties and Visual
Effects of Smokestack Plumes. U. S.-Public Health Service Publica-
tion Mo. 999-AP-30, Cincinnati, Ohio, 1967. 89p.
2. Pilat, M. 0. and 0. S. Ensor. Plume Opacity and Particulate Nhss
Concentration. Atm. Environment. 4:163-173, April 1970.
3. Schneider, W. A. Opacity Monitoring of Stack Emissions: A Design
Tool with Promising Results. The 1974 Electric Utility Generation
Plan Book. McGraw Hill, Mew York, 1974, p. 73.
4. National Council for Air and Stream Improvement, Test Report Pre-
sented West Coast Regional NCASI Meeting, Portland, Oregon, 1972.
5. Buhne, K. W. and L. Duwel. Recording Dust Emissions in the Cement
Industry with the RM-4 Smoke Density Meter made by Messrs. Sick,
Staub. 32(8):19-26, August 1972.
6. Kelly, D. S., Preliminary Test Results of Marquette Cement
Manufacturing Company. Contract Mo. 68-02-0239, Task Mo. 11.
Cottrell Environmental Sciences. Bound Brook, N. J. August 26, 1974.
7. Crider, W. L. and J. A. Tash. Study of Vision Obscuration by Non-
Black Plumes. J. Air Pollution Control Association. 14:161-167,
May 1964.
8. Halow, J. S. and S. J. Zeek. Predicting Ringelrcann number and
Optical Characteristics of Plumes. J. Air Pollution Control
Association. 23:676-684, August 1973.
9. Conner, W. D., J. S. Nader, and C. F. Smith. Development of a
Smoke Guide for the Evaluation of White Plumes. J. Air Pollution
Control Association. 18:748-750, November 1968.
86
-------
C-2
Development of an Improved Smoke
Inspection Guide*
ANDREW H. ROSE, JR., JOHN S. NADER and PHILIP A. DRINKER
U. S. Department of Health, Education and Welfare
Public Health Service
Robert A. Taft Sanitary Engineering Center
Cincinnati, Ohio
Present methods of black smoke in-
spection generally involve visual ob-
servation of the degree of blackness
of the smoke as it leaves the source,
and comparison to a shade of stand-
ard blackness. Although all methods
of visual bl.ick smoke density observa-
tion share many drawbacks'11, these
methods are, for economic reasons,
the only ones open to most municipal
enforcement officials. Standards used
by air pollution control agencies are
usually of one of 2 basic types: (1)
Reflectance type, which includes the
Ringelmann Chart12,31 and smaller
card sized versions'4' reproduced to
give equivalent shades of blackness;
(2) Transmittance type t5, fll, includ-
ing those comparators employing film
and other translucent reference slides
which are viewed against the sky,
generally through a collimating sys-
tem.
Reflectance and Transmittance
Type Standards
The standards of both categories
have disadvantages as well as advan-
tages. The major drawback of the
reflectance type of smoke inspection
guide11' #) is that a smoke plume,
whose shade of blackness is due pri-
marily to transmitted light, is visually
compared to an opaque reference giv-
ing only reflected shades of blackness.
Variations of the intensity and color
of the incident light will vary the
relative appearance of the smoke and
reference standard. Hence a plume
of constant density may under differ-
ent illumination give quite different
Ringelmann values. The principal ad-
vantage of the reflectance type is sim-
plicity and low cost.
Transmittance type smoke inspec-
tion guides, while overcoming the ma-
Presented at 50th Annual Meeting of
the Air Pollution Control Association,
June 2-6, 1957, St. Louis, Mo.
jor drawback of the reflectance type,
are at present expensive. The use of
translucent reference shades of black-
ness enables a comparison of the
smoke and transmittance guide to be
made, based on the same optical prin-
ciple of operation.
The objective of the present study
is to evaluate the performance of both
guide types and based on this evalua-
tion to develop a smoke inspection
standard embodying the following
•SfSSaKT.
¦ „ . ' Ski* '
Fig. I.
characteristics: (1) The new smoke
guide must be simple, inexpensive,
and suitable for general field use by
smoke inspectors; (2) the reference
guide should correlate with the black
smoke plume under field conditions;
(3) to be useful to enforcement
agencies without requiring extensive
revision of legislation, the new guide
must be capable of correlating with
the nominal standards currently in
Test Apparatus and Procedures
To establish a correlation of the
comparator guides to the black
smoke under an extreme range of field
lighting conditions, a source of
smoke was needed which could be
maintained constant over extended
periods of time. Due to the fluctuat-
ing nature of combustion smoke, a
simulated smoke standard was de-
veloped which exhibited the same op-
tical behavior as black smoke under
field lighting conditions. It con-
sisted of an aqueous suspension of
colloidal carbon contained in a clear
Lucite cell 3 in. thick and 24 in. on
a side (Fig. 1). The use of such a
controlled smoke removed one of the
major variables encountered in the
field and permitted a direct evalua-
tion of the guides for changes in
lighting conditions.
The selection of the guides used in
the study was based on their avail-
ability and ease of manipulation.
The U. S. Bureau of Mines Ringel-
mann Chart was chosen as repre-
sentative of the reflectance type
guide, and exposed photographic
film as representative of the trans-
mittance type guide.
Preliminary Investigation
In practice the inspector's evalua-
tion of a smoke involves a subjective
judgment. The influence of psycho-
logical factors such as learning, light
contrast, and environmental interfer-
ences makes necessary a statistical
treatment of the data for a proper
test evaluation. A preliminary sta-
tistical study was made to evaluate
the performance of a film type guide
as compared to a Ringelmann type
chart. Tests were performed in the
laboratory using simulated smoke,
controlled lighting and background
conditions, and a panel of 4 observ-
ers selected from a group of six sub-
87
AUGUST 1958
JOURNAL
-------
Fig- 3. (Below)
jects tested. The results demon-
strated better correlation between the
smoke and film values and greater
stability of the film readings under
these conditions.
For a performance evaluation of
the guides under field conditions, the
present study was conducted under
atmospheric illumination. To elimi-
nate the subjective elements inherent
in the preliminary study, objective
measurements were made using pho-
tometric instrumentation.
Apparatus and Procedural
The instrumentation used in this
investigation is shown in Fig. 2. A
photometer fPhotovolt, Model 520-M
photomultiplier) selected for its
high sensitivity was used in taking
all light intensity readings. This in-
strument will measure luminous flux
over 4 decimal ranges extending
from a minimum of 5 x 10-8 to
5 x 10-3 ft. candles. To determine
the amount of light transmitted by
the smoke or film strip or reflected
by the Ringelmann Chart, it was nec-
essary to limit the light reaching the
photometer to that emanating from
the specific areas under observation.
For this purpose, a collimating tube
was constructed which permitted a
maximum scanning angle of 1.6°. A
sighting tube mounted on the barrel
of the collimator permitted accurate
selection of the areas to be scanned.
In the experiments the instrument
was set facing approximately north,
with the collimator sighted just
above the horizon. The smoke
standard and Ringelmann Chart were
placed 50 ft. in front of the instru-
ment at a height determined by its
line of sight (Fig. 3).
At 50 ft. the collimator subtends
a field of view 17 in. in diam. To
provide this necessary area, the
Ringelmann Charts as well as the
standard smoke cell were made a
minimum of 24 in. on a side. The
Ringelmann Charts were prepared
by cutting and mounting 12 of the
6 by 9 in. published charts on large
pieces of white poster board. The
smoke cell and chart to be studied
were mounted on a frame which was
set normal to the instrument line of
sight.
The film guide tested consisted of
small film strips of the same neutral
coloration as the standard smoke,
and for testing purposes was
mounted 1 in. from the end of the
collimator. This distance was 19 in.
from the photomultiplier and corre-
sponds to the approximate distance
the chart would normally be held
from the eye. Preliminary tests
proved that the observed transmis-
sion of the film remained essentially
constant for the instrument inde-
pendent of the distance.
In conducting the performance
tests, light intensity readings were
taken at */> hr. intervals between
11:00 A.M. "and 5:00 P.M. To ob-
tain a complete set of readings, light
intensity was measured as received
from the film, the smoke, the Ringel-
mann Chart, and the adjacent sky
background. During tests conducted
in clear weather, a transit set^ along
the projected axis of thq c°lh*nator
was used to determine the position of
the sun at the time of each set of
readings.
Smoke Standard Verification
Atmospheric illumination is both
a complex and variable system, con-
sisting of diffuse sky light and direct
sunlight. The relationship of this
complex system to the light proper.
ties of a black smoke plume will Je.
pend on three conditions: first, the
degree of overcast or haze as it af.
fects the intensity and constancy Qf
diffuse light; second, the intensity
and orientation of the incident sun-
light; and third, the variance of ;n.
tensity ratios of diffuse to direct
light.
An evaluation of the relationship
of the physical to the optical charac-
teristics of the black smoke plume
under atmospheric illumination indi-
cates: first, the suspensoids compris-
ing the black plume, are predomi-
nantly solid carbon particles which
are opaque and have extremely ]ow
reflectance; second, the particle size
spectra lie above the size range nec-
essary to produce significant light
interference phenomena. Consequen(.
ly, under conditions of atmospheric
illumination, the plume can be ex-
pected to exhibit the optical proper-
ties of low light scattering an(j
high light absorption.
In order that the performance of
the 2 guide types be studied un,ier
conditions which are representative
88
PCA
VoL 8' No. 2
-------
20 —
s
M
3
5 w
3
3
m
««.
(0
SO
100
ls- LIGHT DECEIVED FROM SMOKE STAMAM
lf. LIGHT RECEIVED FROM FILM ST*If
l„- LIGHT RECEIVED FROM RINGLEMAM CHART
l8- LIGHT RECEIVED FROM SKY UCKGROUND
"T
M RIMLEMANN
12
35°
61°
27°
«S®
II®
$6®
7°
M®
a*
«S®
2 1
Tl* OF MT
*1® 50° 60° W® 76®
01® 57® 5»° W° tt®
•I®
36°
«7°
J0°
»l®
2S°
SWI'S POSITION, HORIZONTAL DEVIATION AW VERTICAL ELEVATION
Fig. 4. 4/30/57 ind 5/1/57, elur tunny
-------
• J, relating percent blackness (de-
lined as the light intensity received
from the smoke standard or guides
divided by the light intensity re-
ceived from the sky background) of
the inspection guide and the smoke
: tandard relative to the sky back-
ground (Fig. 4) show 3 significant
facts: (I) the percent blackness of
the smoke standard is independent
of the sun's orientation and of the
variation in its illumination intensi-
ty: (2) the percent blackness of the
f Im strip correlates with that of the
s noke standard and is, furthermore,
independent of the variations in the
illumination system; (3) the percent
blackness of the Ringelmann Chart,
however, is dependent on the orien-
tation of the sun and on the varia-
tion of its illumination intensity. Al-
though the variation in the percent
blackness of the Ringelmann Chart
is uniform, it is not predictable un-
der field conditions.
When the results are evaluated on
the basis of a direct comparison,
they are expressed as a ratio of ab-
solute values of light intensity re-
ceived from the guide to correspond-
ing values from the smoke standard
(Fig. 5, 6). The results again
verify the excellent correlation of
the degree of blackness of the film
strip with tha: of the smoke stand-
ard through varying light conditions
as indicated by a maximum deviation
of 4.2',~r. Since the degree of black-
ness of the guides relative to that of
the smoke standard is shown as a
• H-
<10
a
a
8
1
Jr
'•
. II0NT MCCIW* FROM SNMC STAMUO
. IIOMT MKCIVCD FROM IIMLDUM 11
¦ klWT DCCCIKB FHON FILM ITIIP *2
, LIWT KWVtB FRON JIT MHHW
70° ««•««• »«•
sum's vorriui amu or himtio*
Fig. 7. 4/30/57. Blacknait of smoka, film,
Ringlamann ralativa to iliy background.
ratio, it can be expressed as a per-
centage deviation. On this basis
there is indicated a trend which
shows that, under conditions of ideal
atmospheric illuminations, the per-
cent deviation of the Ringelmann
Chart increases with the degree of
blackness of the smoke standard.
A further evaluation of the per-
formance of the guides was made
based on an experiment in which the
horizontal component of the angle
of incidence of the direct sunlight to
the guides was held constant at zero
degrees, and the vertical component
of the angle of incidence was al-
lowed to vary as the sun approached
the horizon (Fig. ?)• * conclu-
sions previously reached regarding
the constancy of the J™'guide ana
the variability of the Ringelmann
Chart continued to hold under these
special lighting conditions. jn
case, however, the Ringelmann Chart
shows a decreasing percentage black-
ness as the sun's position app^^}^
the horizon (angle "' incidence of
the direct sunlight to the guide 8Ur.
face approaches zero). In the pre-
vious case, however, where the guide#
were oriented with the normal run-
ning north and south, the percentage
blackness increases as tht sun ap-
proaches the horizon. Thia differ-
ence in behavior can be explained in
terms of the cosine law of illumina-
tion, which states the illumination of
a surface varies directly as the co-
sine of the angle of incidence of the
direct light to the surface.' When
only the vertical component of the
angle of incidence of the direct sun-
light to the guides decreases, the
horizontal component being held con-
stant at zero degrees, the illumina-
tion increases since the absolute
angle of incidence approaches zero.
When again considering the first
case, in which the chart normal was
oriented north and south, the hori-
zontal and vertical components of
the angle of incidence of the direct
sunlight to the guides were both al-
lowed to vary with the sun's move-
ment. In this case the illumination
decreases since the absolute angle of
*!
m m
g £
E s
3 2
Cft M
si
s
w
to
100
II
¦«
If
>s
LIGHT RECEIVED FROM SMOKE STANOARD
LIQHT DECEIVED FROM KINQLEMAMN CHART 12
LIQKT RECEIVED FRON FIIM STRIP ft
LIGHT RECEIVED FROM SKY MCXGftOIMO
1
1
TIME OF MY
Fig. ». (Abova) 5/S/57. Blackness of smoka, film, and Ringlamann
chart ralativa to iky boekground-brokan clouds ovarcait.
Fig. 9. (Right) 5/8/57 Ratio of impaction guida to imoka black-
nasi, brokan c!ji-n't ovareait.
1.0
8
S
o
3
g o.a
8
0
>-
K
§ 0*0
8
1
| 0.4
0.1
0.0
It
lr
T
I,. LI0HT RECEIVED FROM SMOKE STARDARD
l„. LlflNT RECEIVED FRCN RIRtLENANR CHART «
lr LIMIT RECEIVED FRCN FlUt STRIP «
2 S
TIME OF DAY
of APCA
90
Vol. 8, No. 2
-------
incidence increases as the sun's posi-
tion approaches the horizon.
The general effects of the orienta-
tion of the guides are observed from
these two special case?. First, the
orientation of the film or transmit-
tance type guide has no significant
effect on the constancy of the results.
Second, the orientation of the Ringel-
mann Chart or reflectance type guide
affects the resultant light intensity or
degree of blackness. Fundamental-
ly, the orientation of the reflectance
type guide results in varying degrees
of blackness, either greater or less
than the nominal values specified, de-
pending on the angle of incidence of
the direct sunlight to the guide,
which is established by the orienta-
tion of the guide.
The second phase of the study
covers the evaluation made from
data obtained during a day of cloudy
overcast with occasional breaks in
the clouds. This was a period in
which the configuration of the cloud
cover and, hence, the over-all illumi-
nation pattern varied continuously.
Readings were not taken, however,
during intervals of direct sunlight on
the guides. The effects of atmos-
pheric illumination when evaluated
on the basis of either a direct or an
indirect comparison (Fig. 8, 9) show
3 significant facts: (1) Random vari-
ations in atmospheric lighting have
no effect on the constancy of percent-
age blackness of the smoke standard;
(2) the degree of blackness of the
film strip correlates with that of the
smoke standard and is independent
of variations in illumination system;
(3) the Ringelmann Chart is depend-
ent on (he random diffuse back-
ground and foreground sky light.
Furthermore, under broken cloudy
conditions there does not appear to
#
. "... X3
Fig. 10.
be any trend correlating with the
sun's position that would correspond
to the trend of effects of direct sun-
light observed in the first phase of
the study.
Guide Development
It is realized that all visual com-
parator type guides have an inherent
limitation in that equal optical densi-
ties of black smoke plumes do not
necessarily reflect equal contaminant
discharge to the atmosphere. The
advantages, however, of rapid and
simple application offset this limita-
tion and make this method a practi-
cal enforcement tool. For these rea-
sons it is felt that the development
of a smoke guide which is simple, in-
expensive and suitable for general
field use, and which correlates with
the black smoke plume under field
conditions, would be most useful.
A prototype guide was developed
utilizing exposed photographic film
strips (Fig. 10). The selection of a
transmittance type guide was made
based on performance results. Per-
cent blackness values were chosen
corresponding to the nominal Ringel-
mann Chart values for 2 reasons:
first, to permit the use of the guide
without extensive revision of legisla-
tion ; second, the nominal Ringel-
mann values have been shown
to occur under specific conditions of
atmospheric illumination. Enforce-
ment agencies will be given the op-
portunity to field test prototype
guides as a continuation of this
study. Evaluation of various fotms
of the guide will be made to deter-
mine ease of application, stability,
and effect of environmental interfer-
ences. A final version of a transmit-
tance type guide will be based on
these field evaluations.
Acknowledgment
The authors acknowledge the con-
tribution of Dr. M. W. Bredekamp,
Mr. J. L. Witherow, and Mr. K. A.
Busch, who conducted the prelimi-
nary statistical guide evaluation dur-
ing the summer of 1956.
References
1. Lionel S. Marks. Inadequacy of the
Ringelmann Chart. Mechanical Engi-
neering, 631 (Sept. 1937).
2. Max Ringelmann. Methode d'Eatima-
tion des Funtees Produites par let
Foyers Industriels. La Revue Tech-
nique, 26S (June 1898).
3. Rudolph Kudli:h. Ringelmann Smoke
Chart. Bureau of Mines Information
Circular 7718, (Revised Aug. 195S).
4. Anon. Power's Micro Ringelmann: A
New Aid to Smoke Inspectors. Power,
98: 90, (March 1957).
5. American Society of Mechanical Engi-
neers; Power Test Code 19, Supple-
ment on Instruments and Apparatus,
Part 20. Smolca Density Determinations,
(1951).
6. John P. Strange. The Smokescope;
Paper No. 53-S-3. Presented at the
Spring Meeting of the American So-
ciety of Mechanical Engineers, (April
19S3).
AUGUST 1958
91
JOURNAL
-------
C-3
Excerpt from EPA 650/2-74-128,Measurement of the Opacity and the Mass
Concentration of Particulate Emissions fay Transmlssometrv, W. Conner,
November 1974
OPACITY AND VISUAL EFFECTS OF SMOKE PLUMES
Two different visual effects may be attributed to plumes containing
particles that scatter and/or absorb light: (1) the plumes become
visible, and (2) they obscure the visibility of objects behind them.
These visual effects have been analyzed in this laboratory in terms of
the luminance contrast between objects (or plumes) and their surroundings.^
The work indicated that both effects are dependent not only on opacity but
on the amount of light scattered by plumes into the viewer's eyes and,
consequently, are dependent upon the environmental lighting conditions of
the plumes. In the case of obscuration of visibility by plumes, however,
2
the ambient lighting conditions have been shown to be less important.
It is the obscuration of visibility by plumes that represents the
basis for regulating plumes by a visual effect. The term opacity has been
defined as "the degree to which emissions reduce the transmission of light
and obscure the view of an object in the background."^ A theoretical
hypothesis of vision obscuration by a plume in terms of the reduction in
luminance contrast of an object viewed through the plume can be presented
to illustrate the opacity concept.
VISIBILITY OBSCURATION BY PLUMES
The luminance contrast (C^) between an object of luminance (Bt) viewed
against an extended background such as the sky with luminance (B^) is:
93
-------
If the object is viewed through a plume, its apparent luminance (B *)
and the apparent luminance of the sky behind the plume (Bs*) will depend
on the amount of light that the plume scatters from the surroundings and
sun into the viewer's eyes (B&), and on the amount of light transmitted
through the plume from the background and object. The apparent contrast
of the object (C may be written as:
B 1 R 1
ct' . -£ t- (2)
z B
s
The denominator of equation (2) remains, as in equation (1), the
intrinsic luminance of the sky (Bs) since it remains the dominant lumi-
nance of the scene. Dividing by the intrinsic luminance of the sky
simplifies the analysis by removing the light scattering term (B ) from
cl
the final results. This differs from a more complicated earlier analysis
from this laboratory in which the reduction in contrast between two con-
trasting targets located behind the plume was discussed.^ Dividing by
the intrinsic luminance of the sky would not be realistic for atmospheric
pollution or for extended sources where the scatter and absorption of light
by the particulates would affect the luminance of the entire scene. For
these cases, the denominator of equation (2) should be the apparent lumi-
nance of the sky (Bs*).
If the plume has a transmittance (T) and the amount of light scattered
by the plume toward the observer is Ba> then B^* = B^T + B& and Bg* = B T +
and equation (2) may be written:
I T
Ct 1 2 (3)
s
94
-------
The reduction in object to background contrast due to the plume may
be defined as:
c - C 1
R • -TT"5- CJ
which, after substitution of equations (1) and (3) and rearrangement,
becomes:
CR • 1 " T (5)
Equation (5j indicates that the reduction in luminance contrast and
visibility of an object behind a plume viewed against an extended back-
ground is equivalent to one minus the transmittance (1-T) of the plume.
1-T is commonly called the opacity of the plume. The relationship also
indicates that the visibility reduction is independent of the environmental
illumination, and that there is an equivalency between the opacity of
plumes and their reduction of visibility.
The equivalent opacity concept was studied by the Bay Area Air Pollu-
2
tion Control District in the mid-1960's. In that study, a group of
observers was used to determine the light transmittances (opacities) of
black (carbon) and white (oil) experimental plumes with equivalent vision
obscurations. Because the observers had no previous experience in reading
smoke, they were screened to establish a statistically consistent group.
After instruction in the use of the Ringelmann Charts for evaluating black
plumes, the observers were asked to assign Ringelmann number values to
black plumes of various densities and to observe how much the black plumes
obscured the visibility of a target located behind them. They were then
asked to view a similar target behind white plumes of various densities and
assign "equivalent Ringelmann numbers" to the plumes by equating their
95
-------
visibility reduction to the black plumes. During the tests, the opacity
of both plumes was monitored by identical photoelectric transmissometers.
The Bay Area APCD used the results of the tests to calibrate the trans-
missometers on the black and white smoke generators in terms of Ringelmann
numbers and "equivalent Ringelmann numbers," respectively. These curves
represented the average of 14,400 observer estimates of the Ringelmann
number values of the black plumes (9 observers, 80 series of 20 readings),
and 8,120 observer estimates of the "equivalent Ringelmann number" values
of white plumes (7 observers, 58 series of 20 readings).
Since the two Ringelmann number scales reported by the Bay Area APCD
are related through the equivalency in the vision obscuration produced by
the plumes, the two curves may be combined to show a general equivalency
in the transmissometer-measured opacity and the reduction in visibility
by the two plumes (Figure 1). The averaged data shown in Figure 1 indicate
that, in the important opacity region below 35 percent, black plumes have
slightly greater vision obscuration than white plumes with equivalent
opacities. In the opacity region above 35 percent, white plunes were
judged to have greater vision obscuration than black plumes with equivalent
opacities. Since standard deviations of up to 12 percent opacity are
reported for the data, the averaged curve shown in Figure 1 is within one
standard deviation of the theoretical one-to-one relationship.
Although the opacity of plumes represents a measure of how much they
obscure visibility, the plumes are generally located where they are observ-
able only against the sky, and their obscuration of visibility (obscuration
of objects behind them) is not observable. Consequently, it is the visual
96
-------
I I
OPACITY BLACK PLUME, parctnt
Figure 1. Observer estimates of the opacity of black and white
experimental plumes with equivalent vision obscuration.
appearance of plumes that is usually observed by both the casual observer
and by the trained observer assessing their opacity.
VISIBILITY OF SMOKE PLUMES
Analysis of the visibility of plumes6 in terms of their luminance
contrast with their background (C^) has shown that:
s - r - (1 - T)
s
(6)
where: B, = the amount of light scattered by the plume toward the
observer
Bg = the luminance of the background (usually the sky)
T = the light transmittance of the plume
(1-T) = the opacity of the plume
97
-------
Equation (6) indicates that the visibility of a plume depends on the
amount of light scattered by the plume from the sun and its surroundings
into the viewers eyes (B ) relative to the luminance of the plume's back-
a
ground (Bg) and the opacity of the plume (1 - T).
The dependency of the appearance of plumes on their environmental
lighting, background, and general viewing conditions makes plume visibility
or simple telephotometric contrast measurements undesirable for regulating
particulate emissions. Consequently, visible emissions regulations are
based on the opacity (visibility obscuration) of the plumes, which is an
intrinsic property of the plume and a better indicator of the particulate
emissions. It may be measured instrumentally by transmissometry or esti-
mated by trained observers.
OBSERVER EVALUATION OF OPACITY
The smoke inspector or trained observer receives training at a smoke
inspector training school where he is taught to evaluate the opacity of
black and white training plumes with prescribed accuracies relative to
transmissometer measurement of their opacities. Upon passing the course,
he becomes certified by the school as capable of evaluating the opacity of
smoke plumes from their appearance. This procedure requires the inspector
to evaluate an intrinsic optical property of the plumes (their opacity) by
observing their visibility, which is dependent not only on their opacity,
but also on their illumination and background viewing conditions.
Clearly, the trained-observer method of evaluating opacity has
limitations; e.g., at night when a plume cannot be seen, the method is
g
not generally applicable. Halow and Zeek have analyzed the visibility
of white plumes in daylight and have shown that observer evaluations depend
98
-------
upon the background and environmental lighting conditions of the plumes. A
white plume viewed against a white overcast sky is often invisible because
the amount of light it scatters toward the viewer equals the amount of
light that it attenuates from the background. The visibility of a white
plume viewed on a clear day is particularly sensitive to the viewing direc-
tion relative to the sun. The plume appears much brighter when the sun is
illuminating it from behind. Tests with trained observers^ have shown that
the evaluation of such plumes should be limited to times when they may be
seen with the sun illuminating them from the front. There are many plume
background and illuminating conditions confronting the smoke inspector in
the field that may impose limitations on the visual evaluation of the
plumes; consequently, it is important that the inspectors be trained
under viewing conditions that generally prevail in the field, and that
the inspectors know of any limitations that may result because of
deviations from these conditions.
99
-------
SECTION D: RELIABILITY OF OPACITY STANDARDS
Page No.
D-1. EPA Response to Remand in Portland Cement Association v.
Ruckelshaus, Appendix III, Part A: Opacity Standards, 1973 . 103
D-2. Excerpts from Various Collaborative Test Reports Summarizing
Results of Collaborative Testing of EPA Method 9 145
D-3. Data Graphs 1-5 of Opacity Measurements and Observations
Made During Collaborative Testing of EPA Method 9 and
Summary Tables 151
D-4. Summary of Contract Report--Examining the Properties of
Qualified Observer Opacity Readings Averaged Over Intervals
of Less Than Six Mi-nutes, T. Hartwell, 1976 159
-------
0-1
PORTLAND CEMENT REMAND RESPONSE:
APPENDIX III, PART A: OPACITY STANDARD
In the case of other points at Issue in this Remand Response,
it has been necessary only to supplement the November, 1973 Draft
Response to deal with the comments of the Portland Cement Association
and others on that Draft Response, In order to respond fully to the
comments with respect to opacity, it has been necessary to prepare the
rewritten, extensive response found in this Appendix III. The response
contained herein should therefore be considered to supercede that portion
of the Draft Response dealing with the opacity standards, and the
present Appendix III constitutes the Administrator's full and complete
position with regard to this issue.
EPA promulgated opacity limits as well as mass emission limits
primarily because EPA believes that opacity limits provide the only
effective and practical method for determining whether emission
control equipment, necessary for a source to meet the mass emission
limits, is continuously maintained and operated properly. It has
not been EPA's position that a single, constantly invariant and
precise correlation between opacity and mass emissions must be iden-
tified for each source under all conditions of operation. Such a
correlation is unnecessary to the opacity standard, since the
Administrator consistently develops opacity standards for each class
of sources at a level no more restrictive than the corresponding mass
emission limitation with due consideration given to all conditions
103
-------
12 3
of operation. ' ' Any source which is meeting the mass limit will there-
fore also be meeting the opacity limit; if a source 1s not meeting
the opacity limit, this will be due to the failure of that source to
properly maintain and operate its emission control equipment. And if
the equipment necessary to meet the mass emission standard is not
properly maintained and operated, the source will also fail to meet
the mass emission standard.
Opacity standards are a necessary supplement to mass emission
standards, since there is no reliable instrument to directly
measure mass emissions currently available. If only mass emission
standards were in effect, operators would not be compelled to assure
proper maintenance and operation of pollution control equipment at
times other than during periods of performance testing. Two weeks
or longer are necessary to schedule a typical stack test, due to the
preparation Involved. If relatively minor repairs were all that
were needed in a given case, such as pump or fan repair or replace-
ment of fabric filter bags, a plant operator could delay such
repairs and continue to pollute until shortly before the stack test
was conducted. For some types of control equipment, such as
scrubbers, the operator could reduce the input of energy (the pressure
drop through the system) thereby partially or totally defeating the
air pollution control system when stack tests were not being
conducted and through such action could release significantly more
particulate matter than permissible under the mass emission standards.
104
-------
Furthermore, stack tests can only be conducted with limited frequency,
since they often cost up to $10,000 to perform. Therefore, EPA has
required that operators properly maintain and operate air pollution
control equipment at all times [40 CFR 60.11(d)] and meet opacity
standards at all times except during periods of startup, shutdown,
malfunction, and as otherwise provided in the applicable standard [40
CFR 60.11(d)]. Opacity is established as an independent, enforceable
standard, rather than merely as an indicator of maintenance and
operating conditions, because information regarding those conditions
1s peculiarly within the control of the operator. The time and expense
required to prove that proper procedures have not been followed are so
great that the provisions of 40 CFR 60.11(d) by themselves (without
opacity standards) would not provide a practical means for ensuring
that on a day-to-day basis the pollution control equipment is main-
tained and operated 1n the way necessary to meet the mass emission
limitations. Opacity standards, 1n contrast, require only a qualified
observer and can be performed with no prior notice. In some cases, of
course, a full Investigation of operating and maintenance conditions
will be desirable. Accordingly, the opacity standards and maintenance
requirements were both promulgated, and work in tandem to guarantee
that proper maintenance and operation of pollution control equipment,
the sine qua non of continuous compliance with emission limits, can in
fact be required and monitored.
Two primary aspects of the opacity standards, raised by Petitioner
Portland Cement Association, were identified by the Court of Appeals as
105
-------
the bases of its concern: 1} the Inherent reliability of opacity
standards, i.e. whether measurements of opacity can be made with
reasonable accuracy, and 2) the achievability of the opacity standard
set for portland cement plants.
RELIABILITY OF OPACITY STANDARDS
Accuracy of Opacity Measurements in Detecting Violations
The Portland Cement Association (PCA) contends that opacity
measurements cannot be made with reasonable accuracy. PCA indicates,
in particular, that individual observer readings under certain con-
ditions are often 1n error by 10 percent opacity or more, and therefore
that "equivalent opacity—at any level—is an inherently inadequate
and invalid measure to determine compliance with, or violation of an
ant1-pollution law." In support of this contention, PCA argues that
observed opacity may be affected by such factors as: dark nights,
contaminated water vapor or detached plumes, lack of contrasting back-
ground (e.g., white plume viewed against an overcast sky), and wind
velocity.
The PCA and other commentators contend also that EPA smoke school
requirements and procedures are inadequate to ensure that qualified
observers will make accurate readings. The procedure whereby training
plumes are shown 1n advance of certification runs, and under which
observers are allowed to use fixed (non-sky) contrasting backgrounds
are noted as specific deficiencies.
In response to the questions raised by the Court and the PCA,
EPA has reviewed available data concerning the accuracy of qualified
106
-------
4 5 6
observers; ' ' has reviewed the theory of equivalent opacity,
particularly concerning the effect of variable viewing conditions on
accuracy;5,6 has reviewed the smoke school certification procedures;
and has conducted extensive field tests Involving both specially
generated plumes and industrial plumes.^ On the basis of this extensive
review, EPA concludes that the accuracy of Reference Method 9 has been
well established, that Its error tolerance 1s reasonable and 1s within
limits considered normal by the scientific and engineering coirmunity,
and that when that error tolerance 1s accounted for in assessing the
results of opacity readings, the method 1s valid and reliable for
determining compliance with prescribed opacity standards. EPA also
concludes that the smoke school certification procedures are adequate
to assure proper qualification and accuracy of qualified observers. The
basis for these conclusions 1s set forth below.
PCA contends that plumes of the same actual opacity will appear
to have somewhat different opacities under dissimilar conditions of
viewing. EPA agrees that this 1s true. A plume will appear most
visible to an observer when the plume 1s viewed against a contrasting
5 6
background. ' A white plume therefore will be more visible when
viewed against a black background or a blue sky than it will when
5
viewed against a hazy or overcast sky. Indeed, a white plume viewed
against a white background may well be invisible. PCA concludes that
because a plume will appear of lesser opacity under less contrasting
conditions, EPA's opacity standards, which limit allowable opacity
107
-------
observers;4'5,5 has reviewed the theory of equivalent opacity,
particularly concerning the effect of variable viewing conditions on
accuracy;5,6 has reviewed the smoke school certification procedures;
and has conducted extensive field tests Involving both specially
generated plumes and industrial plumes.7 On the basis of this extensive
review, EPA concludes that the accuracy of Reference Method 9 has been
well established, that its error tolerance is reasonable and 1s within
limits considered normal by the scientific and engineering coirmunity,
and that when that error tolerance 1s accounted for 1n assessing the
results of opacity readings, the method 1s valid and reliable for
determining compliance with prescribed opacity standards. EPA also
concludes that the smoke school certification procedures are adequate
to assure proper qualification and accuracy of qualified observers. The
basis for these conclusions 1s set forth below.
PCA contends that plumes of the same actual opacity will appear
to have somewhat different opacities under dissimilar conditions of
viewing. EPA agrees that this 1s true. A plume will appear most
visible to an observer when the plume 1s viewed against a contrasting
background.5,6 A white plume therefore will be more visible when
viewed against a black background or a blue sky than 1t will when
viewed against a hazy or overcast sky.5 Indeed, a white plume viewed
against a white background may well be invisible. PCA concludes that
because a plume will appear of lesser opacity under less contrasting
conditions, EPA's opacity standards, which limit allowable opacity
108
-------
of a violation when true opacity was below the standard. While
negative errors are undesirable from an enforcement point of view,
they do not place a complying source 1n jeopardy. It is only necessary
therefore to quantify the positive error, which may occur under high
contrast viewing conditions, to demonstrate that the method does
measure opacity with the reasonable accuracy required of an enforceable
standard. As the Court of Appeals notes 1n Amoco Oil Company v. E.P.A.,
501 F2d Z22' (6ERC 1481» May 1. slip op. at pp. 38-39):
Petitioners also contend that any test methods adopted
will involve a certain capacity for statistical error.
We fail to appreciate the force of this argument. The
possibility of statistical measurement error, which is
often unavoidable where regulations set quantitative
standards, does not detract from an agency's power to
set such standards. It merely deprives the agency of
the power to find a violation of the standards, in
enforcement proceedings, where the measured departure
from them 1s within the boundaries of probable measure-
ment error. (Emphasis 1n original.)
The following data provide a basis for such quantification and
lead to the conclusion that, using an average of 25 readings obtained
under the conditions prescribed in Method 9, qualified observers are
able to consistently assign opacity values under relatively Ideal
8 9
conditions, with a positive error seldom exceeding 7.5 percent opacity. '
(a) October 1973 report titled "Average Observation Error Asso-
4
dated with Smoke Plumes at Levels of Known Opacity". This report 1n
addition to showing that each observer met the +_ 7.5 percent require-
ment, shows that when the readings of all observers are averaged, the
largest positive error associated with a given opacity level was 5.0
109
-------
percent opacity which occurred when the true opacity was 10 percent.
For plumes between 15 percent and 90 percent opacity, the average error
did not exceed + 2.8 percent opacity. This is a clear demonstration of
the inherent accuracy of the technique of determining plume opacity
through observation by qualified observers.
(b) In a test conducted November 19-20, 1973, nine qualified
observers were shown plumes generated by a smoke generator.7 The plumes
were presented in runs of 25 each during which each observer assigned
opacity using the procedures set forth 1n Method 9. A total of 133
sets of 25 observations using black smoke and 170 sets of 25 observa-
tions using white smoke were made. The results of this test are as
follows. For white plumes, 99 percent (168 of 170) of the sets of
observations were made with an error of less than +7.5 percent opacity.
N1nety-f1ve percent of the sets of observations were made with an error
of less than + 5.8 percent opacity. For black plumes, the accuracy was
even better: 99 percent of the sets of observations were made with an
error of less than + 5 percent opacity; 95 percent were made with an
error of less than +2.7 percent opacity.
(c) A test was conducted at a power plant 1n Charlotte, North
Carolina, on September 30, and October 1, 1974.7 In this test, 0
qualified observers made a total of 168 sets of 25 observations and
110
-------
assigned opacity according to the procedures of Method 9.* Each run
consisted of 25 individual readings. All observers had been certified
according to Method 9. Two of the 8 observers had not had previous
field experience and two had only limited field experience. The power
plant plume (which was monitored with an in-stack transmlssometer) was
white in appearance and was viewed against a constrasting clear blue
sky (no condensed water vapor was present in the plume). The plume was
held relatively constant during some runs and was varied during other
runs in order to simulate the various conditions which could be
encountered in the field. The test results showed that 99 percent
(166 of 168) of the observations (average of 25 readings) were made with
an error of less than + 6.2 percent opacity; 95 percent with an error of
less than + a.7 percent opacity, and 90 percent with an error of less
than +2.6 percent opacity.
(d) A test was conducted at a sulfuric acid plant on January 29-30,
1974.^ The plant was equipped with an in-stack transmlssometer.**
* This test was conducted with the 8 observers divided Into two groups
of 4 each. One group, the control group, made all readings from an
optimum viewing location. The second group made readings under differing
conditions. Only the readings which were made according to Method 9
criteria are reported here, except that 32 sets of readings are also
Included which did not conform precisely to Method 9. In these sets,
observers were located at an optimum location and assigned opacity 1n
1 percent Increments, not 5 percent increments as specified 1n the
method. The only other data collected but not Included 1n the totals
shown here, was read with the sun outside the 140° sector to the observer's
back. This is not acceptable under Method 9.
