600176023
SCIENTIFIC AND TECHNICAL DATA BASE
FOR
CRITERIA AND HAZARDOUS POLLUTANTS
1975 ERC/RTP REVIEW
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND MONITORING
HEALTH EFFECTS RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, N.C. 27711
u.
-------
DISCLAIMER
This report has been reviewed by the Health Effects Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
11
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES v-vi
LIST OF TABLES vii-ix
PREFACE x
1. SUMMARY 1
1.1 GENERAL DISCUSSION 1
1.2 INSTRUMENTATION 2
1.2.1 Nitrogen Oxides 2
1.2.2 Carbon Monoxide 3
1.2.3 Hydrocarbons 3
1.2.4 Photochemical Oxidants 4
1.2.5 Particulate Matter 5
1.2.6 Sulfur Oxides 6
1.2.7 Asbestos 6
1.2.8 Beryllixm 7
1.2.9 Mercury 7
1.2.10 Methods Standardization and Performance
Evaluation 7
1.3 ENVIRONMENTAL APPRAISAL 7
1.3.1 Emissions 8
1.3.2 Ambient Concentrations 9
1.4 TRANSFORMATION AND TRANSPORT 10
1.5 EFFECTS 13
1.5.1 Human and Animal 13
1.5.2 Materials 27
1.5.3 Ecology 29
1.6 CONTROL TECHNOLOGY 29
1.6.1 Stationary Sources 29
2. INTRODUCTION 35
3. MEASUREMENT 37
3.1 INTRODUCTION 37
3.2 PRODUCTION OF TEST ATMOSPHERES 45
111
-------
11
TABLE OF CONTENTS (continued)
Page
3.3 OXIDES OF NITROGEN MEAStJREMENTS 47
3.3.1 Ambient Air 47
3.3.2 Stationary Sources 58
3.3.3 Mobile Sources 60
3.4 CARBON MONOXIDE MEASUREMENT 60
3.4.1 Ambient Air 60
3.4.2 Stationary Sources 72
3.5 HYDKX&KBCNS 72
3.5.1 Ambient Air 72
3.5.2 Stationary Sources 80
3.5.3 Mobile Sources 81
3.6 PHOTOCHEMICAL OXIDANTS 84
3.6.1 Ambient Air 84
3.6.2 Stationary Sources 99
3.7 SUSPENDED PARTICUIATE 102
3.7.1 Calibration 102
3.7.2 Ambient Air 103
3.8 SULFUR OXIDES 105
3.8.1 Ambient Air 105
3.8.2 Stationary Sources 106
3.8.3 Mobile Sources 107
3.9 ASBESTOS 112
3.9.1 Stationary Sources 112
3.10 BERYLLIUM 113
3.11 MERCURY 115
3.12 METHODS STANDARDIZATION AND PERFORMANCE
EVALUATION 117
3.12.1 General Discussion 118
3.12.2 Particulates 118
3.12.3 Sulfur Dioxide 118
IV
-------
Ill
TABLE OF CONTENTS (continued)
Page
3.12.4 Carbon Monoxide 119
3.12.5 Nitrogen Dioxide .* . . . 120
3.12.6 Stationary Sources 121
3.13 REFERENCES 123
4. ENVIRONMENTAL APPRAISAL 126
4.1 SOURCES 126
4.1.1 National Emissions Report 126
4.1.2 Nitrogen Oxides Emissions (NO ) 129
4.1.3 Carbon Monoxide x 137
4.2 TRANSFORMATION AND TRANSPORT 142
4.2.1 Oxides of Nitrogen 142
4.2.2 Carbon Monoxide 151
4.2.3 Hydrocarbons 152
4.2.4 Oxidant 153
4.2.5 Particulate 154
4.2.6 Sulfur Oxides and Aerosols 163
4.2.7 Mathematical Modeling 174
4.3 REFERENCES 184
5. Et'ttXJlS 191
5.1 HEALTH 191
5.1.1 General Discussion 191
5.1.2 Nitrogen Oxides 191
5.1.3 Carbon Monoxide 229
5.1.4 Hydrocarbons 237
5.1.5 Oxidants 244
5.1.6 Total Suspended Particulates 275
5.1.7 Sulfur Oxides 299
5.1.8 Mercury 314
5.1.9 Beryllium 324
5.1.10 Asbestos 327
5.2 MATERIALS 343
5.2.1 Nitrogen Oxide 343
5.2.2 Sulfur Dioxide (SO,) 345
5.2.3 Ozone (O3) ... 7 346
-------
IV
TABLE OF CONTENTS (continued)
Page
5.3 WEATHER, VISIBILITY AND CLIMATE 348
5.3.1 Air Quality Trends Analysis 348
5.3.2 Atmospheric Turbidity 349
5.3.3 Visibility 349
5.4 ECOLOGICAL 350
5.4.1 Introduction 350
5.5 REFERENCES 360
6. CONTROL TECHNOLOGY 391
6.1 STATIONARY SOURCES VS., POLLUTANTS 391
6.1.1 Particulate Control 391
6.1.2 Sulfur Oxides Control 392
6.1.3 Nitrogen Oxides Control 394
6.1.4 Hazardous and Other Control 395
6.2 STATIONARY SOURCES VS. CONTROL 396
6.2.1 Combustion 396
6.2.2 Industrial Processes 425
6.2.3 Particulate Control Technology 442
6.3 REFERENCES 449
VI
-------
LIST OF FIGURES
Page
3.1 Oxides of nitrogen chemluminesoent analyzer 55
3.2 Flow scheme for calibration of NO, N02, and NO and
O_ monitors by gas phase titration . . . .x 57
3.3 NDIR analyzer 66
3.4 Dual isotope fluorescence analyzer 70
3.5 Block diagram of dual hydrogen flame detector 78
3.6 Electronic diagram of dual flame detector 79
3.7 Typical ozone source, dilution and manifold system 86
3.8 All-glass midget impinger 89
3.9 Schematic of colorimetric oxidant analyzer 94
3.10 Automated gas phase chemiluminescent ozone monitor 97
3.11 Ultraviolet absorption ozone analyzer 100
4.1 Emissions by categories - United States 128
4.2 Seasonal diurnal patterns of nitric oxide concentrations . . 135
4.3 Seasonal diurnal patterns of nitrogen dioxide
concentratioris 135
4.4 Seasonal pattern of nitrogen dioxide concentration
levels, composite graph of the six CAMP sites 136
4.5 Long-term trend line for nitrogen dioxide concentration
levels composite graph of the six CAMP sites 136
4.6 Maximum 1-hour NO- as a function of initial NO at
various initialTMIC levels (Snog chamber data) 145
4.7 Nitrogen dioxide dosage as a function of initial NO
at various initial NMHC levels (Smog chamber data* .... 146
4.8 Correlation of aerometric data on NO (6-9 am) and
maximum - 1-hour-NO,, (8 am - 5 pm)X(Data from all
CAMP sites) 148
vii
-------
VI
LIST OF FIGURES
Page
4.1 Maximum daily 1-hour-average oxidants as a function of
6-to-9-a.m. averages of. nonmethane hydrocarbons at
CftMP stations, June through September, 1966 through
1968, Los Angeles, May through October, 1967 150
4.10 Number, surface area, and volume distributions of a
hypothetical smog. The linear ordinate normalized
by total number, area, or volume is used so that the
apparent area under the curves is proportional to
the quantity in that size range 158
4.11 National and regional trends in sulfur dioxide, 1970-
1973 171
64
5.1 % difference in mortality (N02 - control) 228
5.2 Difference in % mortality versus length of exposure to
.5 or 3.5 ppm N026 . , 228
5.3 The difference in mean survival time between control ,.
of N0~ mice versus length of exposure to 3.5 ppm N0? . 230
5.4 Effect of particle size on. regional respiratory
particle deposition in nan 291
5.5 Deposition vs. particle size of inhaled particles
in the upper respiratory tract and lungs of guinea
pigs and monkeys compared to people 292
5.6 Effects of inhalation of asbestos in rats 335
314
5.7 Effects of inhalation of asbestos in rats 336
Vlll
-------
LIST OF TABLES
Page
3.1 APPROXIMATION OF THE GASEOUS COMPOSITION OF DRY
UNPOLLUTED AIR 38
3.2 PRIMARY GASES AND VAPORS PRODUCED BY HUMAN ACTIVITIES . . 39
3.3 SECONDARY ATMOSPHERIC POLLUTANTS 39
3.4 ATMOSPHERIC PARTICULATES AND MAJOR PRIMARY AND SECONDARY
AIR POLLUTANTS FORMED BY PARTICLE PARTICIPATION .... 40-41
3.5 NATIONAL AIR QUALITY STANDARD3a 42-43
4.1 EMISSIONS BY CATEGORIES - UNITED STATES 128
4.2 EMISSION FROM MISCELLANEOUS SOURCES 129
4.3 ESTIMATED 1972 NO EMISSIONS FROM STATIONARY SOURCES
(Partial Listing of Equipment Type and Fuel Burned
Accounting for Approximately 70% of Total Emissions) . . 131
4.4 RECLASSIFICATION STUDY - 24-HOUR DATA 137
4.5 ESTIMATED CC»O!NTRATIONS OF LOS ANGELES AEROSOL
PARTICLES BY SOURCE18 (Annual Average, 1966) 160
4.6 PARTICULATE POLLUTION INCLUDING CONSTITUENTS LARGELY
OF SECONDARY ORIGIN 161
?Q
4.7 AEROSOL AND OXLDANT REACTIVITY OF HYDROCARBONS 164
44
4.8 SURVEILLANCE CATEGORIES FOR SULFUR DIOXIDE SITES
(Nutrter of Sites) 170
4.9 SO2 DEPOSITION VELOCITIES 174
5.1 ESTIMATED AVERAGE ANNUAL POLLUTANT EXPOSURE LEVELS IN
THREE STUDY AREAS Ii! CHATTANOOGA, TENNESSEE, 1968-
19724 195
5.2 DAILY NITROGEN DIOXIDE CCNCEOTRATIONS OBSERVED AT
CHATTANOOGA AIR MONITORING STATIONS4 196-197
5.3 MAXIMUM HOURLY NO., OONCENTRATICNS BY COMMUNITY AND
STATION O^TIANOOGA, TENNESSEE, 1970-734
(Micrograins per Cubic Mater) 199
IX
-------
VL11
LIST OF TABLES (continued)
Page
5.4 THREE-YEAR FREQUENCY OF "ANY LOWER RESPIRATORY DISEASE"
AND BRONCHITIS: MODEL PREDICTED RATES FOR CHILDREN
AGED ONE TO 12 YEARS FROM HOUSEHOLDS WITH A HIGH
SCHOOL EDUCATION OR GREATER (1969-1971) 200
5.5 RELATIVE RISK FOR TWO STUDY PERIODS OF ACUTE RESPIRATORY
ILLNESS FOR STABLE FAMILIES LIVING IN COMMUNITIES
EXPOSED TO DIFFERING LEVELS OF AIR POLLUTION38 .... 202
5.6 HOURLY NITROGEN DIOXIDE OMENTRATIONS OBSERVED AT
CHATTANOOGA AIR MONITORING STATIONS8 204-205
5.7 SUMMARY OF DATA FOR RATS IN MORPHOMETRIC STUDY36 .... 215
5.8 NUMBER OF PARENCHYMAL AIR-SPACE UNITS (ALVEOLAR AND
DUCTAL) , RATES OF FORMATION, AND ' IVIDUAL VOLUMES
IN THE MATURING CONTROL AND NO -E*. -6ED RAT LUNGS35 . . 216
5.9 RELEASE OF HISTAMENE FROM LUNG FRAGMENTS IN THE
PRESENCE OF 100 nW SALT SOLUTIONS49'52 . 223
5.10 AMBIENT AIR QUALITY STANDARD FOR CARBON MONOXIDE .... 232
5.11 JUDGMENT THRESHOLD ESTIMATES FOR HEALTH JSbMsSUlS FOR CO . 233
5.12 SIZE OF POPULATION AT POTENTIAL RISK OF POiLLUTANT
INDUCED EFFECT 234
5.13 SOME CLINICAL HEMATOLOGIC AND PATHOLOGIC FINDINGS IN
SIX PATIENTS WITH HODGKBJ'S DISEASE 240
5.14 LEUKEMIA EPIDEMIOLOGICAL STUDY 241
5.15 EXPECTED VS. OBSERVED DEATHS FROM LEUKEMIA IN EIGHT
EUROPEAN AFFILIATES 1962-1971 241
5.16 PATHOLOGICAL CHANGES CAUSED BY OZONE 266
5.17 EFFEC1S OF OZONE ON INFECTIVITY 267
5.18 OZCNE INDUCED TOLERANCE 268
5.19 EXTRA-PULMONARY EFFECTS OF OZONE 269
5.20 EFFECTS OF SMOG 270
5.21 EFFECT OF OZONE ON PULMONARY FUNCTION 271
-------
IX
LIST OF TABLES (continued)
Page
5.22 BIOCHEMICAL EFFECTS 272-273
5.23 BEHAVIORAL EFFECTS 274
5.24 U.S. NATIONAL PRIMARY STANDARD FOR TOTAL SUSPENDED
PARTICIPATES 276
5.25 SCIENTIFIC EVIDENCE FOR THE U.S. NATIONAL PRIMARY
TOTAL SUSPENEED PARTICIPATE STANDARD 276
5.26 SUBSEQUENT SCIENTIFIC EVIDENCE RELATING TO THE U.S.
NATIONAL PRIMARY TOTAL SUSPENDED PARTICULATE
STANDARD 279
5.27 BEST JUDGMENT THRESHOLD ESTIMATES FOR HEALTH EFFECTS
OF TOTAL SUSPENDED PARTICULATES 280
5.28 BEST JUDGMENT THRESHOLD ESTIMATES FOR HEALTH EFFECT
SUSPENDED PARTICULATES BASED ON NEW EVIDENCE 285
5.29 ESTIMATED POPULATION AT POTENTIAL RISK TO POLLUTANT-
INDUCED EFFECT 286
5.30 PRIMARY NATIONAL AMBIENT AIR QUALITY STANDARDS FOR
SULFUR OXIDE 301
5.31 INFORMATION BASE FOR ESTABLISHMENT OF NATIONAL
PRIMARY AMBIENT AIR QUALITY STANDARD FOR SULFUR
DIOXIDE 302
5.32 ADDITIONAL SCIENTIFIC EVIDENCE 304
5.33 BEST JUDGMENT THRESHOLD ESTIMATES FOR HEALTH EFFECTS
OF SULFUR OXIDES 309
5.34 SAFETY MARGIN INCLUDED IN SULFUR OXIDE STANDARD 310
5.35 SIZE OF POPULATION AT POTENTIAL RISK OF POLLUTANT-
INDUCED EFFECT 311
5.36 CORRELATION OF INORGANIC FIBERS WITH ASBESTOS BODIES
IN LUNGS OF 3,000 CONSECUTIVE AUTOPSIES IN NEW
YORK CITY, 1966 to 1968 333
5.37 CHRYSOTILE IN 28 CASES STUDIED BY ELECTRON MICROSCOPY . . 333
5.38 NUMBER OF ANIMALS WITH LUNG TUMOURS OR MESOTHELIOMATA . . 339-340
XI
-------
PREFACE
This report was prepared by Environmental Research Center,
Research Triangle Park (ERC/RTP) staff. The report's objective is
to review results of research by and under the sponsorship of ERC/
RTP, related to criteria and hazardous pollutants since publication
of the criteria and background documents. Recent reviews are also
included in a series of documents by the Conmittee on Challenges to
Modern Society, NATO; and by the National Academy of Science.
This report does not duplicate those other reviews. Primary
emphasis here is on results of the ERC/RTP research program. Sig-
nificant results of still other investigators known to the authors,
however, are included. But it is not a complete review and evalua-
tion of all current literature.
Contributing authors from the ERC/RTP include the following:
D. L. Coffin C. Kommineni
J. H. B. Garner J. Elder
T. Ripberger R. Smizlowicz
D. I. Hammer J. Graham
J. G. French A. H. Ellison
M. D. Waters P. Brubaker
K. Bridbord G. Ortman
J. R. Smith 0. Stopinski
J. B. Clements J. Nader
W. Galke W. Stein
F. Benson T. Hauser
Paul Humphrey J. Upham
D. Cahill G. Fairchild
XII
-------
1. SIM4ARY
1.1 GENERAL DISCUSSION
The best scientific data base was unavailable for six air pollu-
tants when National Ambient Air Quality Standards (NAAQS) were set.
This data base is still inadequate. However, evidence gathered since
the NAAQS were set support the standards.
A recent in-depth review by the National Academy of Sciences (MAS)*
indicated that there was no good reason relative to health and welfare
effects to change the standards. In some cases, for example, safety
factors were minimal compared to those factors used to control pollu-
tants such as radiation.
The National Academy of Sciences' report comprehensively reviews
the scientific data base for NAAQS. The details of material included
in the MAS reports will not be duplicated here. Summary material will
be included, however, when appropriate for completeness.
In this review, primary emphasis is on results achieved from the
ERC/RTP research and development program since publication of cri-
teria and background documents.
Air pollutants include complex combinations of reactive sulfur,
nitrogen, inorganic and organic compounds from man-made and natural
sources. Man and other species and materials are exposed to varying
*Air Quality and Automobile Emissions Control, A Report by the Coordi-
nating Committee on Air Quality Studies, National Academy of Sciences,
National Academy of Engineering (4 Volumes), dated September 1, 1974.
Prepared for the Committee on Public Works, U.S. Senate.
-------
amounts of this mixture of pollutants at different times. The biologi-
cal response to exposure is different for different species and pollu-
tion combinations. Physical properties of the environment such as
temperature and humidity may also influence the response.
Parts of this review consider things such as atmospheric transport,
certain analytical techniques and control technology common to each
classification of pollutants. Material on specific pollutants is also
presented with awareness that such pollutants contribute to the total
pollution burden in the atmosphere. A brief general discussion is
included in each section of the report.
1.2 INSTRUMENTATION
Since NAAQS were set, significant progress has been made in devel-
opment of sampling, analytical and instrumental methods to measure
ambient and source concentrations of air pollutants--particularly in
areas of chromatography, spectroscopy and chemiluminescence. Programs
in method calibration, interlaboratory comparison, quality assurance
and station siting have been developed and are being implemented.