** Not required at sulfuric acid plants by EPA standards.
Ill
-------
Observer readings were made according to Method 9 against a contrasting
background and were compared with the transmissometer readings. The
results are as follows: A total of 298 sets of 25 readings were made.
Two sets of readings were made with an average error of + 8 percent
opacity. A total of nine sets of readings were made with an average
error of + 5 percent opacity or greater. The remaining 289 sets of
readings were made with errors of less than + 5 percent opacity.
In summary, a total of 636 runs of 25 readings each were made
using white plumes and 133 runs of 25 readings each using black plumes
for the specific purpose of quantifying the degree of positive error
possible in Method 9. The readings were made against contrasting
backgrounds, the condition under which a plume 1s most visible and is
most conducive to positive observer errors. The results show that
single observers, on the basis of single sets of 25 readings, are able
to assign opacity with errors of less than +7.5 percent with good
consistency (630 of 636 for white plumes; 133 out of 133 for black
plumes). Errors in excess of + 7.5 percent can occur, although very
infrequently. In short, on the basis of this extensive research of
the very sort demanded by PCA, observer readings of opacity have been
shown to be within an error tolerance of + 7.5 opacity with a frequency
of approximately 99.3 percent (763 of 769 total sets of observations).
This indicates clearly that opacity observers do 1n fact meet the + 7.5
percent opacity accuracy limit which is specified explicitly In the
introduction to (amended) Reference Method 9.
112
-------
Finally with regard to observer accuracy and the inherent relia-
bility of opacity readings, much of the concern of the Court of Appeals
is related to PCA's contention, based on an HEW study,® that at low
levels opacity readings cannot be made reliably. The Court notes^0 that
PCA's claim that Table 2 on page 28 of that study shows that the six
trained observers rated a plume known to have 0 percent opacity at more
than 0 percent opacity, and three rated 1t at more than 20 percent.
The Court notes further that a plume known to be 20 percent opacity was
rated higher than 20 percent by five of the six observers, based also
on Table 2.
A close reading of Table 2 shows, however, that Petitioners have
wholly misconstrued the nature of that test and the data presented in
Table 2. In fact, the study specifies that
"tests of trained observers show that the observers that
viewed white plumes on a clear day facing the sun assessed
the plumes at a higher Rlngelmann number (lower trans-
mlttance) than did observers that viewed the plumes with
the sun to their backs. For darker, plumes the effect was
less pronounced. Group assessments showed good agreement
for similar sun-plume-viewer qeometr1es"6 (emphasis added).
EPA agrees that viewing plumes while facing the sun may produce inaccurate
8
readings. It is for this very reason that Reference Method 9 requires
observations to be made with the sun at the viewer's back.
When Table 2 1s examined 1n this regard, the "good agreement" for
apparent similar geometries divides the observers into two groups. The
first, consisting of Inspectors 1, 2, and 3, had similar readings among
113
-------
themselves. The second group, consisting of Inspectors 4, 5, and 6,
also had good agreement among themselves. The levels of the second
group are significantly higher than those of the first group, indicating
that the second group was viewing into the sun, while the first group (1,
2, and 3) was viewing with the sun to their backs. This probability is
confirmed by the report itself, which states that no restriction was
placed on viewing direction.^ Since only Inspectors 1, 2, and 3 appa,-
ently made their observations with the sun at their backs, it is only
their results that are at all relevant to the question of observational
reliability or accuracy. These results, for the low levels of opacity
concerning the Court, are reproduced below:
Table 2. AP-30, Evaluation of White Plumes^
In-stack transmittance, percent 75% 90%
Equiv. Ring. No. 1.00 0
Insp. No. 1 1.03 0.03
Insp. No. 2 1.13 0.16
Insp. No. 3 0.97 0.17
It is at once apparent that the observations of Inspectors 1,2,
and 3 reflect a very high degree of accuracy, contrary to PCA's conten-
tions. Closer examination of this table, moreover, shows that the
data itself is presented in a way not corresponding with EPA's
standards. It will be noted that each column 1s headed by two
figures: the "equivalent Rlngelmann No." and the "In-stack Trans-
mittance, Percent." Transmittance, of course, represents the
percentage of light not blocked by the smoke, so the 75 percent
mittance is a 10 percent opacity. As the Court notes,10 the presently
114
-------
accepted equivalency between Rinaelmann numbers and opacity of white
smoke 1s 20:1, i.e. Rlngelmann No. 2 ¦ 40 percent opacity, Rlngelmann
No. 1 =20 percent opacity, Ringlemann No. .5 » 10 percent opacity,
Rlngelmann No. 0 * 0 percent opacity, etc. The Table 2 column headings,
however, reflect a different equivalency: there, Rlngelmann No. 1
corresponds to 25 percent opacity (75 percent transmlttance), and
Rlngelmann 0 corresponds to 10 percent opacity (90 percent transmittance).
Since EPA's opacity standards are predicated upon and expressed 1n terms
of the transmission and obscuration of light, it is therefore only the
transmittance figures which are at all relevant to evaluating EPA's
standards.
The discrepancy between the equivalency used 1n Table 2 and the
presently accepted Ringelmann-opaclty equivalency exists because during
the middle nineteen-slxties when the data were obtained, there was little
interest 1n reading low opacity plumes, since visible emissions stan-
dards were generally at about forty percent opacity. The tests were
conducted with smoke readers at a smoke school where^the Rlngelmann
Chart was considered the primary standard, and the school's smoke genera-
tor transmlssometers were calibrated 1n terms of Rlngelmann numbers (a
procedure not used today). This nonlinear calibration resulted 1n
white plumes below thirteen percent opacity being assigned a Ringlemann
number of zero. Black plumes below five percent opacity were also
assigned a Rlngelmann number of zero.1^ It 1s clear therefore that this
equivalency (now replaced by the standard 20:1 ratio) was not intended
115
-------
to deal with low opacity plumes, since all six inspectors 1n fact
reported seeing smoke at what was supposed to be Rlnglemann zero.
As a result of these flaws, the data in Table 2 can only be
partially converted to opacity percentages for purposes of evaluating
observer accuracy and are not completely reliable for this purpose as
the observations were not made in accordance with current procedures
specified by Method 9. Since a definite ratio of 25:1.00 was used to
calibrate the observers at the 25 percent level, those figures
can be expressed as opacity percentages. The conversion for the 10
percent oDaclty column 1s less exact, but the fact that all observers
did observe smoke with a known 10 percent opacity plume allows the use
of the standard ratio (Rlngelmann .5 * 10 percent opacity) as a close
approximation for purposes of conversion. The results of this conversion
appear below.
Table 2. AP-30. Evaluation of White Plumes6
In-stack transmlttance, percent 75% 90%
Opacity, percent 25% 10%
Inspector No. 1
Ringelmann observed 1.03 0.03
Equivalent opacity 25.75% 0.6%
Inspector No. 2
Rlngelmann observed 1.13 0.16
Equivalent opacity 28.25% 3.2%
Inspector No. 3
Ringelmann observed 0.97 0.17
Equivalent opacity 24.25% 3.4%
It is apparent from this more properly presented data that Table
2 in fact refutes, rather than supports, PCA's claims of unreliably
116
-------
high readings at low levels opacity. Indeed, the column which PCA
treated as 0 percent opacity is in fact 10 percent opacity; 1n this
column all three Inspectors rated the opacity at well below its actual
value. Likewise, the column which PCA treated as 20 percent is actually
25 percent opacity. In that column, all observers rated the opacity at
values very close to the actual level, and well within the 7.5 percent
opacity tolerance to which observers are limited by Method 9.
Because of the fundamental differences in the method of relating
and calibrating Rlngelmann numbers with equivalent opacity between the
method used at the smoke school where the above data were gathered and
the present approach used by EPA (as reflected In Method 9), the
author of AP-30 states that the data in Table 2 cannot be used to
assess the reliability of EPA's method or opacity readings 1n general.^
The above discussion is therefore presented solely to demonstrate that,
contrary to PCA's belief, those data do not 1n any way Indicate the
unreliability of EPA's use of qualified observers. The Agency is not
relying on those data as part of its demonstration of the accuracy of
Method 9.
In accordance with the Amoco case discussed above, this error
tolerance of + 7.5 percent opacity must be accounted for in the enforce-
ment process, so that no source which 1s 1n fact 1n compliance with the
applicable opacity standard is cited for a violation due merely to
normal observational error. EPA has consistently maintained that 1t
does in fact take this error tolerance into account in the enforcement
117
-------
process. In an earlier draft response to this remand (November, 1973),
for example, the Administrator stated that "the agency would not take,
and the courts would not sustain, enforcement action based upon opacity
observations that do not exceed the standard by more than the average
opacity error associated with the particular opacity standard." PCA,
however, has been unsatisfied by this assurance, apparently fearing that
EPA in its enforcement decisions would ignore the established error
tolerance. To provide the certainty desired by PCA and others, EPA is
therefore now amending the applicable regulations (40 CFR 60, Appendix
A, Reference Method 9) by adding an introductory section which discusses
the concept of visible emission readings and describes the effect of
variable viewing conditions. This section presents pertinent portions
of the above data and notes specifically that the accuracy of the method
must be taken Into account when determining possible violations of
applicable opacity standards. A copy of the revised regulation appears
in Part B of Appendix III of this Response.
EPA 1s also now amending the provisions of Reference Method 9 to
make clear that the determination of an opacity level reportable as a
violation involves averaging 24 readings taken at 15-second intervals.
This will make clear that a single high reading will not be cited as a
violation. It will also ensure, as PCA requests, that such individual
readings, which may result under normal plant operations from brief
events such as shaking of fabric filter bags in baghouses or rapping
of electrodes in electrostatic precipitators, will not result 1n
118
-------
determinations of violation. This amendment to Reference Method 9 is
set forth in Part B of Appendix III. (This approach 1s a satisfactory
means of enforcing opacity standards in cases where the violation 1s a
continuing one and time exceptions are not part of the applicable
opacity standard. However, the opacity standards for steam electric
generators in 40 CFR 60.42 and fluid catalytic cracking unit catalyst
regenerators In 40 CFR 60.102 and numerous opacity standards in State
Implementation Plans specify various time exceptions. Many State and
local air pollution control agencies use a different approach in
enforcing opacity standards than the six-minute average period specified
1n this revision to Method 9. EPA recognizes that certain types of
opacity violations that are intermittent In" nature require a different
approach in applying the opacity standards than this revision to Method
9. It is EPA's Intent to propose an additional revision to Method 9
specifying an alternative method to enforce opacity standards. It is
our intent that this method specify a minimum number of readings that
must be taken, such as a minimum of ten readings above the standard 1n
any one hour period prior to citing a violation. EPA 1s in the process
of analyzing available data and determining the error Involved 1n
reading opacity in this manner and will propose this revision to Method
9 as soon as this analysis 1s completed. The Agency solicits comments
and recommendations on the need for this additional revision to Method
9 and would welcome any suggestions particularly from air pollution
. control agencies in how we might make Method 9 more responsive to the
needs of these agencies.)
119
-------
PCA also contends that the wind velocity during a given se<. of
observations can affect the opacity levels read by observers. EPA
agrees that this is true. However, as is the case with lack of contras-
ting background in viewing a plume, the result of high winds is a
negative error in reading the opacity of the plume; this results from
the dissipation of the plume caused by the wind. Since a negative
error in reading the opacity of the plume cannot prejudice an operator,
this consideration is not relevant to the reliability of the method for
assessing violations of the opacity standards.
PCA contends that the viewing angle of the observer with respect
to the sun and the plume can produce erroneously high readings; if the
observer reads the plume while facing the sun, PCA says, such high
reading may result. EPA agrees, as was noted above, and for this reason
Reference Method 9 has always required the observer to read the plume
with the sun at his back. Reference Method 9 1s now being amended to
Include even more specific criteria concerning observer position with
respect to the sun. The method will now require that the sun shall be
within a 140° sector to the observer's back. This amendment appears 1n
Part B of Appendix III.
The final aspect of opacity observations with which PCA and the
Court of Appeals were concerned is the procedure employed by qualified
observers to evaluate plumes containing condensed (visible) water vapor,
or "steam plumes." When stack emissions contain appreciable water vapor,
the vapor becomes visible when the gases are cooled below the dew point
120
-------
and water condenses in fine droplets. Some steam plumes are detached
while other? are visible as released from the stack. However,
since ambient air is normally much drier than stack emissions containing
visible water vapor, dilution causes the gases at some point to become
unsaturated. At this point, condensed water evaporates and is no
longer visible in the plume. PCA contends that these plumes cannot be
read properly, and the Court of Appeals also notes concern over the way
in which steam plumes are to be read. This issue was addressed in part
24
in the report titled "A Report on Contaminated Water Vapor Plumes." As
EPA has previously noted, instruction is given in the smoke school as
to how such plumes are to be read; the above report Illustrates how
qualified observers can readily distinguish the point at which all visible
water vapor has evaporated from the plume. It is this point at which
the opacity reading is made. Since the plume has been somewhat dissi-
pated by that point, the reading 1s often lower than that which would
have been found at the Hp of the stack, had the plume been free from
visible water vapor. This error is negative, however, and in no way
prejudices the plant operator. EPA agrees, however, that the prior
language of Method 9 did not sufficiently make this procedure clear.
Method 9 is therefore now being amended to remedy that deficiency;
*** The phenomenon of a "detached plume" 1s often observed when a warm,
moist gas stream 1s released into a cooler atmosphere, e.g. when hot,
wet exhaust gases from a wet process kiln are released to the atmosphere.
Here the gases are clear at the point of discharge but become opaque as
the wet gases mix with cooler air in the atmosphere.
121
-------
when the new guidelines are adhered to (as they were In the past, though
not formally articulated in the Method), any errors due to the presence
of water vapor can be eliminated. These amendments appear in Part B of
Appendix III.
In addition to proposing specific procedures, EPA is currently
investigating techniques which could be employed in the smoke school to
further enhance an observer's capability to accurately identify
contaminated water plumes. This investigation 1s being undertaken by
a contractor—Pacific Environmental Services, Inc.—and by EPA staff
as part of an overall program to develop further guidelines concerning
12
opacity training and field observation. Procedures used by State and
local regulatory agencies to observe contaminated water vapor plumes
are being reviewed, and the use of such training aids as movies and
apprenticeship training will be considered.
The basic question involved 1s not whether an observer can accurately
assign opacity to a contaminated water plume, but rather 1s whether an
observer can distinguish between condensed water vapor and visible
particulate matter. To accurately evaluate wet plumes pursuant to
revised Method 9, 1t is necessary only for an observer to distinguish
between condensed water vapor and visible particulate air pollutants.
In no case is the observer required to assign an opacity to a plume
13
which contains visible (condensed) water droplets.
As noted in the previously cited report, water plumes present an
entirely different appearance to an observer than a dust plume. A
122
-------
trained air pollution inspector is familiar with both types of plumes,
and has no difficulty distinguishing between the two. While much of
his ability in this area stems from smoke school training, his background
in inspecting industrial processes is also a factor. Inspectors
employed by control agencies are required to become familiar with a
broad spectrum of sources and to evaluate opacities from them under
all weather conditions. For instance they know which sources employ
scrubbers, which cement plants employ wet and dry process systems and
generally which sources tend to produce wet plumes. They also know that
extreme weather conditions can create visible water plumes at sources
which release only negligible particulate matter, e.g. natural gas-fired
boilers and water heaters. Thus, when a trained air pollution inspector
evaluates a wet plume source he relies on all of this background 1n
making a judgment. In following Method 9 he first identifies that
portion of the plume which contains condensed water. As explained in
the previously cited report, fluffy water plumes may trail 100 feet or
more from the stack.24 When the water plume disappears, he evaluates the
opacity of the remaining dust plume which, if of violating opacity, will
persist for a much longer distance. At this point, when water droplets
evaporate and cease to be visible, the dust plume has been diluted
considerably below the concentration at the stack exit. Only if the
dust plume is still of greater than allowed opacity at the point of
complete evaporation can an Inspector consider a citation of violation.
His normal procedure is to enter the plant and evaluate the process
123
-------
operation before in fact issuing the notice of violation. Thus it is
obvious that the procedure allows a great deal of latitude to the
operator of a wet plume process. Far from citing unjustly for violative
visible emissions, the air pollution inspector will often be forced to
ignore periods of marginal violations simply because the visible dust
plume is masked by visible water droplets.
The revisions in Reference Method 9 make clear that the above
Q
procedure is to be followed. For attached steam plumes, the observer
is to make his readings at the point where all visible water vapor has
evaporated and only particulates remain in the plume. He is to record
the reading made at that point in the plume, and 1s not (as PCA fears)
to extrapolate back to any hypothetical value at the Hp of the stack.
For detached steam plumes, the observer is to read the opacity value as
usual at the lip of the stack, providing that no visible water vapor 1s
present there, i.e. providing that there 1s a gap between the lip of the
stack and the point of initial condensation into steam.
124
-------
EVALUATION OF SMOKE SCHOOL PROCEDURES
The purpose of a smoke school 1s two-fold. It provides a
rreans for training an observer, and for certifying that an observer
is, in fact, capable of making opacity observations with reasonable
accuracy. In this sense, the school serves to calibrate (or re-calibrate)
observers, and rejects candidates who are not capable of accurately
reading opacity. The question of whether the school adequately serves
these purposes cannot primarily be answered by examining the procedures
used. Rather, it is necessary to examine the accuracy of observers
who have been trained and certified in the school, I.e., the results
of those procedures. If, as the data discussed earlier demonstrate,
qualified observers can read plumes with reasonable accuracy, the
procedures and instruction used at the smoke school must, by
definition, be accomplishing the desired results.
With regard to PCA's criticism concerning the ability of an
observer to accurately read emissions under various field conditions
which were not encountered in the training school, 1t 1s necessary
to refer again to the factors affecting apparent opacity. As discussed
above, observations made under conditions of high contrast are most
conducive to positive errors. To the extent possible, observers employ
contrasting backgrounds during qualification. It is true that this
procedure may tend to limit the accuracy of an observer under less than
ideal conditions. However, since a plume will present an appearance of
lower opacity under such conditions, the resulting bias will be negative.
As discussed above, this effect is confirmed by theory and by field
observation.
125
-------
PCA has noted the fact that smoke school trainees are given
repeated opportunities to pass the certification test. Thus, an
observer may be certified on one run even though the certification
criteria were not met on previous runs. Presumably, he could
also fail to meet the criteria on subsequent runs. Petitioners
are correct in this description—particularly when new or relatively
inexperienced persons are being tested. It does not follow, however,
that the smoke school procedures are deficient or inadequate in this
respect. First, as noted above, the adequacy of the smoke school
cannot be determined solely by reviewing the procedures; evaluation
can also be made by testing the ability of observers who have
passed the certification test. Secondly with respect to the
certification procedures, it should be recognized that the +7.5%
criterion is particularly restrictive in that it involves the absolute
error, not simply the average taken over 25 readings. Thus, negative
and positive errors are not compensating (I.e., a -5% error and
a +7 error result in an average absolute error of +6%). This 1s
unliks the field observation procedure where negative and positive
errors are compensating (a -5% error and a +7 error result 1n an
average error of +1%). During an EPA smoke school conducted on
October 9, 1974, a record was kept for all observers who attended and
were certified.9 This record includes both runs during which the
observers passed the certification criteria and during which observers
failed to meet the criteria. These data show that even on these
runs when observers failed to qualifyt over two-thirds of them met
126
-------
g
7.5% average error limit Imposed by Method 9. Since the Reference
Method 9 involves an average of 24 readings as a means of determining
plume opacity, failure to continuously meet the more stringent
certification requirements is relatively unimportant as long as
the average reading is consistently within the prescribed criterion
of less than +7.5% average opacity. For example, the requirement
that no single reading be missed by more than +15 percent is somewhat
arbitrary, and could be relaxed to +20 percent to reduce the number
of observers failing the certification test. EPA has concluded,
however, that the stringent certification criterion serves the intended
purpose of forcing a trainee to learn to assign opacity within a
reasonable range of precision and therefore should not be relaxed.
EPA has further concluded, in summary, that while any changes
recommended by the contractor Pacific Environmental Services will
be carefully considered, the training program does 1n fact effectively
train observers to read plumes under field conditions with
reasonable accuracy.
Summary
EPA believes that the above data have convincingly demonstrated
that the opacity standards are inherently reliable.**** The data show
that observers can, with a very high confidence level, read plumes
to within the certified error tolerance of 7.5%. That tolerance
is taken into account in the enforcement process, as provided
**** A New Jersey court has recently reached the same conclusion. See
New Jersey v. LLoyd A. Fry Roofing Co., Docket no. C-3682-72
(Superior Court, Sept. $6, 1974).
127
-------
9
7.5% average error limit imposed by Method 9. Since the Reference
Method 9 involves an average of 24 readings as a means of determining
plume opacity, failure to continuously meet the more stringent
certification requirements is relatively unimportant as long as
the average reading is consistently within the prescribed criterion
of less than +7.5% average opacity. For example, the requirement
that no single reading be missed by more than +15 percent 1s somewhat
arbitrary, and could be relaxed to +20 percent to reduce the number
of observers failing the certification test. EPA has concluded,
however, that the stringent certification criterion serves the intended
purpose of forcing a trainee to learn to assign opacity within a
reasonable range of precision and therefore should not be relaxed.
EPA has further concluded, in summary, that while any changes
recommended by the contractor Pacific Environmental Services will
be carefully considered, the training program does 1n fact effectively
train observers to read plumes under field conditions with
reasonable accuracy.
Summary
EPA believes that the above data have convincingly demonstrated
that the opacity standards are Inherently reliable.**** The data show
that observers can, with a very high confidence level, read plumes
to within the certified error tolerance of 7.5%. That tolerance
is taken into account in the enforcement process, as provided
**** A New Jersey court has recently reached the same conclusion. See
New Jersey v. LLoyd A. Fry Roofing'Co.. Docket no. C-3682-72
(Superior Court, sept. 30, 19/4}.
128
-------
EPA will of course continue to review the degree of accuracy
and need for opacity measurements as new data and Information are
acquired. For example, at the request of the Department of Commerce,
EPA solicited such information in the preamble to new source
performance standards recently proposed (39 FR 37466). However, at
this time the data and information available to EPA support the
position taken in this response.
ACHIEVABILITY OF THE OPACITY STANDARDS FOR PORTLAND CEMENT PLANTS
Under Section 111 of the Clean Air Act, EPA has established
10 percent opacity as the standard which may not be exceeded by
new kilns at Portland cement plants. PCA contends that some new
plants will or may be unable to achieve compliance with this
visible emissions limit. EPA policy is that opacity limits should
be set at such levels that any source which meets the mass emission
limits will also meet the opacity limits. EPA has conducted an
intensive review of its opacity standards, Including conducting new
tests and gathering new data, and has concluded that although the 10%
opacity standard is achievable by almost all new Portland cement plants,
that standard should nevertheless be relaxed to 202 opacity to accomo-
date certain extreme circumstances. EPA further concludes that at this
new level, practical circumstances under which any combination of the
factors cited by PCA or the Court as having a possible effect on opacity
(transmittance) of smoke plumes that could cause a new source to be
unable to meet the opacity standard is exceedingly rare. The
129
-------
EPA will of course continue to review the degree of accuracy
and need for opacity measurements as new data and information are
acquired. For example, at the request of the Department of Commerce,
EPA solicited such information in the preamble to new source
performance standards recently proposed (39 FR 37466). However, at
this time the data and information available to EPA support the
position taken in this response.
ACHIEVABILITY OF THE OPACITY STANDARDS FOR PORTLAND CEMENT PLANTS
Under Section 111 of the Clean Air Act, EPA has established
10 percent opacity as the standard which may not be exceeded by
new kilns at Portland cement plants. PCA contends that some new
plants will or may be unable to achieve compliance with this
visible emissions limit. EPA policy is that opacity limits should
be set at such levels that any source which meets the mass emission
limits will also meet the opacity limits. EPA has conducted an
intensive review of its opacity standards, Including conducting new
tests and gathering new data, and has concluded that although the 10%
opacity standard is achievable by almost all new Portland cement plants,
that standard should nevertheless be relaxed to 20% opacity to accomo-
date certain extreme circumstances. EPA further concludes that at this
new level, practical circumstances under which any combination of the
factors cited by PCA or the Court as having a possible effect on opacity
(transmlttance) of smoke plumes that could cause a new source to be
unable to meet the opacity standard is exceedingly rare. The
130
-------
0.03 grains per standard cubic foot which is necessary to meet
the mass emission standard.' These collectors, fabric filters or
electrostatic precipitators, are more effective in removing coarse
particles than fine particles. Thus emissions from collectors, i.e.,
visible emissions, tend to be made up of much greater fractions
of fine particulates than coarse particulates.
In typical high efficiency collector exhaust gases there are
generally few particulates larger than 40 microns diameter.16, 17 The
predominant number of particles are between 0.5 and 10 microns with
2 17
the average size being about 2-4 microns. * Maximum light scattering
is generally acknowledged to be caused by particles 1n the size range
of 0.2 to 2.0 micron.6,17 Available data indicate that the size dis-
tribution of particulates released from well controlled cement kilns
are similar within a narrow range (approximately 2 to 6 microns) from
one kiln to another, and therfore from one plant to another.17
What the above data and studies Indicate, 1n short, 1s that the
size of particles emitted by plants with such control equipment varies
only within a very narrow range. This variability in average size
is theoretically not sufficient to cause more than a +_ 5 percent
variation 1n opacity for typical cement kilns.6 These data are
wholly Incompatible with PCA's speculation that widely varying parti-
cle sizes might produce much larger variations 1n opacity at cement
plants meeting the mass emission standard.
The shape of particles emitted from well-controlled
Portland cement plants may vary as PCA contends. However, we disagree
with their conclusions that such variations 1n shape would affect the
131
-------
opacity of emissions perceived by a qualified observer or iri-stack
transmissometer. Any irregularly shaped particles would be randomly ori-
ented thereby smoothing or averaging out any significant variations of
opacity that could possibly result. In addition, it is likely that the
particles will be abraded or will condense into spherical shapes for the
following reasons: 1) feed materials are ground to less than 75 microns
in almost all cases; 2) further attrition and size reduction occurs
as the feed material passes through the high temperature rotating
kiln; and 3) alkaline material tends to vaporize and condense as
. fume.15-'6-17
Also, the recent microscopic examination of particles from
the exhaust gases of a precipitator controlled kiln showed that
almost all particles were approximately spherical 1n shape.2
Given that particle size and shape do not vary to the extent
necessary to substantiate PCA's speculation, 1t 1s apparent that
normal operations even from plant to plant will not produce vari-
ations sufficient to significantly affect the opacity observed
at the point of emission from the stack. Therefore, 1f the opacity
standards are set 1n accordance with the opacity actually produced
by normal operations, then with proper maintenance and operation
of control equipment the very slight variations which may occur
from plant to plant or from time to time within an individual plant
will not cause the source to exceed the opacity standard. And 1f,
for any reasons, some unusual design feature of a given plant
caused it to exceed the opacity standard but not the mass emission
132
-------
standard, the variance procedure described below would elimi-
nate any prejudice to such source.
The second major factor which PCA claims affects opacity
(transmittance) to the extent that some plants may exceed the
opacity standard while meeting the mass emission standard 1s
the diameter of the duct Involved; I.e., the path length through
which the light must pass. EPA agrees that the greater the path
length through which the light must pass, the more light will be
attenuated, and the greater will be the opacity. EPA again dis-
agrees, however, with PCA's conclusion, which 1s again 1n disregard
of the actual circumstances of plant design and operation, that
some plants will therefore be unable to meet the opacity standards,
though meeting the mass standard. EPA believes that the proper
issue in this regard 1s the actual range of stack diameters
resulting from the engineering and economic necessities of plant
design. This question 1s addressed below.
Stack diameter affects opacity directly because 1t governs the
length of the path through which the light 1s attenuated before 1t
reaches the observer viewing the plume. The larger the diameter
of the stack, and therefore the plume, the more light will be
attenuated as 1t passes through the plume, and the greater will
be the opacity of the plume. Stacks venting high efficiency control
equipment, such as that serving kilns at portland cement plants, are
designed to ventilate the process to the degree necessary with the
minimum pressure drop and at the lowest cost. These engineering
and economic factors result in the construction of stacks in which
133
-------
the gas velocities approximate 30 to 100 feet per second, since at
higher velocities friction losses become excessive, necessitating
much larger horsepower fans. At lower velocities, the size of the
stack required would result in disproportionately large capital
costs. Thus, stack diameter tends to be directly related to and
characteristic of plant size, varying only within a relatively
limited range.
Stack gas exit velocity is controlled by stack diameter.
Exit velocity 1s directly proportional to the volume of gases
vented from a kiln and Inversely proportional to the cross-sectional
area of the stack. For plumes of a given cross-section, velocity
has no effect upon opacity, since opacity is a function of path
length and is proportional to the concentration and characteris-
tics of particles 1n the gas stream. The velocity at which the
gas stream moves through the viewer's line of vision has no
effect on opacity.
It follows, therefore, that an extremely large diameter
stack could theoretically be Installed at a cement plant, even
though Its cost would be much greater than normal. In the course
of EPA's survey of the cement Industry, the largest size kiln
stack that has come to our attention 1s 15 foot in diameter. In
the unlikely event that larger stacks were used or other abnormal
conditions of design or operation were encountered that could
cause a violation of an applicable opacity standard, EPA would evaluate
opacity during the performance tests for the mass emission standard.
If 1t were shown necessary, a suitable opacity level greater than
134
-------
20 percent would be established for that specific kiln installation.
Although such an eventuality is highly unlikely, appropriate lan-
guage is being promulgated 1n the general provisions of the regulation
to accomodate such situations and appears in Part B of Appendix III.
EPA concludes, on the basis of the actual stack diameters
employed by the industry, that stack diameter will not cause a
plant meeting the mass emission standard to exceed the opacity
standard. To deal with the uncertainties cited by PCA as to
how plumes from baghouses with stub stacks are to be read, EPA is
revising Method 9 to specify that a row of such stub stacks, or
any non-circular stack, is to be read along its shortest axis
(perpendicular to its longer axis). This ensures that the
shorter path length will be used to determine opacity, and will
eliminate any concern that a long row of stub stacks, common on
baghouses, might be read along its long axis, producing higher
opacity.
Given that particle size-shape and stack diameter do not pro-
duce opacity variations sufficient to cause a source which might
otherwise meet the opacity standard, and which 1s meeting the mass
standard, to exceed the opacity standard, the only remaining issue
is whether well-controlled portland cement plants can, in fact,
meet the opacity standard. EPA has accumulated data for portland
cement plant kilns that are vented to high efficiency control
equipment that 1s properly designed, operated and maintained. These
data are summarized in Table I. Opacity readings were also taken of
135
-------
Table r
Particulate Emissions2,19'23
Name of the Company,
Date of Test and
Type Control Equipment
Dragon Cement Co.
Northampton, Pa.
4/29-30/71
Baghouse
Oregon Portland Cement Co.
Lake Oswego, Oregon
10/7-8/71
Baghouse
Santee Portland Cement Co.
Holly Hill, S. C.
5/23-25/74
Electrical Precipitator
United Portland Cement Co.
Artesla, Miss.
6/25-27/74
Electrical Precipitator
Monarch Cement Co.
Humboldt, Kansas
6/12-14/74
Baghouse
Marquette Cement Co.
Neville Island, Pa.
6/19-21/74
Electrical Precipitator
Test Duration
6 hours
6 hours
8 hours -
2 minutes
8 hours -
23 minutes
9 hours -
0 minutes
lbs/ton
of dry
feed
0.070
0.272
0.107
0.209
12.5 minutes 0.196
0.339
Stack D1a.
1110"
4'4"
(velocity
cone)
12'0"
11" 3"
7*0'
12'6"
Max.
Average Single
Opacity Reading
<10
-------
several additional well controlled portland cement plant kiln stacks
18
by an EPA contractor. Mass emission tests were not conducted 1n
these Instances, however, in light of the petitioners contention
that the mass standards are overly restrictive, it would seem
unlikely that these plants would be ODeratlng significantly below
the mass emission standard. A summary of these Installations 1s
shown in Table II.
Table II
Opacity of Visible Emissions from Cement Kilns Controlled by Fabric Filters
18
Name of Company
Date of Test and
Control Equipment
Whitehall Cem
Cementon, Pa.
7/29 & 8/1/74
Baghouse
Coplay (Egypt)
Cement Co.
Egypt, Pa.
7/29-31/74
Baghouse
Hercules Cement Co.
Stockerton, Pa.
7/29 & 8/2/74
Duration
of Test
(hours, min.)
8h 2m
7h 15m
5h 16m
Penn-D1xie Cement Co.* 7h 55 m
Nazareth, Pa.
7/29-30/74
Baghouse
Stack Size
(feet)
15' dia.
(1 stack)
4'10" d1a. each.
(7 stacks)
13' dia.
(1 stack)
2'xl' each.
(12 stacks)
5'xlV each with
center divider.
(8 stacks)
Average
Opacity
(percent)
Maximum
6-M1nute
Average
Opacity
0.2
3.8
Maximum
Single
Reading
5
0
0
10-15
0
0
15
0
0
25
•These observations were taken across the shorter axis of ±he eight stacks (40' path-
length). If the observations had been taken across the longer axis of any one stack,
pathlength would have been 11 feet. The estimated maximum 6-m1nute average opacity
under these latter conditions would be 4.4 percent rather than 15 percent. Like-
wise, the maximum single reading would be 7.5%. Reference Method 9 now requires
these observations to be made across the 11' axis.
137
-------
With the exception of Marquette Cement Co. all of the kilns
Identified in Table I were1 operating below the mass and opacity
standards promulgated in 1971 (i.e., 0.3 lb/ton of dry feed and
10%). The data for Marquette Cement Co. Indicate that the kiln
was not in compliance with the mass emission standard (0.339
versus 0.3 lb/ton of dry feed). The opacity data were collected
with an 1n-stack transmissometer. The stack diameter is 12'6", the
average opacity during the mass emission tests was 18%, and the
average size of emitted particles was determined to be approximately
2 microns diameter- If the stack diameter were 15', which 1s the
largest diameter stack EPA has observed on cement plants, the average
opacity would have been 21%. The mass emission tests for Marquette
Cement Co. consisted of three separate runs. The results of the three
runs are:
Table III
2 3
Opacity and Mass Emissions for a K1ln '
Particulate Average opacity Average opacity
Emissions, (percent) at (percent) extrap-
Run lbs/tons of actual stack con- olated to 15' diameter
Number dry kiln feed dltlons (1216") stack conditions
1 0.386 22 25
2 0.339 17 20
3 0.292 15 18
The particulate matter emissions from Marquette Cement Co. were
analyzed to determine the particle size distribution. This analysis
Indicated that the average particle diameter was approximately 2 microns
138
-------
which is very low for portland cement plants. These data indicate
that for a plant with an unusually large stack diameter and
with particulate emissions having very small average particle diameter
properties, a 20% opacity standard is more reasonable than a 10% standard.
The achievabiHty of a 20 percent standard 1s amply demonstrated by
the data in Tables I and II which Illustrate that most portland cement
kilris have no difficulty in meeting even a 10 percent opacity standard.
The Marquette Cement Company test results Illustrate that even under
the combination of conditions that would produce the maximum obscuration
of light (i.e., large stack diameters, small average particle sizes,
and concentrations near the mass emission standard), a 20 percent
standard (Table III, run number 3) would be achieved. It should be
additionally noted that the combination of conditions for maximum
opacity, even though accommodated by the 20 percent standard, will
occur Infrequently. Therefore, 1t 1s concluded that a plant with a 15'
diameter stack which meets the mass emission standard will also meet a
20% opacity standard and when it exceeds the mass emission standard
will exceed the opacity standard. For stacks smaller than 15 feet 1n
diameter, a portland cement kiln that violates a 20% opacity
standard will also violate the mass emission standard by a sub-
stantial amount. Data 1n Table II show that 1n several cases,
opacity of emissions from well controlled cement kilns are
appreciably less than 20% even where larger stacks are employed
e.g., 0.2% from a 15 foot dlamater stack and zero from a 13 foot
diameter stack. Based on these data and EPA's policy that opacity
standards be established at levels which require proper operation
139
-------
and maintenance of control systems, the opacity standard for
Portland cement plant kilns is being changed from 10 to 20 percent.
This change appears in Part B of Appendix III.
In addition to considerations of the average opacity standards,
EPA has evaluated the need for possible time exemptions in the
opacity standard. It is recognized that during normal operation of
Portland cement kilns, process variations, shaking of bags in bag-
houses, rapping of electrodes in electrical precipitators, etc.,
can produce mass emissions that are not perfectly constant. It
follows that slight variations in opacity can also occur under
these conditions. However, with the clarification to Method 9
wliereby a minimum of 24 opacity readings must be taken at 15
second intervals over a 6 minute period and with the averaging
of observations for determining compliance with the opacity
standard, it is apparent that such instantaneous opacity readings
will not constitute a violation of the standard. Changing the opacity
standard for portland cement kilns from 10 to 20 percent further
ensures that such temporary incidents will not cause the opacity
standard to be exceeded. Since the change 1n the opacity standard
merely represents a relaxation of the standard, it will be appli-
cable to all sources which commenced construction subsequent to
the promulgation of the original, more stringent level. To an even
greater degree than the 10% standard, this 20% standard is
sign1f1ca. t1y less stringent than the mass emission TfmU for
Portland cement kilns.
140
-------
As discussed 1n this remand response, the change of the
opacity standard from 10 percent to 20 percent was made to Insure
that no plant that is meeting the mass emission standard is cited
for violating the opacity standard. It does not change the
degree of particulate control required of new port!and cement
plants. Accordingly, this relaxation of the opacity standard
does not require an environmental impact statement.
141
-------
APPENDIX III
References
1. K.W. Buhne and L. Duwel. Recording Dust Emission Measurements 1n
the Cement Industry with the RM4 Instrument. Staub. Vol.32, No. 19.
1972 (in English).
2. Preliminary Test Results of Marquette Cement Manufacturing
Company for Task Number 11. Cottrell Environmental Sciences.
EPA Contract No. 68-02-0239. August 26, 1974.
3. Robert L. Ajax. Memo to Jack R. Farmer, Credibility of Emission Test
Data Contained in the Marquette Cement Manufacturing Company's Report.
October 24, 1974.
4. Average Observational Error Associated with Smoke Plumes at Levels of
Known Opacity. EPA, OAWP, OAQPS, ESED Report. October 1973.
5. W. D. Conner. Memo to R. L. Ajax, Measurement of Opacity by
Transmissometer and Smoke Readers. October 16, 1974.
6. Optical Properties and Visual Effects of Smoke Stack Plumes. PHS
Publication No. 999-AP-30.
7. Henry F. Hammlll, Richard E. Thomas, and NolUe F. Swynnerton.
Evaluation and Collaborative Study of Method for Visual
Determination of Opacity of Emissions from Stationary Sources:
Interim Report. October 1974.
8. Federal Register: 39 FR 32857, September 11, 1974.
9. Evaluation of EPA Smoke School Results. EPA, OAWM, OAQPS, ESED Report.
October 9, 1974.
10. Portland Cement Association v. Ruckelshaus. 486 F.2d 375, No. 72-1073.
June 29, 1973.
11. Coons, J. D., et al. Development, Calibration, and Use of a Plume
Evaluation Training Unit. Journal of A1r Pollution Control Association.
Vol.15. May 1965. pp. 199-205.
12. Task Order: Preparation of Guideline for Evaluation of Visible
Emission Opacity by Trained Observers. EPA Contract No. 68-02-1390,
Task 2, June 5, 1974.
13. Federal Register: 39 FR 32854, September 11, 1974.
14. Federal Register: 39 FR 32852, September 11, 1974.
15. R. Emmet Doherty. Current Status and Future Prospects - Cement Mill
A1r Pollution Control. Proceedings, The Third National Conference
on Air Pollution, Washington, D. C. December 12-14, 1966.
142
-------
16. A Manual of Electrostatic Precipitator Technology, Part II —
Application Areas. Southern Research Institute. Contract Mo.
22-69-73. 1970. pp. 609-640.
17. Particulate Pollutant System Study, Volume II~F1ne Particulate
Emissions. Midwest Research Institute. Contract No. CPA 22-69-104.
1971.
18. Visible Emission Testing at Four Portland Cement Plants. Scott
Environmental Technology, Inc. Contract No. 68-02-1400. 1974.
19. Emission Testing Report (tested by company personnel). United
Cement Company, Artesla, Mississippi. 1974.
20. Emission Testing Report (kiln No. 2). Santee Portland Cement
Corporation, Holly H111, South Carolina. Pollution Control-
Walther Inc. 1974.
21. Emission Testing Report. Monarch Cement Company, Humboldt, Kansas.
Fuller Company. 1974.
22. Emission Testing Report. Dragon Cement Company, Northampton,
Pennsylvania. EPA, OAWM, OAQPS, ESED Report. March 1972.
23. Emission Testing Report. Oregon Portland Cement Company,
Lake Oswego, Oregon. EPA, OAWM, OAQPS, ESED Report. March 1972.
24. Contaminated Water Plumes. EPA, OAWM, OAQPS, ESED Report. 1973.
143
-------
D-2 Ca)
Excerpt from How EPA Validates NSPS Methodology, R. Midgett, ES&T,
July 1977
Opacity
Collaborative testing of EPA Method 9 for visual determination
of opacity of emissions from stationary sources was conducted
by certified observers at three collaborative test sites: a training
smoke generator, a sulfuric acid plant, and a fossil fuel-fired
steam generator. The initial test on the training smoke generator
was conducted to provide background information on the use
of the method, while the test at the sulfuric acid plant and the
fossil fuel-fired steam generator was conducted to obtain in-
formation on the use of the method on applicable sources under
field conditions.
Composite precision estimates based upon the results of all
the tests were derived, and these are shown In Table 3. With data
from the training generator and from Test 3 at the steam station,
a composite estimate of the accuracy of Method 9 was derived
for ideal (clear sky) conditions. This estimate compares the
expected deviation of the observer from the average metered
opacity, and is given by the equation, deviation = 3.13 — 0.31
(meter average), for the range from 5-35%, average opacity.
With respect to the other experimental factors and variables
studied, it was concluded from the clear-sky data of Test 3
that:
• The angle of observation does affect the observer's de-
terminations. and, in this study, the most accurate readings were
made when the group was at an approximately 45-degree angle
to the sun.
• The experienced observers were able to read average
opacity more accurately than the inexperienced observers, but
the difference occurred mainly in the high opacity range
(>25%).
• Attempts at reading opacity in increments of 1 % produced
greater within-observer variability and was less accurate than
reading in 5 % increments.
• Averaging the results of two observers yielded increased
accuracy over the result of a single observer.
Based partly on the results of these studies, Method 9 was re-
vised and improved and has now been repromulgated to replace
the original method of 1971.
TABLE 3
Precision estimates (standard deviation
independent of mean value)
Standard deviations,
Metho- Parameter, par^Tneter unitsa
Od =r
units
a
Oh
OT.
3
co2, *
0.20
0.40
0.35
3
o2, *
0.32
0.61
0.52
3
Dry mol wt,
g/g-mol
0.035
0.048
0.033
5
Moisture
fraction
0.009
0.912
0.008
8
S02, mg/m^
123
115
99
9
Opacity, %
2.05
2.42
1.29
10
CO, rag/n^
14.3
32.3
29.0
a ,
o = within-laboratory deviation;
a
b = within-laboratory deviation;
a
L » laboratory bias.
145
-------
D-2 lb)
Excerpt from EPA 600/4-76-044, TFte EPA Program for Standardization of
Stationary Source Emission Test Methodology - A Review, R. Midgett,
August 1976
OPACITY OF STACK EMISSIONS
Collaborative testing of EPA Method 9 (Ref. 1, p. 24895) for visual
determination of opacity of emissions from stationary sources was conducted
using certified observers, to obtain data that would allow statistical
evaluation of the method.^ Three collaborative test sites were used: a
training smoke generator, a sulfuric acid plant, and a fossil fuel-fired
steam generator. The initial test on the training smoke generator was
conducted to provide background information on the use of the method,
while the test at the sulfuric acid plant and the fossil fuel-fired steam
generator were conducted to obtain information on the use of the method
on applicable sources under field conditions. At no time during any of
the tests were warm-up or practice runs allowed prior to the test itself.
These tests required the determination of average opacity, defined as the
average of 25 readings taken at 15 second intervals. For the purpose of this
study, one set of 25 readings was designated a "run". The collaborators began
taking readings on a signal from the test supervisor, and thereafter at 15
146
-------
second intervals until the required 25 observations were obtained. Concurrent
with the observers' readings, plume opacity readings were taken from the in-
stack transmissometer. The accuracy of the method was judged by the devi-
ations of the observers' readings from the actual opacity as measured by this
calibrated in-stack transmissometer.
Five separate tests of EPA Method 9 were conducted, if both the
white smoke and the black smoke phases cf the training generator study are
considered as comprising one test. For each test, Table * presents the
pertinent information on the number of runs completed, the number of
observers participating, and the range of opacity studied. The studies
were deliberately restricted to the lower opacity ranges within which the
CPA standards lie.
While the smoke generator and the sulfuric acid plant tests were
designed to evaluate the accuracy and precision of Method 9 as written,
the steam station studies, in addition to this, were designed to in-
vestigate the effects of various factors on the performance of the
method. The experimental factors studied included the angle of obser-
vation and the relative experience of the observer. Variations to the
method to be evaluated included reading 1n 1 percent rather than in 5
percents, and using the average responses of two observers as opposed to
a single observer's result to determine whether these yielded increased
accuracy. The observers at each test were divided into two groups for the
test, a control and an experimental. The control group observed the plume
at all times from a position consistent with the method as written and read
in increments of 5 percent. The experimental group either read the plume
from a more extreme angle 1n increments of 5 percent or from the same angle
as the control but in increments of 1 percent. Each group was composed
both of observers who had considerable field experience with the method and
147
-------
of observers who had relatively little such experience.
Due tc the adverse sky and wind conditions during Tests 1 and 2 at
the steam station, not all of the planned evaluations were useful. There
was an inability to read the low opacity plumes against the type of back-
ground that existed, and as a result, the determinations were generally
well below the concurrent meter average. The precision estimated, however,
is independent of the accuracy of the determination. Separate precision
estimates were therefore developed for these tests, and for the tests at
the other sites. Composite estimates based upon the results of all the
tests were also derived, and because the individual estimates were similar
from one test to another, only the composite estimates shown in Table 3
will be presented in this report. Using data from the training generator
and from Test 3 at the steam station, a composite estimate of the accuracy
of Method 9 was derived for ideal (clear-sky) conditions. This estimate
compares the expected deviation of the observer from the average metered
opacity and is given by the equation, deviation 3 3.13 - 0.31 (meter average),
for the range from 5 to 35 percent average opacity. As the equation Indicates,
observers tend to read slightly high at the very low opacities, exhibit good
accuracy at around 10 to 15 percent average opacity, and acquire a definite
negative bias at the higher opacities.
With respect to the other experimental factors and variables studied, it
was concluded from the clear-sky data of Test 3 that (1) the angle of observation
does affect the observer's determinations, and 1n this study, the most accurate
readings were made when the group was at an approximately 45 degree angle to
the sun; (2) the experienced observers were able to read average opacity more
accurately than the inexperienced observers, but the difference occurred mainly
in the higher opacity range (-25 percent); (3) the 1 percent Increment data
exhibited greater withln-observer variability and was less accurate than the
J.-+8
-------
5 percent increment data; and (4) averaging the results of two observers
yielded increased accuracy over the result of a single observer. Based
partly on the results of these studies* Method 9 was revised and improved
and has now been repromulgated replace the original method of 1971.26
TABLE 2. PRECISION ESTIMATES FOR THOSE PARAMETERS WHERE STANDARD DEVIATION
WAS PROPORTIONAL TO Th£ MEAN VALUE. 6
Standard deviations, ,
percent of mean value (<$)—
Method No.
Parameter, units
a
ab
°L
2
Velocity, ft/sec
3.9
O
to
3.2
2
3
Volumetric flow rate, ft /hr
5.5
5.6
1.1
5
3
Particulate matter, mg/m
10.4
12.1
6.1
6
3
S02, mg/m
4.0
5.8
4.2
7—
N0X, ng/m3
fi.6
9.5
6.9
7—
N0X, mg/m3
14.9
18.5
10.5
8
HgSO^ mist (including SO^),
mg/m3
58.5
66.1
30.8
104
Be, g/day
43.5
57.7
37.9
— a * within-laboratory deviation; ah » between-laboratory deviation,
= laboratory bias.
-^Pooled power plant/pilot combustion plant data.
-^Nitric acid plant data.
149
-------
D-3
Charts Stowing tn-Stack Opacity Measurements and Individual
Opacity Readings of Participants in Collaborative Testing of
EPA Method 9. (Data taken from tables in EPA Report No. 650/4-
75-0Q9, Evaluation and Collaborative Study of Method for Visual
Determination of Opacity of Emissions from Stationary Sources,
H. Hamil, January 1975)
151
-------
AFTER HAMIL AND THOMAS
EPA-650/4-75-009
JANUARY 1975
o MULTI-POINT
>
TRANSMISSOMETER OPACITY, psrcent
BLACK SMOKE GENERATOR
VIEWING CONDITIONS: IDEAL
TOTAL NO. OF POINTS: 132
NO. OF DETERMINATIONS PER OBSERVER: APPROXIMATELY 16
152
-------
AFTER HAMIL AND THOMAS
EPA-650/4-75-009
JANUARY 1975
o MULTI-POINT
v
•• •
> 15
M
«
TRANSMISSOMETER OPACITY, percent
WHITE SMOKE GENERATOR
VIEWING CONDITIONS: IDEAL, BLUE SKY, BRIGHT SUNSHINE
TOTAL NO. OF POINTS: 170
NO. OF DETERMINATIONS PER OBSERVER: APPROXIMATELY 20
153
-------
AFTER HAMIL AND THOMAS
EPA-650/4-75-009
JANUARY 1975
TRANSMISSOMETER OPACITY, percent
RIVERBEND STEAM STATION TEST NO. 1
VIEWING CONDITIONS: CLOUDY, LOW HAZE
TOTAL NO. OF POINTS: 45
NUMBER OF DETERMINATIONS PER OBSERVER: 19
154
-------
AFTER HAMIL AND THOMAS
EPA-650/4-75-009
JANUARY 1975
O MULTI-POINT
>
h-
O
cc
IX)
V)
eo
Q
TRANSMISSOMETER OPACITY, percent
RIVERBEND STEAM STATION TEST NO. 2
VIEWING CONDITIONS: MARGINAL, SOLID OVERCAST
TOTAL NO. OF POINTS: 85
NO. OF DETERMINATIONS PER OBSERVER: 17
155
-------
50
AFTER HAMIL AND THOMAS
EPA-650/4-75-009
JANUARY 1975
o MULTI-POINT
40
e
»
u
u
<
a.
O
a
LU
>
cc
LU
M
J
20 30 40
TRANSMISSOMETER OPACITY, percent
RIVERBEN0 STEAM STATION TEST NO. 3
VIEWING CONDITIONS: IDEAL, CLOUDLESS, BRIGHT SUNSHINE
TOTAL NO. OF POINTS: 96
NO. OF DETERMINATIONS PER OBSERVER: 24
SO
156
-------
NUMBERS=>1 DETERMINATIONS AT POINT
N"
•2 *2
•5 •
/.! |
•4
TRANSMISSOMETER OPACITY, percent
STAUFFER CHEMICAL (ACID MIST PLUME)
VIEWING CONDITIONS: GOOD. BLUE SKY, LIGHT HAZE
TOTAL NO. OF POINTS: 297
NO. OF DETERMINATIONS PER OBSERVER: APPROXIMATELY 30
157
-------
D-4
EXECUTIVE SUMMARY *
EXAMINING THE PROPERTIES OF QUALIFIED OBSERVER OPACITY
READINGS AVERAGED OVER INTERNALS OF LESS THAN SIX MINUTES
~From EPA Contract Report 68-02-1325 (RTI 893-66), Examining the Properties
of Qualified Observer Opacity Readings Averaged Over Intervals of Less
than Six Minutes, T. Hartwell, Research Triangle Institute, Research Triangle
Park, North Carolina 27709, July 1976
159
-------
EXECUTIVE SUMMARY
EXAMINING THE PROPERTIES OF QUALIFIED OBSERVER OPACITY
READINGS AVERAGED OVER INTERVALS OF LESS THAN SIX MINUTES
In this study visual emission observations (VEO's), collected by
procedures set forth in EPA Reference Method 9, were compared with con-
current opacity measurements made by an instack transmissometer to assess
the impact of using other than the prescribed 6-minute averaging time for
determining the opacity of smoke plumes. A 6-minute average is defined as
the mean of 24 consecutive opacity observations made at 15-second intervals.
The effect of determining opacity as the average of non-consecutive 15-second
opacity readings also was examined. Data from both white and black smoke
plumes were used.
The data used in the analyses were obtained from two field tests in
which qualified observers assigned opacities at 15-second intervals to
white and black plumes produced by smoke generators and to a power plant
plume. Concurrently instack opacities were measured by a transmissometer.
The test using the smoke generator conducted at the Research Triangle Park
in 1973, nine observers were used; the other test was conducted in 1974 at
the River Bend Steam Station near Charlotte, North Carolina, 4 observers
participated. At the Riverbend test, the plume was essentially white in
color. At both of these tests, usable data were collected on 266 runs
with white smoke and 133 with black smoke.
To determine the effect of different averaging times, several statis-
tics were computed using the VEO's and instack opacity readings. Separate
analyses were made for both white and black smoke plumes using averaging
times of 30 seconds, 1 minute, 2 minutes, 3 minutes, 4 minutes, 5 minutes
160
-------
and 6 minutes. An examination of these statistics indicates that:
• In the cases examined, average opacity can be estimated using 3-
or 4-minute averages of consecutive 15-second VEO's about as well as
using a 6-minute average.
• A 30-second average of two consecutive VEO's is inferior to using
a 3- or 4-minute averaging time.
• The averaging of nonconsecutive VEO's appears to be a relatively
poor method for calculating the average opacity of smoke plumes.
• Observer bias becomes negative as opacity increases.
• Observer measurement error varies directly with opacity level.
In arriving at these determinations, the following analyses were
made:
• the frequency by which observed opacity varied from transmisso-
meter measured opacity by units of 5, 10, and 20 percent for each
of the averaging times studied.
• the estimated observer bias
• the estimated observer measurement errors
The analyses indicated that:
1. For averages computed from consecutive readings as averaging time
increases:
a. The frequency by which observed opacities varied from true
opacity (as measured by a transmissometer) decreases with
161
-------
little change between the 3-, 4-, or 5-minute and 6-minute
averaging time.
b. Little effect is noted on observer bias.
c. Estimated observer measurement error decreases
2. For averages calculated from non-consecutive observations,
a. The number of times in which observed opacities varied
from true opacity increased, especially for the averaging
of the highest 2 and 4-minute non-consecutive readings.
In most of the cases analyzed, the opacities were greater
than true opacities.
b. Observer biases are less negative than for consecutive
readings.
3. In general as the level of true opacity increases -
• number of times that observed opacities vary from
true opacity decreases
• observer bias becomes more negative
• observer measurement error increases
The study has two limitations that must be considered. Many of the
statistics are based on statistically insufficient data; especially for
the 4-, 5-, and 6-minute averaging times and the non-consecutive averaging
times. Accordingly, if accurate estimates of the statistics are desired,
more data should be examined. Also the data did not permit direct investi-
gation of State opacity regulations such as "opacity cannot exceed 'X'
percent for more than 'Y1 minutes per hour."
162
-------
SECTION E: MEASUREMENT OF OPACITY AND PARTICULATE EMISSIONS USING
INSTRUMENTAL METHODS
Page No.
E-1. "The Correlation of Plume Opacity to Particulate Mass
Concentration," D. S. Ensor and M. J. Pilat, reprinted
from AIChE Symposium Series: Air Pollution Control and
Clean Energy (Vol. 72, No. 156), 1976 165
E-2. "Instrumental Method Substitutes for Visual Estimation of
Equivalent Opacity," Herbert C. McKee, APCA Journal,
August 1971 175
E-3. "Texas Regulation Requires Control of Opacity Using Instru-
mental Measurements," Herbert C. McKee, APCA Journal,
June 1974 179
E-4. "Measurement of Opacity and Particulate Emissions with an
On-Stack Transmissometer," Heinz P. Beutner, APCA Journal,
September 1974 183
E-5. Chart Showing Opacity-Mass Emissions Correlations for
Various Source Categories, W. Conner, 1977 190
-------
Reprinted From AIChE Symposium Series; E-l
Air Pollution Control and Clean Energy
(Vol. 72, No. 156) 1976
THE CORRELATION OF PLUME OPACITY TO
PARTICULATE MASS CONCENTRATION
David S. Ensor
and
Michael J. Pilat
Plume opacity measured with an in-slack transmissometar is related to other particu-
late properties, such as mass concentration. The in-stack transmissometar and cascade
impactor data from a Kraft paper pulp recovery furnace and aluminum refinery are
presented and discussed. The data are analyzed statistically and explained with light
scattering theory.
OBJECTIVE
The visual appearance of smoke plumes
has been used for at least 60 yr. to regulate
air pollution sources. The inability to pre-
dict plume opacity or to relate to other prop-
erties of the aerosol has been an important
limitation in design of control equipment and
determination of the impact of the source on
the environment. The work described in this
paper began in 1968 at the University of
Washington from requests from local indus-
try for a detailed analysis of the opacity
problem. Over the years, insight has been
gained into the problem, and the efforts have
resulted in many journal publications, theses,
reports for clients, and unpublished data. The
objective of this paper is to review the prog-
ress we have made in the study of opacity.
APPROACH
The visibility of smoke plumes is a
very complex phenomenon. It was recog-
nized early in our work that in order to make
David S. Ensor is with Meteorology Re-
search, Inc., Altadena, California. Michael J.
Pilat is at the University of Washington,
Seattle, Washington.
progress towards practical methods of pre-
dicting the visual impact of the plume, the
problem had to be reduced to segments which
could be solved in a reasonable amount of
time. They are interior opacity (in-stack),
exterior opacity (outside of stack), and down-
wind visibility.
The interior opacity approach involves
using the size distribution, particle refrac-
tive index, and particle density as parame-
ters in relating the mass concentration to
light transmittance. The light transmittance
is an intrinsic property of the plume (3^) and
can be measured with commercially available
in-stack instrumentation. The fractional
light transmittance is related to opacity by
Opacity = (1 - fractional transmittance)
X 100 (1)
The exterior opacity is the description
of plume appearance at the exhaust of the
stack. While the equations describing the
ambient light scattered from a plume are
fairly straightforward, the interpretation of
opacity from a legal standpoint is rather con-
fused. In the case when the definition of
opacity by Equation (1) is accepted and there
is no aerosol formation outside of the stack,
the interior opacity and experior opacity will
165
-------
72
No. 156, Vol.
be identical.
The downwind visibility problem in-
volves combining dispersion equations with
the aerosol relationships developed to de-
scribe opacity as reported by Ensor et al.(8).
The interior opacity studies are basic
to the exterior opacity and downwind visibil-
ity and must be understood first. The follow-
ing approach was used: theoretical analysis
and simplification, suitable data to indicate
regions of error, application to predict
plume opacity.
THEORETICAL ANALYSIS
Basic Equations
The most common optical parameter
measured in a smoke stack is the light trans-
mittance. It can also be measured in the ex-
terior plume with telephotometer s (4). The
fractional light transmittance through a vol-
ume of aerosol is given by Beer's law:
I/I ~ 8Xp - b L (2)
o
The starting point in the analysis was to pro-
pose the following formula:
W
yt = K p b (3)
The parameter K has the units of length and
can be thought of as an effective particle ex-
tinction size. A similar earlier approach
was reported by Charlson et al. (2) to analyze
atmospheric visibility data.
The parameter K is given by
ENSOR and PILAT
K = 4/3
/r3f(r) dr (4)
/Qext
-------
POLLUTION CONTROL AND CLEAN ENERGY
AIChE SYMPOSIUM SERIES
refractive index and relatively insensitive to
particle size. An important result is that
stack aerosols are usually polydispersed
enough (
-------
No. Vol. 7 2
A. b
b
AX/I
I/I nl/I
A L
L
This error in the extinction coefficient will
become large for very dilute emissions when
I/I approaches 1.0.
Mass Concentration. The important
consideration is to measure the concentration
in the same conditions as measured by the
transmissometer. An in-stack measurement
is desirable. However, a method 5 determin-
ation is suitable if there is no particle form-
ation or growth in the external probe and fil-
ter.
ENSOR and P1LAT
2. The average measured K's were
19% greater in magnitude thaj\ the K's cal-
culated from the measured size distribution
by the use of the computer program, on the
assumption of a particle density of 2.5 g/ctn
and a refractive index of 1. 50.
These results substantiate the theorem
ically predicted averaging effects of polydis-
CORRELATION OF IN-STACK OPACITY TO
MASS CONCENTRATION
Kraft Recovery Boiler
One of the earliest extensive studies of
in-stack opacity and mass concentration was
for a Kraft recovery boiler reported by Lars-
sen et al. (II). In this research project, suf-
ficient size distribution, mass concentration,
and light transmittance data were obtained to
allow extensive statistical and theoretical an-
anlysis.
The mass concentration was measured
with an alundum thimble from source test
methods described by Haaland (9). The par-
ticle size distributions were measured with a
University of Washington cascade impactor.
A Bailey smoke meter (bolometer) was used
to measure the in-stack light transmittance
over a 3 ft, illumination distance. Sections
of the electrostatic precipitator were shut off
to vary the concentration of particulate mat-
ter over a wide range.
The two most important findings are:
1. Although large variations were
measured in Hie particle mass concentration,
the magnitude of the parameter K was rea-
sonably constant. The measured light trans-
mittance and mass concentration were found
to fit a modified form of Beer's law W = A +
B log (I/I ), as shown in Figure 5, with a
correlation coefficient of 0. 991.
lcax-iy b"- ~ .
persity as indicated in Figure 1, Quite for-
tuitously, the Kraft recovery furnace K's
were in the valley of the curves describing
K as a function of the geometric mass mean
radius. Additionally, the close agreement
of the K's determined from 1he light trans-
mittance and particulate mass concentration
suggests that the cascade impactor size dis-
tribution results are representative of the
real particle size distribution. It also indi-
cates that the assumptions of spherical par-
ticles and the lognormal particle size distri-
bution used in the theoretical analysis are
valid for Kraft recovery furnace emissions.
Aluminum Refinery Emissions
The emissions of a horizontal spike
Soderberg aluminum refinery were studied
in a combined scrubber evaluation and in-
stack opacity project. The scrubber evalu-
ation and the source test data were reported
by Hofer
Simultaneous measurements of the
light transmittance, particulate mass con-
centration, and particle size distributions
were conducted at 1he exhaust of the ventila-
tion system for the hoods of fourteen hori-
zontal spike Soderberg aluminum reduction
cells. Source tests were conducted at the
inlet and outlet of a cyclone scrubber. The
in-stack light transmittance of the aerosol
was measured with a Bailey smoke meter
over a distance of 6 ft. Both the particle
size distribution and mass concentration
were measured with a University of Wash-
ington Mark II Cascade Impactor. The par-
ticle concentration and the size distribution
results are not strictly independent; how-
ever, this should not be an important bias
of tihe data.
The only criteria for the rejection of
individual runs were errors in the procedure
168
-------
POLLUTION CONTROL AND CLEAN ENERGY
AIChE SYMPOSIUM SERIES
noted by Hofer and deviation from the isoki-
netic sampling velocity at the nozzle by more
than + 15%. There were a total of eighty runs
judged to be suitable for data analysis; fifty-
six were measured upstream (inlet), and
twenty-four were taken downstream (outlet) of
the wet scrubber.
The analysis of the plume opacity and
particulate mass concentration and size dis-
tribution data consisted of the following steps:
1. Correlation of measured concentra-
tion and light transmittance to the design
equation, W = A + B log (I/Iq).
2. Correlation of the measured K to
the measured mass mean particle diameter.
3. Correlation of the measured K to
the theoretical K computed from the mea-
sured size distributions with assumed par-
ticle densities and refractive indexes in the
scheme shown in Figure 3.
The data analyses are summarized in
detail in a report by Pilat and Ensor (14).
The important results follow.
The mass concentration was very poor-
ly correlated to the logarithm of the mea-
sured transmittance as shown in Figure 6.
The correlation coefficient was 0. 0244 for
the inlet site and 0, 204 for the outlet of the
scrubber.
This poor correlation of W = A + B
log (I/I ) was due t0 the variations in the
magnitude of the parameter K (K should be
constant for a good correlation) from changes
in the mean particle size. The aerosol was
composed of two populations as determined
by examination with a microscope. One pop-
ulation appeared to be tarry material boiled
out of the anode and carbon particles, while
the other was alumina dust from the ore feed.
The alumina dust in the system was
highly variable. There were highly statisti-
cally significant correlations between K and
the geometric mass mean diameter (correla-
tion coefficients from 0. 6 to 0. 9). This sub-
stantiates the observation that Hie variation
was due to particle distribution variation.
The scheme in Figure 3 was used with vari-
ous assumed values of particle density to try
to determine the best average material.
The assumed particle properties of
density (1.0, 2.0, and 4.0 g/cm3) and re-
fractive index (1.70, 1.5-10"4 i, and 1.96-
0.66 i) did not have statistically significant
improvements in the description of the data
(in the correlations of measured K with the
theoretical K computed from the measured
size distribution), although the absorbing re-
fractive index (1.96-0.66 i) produced the
best correlations.
The K measured and the K computed
from the size distributions measured with
the cascade impactor were correlated with
high significance (correlation coefficients
up to 0.87). An example scattergram is
shown in Figure 7. For these data, the re-
fractive indexes and densities were of sec-
ondary importance compared to particle
size.
The validity of the lognormal assump-
tion was tested by using two computer pro-
grams, one with a lognormal distribution
and a least-squares fit to the data and the
other with the points of the size distribution
to compute K. The deviation of the mea-
sured size distributions from the lognormal
model was statistically insignificant by com-
paring the two results.
SUMMARY
Over the years a systematic approach
to the understanding of in-stack plume opac-
ity has been developed. The basis of the
theoretical work has been the calculation of
graphs showing the parameter K as a func-
tion of the particle size distribution is a
basic part of the approach. The use of
stack test equipment and opacity monitors
is useful to gain knowledge of the error
limits in the application to design.
The data from a Kraft recovery fur-
nace and a aluminum refinery are reported
as examples.
169
-------
No. 156, Vol. "2
ENSOR and PILAT
acknowledgment
M
p
b
c
D
f(r)
The research reported in this paper
was supported over the years by training
grants, special fellowships, and a research
grant from the Environmental Protection
Agency, and research grants from private
industry. The publication of the paper was
supported by MRI funds.
NOTATION
org = geometric standard deviation
= viscosity
= specific gravity of the particles,
g/cm3
2 3
= extinction coefficient, m /m
= Cunningham correction factor
- hole diameter
= number fractional frequency par-
ticle distribution which, multiplied
by the radius increment dr, is the
fraction of particles between, r and
r + dr
= parameter formed by a ratio of the
particulate specific volume divided
by the extinction coefficient, cm /
m3/m'
= length of path, m
= particulate mass concentration,
g/m3
= light extinction efficiency factor, a
function of particle radius r, wave-
length of light and refractive in-
dex m
= particle radius
= geometric mean radius by number
= square root of the Stokes number at
the 50% collection efficiency [0,38
from the results of Ranz and Wong
-------
POLLUTION CONTROL AND CLEAN .ENERGY
AIChE SYMPOSIUM SERIES
(11) Larssen, S., D.S. Ensor, and M, J.
Pilat, "Relationship of Plume Opacity
to the Properties of Particulates
Emitted from Kraft Recovery Fur-
naces, " TAPPI_J., 55, 88-92 (1972).
(12) Nader, J.S., F. Jaye, and W. D.
Connor, "Performance Specifications
for Stationary-Source Monitoring Sys-
tems for Gases and Visible Emissions,"
EPA-650/2-74-013 (1972).
(13) Pilat, M.J., and D.S. Ensor, "Plume
Opacity and Particulate Mass Concen-
tration, " Atm£sJ_Environ., 4, 163
(1970).
(14) , "Opacity
of Aerosol Emissions from a Horizontal
Spike Soderberg Aluminum Reduction
Plant, " Water and Air Resources Divi-
GE0METRIC
STANDARD
DEVIATION, 1.96 - 0.66 i
WAVE LENGTH OF LIGHT ¦ 350 nm
GEOMETRIC MASS MEAN RADIUS , r,w (MICRONS )
Fig. 1. The parameter K as a func-
tion of the parameters of a
lognormal size distribution
for a white aerosol [after
Ensor and Pilat (7)].
sion, Department of Civil Eng., Univ.
of Wash., Seattle, Wash., Research
Report (1971).
(15) . and J. C.
Bosch, "Source Test Cascade Impac-
tor, "Atmos. Environ., 4, 671-679
(1970).
(16) Ranz, W. E., and J. B. Wong, "Impac-
tion of Dust and Smoke Particles, "Ind.
Eng. Chem.. 44, 1371-1381 (1952).
(17) Smith, J.E., and M. L. Jordan, "Math-
ematical and Graphical Interpretation of
the Lognormal Law for Particle Size
Distribution Analysis, "J. Colloid Sci.,
19, 549 (1964).
(18) Sparks, L. E., Personal communication
(1971).
GEOMETRIC
STANDARD
DEVIATION, a.
§
cc
B
3
K
2
REFRACTIVE INDEX » 1.50
WAVE LENGTH OF LIGHT * 550 nm
<"2
GEOMETRIC MASS MEAN RADIUS, r,„ (MICRONS)
Fig. 2. The parameter K as a func-
tion of the parameters of a
lognormal size distribution
[after Ensor and Pilat (jf)].
171
-------
No. 156, Vol. 7 2
ENSOR and PILAT
MEASURE
Wl0, L, W
k CORRELATE
K = -
P
WL
lci/t0
P
MEASURE
SIZE
DISTRIBUTION
r9» "<3
COMPUTE
K WITH MIE
EQUATIONS
compare
"K'S"
REFRACTIVE
INDEX
wavelength
Fig. 3. Analysis scheme for opacity-
data.
Jet Orifice
Streamline*
J«t Exit
/ J / S / s -T-Tl
Impaction
Plate
Trajectory of Trajectory of ParttcU
Impacted Particle too Small to Impact
Fig. 4. Schematic of an impactor
stage.
too
90
— 80
I-
2
Id
O
cr
UJ
5 70
UJ
o
z
<
to
<
oc.
60
I
o
~ 50
40
KRAFT RECOVERY FURNACE
LIGHT PATHlENGTH 3 FEET
» ELECTROSTATIC PRECIPITATOR ON
• ONE EL. PRECIP. FIELD OFF
¦ BOTH EL. PRECIP. FIELOS OFF
90°/. CONFIDENCE LIMITS
0.5
1.5
2.0
2.5
PARTICULATE MASS CONCENTRATION (CRAiNS/SDCF!
Fig. 5. Relationship of light trans-
mi ttance to particulate mass
concentration of a Kraft re-
covery [after Larssen,
Ensor and Pilat (11)].
172
-------
POLLUTION CONTROL AND CLEAN ENERGY AIChE SYMPOSIUM SERIES
INLET OF SCRU8BER TOWER
56 SOURCE TESTS
O
o
-N.
Q24
in
c
o
o>
Q20
O
5
CC
I—
z
0.16
CORRELATION COEFFICIENT « Q0244
y QI2
m QOS
,.,.nn636 >00Sg6JSai^l
UJ
CD
d
fe
g
004
Q7
Q6
0.4
Q5
0l3
0.2
MEASURED LIGHT TRANSMITTANCE , VI*
Fig. 6. Regression of particle mass
concentration to light trans-
mittance at scrubber inlet.
2.0-
— 1.6 -
™E
E
u
SC.
a
1.2 -
Q: 0.8
ft
<
UJ
INLET OF SCRUBBER
28 SOURCE TESTS
O m ¦ 1.70, CORRELATION COEFFICIENT » 0.626
• m »l.96-0.66i , CORRELATION COEFFICIENT §0.635
PARTICLE DENSITY • l.0flm/cm3
0.4-
CALCULATED Ke (cm3/m2)
Fig, 7, Correlation of measured K
to K calculated from par-
ticle size distribution at
scrubber inlet.
m
173
-------
E-2
Instrumental Method Substitutes for Visual
Estimation of Equivalent Opacity
Herbert C. McKee
Southwest Research Institute
Many air pollution control regulations limit the emission of visible effluents, based
on the visual observation of "equivalent opacity." Because of difficulties en-
countered in using visual observation, the Texas Air Control Board developed a
method of calibration which made it possible to use an instrumental method for
measuring visible emissions. A legal regulation based on this instrumental method
has been in effect for almost two years. Despite minor difficulties in calibra-
tion and maintenance, results have been satisfactory. The use of the instrumental
method avoids many of the difficulties inherent in using a regulation based on visual
observation, and continued use of the instrumental method is anticipated.
The regulation of emissions of dust
and other particulate matter is a major
portion of most air pollution control
programs. This is usually done by
either or both of two different methods;
(1) regulations based on the weight of
material emitted, and (2) regulations
based on the optical properties (opacity)
of the emission.
The visual appearance of smoke emis-
sions has been regulated for over 70
years by the use of the Ringelmann
chart to evaluate the density of gray and
black smoke.1-' More recently, the
same principle has been extended to the
control of emissions other than gray or
black bv utilizing the concept of "equiv-
alent opacity."' It has been estimated
that between 75 and 100 governmental
control agencies now use this concept to
control visible emissions. These regu-
lations usuallv specify that the emission
should not obscure the view of an obser-
ver to a greater degree than the obscura-
tion which would be caused by gray or
black smoke of a specified density as
measured by a Ringelmann smoke chart.
To aid in evaluating equivalent
opacity, good correlation has been shown
between visual estimates by a human
observer and instrumental readings
obtained with a transmissometer
installed in the stack of a special smoke
generator.4 This principle has been
used to train air pollution inspectors in
the estimation of equivalent opacity.®
It has been shown that the errors inher-
ent in a subjective observation can be
greatly minimized in this way so that
visual estimates are sufficiently repro-
ducible to form a basis for legal control.
This principle has also been upheld in
the courts.
Despite the use of smoke generators
as training devices, however, there is
still some uncertainty in making visual
observations of equivalent opacity.
Furthermore, such observations can be
made only with extreme difficulty at
night or during cloudy or rainy weather.
Therefore, the Texas Air Control Board
undertook to develop an alternate
method of measurement which could
overcome at least some of these limita-
tions and could provide more accurate
and reproducible results. This method
has now been in use for some time and
is specified by a legal regulation of the
Board as a substitute for visual esti-
mation.'
Principle of Method
Since the reason for controlling emis-
sions on the basis of equivalent opacity
is to limit the emission of materials
which cause absorption or scattering of
light, the direct measurement of light-
scattering proi>erties was selected as the
basis for developing an instrumental
method of measurement. Instruments
have been in use for many years which
measure light transmission across a
stack as an aid in estimating the amount
of material emitted or the density of
black smoke. Most of these instru-
ments include a light source and also a
photocell, bolometer, or other optical
measuring device; these are located on
opposite sides of the stack in such n way
that the light beam passes through the
stack gases which are being emitted.
The amount of light which passes
through the plume is measured, and the
result is expressed as |>orceiit transmit-
tanee compared to a measurement with-
out stack gases present.
Instruments of this nature are avail-
able from several suppliers, and there-
fore it was decided to utilize this type of
4M
175
Journal of the Air Pollution Control Association
-------
.ii-trument a- a -ub-titnte for tlx- vi-
sual e-tiiuatioii of f•'111iv.al<-j11 opacity.
By iti-talliii^ the m-truriii-iit in a duct or
stack where tin- optical properties of the
gas stream can be measured prior to
leaving the stack, a continuous record of
optical transmittance can l>e obtained.
Since the optical properties depend
primarily on the smaller particles pres-
ent in the gas stream, considerations of
isokinetic sampling do not appear to be
important; therefore, the instrument can
be located either in the stack or in a duct
leading to the stack. To help minimize
differences between instrumental read-
ings and visual estimates, the regulation
provides that the light source should
emit spectral energy approximately
equivalent to normal daylight, with no
more than 10% of the total energy in the
region of the spectrum above two
microns wave length. A calibration
procedure was developed to provide
instrument readings which could be used
as an approximate substitute for equiva-
lent opacity, so that the instrumental
method could be used as a legal basis
for regulation and control. Exactly
equivalent measurements cannot be
obtained because of variations in visual
estimates due to changes in sky back-
ground, emission conditions, or other
factors which influence visual observa-
tions.
Calibration
In order to avoid orrors due to changes
in exit velocity or other emission condi-
tions, the |>ermis.sible optical proi>erties
of the emission were established oil the
basis of the total volume emitted. In
this way, the total contribution of the
emission to the optical properties of the
community atmosphere was the primary
controlling factor.