Although the over-all monitoring program is still inadequate, the above
efforts have added much to the quality of aerometric data.
1.2.1 Nitrogen Oxides
An improved 24-hour wet chemical method of sampling and analysis
(TGS method) for ambient NO-, and an instrumental method (Chemilumi-
nescent) for NO, N09, NO , has been developed over the past three years.
If A-
The chemiluminescent method is now widely used for continuous monitoring
of ambient oxides of nitrogen along with the TGS and other methods and
is being considered as a replacement for the current Federal Reference
method.
-------
Guidelines have been proposed for selection and use of continuous
monitoring systems for NO emissions from chosen stationary sources.
J\,
The criteria and procedures for determining acceptability of a
candidate ambient air-monitoring method, equivalent to the prescribed
reference method, have been proposed.
Chemiluminescence is widely accepted as the most selective, accu-
rate and rapid means of NO analysis of mobile source emissions.
JC
1.2.2 Carbon Monoxide
Automated gas-chromatography instruments capable of semi-continu-
ously monitoring ambient air CO concentrations over a range of 0.1 to
1000 ppm are now commercially available. Improvements in methods for
generation of known concentrations of CO are under development.
The EPA reference method for measurement of CO from stationary
sources (continuous sampling coupled with a Non-Dispersive Infrared
analysis) is considered the most reliable of available methods. Pri-
mary problems involve particulates, water vapor and CO-. But recent
advances in NDIR technology have minimized these problems.
Gas correlation spectroscopy has been developed as an improved
method to measure CO emissions from mobile sources. The technique can
measure the expected low concentrations from light-duty vehicles of
FY 1976 and later.
1.2.3 Hydrocarbons
A commercial non-methane reactive hydrocarbon analyzer is now
available. The system converts all hydrocarbons except methane to FLO
and CO.-,--the methane is continuously monitored with a flame ionization
detector (FID) and total hydrocarbons are monitored with a second
-------
detector. A chemiluminescent method to monitor reactive hydrocarbons
and olefinic organics (such as ethylene and vinyl chloride) has been
partially developed. EPA has not yet required the states to monitor
hydrocarbons in support of the State Instrumentation Plans (SIP).
Emission standards for hydrocarbons from stationary sources have
not been set. No measurements are therefore required. Both total hydro-
carbons and non-methane hydrocarbon instrumentation developed for ambient
air measurements have been used to measure emissions from stationary
sources. These systems do not settle the question of their applicabil-
ity to emission standard settings. The present research and development
effort involves grab-sampling systems for class fractions that are then
analyzed in the laboratory using chromatographic methods. The analyti-
cal methods for polynuclear aromatics are not completely satisfactory.
The procedures are complicated and reproducibility and sensitivity are
problems.
Current Federal regulations of total hydrocarbon emissions from
mobile sources utilize flame ionization analysis. Techniques have been
developed to measure detailed composition, total non-reactive hydro-
carbons and methane. Work is in progress to develop methods to measure
other reactive components such as osygenates.
1.2.4 Photochemical Oxidants
The Federal Reference method for ozone measurements in ambient air
is based on the chemiluminescent reaction of 0_ with ethylene. A number
of chemiluminescent methods are now available commercially. These ozone
monitors are calibrated with ozone generators whose output has been
determined using the neutral buffered KI method. The calibration
-------
procedure is not completely satisfactory, but work is underway to improve
the procedure.
1.2.5 Particulate Matter
New procedures to collect aerosols in the fine and coarse particle
size-ranges are in final stages of development and will be tested in
CY 1975. With these collection procedures, sample deposit profiles are
well suited for rapid mass analysis by the beta gauge for mass and by
x-ray flourescence for elements with atomic numbers above 12.
Methods to determine the rate of particulate matter emissions, mass
concentrations, visible emission (opacity) selection of sampling ports,
stack gas velocity, stack gas molecular weight and stack gas moisture
from certain types of stationary sources have been published. To date,
most measurements methods have been concerned with mass emissions and
valid sampling techniques. Work is underway to develop satisfactory
methods to determine size distribution using inertial and optical tech-
niques .
Methods to determine the composition of particulates emitted from
stationary sources are being studied. The most attractive method cur-
rently for routine multielemental analysis of particulate matter from
stationary sources is x-ray fluorescence (XRF) spectroscopy. A wave-
length dispersion multispectrometer instrument, recently set up at
ERC/RTP, permits rapid analysis of all elements higher than oxygen in
atomic number with the high degree of spectral resolution needed for
airborne particulate samples.
Particulate measurement techniques have been developed for light-
duty mobile sources using a dilution technique. Measurements are
-------
currently made gravmetrically. Instruments to do that measuring are
being developed.
1.2.6 Sulfur Oxides
Several major improvements in methods to measure ambient air con-
centrations of sulfur dioxide and sulfate have been developed over the
past three years. Sensitivity, specificity and stability of field
monitoring systems for SCL offered commercially during the 1970's have
been greatly improved over those available in the 1960's. Improved
methods to measure sulfate and sulfuric acid in ambient air are being
developed.
Specifications for SO--continuous-monitoring systems for stationary
sources and ambient air have been published or proposed.
Sulfuric acid emissions from vehicles with exhaust oxidation cata-
lyst systems for CD and hydrocarbon control have stimulated work to
develop measurement techniques for sulfur in fuel and for sulfur oxides,
sulfuric acid and sulfates in exhaust emissions. Techniques have been
developed and are in use. Work is in progress to develop real-time
methods.
1.2.7 Asbestos
The present National Emissions Standards for Hazardous Pollutant
(NESHAP) regulation for asbestos emissions is based upon prescribed con-
trol practices. The most common measurement technique involves electron
microscopy to identify, count and size fibers. Development of a new and
more rapid measurement method is under way using the x-ray diffraction
technique on a. sample pretreated for fiber alignment.
-------
1.2.8 Beryl Hum
A measurement method for beryllium emissions involving isokinetically
sampling with a participate sampling train and analyzing the sample by
atomic absorption technique is described in the Federal Register. A
continuous monitoring technique is not available as a proven method
applicable to emission sources.
1.2.9 Mercury
The measurement method for mercury prescribed in the Federal Register
involves isokinetically sampling emissions in a solution of iodine mono-
chloride with analysis by spectrophotometry. A continuous monitoring
system can be developed through use of available analytical systems in
conjunction with an appropriate automated sampling interface system.
Development of such an interface system is under way.
1.2.10 Methods Standardization and Performance Evaluation
EPA is standardizing reference methods for individual air pollu-
tants in ambient air and stationary and mobile sources. The program
includes in-house methods evaluation and ruggedness testing, method
inter-relatability studies and collaborative testing. All criteria pol-
lutants are under study and some results have been published for all six
criteria pollutants.
1.3 ENVIRONMENTAL APPRAISAL
Essentially all man-made atmospheric matter classified as pollu-
tants are also produced by natural sources. Emissions produced by man-
made sources are superimposed on the "natural background." Trace gases
are naturally emitted from geophysical and biological sources. Natural
sources produce a major portion of particulate matter found in the
7
-------
atmosphere. Crude models for sulfur and nitrogen cycles are available,
but they are not sufficiently accurate to determine the contribution of
air pollutants from man-made activities to atmospheric loading. At
present, only limited attention is given to problems relating to natural
sources and removal mechanisms that impact on global or regional bal-
ances of air pollutants.
Limited data are available on emissions from selected classes of
stationary and mobile sources. They only permit crude estimates of the
contribution to atmospheric loading. An effort is under way to get
improved emission factors for both stationary and mobile sources.
1.3.1 Emissions
1.3.1.1 Oxides of Nitrogen Emissions--It is estimated that stationary
sources contribute at least half of the NO emissions of man-made
X.
sources in the U.S. The 1972 estimate was over eleven-million tons.
About 98% of total emissions were produced by combustion of fossil fuels.
The remainder was from industrial processes such as manufacture of nitric
acid, sulfuric acid and explosives.
Nationwide, mobile sources contribute about 50% of the total man-
made emissions of NO . In some urban areas such as Los Angeles, however,
x
the percentage is much higher. Oxides of nitrogen are formed in mobile
sources primarily by the fixation of atmospheric nitrogen in the combus-
tion process. Exhaust gas recirculation (EGR) and water injection
schemes reduce the peak temperature attained and this reduces the NO
J\.
emissions. Use of these methods on light-duty vehicles have reduced NO
A.
emissions significantly (50% between 1968 and 1973 models).
1.3.1.2 Carbon Monoxide Emissions--Preliminary results of tests on 1975
prototype light-duty vehicles equipped with oxidation catalyst-converters
8
-------
indicate that these vehicles will meet the 1975 standard of 15 gm/mi. In
some cases, the results were better than the 1976 standard of 3.4 gm/mi.
Reported values of CO emissions from the only commercially available car
equipped with the stratified charge engine (Honda CVCC) was 2 gm/mi.
1.3.1.3 Hydrocarbon Emissions--The Clean Air Act requires a 901 roll-
back from the 1970 levels of hydrocarbon emissions from light-duty
vehicles. The standard for 1976 models will be 0.41 gm/mi. Preliminary
tests indicate that this standard is being achieved in. prototype vehi-
cles. Tests indicate that the relative percentage of methane in emis-
sions from vehicles equipped with catalytic controls is increased--
approximately 801 of the hydrocarbon emissions with catalytic controls
may be methane. Ford Company has asked that emission standards be
changed to exempt methane. This proposal is currently under review.
1.3.1.4 Sulfur Oxides Emissions--Trace qualities of sulfur primarily
in the form of organic sulfur compounds are present in gasoline. Dur-
ing normal combustion processes, this sulfur is converted to SO- and
S03 (about 98% S02). The S03 reacts rapidly with water in the exhaust
and is emitted as sulfuric acid or sulfate particulate. The amount of
SO- and SO, emitted from mobile sources is negligible (less than II)
compared to that of stationary sources. Light-duty vehicles equipped
with catalytic converters emit higher concentrations of sulfuric acid
and sulfates than those vehicles without these devices. This may
result in significant concentrations of sulfur oxide emissions in high
traffic density areas.
1.3.2 Ambient Concentrations
Because of uncertainties in emission and air quality data, it is
difficult to establish a quantitative relationship between trends.
9
-------
However, some qualitative estimates can be made. Since the national
ambient air quality standards were set and control strategies imple-
mented, air quality data indicates a downward trend in CO and particu-
late matter pollutants. Oxides of nitrogen concentrations have increased,
but urban concentrations have decreased. No significant change in sulfate
concentrations is evident despite reduction in urban SO- levels. There is
some evidence that non-urban sulfate concentrations may have increased.
Ambient CO trends should directly reflect exhaust emissions from
motor vehicles. Since introduction of CO emissions controls, the average
CO emission rate has decreased. The annual average CO concentration for
selected urban areas appears to have decreased. Instrumentation problems
and operating procedures have biased the data in some areas.
The annual average concentration of nitric oxide (NO) has increased.
The pattern is less consistent for N02 and total oxides of nitrogen.
This increase is attributed to increase in fuel combustion and introduc-
tion of control systems for CO and hydrocarbons on mobile sources that
in some cases produced increased emissions of nitrogen oxides.
Trends for photochemical oxidants are not well established. The
data in general suggest a decrease in urban concentrations. However,
there is now much evidence that ozone concentration in certain non-urban
regions is increasing (Northeast U.S.).
There are few data available on ambient concentrations of mercury,
asbestos and beryllium.
1.4 TRANSFORMATION AND TRANSPORT
The relationship between emissions and atmospheric loading or air
quality is dependent on transformation, transport and removal processes
10
-------
in the ambient atmosphere. These processes are greatly affected by
meteorology and terrain. The extent that these processes contribute
to chemical characteristics and temporal and spatial distribution of
criteria and hazardous pollutants is often not well known.
Once pollutants are emitted into the atmosphere, they are subject
to various chemical reactions. A polluted atmosphere is a very
dynamic mixture of reactive gases and aerosols. It is this mixture
that man and materials are exposed to.
Although the photochemistry of nitrogen oxides has been widely
studied, only about 501 of the nitrogen emitted into the atmosphere has
been accounted for. Recent investigation suggests that nitric acid is
produced in the atmosphere by photolytic processes. Nitrous acid is
not formed in as great a quantity as nitric acid, but is important in
photochemical smog production because of the highly reactive radicals
produced. Many questions remain to be answered about quantitative
relationship between NO , oxidant and hydrocarbons in urban and non-
.X
urban atmosphere.
New findings on pollutant transport and conditions of hydrocarbon
oxidation have led to a new reactivity classification for organic com-
pounds, under the prolonged irradiation conditions of transported air
masses, organic compounds formerly classified as "less reactive" can
yield as much oxidant as some classified "more reactive." Recent smog
chamber experiments have produced quantitative oxidant-HC-NO relation-
J\.
ships. The applicability of these findings to the real atmosphere,
however, remains questionable. These results suggest coordinated con-
trol of HC and NO as the most effective means of oxidant reduction.
.A.
11
-------
Major advances have been made in understanding the sources, proper-
ties, chemical composition and dynamics of atmospheric aerosols.
Effective control techniques must consider particle size and particle
composition as well as precursors (S09, NO and certain organics).
Li Jv
Research work has identified a multimodal distribution of the atmo-
spheric aerosol.
Secondary sulfate, nitrate and organic aerosols comprise the major
part of fine particulate (one mode) matter. Road dust, tire debris and
other materials generated by vehicular traffic are a major source of
coarse particles in urban areas. Organic material is a major primary
and secondary component of fine particulate matter.
The transformation that converts S02 to sulfate is better under-
stood and measurements of transformation rates in power plant plumes and
urban atmosphere have been made. Progress has been made in chemical
characterization of atmospheric aerosols and further work is under way.
A solution to regional sulfate pollution problems will require an
understanding of how sulfates are formed in the atmosphere. Chemical
mechanisms for the transformation are not well known. Chemical oxida-
tion by photochemically generated intermediates and catalytic oxidation
are currently being emphasized as the most likely mechanisms. Both
laboratory and field studies are under way.
No significant oxidation reaction of CO has been found for urban
air.
Models have been developed to estimate pollutant concentrations
over an urban area. A User's Network for Applied Modeling of Air Pol-
lution (UNAMAP) was put into operation in May 1973. This program gives
12
-------
interested agencies current air-quality simulation models via a teletype-
processing network. A fluid modeling facility is being established to
study the flow and diffusing of pollutants around buildings, highway
configurations and complex terrain. A regional air pollution study
is now under way in St. Louis, Missouri that provides data for valida-
tion and improvement of pollution simulation models.
1.5 EFFECTS
National ambient air quality standards were established in refer-
ence to an incomplete data base. The standards were designed, based
upon available information, to include a margin of safety sufficient to
protect the health of the most sensitive segments of our society.
Since setting the standards, a continuing research effort has provided
additional effects information. Although the data base is still inade-
quate, recent reviews by the National Academy of Sciences (MAS) and the
National Institute for Environmental Health Sciences (NIEHS) have con-
cluded that the standards should not be changed at this time. However,
the NAS suggested that consideration be given to establishment of an
hourly standard for NCL.
Recent evidence suggests that secondary aerosols or combinations
of aerosols and gases are better correlated with adverse health effects
than are gaseous pollutants. The size distribution and chemical compo-
sition of the aerosols appear to significantly influence the biological
response.
1.5.1 Human and Animal
1.5.1.1 Nitrogen Dioxide and Suspended Nitrates--The limited number of
studies on human health effects of NCL reveal many gaps in the information
13
-------
base for the primary national standard. The few studies that have been
conducted are confounded by the presence of high levels of other pollu-
tants that have not been appropriately disentangled. Measurement data
used in these studies was also limited either by use of too few sampl-
ing stations or use of an unreliable measurement method for NCL or both.
A continuous monitoring instrument using the chemiluminescence
reaction between ozone and nitrogen oxide is considered more reliable
than either the Jacobs Hochheiser or the Saltzman method for measuring
nitrogen dioxide (that can be reproducibly converted to nitric oxide).
The chemiluminescent reaction is currently being used in new health
studies but information from such studies is not yet available.
In the Chattanooga study £>nwhi.chthe NCL standard was based), the
high NCL levels also were accompanied by relatively high levels of
suspended nitrates. The recent linking of asthma attacks to suspended
nitrates raises the question as to whether or not suspended nitrates may
have contributed to the observed adverse health effects in the Chatta-
nooga study. The observed human effects support reasonably well the
biological effects observed in experimental animal studies using NCL as
a single pollutant.
However, it has been reported that in some NCL rich atmospheres of
animal exposure chambers, aramonum nitrate may be generated in fairly
large quantities. This suggests that in some instances suspended
nitrates as well as NCL may have been contributing to atmospheric expo-
sure in animal studies. In any study involving exposure of animals to
NCL, the extent of reduction of NCL must be considered.
Five categories of human health effects have been identified in
association with NCL; acute toxic, often fatal reactions; diminished
14
-------
lung function; structural and biochemical changes in lungs that are most
likely to promote development of chronic respiratory disease; suscep-
tibility to infection; and systemic effects.
There is enough information from both human and animal studies to
determine the levels of N09 that may contribute to toxic reactions.
£
However, there are very few points on the lower part of the dose-response
curve to provide information on the levels of NCL that may contribute to
susceptibility to infection and development of chronic respiratory
disease. Further human studies are necessary that address the contribu-
tion of N02 and its degradation products to insidious chronic disease.
Until such information is available, it seems appropriate to maintain the
present standard. It provides a reasonable margin of safety based on
both animal and human studies to date.
In the New York-New Jersey metropolitan area, a study of 373
asthmatics followed daily for eight months showed that increased asth-
matic attacks were significantly associated with ambient suspended
nitrates. The observed increased risk of asthmatic attacks related to
increased concentrations of nitrates occurred only at a temperature of
10° centigrade or above. In another study in two southeast communi-
ties there was evidence of excess risk of asthmatic attacks with ele-
vated levels of suspended nitrates at higher temperatures. But the
findings were less consistent than those observed in the New York-New
Jersey area.