The instrument response also varies
depending on the light path across which
the measurement is made. In conven-
tional installations, the light path usu-
ally is not extended the full width of the
stack or duct in which the instrument is
installed since disturbances iu the flow
pattern next to the wall might cause
erroneous readings. To calculate a
calibration curve, an exit velocity of 40
fps was arbitrarily assumed so that a
direct calculation could be made of the
assumed diameter of the plume, and
thus establish the length of the light
path across which the required trans-
mittance would be measured. The
required minimum transmittance was
selected as 70%, this being somewhat
more restrictive than the usual opacity
limit equivalent to a Xo. 2 Ringelmann
reading. Figure 1 illustrates the re-
sults of this calculation, and shows the
light path for a minimum transmittance
of 70% at any flow rate between 10,000
and 1,000,000 cfm. This volume is
based on the total volume of gases emit-
ted including air, combustion gases,
gaseous impurities, and all gaseous mat-
ter combined, but excluding water or
water vapor, calculated at the pressure
and tem|>erature existing at the point of
measurement. This graph, then, speci-
fies the light path across which the mini-
mum acceptable transmittance to deter-
mine legal compliance is 70%.
Since the dimensions and the geome-
try of some duct and stack installations
will not permit the use of this length of
light path, a method was needed to con-
vert the 70% transmittance to an equiva-
lent figure over some other lialit path.
This was done by the use of the follow-
ing equation:
Not*
4- Entraooiation beyond i.m'its I
shown may be erroneous
— log T = kl
where
5 2 5
10,000 100.000 1.000,000
volume t..vtted, cfm
Figure 1. ligti. path required for 70% trans-
mittance.
T — transmittance
I - length of light path
k — constant
In this equation, k can be evaluated
using a transmittance of 0.70 and the
light path indicated by Figure 1; this
value can then be used to calculate the
required transmittance for some other
length of light path.
Figure 2 is based on this equation and
can be used to determine the permissible
transmittance across some light path
other than the length specified by Figure
1. Experience has shown that this
equation is at least applicable over the
range of 50% to perhaps 250% of the
10 ,. 20
*ctuai light path
Light path for 70* transmittance
Figure 2. Transmittance versus light path.
light path indicated by Figure 1 in
relating transmittance to length of light
path. However, until more experience
is obtained, it is possible that extrapola-
tion beyond these limits may introduce
some degree of error.
Precaution*
Certain precautions should be kept in
mind in the use of this instrumental
method. For example, gases high in
moisture content should be measured at
a point where the gas temperature is
above the dew point to prevent conden-
sation; otherwise, loss of light transmit-
tance may occur due to the presence of
condensed water droplets in the gas
stream. Certain situations may also
arise in which the method cannot be
used satisfactorily, such aa the following:
1. Emissions may change in char-
acter after emission to become more vis-
ible (so-called "detached plume"), so
that an emission which is invisible in the
stack may in fact cause deterioration of
visibility in the ambient atmosphere
following emission.
2. Different emissions may mix and
react in the atmosphere to form aerosols
which restrict visibility, although each
emission may be essentially invisible
when first emitted. A familiar example
is the formation of a highly visible aero-
sol of ammonium chloride by the reac-
tion of ammonia with either chlorine or
hydrogen chloride.
Where these or other conditions make
it impossible to use this method, visual
inspection or some other alternate must
be found. However, so far these excep-
tions have occurred only infrequently.
Illustrative Example
Through the cooperation of the Dow
Chemical Company, data are presented
here to show a typical illustration of the
August 1971 Volume 21. No. 8
176
-------
use of this method of measurement. An
instrument supplied by the Bailey Meter
Company was installed in the stack
carrying flue gases from a series of lime
kilns at the Freeport, Texas plant of
Dow Chemical Company. Figure 3
shows a typical chart recording during a
24-hr period, illustrating the minor var-
iations in transmittance during this time
when the average transmittance across
the 5-ft light path of the instrument was
around 90%. Note that the recorder
was set to record in the range of 70 to
100% transmittance, with 70% at the
periphery of the chart and 100% at the
center.
Total volume of gases emitted is
approximately 130,000 cfm, although
this naturally varies to some degree with
changes in operating conditions. At
this flow rate, Figure 1 indicates that a
path length of 8.2 ft would be correct for
measuring the required 70% transmit-
tance. Since the instrument was
installed with a 5-ft light path, Figure 2
was used to determine the equivalent
reading with this installation; this indi-
cated that the transmittance should not
be less than approximate!}' 80%, which
corresponds to a 70% transmittance
over a light path of 8.2 ft. Thus, the
90% transmittance shown in Figure 3
illustrates better precipitator perform-
ance than the minimum required to
meet the regulation. This instrument
has been operated for several months
and has provided a convenient means of
monitoring the operation of the precipi-
tator to detect any minor changes indica-
ting a need for maintenance or adjust-
ment. The only instrument mainte-
nance has been an occasional cleaning of
the windows on the light source and bo-
lometer, and instrument stability'and
reliability have been reported as
satisfactory.
Advantages of Instrumental Method
lu the past two years, 40 to 50
instruments have been installed in vari-
ous industrial plants in Texas, for use in
meeting this new regulation. Processes
include lime kilns, cement plants, fertil-
izer driers, regenerators in catalytic
cracking units, and others. Based on
the experience which has been accumu-
lated, the instrumental method appears
to possess the following advantages:
1. Tlio instrumental method is com-
pletely objective and therefore is not
subject to the errors inherent in subjec-
tive evaluations bv human observers or
to variations among different individ-
uals making visual estimates.
2. The instrumental method appears
to be more accurate than visual observa-
tions. With adequate maintenance and
calibration of the instrument, it appears
to be possible to maintain an accuracy
of a few percent. This is better than
the accuracy generally obtained with
visual estimation, especially with high
Figure 3. Twenty-four hour chart record,
showing typical transmittance measurement.
transmittance values above 60 percent
which are required by many regulations
now being written. Previous experi-
ence has indicated that an equivalent
opacity of Ringelmann No. 1 is more
difficult to estimate than No. 2, which
means that transmittance values in
excess of 60 to 70% may be subject to
considerable inaccuracy if measured by
visual estimation.'
3. An automatic instrument can be
used at all times, and the results are not
affected by sunlight, cloud cover, dark-
ness, or other conditions affecting visual
observations. Readings can be ob-
tained on a 24-hr basis.
4. Continuous measurements can be
recorded automatically at less cost than
the cost of frequent observations by
trained human observers.
5. Since results are available on a
continuous basis, they can be used as a
means of process control for continuous
processes. Thus, the plant operator
can be aware of any malfunction of pol-
lution control equipment immediately
in order to take necessary corrective
action. This is not possible with regu-
lations based on the weight of dust
emitted since stack sampling to measure
weight is tedious and time-consuming
and subject to a time lag in obtaining
measurements which may be as much as
several days.
The exi>erience to date has not revealed
any serious problems in the operation or
maintenance of the instruments. The
major problem is that of rechecking cal-
ibration on production units which are
operated continuously for many months,
and various ways to accomplish this are
under study. Glass surfaces exposed to
flue gases must be cleaned occasionally
although this can be reduced by the use
of an air bleed which minimizes the con-
tact of dust laden gases with the window
surfaces. Plant operators have re-
ported no unusual maintenance problems
which cannot be handled adequately
by the average plant instrument depart-
ment.
Because of the advantages of this
method, the Texas Air Control Board is
considering making this method manda-
tory for all industrial sources subject to
regulation on the basis of opacity,
except for very small operations where
the cost of installation and maintenance
of an instrument would be out of propor-
tion to the benefit received. A figure
of 10,000 cfm has been mentioned as a
minimum size for mandatory installa-
tion, but this factor is still under consid-
eration.
Acknowledgment
The author wishes to express appre-
ciation for the assistance of the Texas
Air Control Board staff in the develop-
ment of calibration procedures and
other details in connection with the new
method which is outlined. In addition,
appreciation is also expressed to the
Texas Division of Dow Chemical Com-
pany, for permission to describe the lime
kiln installation which illustrates the use
of this method, and for the illustration
furnished.
References
1. Ringelmann, M., "Method of esti-
mating smoke produced by industrial
installations," Rev. Technique, 268
(June 1898).
2. U. S. Department of the Interior,
Bureau of Mines Information Circular
7718 (August 19.1o).
3. State of California, Health and Safety
Code, Chapter 2, Division 20, Section
24242 (1947).
4. Conner, W. D. and Iiodkinson, J. R.
Optical Properties and Visual Effects
of Smoke-Slack Plumes, Publication No.
999-AP-30, U. S. Public Health Ser-
vice (1967).
5. Coons, J. D., James, II. A., Johnson,
H. C., and Walker, M. S., "Develop-
ment, Calibration, and Use of a Plume
Evaluation Training Unit." J. Air
Poll. Control Assoc., IS, 199 (Mav
1965).
6. Regulation I. "Control of Air Pol-
lution from Smoke, Visible Emissions
and Suspended Particulate Matter"
(July 30, 1969). Texas Air Control
Board, Austin, Texas.
Dr. McKee, who is Assistant Di-
rector, Department of Chemistry
and Chemical Engineering of the
Southwest Research Institute—
Houston, also >erves as Chairman of
tlie Texas Air Control Board- This
paper was presented at the
-------
E-3
Texas Regulation Requires Control of Opacity
Using Instrumental Measurements
Herbert C. McKee
Southwest Research Institute, Houston, Texas
For over four years, a Texas air pollution control regulation has been in force
which permits use of a stack mounted transmissometer instrument to measure visible
emissions, thus avoiding the necessity for visual observations. Many types of
industrial plants have installed instruments, and extensive experience has been
obtained with this method of measurement and control. Plant operators and
enforcement officials have expressed a strong preference for this method, both to
demonstrate compliance and to provide data to aid in operating control equip-
ment. Some operational problems have been encountered, but these usually can
be overcome with proper care. Because of the advantages obtained with this
regulation a new regulation has been adopted which makes the instrumental
method mandatory for large sources of visible emissions.
In 1969. the Texas Air Control Hoard
adopted a regulation which permitted
the use of a stack mounted traiwmis-
someter instrument as a substitute for
the "calibrated eyeball" estimation of
equivalent opacity to evaluate visible
emissions.1 At the option of the plant
owner, the instrument could be in-
stalled, calibrated in accordance with
the calibration procedure included in the
regulation, and operated on a continuous
basis. If chart recordings or other
data were made available to the state or
local air pollution control agency, no
enforcement action would be attempted
based on visual observations. The basic
principles of this regulation and details
of the method of measurement have
been described in A previous publica-
tion.1
Instruments of this type had been
used for many years to measure the
opacity of smoke from coal fired boilers,
and the results were used to show
compliance with regulations based on
Ringelmann numbers. However, this
regulation marked the first time that a
lejal regulation specified an instru-
mental method as a substitute for the
visual estimation of equivalent opacity
to evaluate emissions other than black
smoke. This regulation did not require
much pioneering in a scientific sense
since the instruments were commercially
available and since the method of
calibration relied on well known princi-
ples of optics. It did represent a
pioneering effort in the legal sense, in
the writing of an enforceable regulation
in which the optical properties of the
gas emitted were measured to determine
compliance, rather than measuring the
weight of material emitted or the con-
centration in the flue gas.
During the past three years, many
companies have installed instruments
and have obtained experience in the
use of this regulation and the method of
measurement specified. Results have
been favorable for the most part, and
increasing use of this principle is antici-
pated to control emissions from sta-
tionary sources.
Previous Experience
Since this regulation was adopted,
more than 100 installations have been
made in various industrial plants in the
state of Texas. Some of these were
installed as an aid in process control
June 1974 Volume 24, No. 6
179
Ml
-------
with only :m indirect interc-t in meeting
iin ;ur pollution emission standard.
Mo-t, however. w-r'• installed primarily
a- :i mean* of t wet nr.: 'applicable emis-
sion -tandards. Tin; types of processes
include the following:
Alumina Kilns
Carbon lilack (furnace process)
Cement Kilns
Copper Smelters
Fluid Catalytic Cracking Units
Glass Furnaces
Incinerators
Instant Coffee Driers
Lime Kilns
Recovery Units in Paper Mills
Power Generation Boilers
(lignite-fired)
Sulfuric Acid Plants
Some practical operating problems
have been encountered, hut the results
have demonstrated definite advantages
for this method of regulation and con-
trol for mo-t processes producing visible
emissions. While there are some excep-
tions, the advantages outweigh the
disadvantages for most processes, both
from the standpoint of the control
agency and the plant operator. Two
major advantages can be cited: ll)
The instrument operates continuously,
and,the reading are not affected by
nighttime, cloud cover, or other cir-
cumstances that may make the visual
estimation of equivalent opacity in-
accurate or int|>o*sible. (2) Results arc
available instantaneously on a recorder
chart or other permanent record, and
enable the operator to use the measure-
ment as a means of process control to
avoid or correct excessive emissions to
the maximum extent possible.
Both of these are major advantages
which are important in controlling
processes which produce visible emis-
sions. From the standpoint of a control
agency, the visual method of measure-
ment is not completely satisfactory
because it can lie used at night only
with great difficulty or not at all, and
because it is subject to serious errors
when used to measure white dust emis-
sions during cloudy or overcast weather.
From the standi»int of the plant
operator, emission standards based on
weight measurements arc difficult to
meet on a continuous basis because of
the time lag involved in taking and
analyzing samples by manual methods.
The plant ojwrator cannot correct a
malfunction or process upset if it takes
three or four days to make a measure-
ment to determine whether or not
emissions have exceeded the acceptable
limit.
602
Theoretical Advantages
In addition tn these practical advan-
tages for both the enforcement agency
and the plant operator, there :ire theo-
retical reasons why an optical measure-
ment may in many cases be superior to
a weight measurement as a basis for
controlling particulate emissions. After
all, the reason for controlling air pollu-
1 t
one 11^.1 —
'.ion is not to limit the number of pounds
of dust in the air but to eliminate or
reduce the adverse effects of this dust
and particulate matter. In many ways,
it would appear that measurements of
optical properties may relate more
closely to these adverse, effects than
measurements of the weight of material
emitted. Some of the pertinent factors
can be summarized as follows:
¦ Adverse health effects are caused
primarily by smaller particles that
make up the so-called "respirable
fraction," while larger particles do
not ap|>ear to be nearly as important.
The size range of 1-5 n is frequently
mentioned as the respirable fraction,
although there is some evidence that
much smaller particles may also be
important. If weight measurements
are used to evaluate emissions, then a
large portion of the total weight
measured may be due to particles
larger than five microns which are
relatively unimportant in causing
adverse health effects. A measure-
ment based on optical properties
may be better, since optical proper-
ties arc determined to a large degree
by particles in the micron and
submicron range; this size range
overlaps a large ]>ortion of the size
ranpe that appears to be of most
concern in causing adverse effects on
human health.
¦ The problem of haze in the atmo-
sphere is directly related to the
optical [>ro|>erties of the emission,
much more so than to a weight
measurement which may be due
primarily to large particles that
contribute very little to light scat-
tering. The only obvious situation
which would be an exception is that
in which particles change following
emission due to water absorption,
atmospheric reaction, or other unique
characteristics.
¦ The tendency of particles to influence
gas phase reactions in the atmosphere
is most likely to be a surface phe-
nomenon, and therefore a measure-
ment which relates to surface a.?a
is more likely to be related to this
effect than a total weight mea-ire-
meni. Here again, a measurement
ba* and
over) will tend to fall out close to
the source, with some portion of the
fallout occurring on plant property.
Smaller particles (micron and snb_
micron) remain suspended for longer
periods of time, and thus it is the
smaller particles tliat make the
major contribution to area-wide
problems. This may be particularly
important with existing sources that
are not equipped with the most
modern control equipment. Plants
constructed in the future will almost
certainly lie equipped with control
equipment that provides essentially
complete removal of larsrer particles.
Regardless of the degree of removal
achieved, however, it would appear
that an optical measurement gives a
better indication than a weight mea-
surement of the contribution of the
source to the air pollution problem*
of an urban area.
Many air pollution regulation* (in-
cluding those used in Texas) recognise
t.iis distinction by establishing both
weight limits and opacity limits on
particulate emissions. As a practical
matter, the opacity limits nearly always
prove to Ik? more restrictive than the
weight limits. This suggests the possi-
bility that opacity limits alone would
provide adequate control, and that
regulations based on weight measure-
ments could be eliminated. The advan-
tages which have been shown for a
regulation based on optical measure-
ments add additional reasons for con-
sidering this course of action. The
major stumbling block at present is the
fact that equipment suppliers cus-
tomarily design and guarantee control
equipment on the basis of weight
measurements. Knowledge is not now-
available to develop desifin criteria
relating directly to optical measure-
ments, but such knowledge can he
obtained and applied if adequate incen-
tive exists. With more experience, it
may be feasible to eliminate particulate
regulations based on weight measure-
ments, and avoid the tedious and time
consuming manual sampling and result-
ing time delay,
180
Journal of the Air Pollution Control Association
-------
Operating Problems
Most of the problem? encountered in
using transmis«ometer instruments occur
in their installation and operation.
These problems are gradually being
overcome, but some trouble and expense
has been caused in certain installations
due to these operating difficulties.
One of the major problems in some
applications occurs in rechecking the
zero and span adjustments of the instru-
ment. With the instruments which
have been available in the past, the zero
reading is checked with the light source
turned off and the 100% transmittancc
value is checked with the light on but
with no dust or particulate matter in the
light beam. If the process is only
operated intermittently, this 100%
value can be checked when the unit is
shut down with little or no difficulty.
However, some industrial operations run
continuously for long periods of time,
and such a check is not possible. For
example, catalytic cracking units in
petroleum refineries may run for 3 to 4
years between shutdowns; obviously,
an extra shutdown costing several
million dollars is out of the question
merely to recheck an instrument calibra-
tion.
^ arious methods have been used to
overcome this problem, all of which have
some limitations. One method is to
provide an empty parallel pipe across
the stack which does not have siots to
permit the passage of flue gas. The
light source and detector are then re-
moved from the slotted pipe, placed on
the parallel pipe, and calibration
reading* obtained. An alternate to
this is to provide another pipe of
similar dimensions outside of the stack
so that the light source and detector can
be removed from the stack and placed
on the calibration pipe for checking
purposes. However, any time the light
source and detector are removed from
their permanent location, problems of
misalignment or temperature changes
may cause an inaccurate calibration
reading. In some cases, uneven heating
of the stack or of the calibration pipe by
sunlight can be great enough to cause
an alignment problem, especially in
large stacks 15 or 20 ft in diameter or
greater.
Some problems have also occurred
because of dirt in the flue gase* deposit-
ing nu the optical surfaces of the instru-
ment. giving unrealistically low readings.
If the^in-truinent is readily accessible,
cleaning the optical surfaces every few-
hours to insure accurate readings is not
an excessive burden although obviously
some labor cost is involved. However,
if the instrument is mounted several
hundred feet up oil a stack and is acces-
sible only by an exposed ladder, frequent
cleaning becomes an even greater
burden. This problem can be reduced
by proper design of the installation
although it may never be completely
eliminated. The common practice with
a negative pressure stack is to leave a
small space between the mating flanges
where the light source is bolted to the
pipe which extends through the stack
so that outside air is aspirated into the
slot, passes across the optical surface,
and then through the pipe, into the
stack, and out the top of the stack.
A similar arrangement is provided on
the other side where the bolometer or
other detector is mounted. With a
positive pressure stack, a sparging ring
is placed in the flanges and clean air is
forced in under pressure to sweep across
the optical surfaces in the same manner.
Some installations of this tyjie have been
quite successful and lens cleaning is only
required at very infrequent intervals,
sometimes as long as a month or more.
In other cases, especially where the
particulate matter in the stack is sticky
or gummy, these measures have been
only partially successful.
Still other problems have been en-
countered with process emissions at or
near the dew point where condensation
of moisture is a problem. Such prob-
lems can be avoided by withdrawing a
portion of the flue gas, passing this
stream through a heated chamlxr to
raise the temiwrature above the dew
point, and making a measurement in this
chamber, after which the sample stream
is then returned to the stack. How-
ever, very few installations of this type
have been constructed and operated
successfully, and moisture condensation
remains a problem in certain types of
industrial operations.
Future Improvements
Work is now in progress to overcome
some of the disadvantages which have
been encountered in the use of this
method of measurement. Two recent
advances in the instrumentation avail-
able have made the transmissometer
method more practical lor routine
application. The first, of the»e is the
availability of an instrument with an
automatic calibration sequence. This
avoids the problem referred to previously
of making a calibration check on
processes which operate continuously
for long periods of time. This instru-
ment contains a mirror which i>eriod-
ically diverts the light beam down
through a calibration cell and back to the
detector without going through the
stack, thus obtaining a check of the
100% transmittancc value. Since this
is accomplished automatically, the inter-
val can be set as short as desired and a
check is usually done routinely every
hour while the instrument is o|>erating.
The second development which
promises to be quite useful is the fab-
rication of a portable instrument which
can be inserted into a stack through a
3 in. pipe or other conventional sampling
port. This enables the air pollution
control inspector to make readings and
provide an independent cheek of the
permanently installed instrument. A
socially fabricated probe is mounted on
the instrument with a slotted tube con-
taining a corner cube reflector at one
end. This probe is then inserted in the
stack so that the flue gas passes through
the slot and any entrained particulate
then reduces the light measurement.
The source and detector are l>oth in the
instrument package which remains out-
side the stack, with the light beam
passing into the slotted tube through
the flue gas, being reflected by the
corner cube reflector, and passing back
through the slotted tube again to the
detector.
Field tests of this portable instrument
are now in progress by the staff of the
Texas Air Control Hoard.1 Preliminary
results have been inconsistent, but with
more experience the previous difficulties
likely will be overcome. This instru-
ment shows groat promise for extensive
use by air pollution control ins|)cctors,
and should also prove useful for a
company to provide an inde]>endeiit
means of checking the calibration of
permanently installed instruments.
Modification of Regulation
Because of the mostly favorable ex-
iwience with the previous regulation,
despite some of the practical difficulties,
the regulation lias been amended to
make the use of the transmissometer
instrument mandatory rather than
optional. Effective in December 1073.
all industrial process producing visible
emissions must install a tran>missonieier
instrument in the stack if the total flow
rate is in excc-s of 1(10.000 cfin. Con-
sideration is being given to reducing this
niuuninm flow rate to ">0.000 din in
order to cover more of the -ources of
visible emissions in the state of Texas.
If this is done, data obtained in a 10ti9
e-.iission inventory indicate that the
mandatory installation will apply to
June 1974 Volume 24, No. 6
181
603
-------
between 350 and 400 stacks in the state.
These stacks are responsible for approxi-
mately 70% of the total atmospheric
emissions in the state which have been
identified through this emission inven-
tory, omitting small sources of less than
10,000 cfm.
Regulations
Air pollution control agencies in other
jurisdictions are also considering various
regulations based on instrumental
measurement of optical properties.
Some brief mention of these will be
made here and more complete details
can be obtained from the agencies
involved.
In West Germany, the installation of
& transmissometer instrument is re-
quired to demonstrate compliance with
emission limits applicable to several
types of sources. A detailed evaluation
of various instruments was conducted4
leading to certification of particular
instruments aa meeting the requirements
of this regulation. According to present
practice, the governmental agency con-
ducts stack sampling periodically to
measure emissions on a weight basis and
these results are then compared to
transmissometer readings to establish a
relationship between weight and opacity
for that particular source. This rela-
tionship is then used to evaluate trans-
missometer readings and thus determine
compliance with the weight regulation.
The sampling is repeated by the
governmental agency at certain intervals
in order to reestablish the correlation
between emission limits and optical
properties.
The state of Connecticut has adopted
regulations which require the installa-
tion of monitoring instruments based on
opacity in certain fuel burning equip-
ment, incinerators with a capacity
exceeding 2000 Ib/hr, and process
sources that will emit in excess of 25
lb/hr of particulate matter. A set of
specifications has been developed, to
establish instrumental and operating
requirements for compliance with this
regulation.1 Specific instruments sup-
plied by various manufacturers have
been approved by the state agency for
this purpose.
Other state and local agencies have
been considering similar regulations, and
some have required instrument installa-
tions on individual sources where special
circumstances made such installations
desirable or necessary. Also, the En-
vironmental Protection Agency has
specified that opacity instruments musi
be installed in certain sources to deter-
mine compliance with opacity limits
established as a part of the New Source
Performance Standards.1 Specific re-
quirements and instrument specifica-
tions have not been promulgated.
Acknowledgment
The author wishes to express ap-
preciation to Mr. liill Stewart, \Ir.
Richard Brown, and other staff mem-
bers of the Texas Air Control Hoard,
for assistance in providing information
used in the preparation of this paper.
Many staff members have visited various
industrial plants to confer with plant
personnel on the problems encountered
in installation and operation of traits-
missometer instruments. Also, appre-
ciation is expressed to the many in-
dustrial organizations that installed the
instruments, and then shared their ex-
periences so that the advantages and
disadvantages of this new method of
regulation could be evaluated.
IMtrincti
1. Regulation I. "Control of Air Pollu-
tion from Smoke, Visible Emissions
and Suspended Particulate Matter"
(July 30, 1969). Texas Air Control
Board, Austin, Texas.
2. Herbert C. McKee. "Instrumental
method substitutes for visual esti-
mation of equivalent opacity," J.
Air Poll. Control Assoc. 21(H): 4M8
(1971).
3. 11. 1). Smith and C. Dean Wolbach.
Technical Programs, Air Pollution
Control Services, Texas State Depart-
ment of Health, Austin, Texas. Pri-
vate Communication, December 1072.
4. "Summary of Results of the Certifica-
tion Test for the HM4 Precision
Smoke and Dust Measuring Instru-
ment," Technischer Uberwachungn-
Verein, for the Secretary of Health and
Labor of the State of North Rhine
Westfalia, Federal Republic of fier-
many. 1070.
5. "Performance and Installation Specifi-
cations for Photo-Electric Monitoring
Instruments for Smoke Density and
Opacity," State of Connecticut, De-
partment of Knvironmental Protection.
1972.
6. Knvironmental Protection Agency,
"Standards of Performance for New
Stationary .Sources," Frdrral Rrgiilrr,
36(247): (Dec. 23, 1071). Also, "Pro-
posed Standards for Seven Source
Categories," Frdernl liefiatrr, 38(111)
(June 11, 1973).
Dr. McKee is assistant director,
Department of Chemistry and
Chemical Engineering, Southwest
Hesearch Institute, 3600 Yoakum
Blvd., Houston, Texas 77006. He
is a former chairman of the Texas
Air Control Board. This is a re-
vised version of Paper No. 73-24.)
presented at the APCA Annual
Meeting in Chicago in June 1973.
(44
182
Journal of the Air Pollution Control Association
-------
E-4
Measurement of Opacity and Particulate
Emissions with an On-Stack Transmissometer
Heinz P. Beutner
Lear Siegler, Inc., Environmental Technology Division
An on-stack transmissometer system which is designed to
provide a precision measurement of the opacity of visible
emissions is described. The sources of error in opacity
measurements with regard to recent EPA emission monitor-
ing requirements and planned specifications are discussed.
Sources of error are voltage changes, temperature changes,
light source and detector aging and effects of ambient
light. Other major operational errors are caused by
alignment drift and soiling drift. The methods employed to
minimize these errors achieve an accuracy of ±3% of
span and a maintenance free operational period of 3
months. The relationships between optical density, opacity
and transmittance are described. The instrument measure-
ment can be correlated with dust loading piovided the
particle size distribution is constant. Examples are given
of correlations obtained between optical density and par-
ticulate concenti ation in the gas on various types of emis-
sion sources and the observed error margins are sum-
marized.
TriMCflvCP I***
NflMtor Unit
'Sryi
•v,...»
."~'J
••—-a
Figure 1. Typic»' on • stack.
September 1974
Volume 24, No. 9
Two standard methods for the continuous measurement of
particulate emissions are:
Determination of the Ringelmann number or the equiv-
alent opacity of visible emissions.
Manual sampling using particulate sampling trains.
The problems of judging visual emissions by human observers
are well known. The original Ringclmann scale, comparing
the shade of gray of smoke with that of a chart, is useful only
for black smoke emissions. Today, smoke readers are trained
to judge the equivalent opacity of emissions of any color.
However, the results are de|ieudcnt on the jiosition of the sun
relative to the observer, errors arc made on overcast days, and
no observations can be made at aight. The reading errors
were acceptable for Riugclmaun numbers 3 aiul 2, but are
unacceptably large for today's requirement of Ringelmann
number 1 (20% opacity), or (10%opacity).
As currently practiced, smoke shade observations bear
little relationship to particulate loading in the gas or to partic-
ulate mass emissions. The most obvious inadequacy is the
dependence of the observed smoke shade on the stack exit
diameter. The same smoke is darker for a large diameter
than for a smaller one. In addition, smoke density is a funo-
tion of the density of the emitted material, the particle size
distribution, and the optical properties of the particles. In
fact, visible emission regulations are primarily inteuded to
control the appearance of a stack plume, not the quantity of
emissions. However, indirectly and in a very inadequate
way, they also control particulate mass emissions. For ex-
ample, a modern plant may comply with applicable particu-
late emission standards, vet it may be forced to reduce emis-
sions even further to meet opacity requirements.
The second method mentioned is the reference method for
measuring particulate emissions (El'A method 5 sampling).
It is a laborious manual procedure which, in large stacks, may
require 2 to 4 man-weeks of effort involving a cost of from
S3000 to 310,000. Typically, such sampling is |)erformed
annually or biannually, and as a result, it does not provide in-
formation about the performance of control equipment during
the interval between measurements.
183
MS
-------
Figura Z. RM4 transmiisomtter optical system.
Clearly, a better method for measuring and continuously
monitoring particulate emissions is needed. This paper dis-
cusses an oil-stack optical transmissometer system for pre-
cision measurement of the opacity of emissions and correlation
with the particulate concentration in the gas. The RM4
Visible Emission Monitoring System was designed and de-
veloped in Germany and complies with the very strict long
term operational performance specifications established by
the German pollution regulatory agency for continuous partic-
ulate monitoring instruments. It also meets the require-
ments of the United States Environmental Protection Agency
(EPA) for opacity measuring instruments. There are over
1000 of these instruments now operating in Europe and the
United States. The instruments are marketed and produced
in the United States, under license, by the Environmental
Technology Division of LearSiegler, Inc.
Transmissometer System Description
Figure 1 illustrates a generalized installation of the instru-
ment system, excluding the readout unit in the control room.
This instrument consists of an optical traasmitter/receiver
(transceiver) unit which is attached to one side of the duct or
stack, and a similar housing for the retroreflector on the other
side.
The transceiver contains a light source, a photocell/detec-
tor, and electronic signal-processing circuitry. The passive
reflector unit requires no electrical connections except for the
air blower provided for air Mushing oil that side.
Both the transceiver and the reflector arc equipped with
specially constructed air /low attachments to keep the optical
windows free of dirt deposits. The purge air normally is
supplied by two high-volume air blowers with filters. They
provide up to SO cfm each, depending on pressure conditions
within the stack and pressure drop through the filter. The
air-purging system typically provides 3 to 6 months of main-
tenance-free operation. Automatic calibration checks notify
the operator when cleaning of the optical surfaces is required.
This represents a significant advancement over other instru-
ments of this type available so far, since many others require
daily cleaning of the optical surfaces to maintain useful read-
ings, and their calibration cannot be verified during operation
without shutting down the process or employing time con-
suming manual procedures.
Optical Design
The instrument employs a chopped, dual-l>«am> optical
system (Figure 2). This design automatically compensates
for the effects of temi>crature changes, voltage changes, and
component aging. The instrument is non-rcspoO-'ive to am-
bient light, since only pulsed light at either of the two chopped
frequencies is measured.
Light from a single source in the transceiver is divided into
a measuring beam and a reference beam. The measuring
beam is transmitted across the entire width of the smoke
channel to a corncr-cube retroreflector which directs the beam
back through the smoke channel to the photocell detector, re-
gardless of small alignment variations between the two units.
The reference beam is projected directly onto the same photo-
cell detector. A rotating disc modulates the two light beams
at different frequencies so that both beams can be measured
by a common photocell, and electronic circuity then com-
pares the signals generated by the two beams. The reference
beam provides automatic gain control and, therefore, com-
pensates for any change in light output or photocell response
as a result of temperature variations, voltage fluctuations,
or component aging. As a result, the instrument is free of
drift errors usually caused by these variables.
Tolerance for Misalignment
Another common source of error in smoke measuring in-
stallations is a variation in alignment as a result of buckling
and temperature movements of the stack or duct walls to
which the instrument is attached. As a result, some instru-
ments require the installation of a slotted pipe across the duct
or stack to maintain rigid alignment. Among the disadvan-
tages of this arrangement are the possibility of measuring a
non-representative sample of the smoke or dust, a reduced
instrument sensitivity as a result of measuring only the
smoke passing through the slotted section of the pipe, and the
possibility of light vignetting, i.e., scattered light is reflected
off the walls and reaches the detector. In large installations
and at high operating temperatures, such a pipe can sag,
causing a large measurement error, and the pipe usually is not
recommended for stack widths greater than 20 ft. TheRM4
instrument system does not utilize a pipe across the smoke
channel, and it can be used over distances as great as 52 ft.
The instrument system tolerates commonly encountered
alignment changes up to ±0.4°, without loss in accuracy, by
utilizing a simple patented concept.1'* The light beam is
focused ou the retroreflector for each specific distance and has
a uniform intensity (±2%) throughout its cross section at the
focal distance. For each sjwcific distance, a reflector aper-
ture size is chosen that will maintain a largo ratio of beam
cross sectional area to reflector area. Since the beam is uni-
form, the retrorcllector always returns a constant amount of
light, regardless of what portion of the beam impinges upon'
the retroreflector. Furthermore, the corner cube retrore-
flector always returns light directly to the source, regardless of
the angle of incidence. Within its accuracy specifications,
this instrument system tolerates alignment changes of ±0.4°,
which is equivalent to measuring-beam movements of I to 3
in. at distances of 10 and 30 ft. This is sufficient to account
for all alignment variations encountered in actual practice.
In addition, the instrument is equipped with an optical bull's
eye which allows verification of the alignment at any time
during operation. If necessary, the alignment can be ad-
justed using two fine-adjustment screws on the instrument.
Automatic Calibration
Temperature changes, voltage changes, accumulation of dirt
on the optical windows, and alignment changes are the major
sources of drift in smoke or dust density measuring installa-
tions; the calibration of such installations must be checked
frequently. New source performance standards, published
by the EPA,' require opacity monitoring instrumentation in
all new steam electric power generating stations and further
specify that such instruments be calibrated at least once every
day. None of the two-ended smoke-density monitoring in-
struments so far available can be calibrated during operation
on the stack since, in any two-ended system (light source
and detector on opposite sides of the smoke channel), it is
necessary to stop the exhaust gas flow in order to obtain a
MC
184
Journal of the Air Pollution Control Association
-------
calibration for zero percent opacity. The single-ended RM4
system checks its zero and span calibration automatically at
regular intervals without interrupting stack operation. The
calibration values obtained are recorded directly on the chart
paper. The calibration checks can also be activated man-
ually at any time from the control room.
The calibration checks in the system arc accomplished by
insertinz, in the liaht path outside of the optical window of the
transceiver, a zero reflector which simulates the retroreflector
on the opposite erformcd by
the Techiiischer I berwachungs-Vorein (TCV),' demonstrate
that none of the following variables causes an error exceeding
±3% of span for a full scale range of 0 to 0.45 double-pass
optical density* (0 to 0.225 single pass or 0 to 40.4% opacity),
and most ;irp well under this limit:
Voltaic variations within ± 10% of nominal.
Temperature variations from -30°C to + 55°C (-22°F
to + 131°F).
Component aging and effects of ambient light.
Alignment variations within ±0.4° of the optical axis.
Calibration and linearity errors in measurement scale.
Operational tests on stack emissions of cement plants and
power plants confirm that none of the following errors exceeds
±3% of span for the 0 to 0.45 double-pass optical density
full-scale range:
Zero and »pan drift over a maintenance-free operational
period of 3 mo.
Soiling ami alignment drift over a maintenance-free o|M?ra-
tional period of 3 mo.
I—. light he»m pumi through th« #mok« twiee. therefor#
T 'Th»n 'hi op«.«> •«» hy .n obwrv., <>M (urth.r b.low>.
mewurn more m»n llt- »
— ,,
1 «... ¦¦
r- - : 1
| I
1 H f—
1.
i
i >n
H—h—1
I— riiuar J
4—
r--
¦ i-.— i ¦¦ ¦
i
I ;
1 !
1
1
1 1
—)
s t !
i .
1
i
! !
Figure 3. Example of emission trace from cement plant showing
hourly zero and span calibration checks.
A choice of five full-scale measuring ranges which correspond
to stack opacities of 0 to 9.8%, 0 to 18.7%, 0 to 40.4%, 0 to
64.5%, and 0 to 87.4% is provided. The instrument is sensi-
tive to changes as low as 0.05% opacity, and, on its lowest
range, it can accurately measure opacities of 1 to 2%. This
high sensitivity is a result both of the double pass of the light
through the entire stack or duct width and of the expanded
measuring scales.
The absolute accuracy of the transmissometer is based on
the calibration of the linear scale of optical density against
neutral density filters of known transmittance. The mea-
sured values must be within ±2% of the calibrated values of
the filters.
The calibration against neutral density filters is acceptable
if the angles of projection and view are small. A neutral den-
sity filter is not strictly equivalent to a smoke or dust aerosol,
since light attenuation in the filter is due primarily to absorp-
tion, while attenuation in the aerosol is the result of both ab-
sorption and scattering. Instruments with large angles of
view measure a portion of the forward (or backward) scat-
tered light and therefore measure a lower apparent o|>acity
for the aerosol than for a neutral density filter of equal opac-
ity .••* Since, for particle sizes under 2 microns, nearly an
order of magnitude less light is backscattered than forward
scattered, and since the angles of view and projection are
very small, the folded beam system utilized in the trans-
missometer has an absolute accuracy equal to any equivalent
two-ended transmissometer. The angle of view for this in-
strument is ±1.8° and the angle of projection is ±0.7° from
the optical axis. Proposed EPA specifications for visible emis-
sion measuring instruments suggest limits of ±2.5° for the
angles of view and projection in order to assure an accurate
absolute opacity measurement by instruments from different
manufacturers.