Preliminary findings of a study of 512 asthmatics in seven communi-
ties in the Los Angeles Basin also show significant association of
asthmatic attacks with suspended nitrates. But the full data analyses
15
-------
of this study has not yet been completed. Nitrogen dioxide was not
significantly associated with asthma attack rates in any of these
studies.
Very little is known about inhalation toxiciology of inorganic
nitrates. If the associations between airborne nitrates and asth-
matic attacks are causal, airborne nitrates may be an important
respiratory irritant.
Pathologic effects of NCL on the respiratory tract have been
induced in experimental animals at concentrations at and above 2 PPM.
Exposure to concentrations of 10-25 ppm on a continuous basis has pro-
duced pulmonary emphysema-like disease in rats, mice, dogs, and
rabbits. The biological response appears to be different when the
experimental animal is exposed to oxidant and NCL, although the results
are not yet conclusive.
The NCL effect was obscured when equal parts of NCL and 0, were
used. The physical appearance of emphysema varied with the species
studied and was possibly influenced by differences in the size and
frequency of inter-alveolar pores.
The biochemical mechanisms by which NCL reacts with tissue resem-
bles CL to some degree because of their common oxidizing properties.
A biochemical study is being done to examine the prospect that toxicity
of NCL and CL may be an "ultimate" toxicant produced by a reaction
between the oxidant and unsaturated fatty acid. Ozone is known to
react with unsaturated fatty acids to produce fatty acid ozonides. NO
t+
similarly can react with unsaturated fatty acids to produce nitroxide
fatty acids that are likely to be very reactive.
16
-------
A second biochemical study is being done to examine the mechanism
of action to nitrates and sulfates. The results of these studies should
provide additional insight into comparative toxicity of these pollutants.
Studies are being conducted that employ interaction of pathogenic
organisms and air pollutants to demonstrate increased susceptibility to
infectious pneumonia. The model has been used effectively in environ-
mental toxicological studies of ozone, NCL, radiated automobile smog and
exposure to some trace metals. The results have shown that the role of
infectivity can be enhanced in certain species of animals studied.
The model has also been used to study mechanisms of action of pol-
lutants. An important problem in NCL and CL toxicity involves relative
toxicity of mode of dose-concentration or a function of time over given
time periods. The results of these preliminary studies show greater
mortality in experimental animals from pulmonary infection produced by
the same concentration multiplied by time for different time periods
(for example: 14 ppm x 0.5 hr. versus 1 ppm x 7 hr.). The study indi-
cates that actual concentration has more influence than duration of
exposure. An inverse relationship was found between survival and length
of exposure at constant concentrations,
1.5.1.2 Carbon Monoxide--Both the present ambient air quality standard
for carbon monoxide of 10 mg/m (9 ppm) averaged over eight hours and of
the 40 mg/m (35 ppm) one-hour average coincide with carboxyhemoglobin
levels of approximately 1.51. Since the standard was established, new
evidence has appeared supporting the current standard in three areas.
These areas include:
(1) Effects on individuals with arteriosclerotic vascular disease
(2.5-3.01 range during exercise); (2) effects on exercise capacity in
17
-------
healthy individuals (diminished exercise performance at COHb levels of
approximately 5-6%); and (3) additional evidence on alertness effects
(impairments on vigilance tasks at COHb levels ranging from 3.0-6.5%).
1.5.1.3 Oxidants_--The oxidant thresholds for cough and chest discomfort
were found in one study to be 500-600 yg/m as a maximum one-hour
average. These levels are similar to those observed in studies of
occupational exposure of men. These thresholds should be generally
applicable to healthy adults and would probably represent an upper
bound for persons with pre-existing cardiovascular disease or asthma.
The synergistic effect of ozone and sulfur dioxide on human pul-
monary function has been reported in the past two years. There is also
concern that interaction of these two pollutants on lung function is
impaired in a non-linear fashion to any stimulus. The "extra" decre-
ment attributed to the combination of SCL and ozone could be an additive
effect in the non-linear part of lung-function curves.
Chromosome breaks in hamsters following ozone exposure have been
observed. But this may only be a peripheral and not a stem cell effect.
If so, it is not properly described as a mutagenic effect, but more data
are needed to clarify this point.
There is much evidence that exposure to ozone at concentrations
-------
number of studies have shown that ozone exposure induces changes in
pulmonary function.
Studies have consistently shown deleterious effects (decrease in
pulmonary function, chest discomfort, cough and eye discomfort) in
people exposed to less than 1 ppm (0.05 to 0.75 ppm). Healthy adult
males exposed to 0.50 ppm of ozone for two hours during intermitten
light exercise developed measurable physiological and biochemical
changes.
1.5.1.4 Particulates--The theoretical definition of suspended fine
particles as they relate to health are necessarily broad and somewhat
arbitrary. Deposition of particles anywhere in the respiratory tract,
from nose to alveoli, may be harmful to health. There appears, how-
ever, that non-viable particles deposited below the trachea may be
more harmful than those deposited in the nasal cavity. Whether parti-
cles must reach the alveoli to exert harmful effects is a more contro-
versial issue. It is likely that particles deposited in the bronchi
or bronchioles are more important in producing bronchitis than those
deposited in the alveoli. At present, there is little value in con-
fining the category of suspended fine particles only to those parti-
cles that reach the alveoli.
The Task Group on Lung .Dynamics has calculated that up to 10% of
inhaled particles of diameter 2-5 ym are deposited in the alveolar
compartment and up to 30% are deposited in the alveolar compartment.
Most of the particles of diameter above 5 urn are deposited in the
nasal cavity. So a size range of fine particulates, up to 2-5 ym
diameter, measured aerodynamically, is of principal interest from a
health standpoint.
19
-------
The toxicological properties of inhaled particles are related to a
complex of factors that include the specific physical-chemical proper-
ties of the particles themselves and host conditions that affect response
to exposure to these particles. As for toxicity of certain specific
particulates that may exist in the atmosphere, moderate amounts of infor-
mation are known. But such information is not usually available in a
form suitable for accurate assessment of minimum time-concentrates for
exposure that will lead to adverse health effects.
Even less information is known about how physical environmental
variables (weather conditions) may interact with suspended particulate
matter and alter its chemical composition and affect its toxicity.
Weather conditions per se may affect people in such a way as to alter
their exposure to inhaled particulates. The information that is avail-
able, however, provides a basis for estimating the relative toxicity of
certain substances and provide insight into problem areas where our
information is inadequate.
The main way air particulates affect health is presumably through
breathing and consequent effects on the respiratory system. This pre-
sumption is acceptable in determining short-term effects of irritant
aerosols, but may be less tenable in considering long-term exposure
effects where such responses are carcinogenesis, mutagenesis and subtle
metabolic effects result from whole body burdens. In these cases, routes
of entry besides breathing may be important in establishing dose-response
relationships, However, inhaled particles may become non-respiratory
toxicants where substances are translocated to the gastro-intestinal
system by muco-ciliary transport and swallowing. There they may exert
20
-------
a primary toxic effect or be absorbed through the lung into the blood
stream and translocated to other tissues, eliciting an adverse health
effect.
Deposition rates and retention time of inhaled particles are sepa-
rate phenomena with overlapping biologic effects. Particle retention
time is dependent upon: (1) the site of initial deposition in the
respiratory tract, and (2) the chemical composition and properties of
the particles.
If particles are deposited in the ciliated epithelium of airways,
particle clearance is reasonably rapid. Toxicity then depends on
solubility of particles in mucus while in transit up the muco-ciliary
escalator. If the particle is highly soluble, toxic inflammation in
the airway epithelium may occur.
Particles deposited in the alveoli of the lung may only slowly
clear from this location. The extent to which alveoli are cleared
reflects their solubility and subsequent translocation via lymphatic
drainage or macrophage phagocytosis and clearance via the muco-ciliary
transport mechanisms.
In some cases, especially with fine suspended particle material,
numbers of particles may be more relevant than particle mass. To
relate the number of particles to the mass size distribution requires
determination of frequency distribution. Particle number can be
toxicologically important since the large surface area may promote gas-
particle or volatile-substance particle interaction. Coupled with
small size, these interactions may result in deeper penetration of
gaseous or volatile substances in the respiratory tract and more rapid
21
-------
dissolution. Surface area is also important for highly insoluble parti-
cles that express their toxic potential only at the surface.
Another important factor in the regional deposition of inhaled
particles is hygroscopic!ty. Particles, as usually measured externally,
are not saturated with water. The growth of sodium chloride particles
is a function of relative humidity, with particle size increasing when
relative humidity exceeds 75%. The respiratory tract, with its satu-
rated atmosphere, may considerably increase effective particle size over
and above that measured in ambient air. However, hygroscopicity is of
less significance with particles of high density.
1.5.1.5 Sulfur Oxides, Sulf uric Acid and Sulfate--0ne of the primary
problems in determining health effects of SO- continues to be develop-
ment of an understanding of the way this gas interacts with other sub-
stances in the atmosphere. Laboratory studies have demonstrated that,
in an experimental situation, the levels of SO- found in ambient air
are innocuous until combined with other substances.
Similar studies with animals demonstrated increased toxicities of
sulfuric acid mists and various sulfate compounds that are products of
atmospheric transformation of S0?. The importance of high relative
humidity (580%) to enhancement of the effect of sulfur dioxide and
sodium chloride aerosol is demonstrated through sensitive respiratory
physiologic responses. Observed initial associations of adverse health
effects with ambient levels of suspended sulfates in epidemiologic
studies tend to support these laboratory studies.
Studies of three New York neighborhoods showed that cold tempera-
tures were directly related to increased symptom rates in people with
22
-------
combined heart and lung disease. Elevated suspended sulfates were the
only pollutant consistently associated with symptom aggravation. Daily
sulfur dioxide and total suspended particulate concentrations could not
be associated with symptom aggravation in the "heart and lung" panel,
the most sensitive to variations in daily pollutant concentrations.
Also, asthma attack rates were more consistently associated with
daily suspended sulfate levels than with either sulfur dioxide or
total suspended particulates.
Studies of daily asthma attack rates in the Salt Lake Basin showed
that attacks were more consistently correlated with colder outdoor
temperature than with any measured pollutant. So an analysis of asthma
attack rates measured against daily pollutant concentrations was car-
ried out within two ranges of minimum temperature, 30 to 50°F and
greater than 50°F. Virtually no relationship between sulfur dioxide
and attack rates appeared, but total suspended particulates and sus-
pended sulfates were positively correlated with daily asthma attack
rates. A higher frequency cf asthma attacks was observed at the same
daily total suspended particulate concentration when a high sulfate
fraction was present in the atmosphere so that it appeared sulfate
levels were a stronger determinant of asthma attack rates than total
suspended particulates.
The breathing ability of elementary school children in New York
City was affected in areas with high levels of SCL. In Cincinnati,
this same indicator was associated with sulfate exposure.
The pattern of chronic bronchitis rates is consistently higher
among residents of more polluted communities.
23
-------
1.5.1.6 Hydrocarbons - -The National Ambient Air Quality Standard for
hydrocarbons was established as a guide in regulating oxidant standards.
The reference method measures gas-phase hydrocarbons corrected for
methane. Particulate hydrocarbons, especially polynuclear hydrocarbons,
were not considered in establishment of the standard. Current ambient
air data are limited primarily to non-methane gaseous hydrocarbons and
benzo-a-pyrene.
When the criteria document of hydrocarbons was published (March
1970), there was insufficient evidence to demonstrate any bad health
effects from gaseous hydrocarbons at levels found in the ambient air.
No new evidence to the contrary is available. However, little research
has been done on the question.
Many gaseous hydrocarbons are directly involved in atmospheric for-
maldehyde, other aldehydes, ketones, peroxyacetyl nitrate (PAN) and
other oxidants. Some of these products produce adverse health effects.
A limited amount of study in cases of occupational exposure has been
devoted to some of these compounds. In reported studies Hodgkin's
disease has been associated with exposure to benzene at concentrations
of 150 to 210 ppm in the work place. Toxic effects attributed to for-
maldehyde concentrations of 2 to 10 ppm have been reported. A study of
the incidence of leukemia involving petroleum and chemical products in
the work population produced no conclusive increased risks. The limi-
tations inherent in each of these studies are such that results should
be viewed with caution--particularly for quantitative results.
Occupational evidence and laboratory experiments for other classes
of hydrocarbons such as PVC/VC, PCB's and polynuclear hydrocarbons show
24
-------
that some of these compounds are carcinogenic. The dose response data,
however, are incomplete.
1.5.1.7 Mercury--Several recent animal exposure studies using methyl
mercury given in repeated oral doses over relatively long time-periods
show that a "no-effect" level for neurological sign development and/or
histopathological evidence of mercury poisoning is from 0.03 mg/kg body
weight (monkeys, cats) to 0.1 mg/kg body weight (calves).
Daily oral doses of alkyl-mercury compounds as small as 0.4 mg/kg
body weight produce neurological disorders in rabbits, cats and dogs.
Daily oral doses of 0.25 mg/kg body weight of methyl mercuric chloride
given throughout most of the gestational period have caused increased
abortions and fetal abnormalities in cats. Levels above 0.3 mg/kg
given during gestation also increased fetal abnormalities in rat fetuses.
Most animal experiments were carried out with daily dose levels
much higher than the foregoing experiments. The most chronic exposure
found was that of monkeys exposed for two years. Low level, long-term
exposures were notably absent.
Behavioral changes have been produced in rodents where no histo-
logical lesions were found from single oral doses of methylmercuric
chloride. A relatively high 25 mg/kg dose level was required since
doses at 20 mg/kg level did not produce significant changes.
1.5.1.8 EeryIlium--A long latent period of 15 or more years has been
observed between the onset of disease and the last exposure to beryl-
lium, with evidence indicating that beryllium may accumulate in a
latent form in lyosome. A latent period could persist for years until
declining adrenal stress, reserve and output with age caused lyosomal
25
-------
instability and the onset of beryllium disease. Under these conditions,
especially if the adrenal reserve was also low, the adrenal output may
not be adequate to meet the needs.
In recent experiments, the biological half-life of Be in mice was
found to be about twice the value cited by the ICRP (540 days). There
is now some evidence of auto-immune response subsequent to beryllium
exposure.
1.5.1.9 Asbestos--The association of impaired health with industrial
exposure to asbestos, principally asbestosis and pulmonary cancer is
well known. There is little evidence that exposures to low levels of
asbestos and asbestos-like fibers that may be present in ambient air
have resulted in either asbestosis or pulmonary cancer in people. How-
ever, the discovery of mesiothelioma of the pleura associated with low
exposure or casual contact with chrysotile asbestos has raised suspi-
cion about possible health hazards of air pollution.
Evidence indicates that all major types of asbestos may cause lung
cancer. There are clear differences in risk with type of filter and
the nature of exposure, however. All commercial types of asbestos,
except anthophylite, may cause induction of mesathelioma. However, the
risk is greatest with crocidalite, less with amosite and apparently
less with chrysotile.
Some cases of mesothelioma have no known association with exposure
to asbestos. There is some evidence of association of mesothelioma
development with air pollution near croidalite mines and factories work-
ing with mixed fibers. There appears to be no extra risk from air
pollution near chrysotile and amosite mines.
26
-------
At present, there is no evidence of extra cancer risk to the general
public from asbestos in air, water or food. Cigarette smoking enhances
the risk of developing bronchogenic cancer in workers exposed to asbes-
tos. There is evidence that health effects associated with exposure to
asbestos may not be unique to commercial forms, but may also be associ-
with asbestos-like fibers found in other rocks and possibly man-made
fibers.
1.5.2 Materials
The soiling effects of particulate matter and most cause-effect
relationships for common gaseous pollutants have long been known. What
has been lacking is information on dose-response relationships to use in
cost-benefit studies, support for primary air quality standards and set-
ting of secondary standards.
In past years, EPA scientists have conducted both laboratory and
field studies to better define cause-and-effect. Recently they used
specially-designed controlled environment chambers to generate needed
dose-response data for economically vulnerable materials. Although some
preliminary information was produced, all (materials) research for FY-75
and after, including dose-response studies, was terminated because of
budget limitations.
1.5.2.1 Nitrogen Oxides--The damaging effects of atmospheric nitrogen
oxides mainly concern the textile industry. Nitrogen dioxide (N07) is
LJ
the troublesome pollutant; nitric oxide (NO) appears harmless. Atmo-
spheric N0~ can react with a number of textile dyes and produce
noticeable color change or faiding, causing a problem to the textile
industry for decades. Concerning the yellow discoloration of undyed-
27
-------
white and pastel-colored fabrics, chemists have traced this problem to
the action of NCL on additives used to enhance fabric marketing proper-
ties. In the United States, annual gross damage to materials by N0~ is
estimated at three-quarters of a billion dollars.
1.5.2.2 Sulfur Oxide--The fact that SO- speeds deterioration of many
materials has long been recognized. Both field and laboratory studies
have shown that SCL-polluted environments (1) increase the corrosion
rates of most metals, especially iron, steel and zinc; (2) damage build-
ing materials such as limestone, marble, roofing slate and mortar; (3)
discolor and cause deterioration of statuary and other works of art;
(4) increase erosion rates of oil base paints; (5) degrade textile
fibers, especially cotton and nylon and fade-dyed fabrics; (6) embrittle
and discolor paper and (7) weaken leather.
1.5.2.3 Ozone--Damage by atmospheric ozone occurs mainly to elastomers
and textile dyes. Not all elastomers, however, are subject to ozone
damage; synthetic elastomers with saturated chemical structures have
inherent resistance, although they are relatively expensive and account
for only a fraction of the market. The great bulk of elastomers, includ-
ing natural rubber and synthetic polymers of styrerebutadiene, poly-
butadiene and polyisoprene, contain double bonds prone to oxidation and
thus sensitive to ozone. Researchers have developed antiozone additives
capable of protecting elastomers, but such protection is not permanent.
For certain items, such as tires, it usually lasts for the life of the
item.
Some important textile dyes are susceptible to fading from ozone
exposure. Blue dyes used on acetate and polyester/cotton fabrics and
28
-------
nylon carpets have been particularly troublesome. To prevent or mitigate
ozone fading, the textile industry must use more resistant dyes and/or
inhibitors, increasing the cost of finished goods.
1.5.3 Ecology
1.5.3.1 Nitrogen Compound:;--The effect of nitrogen oxides on vegetation
has been noticed only near industrial sources, as in the effect of nitric
acid on plants where ambient levels of nitrogen dioxide have been high.
Effects have been identified chiefly through laboratory experiments.
Indirect damage to vegetation has occurred from formation of acid rain
and photochemical reactions.
Photochemical reactions result in the formation of peroxyacyl
nitrate, the most abundant of the compounds formed, and responsible for
serious plant damage.
1.6 CONTROL TECHNOLOGY
1.6.1 Stationary Sources
Stationary combustion sources contribute more to the ambient atmo-
spheric loading of SO , NO and total suspended particulates (TSP) than
,X j\.
any other anthropogenic category, causing approximately 80% of SO
.A.
emissions, 501 of NO emissions, and 50% of ambient fine particulates.
.X
Advances in the development of technology to control SO, NO and
.A. .A.
primary particulates have been, achieved, and additional improvements are
expected in coining years.
Industrial processes and open sources (e.g., open burning, mining,
incineration, fugitive emissions) account for approximately 16% of SO
J\.
emissions, 44% of hydrocarbon emissions, 22% of CO emissions and 25% of
ambient fine particulates. They are also responsible for most hazardous
29
-------
air pollutant emissions (i.e., asbestos, beryllium,and mercury) as well
as vinyl chloride and other potentially hazardous carcinogenic pollu-
tants.
1.6.1.1 Particulate Control--Most pressing is the development of commer-
cially available and widely applicable fine particle (less than 3 microns)
control technology. It is currently in very early development under EPA
sponsorship. This technology is also needed to control detrimental trace
metals.
Scrubbers, electrostatic precipitators and fabric filter collectors
are the principal particulate collection devices. They are being
improved by attention to lowering costs, increasing their applicability
and extending the range of collection to smaller particle sizes. In
addition, entirely new concepts of particle collection are being
investigated.
System studies have been completed for each conventional device to
define areas for improvement and show new applications. Long-range
projects include: development of models for existing collectors; iden-
tification and laboratory testing of new concepts in particle technology
and new work in small particle detection and generation equipment.
1.6.1.2 Sulfur Oxides Control--In the area of sulfur oxides control,
achievement of Ambient Air Quality Standards demands control of indus-
trial combustion, industrial processes and area sources as well as
utility combustion. Natural low-sulfur fuels are only sufficient for
about 40% of the demand of State Implementation Plans. The remainder
must be supplied by technology in the form of effluent desulfurization,
fuel cleaning or process modification. Although the currently defined
30
-------
sulfur oxides control technology program contains projects diverse enough
to ensure achievement of Ambient Air Quality Standards, substantial EPA
research and development investment will be required to bring this
technology to commercial fruition. Only effluent desulfurization tech-
nology for electric utilitlies should achieve widespread commercializa-
tion within the next few years.
Although the program has emphasized stack gas cleaning systems, with
five or six such processes to be demonstrated before 1976, it has also
included work in areas of fuel modification and desulfurization, indus-
trial process modification and advanced combustion processes.
Most of EPA's early coal desulfurization work was to evaluate and
improve physical cleaning methods for coal. These methods should now
have moderate applicability in reducing sulfur by 30 to 701 in certain
coals. Ongoing work here includes washability testing, coal desulfuri-
zation and cleaning and pollutant control evaluation and optimization.
Most recent work is aimed at chemical desulfurization and the problems
of pollutants in fossil fue].
Laboratory-scale work is under way on processes to remove sulfur
totally from coal. The two processes under development are unique
since one is completely selective to inorganic sulfur in coal and the
other is preferential organic sulfur removal, but both produce a solid
fuel retaining the basic physical characteristics of coal. The tech-
nical feasibility of the inorganic sulfur removal system has been
proven on the laboratory and bench scale and is undergoing design to a
pilot-plant-size operation, with construction of this plant planned for
early 1974. The degree of sulfur removal from the organic process is
31
-------
currently being studied in a laboratory apparatus and should be ade-
quately developed by mid-1974 to start scale-up design to continuous
pilot operation by September 1974. The two systems will be examined
in detail as to possible integration into one scheme for complete
removal of sulfur from coal. After successful pilot plant operation,
plans call for scale-up to a demonstration-size plant.
Two factors—the major contribution of small coal-and-oil-burning
sources (area sources) to the pollution of ambient air and severely
limited supplies of low-sulfur fuels--indicate a need for clean fuel
or coal conversion processes. EPA has provided funds to both the
bureau of Mines and the Office of Coal Research to help develop conver-
sion and gasification processes for providing clean fuels and help
determine the environmental controls these techniques will require.
Work on advanced processes involves primarily the development of
fluidized-bed combustion. Fluidized-bed combustion options include:
pressurized coal combustion, coal gasification for the production of
low-sulfur, low-Btu gas and oil gasification. EPA also has conducted
systems studies for the development of advanced power cycles.
Work is also under way to control emissions from smelters, iron
and steel operations, Kraft wood pulping, iron foundries, secondary
aluminum furnaces and other industrial sources.
1.6.1.3 Nitrogen Oxides Control--On a national basis, 98% of the nitro-
gen oxides (NO ) from stationary sources are from combustion processes.
A.
Combustion control technologies under EPA investigation include: com-
bustion modifications such as flue gas recirculation, off-stoichiometric
combustion and low excess air operation; identification of alternate or
32
-------
novel combustion systems with inherently low NO emission levels; changes
-X
in burner design and use of alternate fuels. Although flue gas cleaning
processes and advanced power cycles are also being investigated, they
may offer only longer-term solutions to the problem. Major emphasis has
been on coal combustion in utility and industrial boilers, since this
poses the most difficult problem.
The main approach has been modification of the combustion process
to reduce NO formation, with bench-scale and field work indicating
A.
that control for the immediate future can be achieved. In addition,
promising stack gas cleaning processes are being investigated on a
small scale. Results in this area are not too encouraging to date,
however.
A complicating factor is that NO are formed during combustion
jC
through two separate mechanisms--the fixation of nitrogen in combus-
tion air, and the conversion of fuel-bound nitrogren contained in the
fuels. Although some control of both mechanisms has been achieved by
combustion modifications, it remains to be determined whether or not
combustion modification alone will adequately control the NO from
.A
fuel nitrogen.
1.6.1.4 Hazardous and Other Control--The number of pollutants requiring
control has expanded to include more than 30 potentially harmful and
annoying air pollutants. Since most of them are likely to be emitted as
particulate matter, there is strong interaction between these two pro-
grams. Hazardous and other pollutants under consideration include:
asbestos, beryllium, cadmium,, carbon monoxide, fluorides, hydrocarbons,
lead, mercury, and odorous materials from rendering plants, pulp mills,
33
-------
sewage treatment plants and chemical process industries. Because of the
paucity of data now available, many problem definition and source
characterization studies are required.
34
-------
2. INTRODUCTION
National Ambient Air Quality Standards established under the pro-
visions of the Clean Air Act amendments of 1970 were designed on the
basis of the best available information. Since this information was
incomplete, the Environmental Protection Agency began a continuing
research program to improve the data base and assess its adequacy. This
document is the first in a series of annual reviews by the Environmental
Research Center/RTF, as to results of its research efforts. It is not
intended to be a update of the criteria documents, which will be accom-
plished under the Pollutant Characterization Program.
Since this is the first of a series of reviews, it must cover new
information attained by ERC/RTP since promulgation of the Standards. A
review of the basis for the National Ambient Air Quality Standards and
automobile emission controls has recently been completed by the National
Academy of Sciences (MAS). For convenience, some material presented
here duplicates that presented by NAS.
Maintaining air quality and the consequences for human health and
welfare involves understanding many scientific, economic and social
variables. Since our present understanding is extremely limited, stan-
dard-setting and environmental control must be based in large measure
on judgment. Man-made environmental pollution is a consequence of our
unwillingness or inability to control human population and the products
of our natural resources, so that decisions on control are compromises
with impact on all of society's values. There is an important element
of risk resulting from incomplete information and understanding.
35
-------
Material presented here represents only a small part of the problem.
It is limited by our ability to: (1) measure gaseous and aerosol com-
pounds which have been classified as deleterious or hazardous; (2) quan-
tify and characterize emissions of pollutants from stationary and mobile
sources; (3) monitor and characterize atmospheric loading; (4) determine
the effects of a polluted atmosphere on human health and welfare and (5)
control man-made emissions necessary to maintain a given air quality.
36
-------
3. MEASUREMENT
3.1 INTRODUCTION
Measurement of an air pollutant begins with isolation and identifi-
cation. This is the qualitative step. Technique is developed to
generate the known substance in a controlled atmosphere free of unknown
substances. The quality of a sensor is then judged based on freedom
from response to other substances, sensitivity to the substance to be
i
measured and other characteristics.
Equally important is how well the measuring method reflects con-
centrations of the substance sensed. This is the quantitative step.
Before assessing the quantitative capability of a method, it is essen-
tial to generate a substance in known concentrations. Determining the
response of a sensor to a series of known concentrations constitutes
the calibration of the sensor. The degree to which the sensor repro-
duces the calibration reflects the stability of the methods.
The scope of the measurement problem is shown in discussion
following.
The composition of unpolluted air is not known precisely, but
Table 3.1 represents an approximation.
37
-------
Table 3.1. APPROXIMATION OF THE GASEOUS COMPOSITION OF
DRY UNPOLLUTED AIR
Constituent
Nitrogen
Oxygen
Water
Argon
Carbon Dioxide
Neon
Helira
Methane
Krypton
Nitrous Oxide
Hydrogen
Carbon Monoxide
Xenon
Ozone
Organic Vapors
Sulfur Dioxide
Nitrogen Dioxide
Ammonia
Molecular
Formula
N2
°2
H2°
Ar
co2
Ne
He
CH4
Kr
N20
H2
CO
Xe
°3
--
S02
N02
NH,
Volume
Fraction
78.09%
20.94%
—
0.93%
0.0318%
18 ppm
5 . 2 ppm
1.0-1.2 ppm
1.1 ppm
0.5 ppm
0.5 ppm
0.1 ppm
0.09 ppm
0.02 ppm
-0.02 ppm
-0.001 ppm
-0.001 ppm
-0.001 ppm
Major gases and vapors produced by man are shown in Table 3.2.
38
-------
Table 3.2. PRIMARY GASES AND VAPORS PRODUCED BY
HUMAN ACTIVITIES
Compound Formula
Carbon dioxide CO-
Carbon monoxide CO
Sulfur dioxide S09
Li
Nitric oxide NO
Hydrogen sulfide H2S
Ammonia NH.,
Hydrogen chloride HC1
Hydrogen fluoride HF
Paraffins
Olefins
Aromatics
Principal secondary pollutants produced in the atmosphere are shown
in Table 3.3.
Table 3.3. SECONDARY ATM3SPHERIC POLLUTANTS
Compound Formula
Nitrogen dioxide NO-
Ozone 0,
Peroxyacetyl nitrate H3C-C-0-0-N02
Aldehydes RCHO
Sulfuric acid H2S04
39
-------
Organic or inorganic particles found in the atmosphere frequently
are participants in the formation of gaseous or particulate secondary
pollutants shown in Table 3.4.
Table 3.4. ATMOSPHERIC PARTICULARS AND MAJOR PRIMARY AND
SECONDARY AIR POLLUTANTS FORMED BY PARTICLE
PARTICIPATION
Suspended Particulate Matter
Inorganics
Metals, (Pb, Cd, Be, Hg, etc)
Fluorides
Nitrates, Sulfates, Phosphates
Asbestos
Mineral Dusts (Silicates, Silica, etc)
Organics
Polycyclic Organics (Benzo [a] pyrene, etc)
Oxygenated Compounds
Pesticides
Aeroallergens
Gases
Inorganics
Sulfur Oxides (S02, S03)
Nitrogen Oxides (NO, N02)
°3
CO
H2S
HC1, C12, Fluorides, etc
40
-------
Table 3.4 (continued)
Ammonium Compounds (NH,, etc)
W2
Organics
Hydrocarbons (Paraffins, Olefins, Aromatics)
Oxygenated Compounds (Aldehydes, Ketones)
Sulfur Containing Compounds (Mercaptans, etc)
Nitrogen Containing Compounds (Peroxyacetyl Nitrate
(PAN), etc)
Halogenated Compounds
Table 3.5 shows pollutants for which the federal government has
established National Ambient Air Quality Standards
41
-------
Table 3.5. NATIONAL AIR QUALITY STANDARDS
Averaging Primary Secondary Reference0
Pollutant Time Standards Standards Method
Sulfur Dioxide
Particulate
Matter
Carbon
Monoxide
Annual
Arithmetic
Mean
24 hours
3 hours
Annual
Geometric
Mean
24 hours
8 hours
1 hour
80 iio/m
(0.03 ppm)
365 yg/m
(0.14 ppm)
--
75 yg/m3
260 yg/m3
10 mg/m3
(9 ppm)
40 mg/m
(35 ppm)
--
1300 yg/m3
(0.5 ppm)
60 yg/m
150 yg/m3
Same as
Primary
Standards
Pararos aniline
Method
High Volume
Sampling Method
Non- Dispersive
Infrared
Spectroscopy
Photochemical
Qxidants
(corrected for
N0~, and S0~)
1 hour
160 ^/m PrUS
(0.08 ppm) J™2*
Standard
Gas Phase
Chemi luminescent
Method
Hydrocarbons ,6Q / Same as Flame lonization
(corrected for 3 hours ff] -$' -> Primary Detection Using
methane) ^ pp J Standard Gas Chromatography
Nitrogen
Dioxide
Annual ,nn / 3 Same as
Arithmetic ^°nf/l. Primary
Mean (0-05 ppm) Standa;d
New Reference
Method Under Study
42
-------
Table 3.5 (continued)
a. National standards other than those based on annual arithmetic means
or annual geometric means are not to be exceeded more than once per
year.
b. National Primary Standards: The levels of air quality necessary,
with an adequate margin of safety, to protect the public health.
c. National Secondary Standards: The levels of air quality necessary
to protect the public welfare from any known or anticipated adverse
effects of a pollutant.
d. Reference method as described by the EPA. An "equivalent method"
means any method of sampling and analysis which can be demonstrated
to the EPA to have a "consistent relationship to the reference
method."
e. For use as a guideline in assessing implementation plans.
EPA plans to examine emissions from industrial categories that may
lead to additional standards. Stationary sources under consideration
are:
*Alurainum Reduction Plants
Animal Feed Defluorination Plants
Automobile Assembly Plants
Chlorine and Caustic Plants
*Coal Cleaning Plants
Coke Plants
*Copper Smelters
Cotton Gins
*Ferro-Alloy Plants
Gas Turbines
Grain Millins and Handling
Graphic Arts Plants
Grey Iron Foundries
Hydrochloric Acid Plants
Lime Plants
Paint and Varnish Plants
Petrochemical Plants
*Phosphate Fertilizer Plants
*Phosphoric Acid Plants
*Phosphorous Reduction Plants
Pulp and Paper Mills
Rendering Plants
*Sewage Sludge Incinerators
Soap and Detergent Manufacturing
Facilities
Sulfur Recovery Units
Vinyl Chloride Plants
*Zinc Smelters
*Promulgated or proposed.
43
-------
Emission standards for asbestos, beryllium and mercury are contained
in Federal Register ^>8_ (6) 8820, April 6, 1973. These substances have
been identified as hazardous air pollutants under the Clean Air Act as
amended.
In 1970 Congress passed legislation requiring 1975 automobiles to
show a 90% reduction in hydrocarbon and carbon monoxide emissions over
1970 models and a reduction in nitrogen oxide emissions by 1976 of 90%
from that in 1971 models. Regulations covering 1972-1975 light duty
vehicles and heavy duty engines were published in Federal Register 37
(22) 24250, November 15, 1972, and amended in Federal Register 3£ (124),
June 28, 1973. The Administrator granted a one-year extension for the
1975 standard in Federal Register 38_ (126) 17441, July 2, 1973, and a
one-year extension for the 1976 NO standard in Federal Register 38 (161)
,x
22474, August 21, 1973.
Additional regulations for special engine types beginning with 1973
model year are contained in Federal Register 37_ (221) 24250, November 15,
1972, and interim standards for specific truck types in Federal Register
38^ (151) 21398, August 7, 1973.
Standards for control of air pollution from aircraft and aircraft
engines were promulgated by EPA in Federal Register_38 (136) 19088,
July 17, 1973.
Thus far EPA has promulgated methods for measuring six ambient air
pollutants, one of which (NO) has been rescinded for further study; five
pollutants for five sources under standards for new stationary sources,
two hazardous pollutants and four pollutants from mobile sources.
44
-------
The objective of the Clean Air Act is ensuring air quality. Since
knowledge of atmospheric loading is important to air quality, it is to
this subject that the rest of this section will be largely devoted.
Generation of a controlled atmosphere for purposes of studying perfor-
mance characteristics and establishing calibration of a manual or
instrumental method will be covered. Measurement methods in common use
as well as those having promise will be discussed as well as EPA efforts
to upgrade measurement methods by promulgating performance criteria for
reference and equivalent methods. Validation techniques, auditing and
interlaboratory and intralaboratory evaluation methods will be discussed.
3.2 PRODUCTION OF TEST ATMOSPHERES
The adequacy of a manual or instrumental method designed to measure
a gaseous pollutant is proven by the manner in which it respons to a
test atmosphere artifically produced by introducing into clean air a
known amount of the pollutant. Knowing the volume of clean air (diluent
gas) and the volume of the pollutant, the volume-to-volume relationship
can be calculated. In ambient air measurements this is commonly refer-
red to in terms of parts per million (ppm), parts per hundred million
(pphm) or parts per billion (ppb) by volume. Air pollution technolo-
gists assume the "by volume."
The reading obtained as a result of the signal (the pollutant) is
then a measure of the sensitivity of the method.
The same procedure is used for generating a test atmosphere that
contains an interferent sepai-ately or in addition to the pollutant to
be measured.
45
-------
There is no technique for providing an optimal test atmosphere for
all measurement methods used in ambient monitoring.