Photopic Light
An additional requirement for absolute accuracy of the
measurement is a specified spectral range, since light attenua-
tion by a polydisperse aerosol with mean particle size of less
than 1n is a function of the wavelength of the light. The
s|>ecific(i spectral distribution is photopic, the same as that of
the light-adapted human eye, because one objective is the
measurement of the opacity of visible emissions as observed
by the human eye.
The instrument utilizes a photocell filter combination to
achieve "photopic" s|>ectral sensitivity. Figure 4 shows the
spectral sensitivity of the instrument compared to the output
of an incandescent light source. The maximum light in-
tensity is at 0.53 micron wavelength compared to about 1.0
micron for the incandescent source. Instruments sensitive to
the entire output of the light source are subject to two serious
September 1974 Volume 24, No. 9
185
U7
-------
LSI Photopie
SpMlril HMponl*
TunS*i«n
nc«>dmc*nl Lifl"' MO® "
-""I ' I 1 1 .
1.0 1
4 Mrirad ¦ m
VitiW*
JUrmoiM
WfewlmgtlfMicrdn*
fig" r« 4. Photopie jp»ctr»l s*n*itivity of tha RM* camp«(«a
Co incandescent light tpvctrum.
errors, first, water absorption bands in the near infrared
portion of the spectrum cause large measurement errors for
stack fcasfts containing high humidity. Second, the instru-
ment response to submicron particulate matter is significantly
lower than the response of the human eye, resulting in low
readings when the stack emissions contain substantial pottum
of submicron particulate matter.
Figure 6 shows the light attenuation of a polydisperse aero-
sol for two different spectra! maxima as a function of the mean
particle sue.' The attenuation of the 1.0 m light falls off
significantly at about 0.8 to 0.6 n particle size, while the 0.55
n light is attenuated by particulate matter down to 0.3 to
0.2 it size.
In the absence of photopie light and especially when the
major instrument response is within the infrared region, as in
instruments utiluicig a bolometer detector eelI, larjre errors
result when the emissions contain large portions of submicron
particulate matter. This generally occurs in modern plants
with strictly controlled emissions, since the control equipment
primarily removes the larger particle sixes, while submicron
particles pass through.
Instrument Output
The instrument has a 0 to 20 milliampere output directly
proportional to optical density units, with an elevated in-
strument zero at 10% full-scale (2 milliamps).
Figure 6 shows the relationship between the optical density
output of the instrument and the more familiar scales of opac-
ity or transmittance. Optical density (D) is defined as the
logarithm (bas« 10) of one over transmittance (T):
D - log 1 /r - -log T (1)
Since opacity (O) is defined by O - I — 7\ the relationship
with optical density can be expressed as:
D - -log T « -fog (I - O), or
T - 1 - O - 10-o (2)
If T or (1 — 0) is plotted on a logarithmic scale, a linear correla-
tion is obtained as shown in Figure 7. The above relationship
is known as fiouguer's Law, equivalent to the well known
Lambert-Beer's Law for the measurement of absorbance by
colored solutions. It can be shown that the optical density
for the case of light attenuation by a smoke or dust aerosol,
like the absorbance for the case of light absorption in solu-
tions, is proportional to the pathlength (a) and the concentra-
tion (e) of the light attenuating material.
D «¦ k ¦ c • d
(3)
This relationship is used to correct the measured optical den-
sity values .« Micron nr—wmw U«M
l.» M *• M
111
»u
MW « •»
§, Jtt«fW«N0A ot * ••roiol as » function of pjrticf* six*.
^ g g Journal of the Air pollution Control Association
-------
is available. It allows the selection of a meter readout in
terms of either optical density at the measuring site or opacity
at the stack exit, and either or both of these outputs can be
recorded on a strip chart recorder. The available measure-
ment ranges indicated in Figure 0 are single pass equivalents of
the instrument's double pass measurements.
Correlation with Dust Loading
According to Bouguer's Law, the optical density instrument
readout is directly proportional to the dust concentration in
the gas. The constant of proportionality differs with the
density of the particulate matter, its size distribution, and its
optical properties. The constant can be calculated if all data
are known." • In practice, it is easier to experimentally
determine the proportionality constant for each installa-
tion.111'"'"
Empirical calibration of the transmissometer in terms of
particulate concentration requires at least one data point ob-
tained by stack sampling, using an approved sampling train
and method (EPA method 5). The average of the trans-
missometer readings over the period of the stack sampling
test equals the measured dust concentration in mg/m' or
gr ft1 at actual conditions. A straight calibration line is
drawn from the origin through the experimental point.
Calibration in terms of dust loading is valid only as long as
the particle size distribution and other particulate properties
do not change significantly. This condition is generally ful-
filled in emissions from sources with control equipment of high
efficiency. To establish the accuracy of the calibration over
a range of emission rates and process conditions, a large
number of calibration points (at least 12 to 15) must be ob-
tained, and the regression line and its confidence limits de-
termined. Figure 8a illustrates a calibration curve obtained
on the stack of a cement kiin equipped with an electrostatic
precipitator. The 95% tolerance intervals at the 95% con-
fidence level are indicated in the figure. Figure 8b shows the
same calibration curve after correction to standard conditions
(temperature, pressure) and dry gas. The corrected calibra-
tion curve can be used when gas temperature, pressure, and
humidity are constant. Figure 9 shows a similar calibration
13
1.0
0.0
0.0
*
t
o.r
o
0.9
i
O
OS
0.4
0.3
01
0.1
J
k
/
f\
t
»
/
/
i
l
i
~
/
1
1
1
t
i
i
i
i
...
1
1
1
1
•
i
Opacity, %
21
I
M
_i_
70 M K ,00
-J I I I
rrwwMIMiK*. * «0 M «•»«•••»¦ « n >« •
Maw OMhea l-H>M •* *••• '"»¦"«»«
MMMnflW KsngM
riaura < Relationship between optical density and
opacity or trsnsmittance (linear scale).
% T/aMMWittw*
10 9 0 7 9 9 4 J S
100 90 00 70 00 90 40 30 M
La
.s
..f
s
4
0 10 20 M 40 N M
N 11 II M M
M IT
% omchr
Figure 7. Relationship between optical density and opacity or trans-
mittance (logarithmic scale)
curve obtained on a lignite-fired boiler in Germany, and
Figure 10 shows a similar calibration curve for a bituminous
coal-fired power station.
The system is being used in numerous installations in the
United States and in Europe for the continuous monitoring
of both opacity and particulate concentration, based on the
correlations established between the optical reading and the
dust loading. In some states of the German Federal Re-
public the instrument is approved for compliance monitoring
on specific large emission sources. 'In this application the in-
strument calibration in terms of particulate concentration is
checked by manual testing once a year. In evaluating the
monitoring record, credit for the possible measurement error
is given.
Table I summarizes the levels of error observed in a number
of correlation lines for typical emission sources. Four dif-
ferent definitions of error have been applied. Values have
been calculated from the available raw data using a computer
program.* The error definitions are as follows:
A. Standard deviation is the root of the mean.
B. Type 1 error is the 95% confidence level that the true mean
of all observed optical data for a given particulate concen-
tration will lie within the limits.
C. Type 2 error is the 05% confidence level that the next one
observation will fall within the limit.
D. Type 3 error is the 95% confidence level that 95% of all
possible observations (95% tolerance) will fall within the
limit.
In Europe type 3 error calculation is typically applied, while
in the U. S. the type 2 calculation is most common. The
smallest error level is obtained by the type 1 method. As a
percentage of the measured mean particulate concentration,
the type I error level ranges from 5 to 21.1%.
Because transmissometer data are typically correlated with
1 hr stack sampling data, an individual data point in the con-
tinuous transmissometer recording can be defined as a 1 hr
* Tbo author thanks Gerald McGowan for hit aaaiotaneo in providing th«
computer data. The error calculation in available aa a eervict from Lear
Stagier, Inc. Environmental Technology Division.
September 1974 volume 24. No. 9
187
169
-------
CvffMM Kill*
• Normal
A Norm»( Moini..nt:
easily allows determining 8 hr or 24 lir averages. The r pr^t
on the error probability is therefore equivalent
-------
Standard Condition*
0.0S 0.10 0.11 OJO 0.3 >
Optical D«n»lty - Sln«l« Pat*
Figure ». Relationship between optical density and par-
ticulate mass concentration for a lignite fired boiler
(standard conditions).
Bituminous Coal Fired Sailer Emiaaiena
I
z
¦ ISO
J
S joo
c
i
e
u
- 100
y
t-ff
-0.1s
n
<
9
-®.10 j
/A
PJ
/¦
•
/A
¦0.0s1 j
7/
• ooi (,i o.ii o.a (.»
Optical Penally — Simla Mat
"®"r* Relationship between optical density and
"'ass concentration for a bituminous coal
fired boiler (standard conditions).
The instrument's continuous output signals for both opac-
ity and optical density, corresponding to particulate mass
concentration, can be used as inputs to a total emission moni-
toring system. Other sensor inputs required include tem-
perature, humidity, pressure, and gas How rate, as well as other
emission parameters such as sulfur dioxide and nitric o.\ide
content. A small data processor or controller can then cal-
culate mass emi.-sion data (lb- lir or lb MUtu) and applicable
average* (e.g., 10 min, 2 hr, and 8 lir averages) and provide
continuously updated printouts. Such complete monitoring
systems will be used in the future.
There are some limitations to the applicability of the on-
stack transmissometer. Water droplets in the gas are mea-
sured as if they were particulate matter. Unless mist emis-
sions are to be measured, efficient demisters must be employed
after wet scrubbers, or a portion of the gas must be reheated
prior to measurement in a by-pass duet.
Similarly, the instrument cannot correctly measure the
opacity of emissions which form visible plumes after the gas
exits from the stack (such as sulfuric acid mist or va|*>rs
which conden^ on contact with the cooler air).
Table II. Comparison of stack opacities for different emission
sources under equalized conditions.
Calcul. for
Single 150 mg/Nm1 and
Mean passO.D. Diam.at 3 meter stack diam.
part. at mean measur-
Emission conc. of part, ing site %
source (mg/Nm1) conc. (m) O.D. Opacty
Cement
Bituminous
coal
Lignite
Municipal
incinerator
Kraft
recovery
Hogged fuel
Bituminous
coal
179.9
322.3
111. 2
90.3
76.4
206.0
252.3
0.1029
0.0944
0.0580
0.0235
0.0746
0.1290
0.1432
1.95
3.25
4.00
3.66
1.52
0.1320
0.0722
0.0293
0.1200
0.1854
4.88 0.0523
26.2
15.3
6.6
24.2
34.8
11.3
References
1. E. Sick, U. S. Patent S'o. 3,617,7.">6, issued November 2, 1971-
2. K. W. Biihne, "Investigations into the directional dependence
of photoelectric smoke density measuring instruments."
Staub 31: 22 (1971), (in English).
3. "Standards of performance for new stationary sources,"
Fed. Reg. (Dec 23, 1071).
4. K. W. Biihne and L. Diiwel, "Recording dust emission mea-
surements in the cement industry with the RM4 Instru-
ment," Staub, 32: 10 (1072) (in English).
D. S. Ensor and M. J. Pilat, "The effect of particle size dis-
tribution on light transmittance measurement," Am. Ind.
Hyg. Assoc. J., 32: 287 .1971).
6. C. M. Peterson and M. Tomaides, In-Slack Transmissometer
Techniques for Measuring Opacities oj Particulate Emissions
from Stationary Sources, Report to EPA under Contract No.
6S-02-0309, April 1072 (Environmental Research Corp., St.
Paul, Minn.).
7. W. L). Connor and J. R. Hodkinson, Optical Properties and
Visual Effects of Smoke Plumes, Environmental Protection
Agency, Office of Air Programs Publication Xn. AP30 (Co-
operative Study Edison Electric Institute and Public Health
Service, 1967).
8. M. J. Pilat and D. S. Ensor, "Plume opacity and particulate
mass concentration," Atmos. Environ., 4: 163 (1970).
0. O. S. Ensor and M. J. Pilat, "Calculation of smoke plume
opacity from particulate air pollutant properties," J. Air Poll.
Control Assoc., 21: 4SJ0 (1971).
10. T. F. Hurley and D. L. R. Bailey, "The correlation of optical
density with the concentration and composition of smoke
emitted from a Lancashire boiler," J. Inst, Fuel, 31: 534
(IO08).
U. L. Diiwel, "Comparative Studies of Different Measuring
Principles for the Continuous Monitoring of Particulate
Emissions from Lignite Fired Boilers," Proceedings Second
Int. Clean Air Congress, Edited by H. Si. Englund and \V. T.
Berry. Academic Press, New York, 1971, pp. 437-448.
12. >S. Larssen, D. 3. Ensor, and M. J. Pilat, "Relationship of
Plume Opacity to the Properties of Particulates Emitted
from Kraft Recovery Furnaces," Tappi, 55: 88 (1072).
13. National Council for Air and Stream Improvement, Test
report to be published.
14. M. J. Pilat and I). J. Lutrick, Relationship of Plume Opacitii
to Properties of Aeroso/ Particles Emitted from a Hog Fuel
Hoiler. University of Washington, Department of Civil
Engineering Research Report, 1073.
1.'i. \V. A. Schneider, "Opocity monitoring of stack emissions:
A design tool with promising results." The 1974 electric
ulilitu ... Generation I'lankook. McGraw-Hill, New York,
1974, pg. 73.
Dr. Beutner is a Vice President of the Environmental
Technology Division of I.ear Siegler, Inc., One Inverness
Drive East, Englewiwid, Colorado 80110. Since 1071, Dr.
Ueutner has f>ecn responsible for the marketing and appli-
cations development efforts for the division's pollution-
monitoring instrumentation. His present responsibility is
business manager for Lear Siegler's air pollution control
equipment. This is a revised version of Paper No. 73-169
presented at the With Annual Meeting of APCA at Chicago
in June 1073. The author thanks Ernest E. Mau for hut
assistance in the preparation of this poper.
September 1974 Volume 24. No. 9
189
171
-------
<
ATTENUATION
MASS CONCENTRATIO
l£i
s)
m
CM
l-U» 'lN3l3ldd303"N0llVnN3UV
190
-------
SECTION F: REMOTE MEASUREMENT OF PLUME OPACITY
Page No.
F-l. Remote Measurement of Plume Opacity, W. Conner, Excerpt
from EPA 650/2-74-128, November 1974 193
F-2. "Remote Measurement of Smoke Plume Transmittance Using Lidar,"
Charles S. Cook et al., Applied Optics, August 1972 .... 197
F-3. Amendment to NSPS to Allow Methods Other Than Reference
Method 9 for Measuring Plume Opacity, Federal Register,
Vol. 42, No. 99, May 23, 1977 205
-------
F-l
Excerpt from EPA 650/2-74-128, Measurement of the Opacity and
Mass Concentration of Particulate Emfssions by Transmlssometry,
W. Conner, November 1974
REMOTE MEASUREMENT OF PLUME OPACITY
There are three instrumental methods that have been described for
remote measurement of plume opacity. Two of the methods use photometry of
the sun or of contrasting targets through the plumes for the measurements.^
These methods are considered methods of opportunity since they are applicable
only under certain conditions. The third method is more general and uses a
laser radar (lidar) technique. With this method, the plume opacity is
determined by shooting a pulse of light through the plume and measuring the
ratio of the light backscattered from the pulse by the atmosphere before and
after the plume.
SUN PHOTOMETRY
Measurement of plume opacity by sun photometry is restricted to times
when the sun is unobstructed by clouds and may be viewed through a discrete
cross section of the plume at the stack exit. Conditions for the measure-
ment are usually best on clear days when the sun is relatively low in the
sky, the sun can be viewed through the plume at the stack exit, and the
direction of flow of the dispersing plume in the atmosphere is away from
the sun. A sun photometer specifically designed for plume opacity measure-
ments is available from the Shell Development Company. The measurement can
19
also be made with a standard sun photometer, e.g., the Yolz sun photometer.
However, it may be necessary to modify the spectxal response of the standard
sun photometer to obtain good sensitivity to visible light. A Volz sun
photometer modified for plume opacity measurements has been described.^
193
-------
CONTRASTING TARGET TELEPHOTOMETRY
Measurement of plume opacity by contrasting target telephotometry is
restricted to times when contrasting targets can be viewed through the plume
at the stack exit and to times when the ambient illumination of the plume is
stable during the measurement. The contrasting targets may be distant hills,
tall buildings, or towers and the sky adjacent to them. For the plume
illumination to be sufficiently stable during the measurement normally
requires that the sky be clear or uniformly overcast. The plume opacity is
determined by using a narrow-angle-view (less than 1/2 degree) telephotometer
to measure the ratio of the luminance difference between targets when viewed
through and beside the plume. Narrow-angle telephotometers suitable for the
measurement are commercially available.
It is also feasible to determine the opacity of a plume by photo-
graphing the contrasting targets through them, e.g., from a helicopter. The
ratio of the luminance differences between the targets is then measured from
the film images with a laboratory densitometer. Plume opacity measurements
by telephotometry and photography of contrasting targets have been described
in detail.6'20
LIDAR
The technique of determining plume opacity by lidar was first proposed
to the Edison Electric Institute and the U. S. Public Health Service hy the
late M.G.H. Ligda of Stanford Research Institute (SRI) during the middle
1960's. EEI and PHS pursued the development of the technique as part of a
cooperative research study in the measurement of the opacity of plumes from
stacks of steam electric power plants. They contracted SRI to conduct a
laboratory and field study of the method using existing SRI lidar equipment
194
-------
designed for atmospheric studies. The results indicated the method was
feasible and defined many of the design parameters needed for a lidar
system for plume opacity measurements.^* The General Electric Company
was then contracted to develop a mobile lidar system specifically designed
20 22
for remote measurement of plume opacity. ' The performance of the lidar
is presently being evaluated by the Environmental Protection Agency in con-
junction with studies to develop and evaluate opacity measurement methods
for various sources.
Application of the pulsed lidar system described above is primarily
for opacity research studies and for further development of the method. In
its present form the instrument would have little application for surveil-
lance work by a control agency because of cost, operating power, and com-
plexity. For the development of a smaller, low-powered, and less expensive
instrument for remote measurement of plume opacity, a lidar technique using
a frequency modulated, continuous wave (CW) laser is being investigated.
The initial investigation of the CW lidar indicates that the method is
feasible, and further development is planned.^
195
-------
F-2
Ht |>rimc«l Iri.m AIMM.IKI) OPTICS. Vol. 11. pure 1742. AiiiaiM I'.'Ti!
Copyright 1!172 liy lI'c ('piu .il Smii-ly el Aiirtk,« and rcprinUit liy permission nl (lie copyright uwnvr
Remote Measurement of Smoke Plume Transmitt'ance Using Liciar
Charles S. Cook, George W. Bethke, and William D. Conner
A van-mounted rubv lidar system has been developed for the remote single-ended measurement of
smoke plume transmitt.mce at ranges of 130-.100 m. The measurement of transmittance Is obtained by
comparing the dear air lidar return from the near side of the plume with that from the far side. The
lidar system performance h:u lieeti tested using synthetic targets of known transmittnnce, and its Ac-
curacy limitation has been found to be pliotomultiplier tube afterpulsing. Results from these target
tesU as well as tield testa using real power plant, smoke plumes where comparison tclcphotomoter mea-
siiremeius were made are discussed in terms of accuracy for the determination of plume traluminance.
Introduction
The.use of laser radar (lidar) systems for mete-
orological and air pollution measurements has been de-
scribed extensively by Colli.*1 and others.5-4 The sub-
ject of this discussion is the specific use of a pulsed ruby
laser lidar system lor remotely measuring the trans-
mittancc (or opacity) of a smoke stack plume. The
impetus for this effort was t he desire. on the part of the
halloa Llectric Institute and the Environmental Pro-
tection Agency, to obtain a remote technique giving an
objectn e measurement 01" plume transmittancc. Early
uoik b\ E\ansJ explored the various possible lidar
techniques for making such a rejnote transmittancc
measurement and described some experimental results
obtained using an oxi .tin- lidar system. This effort
pointed out a principal ditiiculty in the single beam,
single shut technique that was described as photomulti-
plier tube afterpulsing. in ^jl0 pvo.-soi,t system, after-
pulsing >till remains the accuracy limitation in the
transniittancc measurement.
Ihc ba-ic technique for obtaining the lidar deter-
mined transmittanco. Tt, is illustrated in Fig. 1 which
shows a drawing of an ideal lidar signal return through
a smoke plume. The signal return begins to decrease
ns (range) ~®, corn-ponding to the gt iifial lidar equa-
tion,1 at a range. >•, <>t about 1 ."-10 in where tin- laser and
rece iver fields achieve lull overlap. This 1 /Vs perform-
ance will continue, providing the atmospheric back-
scattering is homngeiifjU'. ami the range interval is
suflieiciitly small that extinction can be neglected.
When the laser pulse intercepts the plume the back-
Tl.e fir.-t twit ailtUr* iruni-l :trr «itl> tlrncral lllnliif C'uin-
pativ, Kini; ..f l'ru-ia, r«-nn-vlv:.:ii;i I'Mu.; tin- i.llicr aiitlim i<
v\itli l^iviiMiiiui'tit'il Atti-M-y, A:: I*. l!;Hinn Ci,nli.,l
Oil',,Uv-t'.Mitli »\n!;. N. rlli Cainiina.
lii!Ci'ivttl 111 January l'.>7—.
174? APPLIED OPTICS / Vol. 11, No. 8 / August 1972
scattered signal may increase as much as 40 dB above
the ambient atmospheric scattering, appearing as a
spike on the lidar signal. Light backscattered to the
receiver from ranges beyond the plume will have tra-
versed the plume twice and thus will have been at-
tenuated twice by the plume. Thus, if the signal from
just in front of the plume (B) and the signal from just
bevoncl the plume (.4) are both extrapolated in a 1/r®
manner to a common range as indicated in Fig. 1, the
one-way lidar determined transmittancc is given by
Tl - (l)
In actual practice, the intense backseat tering received
from the plume itself causes photomultiplier tube after-
pulsing to occur in the region following the plume in
addition to causing the system's amplifying electronics
to be driven hard enough to produce a slow signal re-
covery in that region. Both these factors contribute
to the error in the extrapolated signal value, .1, causing
the experimental value of transmittancc, TL, to bo too
large.
Mobile Lidar System and Deteclor Gating
A block diagram of t he lidar system is shown in Fig. 2.
It is a side-bv-side refracting system mounted on an
elevating pedestal mount in a hinged roof panel truck
and is shown in operating configuration in l'ig. 3. The
basic system characteristics are listed in Table I.
It had been pointed out by F.vans* that receiver
paralysis caused by the intense plume signal return
would be a problem, and that this problem could be
avoided by off-gating the phuttimultiplier tube, in par-
ticular an Amperes ."»t> TVI* tube, by applying a voltage
pul-e to its focus electrode. Fur this reason, we in-
vestigated ,-everal u,;iliii^ cnulivmrutions ami two braml-
of phiitimiultiplicr tube (Amperex .*t» T\T ami l'l'.V
7'JIm), bolh in the laboratory and in the lidar sy.-lem, in
an attempt to limit tins effects of the intense plume rc-
197
-------
LIOAH RETURN FROM *
SMOKE PLUME
PLUME
RETURN
AMBIENT
AIR RETURN
tl» ja/T
UL3E
r s«c
r»cfp/2
Fig. 1. A sample hilar return through a smoke plume showing
the single beam, single *hot technique for obtaining the one-way
lidar determined transmittance, Tl.
' PULSE C€N. U—|
— (FC^ PMT i
GATEl i I
4*12 X AIMING SCOPE
• IN. 01AM., U 5 •
~-FIELD STOP
PM TUOC
OETECTOR
OCU3l£ 3CAM
5CCPC a S3
CAMERA
FROM LASER MONITOR
RUBY LASgft
LASCR PS
!
' r WATCH
— COOLING
| i system
9 IN. 01 AM., f/9"
-L
MOTOR
OEM.
tr- rH«iM4ixr controlled
intcmckcnci ritrtw
Fig. 2. Block diagram of mobile lidar system.
turn. When gating pulses were applied to tube ele-
ments near the pbotoeathode (external mesh grid, focus
electrode), they resulted in no reduction of nfterpulsing
and/or unacceptabiy slow tube on-responsc following
removal of the off-gate for off-gating pulsp widths of
practical interest (> 100 usee). The slow on-rcsponse
of the tube was suspected to be due to slow redistribu-
tion of charges in the semiconducting (rialkali photo-
catliode following the application of a large electric
field for long enough (>100 nsec) to have upset the
.lo-licld charge distribution.
In an effort to prevent large tube outputs from over-
driving the electronics and to try to reduce afterpulsing,
otT-gating pulses were simultaneously applied to dy-
nodes 2 and 6 of both the Amperex oG TVP and UCA
~'2G3 tubes. For both tubes this method of double
dvnode off-gating allowed an on-to-otT ratio of about 45
illi with good on-recovery response « j0 nsec), but it
had no effect on nt'terpulsing. This led to the conclu-
sion, supported by RCA," that afterpulsing originates
between the photooathode and the 'iring starting immedi-
ately after the plume return, while the Ampcrex 50
TVP showed two distinguishable afterpulsing peaks
r*
? i
fci/r n
i.v 4^/ri [ I
wtejafr -i
.V<>. I
Fig. 3. Mobile lidar system shown in operating configuration
at tlic Barbadoe* bland Station of tho Philadelphia Electric
Company.
Tabto I. Mobile Lidar Systom Characteristics
Transmitter
Laet-f
l.O-cm X 7.3-cm ruby
Q-switcli
Pouting prism
l'uUe width (FWJIII)
<30 usee
Max. output
1.0 J
Ilcpetitkm rate
3 pprn
Cooling
IVionizcd <111
Gating rt>p»iu
-------
GATED PLUME SIGNAL MAIN AFTER PULSE
LIDAR r
RETURNS
t.
GATE POS.
0.5 V/cm
0-05 V/cm
0.05 V/cm
0.5 /x scc/cm
Fig. 4. A lidar return from a white building wall at a raiiRP of
about 200 m. This >iiiiul:ito* a plume of zero tr:in»mittancc and
high reflect .nice and was obtained using an Amperes "16 TVP
pliutomultiplicr tube niih dyuodc gating.
Tablo II. Synthetic Targets Used for Lidar Evaluation
Target Material
Laboratory Transm. (09-13 A)
Plexiglas-Cl (.two)*
0.915,0.9tj
Gla»s(two)'-'
0.S01,O.SOl
A1 screens (two)
0.6-15, 0.040
A1 screen (black)
0.325
A1 screen
0.515
A1 screen
0.2SS
White paint
0
Black felt
0
—v.* mm in
* Paratlel-O-Float glas«.
11 lmi«j region of minimum afterpulsing between
_00 n>ec and cmO usee following the plume return. This
can best be seen in Fig. 4, obtained when a while build-
ing wail was used to simulate a plume of zero trans-
mit taticc and high reflectance. The main aftcrpulsc is
clearly visible on the upper, low gain trace, while the
lower, high gain trace shows immediate early afterpuls-
ing as the ofT-gating pulse is removed. It i.s clear that
since the minimum signal between those peak* was not
quite zero, a transmittance measurement error was in-
troduced. '1 lie double dynnde-gated Amperex tube
was chosen for Final system u>e, because its afterpulsing
characteristics allowed transmittance measurement* to
be madi- in u minimum afterpul.-ing region and because
the gating method provided last response.
Synthetic Target Tests
The lidar system wa-« evaluated using synthetic tar-
gets made nf wliilr-painted plywood, frit, aluminum
screen, black aimdizrd aluminum screen, glass, and
Plexiglas. The transmittance of these targets was de-
termined at the laser wavelength A) in the labora-
tory u-ing a rollimated white liylit source and the lidar
receiving ,-ystein. The targets and their transmittance
value- are -Ituwn in 1 able 11.
These targets were u-etl .-incly and in pairs at a test
range of 'ill m by jjlacing them in each of two target
holders having 1.07-m diain apertures. The target
holders were placed on top of a building with the lidar
optical axis centered in both apertures, while the holders
were spaced about 7 m apart along the path to simulate
plume depth. The holders were also tilted sufficiently
with respect to each other and the lidar axis to elimi-
nate specular reflective effects. The dear target aper-
ture subtended o.O? mrad at the target range assuring
>0.'.H) transmittanee for the O.'i-mrad rollimated laser
pulse. This was verified by taking lidar shots through
the clear aperture of both targets using gating at the
target range that vielded a measured transmittanee of
rL - i.oo.
Since the reflectance of the Plexiglas and the glass
targets was specular, account had to be taken, in princi-
ple, of the atmospheric light scattering return from the
reflected beams in making a comparison between
laboratory and lidar determined transmittance. Nu-
merical analysis, however, showed the error in omitting
this correction to be <0.4% -®o the data were not ad-
justed. Because of their olT-axis specular reflectance,
these targets were not expected to produce afterpuls-
ing, and in general they did not. Figure 5 shows the
lidar return and the reduced data from a double glass
target with a laboratory determined transmittance, T
«¦ 0.741. No afterpulsing is visible on the oscillogram,
and the data on both sides of the target indicate the
expected l/r* system performance, the sloped lines
having been constructed with a slope of —2. Further,
»T-r?-ir'-'f—;r~v "" '
. ft V. . K
2.0
tfl
0
>
1
< 1.0
z
<9
<
a
-j 0.5
ENERGY
MONITOR
LIDAR
RETURN
GATE POS.
H—i—r-
Tl» 0.757
TARGET
0.3
100
200
J I L.
400 600 1000
RANGE-METERS
l'ij* 5. Lidar shot through two ula.-* tartels (7' ~ 0.711) nt
21 l-m ruiiKC. Tliv oscillogram sweep j.pced is 0.."» p>cr/cm.
17<4 APPLIED OPTICS / Vol. 11. No. 8 / August 1972
199
-------
1.0
0.6
0 0.4
>
1
<£.
5
w 0.2
a:
<
o
0.1
0.06,
1
: \
, -1 1 1—»
B
\ :
-
^TARGET
\
Tl » v/aTF «0.431
\
\
i
\ :
' ribx ' ' ,U
100
L
200
RANGE - METERS
" 'lt .UL.i lENERGY
" MONITOR
— r "• lUDAR RETURN
J
GATE POS.
Fig. 6. Lidar shot through two screen targets (T - 0.412) at
'21 l-m range. The oscillogram s«*eep speed is 0.3 fisec/cm.
r i i | I
SYNTHETIC TARGET TEST RESULTS
211 METER RANGE
oNON-BLACK TARGET! GATING
o BLACK TARGET J USE0
~ NO GATING
< 0.8
d 0.6
5 AVG. DEVIATION
0 02 0.4 0.6 OS IjO
LABORATORY TRANSMITTANCE-T
Fig. 7. Averaged lidar determined target tranamittaace, Tt,
vs laboratory determined transmittance, T, with and without
oCf-gating.
the lidar clotcrmined value of transmittance, TL = 0.757,
is iti good Agreement with the laboratory value.
Figure G shows a lidar shot through two bright alumi-
num screen targets with high diffuse reflectance. Hero
the main at'terpulse is clearly visible on the oscillogram,
and evidence of the early, lower level afterpulsing is
clear in the lack of \/r"- performance on the far side of
the targets. The line, of *!opc —2, drawn through the
far side data points favors the points closest to the target
fur reasons of atmospheric homogeneity even though in
this case it yields too high a transmittance. Experience
with real plumes, in which thermal lensing occasionally
causes deviations from l.'r1 far-side performance, has
indicated that this choice of data weighting is eonsi*.
ti j«tly best. In any ca*«\ for the extreme example
shown in Fig. f>. choosing the smallest possible value for
.1 still yields a tran-mittance, 1\ =* 0.-10. t
The results of the svnthetic target testing are pre-
sented in Fig. 7 as lidar determined tran>mittance, ftl
v.s laboratory determined transmittance, 7. I'.aeh
data point shown represents the average of two to four
lidar shots except lor the points at / * 0 and 7 a 1.00
that were single shots, ll is clear that off-gating im-
proved the accuracy of the results, particularly for the
kPLUME\
©I© SKY
c«
T»,
Up
©
Up" *hp
• »- 'h
STACK
HILLSIDE
POWER PLANT
Fig. S. The teU'photoniMer toolminue of meaMiring plumr-to.
tky contrnrt, C, anil pin mo . r:>n»mitt:itiiv, by mi»f»ri>on of
the of xlir circled nro:v» shown.
200 Augutt 197? / Vol. 11, Mo.» / APPLIED OPTICS 1745
-------
LIDAR |
RETURN
MOW.
GATE
POS.
<
z
u
CO
a:
<
o
0.20
0.10
0.07
0.05
0.03
\
. Tl» SaTb « 070 >
: \
—| f
®^L
: ^
¦ L f . . i
\ :
t i i i. i _
100 200 400 600
RANGE-METERS
1000
Fig. 9. Lid.ir shot through a power station plume; Fort .Martin
Station, Allegheny Power and Light Co., Point Marion, Pennsvl-
^,'?.C|>tc"1'>or '2" '• This is a coal fired plume operating
a ° J " w'''11lidar range of 4S7 m. The oscillogram sweep
speed is 0.5 /isec/cm.
low transmit tanee targets where the target return sig-
nal was large. Further. even when off-gating was \isi-cl,
tin* error in transmit tnnri' increased as tlu* target trans-
mittance decreased, alt lit miilt at a given transmittance
the error decreased for a decrease in diffuse reflectauro,
This residual error is due to aftcrpulsing that the otT-
gtvtini? did not prevent. The lidar data for T >0.70
agree quite well with the laboratory transmittance val-
ues with the exception of one point at T = O.SoT that
was obtained using two Plexiglas targets. These tar-
gets had been cleaned repeatedly, and it is thought that
their relatively soft surfaces may have become suili-
cientlv scratched to give a larger than expected dilTu-e
component of reflectance. Slight aftcrpulsing was
visible on these oscillograms although not on any of the
other glass or single Plexiglas target results. The target
tests, then, indicate that lidar measurements always give
an upper limit transmittance value with an absolute
error that decreases as (a) target transmittance in-
creases and (b) target diffuse reflectance decreases.
Field Tests on Real Plumes
The lidar system was evaluated in western Pennsyl-
vania and West Virginia using coal burning power plant
smoke plumes at three Allegheny Power and Light Com-
pany stations: Albright Station at Albright, West
Virginia; Fort Martin Station at Point Marion, Penn-
sylvania; and Western Penn Station at Springdale,
Pennsylvania. Separate evaluations of plume-to-sky
contrast, C, and plume transmittance, T„ were made
with a telcphotometer as shown in Fig. 8. The tele-
photometer transmittance technique1 provides an ab-
solute measurement of transmittance but can only be
used where the local plume-hill-sky geometry allows.
For these comparison measurements the tclcphotom-
cter's spectral response was centered at 6510 A with a
GOO-A bandpass (FWHH).
Two of the power stations, Albright and Western
1.00
0.80
ui
o
z
?
t
z
in
z
<
X
H
U
s
3
0.60
0.40
0.20
O LIDAR - TL
aTELEPHOTOMETER-Tt
* TELEPHOTOMETER-C
k
0.40
-"•2
-#3 —
140 MW
-*3 —
100 MW
0.50
u
h-
V>
<
K
030 o
o
0.20?
bl
2
0.10 =>
1300
1400
1500
-0.10
Fig. 10. Lidar and telcpho-
tometer determined plume traiervation time ami
power plant ^tack operating
load. Data taken at the Al-
bright Station, Allegheny l'mvcr
and Light Co., Albright, WcM
Virginia, IS September I'JTI.
Tin* lidar range wa-> .100 m.
TIME - HRS.
1746 APPLIED OPTICS / Vol. 11. No. 8 / August 1972
201
-------
hi
U
<
I-
I OOr
0.8QL
0.60 r
s
in
5 0.40-
2 |
tii i
i 020 r
OUOAR-TL
00 » telephotometer-c
°A
s
/
A/V
#2
~535MW~
#1
480MW
_u
^2
"3G0MW
qo.30
¦40.20
040
»-
t/>
«
c
»-
z
o
u
>-
*
If)
0.10 o
toJ
a
3
* 400MW
1100
1200 1400
TIME-HOURS
1500
-0.10
Pig. 11. Lidar dctoruuu«s|
plume transmit tance niici trie-
photometer determined pltinus.
to-sky conlr.vt v* otiacrvatiou
time and power plant «tack
operating lonU. D*ta taken .11
Fort Martin Station, Alleglu>tiy
Tower nnd Light Co., Point
Marion, Peiiiwylv.viin, 10 Sep-
tember 107). The lidar rang*
wu 4S7 m.
Pcnn, were chosen because of favorable plume-hill-sky
geometry for the telephotometer transmittancc mea-
surement.-', while the Fort Martin'Station was chosen
because it is a relatively new plant with good particulate
collection efficiency and high transmittance plumes.
The power company had agreed to change the operating
load on individual stacks to vary the plume traadmit-
tance. Although this was done, only a limited amount
of data were taken because oL uniformly bad weather /or
all three te.it days.
An example of a lidar .shot through a real plume is
shown as l-*ig. 9. The system performance is 1 /r* on
both sides of the plume, and no afterpulsing is visible
as might be expected from .-such a high transmittancc
plume. In a limited number of cast* of lidar shot."
through high transmittance plumes, somewhat steeper
than l/i " performance was observed on the fur side of
the plume in the absence of any afterpulsing. This is
believed to have been due to thermal lensing by the hot
plume gases that caused some of the laser pulse to di-
verge beyond the field of view of the receiving system.
It is implicit in this argument that the field of view of
the receiver is much larger than the collimated laser
divergence sucli that the two see different mean portions
of the plume lens thus making differential defocusing
possible. As mentioned earlier, it is for this reason, in
addition to atmospheric homogeneity consideration*,
that data points closest to the plume arc most heavily
weighted..
The results of the field tests arc shown in Figs. 10-12.
It can be seen that no general correlation exists between
plumc-to-sky contrast and plume transmittance ar.d
that, within the accuracy dictated by the synthetic tar-
get tests, there was good agreement between tlu« two
transmittance techniques. Specifically, the lack of cor-
relation between C and T is best illustrated by the last
I.OOr
o.eor
-1O.50
UJ
u
2
2
t
s
-
£
H0.30
H0.20
u
X
3
Fig. 12. Lidar and telepho-
tometer determined plume trsin»-
mitt mice ami tclepltoiomcter
determined pluroe-Uii>ky ewi»-
lrn>l v.x uhiicrviuiim time. I>at»
taken at the Western lVnu Sta-
tion, Allegheny Power ami l.inlit
Co., SpriiiR«l:»lP, l,eim*>iv:wi.i.
2U Septrinlwir 1971. The Hilar
ruiigc wiM 301) m.