The permeation technique as originally reported described how a
condensible gas (in its liquid state) would be sealed in a section of
inert plastic tubing. Following an initiation period of a few hours to
several weeks, permeation of the substance through the tubing proceeds
at a constant rate until the enclosed gas is nearly exhausted. The rate
of permeation is calculated by gravimetric determinations of weight loss
in a unit of time. Using apparatus to control the temperature of the
permeation tube and regulate the flow of the diluent gas stream over the
tube, a wide range of known concentrations are obtainable from parts per
billion to parts per million with achievable accuracies of within plus
or minus 2%. The concept has been extended to include the controlled
dispensing of non-liquifiable gases, and some instrument suppliers are
including permeation dispensing devices as a self-contained calibration
source or an optional ancillary module.
There are installations where gas mixtures pressurized in metal
cylinders have advantages. Methane in air is an example of convenience
dictating the use of a certified concentration of the gas for calibra-
tion of hydrocarbon analyzers. While not long ago the reliability of
such specialized gas mixtures left much to be desired, recently spe-
cialty gases from industrial suppliers have improved greatly. Nonethe-
less EPA plans a quality control program for these mixtures.
The National Bureau of Standards (NBS) is more and more supplying
standard gases for use in air pollution laboratories. Currently it sup-
plies nitric oxide, carbon monoxide and methane in nitrogen at one or
46
-------
more concentrations. These concentrations are considerably higher than
those found in the average atmosphere. The use of nitrogen rather than
air as the diluent gas renders the mixture unusable for certain applica-
tions. NBS is cooperating with EPA in exploring the feasibility of sup-
plying a greater variety of gases and concentrations and the use of air
as a diluent.
Very promising is the use of aluminum cylinders for pressurized
gases. Preliminary evidence indicates that changes to a contained gas
due to wall effects occur at a much slower rate (if at all) in aluminum
rather than steel cylinders. It has been established that aluminum
cylinders are superior for storage of ultrapure air, a supply of which
becomes increasingly important as measurement technology progresses.
Having a pollutant-free gas for use as a diluent to provide step con-
centrations of a pollutant is useful in producing a calibration curve
characterizing the response of any manual or automated method.
3.3 OXIDES OF NITROGEN MEASUREMENTS
3.3.1 Ambient Air
This section discusses the measurement of nitric oxide (NO),
nitrogen dioxide (N09) and total oxides of nitrogen (NO ). Most man-
^ X
made nitrogen oxides are in the form of NO. In the atmosphere, NO
reacts to form the more irritating and toxic N0~. Other oxides of
nitrogen, nitrogen pentoxide (N^Or), nitrogen tetroxide (N^O-^) and
nitrogen trioxide (N-O?) are present in the atmosphere in extremely
low concentrations if at all. Their involvement in photochemical
reaction is possible. Laughing gas, nitrous oxide (N~0), if present
in the atmosphere, is harmless. NO and N09, referred together as NO ,
L* -X
are the compounds of primary interest.
47
-------
3.3.1.1 Calibration--Permeation tubes are the preferred source for
calibration of NCL analyzers. NO is supplied in nitrogen, typically
at concentrations of 50 or 100 ppm, in pressurized cylinders. The con-
centration in the cylinders remains stable for months.
The calibration accuracy of NO analyzers is dependent upon a true
,A-
zero. Residual NO and N02 are removed by converting the NO to N02 using
ozone to oxidize the NO and then removing the N02 by an N02 absorbent
like activated charcoal.
3.3.1.2 Manual Measurement--Manual N02 analytical techniques usually
are performed in a laboratory remote from the sampling site; hence the
requirement for a method to efficiently collect and retain the N02 sam-
ples. The sampling time is generally 24 hours. Variables are sampling
rates, duration of sampling period, collection efficiency and analytical
method. Absorption in liquids is the most common method of sampling.
3.3.1.2.1 Jacobs Hochheiser Method--This method, contained in Federal
Register 36_ (84) 8186, April 30, 1971, is a procedure where an ambient
sample containing NO- is drawn through an aqueous sodium hydroxide solu-
LJ
tion producing a nitrite ion from N02- The resulting alkaline solution
is treated in the laboratory to produce a color proportional to the N02
sampled so that optical absorbance is linearly proportional to the con-
centration of the colored species and is measured spectrophotometrically.
This method, declared the reference method for measurement of N02,
was used in the National Air Sampling Network (NASN). However, in 1972,
EPA studies of the procedure showed nonlinearity of collection efficiency
varying from approximately 15 to 70%, depending on the concentration of
N02 sampled. It was also found that NO in the presence of N02 caused a
positive interference in the measurement.
48
-------
EPA published the following remarks on June 8, 1973:
" EPA's analysis indicates that the reference
method is deficient in two respects. First, the
method overestimates nitrogen dioxide concentra-
tions at low levels and underestimates them at
high levels because the collection efficiency of
the abosrbing reagent is dependent upon nitrogen
dioxide concentration being measured. Second,
the method is subject to positive interference by
nitric oxide. Since the variable collection
efficiency problem cannot be resolved, this
method can no longer serve as the reference
method."
3.3.1.2.2 Saltzman Method—One of the most widely used chemical methods
for determining NCL in ambient air is also known as the Greiss-Saltzman
method. This coloriraetric method is based on the reaction of NO- with
sulfanilic acid to form a diazonium salt which couples with N-(l-naph-
thyl)-ethylene-diamine dihydrochloride to give a red dye. The color is
measured spectrophotometrically at 550 nm and the value obtained is
proportional to the NO- concentration.
Li
This method is sensitive from 0.005 ppm to about 5 ppm when sam-
pling is conducted with bubblers. Sampling time is usually 10 to 30
minutes.
The Intersociety Committee (ISC) has adopted the Saltzman method
as the tentative manual method of analysis for NO,,. The American
Society for Testing and Materials (ASTM) has also selected the Saltzman
method as the standard method of test for nitrogen dioxide content of
the atmosphere.
Ordinarily interferences are not a problem. A tenfold ratio of
S02 to NO- produces no effect. A 30-fold ratio bleaches the color to
a slight extent. Interferences from other oxides of nitrogen and ozone
at typical ambient levels are negligible. Peroxyacetylnitrate (PAN)
49
-------
can give a response of up to 35% of an equivalent molar concentration of
NCL, but in ordinary ambient air, PAN concentrations are too low to
cause any significant error.
The Saltzman method for NCL measurement cannot be used successfully
when the delay between sample collection and color measurement is more
than four to six hours or when sampling periods of longer than one hour
are required because of color fading.
3.3.1.2.3 Tentative Method for the Determination of Nitrogen--Dioxide
TGS ANSA--This method has its basis in the Levaggi (TEA) Method. It is
currently one of several prime candidates for a reference method.
NCL is drawn through a solution of triethanolamine, o-methoxyphenol
and sodium metabisulfite. The nitrite ion produced during sampling is
determined coloriraetrically by reacting the exposed absorbing reagent
with sulfanilomide and 8-anilino-l-naphthralenesulfonic acid. The method
is applicable to collections of 24-hour samples in the field followed by
analysis within three weeks in the laboratory. The range of analysis is
about 0.01 to 2 ppm. Beer'slaw is obeyed throughout this range. The
method has a predictable collection efficiency of between 92-94% from
20-740 yg/m5 NCL and virtually no interference from NO, SC>2, 03 and C02.
3.3.1.2.4 Tentative method for the determination of nitrogen dioxide
in the atmosphere (Sodium arsenite procedure) - -This method has its basis
in the Christie (arsenite) Method. It is also under investigation and
is a prime candidate for the reference method.
NCL is drawn through a solution of sodium hydoxide sodium arsenite
to form a stable solution of sodium nitrite. The nitrite ion produced
during sampling is reacted with phosphoric acid, sulfanilamide and
50
-------
N-l-(naphthyl) ethylenediamine dihydrochloride to form an azo dye. The
method is applicable to collection of 24-hour samples in the field fol-
lowed by analysis in the laboratory within six weeks. The range of
analysis is about 0.015 to 1.0 ppm. Beer's law is obeyed through this
range.
3.3.1.3 Automated Analyses--Commercially available instruments for the
automatic analysis of NO are based on perhaps a dozen principles. Some
J\.
of these are colorimetric, amperometric, ion-selective electrode,
electrochemical, chemiluminescence and a variety of spectrometric tech-
niques. The cost range is from about $1,000 to more than $10,000, with
a size range from a portable 7 kilogram module to a room-sized, stationary
450 kilogram system. Skill required to operate an analyzer varies
greatly. Selected analyzers suitable for ambient air field monitoring
applications are discussed here. Complex techniques limited to the
laboratory or special field applications are not included even though
they may have superior performance characteristics. Also, certain
promising field techniques such as second derivative spectrometers are
excluded here owing to insufficient performance information, too rigor-
ous operational requirements or too few instruments sold.
Analyzers representing breakthroughs in the art of measurement are
coming to the forefront but have not been proven.
3.3.1.3.1 Colorimetric analyzers--Colorimetric analyzers usually
represent automating of a manual technique. As with manual methods,
such analyzers measure the optical absorbance of a solution following
reaction with NCL in an air sample. The absorbance is linearly propor-
tional to concentration of the colored species.
51
-------
This type of anlayzer has been in use for nearly 20 years; improve-
ment lately has been in the area of modifications to reagents, improved
electronics, updated optics and so on.
Ten or more manufacturers are selling these analyzers, and an auto-
mated colorimetric method is under consideration for a reference method.
Currently colorimetric analyzers are the most prevalent instrument for
measuring NCL, NO and NO although the chemiluminescent method is
LI X.
rapidly growing in usage. Advantages of colorimetric analyzers are
simplicity, high sensitivity, and with effect controls, specificity.
Problems include reagents leaking at fittings, temperature and pH sensi-
tivity, long response time, reagent purity and interferents in the
sampled air.
Analyzers almost exclusively use a modified Saltzman absorbing
reagent.
Colorimetric NO anlayzers are designed to operate in one of three
J\.
modes: series, parallel and simultaneous.
3.3.1.3.1.1 Series mode--The air sample is drawn through the reagent
wherein N02 reacts to form the azo-dye which is measured and recorded
as NO-. The sample stream still containing NO passes through an oxi-
dizer (e.g., acid permanganate bubbler) becoming N0~ and is then
measured. If the analyzer is properly calibrated, an accurate reading
of N0? and NO is recorded. Summing the two readings provides the NO
LI X.
value.
3.3.1.3.1.2 Parallel mode--The air sample is drawn into the analyzer
where it is split. One flow is passed through the reagent where the
NO- reacts to form the azo-dye which is measured and recorded. The
52
-------
other stream is drawn through an oxidizer where the NO is converted to
N02. The N02 present in the sample stream plus the NO (now N02) reacts
to form the azo-dye which is measured and recorded. This reading is the
NO^ value (N02 + NO). By subtracting the N02 reading the NO value is
obtained. Again, the best results are obtained by a dynamically cali-
brated analyzer.
3.3.1.3.1.3 Simultaneous node--The air stream is split into two por-
tions as it is drawn through the anlyzer. One stream is scrubbed of its
N02 content by drawing it through a filter containing triethanlomine
(TEA) impregnated on firebrick for the removal of NO,, which reacts with
the azo-dye which is measured for the NO reading. The other stream is
drawn through the reagent where the N02 reacts and the azo-dye developed
is measured and recorded as N02. Combining the two readings gives the
NO value.
J\.
Each of three modes has certain advantages and disadvantages.
Generally an instrument using one of the modes, properly attended,
maintained and most important, carefully calibrated, will provide
creditable data. Electronic modules can be added to the systems for a
variety of signal presentation such as direct digital readings of NO,
N07 and NO .
£ J\.
3.3.1.3.2 Chemiluminescent analyzers--EPA has under consideration the
chemiluminescent method as a reference method for N02. Approximately
seven years ago, EPA inaugurated in-house and contract investigations
into measuring oxides of nitrogen by application of chemiluminescent
technology. Light energy—with or without flames--of specific wave-
lengths can be generated by bringing into a chamber a given substance
53
-------
that burns or reacts with another substance to emit light energy detected
by a photomultiplier tube. Application of the principle of measurement
for oxides of nitrogent has passed through several stages.
Today about ten manufacturers are selling oxides of nitrogen analy-
zers using the chemiluminescent principle. They operate over a wide
range of ambient temperature and humidity with minimal effect on accuracy.
The oxides of nitrogen analyzer is designed to measure NO, NCL and
NO in a continuously exchanged ambient air sample. The system reacts
A.
quickly to quantitative changes in NO, N09 and NO content of the sample,
I* X.
while providing a highly stabilized measurement capability for extended
periods without repeated adjustments. An output signal proportional to
the measurement is produced for recording.
The oxides of nitrogen analyzer uses the principle of photometric
detection of chemiluminescence resulting from the flameless gas phase
reaction of NO with ozone (0,). Since this reaction only occurs between
NO and 0,, the amount of NO., is shown by the difference between NO and
o Z .Jt
NO, accomplished by dividing the input sample into two flow streams.
One stream goes directly into the reaction chamber and results in a
measurement of NO. The other stream, defined as NO , goes through a
JO.
catalytic converter changing N02 to NO but having no effect on incoming
NO, and then into the reaction chamber for measurement. The amount of
N09 is determined by subtracting the NO from the NO measurement. The
b X.
analyzer provides automatic cycling through the NO and NO measurements
J\.
with the output difference (N0?) up-dated after each cycle. The normal
cycle time is one or two minutes.
54
-------
Analyzers generally have several ranges. In most cases a range is
0-0.5 ppm full scale with a minimum detectable sensitivity of 5 ppb.
Zero and span drift usually do not exceed one percent in 24 hours.
Nominal weight of the instrument is 25 kg.
The variation produced by one manufacturer is shown in Figure 3.1.
This analyzer measures NO and NO simultaneously by using two cells,
.A.
negating the necessity for cycling between measurements. The manu-
facturer claims improved stability with use of the optical chopper for
automated drift compensation and an improved NO,., to NO converter. The
EPA St. Louis Regional Air Pollution Study (RAPS) has 25 such analyzers
in the Regional Air Monitoring Stations (RAMS). As of November 1974
they are undergoing the Systems Acceptance Test (SAT).
Air
Air
r
Opficol
chopper
Photo -
multiplier
tube
Slgnol
Phase-
sensitive
amplifier
Phose sensing freq. -Tl-ri-n_
NOX
Flow '
constrictors
Diff
amp
Photo -
multiplier
tube
Signal
Phase -
sensitive
amplifier
1^
NO
°NO.
>N09
oNO
Figure 3.1. Oxides of nitrogen chemiluminescent analyzer
55
-------
3.3.1.3.3.1 Tentative method for the calibration of nitric oxide,
Nitrogen Dioxide and Ozone analyzers by gas phase titration--This tech-
nique developed by EPA for calibration of chemi luminescent oxides of
nitrogen and ozone analyzers is a milestone in ambient air quality
measurement. Reported in the Environmental Protection Technology
Series, March 1974 (EPA-R2-75 246), its application substantially
improves rapid calibration ability in the field. Integrity of the
technique can be affirmed in the laboratory by certification using NO-
primary standard permeation tubes.
The gas phase titration (GPT) technique uses the rapid gas phase
reaction between nitric oxide and ozone so that with concentration of
one of the three gases known, the concentration of the other two may be
determined. Cylinder nitric oxide at 50 to 100 ppm in nitrogen is used
as a secondary standard for routine field calibrations. Ozone is added
to excess nitric oxide in the dynamic calibration system, and a chemi -
luminescent nitric oxide monitor indicates changes in concentration.
The decrease observed on the spanned nitric oxide analyzer with addition
of ozone is equivalent to the concentration of nitric oxide consumed,
the concentration of ozone added and the nitrogen dioxide concentration
produced. The advantages of the procedure are that a primary standard
for only one gas is required and that rapid and routine calibrations of
ozone, nitric oxide and nitrogen dioxide monitors may be performed at a
common manifold as shown in Figure 3.2.
56
-------
NEEDLE
VALVE
JLEAN AIR
SOURCE
VENT
AIR FLOWMETER
0 TO 10 LITERS MIN
"A
t
Pll 1 tLRvJ
1/3 SOURCE
-£&»•
QJ
^
RESTRICTION
STANDARD
NO
CYLINDER
REACTION
CHAMBER
FLOWMETER
SAMPLE MANIFOLD-^
MIXING BULB
J
CHEMILUMI-
NESCENT N02
MONITOR
CHEMILUMI-
NESCENT 03
MONITOR
CHEMILUMI-
NESCENT NO
MONITOR
Figure 3.2. Flow scheme for calibration of NO, MX,,
and NO and 0,, monitors by gas phase
titration
3.3.1.3.3 Electrochemical NO and N00 analyzer--Fjiergetics Science, Inc.
" "' ---«-" ~ ~~ --"--•-! i T •"' — -jT, -'-— - -i—rj-i- J-~ -•
has produced a fully protable analyzer to measure NO and N02 separately
and simultaneously.
The detection principle used is an advanced electrochemical method
whereby NO is measured by oxidation and N02 is measured by reduction.
The analyzer has dual ranges for both NO and N02 with the high sensi-
tivity range for N02 being 0-1.0 ppm and for NO, 0-10 ppm full scale.
57
-------
Prototypes with a full-scale sensitivity of 0.5 ppm are being engineered.
Limitations include pronounced temperature sensitivity; the principal
advantage of the analyzer is that it is portable and can be operated with
a minimum of training.
3.3.1.3.4 N0? Fluorescence monitor—A laser-induced fluorescence monitor
for NO- has been developed for EPA. A recent 30-day field study in St.
Louis, Mo. comparing the fluorescent and chemiluminescent methods for
NO,, showed almost no difference between the two procedures. EPA is work-
ing on a system to induce NO- fluorescence with a pulsed high energy
broad bank xeon source. Such a system would be simpler to operate and
considerably less expensive than the laser method.
3.3.1.3.5 Nitrates--An automated instrument for nitrates has been devel-
oped. The system couples a wet microelectrostatic precipitator (Leap
sampler) with a nitrate ion selective electrode. The solution from the
Leap sampler is automatically pumped into a cell containing the ion
selective electrode where the nitrate ion concentration is measured.
The system is capable of measuring 1 yg/m of NO, for a one-hour collec-
tion period. EPA is conducting field and laboratory tests on interfer-
ences and general performance characteristics.
3.3.2 Stationary Sources
The prescribed reference method for NO uses the EPA sample train
J\.
to collect a sample from the stack source and the phenoldisulfonic
1 2
acid method for analysis. ' The method is time-consuming (with respect
to sample analysis) and needs improvement to increase rep ducibility.
Regulations allow for equivalent and alternative methods, but criteria
and procedures for determining equivalency and acceptable alternative
methods have not yet been proposed.
58
-------
Guidelines have been developed for continuous NO monitoring
.X
systems and performance specifications proposed based on evaluation
of commercially available systems. ' ' Commercial systems are avail-
able using various sampling and analytical techniques. Both extractive
and in-situ techniques are available. Analytical methods include:
electronchemical, coulometric, UV, IR and visible absorption spectres-
copy, colorimetric and chemiluminescent.
The reference method procedure is being evaluated to improve its
reproducibility, minimize executive time and delineate critical per-
formance parameters to be used as criteria for equivalent and alterna-
tive methods.
The lead dioxide method for NO developed on contract is being
X.
evaluated as an alternative reference method for power plant emissions
and will be evaluated also for application to other NO sources.
JC
Evaluation studies have been done comparing various sampling
approaches to continuous monitoring of NO emissions, namely, extrac-
-A.
tive, in-situ and remote sensing. A report on these studies is under
way. A study has been completed on delineation and evaluation of
criterical design and performance parameters of interface systems
7
required in the use of extractive monitoring systems. The study
considered various pollutant gases including NO . An in-situ gas
A.
filter correlation NO monitor has been developed as a field proto-
JL
Q
type and is undergoing field evaluation. Contract effort has been
started to evaluate the application of the same technique to remote
sensing of NO by IR emission compared with IR absorption in the in-situ
J\*
g
device. An inexpensive, simple NO monitor is being developed in-house.
-A.
59
-------
Within a year or two enough data will be available to allow
reassessment of research and development activities. At this time
no additional activities are advised beyond completion of ongoing
tasks and two tasks in future plans. One is development and evalua-
tion of a myoglobin method as an alternative method for NO; the other
is the evaluation of conversion efficiency of N0? in the chemilumi-
nescent method for NO monitoring as applied to nitric acid plants.
-J\.
3.3.3 Mobile Sources
Standards and techniques of measurement for nitrogen oxides have
been promulgated for most mobile sources. These methods and technique
result from research and development efforts of the Chemistry and Physics
Laboratory. The analytical system specified throughout was the chemi-
luminescent nitric oxide plus ozone reaction with prior conversion of
nitrogen dioxide to nitric oxide. Standardization and efficiency
determinations are also made with CPL-developed procedures.
3.4 CARBON MONOXIDE MEASUREMENT
3.4.1 Ambient Air
Carbon monoxide (CO) is by far the largest single pollutant found
in urban atmospheres. Although the background level is about 0.01 ppm,
concentrations of 300 ppm have been reported in congested industrialized
urban areas during prolonged periods of air stagnation.
CO is a product of incomplete combustion of carbonaceous fuels.
The largest man-made source comes from the combustion of motor vehicle
fuels. Estimates vary, but CO introduced into the atmosphere in the
United States is from one-half to three-fifths of the global input, cal-
culated to be about 2.25 x 10 kg (250 million tons) annually. Without
60
-------
removal processes this input would be sufficient to raise the background
0.03 ppm per year. The mean residence time has been estimated to be
between 0.2 and three years.
3.4.1.1 Calibration--Calibration of methods to measure carbon monoxide
has probably been the hardest problem faced by the analyst. Probably it
also accounts for measurements discrepancies and arguments that have
arisen from comparing data. When there are problems in calibration,
usually they can be traced to errors in concentration values of calibra-
tion gas sources or differences in the diluent gas. In many instances,
these gases have a CO content higher than the ambient concentration the
analyst has attempted to measure.
It appears that lacking a primary standard, manufacturers have
assayed the concentration of a carefully blended mixture in a reference
cylinder. This cylinder subsequently has been used to calibrate instru-
mentation to determine the content of subsequent mixtures blended to
approximate the concentration requested. The reference cylinder might
be used for many months, and during that period of time, the CO concen-
tration is changing. Depending on the concentration of CO in the
cylinder, the diluent gas and metallurgical aspects of the cylinder,
the decay rate ranges from dramatic to tolerable. Some vendors peri-
odically reassay the reference cylinder or prepare new mixtures to
minimize the problem. Meanwhile, the cylinder concentration in the
"certified" mixture is shipped to the consumer and change in the
mixture occurs. If the CO concentration is very low, the cylinder wall
oxidized, or reactive impurities present, the CO level will decrease
significantly within the first month before tapering off. The purchaser
61
-------
may use this "standard" cylinder for several years. It is not surpris-
ing, therefore, if standard cylinders are brought together from various
sources and introduced into a single analyzer, that a difference of 7 ppra
between cylinders certified at 25 ppm might be seen.
Aluminum cylinders for pressurized specialty gases introduced within
the last year promise to substantially reduce the calibration problems
of CO methods. The original assay of the mixture must be accurate, and
this can only be accomplished now by rigorous attention to detail and
use of expensive and delicate equipment by a very competent analyst. The
purchase of such blends would be very expensive.
EPA is supporting investigation of generating precisely known con-
centrations of CD according to the scheme:
A -> B + CO,
where A is a compound capable of unimolecular decomposition (e.g., by
pyrolysis) to produce a stoichiometric quantity of CO, and further being
stable and volatile enough to be precisely dispensed from a permeation
device. Metal carbonyls are excluded from consideration because of
toxicity. Support is also being given to improvements in permeation/
diffusion devices expected to facilitate the technique for dispensing
precise small quantities of CO.
3.4.1.2 Manual Measurements--No new manual methods have become popular
in the last five years. There may have been laboratory work on new
manual techniques for ambient air analysis, but it has not been docu-
mented in air pollution literature.
3.4.1.2.1 American Public Health Association (APHA) Method--About two
years ago the Intersociety Committee adopted the method for colorimetric
62
-------
determination of CO that uses a solution of the silver salt of p-
sulfaminobenzoic acid. The solution is made by adding to 20 ml of
p-sulfaminobenzoic acid, 20 ml of 0.1 M AgNo3 and 10 ml of 1.0 M HaOH
with shaking. Ten ml of the solution is transferred to a 125 ml
Erlenmeyer flask with a standard taper joint at the mouth. A stopcock
adapter is placed into this joint and the assembly is evacuated until
the solution begins to bubble, after which the stopcock is closed and
the assembly is removed. The sample is introduced simply by opening
the stopcock. Total volume of flask plus stopcock adapter is deter-
mined by measuring the water required to fill the apparatus.
After introduction of the sample, the stopcock is again closed and
the flask allowed to shake on a wrist shaker for two hours. A portion
of the solution is transferred to a 1 cm spectrophotometer cell and the
absorbance is determined at 425 nm for 0-400 ppm samples. The absorp-
tion at this wavelength results from the formation of a colloidal sus-
pension of silver produced when the silver salt is reduced by CO.
Acetylene, olefins, H.?S and aldehydes in high concentration will
interfere with this method either by reducing the silver or by forma-
tion of insoluble silver sulfide. These interferences can be removed
by passing the sample through a silica gel filter containing HgSO..
When standard samples are analyzed with each series of unknowns, repli-
cate analyses agree to within ± 5%. Because this method is empirical,
it is important that all conditions be maintained for the analysis of
samples and standards. These include shaking time, the interval of
time between completion of shaking and measurement and the temperature.
63
-------
A long (two hours) shaking period is required because the reaction rate
is slow and shaking increases the contact between reagent and gas sample.
3.4.1.2.2 Draegor Type Tubes--There has been marked improvement in the
last year or two in lower detection limits for devices manufactured by
several firms based on the adsorption of CD on solid surfaces with a
simultaneous color reaction. The method employs a hand pump which sucks
air through a specially designed glass test tube containing the reagent.
Depending on the kind of tube, the length of stain is either a measure
of CO concentration in the tested air or a colorimetric comparison
between standards and sample. Formerly these Draeger-type tubes were
considered a semi-quantitative means for rapidly detecting dangerous CO
concentration in an industrial environment. Now, with care, concentra-
tions as low as 2 ppm can be estimated with the length of stain tubes.
The tubes are sealed and just prior to use, the glass tips are broken
off the tubing ends, with the designated end coupled to the pump. The
tubes are not reuseable. While it is not appropriate to use this
technique to obtain a mean concentration (e.g., hourly average), it is
useful for determining the approximate CO in an indoor/outdoor surround-
ing, a vehicle or other type enclosure.
3.4.1.3 Automated Analysis--A number of instrumental methods for auto-
mated analysis of ambient CO concentration are very satisfactory.
Several instrumental methods that have been marketed have not been
suitable, including amperometric, mercury vapor and certain species of
electrochemical analyzers.
3.4.1.3.1 NDIR--The most popular technique used in commercial instru-
mentation for CO monitoring is the nondispersive infrared (NDIR)
64
-------
technique. This is also the basis for the reference method specified by
EPA in Federal Register 36 (228) 22384, 1971.
NDIR analysis is based on broad-band spectral absorption sensitized
for CO by means of a selective detector and optical filters. Commercial
instruments for ambient CO concentrations are available from about four
or five manufacturers in the U.S. and about the same number abroad.
Typical nondispersive infrared analyzers use a double-beam arrange-
ment of the type shown in Figure 3. Infrared radiation from the fila-
ments is directed onto two cells: (1) a reference cell filled with a
non-infrared absorbing gas, such as nitrogen or argon, and (2) a sample
cell through which the sample air containing the gas of interest is con-
tinuously drawn. The detector consists of a gas-impermeable double
chamber with a flexible metal diaphragm as a partition. Both chambers
are filled with the gas of interest. The infrared radiation passing
through the reference cell enters one chamber of the detector, and
radiation passing through the sample cell enters the other chamber. The
gas in each chamber is heated by incoming energy, causing a pressure
increase in the two chambers. The pressure rise is greater in the
chamber receiving radiation from the reference cell, since a portion of
the radiation transmitted through the sample cell has been absorbed by
the gas of interest before entering the chamber on the sample cell side.
This difference causes a diaphragm displacement which is electronically
measured as a capacitance ctemge. The infrared radiation is chopped by
an optical chopper to cause a periodic capacitance change, which modu-
lates a radiofrequency signal from an oscillator. The signal is subse-
quently demodulated, amplified, and the output fed to a meter or a recorder.
65
-------
NDIR instruments are usually subject to interference because other
gases (e.g., H~0 and CCL) absorb at the wavelengths of CO. To eliminate
or minimize interference, auxiliary absorption cells or optical filters
are incorporated to remove most radiation at wavelengths at which the
interfering substances absorb. This could be achieved, for example, by
placing cells containing CCL in both light beams pictured in Figure 3.
For monitoring CO in ambient air, water vapor is the main interfer-
ing substance. Even using corrective filters, the discrimination ratio
may be as much as 4000-7000:1 (i.e., 4000-7000 ppm of water vapor causes
the meter to respond as if there were 1 ppm of CO). This could result
in an error of as much as +5 ppm under humid conditions, although +3 ppm
is more typical. Use of refrigerated dehumidification or drying traps
reduces error to <1 ppm. Other sources of error include gas leaks in
detector and reference cells, nonlinear response and drift in the elec-
tronic circuits. However, if conditions are optimum most analyzers set
to span at 50 ppm full scale will reproduce to ±0.5.
Infrared
source
uu
I
uu
I
Reference
Chopper
Sample
cel
Sample in*-
o_om
Somple out
o^o-A^- ^
Diophagm .
distended?
. • ••••I
-^Component
* of Interest
/Other
molecules
Control unit
Figure 3.3. NDIR analyzer
66
-------
3.4.1.3.2 Catalytic Oxidation--Detection of CO by catalytic oxidation
has been reported in the literature over the years and analyzers incor-
porating the principle have been marketed. In the past, full-scale
sensitivity has been about 0-500 ppm.
Within the last two years Mine Safety Appliances, Inc. (MSA) has
converted CO to C02 by a catalyst called Hopcalite, and temperature
increases as a result of the oxidation proportional to the CO concentra-
tion. Present MSA analyzers are measuring 0-10 ppm full scale reliably
with drift tolerances for zero and span about ±1% per day. Shortcomings
of this type instrument such as sensitivity to interferences have been
practically eliminated in MSA's most recent design. The cost of the
analyzer is competitive with NDIR analyzers. The same sensitivity can
be achieved using gas chromatography techniques, but this analyzer is
continuous and provides a real time analog signal. Unknown is how well
the instrument will perform over a period of years.
3.4.1.3.3 Gas Chromatography—Gas chromatographic (GC) analyzers
encompassing the automated measurement of CO have passed from the
prototype stage to commercial production. Several hundred GC analyzers
have been sold in the U.S. by a half-dozen manufacturers. Twenty-five
St. Louis RAMS stations are equipped with the Beckman model 6800 which
is designed to measure total hydrocarbons, methane and carbon monoxide.
The first automated GC analyzer for analysis of methane and carbon
monoxide was developed by EPA. Commercial systems essentially are
copies of the original design except for augmented analytical capability
to include total hydrocarbons and in some cases other hydrocarbon species
such as ethylene and acetylene. The basic system has an electronic
67
-------
program to cause the introduction of two sample aliquots. One is ana-
lyzed for its total hydrocarbon content by direct introduction into a
flame ionization detector. The second aliquot is first transferred
through a stripper column via a carrier gas. C02, water and all
\inwanted species of hydrocarbons are removed by the stripper column
before it is backflushed for the next analysis. The eluting hydrocar-
bons and CO pass through a molecular sieve column, where the substances
are separated, and through a catalytic reactor to a flame ionization
detector (FID). The hydrocarbons pass through the catalytic reactor
quantitatively and the CO is quantitatively converted to methane which
is then measured by the FID.
The system is designed for semi-continuous operation and can per-
form an analysis about every five minutes. It has a linear output for
measured substances and a wide operating range that suggests its use
both in heavily polluted areas and relatively clean locations. Concen-
tration of CO from about 0.03 pm to several hundred ppm can be
measured with this instrument. The major disadvantage of this analyzer
is that its complexity requires a highly trained, skilled operator.
Data is not available for from several hours to several days following
installation while the system stabilizes.
3.4.1.3.4 Dual Isotope Fluorescence--Modification of the NDIR method
has been made using the principle of dual isotope fluorescence (DIP).
The technique consists of absorption spectroscopy in which the source
of energy is fluorescence of CO. This gives excellent wavelength cor-
relation, high signal-to-noise ratio and discrimination. The system
is illustrated in Figure 3.4. Infrared radiation from the Broadband IR
68
-------
source stimulates the CO molecules in the sealed fluorescent cell that in
turn provide fluorescent radiation as the source of energy for absorption
1 f\ T ft
measurement. The chopper cells containing C 0 and C 0 are part of the
analytical scheme whereby the measurement is uniquely sensitive to the
1 f\ 1 ft
presence of CO in the sample tube. If the C 0 and C 0 signals are
defined as the "A" and "B" signals, respectively, only the "A" signal
will be attenuated by the presence of CO gas in the sample tube since
the natural isotopic abundance of C 0 is only 0.2% of C 0. The pro-
cessing electronics generate an output proportional to the quantity *
B-A
This expression shows sensitivity to differential absorption of the two
signals. In addition, maintenance of the "B" signal at fixed amplitude
by automatic gain control allows the measurement to be independent of
variations in source power of detector response.
Relatively high cost and a reported high incidence of malfunction
and temperature sensitivity have held down widespread acceptance of the
analyzer, but the latter two problems are being overcome. The license
to manufacture the analyzer was purchased about a year ago by Beckman
Instruments, Inc. for Andros, Inc.
69
-------
Sample Gas
Inlet
Signal
Processing
Electronics
A. Synchronous
Spin Motor
G. Isotope Outlet
Chopper Wheel
CO Concentration
Meter
B. Broadband IR H. Filter, Cell
Stimulation Source (CO)
C. Broadband IR Inlet I. IR Outlet
Chopper Wheel Window
D. Inlet Windows
J. Magnet Pickup
For Sync Pulses
E. Dual-Isotope-
Fluorescent Cell
F. Filter Cell
(CI60)
Figure 3.4. Dual isotope fluorescence analyzer
3.4.1.3.5 Electrochemical Analyzer--At the time of this writing, Ener-
getics Science, Inc. (ESI) had sold about 2500 "Ecolyzers," with the
number sold now over 100 per month. In the future it could be the most
prevalent instrument in the U.S. for ambient CO analysis. Because of
the Ecolyzer's importance, we are quoting the entire conclusion of an
EPA, ERG, Environmental Monitoring and Support Laboratory manuscript
presented at the Instrument Society of America Symposium, May 13-15, 1974:
"ESI has made available to the air pollution
analyst a low cost useful instrument. The ana-
lyst must be alert to the analyzer's limitations.
He should keep them in mind in using the analyzer
70
-------
and exercise caution in assessing the values
obtained in terms of absolute accuracy.
We view the Ecolyzer as a portable ana-
lyzer that gives us a capability for rapidly
sighting potential or real dangers be it in
a garage, tunnel, intersection, industrial
plant or inside a conveyance. It is an ana-
lyzer that is suited for that type of
emergency episodes where one need not be as
concerned about accuracy as about the con-
centration in terms of a hazardous situation.
Portability of the analyzer makes it useful
for locating a "hot spot," a place where
higher concentrations occur and if interest
lies in that direction, locate a permanent
sampling site. The Ecolyzer should be kept
in a condition of readiness for the activi-
ties mentioned above. A weekly check of
its operational condition and calibration
would be in order. A spare sensor on hand
would likely be desirable.
The Ecolyzer is not looked upon as
desirable for continuous operation at a
fixed monitoring site. Problems of lin-
earity, drift, temperature sensitivity,
rise vs. fall time are potential sources
of error that could prove misleading.
A person with little training can take
an operating instrument into the field and
obtain satisfactory results."