TIME (HRS)
August 1972 / Vol. 11. No. t / APPLIED OPTICS 1747
-------
in.-I un period in I'iu. 11- H'-re the lighting con-
di:i i!i. v.rri' i rapidly while tIn- plume trans-
mit tanec remained e-setitiaily eon-taut. leading at one
point ii• an invisible plume with a t r:m.~mit tniifi* of only
O.I'pD. Tii1' agreement between the two tnmsmirtunce
t< i* i)K>~t evident from Fig. 12 where tho
plum'' iraiiHiiittanci* was large enough that the error in
Tl should have hiv n negligible. Thr good agreement
is particularly apparent after 1400 h where* there was
reasonable data overlap. Tin* fluctuations in T, were
real and lvprc-cnted variations caused by rapping of the
electrostatic precipitators that allowed additional par-
ticulate matter to go up to the stack.
Conclusions
The synthetic target testing demonstrated the ac-
curacy of the single beam, single shot lidar technique
for the remote single ended measurement of plume
transmittance anil showed the major accuracy limita-
tion to be due to photoinultiplier tube aftorpulsing.
This caused a maximum systematic error (assuming
maximum plume reflectance at a given transmittance)
of < + 12% for T > O.oO and <2.5% for T > 0.S0.
This error decreased as plume reflectance decreased at a
given,transmittance and should, in principle, vanish for
a black (zero reflectance) plume. Since any error is
always positive, the lidar measurement inherently
yields an upper limit value for the transmittance.
It is clear that plumc-to-sky contrast (visibility) can
hi* correlated with plume transmittance only under con-
stant plume illumination conditions that in general are
not the case. This makes the lidar technique attrac-
tive since it provides an absolute, remote measure of
plume transmittance that requires only a clear optical
path to the plume and is independent of ambient light-
ing conditions.
Finally, field operation of the lidar system using real
smoke plume targets has demonstrated agreement,
within the present accuracy of the lidar technique, be-
tween simultaneous lidar and telcphotometer transmit-
tance measurements.
The work upon which this publication is based was
performed under Edison Electric Institute Research
Project 39 and contract GS-02-00P3 of the Environ-
mental Protection Agency.
References
1. It. T. H. Colli*, Appl. Opt. 9.17S2 (1970).
2. V. 1£. Derrand C. G. Little, Appl. Opt. 9, 1970 (1970).
3. P. A- Davis, Appl. Opt. 8, 2009 (I960).
4. W. 11. Johnson, Jr., J. Appl. Metcorol. 8,443 (1909).
5. \V. E. Evans, Final Report, Stanford Research Institute
Project 6529 (1967).
6. RCA Photomultiplicr Manual, Tech. Scries PT-G1, 47 (1970).
7. 11. Shaffer, RCA Lancaster, Pa., Private communication.
8. W. D. Conner and J. R. Ilodkinson, Public Health Service
Publicatiou 999-AP-30(19G7). ¦
203
-------
60
Title 40—Protection of Environment
CHAPTER I—ENVIRONMENTAL
PROTECTION AGENCY
(FRZ.71fr-8)
PART 60—STANDARDS OP PERFORM-
ANCE FOR NEW STATIONARY SOURCES
Compliance With Standards and
Maintenance Requirement*
AGENCY: Environmental Protection
Agency.
ACTION: Pinal rule.
SUMMARY: This action amends the
general provisions of the standards of
performance to allow methods other
than Reference Method 9 to be used as a
means of measuring plume opacity. The
Environmental Protection Agency (EPA)
is investigating a remote sensing laser
radar system of measuring plume opacity
and believes it oouid be considered as an
alternative method to Reference Method
9. This amendment would allow EPA to
propose such systems as alternative
methods in the future.
EFFECTIVE DATE: June 23,1977.
FOR FURTHER INFORMATION CON-
TACT:
Don R. Goodwin, Emission Standards
and Engineering Division, Environ-
mental Protection Agency, Research
Triangle Park, North Carolina 277 It,
telephone no. 919-688-81M, ext. 371.
SUPPLEMENTARY INFORMATION:
As originally expressed, 40 CFR 60.11(b)
permitted the use of Reference Method 9
exclusively for determining whether a
source complied with an applicable
opacity standard. By this action, EPA
amends i 60.11(b) so that alternative
methods approved by the Administrator
¦*y be used to determine opacity.
When 160.11(b) was originally pro-
mulgated, the visible emissions (Method
9) technique of determining plume
opacity with trained visible emission ob-
servers was the only expedient and accu-
rate method available to enforcement
personnel. Recently, EPA funded the de-
velopment of a remote sensing laser ra-
dar system (LIDAR) that appears to pro-
duce results adequate for determination
of compliance with opacity standards.
XPA to currently evaluating the equip-
ment and is considering proposing Its
ue u an alternative technique of meas-
uring plume opacity.
This amendment will allow EPA to
use of the LIDAR method of
determining plume opacity and, if ap-
propriate, to approve this method for en-
forcement of opacity regulations. If this
method appears to be a suitable alterna-
tive to Method 9, it will be proposed In
the Txbzkal Regis tir for public com-
ment. After considering comments, EPA
wQl determine if the new method will be
an acceptable means of determining
opacity compliance.
(Sew. 111. 114. S01 (a). Clean Air Act, mc. 4(a)
ef Pub. L. Bl-604. 84 Stat. 1663; sec. 4(a) or
rub. L. 91-404, 84 Stat. 1*87: me. J at Fob. L.
We. 10-148, 81 Stat 804 (48 UJS.C. l88?e-«.
18S7o-« and 1881g(a)).)
Worm.—loonomlc Impact Analysts: The
environmental Protection Agaoey baa deter-
mined that tbla action does not contain a
major proposal requiring preparation of an
¦eonomle Impact Analyse ondcr Kseeuttve
Orden 11831 and 11948 and OMB Circular
A-107.
Dated: May 10,1977.
Dottqlas M. Cos tlx,
Administrator.
Part 60 of Chapter I. Title 40 of the
Code of Federal Regulations is ammrted
•s follows:
L Section 60.11 Is amended by revising
paragraph (b) as follows:
§ 60.11 CompHiiy with standards n4
maintenance requirement*.
• ••••
(b) Compliance with opacity stand-
irii tn this part shall be determined fey
conducting observations in accordance
with Reference Method 9 in Appendix A
of this part or any alternative method
that is approved by the Administrator.
Opacity readings of portions at plumes
which contain condensed, uncombined
water vapor shall not be used for pur-
poses of determining compliance with
opacity standards. The results of ooo-
tinuous monitoring by transmlssometer
which indicate that the opacity at the
time visual observations were made was
not in excess of the standard are proba-
tive but not conclusive evidence of the
actual opacity of an emission, provided
that the source shall meet the burden of
proving that the Instrument used meets
(at the time of the alleged violation)
Performance Specification 1 In Appendix
B of this part, has been properly main-
tained and (at the time of the alleged
violation) calibrated, and that the
resulting data have not been tampered
with In any way.
(S«c*. ill, 114. 801(a), Clean Air Act, gee. 4
(a) of Pub. L. 91-604. 84 Stat. 1888; sse. 4(a)
of Pub. L.01-804, 84Stat. 1687; He.IotPub.
L. No. 80-148 81 Stat. 804 (43 O.S.C. 1887C-6,
1887C-8.1887«(a)).)
(Fit Doe.77-14869 FUed (-30-17:8; 41 sal
ruaui Mwn, vol «, no. n mommy. auv ta. itw
205
-------
SECTION G: VISIBLE EMISSION TRAINING CERTIFICATION
Page No.
G-l. Smoke School Evaluation Checklist, R. Missen, 1977 209
G-2. "Factors Affecting Accuracy in Visual Emissions Evaluations,"
R. E. Hague et al. , 1 977 223
G-3. Affidavits of Visible Emission Course Attendees Describing
Experiences at Training Courses, 1974 235
207
-------
G-l
SMOKE SCHOOL EVALUATION
CHECKLIST*
~Draft of evaluation guide to appear in EPA Quality Assurance Manual for
Reference Method 9 (in preparation).
209
-------
1. TRAINING SCHOOL LECTURE COURSE
No.
Item
Remarks
1.1 I Description of Ringelmann Chart. Defi-
nition of equivalent opacity.
1.2 | VE aids
1.3 I Description of training procedures
1.4 I Description of smoke generator
1.5 | Basic meteorology; light scattering
1.6 | Field procedures
1.7 | Wet plumes
1.8 | Testimony procedures
210
-------
2. EVALUATING PROCEDURES
Ho.
Item
Remarks
Are students told to read:
2.1
With Sun within 140° sector behind the
observer
2.2
Perpendicular to the wind direction
2.3
Through the densest part of the plume
2.4
With a contrasting background 1f possibl
e
2.5
When the appropriate signal Is made
2.6
Not to stare at the plume
2.7
Does the agency certify observers under
nighttime conditions
)
\
i
i
\
i
211
-------
3. QUALIFICATION PROCEDURES
No.
Item
3.1 What do operators do when smoke
doubles over
3.2 How 1s wind speed determined
3.3 How is wind direction determined
3.A Record name
3.5 Record affiliation
3.6 Record date
3.7 Record time of run
3.8 Record sky condition
3.9 Record wind speed
3.10 Record wind direction
3.11 Record observer's location
3.12 Which units are used
3.13 Do observers read to nearest 5%
3.14 Who checks readings against trans-
missometer
3.15 Is Black-White and White-Black accepted
3.16 Do students complete all parts of form
3.17 Do students double check transmisso-
meter readings after qualifying
3.18 Do students check arithmetic
3.19 Does agency check transmissometer
values on form
Remarks
212
-------
3. QUALIFICATION PROCEDURES (continued)
No.
Item
Remarks
3.20
Does agency check arithmetic
3.21
Does agency send written affirmation
3.22
Does affirmation include expiration
date
3.23
Does agency retain forms on file
3.24
Does agency retain transmissometer
trace
3.25
Does agency retain transmissometer
readings
3.26
Is plume attached or detached
213
-------
4. SMOKE GENERATOR
No.
Item
Remarks
4.1
Who made the smoke generator
4.2
Model type
4.3
Any modifications made to 1t
4.4
Was operating manual supplied
4.5
Does agency follow procedures in
operating manual
4.6
When received
4.7
Any problems
4.8
What are solutions to these problems
4.9
Who made transmlssometer
4.10
Model type
4.11
Type of photocell
4.12
When received
4.13
Any modifications made to 1t
4.14
Was operating manual supplied
4.15
Do they follow procedures 1n
operating manual
4.16
Any problems
4.17
What are solutions to these problems
4.18
Are calibration and maintenance pro-
cedures followed
4.19
Type of fuel used for black smoke
214
-------
4. SMOKE GENERATOR (continued)
No.
Item
Remarks
4.20
What is the pathlength of the trans-
mi ssometer
4.21
Can opacities from 0 to 100% be
obtained
4.22
Does the agency make 5 non-consecutive
readings for each filter
4.23
Is the maximum error %3% opacity?
4.24
What is zero and span drift
4.25
Is trace checked for noise after
each run
215
-------
5. FIELD PROCEDURES
No.
Item
Does agency describe
5.1
Equipment needed
5.2
Pre-entry procedures
5.3
Entry procedures
5.4
Evaluation form
5.5
Inspection report
5.6
Available publications
Remarks
216
-------
METHOD 9 REQUIREMENTS
METHOD 9
SECTION
ITEM
REMARKS
2.
PROCEDURES
2.1
POSITION
Are observations made:
With a clear view of emissions
With the sun oriented in the 140°
sector behind the observer
Such that the line of vision is approxi-
mately perpendicular to the plume
direction
For non-circular stacks approximately
perpendicular to the longer axis of
the outlet
Such that the line of sight does not
include more than one plume.
2.2
FIELD RECORDS
Does the observer record the following
on a field data sheet:
Plant name
Emission location
Type of facility
Observer's name and affiliation
Date
Are the following items recorded at
the beginning and at the end of the
opacity evaluation:
Time
Estimated distance to the emission
location
Approximate wind direction
Estimated wind speed
Sky condition (presence and color of
clouds)
Plume background
217
-------
METHOD 9 REQUIREMENTS
METHOD 9
SECTION
ITEM
2.3
2.3.1
2.3.2
2.4
2.5
OBSERVATIONS
Are observations made at point of great-
est opacity where no condensed water
vapor is present
Is plume observed momentarily at 15-
second intervals
ATTACHED STEAM PLUMES
Are observations made beyond breakpoint
Is approximate breakpoint distance
recorded
DETACHED STEAM PLUME
Are observations made prior to formation
of steam plume
RECORDING OBSERVATIONS
Are observations recorded to nearest 5%
Are observations made at 15-second
intervals
Is a minimum of 24 observations recorded
DATA REDUCTION
Are the observations divided Into sets
of 24 consecutive observations
Are the sets mutually exclusive
Is the average opacity determined for
each set
Is the average opacity recorded on data
sheet
Does applicable standard specify averag-
ing time requiring more than 24
observations.
If so, is the average opacity determined
for the specified time period.
REMARKS
218
-------
METHOD 9 REQUIREMENTS
METHOD 9
SECTION
3.
3.1
3.2
3.3
ITEM
QUALIFICATIONS AND TESTING
CERTIFICATION REQUIREMENTS
25 black smoke readings
25 white smoke readings
5% increments
Maximum error 15% on one reading
Maximum average error <7.5%
CERTIFICATION PROCEDURE
Are candidates shown a complete run of
50 plumes -- 25 black and 25 white —
at one time
Is opacity selected randomly
Are readings recorded on a suitable form
Is black run made first
Is complete 50 readings test repeated
if observer fails to qualify
Are familiarization runs given
SMOKE GENERATOR SPECIFICATION
Is opacity based on pathlength equal to
exit diameter
Do readings range from 0 to 100%
Does smoke meter meet the specifications
shown 1n Table 9-1 (See Section 3.3.2
Smoke Meter Evaluation).
Is the smoke meter calibrated according
to Section 3.3.1 prior to the beginn-
ing of each test
At the completion of each test is the
zero and span drift checked
Is the drift less than + 1 percent
opacity
If out of spec, is the situation cor-
rected prior to conducting any subse-
quent test runs
REMARKS
219
-------
METHOD 9 REQUIREMENTS
METHOD 9
SECTION
3.3
(cont'd)
3.3.1
3.3.2
3.3.2.1
3.3.2.2
3.3.2.3
3.3.2.4
ITEM
Did the smoke meter meet the Table 9-1
specifications at time of installation
Is the meter rechecked for compliance
with Table 9-1 requirements (details
for determining compliance with Table
.9-1 requirements are listed under
Section 3.3.2 - Smoke Meter Evaluation
following any subsequent repair or re-
placement of the photocell or associa-
ted electronic circuitry Including
chart recorder or output meter or
every six months, whichever occurs
first.
CALIBRATION
Is warmup time 30 minutes
Is 0% and 100% opacity obtained by
switching power to light on and off
SMOKE METER EVALUATION
LIGHT SOURCE
Is lamp operated within + 5% of nominal
rated voltage
Type of lamp (manufacturer, voltage,
model no.)
SPECTRAL RESPONSE OF PHOTOCELL
Does photocell have acceptable photoplc
response
ANGLE OF VIEW
Is angle of view si5°
0- tan
d = photocell diam. + limiting
aperture diam.
L 3 distance from photocell to
limiting aperture
ANGLE OF PROJECTION
Is angle of projection <15°,
6= tan"^ |j-
-1
d
2L
REMARKS
220
-------
METHOD 9 REQUIREMENTS
METHOD 9
SECTION
ITEM
REMARKS
3.3.2.4
(cont'd)
d = length of lamp filament
+ limiting aperture diam.
L = distance from lamp to limiting
aperture
3.3.2.5
CALIBRATION ERROR
Is calibration attained at 20%
Is calibration attained at 50%
Is calibration attained at 75%
Are neutral density filters calibrated
to + 2%
Are five non-consecutive readings taken
with each filter
Is the maximum error + 3% on any one
reading
Is calibration performed before each
school
3.3.2.6
ZERO AND SPAN DRIFT
Is the zero and span drift checked after
the smoke generator has been operated
in a normal manner for at least one
hour
Is the zero and span drift checked over
a 30-minute period and after at least
an hour of normal operation
Is this zero and span drift less than +
1 percent opacity
3.3.2.7
RESPONSE TIME
Using five simulated 0 percent to 100
percent opacity values, is the time
required to reach a stable response
less than 5 seconds.
221
-------
6-2
Prepared for Presentation at the
Air Pollution Control Association
National Conference
June 20-25, 1977
Toronto, Canada
FACTORS AFFECTING ACCURACY IN
VISUAL EMISSIONS EVALUATION-'
R, E, Hague, D. A. Deieso and F.B. Flower
Department of Environmental Science,
Cook College, Rutgers University
P.O. Box 231, New Brunswick, New Jersey 08903
ABSTRACT
The visual determination of smoke plume opacities has
long been used by all levels of government as a method of
air pollution control and enforcement. Present Federal
certification regulations require recertification training
every six months, with the assumption that this provides
an adequate safety margin for reader accuracy. This paper
investigates the ability of observers to read accurately at
the end of this six-month period. Trained observers employed
by both public agencies arid private industry are tested for
opacity-reading ability on both black and white plumes before
any retraining takes place. Accuracy is estimated by
determining both the average error and the frequency with
which an observer makes large (greater than 15%) errors in
opacity estimation. Factors including age, number of years
experience with the method, frequency of method use and the
use of glasses or contact lenses are evaluated statistically*
223
-------
INTRODUCTION
The use of traine'd observers has long heen accepted in
Che United States and elsewhere as a useful and sufficiently
accurate means of determining smoke emission density, In^
deed for many years, smoke ordinances were essentially the
only air pollution control laws available. Although other
methods are now available for determining plume opacity,
these are limited in their usefulness by their high cost and
difficulty of operation, at least at the present time. As
a result, enforcement of opacity regulations is very often
based on the judgement of one or more trained observers.
Studies (3,5)have been made of observer accuracy and precision
and the method determined by the United States Environmental
Protection Agency to be acceptably accurate for use in
enforcement of regulations. There have not been, however,
any studies to the present time of the effect on reading
accuracy of such variables as the age of the observer, his
experience with the method or the frequency with which he is
required to use the method. Since very often it may be up
to one individual inspector to determine whether a given
source is in violation of a regulation, such factors may
have a very real impact if they are indeed significant.
The present report represents an evaluation of some
of these effects, based on data collected in 1976-77 from
the visual emissions training program run for the state of
New Jersey by Rutgers University.
Background
It has been more than 75 years since Ringelmann devised
the grid chart bearing his name which provided an easily
reproduceable comparator for shades of gray smoke. The
practice of defining the opacity of a given plume on the
basis of its Ringelmann Number has been very widely accepted
legally. In the 1940's the concept of "equivalent opacity"
was developed for non-black plumes. This was defined as
the density of smoke or dust which obscures light to a degree
equivalent to a given Ringelmann Number and has achieved the
same degree of acceptance as the older standard.
Regulations covering the visual determination of
plume opacity are contained in Federal Reference Method 9.
(9). These regulations, cover both the use of this procedure
in the field and the training and certification of observers,
and were last revised in November, 1974. To be certified
under Method 9 as a qualified observer;
"a candidate must bo tested and
demonstrate the ability to assign
224
-------
opacity readings in 5 percent
increments to 25 different black
plumes and 25 different white
plumes, with, an error not to
exceed 15 percent opacity on
any one reading and an average
error' not to exceed 7.5 percent
opacity in each category....
The certification shall be valid
for' a pei*iod of six months , at
which time the procedure must be
repeated by any observer in order
to retain certification."(9 )
The provision here of a requirement for recertification
every six months is significant. It would seem to imply
that the majority of the users can maintain an acceptable
level of accuracy over this entire period, or, to put it in
instrumental terms, that there is negligible or at least
tolerable drift away from the accuracy displayed immediately
after calibration. It is, of course, true that in any
method of measurement using one of the senses that there
is necessarily an element of subjectivity. The ability of a
given individual to read opacity accurately may vary some-
what from month to month or even from day to day, but the
assumption -used in training and certification is that the
observer will read at close to the same accuracy throughout
the six-month period. To evaluate this assumption, 100
individuals were tested for accuracy before retraining took
place, and the resulting test scores related to other data
pertinent to the individuals use oE Method 9.
Testing Procedure
The subjects, all experienced in the use of Method 9,
were tested for opacity-reading ability using a mobile smoke
generating machine. Black or gray plumes were generated
by the combustion of xylene under limited air supply conditions
and non-black plumes by the condensation of vaporized diesel
fuel. The stack height of the smoke generator is approximately
16 feet with a diameter of one foot.
The test site was an open field of sufficient size to
allow free movement of the observers around the stack, so
as to best adjust for sun position, contrasting background
and wind direction. The backgrounds used by the observers
were open sky Cfor black smoke) and either open blue sky or
a dark tree line about 150 feet beyond the test site for
white smoke. The observers were allowed to select their own
position for reading, within the constraints imposed by
Method 9, that is, a minimum observer - stack distance of
two stack heightswith sun in the 140 degree angle to the
observer's back and at right angles to the direction of plume
travel.
225
-------
As specified in fteth.Qd teat, congested of a set of 25
black and a set of 25 ^h.ice plumes giyen in r^ndoja order.
Testing of the individuals was performed "cold'1, th.at is
with no prior recalibration before either set, The runs
were administered as the first test of the day to minimize
eye fatigue, a factor which can have a great effect on
reader accuracy.
The sample population represented a cross-section
of the types of individuals using Method 9. Of the 100
test subjects, 45 are employees of private industry and
55 work for a state, regional or municipal ai.r pollution
control agency. The observers' ages ranged from 21 to 70,
with the mean age being 39.6 years.
Test scores for black and white smoke for each
individual were correlated with four variables: the
subject's age, whether or not he required glasses or
other vision correction and two variables which are referred
to in the data as "use" and "experience". "Use" was
defined as the average number of times per month a given
individual made use of Method 9 in his work, while "experience"
was defined as the total number of years which the individual
has used the technique.
Since it is possible under Method 9 to fail a given
certification test run by either exceeding the allowed 7.5%
average error or by having one or more individual readings
with an error exceeding 15%, two additional terms, "adjusted
black score" (BSCA) and "adjusted white score" (WSCA) were
developed. These terms added a variable amount to the total
black or white test score, which was directly proportional
to the number of times in a given set the observer made a
greater than 15% error. This greatly facilitated the
comparison of the scores of those who passed with those who
failed for one or both of the above-listed reasons.
Data Analysis
Employing a packaged statistical computer program
(Statistical Analysis System, 1972) regression analysis and
discriminant analysis were performed on the data. The
variables; age. experience, glasses and use were tested
against an observed response, i.e. the adjusted score.
In regression analysis, the contribution of each variable
to the total variation of observed score was provided. The
results of the discriminant analysis were not revealing
and consequently this information has not been presented.
It is worthy to note that future investigation of these
factors will require multi-discriminant analyses, a procedure
not readily accessible at the time this study was done.
226
-------
Res ults
The finding of perhaps the greatest' significance
was the pass->fail rate among those sampled. Of the 100
observers, 58 passed on th.e run using black and gray
plumes, and 42 failed. On white plumes, 46 passed and 54
failed. Only 28 were able to pass both at the same time.
An objective of the statistical analysis was to
provide insight into the relative importance of the variables,
i.e. age, use, experience and corrected vision. Although
the correlation coefficients were very low, some valid
conclusions may still be drawn, at least to the relative
importance of each of the variables. It is strongly suspected
that many other factors not considered and difficult to quantify
are responsible for the poor correlation. Each variable will
be discussed individually.
Age
It became obvious from the analysis that the age of
the observer seems to have little or no effect on accuracy.
Correlation coefficients were extremely low in all cases .
Vision Correction
As with age, glasses and contact lenses were shown
to be weakly correlated to scores and no evidence was found
that such correction had any effect on ability to read
smoke plumes.
Frequency of Use
An interesting result of the analysis was that
frequency of use seemed of relative importance in the reading
of white smoke and showed some correlation with adjusted
score, while, oddly enough, the correlation between use
and black smoke test score was rather poor. A comparison
of the plots of the white and black test scores versus
use illustrate this as well, (Figure I and II). As. would
be expected, both diagrams indicate a definite decrease in
test score with increasing use, i.e. accuracy is greatly
dependent on frequent practice.
Exp erience
Somewhat unexpectedly an examination of the correlation
coefficients indicated that this variable seemed to show a
relationship the opposite of that observed for frequency of
use. That is, years experience with Method 9 seemed more
likely to indicate accuracy in reading black plumes than
white plumes.
227
-------
Again, an examination of the plots of black and white adjusted
scores versus expedience giv© some evidence of this, (.figures
III and I,y) , Again, as in the previous case, there is a
definite trend to the relationship between high score Clow
accuracy) and short experience.
S ummary
The data obtained from previously trained observers
with regard to reader accuracy seems to be significant.
The ability of the test subjects to read accurately within
Method 9 guidelines after a six-month period is poor. For
both black and white plumes only about half of those tested
were able to pass a 25-reading test, in spite of fairly
good viewing conditions on all test dates. As might be
expected, the group as a whole did slightly better reading
black smoke plumes than on white plumes.
The low degree of correlation between the chosen
variables and test scores indicate that some other factors
may be of greater effect than those considered. Recent
studies CBaist, et al 1976), (Baist, 1977) refer to the
importance, especially for non-black plumes, of such factors
as solar elevation angle, and these may certainly have some
bearing on the matter.
What seems to emerge from the data collected thus
far seems to be that the present six-month interval
between training sessions may be too long for maximum
possible accuracy using Method 9, and that perhaps this
provision should be reassessed.
228
-------
REFERENCES
1. CONNER, W.D. and J. R. Hodlcingon, Optical Properties
and Visual Effects of Smoke Stack Plumes.
U, S. Environmental Protection Agency Publication
AP-30, (1972).
2. HALOW, J.S. and S. J. Zeek, Predicting Ringelmann
Number and Optical Characteristics of Plumes.
Jour. Air Pol. Cont. Assoc., 23,8 pg. 677 (1973).
3. HAMIL, K.F., R. E. Thomas and N.F. Swynenton, Evaluation
and Collaborative Study for Visual Determination
of Opacity of Emissions from Stationary Sources.
EPA Report No. EPA-650/4-75-009 , EPA Office of
Research and Development. (1975) .
4. GOODWIN, D.R., Opacity-a federal view. Env. Sci. and
Tech . , 11,1. Pg. 11 (19 77)
5. HOWES, J.E., R. N. Pesut, and J. F. Foster, Final Report
on Interlaboratory Cooperative Study of the
Precision and Accuracy of ASTM D 3211-73T, Method
for The Determination of the Relative Density of
Black Smoke (Ringlemann Method) to ASTM,
(January 18, 1974).
6. WEIR, JR.,A., D. G. Jones, L.T. Papay, S. Calvert and
S.C. Yung, Factors Influencing Plume Opacity.
Env. Sci. and Tech., 11,6 pp. 561-563
7. WEIR, JR.,A, Clearing the Opacity Issue. Env. Sci. and
Tech., 11,6 pp. 561-563
8. SER.VICE, J., Users Guide to the Statistical Analysis
Systern. North Carolina State Univ., Raleigh, N.C.
(1972).
9. U.S. ENVIRONMENTAL PROTECTION AGENCY, "Stationary Sources
Emission Monitoring and Performance Testing
Requirements", Federal Register, Vol. 39, No. 177
November 11, 1974, pp. 32852-82874.
229
-------
PLOT..OF.. WSCA..VS _USE
UOOoOrif.u
A
7U.ll00ulM.lh) *
1
« .
!«
h
v. si: a i
A
1 »
1 A
'•¦j.OKiCiOtiUU ~
1
A
1
AA
A
A A
A
"a
A •
1
A
A
1
1
A H
1
l» f
\
C
,-o.OGOoniiun ~
1
H *
A
A
A
1
r
H A
A
A
ft
A
rt
1
1
A *
•< r
r
H
H
nh 4 !1 C C B ; A A A
»S.U
-------
--PLOT-OF-B-S CA—VS—US E
I
i_
'»{>. CPOOfiPili) •
I
i
h
h
A
A
i
f
U.OlHtHlfii'H •
1
---
FIGURE
II
-------
I Mil III -.MA vs t AI'
'finiliir,in
— -
, PLOT..OF WSCA.. VS. EXP _
iillliliK-im
¦ - -- • - -
.... _. .. . .
IIOO'IOIM'II
A 4 t«
11
A
1 ».
1
If. (Ill (I<1 Ml
•* _
/.
•«
c; A H
A
ii
. irn"«vfifif:.rr.'
i- A t* A A
ft
«\ a .i • A 4
III. ^ A A A
1 - AH
A A
f (. .» 1. C •* A 0 A A
i* »i i; ii t t* h A A
. t|f*'i iiiIiuic
FIGVRE III
-------
H ill I It- v> r.AK
11.~ """KlfVuMni'iiDinr
-
-
•i
PLOT OF BSCA VS EXP
1
/.n.oonoonnu
A
A
. .
1 A
.»i..uon(inooo «
A
nsr/i
X <
A
1 A
/"•.O^OOOOilO
A
A
A
A
A
n
1
A
A
h H
II
A
C
A .
a
lr'.onnnonno
A
/¦
,1
A
~_
II
A
rl
» a
c
A A
rt H A A
II
a
A
A
•1
A 1) A A DA
n H A A
A
H
'
—
£IGURK_iy
-------
G--3 (a)
AFFIDAVIT OF MORTON L. MULLINS
ON OPACITY LIMITS AS EMISSIONS STANDARDS
City of Creve Coeur )
) ss. :
County of St. Louis, Missouri )
MORTON L. MULLINS, being duly sworn, deposes and states
as follows:
I hold a bachelor of arts degree in electrical engineering, which
I obtained in 1956 at the University of Missouri at Rolla, Missouri. I
have been in the employ of Monsanto for the past 14 years and for the past
three years I have been Manager of Environmental Protection and Utilities,
Monsanto Industrial Chemicals Company. During these years I have had
substantial experience with efforts to control air emissions from industrial
sources, including efforts to make visual measurements of opacity.
In my experience in the Environmental field, I have encountered
substantial difficulty with what I regard as wholly subjective and imprecise
efforls by pollution control officials to make visual measurements' of opacity
and to premise enforcement, action on such measurements. In order to
obtain a better understanding of these problem s, I attended an EPA Visible
Emissions course in Boston, Massachusetts, on September 12-14, 1973.
The purpose of the course was to qualify observers under EPA
Test Method 9 for visually measuring opacity. There were around 40 to 50
students in the course; the great majority oi them were pollution control
officials from federal, state and local governments.
The first day and one-half of the course consisted of classroom
instruction and discussion. On the afternoon of September 13, 1973, the
instructor, who said he had developed Tost Method 9 and who indicated he
ran all of the various EPA schools on Visible Emissions and had helped
set up many of the comparable state schools, showed the students series
of white and black plumes of known opacity emitted from a stack about
one foot in diameter and about fifteen feet high. According to the instructor,
the plumes varied in opacity from 0 to 100%, in increments of 5%. On the
same afternoon the students were given a practice test. To pass the test
one had to have an average error no greater than 7. 5% and no one error
tru'ro th.in 1^"! (i.o. , 20"', or more). Tht> test consisted of visually read-
ing 25 white smoke plumes and then 25 black smoke plumes. I passed
the practice test.
235
-------
- 2 -
On the morning of September 14, 1973, the students were shown
a number of plumes emitted from the same stack and were told of the
known opacity of each such plume. They were then given a test, as de-
scribed above. Many, and I believe most, of the students had difficulty
with the test and failed to pass. Those who failed and who wished to try
again were shown more plumes and were told of the known opacities of
each, and were promptly retested. Many of the students failed several
times; I took the test four times in about four hours, but did not pass it.
A number of the students left the school without having passed the test,
presumably returning to their enforcement positions.
The sky was overcast at the time of the testing on the 14th and
students were allowed to choose the position from which they observed the
plumes. Depending on positions chosen, the background would be sky,
trees, or buildings. During the testing, the variables that are important
in actual enforcement situations, such as stack diameter, exit velocity,
sky cover, lighting, smoke color, background, and particle size were
relatively constant; the wind, however, was variable.
The course included some classroom discussion of condensed
water vapor in plumes, but no such plumes were generated or used in the
testing. I have encountered widely varying understandings and interpreta-
tions of such plumes among enforcement personnel. For example, two
officials with the same enforcement agency, although agreeing that they
should measure opacity at the point where they thought the steam had dis-
sipated, disagreed on whether the reading at that point was to be used for
enforcement purposes; one said he "mentally corrected" the reading to
what it would have been at the stack exit or other point of discharge, which
could be hundreds of feet distant from the end of the steam plume. The
other official made no such correction.
In my opinion the conditions that prevailed for the testing at the
EPA Visible Emissions course were not representative of actual field con-
ditions. Those being tested had been repeatedly told of the known opacity
of the plumes being observed, and there was relatively little change among
the variable factors that affect the reading of opacity. That is quite different
from actual field conditions in which one is required to assess, months
later, widely varying plumes for which the physical circumstances diverge
substantially.
These variables are particularly important where opacity is
caused by acid mist, because there opacity is almost entirely a function
of Particle size. Two acid plants with ihb same pounds per hour of mist
can, because of difXcienccs in combustion turbulence, particle growth
236
-------
- 3 -
conditions, and type of collection used, have widely varying particle size
distribution. Concentrations of large particles would be almost invisible,
whereas an identical mass loading with small particles (less than one
microti) could bo opaque. Indeed, lor eve . a single plant, particle size
will vary (even chough the pounds per hour emitted remains constant) from
time to time due to different operating conditions.
In summary, it is my considered opinion that a reading of stack
opacity, even by a recently certifi. d Visible Fmissions Course graduate,
is very imprecise and subjective, and totally inappropriate evidence for
enforcement purposes.
Further affiant saith not.
Morton L-. Mullins
Subscribed and sworn to me
this 14th day of February, 1974.
•Tna^ iLJ Q.&
My commission expires: /97^7
237
-------
AFFIDAVIT OF D. T. KAVALEW
CONCERNING OPACITY
City of Wilmington )
) ss.:
County of New Castle, Delaware )
D. T. KAVALEW, being duly sworn, desposes and states
as follows:
I received a bachelor of science degree in chemical
engineering in 1963 from Michigan Technological University. I also
hold an MBA degree from University of Chicago which I received in
196 7. I have been in the employ of E. I. Du Pont de Nemours and Co.
Inc. since June of 1963. During the period 1963-1972, I was assigned
to the East Chicago, Indiana Plant where I held various engineering
and production supervisory positions, most of which were related to
sulfuric acid manufacture and environmental control. During the
past two years I have been assigned as a Senior Engineer in our
Wilmington offices, specializing in sulfuric acid manufacture and
related environmental control.
The relatively new environmental control regulations,
which apply to sulfuric acid plants, include opacity, .or plume
visibility, as one of the restrictions. In order to become more
familiar with the basic for, and application of, this regulartory
limitation, I attended the Federal EPA Visible Emissions (Opacity)
Evaluation training school at Research Triangle Park, North Carolina
on February 6-8, 1973. '
Approximately the first day and one-half of the
course was devoted to classroom instruction and discussion of various
items related to the generation of visible emission (smoke) plumes,
measurement techniques, and enforcement procedures.
The course moderator was Dennis P. Holzschuh of EPA.
Mr. Holzschuh indicated that he had been instrumental in writing
Method 9 (Federal EPA opacity determination method) and had worked
with several state and local agencies in the establishment of visibly
emission training schools.
During a portion of one of the lectures concerning
the qualification of opacity observers, I asked Mr. Holzschuh if any
studies had boon conducted to check the consistency and accuracy of
readings bv a group of trainees on stacks other than the smoke
generator (school training device). He indicated that none had but
in all likelihood the results would be more variable than during
the school qualification tests.
238
-------
On the afternoon of the secoad day, the students
were shown successions of white plumes era-'ttea from a stack about
15 feet high and about one foot in diameter. The instructor
Mr, r.'ii ;7c:chuh told the students the measured opacities, as determined
by the tronm i s s iomet e r, at 5?; opacity intervals ranging from 0 to
1G G %. Similar examples of black pium'es were planned but nob conducted
because of d maIfunction of the black-smoke-generating portion of the
smoke generator.
After memorizing the white-smoke opacity levels, we had
three white-smoke practice compliance ruus. This included 25 ob-
servations for each run with no single reading which varied from
the transr.iissiometer reading (known opacity) by 20% or more (15%
was an allowable error) and an average er:ror r.ot exceeding 7.5%.
I found that I was within these qualification requirements on the first
practice compliance run but failed badly on the next- two attempts.
I cannot explain this since thu three runs were made in less than
an hour and the sky conditions (partly cloudly) did not vary signi-
ficantly from run to run.
The third day was devoted entirely to qualification
runs with the brief exception of a short period of black-smoke
memorization runs since that had not been possible on the second day.
To qualify as a trained opacity observer, one had to record 25 white
pluues and 2 5 black plumes within the limits described' in the previous
paragraph.
The weather on the third day was completely overcast
with light rain and variable winds. Several of us were very skeptical
about the validity of becoming qualified under these weather conditions
Mr. Holz schuh acknowledged our concerns but indicated that the course
had to proceed regardless of the weather. I found that the black
smoke was discernable because the sky was uniformly white to light
gray. However, the white smoke plumes "/ere almost totally invisible.
We wore told to position ourselves so that we could view the white plum
against a contrasting background ( -ho crees on an adjacent hill or
the rod brick '.'all of a nearby building) . I chose the building wail
which diu .improve the viewing conditions but further added to my con-
cern about the validity of the whole procedure.
I failed the first official qualification run (three
values in-crror by 2Q-25Z), but passed the second. I was sure that
I could junt as easily have failed any subsequent run, and attribute
my qv "i 1 i: icat ion to luck as ranch as to actual ability. Mr. Holzschuh
stnf-i t ;...t about 75'- of the clan:; norm.il.ly qualifies. Only about
one-third oi our c-l.tss had qualified at the end of the third run with
tir<: lo: about throe more run:; before olass was adjourned.
Our class v.\u> r."or.sps* i ci of 15 students, about 40 v ot
v.-hi •••! v : ¦ 1 • i v.:ric \ - v.- ..-v.'. regulatory a ;or.ci as . Some of
-------
On the basis of my experience at the EPA school
and in the field, all observers would encounter substantial difficulty
in extra001 atincr from the small stack used at the school to commercial
stacks, which are oftc-n several hundred feet high. They would also
counter* substantial difficulty in making adjustment for key variable}
s lighting and background which were a'lmost constant at the
encounter
such a
school.