The calibrated pumps a constant flow of sample air through a filter
and sensor. Carbon monoxide in the sample air is electro-oxidized to
carbon dioxide (CO-) at a catalytically active electrode in an electro-
lyte as shown in the reaction: CO + H20 + C02 + 2H+ + 2e.
The electro-oxidation of CO occurs at a sensing electrode in a
sulfuric acid electrolyte. The counter electrode is the cathode where
oxygen reduction occurs. Additionally, a reference electrode is neces-
sary for operation of a potentiostat to balance the system. In the
electro-oxidizing reaction, current flow between sensing and counter
71
-------
electrodes is approximately proportional to the CO concentration of the
sample and is registered on a meter or an optional recorder.
3.4.2 Stationary Sources
A large number of commercial CO analyzers are available including
manual methods (colorimetric) and instrumental methods (catalytic oxi-
dation, electro-chemical cell, gas chromotograph--flame ionization, NDIR
and DIR). Gas-filter correlation (GFC) is an advanced form of NDIR
which has all but eliminated the interference from water vapor, C02 and
particulates. By combining GFC with an in-situ, or non-extractive cross-
stack measurement, a continuous, accurate CO measurement can be made
without problems of sample extraction and handling. One instrument using
this method is commercially available, and a second is expected to be on
the market in early 1975. The GFC technique is also applicable to remote,
single-ended measurement of CO emissions.
3.5 HYDROCARBONS
3.5.1 Ambient Air
Biological processes, petroleum, coal and natural gas account for
almost the total hydrocarbon loading of the atmosphere.
Complex problems are involved in study of the reactants and products
of atmospheric photochemical reactions, one being the determination of
atmospheric hydrocarbon levels and their relation to hydrocarbon emission
from vehicles and other sources. The greatest interest is in those hydro-
carbons participating in photochemical reactions.
While methane is innocuous and does not participate in photoinitiated
reactions, it does constitute a major portion of hydrocarbons present in
the air, and in certain areas and atmospheric conditions has accounted
72
-------
for more than 95% of hydrocarbons present. Typically CH. makes up
nearly all the nonreactive fraction.
Thus a rough correlation exists between "reactive hydrocarbons"
and "total hydrocarbons less methane." Pending development of a type-
specific method for determining reactive hydrocarbons, a compound
specific method has been found to be useful and is now used in setting
national primary and secondary ambient air quality standards for
hydrocarbons.
The standard as written does not allow for natural sources of
hydrocarbons such as the terpenes, which emit into the global atmos-
phere an estimated 1.8 x 10 kg/yr. The terpenes are associated with
forest areas and the blue haze of the Appalachian Mountain region.
Concentrations of total hydrocarbons less methane have been found in
remote rural surroundings at levels close to the standard. The par-
ticipation of naturally occurring hydrocarbons in photochemical reac-
tions producing oxidants is under investigation.
There are limitations to the categorization of all hydrocarbons
of other-than-methane species into a single grouping. Given the same
conditions for photoreaction, half-life for some hydrocarbons will be
minutes while for others it will be hours to days. It is known that
the reactivity (make-up) of the grouping may differ markedly from one
urban area to another. For example, gas chromatographic analysis has
shown that within the same urban area (the Los Angeles basin), make-up
of the grouping is such that relative reactivity (near the Pacific
coast in the morning) is notably greater than the same carbon concen-
tration of the same air parcel reaching another portion of the city
later in the day (inland Azusa area).
73
-------
3.5.1.1 Calibration--Blends of CH. in air, pressurized in gas cylinders,
have been used over the years for field calibration of HC analyzers. The
mixture even at low concentration is stable for long time periods.
Capability was developed as early as 1965 for preparing concentrations
in the ppm region with a tolerance of ±0.1 ppm. Although demanding and
time-consuming, the technique gave EPA an ability to evaluate the product
of commercial suppliers. Until several years ago, cylinders delivered by
manufacturers had a certified analysis either moderately or grossly
inaccurate, although there was one reliable gas supplier. Recognizing
a need, competitors have now upgraded their products so that accurate
standards of low molecular weight hydrocarbons are available from a
number of them. In the laboratory and for special studies, mixtures of
hydrocarbons, including the reactive species, can be made by skilled
personnel in inert plastic bags or chambers for specialized studies or
calibration of special analytical methods. Included are gas chromato-
graphs and sophisticated 1R, W and fluorescent analyzers. This in-
house capability has helped EPA study atmospheric kinetics and photo-
chemical reactions.
Presently the agency has the problem of an acceptable hydrocarbon
to calibrate analyzers measuring the fraction of other-than-methane
hydrocarbons in the atmosphere.
Very accurate concentrations of one or more hydrocarbons can be
delivered using the EPA permeation technology. At least a score of
permeation tubes containing different species of hydrocarbons have been
successfully fabricated and used for this purpose. Many are commer-
cially available.
74
-------
3.5.1.2 Manual Methods--Improved collection media for hydrocarbons has
been the object of several investigations in recent years. Subsequent
stripping of the hydrocarbons for indentification by mass spectrometry
or quantification by gas chromatography has helped the laboratory
researcher in profiling more than two hundred hydrocarbons found in
polluted atmospheres.
No new wet chemical methods of significance have surfaced in the
last seven or eight years. In Air Pollution, Vol. II, ed. Stern,
A. C. P. 15 Academic Press, 1968, Altshuller described manual methods
for measuring certain hydrocarbon species or groups.
3.5.1.3 Automated Analysis--As stated earlier, EPA's interest in
hydrocarbon measurement is primarily related to reactive hydrocarbons
grouped into a category referred to as hydrocarbons corrected for
methane, which is in turn called hydrocarbons as methane (i.e., total
count of carbon atoms, C, in the measured sample) minus the methane,
CH., concentration. Conventional instrumental systems measure total
hydrocarbon concentration and methane separately.
The Federal Register, 36^ (158) 15491, August 14, 1971 states "it
may be assumed that the degree of total hydrocarbon emission reduc-
tion necessary for the attainment and maintenance of national standard
for photochemical oxidants will also be adequate for the attainment of
the national standards for hydrocarbons." EPA therefore does not
require states to monitor hydrocarbons for compliance with national
ambient air standards.
In 1974 an EPA investigation found a lack of comparability between
different commercial analyzers for non-methane hydrocarbon. Reasons
for the discrepancies are still being investigated.
75
-------
3.5.1.3.1 Gas Chromatography--A procedure for collection and concentra-
tion of trace hydrocarbons and rigidly controlled introduction into a GC
where separation of compounds takes place and compounds elute into a
flame ionization detector. Compounds generally elute in relation to
their mass, with heaviest species last on the chromatogram. The peaks
have qualitative and quantitative meaning gained by systematic empirical
fingerprinting through analysis of measured concentrations of known
hydrocarbons. Advances in chromatography have allowed speed-up of the
analysis by temperature programming automatic peak identification and
quantification using state-of-the-art magnetic tape recorders with
sophisticated peak identification and integration (concentration) print-
out systems.
3.5.1.3.2 Programmed Gas Chromatograph--The reference method for deter-
mination of hydrocarbons corrected for methane is contained in Federal
Register, 36^ (84), April 30, 1971, as described earlier. This reference
method is being reevaluated. Present commercially available instruments
using the technique are expensive, complex and demand considerable
operator skill. They are marginal in ability to measure hydrocarbons
corrected for methane on a routine unattended basis at or below levels
required by the national ambient air standard.
3.5.1.3.3 Reactive Hydrocarbon Analyzer—Bendix Process Instruments
Division has made available its model 8201 Ambient Hydrocarbon Analyzer
designed to measure what Bendix calls "reactive" hydrocarbons (total
hydrocarbons less CH.). In operation, a programmer causes an aliquot
of sample air to be transported through a column via a carrier gas.
Hydrocarbons other than CH. are stripped from the gas stream in the
76
-------
column. Following the elution of CH., the column is back-flushed, and
the CH. is carried to a FID for measurement, amplification and signal
storage. Then a discrete volume of ambient air is passed directly to
the detector for measuring total hydrocarbons; the signal is also
amplified and stored. The concentration of CH. is then electronically
subtracted from total hydrocarbon concentration for the measurement of
"reactive" hydrocarbons. During each programmed cycle, approximately
three minutes long, automatic zeroing of the detector output ensures a
stable base line.
When first evaluated this analyzer had several weaknesses that
were almost eliminated by substitution of more rugged components and
plumbing changes. Currently there are about 300 such units in the
field.
3.5.1.3.4 Dual Channel Hydrocarbon Analyzer--The Mine Safety Appliances
Company (MSA) instrument is nearly a complete system in itself, and is
a recent commercial development in hydrocarbon instrumentation. The
device is independent of programming, mechanical and electromechanical
functions and has a built-in supply of zero gas and purified combustion
air. All that must be supplied is hydrogen and electrical power. (See
Figure 3.5.) Operation entails continuous measurement of two gas
streams: one stream containing only CH., the other THC. The two
streams are simultaneously introduced into a dual flame detector mounted
in a constant-temperature-controlled oven. The CH. stream is ambient
air passed through a catalyst column containing Hopcalite (TM) at 260°C.
Essentially all hydrocarbons except CH. are converted to carbon dioxide
(CO) and water vapor, and all CH. is passed to the detector. The
77
-------
instrument, designated 11-2 by MSA, has a built-in pump and catalytic
oxidizer supplying purified combustion air and is used as a source of
zero air for calibration of the analyzer. The electronic design pro-
vides for continuous voltage generation proportional to the total
hydrocarbon, QFL and nonmethane concentrations. (See Figure 3.6.)
Both sides of the system are calibrated using a known concentration
of CH, in air; range selectors plus manually adjusted potentiometers
permit exact alignment so that hydrocarbon other than CFL will be
accurately reflected as nonmethane.
AMIIENT
AIR ••
DUAL HFID
r
}
1
1
i
-\
j
GAS INLETS FROM
DISTRIBUTION BUCK
Figure 3.5. Block diagram of dual hydrogen flame detector
78
-------
Figure 3.6. Electronic diagram of dual flame detector
3.5.1.3.5 Ghemiluminescence---A new technique has been developed by EPA
which reacts hydrocarbons in air with atomic nitrogen to produce a
chemiluminescence approximately proportional to the photochemical
reactivity of hydrocarbons present in the air sample. The method has
applicability for measuring reactive hydrocarbons species in auto
exhaust as well as for measuring the photochemical potential of air
samples containing hydrocarbons.
Another detector developed by EPA is based on the reaction of ole-
finic organics with an excess of ozone. The method is analogous to the
ozone, ethylene reaction currently used to monitor ambient ozone except
that the roles of ozone and ethylene (olefinic organic) are reversed.
The method gives either a single measure of reactivity or, with a gas
chromatographic column, can monitor individual hydrocarbons.
79
-------
3.5.2 Stationary Sources
Ambient air quality standards for hydrocarbons relate the photochemi-
cal reactivity of these species to oxidant formation so that any future
requirement on measurement of emissions from stationary sources must con-
sider: (1) emissions which react mutually or in the presence of other
atmospheric contaminants to produce oxidant, and (2) emissions which
react photochemically and serve as precursors of oxidant formation.
The use of present analytical hydrocarbon measurement techniques
applied to source emissions is limited. Both total hydrocarbon and non-
methane hydrocarbon instrumentation used for ambient air measurement
have been used, and there is a need for development of adequate sample
interface systems. Monitoring systems based on process gas chromato-
graphy have been shown to be impractical, since separation schemes would
have to be developed on a source-by-source basis. Manual or "compliance-
type" methodology requires a sample collection scheme using bag collec-
tion, freeze-out traps, or charcoal adsorption devices for collecting
samples in the field with laboratory analysis using gas chromatography.
At present, our research and development efforts involve a grab-
sampling system which can collect emissions in class fractions, i.e.,
paraffins, olefins, aroraatics, oxygenates and halogenated compounds.
The class fractions can then be analyzed collectively in the laboratory
by a total class hydrocarbon measurement or by on-column injection into
a chromatograph for identification of each species within a class. The
objective is to provide basic emission data from a variety of source
emissions.
80
-------
3.5.3 Mobile Sources
Procedures for measuring total pollutant hydrocarbons and condi-
tions under which they are measured are defined in the Federal Register
based on their ability to participate in oxidant forming reactions.
Mobile sources thus regulated include light- and heavy-duty over the
raced vehicles and aircraft. Qxidants per se are not emitted directly
from motor vehicles, but formed later after photochemical reactions.
Most hydrocarbons are capable of forming oxidants photochemically in
the presence of NO ; exceptions are methane, propane, acetylene and
Ji.
benzene. Other organic emissions such as oxygenates and nitrogen com-
pounds may also participate while higher hydrocarbons including such
classes as the polynuclear aromatics, i.e., benzo-a-pyrene, may not be
reactive. These compounds may be either gases or particulates depend-
ing upon conditions, molecular weight, concentrations, etc.
Mobile sources are diverse, each requiring special measurement
techniques. In all cases, mobile sources are considered diffuse
sources, i.e., sources in which no individual is a significant con-
tributor but the sum of all sources is of great importance. Mobile
source emissions are directly related to conditions of operation as
well as to other factors, such as environment and fuel.
Sample collection is critical to all sources and is of overriding
importance with mobile sources. The constant volume sampler (CVS) has
been specified in the regulations as the method used to integrate
samples collected over a total operating cycle, allowing the total
mass of pollutant emitted to be measured and mass weighed in relation
to its impact. The procedure causes an average dilution of about 10
81
-------
to 1. As applied to well-controlled vehicles, this extra diffusion
increases the sensitivity requirements of the analytical system to
levels approaching those of ambient air.
Dilution of the CVS stream provides cooling and time for particu-
late formation to occur. The cooling is especially important as
temperature fluctuations of many hundred degrees occur. The CVS system
as specified allows the cycle to be broken into three segments with
each segment analyzed and weighed separately. Routine measurements for
hydrocarbons are made on samples collected in bags, while more reactive
materials such as oxygenates are collected directly from the gas stream.
The specified analytical system uses the Flame lonization Detector
(FID). This measures total carbon concentration in the flame and there-
fore allows an estimate of mass emissions to be made although mole con-
centrations are not defined. It is used in detailed gas chromatographic
(GC) systems as well as in total hydrocarbon analyzers. The operation
in the two systems is somewhat different, with the total FID operating
as a premixed flame while the GC system operates as diffusion flames.
This difference has important consequences when catalytically controlled
vehicles are tested, since individual response factors for hydrocarbons
are dependent on operating conditions. Methane is more refractory to
catalytic composition than other materials and may be markedly enhanced
in vehicles so equipped. Recent work, both in-house and by others, con-
tinues to show the importance of response factors and effects of operat-
ing conditions and oxygen content on them.
GC systems have been designed to routinely measure non-reactive
components. Methane alone may be measured by simplified GC systems or
82
-------
directly by infrared gas correlation spectrometry. Standard infrared
techniques are inhibited by water vapor interferences. A rejection
ratio of 10 or greater is required.
Detailed hydrocarbons are determined by sophisticated GC methods.
In this laboratory, 52 individual compounds are measured by an auto-
mated system programmed from -100°C to 75°C. The specificity necessary
determines the type of separations made and other systems have been
used.
Higher molecular weight hydrocarbons are determined by several
methods. Usually other material is first trapped, isolated, on a short
column containing appropriate packing material such as chromosorb 101
or firebrick and "tris." This precolumn or collection column becomes
the first part of the analytical column. For diesel emissions the most
common columns provide boiling point separations only, since there are
too many isomers present to separate in a reasonable time. These
techniques are also applicable to higher molecular weight substituted
compounds such as oxygenates.
Use of liquid chromatography is becoming widespread for the sepa-
ration of higher organics, and a system combining these techniques is
used to evaluate diesel and jet odor.
Higher molecular weight materials are very susceptible to wall
losses and reactions. All f;urfaces must be heated to prevent such
losses and much earlier work must be doubted due to poor practice.
Oxygenates is a commonly found class of organics in emissions of
all mobile sources. They are usually analyzed after derivative forma-
tion either by spectrophotometic or chromatographic procedures. Cur-
rent methodology is not fully satisfactory because of interferences
83
-------
and difficulty in performing analysis. Spectrophotometric methods may
be applicable to specific compounds such as formaldehyde.
Polynuclear aromatics analysis is not completely satisfactory at
this time. Procedures are complicated and reproducibility and sensi-
tivity are problems. Screening methods have been developed for specific
members of this series, but even these are non-routine, the large number
of compounds in the class further complicating analysis.
Other non-regulated organics are just now being sought. These
involve compounds of nitrogen and sulfur.
3.6 PHOTOCHEMICAL OXIDANTS
3.6.1 Ambient Air
Photochemical oxidants (0 ) are a class of atmospheric pollutants
-A.
arising from reactions between hydrocarbons and oxides of nitrogen in
the presence of sunlight. The class is known primarily for ozone
peroxyacyl nitrates--mostly peroxacetl nitrate (PAN)--and hydrogen
peroxide. The latter two compounds in a highly oxidant atmosphere like
Los Angeles may each represent 10% of the total (0 ) concentration,
.A.
while in a reducing atmosphere like New York City, they each may account
for 2% of the total. Their chemical formulas are:
ozone 0.,
peroxacetyl nitrate (PAN) CH3COON02
hydrogen peroxide H^O^
Photochemical oxidants traditionally have been defined by air pol-
lution agencies as atmospheric substances which will oxidize certain
reagents, the most frequently used being a solution of neutral-phosphate-
buffered potassium iodide (KI). This reagent responds to ozone, nitrogen
84
-------
dioxide and the peroxyacyl nitrates, and, to a lesser extent, to
hydrogen peroxide. Reducing agents such as sulfur dioxide (S09) have
Li
the opposite effect of oxidants on the reagent solution. The KI
method may be used, therefore, to provide a measure of net oxidizing
properties of atmospheric pollutants without complete discrimination
as to species of oxidants or reducing agents.
Various terms describe photochemical oxidant concentrations. Two
of the terms, "oxidant" and "total oxidant," are used to describe the
"net" oxidizing ability of the sampled air. Use of these non-specific
terms is required because measurement (usually by the KI method) does
not always differentiate between oxidizing or reducing agents. The
terms "corrected oxidant" or "adjusted oxidant" show that the measure-
ments have been corrected for certain known responses (usually SCL and
N02) other than ozone or peroxyacetyl nitrate. The term "ozone" and
"peroxyacetyl nitrate" show that the measurement method is specific
for these gases.