I am familiar with the experience of Du Pont personnel
with respect to visually measuring the opacity of plumes containing
condensed water vapor. The consensus among qualified observers within
the comoanv is that one cannot accurately discriminate between con-
densed water vapor and other opacity contributors. The EPA training
school contained no demonstration or testing with respect to plumes
containing condensed water vapor. I am not aware of any published
test results demonstrating the feasibility of measuring opacity in
plumes containing condensed water vapor.
There is no reliable relation between mass emissions and
opacity. Ke have found that with constant mass emissions, opacity
measurements vary widely presumably because of changes in the lighting,
background, stack diameter, exit gas velocity, and mist particle size
distribution.
Further affiant sayeth not.
¦J- yfi
D. T. Kavalew
Subscribed and sworn to
before ne this 31st day
of January 1974.
240
-------
SECTION H: ENFORCEMENT OF OPACITY STANDARDS
Page No.
H-l. "The Validity of Visible Emission & Opacity Standards,"
Mark Mestel, January 1976 243
H-2. "Opacity as a Readily Enforceable Standard," Pamela
Giblin, 1 972 . . . 255
H-3. "Simplified Visible Emission Standard," Emory J. Crofoot,
June 27, 1971 273
H-4. "The Opacity Witness," Kenneth B. Malmberg, EPA, 1976 .... 283
-------
H-l
(Unpublished Paper)
THE VALIDITY OF VISIBLE EMISSION AND OPACITY STANDARDS
Prepared bv
Mark Mestel
Enforcement Proceedings Branch
Division of Stationary Source Enforcement
Office of General Enforcement
U.S. Environmental Protection Agency
401 H. Street, S.W.
Washington, D.C. 20460
January 1976
243
-------
THE VALIDITY OF VISIBLE EMISSION AND OPACITY STANDARDS
The most ardent challenges to the use of visible emission standards
are predicated on the premise that opacity and/or the shade of smoke as
defined by the Ringelmann Chart does not accurately reflect the amount of
pollution contained in the emitted smoke. In Partiand Cement Association
v. Ruckelshaus, 486 F. 2d 375 (D.C. cir, 1975), the petitioner introduced
into evidence the results of a test conducted for the National Center for
Air Pollution Control (U. S. Dept. HEW) in an attempt to assail the validity
of an NSPS regulation which prohibited emissions from portland cement plants
in excess of 10% opacity. The HEW study concluded there was a marked degree
of non-uniformity associated with smoke reading and that it was not unusual
for an observer to err when making the reading. The petitioner argued that
the opacity standard was arbitrary and that the average degree of error was
so great as to preclude the accuracy of measurements used to ascertain
whether opacity exceeded the 10% standard. The court felt this to be a
legitimate concern and remanded this aspect of the case for further argument.
After considering additional information submitted by the parties, the Court
concluded:
"...We have considered the detailed analysis by the
Administrator of numerous factors involved in the use of
plume opacity to determine whether or not a portland cement
plant achieves a prescribed standard of pollution control. We
are not warranted on the basis of his analysis to find that
plume opacity is too unreliable to be used either as a measure
of pollution or as an aid in controlling emissions." Portland
Cement Association v. Train, 513 F.2d 506, 508 (1975).
244
-------
- 2 -
The validity of visible emissions was reaffirmed in Friends of
the Earth v. PEPCO, 419 F. Supp. 528 (D.C. C1r. 1976). The defendant
asserted that the provision in the District of Columbia SIP prohibiting
any visible emissions was not federally enforceable. PEPCO argued that
federal enforcement is limited to the implementation of ambient air quality
standards rather than visible emissions which it contended were not emission
limitations. Although the court felt Portland Cemant Authority v. Train,
supra, was controlling, it did state the necessary standard to be met by
a party seeking to prevail on this theory.
"Furthermore, in order for visible emission regulations
not to be related in any way to the implementation of the
ambient air quality standards, PEPCO must show that visible
emissions are wholly unrelated to the emission of any pollutants
for which there are ambient air quality standards."
Applying the pertinent facts to the standard, to wit PEPCO's admissions
that its visible emissions resulted from particulate emissions during ab-
normal operations, the court concluded that PEPCO had not shown the visible
emission regulation to be wholly unrelated to achievement of ambient air
quality standards.
CONSTITUTIONAL CHALLENGES
Sources subject to the pertinent provisions of the Clean Air Act have
also assailed smoke readings on constitutional grounds. The prevalent pro-
position advanced is that opacity does not necessarily indicate, nor accurately
measure pollution hence it is arbitrary and capricious to control pollution
on the basis of opacity standards. This argument was advanced in Air Pollution
Variance Board v. Western Alfalfa Corp. Colo., 553 P. 2d 811 (1976). Despite
245
-------
- 3 -
the agreement of both parties that equivalent opacity is not directly
correlated to the density, volume or amount of particulate matter being
carried in emission, the court found Western Alfalfa Corp. had failed
to adduce the proof necessary to demonstrate a statute authorizing visual
opacity testing so arbitrary or capricious as to render it unconstitutional.
It concluded the opacity standard provides an accurate reflection of the
amount of visible contaminant being carried in a smoke plume.
A constitutional challenge based on a similar proposition but with
greater visceral appeal is advanced by the authors of "Visual Plume Readings -
Too Crude for Clean A1r Laws", 7 Natural Resources Lawyer 457 (Summer 1974).
It is their contention that conclusive presumption statutes based on visible
emissions constitute a conclusive presumption which irrebuttally links
opacity to pollution and therefore violate due process. In support of
their theory the authors cite three recent United States Supreme Court
decisions: Valadis v. Kline, 412 U.S. 441 (1973); U.S.D.A. v. Murray. 413
U.S. 508 (1973); Cleveland Board of Education v. LaFleur, 414 U.S. 632
(1974). While the general language of these cases do suggest that conclusive
presumptions are constitutionally deficient under certain circumstances,
a closer examination of the decisions (not warranted for this discussion)
reveals a set of facts and legal principles inextricably interwlned between
due process and equal protection which yield a sliding scale application
of equal protection principles. As with all equal protection analysis,
relief is predicated, at the very least, on the classification being unfair
to the individuals affected. The use of opacity does not adversely affect
any identifiable group of Individuals, and it 1s therefore inappropriate
to apply the Supreme Court's due process theory.
246
-------
- 4 -
COURTROOM UTILIZATION OF OPACITY READINGS
Introducing the Evidence
Assuming arguendo that the proceeding discussion concerning the
validity of visible emission and opacity standards accurately reflects
the current state of the law, the focus of attention must be directed
to adducing the proofs necessary to constitute a violation. A large
majority of the Courts which have heard air cases based on violations
of these standards have admitted, over objection, the testimony of
qualified smoke readers.
Admissibility
Inasmuch as the smoke reader's testimony frequently constitutes the
sole evidence substantiating the source's alleged violation of the visible
emission/opacity standard, its admission into evidence is of crucial im-
portance. Although counsel for the source will be adamant in his/her
attempt to prevent admission of the preferred testimony, the court
typically allows it into evidence; its decision is rarely, if ever, re-
versed on appeal.
A challenge to the admissibility of an opacity reading was unsuccessful
in Air Pollution Variance Board v. Western Alfalfa Corp., supra. As with
the challenges to the validity of visible emission regulations, this
challenge was predicated on the inherent inaccuracy of a test in which an
observer is confronted with a number of uncontrolled variables. In re-
sponse to this contention the Court stated:
"This Court has reviewed the theory of equivalent opacity,
particularly as the testing method is affected by such variables
as cloud cover, time of day, humiditv, atmospheric haze, wind
247
-------
- 5 -
velocity, compactness or diameter of the column of smoke,
and position of the observer in relation to the source of
available light. We have reviewed the smoke school certification
procedures and the qualifications of the observer who conducted
the inspection in the case. On the basis of the review, we con-
clude that the accuracy of equivalent opacity readings has been
sufficiently well established to permit their introduction into
evidence. The Department of Health smoke readers are certified
at smoke reading schools to make readings with no more than 7.5%
average margin of error. The readings taken at the three plants
here in issue were well within the 7.5% margin of error factor."
• • • *
"For purposes of hearing before the Air Pollution Variances
Board, the results of visual opacity inspection are wholly re-
levant and admissible. Any Question as to the method of con-
ducting the inspection, the qualifications of the inspector or
the reliability of the testing procedure rests in the sound
discretion of the trier of fact." (at 815)
The introduction, over objection, of smoke readings was admitted in
Llovd A. Fry Co. v. State. Tex. App., 541 S.W. 2d 639 (1976), on the basis
of the magnitude of the violation. The pertinent regulation prohibited
bisible emissions in excess of 3035 opacity averaged over a five minute
period. The Fry Roofing Co. claimed the smoke readers' method.of determining
opacity was inherently inaccurate as tests used to certify persons as smoke
248
-------
5 -
readers allowed a margin of error of 15% for an individual reading and
7.5% for an average of 50 readings. The Court rejected the Fry Roofing Co.'s
claim that this margin of error gave the observer unlimited discretion
in selecting violators. The Court found that experience indicated that
the actual variation was less than 5% and that the violations reported in
the instant case exceeded the 30% opacity standard by more than 5055. The
Court did note that had the reading indicated a opacity within the margin
of error a serious admissibility question would have been presented.
COMPETENCY OF WITNESSES
The Court's are in general agreement concerning the admissibility of
field observations made by smoke readers, though they may differ as to
whether it is first necessary to qualify the witness as an expert.
In State v. Lloyd A. Fry Roofing Co., Ore. App., 495 P.2d 751 (1972), the
source argued that the testimony of two smoke readers' was inadmissible
because they did not possess sufficient qualifications. The court stated
that the smoke readers were not offered as experts, in that they were not
asked to express opinions, but rather facts based upon their observations.
In this capacity the court found that the only necessary prerequisites
to testifying on relevant or material matter were personal knowledge, or
experience, training or education if such be necessary. Without deciding
whether special training is a prerequisite to testimony by smoke readers,
the Court held a two day training course, certification by the Oregon Smoke
Evaluation Training School and recertlfication every six months were
sufficient to justify admission of the testimony. It should be noted
that the Oregon Appellate Court had previously rejected an argument which
would have required a witness testifying on background obscuration to have
249
-------
- 7 -
special training. City of Portland v., Fry Roofing Co., 3 OR.App. 351,
472 P.2d 829 ("1970).
In those cases in which a court has required expert qualification
of a witness as a prerequisite to testifying, valid certification by a
smoke reading school has been deemed sufficient. People v. International
Steel Corp.. 102 Cal. App., 2d Supp. 935 C1951).
Sufficiency of the Evidence
The court's acceptance of the smoke reader's testimony does not end
the challenges. Counsel for the source will typically argue, at the close
of the plaintiff/prosecutor's case in chief and on appeal, that the smoke
readings are insufficient to support a verdict against it. Many of the
air pollution cases are adjudicated fn an administrative rather than judicial
forum. In this setting the standard utilized to determine whether sufficient
evidence was presented to support the regulatory action is more than a mere
sciRtilla of evidence or suspicion of the existence of a fact to be established
ft requires such relevant evidence as a reasonable mind might accept as
adeouate to support conclusion. Bortz Coal v. A*r Pollution Commission.
2 Pa. Cmwlth. 441, 279 A.2d 388.
The Comnonwealth Court of Pennsylvania 1n Bortz Coal. supra., deter-
mined that the testimony of two homemakers and an air pollution control
engineer was insufficient regardless that the latter had observed the
source's smoke plume and had determined, but without the aid of a Ringlemann
Smoke Chart, that it exceeded the applicable regulations. In the Court's
words:
"Visual tests and observations are not adequate evidence of
a violation where recognized scientific tests are available ....
This Court cannot close its eyes to the necessity of a regulatory
250
-------
- 8 -
agency proving its case .... CW)here there are available
established methods for determining violations, those methods
must be used." (A.2d at 398) "This court is not unfamiliar
with the Ringelmann Smoke Chart, and therefore the Court is
puzzled why such an inexpensive method of testing was not
used in this case." CA.2d at 399)
The Bortz case was remanded to the Air Pollution Commission for the
purpose of establishing substantial evidence of the alleged air pollution
violation. At the subsequent hearing the Department of Environmental
Resources presented evidence through four air pollution experts who testified
that the emissions from Bortz*s coke ovens, as measured and determined by
the use of the Ringlemann Chart, MSA smokescope, and Pilbrico Smoke Chart
exceeded the specified limits. Bortz once again appealed questioning the
sufficiency of the evidence, but met with no success. The Commonwealth
Court of Pennsylvania affirmed characterizing the Ringlemann Chart as a
"recognized scientific testing device" and the testimony as sufficient to
satisfy the substantial evidence requirement. Bortz Coal v. Department of
Environmental Resources, 7 Pa. Cmwlth. 362, 299 A.2d 670 (1973).
The Utah Air Conservation Committee on the basis of the contraverted
testimony of smoke readers to the effect that the Lloyd Fry Co. was in
violation of a state regulation requiring plumes to be less than #2 on the
Ringelmann Smoke Chart, ordered Fry to reauest a variance and submit a
compliance schedule or shut down. Fry appealed, Lloyd A. Fry Co. v.
Utah Air Conservation Committee. Utah 545 P.2d 495 (1975) alleging that
the administrative findings of fact were not supported by substantial
evidence due to the inherent unreliability of smoke reading and the
251
-------
- g -
haphazard ciarmer in which the readings were made. The Court did not
agree. Its examination of the record revealed the manner in which the
smoke school was conducted, the methodology of reading smoke, the
qualifications of the individual smoke readers, the procedure utilized,
and the state of the natural conditions at the time the readings were
made. Based on this record the Court concluded that there was sufficient
evidence to create an issue of fact for determination by the hearing ex-
aminers and the weight to be accorded to the testimony was within the
prerogatives of the fact finders. In accord, Lloyd A. Pry Roof Co. v.
State Department of Air Pollution. etc.. Colo., 553P.2d 800 (1976).
CONCLUSION
The validity and admissibility of EPA/State visible emission standards
are well established throughout the country, having withstood countless
challenges. However, care must be taken to insure their continued vitality
If the parameters established by the courts are not adhered to, 1t is
possible that the smoke reading programs will have to be restructured.
The primary area of concern should be those criteria utilized by re-
gulators to determine what constitutes an actionable violation. There
should be a stringent rule prohibiting enforcement proceedings against a
source unless the smoking reading, have been performed in the appropriate
manner, clearly indicates that the plume has surpassed not only the applic-
able limitation, but additionally the average margin of error attributed to
smoke readings as indicated by the court in Lloyd A Fry Roofing Co. v.
State, supra which exceed the standard but are within the margin of error
may present readings serious admissibility problems. Rather than chance
an unfavorable judicial precedent, discretion dictates pursuing only those
252
-------
- 10 -
sources whose emissions significantly exceed the prescribed limit.
253
-------
fl-2
Paper Presented at 65th Air Pollution Control Association Meeting,
Miami, Florida, 1972
72-9£ OPACITY AS A READILY EUFORCZABLE STANDARD
Pamela Giblln, Texas Air Pollution Control Services,
Austin, Taxes
Equivalent opacity as a viable pollution control standard Is not new.
As its use Is expanded it becomes Important to be able to enforce it.
There are three basic elements involved in a suit based on opacity.
The basic legal validity of equivalent opacity is the first point
which oust be established. Under administrative law, which encom-
passes most air pollution regulatory agencies, there is a presumption
of validity attached to the rules or regulations of an agency. In order
to overturn a regulation such as opacity, it la necessary to demonstrate
that it is so arbitrary and capricious that no reasonable administrator
would have adopted it. This is extremely difficult to show in view of
the numerous air control programs throughout the country that have adopt-
ed opacity and the many court cases that have upheld It.
The second basic element of opacity litigation is the presentation
of the methods by which opacity observers are trained. This is accom-
plished by having the smoke school instructor testify as to the detailed
and rigorous training received by all certified smoke readers. In Texas,
smoke schools are conducted by the State Air Pollution Control Services.
In order to qualify, a smoke observer must complete 16 training sets of
emission readings. After that, he must take a set of 50 readings in
which his average deviation is not more than 10Z and no reeding varies
20X or more.
The final factor In an opacity case is the testimony of the ob-
server wlio made the readings by which the source in question has been
found in violation. The reader must have followed all the necessary
steps for accurate observations. Such things as position ef the sun,
txtcnt of cloud cover, and steao content of the plume are extremely Im-
portant and should be documented In the reader's Smoke Observation Forms.
The original foras are admissible as evidence if the information was
filled out at or near the time of the readings and if the reader can
attest to their accuracy.
All three elementa are being dealt with successfully by eir con-
trol programs throughout the country and are pevlng the way for in-
creased use of opacity regulations.
255
-------
APCA NO. 72-90
OPACITY AS A READILY ENFORCEABLE STANDARD
In 1969 the Texas Air Control Board adopted the equivalent
opacity standard as a valid method of grading and measuring smoke,
visible emissions, and suspended particulate matter in the seven
most populous areas of the state. The present opacity limit is
40 percent. The Board's regulations define equivalent opacity
as "The degree to which an emission, other than gray or black
smoke, obscures the view of an observer, expressed as an equiva-
lent of the obscuration caused by a gray or black smoke emission
of a given density as measured by a Ringelmann Smoke Chart."1
After December 31, 1973, equivalent 6pacity will become
a statewide standard in Texas. In addition, it will be used to
read smoke of any color and the limit will be lowered to 30 per-
cent and 20 percent depending on the date the source was con-
structed. The past few years have witnessed a marked increase in
the number of enforcement actions based in whole or in part on
equivalent opacity observations and it is highly likely that
opacity will play an increasingly significant role in future
court enforcement by the Texas Air Control Board.
Such confidence in visual observations made by an observer
is well warranted. Opacity, while sometimes disparaged by
scientists as not being the most technically accurate method of
measurement, is preferred by many air pollution control lawyers
^Section II.E. of Regulation I
-4-
256
-------
APCA NO. 72-90
for its relative ease of enforcement. It is the purpose of
this paper to review and discuss the major elements of a lawsuit
based on equivalent opacity. It is not within the purview of
this presentation to debate the scientific merits or flaws of
opacity, but simply to analyze its effectiveness in litigation.
BASIC VALIDITY OF OPACITY
The first major attack which is usually launched against
opacity in a suit is that the standard as such is invalid by
reason of being arbitrary and^capricious. These charges are
fairly common in the realm of administrative law and some
general background is necessary to understand the thrust of
the "arbitrary and capricious" attack.
Administrative law encompasses agencies, such as the Texas
Air Control Board, which combine legislative, executive, and
judicial functions in a single entity. It is widely recog-
nized that administrative agencies are better equipped than
courts, by specialisation, by insight through experience, and
by more flexible procedure to ascertain and interpret the cir-
2
cumstances underlying certain technical issues.
2U. S. v. Western Pacific B« Co., 352 U.S. 59; Fed. Communi-
cations Commission v. RCA Communications. Inc.. 346 U.S. 86;
Swift * Co. v. U.S., 343 U.S. 373» Far East Conference v. U.S..
342 U.S. 570} Universal Camera Corp. v. VLRB. 340 U.S. 474.
257
-------
APCA NO. 72-90
The most pervasive example of the quasi-legislative
function of administrative agencies is the authority to adopt
rules and regulations. "Rulemaking" is legislation on the ad-
ministrative level, that is, legislation within the confines of
the statute under which the agency was created. Aa such, rules
and regulations have the force of law and must be sustained unless
they are unreasonable and plainly inconsistent with the statute.*
An administrative agency has a large range of choices in
determining what regulations or standards should be adopted.
It.is not necessarily a valid objection that another choice
4
could reasonably have been made , that experts disagree over
the desirability of a particular standard , or that soma other
method of regulation would have accomplished the same purpose
and would have been less onerous.® It is enough that the agency
has acted within the statutory bounds of its authority, and that
its choice among possible alternatives adapted to the statutory
end is one which a rational person could have made.
3Burga v. Commissioner (CA4), 253, F2d, 265, 74 A£R2d 664.
^Federal Secur. Amr. v. Quaker Oats Co., 318 U.S. 218.
SMitchell v. Budd. 350 U.S. 473.
^Man O'Viar Racing Ass'n Inc. v. State Horse Racing Commission.
250A2dl72, 35 ALR 231060.
258
-------
APCA NO. 72-90
With this brief background as to the broad grant of
discretion given administrative agencies, it becomes apparent
that an attack on a regulation such as Ringelmann or opacity
faces a heavy burden of proof. The courts are so strong in
supporting the validity of regulations that an administrative
rule will not be set aside unless it can be shown that no
reasonable administrator would have made such a regulation
and that it is so lacking in reason that is essentially
arbitrary.'
Such a showing of arbitrariness is extremely difficult
in view of the many state and local air pollution control
programs throughout the country which have adopted opacity
regulations and the large number of court cases which have
upheld Ringelmann and equivalent opacity. California, New
Jersey, Washington, and Arizona are just a few of the states
where there have been significant appellate court decisions
soundly supporting the use of these standards.
In People v. Plywood Mfg.'s of California,^ the Superior
Court of Los Angeles County rejected the view that the Los
Angeles Ringelmann and opacity provisions of the California
Code were invalid because a plume of smoke may appear darker
from one position than it does from another. As the court
7See 136 ALR 734 j and Udall v. Tallman. 380 U.S. 1.
®291 P.2d 587.
259
-------
APCA NO. 72-90
stated: "If the contaminant has the substance that, fairly
viewed from any position, gives it a shade as dark or darker
than Ringelmann No. 2 [that was the maximum density allowed
by the California Code}, it is condemned, no matter how light
in color it may look to someone situated at another vantage
point."9
The Ringelmann Chart was used by Loa Angeles County as
early as 1945 and it has been extended to allow the concept
of equivalent opacity. The California courts have led the
way in upholding opacity readings made without the need for
actually having a Ringelmann Chart. As tlje courts put it,
determination of opacity with respect to contaminants is
made not by a comparison with the Ringelmann Chart itself
but by a comparison with the op&city of smoke that has been
shown to match certain numbers on the Ringelmann scale. In
People v. Plywood Manufacturers' of California, the Superior
Court of Los Angeles County upheld the validity of the Los
Angeles Ringelmann and opacity provisions. The court stated:
"Subsection (a)[which refers to the black smoke Ringelmann
Section] only begins to solve the problem of the discharge
of contaminants into the air; it does not touch smoke and
other substances too light in shade to come up to Ringel-
mann No. 2. They may be so substantial in material, however,
'ibid., at 591.
260
-------
APCA NO. 72-90
that they make it Impossible to sea an object on the other
side. We have all seen very white smoke that shuts out
the view completely. Again they may obscure the view to
a lesser degree than totality ... We may, therefore, express
the test of Subsection (b) (equivalent opacity section] in
simple termss it condemns smoke or any other contaminant
that is at least as hard to see through as is smoke which
is as dark or darker than Ringelmann Mo. 2. There is
nothing mystic or incomprehensible about such a statement."*0
Even if the basic validity of opacity 4s conceded, defense
attorneys will often argue that the standard is beyond the statu-
tory authority of an agency such as the Texas Air Control Board
in that there is no relationship between visible particulates
and health. In fact, however, the Texas Clean Air Act under which
the Board was created and from which the Board's power is derived
includes "the esthetic enjoyment of the air-resources by the
people and the maintenance of adequate visibility,"1'1 in addition
to the protection of health, general welfare, and physical property.
The very inclusion of esthetic and visibility considerations rein-
forces the usefulness of standards such as opacity.
10291P.2d 587 at 591-592.
USec. 1.02 of the TCAA, Art. 4477-5, V.T.C.S.
261
-------
APCA NO. 72-90
Yet even if health were the sole' concern to which the Board
could address itself, there is enough evidence to indicate that
the smallest particles are the most opague and are also the
most respirable and del.iterious. This would provide a sufficient
basis for proscribing emissions over a certain percent opacity.
PROPER TRAINING OF AN OPACITY OBSERVER
Establishment of the basic validity of opacity concludes only
parto of the battle. In an opacity enforcement action, it is
imperative to demonstrate that the observer had the proper
foundation and training to make accurate headings.
12
In People v. Plywood Manufacturers of California, the
Superior Court of Los Angeles County rejected arguments that
opacity and Ringelmann provisions were invalid because of a
reference to an "observer", a reference which is found in th«
Texas definition of opacity. The court refused to require a
test "that only a scientist with expensive equipment can make."*-3
14
In another court case, it was held that witnesses who attended
a school where they were trained to read smoke had the competence
to testify as experts on opacity. "Expert" in the legal sense
is more liberally construed than in other areas and it applies
to a person who, through training or experience, has particular
knowledge not possessed by the average person.
12291 P.2dS87.
13Ibid., at 592.
14Pcoplo v. International Steel Corp., 226P2d587.
^Soe Texas Law of Kvidence by McCormick t Ray, Vol.2, Section 1401.
262
-------
APCA NO. 72-90
In relying on visual readings made by an observer it is
extremely important to lay the proper predicate as to his
expertise and objectivity by carefully outlining the procedures
by which he was trained,. An effective method of presenting
this data in court is through the testimony of a smoke school
instructor as was done in City of Houston t State of Texas v.
Lloyd A. Fry Roofing. In that case, the first significant
opacity case in Texas, a temporary injunction was obtained
on the basis of opacity violations.
After the instructor's competence and credentials are
established through a brief summary of his^ professional and
empirical background, his general testimony should set out in
detail how the smoke schools are conducted.
In order to assure uniformity of instruction, Smoke
Reading Schools in Texas are conducted by the staff of the
Texas Air Pollution Control Services for the Board which
is the only certifying authority in the state.
Anyone, whether from industry, the general public, or state
personnel can be certified by meeting the prescribed requirements.
Trainees are taught to read emissions from a smoke generating
device capable of emitting both white and black smoke in varying
degrees of opacity or density. The smoke generator is equipped
with a transmissometer which is calibrated to show accurate
measurements of smoke. To be a recognised smoke reader in Texas,
the observer must have successfully completed the authorised
training program and must have been certified within the previous
six months.
263
-------
APCA NO. 72-90
The initial step in the training' is familiarization with
known densities of black and white smoke. Smoke of various
degrees of obscuration is emitted from the generator and the
instructor calls out the reading from the meter. This is re-
peated until tha trainees are sufficiently acquainted with
the various opacity percentages to be able to recognize them.
Each smoke training run consists of 25 readings. The in-
structor uses the generator to produce smoke emissions of different
densities. Following each emission, the trainee-grades the
opacity he observed and marks it on his smoke evaluation training
form. (See Exhibit 1). At the end of twenty-five readings,
the transmissometer readings are entered and the deviations
are calculated. The trainee must complete 16 sets of training
runs prior to testing for certification. After each run, the
transmissometer on the smoke generator is calibrated to maintain
accuracy throughout the training.
After completing the 16 training sets, a trainee is
ready to take the certifying set which consists of SO readings.
A person may be certified as a visible emissions evaluator when
his average deviation on the certifying set is not more than
10 percent and no reading varies 20 percent or more for a set
of SO consecutive readings. He must also demonstrate to the
instructor a consistency and reliability in his determinations.
The foregoing description of the smoke school, at which
all certified observers in Texas are taught, helps convey to
the court that opacity readings are not haphazard guesses, but
264
-------
APCA NO: 72-90
are educated ratings baaed on specialized training. In addition,
it is helpful if the instructor's testimony stresses the fact
that all trainees are carefully taught to observe the following
directives whenever they are grading smoke:
(1) During periods of bright sunshine, the sun should
be directly behind the observer or within 45*
either side of that direction. Zn cloudy weather,
the position of the sun is relatively unimportant.
(2) Readings should be taken at right angles to the
wind direction and from a distance necessary to
obtain a clear view'of the stack and background.
(3) Readings should be made through'the densest part
of the plume and where the plume is approximately
the diameter of the stack. On plumes containing
steam, opacity readings should be taken immediately
beyond the point where the steam dissipates.
(4) An observer should not study the plume as this
will soon produce fatique and cause erroneous
readings. Instead he should glance at the plume
and record his observation immediately, looking
away from the plume between readings.
(5) Readings should be taken at 15 minute intervals
for a period of six consecutive minutes and recorded
on an appropriate form. (See exhibit 2.)
Under cross-examination by the defense attorney, the in-
structor should be able to re-assert his confidence in the
265
-------
APCA NO. 72-90
techniques of smoke training. Defense counsel will often
select details of the instruction method and attempt to
show that these deviated from accepted practice. For example,
he may zero in on the calibration of the smoke generator trans-
missometer, asking such questions as:
—"Do you have any special training in calibration?"
—"Do you rely solely on the smoke generator manual
in calibrating?"
—"Can you be certain that the smoke generator was
properly calibrated at the time each run was made?"
By pursuing such a line of questioning, the defense hopes
to raise doubts as to the reliability of the smoke generator
readings. If such unreliability can be demonstrated, all the
tests which were based on those readings would be subject to
question. It is therefore imperative to be able to support
and justify each step in the numerous interrelated procedures
connected with smoke school. As an expert, the smoke school
instructor is competent to make judgments based on his experience.
He can state, for example, that his training (from NAPCA, from
smoke generator manufacturers, or from whatever sources) included
instruction on calibration and that he uses the smoke generator
manual as an aid but not as his only source of information
on calibration. He can properly conclude that the transmis-
somcter was accurately calibrated prior to each run because
ho followed correct and established calibration methods.
266
-------
APCA NO, 72-90
An experienced instructor who has been prepared by his
attorney on the Air Pollution Control staff as to what to
expect on the witness stand should have little difficulty in
fielding questions on any facet of smoke school training.
OPACITY READINGS AS EVIDENCE
After the instructor has laid the predicate for the
opacity observer's competence and expertise, the observer
himself takes the stand to testify as to the opacity vio-
lations he has witnessed. It is on his testimony that the
major thrust of an enforcement action depends. Bis cred-
ability will determine whether $50 to $1,000 per violation
and/or an injunction will be obtained.16
It is not necessary that the observer have all the
details of each reading memoriied. It is perfectly permis-
sible and actually preferable to refresh h£s recollection by
having with him the Plume Observation Record Forms. (See
Exhibit 2). The original forms are admissible as evidence
if they were filled out at or near the time of the events in
question17 or at least if the facts were fairly fresh in the
18
memory of the witness when recorded.'' in addition, the
witness must attest to the correctness of the information on
16These are the penalties provided in Sec. 4.02 of the TCAA.
17Rice v. Ward. 56 S.W. 747.
1SFire Association of Philadelphia v. Nami, 77 S.W. 2d 260.
267
-------
APCA NO. 72-90
the forma. The forms when admitted must be made available
to the defense* counsel for inspection and use on cross-examination.
One of the advantages of the opacity standard is that large
numbers of violations qan be presented to the court in a fairly
short period of time. Once the theory and technique of opacity
observations have been explained, it is not a lengthy matter
to introduce the Observation Forms and to have the witness
testify as to the violation percentages. It is helpful to
have a large blackboard in the courtroom with blocks drawn
for the following categories:
READINGS FOR PLUME PRODUCERS, INC.
DATE TIME STACK OR TYPES OF SMOKE DENSITY
INSTALLATION
1/29/72
(example)
9:45
Incinerator
691
268
-------
APCA NO. 72-90
As the witness testifies on each opacity reading he made,
the pertinent facts can be entered in the appropriate blocks.
This aids in fitting the testimony together and in providing an
overview of total opacity conditions at the source over a period
of time. After each smoke density is recorded, the witness
should state that if a reading is 40% or more it constitutes
a violation. He may also point out that a notice of the vio-
lation was sent to the company or source in question.
Although photographs cannot be used as evidence of the
actual opacity percentages which were observed, they are useful
in showing that the meteorological and topographical conditions
were such that the stack could be read. It is a good graphic
aid in conveying to the judge or jury what the stack looks like
and the location from which emissions were observed. For a
photo to be admissible, it is only necessary that the witness
know the scene and testify that the picture correctly represents
the facts.
In cross-examining the observer, the attorney for the
defense will often try to show that the observer failed to follow
proper procedures for reading opacity. He may ask detailed
questions as to the position of the observer, the tine of day,
and the weather conditions. If he can demonstrate that the
sun was shining in the observer's eyes, he can discredit the
readings. He may also try to show that the sky was so cloudy
that there was insufficient background against which to read
the plume. In addition, defense counsel may introduce evidence
269
-------
APCA NO. 72-90
that the plume contained uncombined water which was not discounted
in grading the density.
These are all valid points which if proven could seriously
undermine the credibility of the observer. For this reason, he
should be extremely careful in making his observations and he
should document the conditions surrounding his readings, including,
whenever possible, photographs of the site. If there was any
steam present, the picture should show where it dissipates. Tha
observer should have read a wet plume at that point of dissipation.
Ultimately, the outcome of opacity litigation hinges on
the thoroughness with which the observer was trained and on
the precision with which he made his readings. The standard
is viable and valid but its application to a particular source
must be demonstrated by careful, accurate observations.
270
-------
EXHIBIT *1
TEXAS AIR CONTROL BOARD
1.
2.
3.
4.
5.
6.
SMOKE EVALUATION TRAINING FORM
Name of Obscrvf • ftP.ftTt?m frfJT
Afftllatlon At* P&l/. t'Tj*S>
Wind Speed rn m P tt Direction
Observers Position// g. •+ /oo vas
t° :o°
Sky Condition ^£1?*
Corrected By t=o centire?/h/r/i/-y-
JL&OM -T*>G XTAC-#
(Record BUck of Gray Smoke In Rlngeln&nn Ho. r 1/4 Unit Smallest Division)
(Record All Other Smoke in X Opacity - 5X Smallest Division)
ruh no..
JJL
6
B
en
G
•o
1
a
BC
Observer
Reading
t.
a
1?
* —
as
C ac
a
w
R
O
a
>
CJ
~
fS
0
**
3
1
•
0
01
c
"O
8
ce
Observer
Reading
3
§c*
M 6
VI
1 «
C S
s
K-
e
o
»*•
W
K
>
U
a
~
c
0
w
m
>
O
1
1
2
3
20
13
1/2
1
10
2
? V?
3 V?
20
14
1 V?
2
5
3
3
4
?Q
15
2 V2
3
10
4
1 J'4
2
5
16
s
4 V?
If)
5
1/2
1
10
2 V?
? V?
6
1/4
1/2
s
18
1 V2
? V2
7
1
1 V2
10
19
2 V4
2 V4
8; 2 1
? V?
10
20
3
3 V2
10
9
)
3
—
21
?
2
_—
„
10
* 1/4
1 V?
5
22
1 V?
1 V?
11
4
4
23
3/4
1
5
12
4 3'4
5
5
24
1/2
1/2
_ ,
125
E
5
--
RUN NO.ZJi—
10
11
12
t.
41 Ol
C.S
t> •O
"I •
SS
.23.
IL
Jfl.
35
40
50
60
75
as
80
60
55
T "a
G h
JP.
12.
Jfl.
40
50
£0
70
80
90
80
65
55
13.
&
10
10
10
13
14
15
16
17
18
19
20
21
22
23
24
25
u
«) B«
C-5
tl "O
v< m
x> 4»
OK
JO.
Jfl-
JQ.
15
JL
35
JL
JL
85
70
JL
65
6 «
v> «>
C K
s
JL5.
JL
JZL
JL
JO.
JL
JL
JL
IL
21
M.
70
7.
8.
9.
10.
11.
12.
13.
14.
Run Number
Number Correct
Number of Plus Deviations
(lumber of Minus Deviations
Average Plus Deviations - H"" It El"*
1 No. of Mus Deviations
Average Hlnus Deviations ¦ ?"W °» u|nus Scvli^-lon-
llo. ofriinus Deviations
Average Ocvlatlon ¦ vSuni of Plus Deviations) ~ (Sua of Hlnus Deviations!
, Total tlo. of Reading!
"0. of Readings >02 Deviation and Over
271
-------
TEXAS AIR CONTROL BOARD
PLUXF. OBSERVATION RECORD FORM
EXHIBIT #2
NAME OF SOURCE /V///»/>• PAnfU/ftr/> /A/AY £ 9 . (171.
observer fi&/5re
START TIME 9 VS" /A.
ADDRESS S"/f SOO y S7~
"100 r.jTY -reri*^
SMOKE DENSITY TABULATION
OBSERVATION POINT<70& S .St£•/?!//<•/?
srftT/aA'., ft*—.sggr—s_c—
No.
Units X
Units
Equlv.
No. 0
No. 1 Units
DISTANCE FROM STACK (,S PK
Units
No.
Units
No. 1
WIND SPEED / MPH DIRECTION S. &.
-Units
No.IIjI
/
Units
No. 2
2
TYPE OF BACKGROUND 31 l/E SAT Y
V
Units
No.2^_
/o
s
Units
No. 3
/ST
COLOR OF EMISSION UHlTe
s
Units
No 3*J_
/¦fJS
TYPE OF INSTALLATION /nye./AJlSAAraA
e.
Units
No. 4
*1
*
Units
No.4^.
9
i
Units
No. 5
s
OBSERVATION ENDED 9 S/ A./n.
Total
Units
"N. S«e.
Hlihv
0
is
30
45
0
Fif
?9
fo
1
tt
ft
/oo
2
?<>
PO
10
so
3
JT°
y<9
4
ta
70
to
5
?6
PO
7a
. (/Q ,
Total Equlv. No. 1 Unit*
• M
Averaga Soak* Density •
Equlv. No. 1 Unit* * 201
Total Unit*
69%
REMARKSi
SICKED
s^2d2&zk.
272
-------
H-3
#71-51
Simplified Visible Emission Standard
Emory J. Crofoot
Columbia-Willamette Air Pollution Authority
Portland, Oregon
Presented at the 64th Annual Meeting of the
Air Pollution Control Association, Atlanta
City, New Jersey, June 27, 1971
273
-------
The new simplified visible emission standard replaces the long
used Ringelmann Chart and the ancillary "equivalent opacity" standard.
The new standard is based upon the opacity of a plume expressed in
percent. It is easier for the smoke reader to use as he requires training
and background on one standard only for any visible emission, regardless
of content or color of emission.
The principles of the new standard are much easier understood by
the smoke reader. The confusion of attempting to explain the relation-
ship of the equivalent opacity of a white emission to a black smoke chart
is eliminated, thus making it easier for judges, jurors and members of
general public to comprehend the standard. This makes the standard
easier to work with in litigation.
The new single emission standard for all types of plumes is as
follows:
"Opacity means the degree, expressed in percent, which an
emission obscures the view of an object in the background.
No person maintaining, owning or operating an air contaminant
source of emission shall discharge into the atmosphere from
any single source of emission whatsoever or from any combina-
tion of sources if the emissions combine, any air contaminant
for a period or periods aggregating more than three minutes
in any one hour which is equal to or greater than 20JS opacity.n
Opacity observations are generally made with the sun behind the
observer. This results in a lower opacity reading, and is advisable
practice, but not necessary from a legal standpoint.
274
-------
The weak point of the new standard is that it is enforced by a
subjective test. This can be overcome by a good training program and a
written record logging the smoke reader's actual experience in using
the standard.
If the emission standard exempts uncombined water, the emission must
be read downstream from the evaporation point. The evaporation point
cannot be scientifically established, but the observer can be easily
trained on this point.
Every smoke reader should, and every prosecutor must be fully familiar
with the contents of Public Health Service Publication No. 999-AP-JO.
The publication contains much material when lifted from context would
defeat prosecution of an otherwise valid case. But this material can
be overcome if entire content of publication is fully appreciated.
A scientific instrument, perhaps the laser, should be developed for
enforcement of visible emission standards. This would eliminate the
question of accuracy of the subjective eyeball test.
275
-------
71-51
As time marches on, the "good old days" are constantly being replaced
with progress. After having been of faithful service for 73 years, the
Eingelmann Chart with the subsequently developed "equivalent opacity" standard
based thereon should be laid to rest and make room for progress. Elaborate
ceremony with the shedding of a tear would not be inappropriate, but I urge
you to hasten the writing of the eulogy.
The sometimes difficult to comprehend, and even more difficult to
explain in. simple terms, double standard of measuring both black and white
pluses by the same black smoke chart should be replaced by a simple single
standard that has -no reference to color.
The new standard is as follows:
"Opacity means the degree, expressed in percent, which an
emission obscures the view of an object in the background.
No person maintaining, owning or operating an air contaminant
source of emission shall discharge into the atmosphere from
any single source of emission whatsoever or from any combina-
tion of sources if the emissions combine, any air contaminant
for a period or periods aggregating more than three minutes
in any one hour which is equal to or greater than 20£ opacity."
The standard set out above is better because it is equally effective
for use in limiting black, white or any other color plume. It is more simple
because no one has to explain how to relate the opacity of a white plume to
a chart designed for the measurement of black smoke.
Since the beginning of controlling visible emissions by a regulation
other than one defined in terms of dense smoke, the regulation has been
couched in terras of "equivalent opacity", meaning-an opacity equivalent to th«
opacity of black smoke measured by the Ringelmann Chart.
276
-------
It appears that most people, at least in a vague sort of way, understand
the optical principles involved in the use of the Ringelmann Chart for black
smoke. It also appears that most of these same people can understand, in a
vague sort of way, the application of "equivalent opacity" to a dark blue
emission — one that is similar to black or a shade of gray. But it also
appears without question, when relating opacity of red, yellow, green or any
other color gas stream running the spectrum to white to the equivalent of a
shade of gray or black most people get lost.
Perhaps you are wondering why I think it is important that "people" be
able to understand the workings of a visible emission standard. Please keep
in mind the word people includes all those persons of the public whom you are
serving, if you are employed by a public agency, and who would like to
understand the terms of a regulation enacted for their benefit. The term
also includes the persons regulated by the regulation. The term is of further
importance in that it includes Judges and jurors who are or will be passing
on the guilt or innocence of those charged with violating the standard. And
last, but not least, it is important that the "observer" or "smoke reader"
who makes the reading and gives the testimony understand what he is doing when
he makes the reading, and what he is talking about when he is on the stand
giving the testimony, and especially when he is being cross examined by a
good defense attorney.
I think that I am aware of most every criminal case tried in the United
States involving an accusation of violating an equivalent opacity standard.
My knowledge of the cases comes from either reading them in the law reports
or my personal involvement in the cases. Unfortunately, I must tell you that
in each and every one of those cases the smoke reader had difficulty in
explaining the relationship of a white or other light colored plume to a
number on the Ringelmann Chart designed for measuring black smoke*
277
-------
When the smoke reader falters, or stammers around in an attempt to equate
a white plume to a black smoke chart, it takes only a little ingenuity on the
part of the defense attorney to turn otherwise respectable testimony into
devastating cross examination. It is not then an easy matter to rehabilitate
the witness.
Up to this point the paper has been devoted mostly to criticism of the
"equivalent opacity" standard. Now a'few words as to the new standard.
The first sentence of the single standard is really a definition of the
word "opacity". This is a word of art and as such must be defined somewhere
in the law. By making it a word of art it removes any possibility of specula-
tion by attorneys, judges or anyone else as to what the meaning is. I have
included it as part of the standard for the sake of brevity in the preparation
of this paper. If the rules, regulations, statutes or code that you are work-
ing with contains a section on definitions, this definition properly should be
in that section.
The standard is equally applicable to plumes of any color, particulate
matter whether solid or liquid or any gas streams that are visible. In other
words, it is applicable to any "visible emission".
If you are questioning the retirement of Ringelmann and the application
of the new standard to black smoke, let me here point out that Ringelmann really
measures the closeness of given soot particles in a given quantity of smoke
plume; the greater the number of given particles in a given quantity-of smoke
plume, the blacker the plume. And so with the opacity standard, the greater
number of given particles in a given quantity of ga3 stream, the greater the
obscuration of the background object through the plume and thus the higher
percentage. The same theory is equally applicable to a white or any other color
plume.
27 8
-------
Put in another way, percentage of obscuration simply means the percentage
by which the observers view is blotted out or blocked by the particulate
matter in the plume.
The new standard is advantageous to the control agency from an economic
standpoint. With the old dual standard, it is necessary to train smoke
readers as to both standards; that is Ringelmann for black or shades of gray
and equivalent opacity for any other color plume. It is also necessary to have
a smoke generator capable of producing black and other colored plumes.
Obviously it costs more because of the additional time involved to train for
two standards than for one.
All of the equivalent opacity standards that I am faimilar with limit
emissions to a certain given equivalent Ringelmann shade without qualification
except a short permissible time limit in excess thereof, and generally
uncombined water is exempt. On the other hand, it is general practice for
opacity observations to be made with the sun at the observers back as this gives
the lowest reading. This is not necessary from a legal standpoint; however,
it is always nice from a prosecution's standpoint to make out a violation
and obtain a conviction giving the polluter all the advantages in the case.
For further information on this and other strictly legal questions, I suggest
you read People v. Plywood Mfg's of California, 137 Cal App 2d 859 , 291 P2d
5^7 (1955) • Admittedly th® case is now sixteen years old, but is still good law.
Even though the new standard is much superior in many ways to the equiva-
lent opacity standard, it is not Utopia. Both standards have the same inherent
problem in that enforcement is dependent upon an eyeball test. This problem
is overcome to a large extent, however, in litigation by good straight forward
testimony describing in depth the training program and a recorded history of
many emission observations, or evaluations as they are sometimes called. The
more the better. Without taking the time to point out the detrimental elements
ol" any other type record, I here set out an example of the most acceptable type
of record or log.
279
-------
Observation record of:
Date
Name of Plant
Elapsed time Cumulative total
2/28/71 Joe Bloe Co
45 rain.
45 rain
1 hr. ^5 min
3/8/71
Hanky Panky, Inc.
60 min.
3/18/71 Rinky Dinky Apt3
30 rain
2 hr. 15 rain
It is readily apparent the record is easily made and can be done economi-
cally in a hard bound record book costing a dollar or less. The record is
of tremendous value to the prosecutor In establishing the credibility of the
eyeball test. Of course the record should be permanent, that is, made with
ink so there will be no erasures.
A standard that exempts water vapor, or uncombined water, and they moat
all do, can present a different kind of problem in an opacity violation.
When reading this type of plume, the violation, if any, must be established
downstream from the point of evaporation. The problem is that any person
trained in meteorological sciences can give uncontradicted testimony there is
no scientific method of determining the point of evaporation. This type of
testimony should be anticipated and overcome by two methods.
First of all, the smoke generators used for training purposes should
have a method of injecting water vapor into the emission so that reading wet
plumes, as well as dry plumes, is a part of the standard training procedure.
Secondly, and this is for the prosecuting attorney, by careful cross
examination of the meteorological science witness, he can be made to say he
witnessed a water vapor plume issuing from a boiling teakettle or other vessel
and that he could tell the point where it became invisible. At this point
who cares about scientific methods; the scientific defense witness has just
raised the credibility, expertise if you please, of the testimony of the smoke
reader.
280
-------
I would be remiss if I didn't advise you of the problem of the "famous
green book". This is the title by which it is known in Oregon because the
book has been used extensively in litigation in Oregon. The real name of the
book is "Optical Properties and Visual Effects of Smoke-Stack Plumes" and is
Public Health Service Publication No. 999-AP-30, published in 1967.
The publication is the report of a cooperative study conducted by Edison
Electric Institute and the U.S. Public Health Service. Except for the 1^- page
"Conclusion", the publication goes into detail of the unreliability of the
eyeball test for evaluating smoke plumes. It is my opinion it should be
required that every smoke reader be thoroughly familiar with contents of the
publication. And this is a must for every prosecutor so that he can
effectively overcome the adverse use of it through cross examination and
rebuttal testimony.
It is my opinion the question of the accuracy of the eyeball test of
evaluating either black or non-black plumes will never be actually settled
even though the method has been judicially approved in Oregon, Washington and.
California.
What is really needed is the development of a scientific instrument for
measuring effluent plumes. To be sure there are currently such instruments
but they all have one common fault and this is the necessity of a stack probe.
It is impossible to insert a stack proble without permission from the plant or
through use of a search warrant. When either of these avenues of approach is
employed, management is fully cognizant that sampling is being conducted and
can either alter the operation or shut the plant down and defeat the test.
I certainly don't mean to imply the management of every plant would do this,
but unfortunately, not all management is blessed with a halo.
So when I talk about development of a scientific instrument, I mean one
that will operate effeciently and accurately while situated off the premises
281
-------
of the source of emission. This is a must so that the test can be made
surreptitiously as to those sources where the management does not wear white
hats.
In all probability, the laser should be developed to perform this function.
This recommendation is contained in the "famous green book" and also comes from
other sources.
Because it seems likely sale of the instrument would be in »m»'1i numbers,
probably not in excess of <500, private industry may not be interested in
spending necessary amounts of money for the research and development. fMa
leaves only one source for the money, a grant by the Environmental Protection
Agency, Air Pollution Control Office.
282
-------
THE OPACITY WITNESS
(FINAL DRAFT)
Prepared By
Kenneth B. Malmberg
Division of Stationary Source Enforcement
U.S. Environmental Protection Agency
Washington, D.C.
20460
283
<3-2
-------
THE OPACITY WITNESS
Introduction
This guideline is intended to assist the EPA employee in a civil or
criminal action in which he may be called as an opacity witness. It summarizes
the EPA witness1 presentation of opacity testimony in a format which can easily
be interpreted by judges, juries, and counsel. This testimony can be supple-
mented by exhibits which clarify and present opacity data in its simplest and
most easily understood form.
While other witnesses may be called to testify on rules, regulations,
administrative procedures, process variables, or unit process operations,
testimony on opacity evidence may be the most far-reaching in terms of varieties
of issues discussed. Because opacity is one of the most universally applicable
emission regulations, the EPA employee presenting opacity evidence should be
prepared to address diverse and sometimes controversial issues.
284
J-3
-------
VII. THE OPACITY WITNESS
A person testifying on visible emission evalua-
tions (i.e., opacity evidence) should be familiar with
the development of visible emission evaluation procedures and
their application. As a witness, you may be questioned in
the following areas:
1. Specific observation procedures you followed
in the field.
2. The accuracy of the opacity method as applied to
industrial sources.
3. The accuracy of Method 9 as promulgate^ in the
Federal Register on December 23, 1971, and revised
November 12, 1974.
4« Revisions to the method including the following
specific issues:
a. The averaging approach to determining compliance;
b. The potential effects of changes in observer
position and weather conditions;
c. The potential effects of water vapor on opacity.
285
J-4
-------
12
5. The training you received for certification by EPA.
6. The compilation of opacity data and its presentation.
It may be necessary in presenting opacity evidence that
you discuss some or all of the above issues as an expert witness.
A. ISSUES
A major substantive area into which both sides will
delve is the one of qualifications. You must always present
an accurate summary of your qualifications to your attorney
prior to your testimony. Your summary of experience differs
from a resume in that it describes those specific qualifica-
tions which enhance your credibility and general ability
to discuss opacity issues and to answer questions on the
subject of visible emissions with a high degree of confidence.
Include a summary of all relevant work performed either in
your present position or elsewhere, with reference to previous
expert testimony, writing in the field, and any public
presentations you have made regarding opacity techniques or
procedures. This summary should be brief, and not over one
page in length, in most cases.
On a motion to postpone the trial the judge may change
the original trial date so that witness qualifications can
be examined both by himself and by the defense. Should
286
-------
3
this occur you may meet in the judge's chambers for a
verbal examination, in which the judge may ask various
questions about your experience, training, background, and
perhaps some questions relevant to the case at hand.
Other relevant issues are:
1. Training - The adequacy of observer training is
always subject to scrutiny. In anticipation of these
questions, you can be particularly helpful before trial.
Well before trial you should ask your attorney whether or
not a demonstration of visible emission evaluation proce-
dures could be of benefit to the court. This demonstration
could include a visual demonstration to the judge and the
jury of correct opacity reading technique. It could be
supplemented by films, pictures, and in-court demonstrations
illustrating equipment used in visible emission certifica-
tion procedures. At the very least this area of field
procedures should be fully discussed between you and your
attorney prior to any cross examination by the opposing
counsel.
J-6
287
-------
4
2. Accuracy - The Portland Cement Association's
remand response1 clarified EPA's position regarding the
accuracy of visible emissions evaluations, in this document
EPA tests showed an overwhelming majority of observations
made by certified opacity observers were well within the
standard for certification. You should stress the point
that these data also show that the error factor is biased
downward (that is, in favor of the emission source), when
evaluations are made under meteorological conditions con-
sidered less than ideal. If these points are clearly
presented little doubt can remain regarding the accuracy of
opacity evaluations.
3. Method 9 - The opacity method itself is also a
2
subject for witness examination. The method as promul-
gated in the Federal Register is clear in its applicability
procedural, and testing segments. The averaging concept
Response to Remand by U.S. Ct. of Ap., re "Portland
Cement Association v. Ruckelshaus", ESED, EPA, RTP,
North Carolina 27711. EPA 450/2 74023. November 1974.
2
Visual Determination of the Opacity of Emissions from
Stationary Sorces, 40 CFR 60, Appendix A.
288 J-7
-------
5
is particularly adaptable to continuous emission from
an emission source. As described in Section 2.5, the
averaging process must be thoroughly understood prior
to any attempt at presenting evidentiary data derived
by its application. Therefore your presentation must
stress its basic simplicity. You must make every effort
at presenting your data clearly, accurately, and in a
simple manner. Included among changes to the method
is a six-minute averaging procedure. You must insure
that these revisions are thoroughly explained and
understood by everyone prior to your examination on the
stand.
4. Data Presentation - Implicit in the presenta-
tion of opacity evidence is the citation of opacity data
which is claimed to be over the standard called for in
the regulation. You can assist your attorney by clari-
fying your technical reports and providng a concise format
for presentation of data which clearly shows:
a. the alleged violations;
b. the extent of those violations (both the
number of evaluations and their excess over
the standard);
J-8
289
-------
6
c. the six-minute average of those evaluations,
if applicable and;
d if applicable, the number of evaluations
exempted by the particular state regulations.
In addition, you could supply a narrative report tying
in the particular state opacity regulation to the reading
made. For example, many state opacity regulations have
differing applicability requirements. Clearly under-
standing these regulatory applications is important for
development of an acceptable opacity enforcement case
5. Review of Testimony - Before taking the stand
as a witness, you should review the testimony of prior
witnesses with your attorney if possible. This will assure
your complete understanding of facts testified to previously.
B. TESTIMONY
Since you have presented a description of your qualifi-
cations, questions in this regard will be brief and to the
point. The intent of these questions is to put aside any
suspicions in the minds of the judge or the jury. Your
answers should be as accurate and as well founded as was
your original discussion of this information **he
J-9
290
-------
7
attorney handing the case. You may then be asked to discuss
the procedures you followed in evaluating visible emissions
from this source, with particular emphasis on any deviations
from procedure which you found necessary. As an expert you
may be asked your opinion freely, if asked, based upon your
most professional judgment. On cross examination the defense
attorney will try to discredit testimony you previously gave
by placing you in the position of making a contradictory
statement or taking a position which is not based on your
specific areas of expertise. It is important that you do not
depart the narrowed area of your expertise. You should avoid
discussions based on particle sizing, particle distribution,
particle reflectance and dispersion, process operations,
process effluent characteristis, process efficiencies,
etc., unless you feel fully qualified to discuss these
items.
The defense may also include remarks based on irrelevant
subjects which may be disconcerting to you at the start, but
as an expert witness you must remain calm and confident that
your counsel will not allow questions of this type to detract
from the case. Should it occur consistently, the judge will
0-10
291
-------
8
disallow a line of questioning if it appears inappropriate or
inconsistent. The judge himself may wish to take up examina-
tion of a witness already on the stand, if he sees fit. This
will occur if the judge wants a particular point developed or
expanded to clarify the facts for the jury.
Juries deserve discussion here because they are
asked to decide questions of fact. Your testimony will
for better or for worse influence that decision. Both the
type of evidence regarding opacity evaluations and confusing
defense presentations sometimes confuse a jury. Therefore
your presentation of opacity evidence must not place your
testimony in a negative light try a jury which may already be
confused. Do not try to judge their ability to digest your
testimony. The questions, your answers, the courtroom setting
are all influences on the jury's final decision. Therefore,
your answers should be structured as simple and precise
as possible for a layman's comprehension.
There are many excellent discussions on the correct
demeanor of a witness while on the stand., One which is
outstanding is that contained in Part III of the course
manual for "Air Pollution Field Enforcement," Air Pollution
Training Institute, U.S. EPA, 1972.
Questions in chambers by the judge with both opposing
parties present has become more signficant as a part of
-------
9
judicial proceedings involving technical issues. Judges
must have substantive answers to their inquiries based
on their sometimes limited knowledge of the technical
portion of a case which is based on air pollution viola-
tions. These sessions owe their productivity to the fact
that the unique knowledge of the parties involved will allow
a more freewheeling and intensive discussion on a given
point, beyond the hearing of the jury and courtroom. Since
such questions can assume the same form as questions asked
on the stand, as an opacty witness you should conduct
yourself as if you were under oath without any bias toward
either side and answer the questions as clearly and com-
pletely as possible. Questions may be asked to determine
applicability of certain rules of law, and the judge may
adjourn the court until he can make a decision about a point
of law. A transcript will also be made of these in-chambers
questions and answers.
To avoid any element of surprise you should always
conserve your answers if you suspect even remotely that by
continuing beyond a certain point you would demage the
case. Prosecutors always stress this point with witnesses,
and in giving opacity testimony it is doubly important
because of the nature of the evidence presented, and the
technical reports supporting this evidence.
Cross examination may be followed by redirect examination,
to clarify any questions raised under cross examination. The
primary reason for cross examination by defense counsel is
to refute your testimony previously given and if possible to J-12
293
-------
10
characterize testimony thus given as being untrustworthy or
wrong. When an indirect question is asked by the defense you
should demand clarification before answering. You should
qualify you answer, if the question is not presented completely.
The following question, in various forms, can be expected
by you as an opacity witness:
1. Direct Examination
a. General
(1) What is your name?
(2) What is your present job? (How long have you
had it, etc.)
Various other questions of this type will serve to put
you at ease and introduce your qualifications to the court.
b. Specific
(3) What do you do when you evaluate a source?
(4) Do you evaluate the emissions before you
enter the source?
(5) Do you inform the source of your evaluation
in advance?
294
J-13
-------
11
(6) Do you always enter the source?
(7) Why do you evaluate the emissions from
a source before entering?
(8) Do you have a regularly established procedure
you follow?
(9) What are the procedures followed in Method 9?
(10) What do you do if you cannot follow procedure
(e.g., observer angle to plume)?
(11) Where do you look at the plume to evaluate it?
(12) What do you do when you "read" a plume?
(13) What is "opacity"?
(14) How does "opacity" differ from Ringlemann?
(15) Have you seen the Bureau of Mines publication
#8333? (The Ringlemann Method).
(16) How long do you "read" a plume
(17) Are you usually alone when reading?
(18) What are the weather conditions on the day
specified?
-------
12
Further questions would deal with specifics of your
reporting findings which may have been previously admitted
to the courts as evidence.
2. Cross Examination
a. State the exact dates and scores of your
certification.
b. Did you notify this source prior to these
evaluations.
c. Why not?
d. Does everyone in your office follow the same
procedures?
e. Have you ever made a mistake?
f. Are photographs taken at the same time as
your readings?
Specific questions on various features of your report
may be interspersed with reference to various publications,
including the Federal Register. You may take notes or file
copies with you to the stand to refresh your memory.
296
J-15
-------
¦13
Basically, in giving testimony, remember the following
four points:
1. Be truthful in your answers.
2. If you do not know, say so.
3. Be responsive to the question, and if you do
not understand it, say so.
4. Explain your answer when it is in need of either
clarification or qualification.
C. CONCLUSION
Direct evidence is communicated by those having actual
knowledge of the facts by means of their senses. Your
testimony is regarded as direct evidence if it most certainly
exhibits the true state of facts. Your powers of observation
as a visible emissions observer are among the primary means
available to the prosecutor for providing evidence of opacity
violations.
NSPS and SIP opacity regulations vary in their requirements,
both in minimum opacity requirements over a period of time, as
well as exemptions based on the age of the emission source, type
of process, precise location of the emission source, and process
malfunctions as described in the particular regulations.
J-16
297
-------
14
Supplementing your direct testimony is the presentation of
exhibits, documents, and demonstrations. This could include
your field reports, photographs and other information about
the source. Any sketches of the facility which you drew and
which were enlarged to a size such that they could easily be
shown to a jury from a tripod arrangement may also be useful
as possible exhibits. Drawings are especially helpful in that
most of the requirements and procedures for visible emission
evalutions are based on meteorological and physical characteris-
tics relative to a particular point source. Such parameters
as distance to the stack, direction of the wind, position of
the observer, relative position of the sun, meteorological
conditions, and other relevant visible emission evaluations
criteria can be easily, clearly, and concisely illustrated by
means of a large scale drawing or sketch.
Relevant books in the field of opacity include the course
book for the EPA training course Visible Emissions Certifica-
tion AP-30 - Optical Properties and Visual Effects of Smoke-
stack Plumes; Method 9 of Part 60 - Visual Determination of
the Opacity of Emissions from Stationary Sources; and the
Bureau of Mines Circular No. 8333, title Ringlemann Smoke
Chart. Any of these sources may be used to show the jury
specifications, instructions, and generally accepted methodo-
logy of opacity observation. Familiarity with these publica-
tions will enhance your credibility as a witness.
-------
15
VIII. SUMMARY
The complex nature of many visible emissions does not
reduce their compatibility with opacity evaluation techniques.
Details stressed in this manual are necessary to fully explain
techniques and procedures of successful enforcement of opacity
regulations against stationary source of visible emissions.
The elements of case development, chain of custody
procedures, and other relevant portions of opacity evaluation
techniques have not been included. Guidance on these and
other issues is forthcoming. However, following the above
visible emission procedures should provide a sound foundation
for agency development of a comprehensive, enforceable, visible
emissions program for your area.
299
-------
BIBLIOGRAPHY
1. Policy with Respect to the Presentation of Testimony by EPA
Employees - EPA Order 1000.7, April 6, 1972.
2. A Primer for EPA Employees: Presenting Scientific Evidence by
Tames A. Rogers, EPA, Washington, D. C. September 1974.
3. How to Prepare Yourself for Cross Examination, by Jack E. Horsley,
The Journal to Legal Medicine, January/February 1974.
4. Hints for Expert Witnesses, Paul A. Humphrey, July 1973. Meteorology
Laboratory, RTP, U. S. EPA.
5. Field Operations and Enforcement Manual for Air Pollution Control,
CT7T7 Weisburd. August 1972. FRA Contract iUfA 70-122, APTD 1100,
1101, 1102, 3 Volumes.
6. Air Pollution Field Enforcement, Section III. Air Pollution Training
Institute, U. S. EPA, 1972.
7. Conner, E. D., Hodkinson, J. E., Optical Properties and Visual Effects
of Smoke-Stack Plumes, U. S. Dept. of H.E.W., Public Health Service
Publication 999-AP-30, Cincinnati, 1967.
8. Rinqlemann Smoke Chart, Staff, Bureau of Mines. U. S. Dept. of Interior.
Information Circular 8333, May 1967.
9. EPA Visible Emission Inspection Procedures. S-24, by Kenneth B. Malmberg,
Stationary Source Enforcement Division, U. S. EPA, August 1975.
10. Guidelines for Evaluation of Visible Emissions, R. Missen and A. Stein,
April 1975, EPA Contract #68-02-1390, Publication ^340/1-75-007, Washington
D.C.
300
J-19
-------
SECTION I: MISCELLANEOUS OPACITY INFORMATION
Page No.
1-1. Table: IGCI Reports Consensus on Industrial Emission
Levels Producing "Clear" or "Near Clear" Stacks, IGCI,
APCA Journal, July 1973 ............. .. 303
1-2. Visible Emissions Observation Form, EPA Division of
Stationary Source Enforcement - Proposed Form, August, 1981 . 305
1-3. "Smoke: A Fable," Burl E. G111fland and Robert K. Roney,
Journal of Irreproducible Results, 1978 307
-------
E-l
Reprinted from APCA Journal, Vol. 23, No. 7, July 1973
IGCI Reports Consensus on Industrial Emission Levels
Producing "Clear" or "Near Clear" Stacks
A survey of the member companies
of tin* Industrial (las i'leanuu; In-
stitute (K'iC'I) has developed tin* follow-
ing consensus as to emission levels which
jiMierally produce or result in clear or
near clear stacks for various industrial
applications.*
Understandably, reports IGCI, there
are numerous qualifications as to the
accuracy of the data under all circum-
stances. The reader is therefore cau-
tioned to consider any special factors
applicable to his own operational con-
dition* ill using these quideiines.
The data are for general informational
use to those concerned with the appli-
cation of air pollution control equip-
ment. Neither the IGCI nor any
of its members make any representation
as to the accuracy of the data as a(>-
plied to any specific installation.
Variables such as path length, angle
of incidence of li^ht, moisture content
of effluent, weather conditions, and
proce.-s changes may significantly change
stack appearance irrespective of the
weight of particulate emission.
User* and manufacturers of air pollu-
tion control equipment arc urged to
submit data to IGCI relating to their
own experience* with emission levels
which result in vicar or near clear ntacks.
Connrming or contradictory data, or
• «•!»..incnts in general will be welcomed
i y the Industrial Gas Cleaning In-
stitute, 11 lf> Summer Street, I'.O. Box
1J33. Stamford, Conn. 0G9O4.
Industrial process amission levels generally expected to produce visually dear
(or near clear) stacks, excepting condensed moisture plume.
* A paper wa> presented at the May
IS, l'.iTii, nicotine in Baltimore of the
Smith Atlantic Stale* Section of APCA
outlining the rca-oning which prompted
the KlCI Product Division to prepare
l hi- consensu* data. The paper was
presented by Kiigcnt: P. Stastiiv, manager,
Product Planning and Development,
I .nviK'iiinenlal Systems IJiv., Koppers
Company, Inc. :iuti cinTenity vice pres-
ident, KiCI. TIkm iiI ire pa|H'r will appear
in (lie inectinn priicvolinK*.
Electrostatic precipitators.
Estimated
Wet scrubbers.
fabric filters.
humidity*
Grains/SCF Dry*
industrial process
Grains/ACF at stack exit
Vol. %.
(SCF ® 70-F and
classification
gas temparaturo, *F
HsO
14.7 Psia)
Utilities and industrial
power plant fuel-fired
boilers
Coal-pulverized
0.020 (ft1260-320
5
0.029-0.031
Coal-cyclone
0.010 ($260-320
5
0.014-0.016
Coal-stoker
0.050 @-350-450
5
0.080-0.090
Oil
0.003 @300-400
8
0.0047-0.0053
Wood and bark
0.050 @400
10
0.090
Bagasse
0.040 @400
10
0.072
Fluid coke
0.015 @300-450
Ory
0.022-0.026
Pulp and paper
Kraft recovery
Boiler
0.020 @275-350
30
0.040-0.043
Soda recovery
Boiler
0.020 ©275-350
30
0.040-0.043
Lima sludge kiln
0.020 @400
30
0.046
Rock products (Kilns)
Cement—Dry
0.01S 0450-600
3
0.026-0.030
Wat
0.015 @450-600
25
0.034-0.040
Gypsum
0.020 @500
25
0.048
Alumina
0.020 @400
•20
0.040
Lima
0.020 @ 500-600
8
0.039-0.050
Bauxite
0.020 @400-450
3
0.035-0.037
Mag. oxide
0.010 @550
20
0.024
Steal
Basic oxygan
Furnace
0.010 @450
20
0.021
Opan hearth
0.010-9.015 (7450-600
8
0.023-0.027
Electric furnace
0.015 @400-600
5
0.026-0.032
Sintering
0.025 @300
10
0.040
Ore roa stars
0.02 @400-500
10
0.036-0.040
Cupola
0.015-0.02 @ 250-400
20
0.029-0.036
Pyrites roaster
0.02 @400-500
10
0.036-0.040
Taconite roaster
0.02 @300
10
0.032
Hat scarfing
0.01-0.015 @150-250
25-40
0.019-0.025
Mining and metallurgical
(Non-farrous)
Zinc roaster (Ora)
0.010 @450
5
0.018
Zinc smelter (Melting)
0.010 @400
5
0.017
Copper roaster
0.010 @500
5
0.019
Copper ravarboratory
furnace
0.015 @550
S
0.030
Copner converter
0.010 @500
5
0.819
Aluminum—Hall process
"0.025—0.075 @300
5
0.075
Soderbetg process
0.001-0.003 @200
25
0.003
llmenite dryer
0.020 @300
2S
0.038
TiOj process
0.010 @300
5
0.015
Molybdenum roastar
0.010 @300
5
0.015
Ore beneficiation
0.020 @400
5
0.034
Miscellaneous
Refinery catalyst
0.032
regenerator
0.015 @475
15-20
Municipal incinerators
0.015 @500
20
0.034
Apartment incinerators
0.02 @350
20
0.038
Spray drying
0.01 (f>.400
25
0.022
Precious metal refining
0.01 @400
5
0.017
' Based on estimated water vapor content of hot gases
303
-------
VISIBLE EMISSION OBSERVATION FORM
SOU RC
BSERVATION DATE
START TIME
STOP TIME
ADDRESS
sec
30
4 S
30
45
STATE
ZIP
32
CITY
SOURCE ID NUMBER
PHONE
34
PROCESS EQUIPMENT
OPERATING MODE
35
CONTROL EQUIPMENT
OPERATING MODE
36
DESCRIBE EMISSION POINT
38
HEIGHT ABOVE
GROUND LEVEL
HEIGHT RELATIVE
TO OBSERVER
39
DISTANCE FROM OBSERVER
OIRECTION FROM OBSERVER
40
41
DESCRIBE EMISSIONS
42
EMISSION COLOR
PLUME TYPE: CONTINUOUS ~
FUGITIVE Q INTERMITTENT Q
43
WATER DROPLETS PRESENT
NO C
IS WATER DROPLET PLUME
ATTACHED ~ DETACHCO ~
44
49
AT WHAT POINT IN THE PLUME WAS OPACITY DETERMINED
46
DESCRIBE BACKGRO
BACKGROUND COLOR
SKY CONDITIONS
49
WINO SPEED
WIND OIRECTION
20
SO
ambient temperature
RELATIVE HUMIOITY
21
SOURCE LAYOUT 3KETC
DRAW NORTH ARROW
22
23
S3
S4
24
EMISSION POINT
S 5
2S
96
26
57
27
S«
29
SUN S£A DOW LINE
60
30
NUMBER OF READINGS ABOVE
AVERAGE OPACITY FOR
HIGHEST PERIOD
OBSERVERS POSITION
% WERE
COMMENTS
RANGE O F OP AC IT Y REAOINGS
MAXIMUM
MINIMUM
OBSERVER'S NAME (PRINT)
OBSERVER S SIGNATURE
DATE
ORGANIZATION
HTTvE HECKIVEO A COPY OF THESE OPACITY OBSERVATION*
OATE
DATE
VCRIKIEO OY
TITLE
t,PA TR 0101
305
-------
1-3
Reprinted from Journal of Irreproducable Results, Vol. 23, No. 4, 1978
SMOKE: A Fable
BURL E. GILLILAND and ROBERT K. RONEY
University of Tennessee
Memphis State University
Center for Advanced Graduate Study
Memphis, Tennessee
Once there was a large island inhabited by several tribes,
all of which resided in villages on the periphery of a huge,
dense forest which made up the central portion of the island.
Usa was one of the largest of these village states and was
considered by many to be the most powerful. Although
Usaites considered themselves to be peace-loving people,
they were often in rock-throwing conflicts with their biggest
rivals, the Ussros and the Prochs. Most often these rivals
would fight over other tribes' villages, ostensibly helping a
smaller friend against a larger aggressive enemy.
For some time the Usas had enjoyed a relative period of
peace after having successfully fulfilled a commitment of
assistance to a small tribe in the south-east portion of the
island. In addition to the fact that this commitment had been
costly over a long period of time, in terms of both men and
rocks, a new problem had arisen at home which was begin-
ning to demand the attention of the Usaites; thus, the reso-
lution of the conflict was particularly welcome. Several
years earlier a small column of smoke had been spotted by
the Usaites coming from some distance in the dense forest.
Passing comments were made about it at the time, but
speculation soon subsided and the strange phenomenon
forgotten. From time to time the smoke would reappear,
always to fade into oblivion again. Recently, however, the
column of smoke was continually present and seemed to be
getting larger. A few alarmists in the tribe grumbled that
the smoke was occasionally blocking out the sun in the late
afternoon and that something ought to be done about it.
Naturally, the sensible majority scoffed at these complaints
and went about their usual business, which was mostly
digging for rocks or watching the surfing contests from
their new seaside stadium.
Then, at precisely one leap year's length past the mid-
point of Usa's proudest century, the prevailing wind on the
island shifted; and the smoke, which had become blacker
and more acrid, was blown in some quantity toward the
village of the Usas. Alarm was now widespread, and the
village council was called into hasty session. The village chief
was incensed that the council had failed to react to the need
for a program of smoke elimination which he had proposed
during the election campaign the previous year. Actually,
the chief had merely mentioned the smoke, along with several
other "minor" problems, in a few of his campaign speeches.
Vague promises to do something about it, if he were elected,
had been made. Since taking office nine full moons earlier,
he had not mentioned the topic even though the council met
with him every week whether or not there was a need to
meet. The councilmen, especially those of the opposition
party, fought back with counter-charges; and the debate
raged on for many days.
One day the smoke was so bad that the council meeting
had to be adjourned so the members could go down by the
sea where the smoke was not quite as thick. This dramatic
turn of events caused the council to realize that they had to
take action instead of merely bickering about who was to
blame for procrastinating. One councilman suggested a
study of the source of the smoke; another, Mr. R. Wing,
vowed the smoke was a Proch plot; but their level-headed,
action-oriented colleagues quickly pushed aside these fool-
ish suggestions in favor of the proposed solution of the
village chief. It was easy to see that the wind was bringing
the smoke into the village and that until such time as the
Gods saw fit to reverse the winds, it must be incumbent
upon the people of Usa to artificially create wind currents to
24
307
J.I.R. 1978
-------
lake the smoke awav from the village. The chief commis-
sioned the village scientists to develop a smoke inversion
device. A team of the best minds in the village — the same
group who had developed a new light-weight rock launcher
the previous year — developed a manually operated fan
for this purpose. If a sufficient number of these fans could
be operated along the border of the forest, the scientists
reasoned, the smoke could be blown away. Gleefully the
council passed with only two dissenting votes the appropri-
ation to fund the building and installation of the fans. They
also passed a law requiring those citizens who lived nearest
the forest to man the fans once they were installed.
A small group of the citizens who resided nearest the
forest grumbled at the discrimination in the law which re-
quired them to man the fans while most of their fellow
villagers were going about other pursuits, such as watching
the big surfing matches. Most, however, dutifully manned
their assigned fans, and for a short time it seemed as though
the strategy was helping. The air cleared somewhat, and the
sun could again be seen most of the day. One of the two
councilmen who had voted against the "fan law" pointed
out that the wind had shifted, the smoke was not boiling up
quite as thickly, and (hat the council had confused real and
pseudo issues; but the people knew that this was only sour
grapes and an attempt to cover up his own stupid stand on
the issue.
Soon the wind shifted again, and the smoke became
worse than ever despite the fans. Some of the fan operators
became discouraged, disgruntled, and even claimed to be ill.
Many quit fanning altogether. The council seized upon their
niggardly reluctance to do their patriotic duty and passed
another law making it a felony for those living near the
forest to refuse to operate their assigned fans. Reluctant
fanners were put into jail, and soon the fans were all in
operation again. Once more the council congratulated it-
self for solving the smoke problem. An election was coming
up soon, and they were sure to be reelected now that they
had taken such decisive and effective (cough, cough) action.
Somehow they were (cough, cough) able to ignore that in
fact the fans were not working, and the smoke (cough,
cough) was getting worse.
Very early in the smoke controversy a small group of the
Usaites who lived near the forest had appealed the "fan
law" to Magistrate W. R. Wren, the village seer, who
served as a mediator and interpreter of legal conflicts in the
village. This wise, old man had agreed with the appellants
that the council had been unfair to make them the sole group
to be burdened in connection with the smoke problem. He
had also observed that it seemed to him that there was a
greater problem than merely the smoke, and he felt there
should be a thorough investigation into the cause of the
smoke. The village chief and the council responded rather
curtly that this was typical of the fuzzy thinking which had
been coming from the seer recently. They refused to change
their course in the solution of the smoke problem.
The election campaigns of the village chief and a sub-
stantial portion of the council members now came to revolve
around the smoke as an issue. "We must abolish smoke
(cough, cough) from our land forever, even if it requires a
(cough, cough) constitutional amendment," shouted candi-
dates throughout the village. Very few candidates argued
with this point of view, and those who did were soundly
defeated at the polls. But the smoke was worse. With great
haste the new council and the reelected village chief pre-
pared and passed the proposed constitutional amendment,
outlawing smoke once and for all. Once again the people
cheered (cough, cough), more faintly this time through the
billows of smoke rolling across the village from the forest.
The morning after this final decisive action there were no
sounds in the village — not even a cough. THERE WERE
ONLY EMBERS.
J.I.R. 1978
308
25
------- |