We emphasize that the terms, "photochemical oxidant," "total
oxidant," "oxidant," "corrected oxidant," and "adjusted oxidant," have
no exact meaning since response obtained in the KI method depends on
the presence of various substances responding in a different manner.
3 6.1.1 Calibration--Ozone generator calibration systems are available
from nearly a dozen commercial suppliers. Generally they are quality
products stable, reliable and able to be taken into the field. EPA was
responsible for the basic design of the generators (Figure 3.7). The
ozone source is an ultraviolet mercury lamp which irradiating a quartz
tube through which clean air flows at 5-10 liters/minute. Ozone
85
-------
concentrations from 0 to 1 ppm can be generated by varying UV intensity
with a variable shield around the lamp envelope and the flow rate is
controlled by a needle valve and measured by a rotameter. The ozonized
air passes to a manifold from which the monitor under test draws its
sample.
PEN RAY' PHOTOLYSIS
LAMP
VAHIABI e SHIEID
Figure 3.7. Typical ozone source, dilution
and manifold system
In using ozone generators, care must be taken to eliminate all
impurities including NO in the diluent air stream.
Thus far only a source of ozone has been described. This source,
which has a numerically indexed variable shield for precisely reproduc-
ing settings, must be calibrated using a primary calibration method.
The presently accepted primary standard calibration method is the manual
86
-------
wet-chemical KI method described in Federal Register, _36_ (84) 8196,
April 30, 1971.
3.6.1.2 Manual Measurement--Methods used for the measurement of photo-
chemical oxidants (0 ) range from those responsive to all oxidants and
reductants (and therefore indicate to a degree a "net" oxidizing effect)
to those which respond most to a specific oxidant such as 0,.
It is possible to apply corrections to a value obtained using a
given manual method to improve accuracy. For example, using the (to-be-
discussed) KI method one can correct for the positive interference of
NCL and negative interference of SCL if concentrations of each are known.
An ideal method for 0 does not exist and is not. likely to be found.
.A.
Such a method would have an equivalent response for all photochemical
oxidants including 0,, PAN and other peroxyacyl nitrates, the aldehydes
(formaldehyde and acrolein), peroxbenzoyl nitrate, H?0?, alkyl hydro-
Lt I*
peroxides, organic peroxide and as yet unidentified other compounds
believed to account for certain measurement anomalies. The method would
not respond to any atmospheric reductants such as SCL, or oxidants not
photochemically generated, such as most NCL.
Compounds measured can differ dramatically in reactivity, and
therefore the true measure of potential damage to man, other animals,
vegetation and materials would be unknown. Further, some NCL present in
polluted atmospheres is photolytically generated.
Fortunately the problem is not as formidable or perplexing as it
seems. 0, is at least the most abundant of the 0 compounds and ranks
in the top group of reactive species, so that an accurate fix on 0,
concentration would provide a meaningful measure of harmful 0 in the
.X
polluted atmosphere.
87
-------
3.6.1.2.1 Rubber Cracking--A quantitative method for determination of
CL was developed by Haagen-Smit and Bradley in 1951, the essence of the
procedure being to put a flat rubber strip under stress, bending it to
form a tight "u" clipped together at the ends, placing it in a glass
tube and drawing the raw atmosphere directly over it until cracking
appeared at the outer edge as seen using a 3X eyepiece. The exposure
period preceding the appearance of cracks served as an index of the 0,.
We mention this procedure since high 0., concentrations from the 1950's
through '60's were reported by the Los Angeles County Air Pollution Con-
trol District (LAAPCD) using rubber cracking.
Variations have been investigated including microscopic measure-
ment of crack depth, elongation of a weighted rubber strip and continu-
ous unattended open exposure of a strip for a prolonged time with
subsequent optical examinations to estimate total material effect. The
original method required a skilled operator, was time-consuming and
necessitated close scrutinty during analysis. The other procedures had
shortcomings limiting their use to where error tolerances were not
stringent.
3.6.1.2.2 Colorimetric Procedures--As with colorimetric procedures dis-
cussed earlier, optical absorbance of a solution is measured spectro-
photometrically: the absorbance is linearly proportional to concentration
of the colored species which ideally is directly proportional to the con-
centration of pollutant.
There are about eight procedural!zed manual methods. Manual methods
have various advantages and disadvantages, depending on the information
desired, personnel available and other considerations.
88
-------
Ambient manual oxidant sampling using wet chemistry depends on a
collection device as illustrated in Figure 3.8. Sampling precautions are
strict owing to reactivity of the pollutant. Sample systems should be
fabricated preferably from glass or teflon. Location of the intake
probe can be critical, and consultation with an expert is advised.
3
E mm
— -
(
\
(
E
E
CO
j
INSIDE
CLEARANCE
D
*-
\_, \
10mm OD
30
25
20
15
10
'
UU-
. t
) 24/40,
OUTER
^ WITH ^
IN IER\
s* WAY A
NOZZL
1mm, P
cfm AT
PIECES
CHANG
'
25 mm
OD
CHANGEABLE, MAINTAINING
Figure 3.8. All-glass midget impinger
3.6.1.2.2.1 Neutral Buffered Potassium Iodide Method—This has been
adopted as the tentative standard manual reference method by the
Intersociety Committee and the ASTM Committee D-22 on Sampling and
Analysis of Atmospheres and is also the EPA reference method for
calibration of ozone generators.
89
-------
In this method oxidants are absorbed in a one percent potassium
iodide (KI) solution buffered at pH 6.8 ± 0.2. For ozone the reaction
can be written:
07 + 3KI + H00 -*K+I~ + 2KOH + 0
3 23 2
The liberated tri-iodide ion (I.,) is measured colorimetrically at 352 ran
and is directly related to the concentration of ozone in the air sample.
A 1:1 stoichiometry between moles of 0, absorbed and moles of tri-iodid
ion released is accepted. Sampling efficiency is very high when two
impingers are used in series.
Qxidant measurement in a range from about 0.01 to 10 ppm is possi-
ble. SCL produces negative interference equivalent to 100% of an
equilmolar concentration of CL, but the interference can be removed with
a U-tube containing an absorber located in the sampling train upstream
from the impinger. The absorber usually is of glass fiber paper impreg-
nated with chromium trioxide. Since the SCL absorber oxidizes any nitric
oxide in the air sample to nitrogen dioxide, a correction for nitric
oxide will be required. The chromium trioxide scrubber can, if wet,
cause loss of 0, and NCL- NO- is known to give a response in one percent
KI equivalent to 10% of that for an equimolar concentration of CL. The
contribution of NCL to the oxidant reading is eliminated by concurrently
analyzing for NCL and subtracting one-tenth of the NCL concentration from
total oxidant value.
PAN gives a response approximately equivalent to 50% of that of an
equimolar concentration of 0,. Other oxidizing agents will cause posi-
tive interferences with this method. Other reducing agents, in addition
to S0?, can cause negative interference.
90
-------
Color measurement is recommended within 30 to 60 minutes after
sample collection. Time is required for color development (10-20
minutes), but color will fade; with more than 60 minutes. Evidence is
that the amount of tri-iodine ions liberated by CL depends on design
of the gas bubbler and phosphate concentration of neutral pH. Error
due to these variables is probably about plus or minus 10% of true
value.
The neutral buffered KI procedure was the EPA reference method
for determining oxidants until its replacement in 1971 by the continu-
ous chemilurainescent method. The KI method remained in the primary
dynamic calibration method for instrumental analysis methods.
3.6.1.2.2.2 Alkaline Potassium Iodide Method--This method is for
manual determination of oxidants in the range of about 0.01 ppm-20 ppm.
An absorbing reagent of potassium iodide in sodium hydroxide solution
is used. After sampling the solution is treated with H-CL to remove
SCL interference and is finally acidified with phosphoric acid contain-
ing a small amount of sulfamic acid to release iodine for spectrophoto-
metric measurement.
The advantage over the neutral-buffered potassium iodide method is
that delay between sampling and completion of analysis does not involve
color fading. This procedure requires a stoichiometric correction fac-
tor- -0.65 mole of iodine is liberated for each mole of ozone absorbed.
Rapid acidification is necessary because of certain irreversible losses
of microquantities of iodine. The neutral-buffered KI method is prefer-
red for its simplicity over the alkaline KI method.
91
-------
3.6.1.2.2.3 Phenolphthalin Method--This method is based on oxidation
of phenolphthalin to phenolphthalein. Phenolphthalein is pink in alka-
line solution.
Because the technique is very sensitive to pH and temperature
changes, it has dropped out of usage. Concentration measurements
obtained by it are about twice those obtained by the neutral-buffered
KI method so that it is necessary to correct data for the difference.
3.6.1.2.2.4 Other Manual Methods--Miscellaneous acceptable methods were
developed in the late fifties and throughout the sixties, with EPA
researchers responsible for the l,2-di-(4-Pyridyl) ethylene method -
(DPE) method, Ferrous Ammonium Sulfate method and Saltzman and Gilbert
method based on the reaction of 0, with NO to form NO- with wet chemi-
cal analysis for NO™. Outside of EPA were the 4-allyl-2-methoxyphenol
method - (Eugenol method, sodium diphenylamine sulfonate method and Acid
Potassium Iodide method) - (Deutsch) method.
It is clear why some methods have not survived. The Ferrous
Ammonium Sulfate method was used in EPA's National Air Sampling Network
(NASN) for 24-hour sampling of total oxidant. Correlation between
measurements by this method and those of the continuous colorimetric
analyzer were poor and NASN discontinued its use. All methods used in
the field lacked some desirable characteristic. Part of the problem
may be lack of a clear-cut description of the substance to be measured;
publication of the "equivalency document" to be discussed later should
provide needed criteria.
3.6.1.3 Automated Analysis
92
-------
3.6.1.3.1 Colorimetric Analyzers--Colorimetric continuous analysis of
total oxidants almost always involves a 10 or 20% solution of neutral-
buffered potassium iodide as the absorbing medium. The first instru-
mental approach in general use was the Colorimetric oxidant recorder
developed in 1953 by Littman and Benoliel. Instruments of this design
using a 201 neutral-buffered KI solution were incorporated into the
Los Angeles County Air Pollution Control District (LACAPCD) air moni-
toring station network in the early 1950's. About 1957 the KI was
reduced to 10% because of objectionable crystal build-up. The instru-
ments continue in use.
Sample air is passed at a known rate through a liquid absorbing
reagent with a vacuum-pump arrangement. Oxidants in contact with the
KI react to liberate the tri--iodine ion (I.,) which is determined
spectrophotometrically at 352 nm and recorded on a strip chart recorder
as ppm oxidant. Figure 3.9 illustrates a typical Colorimetric
analyzer. The reaction with ozone at pH 6.8 ± 0.2 is approximately:
03 + 3KI + H30 + K+I3 + 2KOH +02
Colorimetric oxidant analyzers operate best at atmospheric oxi-
dant levels ranging from 0.01-1.0 ppm of ozone or other equivalent
oxidant.
SO- produces a negative interference equivalent to 1001 of an
equimolar concentration of oxidant. Response to NO- varies with the
reagent formulation and scrubber design. For 10% KI, N02 produces
positive interference of approximately 20%; for 20% KI, an interfer-
ence of approximately 30%.
93
-------
Filters and scrubbers for removing SCL have been used, but informa-
tion as to performance is not available. One device often used consists
of glass fiber strips impregnated with chromium trioxide (CrO,) and
O
sulfuric acid (PLSO.). As discussed in 3.6.2.2.1 (Neutral Buffered
Potassium Iodide Method), these scrubbers convert a portion of the
ambient NO to NO-, necessitating an additional correction factor.
Despite these drawbacks, CrO, scrubbers are the most generally used for
removing S02 interference, and they are particularly useful during
periods of high SO- concentration. (When S02 levels are less that 20%
of the NO concentration, use of this filter is not recommended.)
The contribution of nitrogen dioxide to the oxidant reading is
determined by measuring the instrument response to a known stream of
N0? and concurrently analyzing for NO,, in the atmosphere. The appro-
priate subtraction can then be made to correct for N0~ interference.
r
MAGENT 5TOR.AG-E
A&tNT ST
TWT. KIJ
Figure 3.9. Schematic of colorimetric oxidant analyzer
94
-------
Disadvantages of colorimetric oxidant analyzers are need for rigid
maintenance including changing of reagent, carbon column, pH adjustment,
lubrication of moving parts (pumps) and leak checks. Other disadvan-
tages are sensitivity of flow systems, non-specificity and fragility of
components.
3.6.1.3.2 Amperometric Analyzers--Amperometric analyzers measure the
current necessary to maintain halogen concentration at a fixed level,
the current being proportional to the amount of absorbed reactant.
Analyzers operating on this principle are often referred to as "coulo-
metric" analyzers. In coulometry, the quantity of electons necessary
to oxidize or reduce a substance is measured. Because it is the cur-
rent and not the charge which is measured, "coulometric" analyzers are
more properly called amperometric analyzers.
Oxidant amperometric analyzers operate on the principal that 0^ and
other atmospheric oxidants in an air sample react with iodide ions in
the sensing reagent to produce iodine which forms the tri-iodide ion
(I,). The tri-iodide ion in turn is electrically converted back into
iodide ions through cathodic reduction using either an electrolytic
(Brewer or Schulze) cell or a galvanic (Hersch) cell.
3.6.1.3.3 Electrolytic Analyzers--For about fifteen years Mast Devel-
opment Company has sold an 0, analyzer with Brewer-type cell. It is
portable, can be operated off batteries or line voltage and costs about
$1,000. Hundreds have been sold.
In the Mast 0, analyzer an iodine-containing electrolyte is used.
A potential (approximately 0.25 V) is applied between electrodes. The
initial current polarizes the cathode with a layer of hydrogen (l-
95
-------
which prevents further current flow. Tri-iodide ion produced by the
reaction between oxidant or ozone and KI immediately reacts with the H~
to depolarize the cathode. Consumption of hydrogen causes flow of
polarization current to replace the reacted hydrogen and reestablish
polarization.
The polarization current is proportional to the tri-iodide ion con-
centration according to the configuration of the electrodes. The reac-
tion at the electrodes gives an electron yield less than 1001 of
theoretical and must therefore be calibrated with a known ozone source.
This technique is now used in the Intertech, Mast and Welsbach ampero-
metric analyzers.
3.6.1.3.4 Chemiluminescent Analyzers--In the early 1960rs a detector
based on measurement of Chemiluminescent emissions from the reaction
between 0, and Rhodamine-B was constructed by Regener. The concept of
chemiluminescence for 0, detection went through a number of stages as
variants were explored. Nederbragt did initial work on a gas-phase
analyzer which is the choice approach based on number of analyzers in
use (over five hundred) and manufacturer numbering about 10. It is also
the reference method designated in Federal Register, 36^ (89), April 30,
1971.
The reaction between ozone and ethylene yields Chemiluminescent
emission in the 300 to 600 nm region with maximum intensity at 435 nm,
intensity being proportional to the ozone concentration. Figure 3.10 is
a diagram of a typical gas-phase (ozone-ethylene) Chemiluminescent ozone
analyzer. The detector resonds linearly to ozone concentrations from
0.002 to above 5 ppm; there are no apparent interferences. Small
96
-------
improvements in sensitivity result from cooling the phototubes used to
detect emitted light or by increasing the ethylene flow rate.
Using pressurized cylinders of ethylene is prohibited in many
locations by OSHA regulations and local fire codes. Using the cylin-
der may require construction of an external housing that is well-
ventilated, adequately secured and protected from flame or excessive
heat, with flow-limiting orifices at the cylinder outlet. Leak-free
corrections are essential. An alternative is a blend of ethylene and
inert gas that is flammable but not explosive. Requirements for stor-
age and use of this gas are less stringent, and satisfactory analyzers
are available.
C2H4
QUARTZ.
WINDOW
SAMPLE
INLET
^°
;v'*
•ft
CONTROLLER
\rU,
CTION '
MBER J
/
', PM
— runcn
i— SUPPLY
N
AMP
METER
RECORDER
THERMO-ELEC.
COOLER
Figure 3.10. Automated gas phase chemiluminescent ozone monitor
Chemiluminescent methods for monitoring ozone other than the 0.,:CJ-L
reaction are now available commercially. Philip Eindhoven of the Nether-
lands has developed an ozone monitor based on the classical Rhodamine-B
ozone reaction. His system is automatically calibrated one an hour due
97
-------
to the change in response characteristic of the Rhodamine disc used.
This monitor is being sold in Europe but not yet in the U.S.
3.6.1.3.5 Ultraviolet Absorption Ozone Analyzer--Dasibi, Inc. sells an
0, analyzer that continuously monitors CL in the air with parts per mil-
lion readable directly from the front panel. An analog output is avail-
able for continuous strip-chart recording, as well as a binary-coded-
decimal (BCD) output for direct interfacing with a computer or a printer.
It is sensitive to 0.01 ppm 0, and reads to 9.99 ppm. A schematic of the
instrument is shown in Figure 3.11. This unit employs modern electronic
circuitry, including a stabilized mercury W light source, dual detec-
tors, digital electronics, and signal integration. The sampled air is
pumped into the instrument at a flow rate of approximately 1 liter/
minute. The sample stream alternately bypasses and then passes through
the catalytic converter, which removes (X selectively. This air sample,
stripped of CL, enters the absorption cell in which the photodetector
O
measures the amount of UV light transmitted through the cell, measured
at 253.7 nm. The result is stored as the reference measurement and
includes effects due to other pollutants present in the sample stream.
The next cycle brings the air sample directly into the absorption cell.
The detector now measures attenuated ultraviolet light due to 0, in the
sample stream and a subtraction is made from the stored reference mea-
surement. The difference is ozone concentration.
The ultraviolet absorption technique has the advantage over chemi-
luminescence that it requires no reagent; it has the disadvantage that
very high concentrations of certain hydrocarbons or mercury vapor may
be interferences.
98
-------
3.6.2 Stationary Sources
Since publication of the "Air Quality Criteria for Particulate
Matter" in January, 1969, EPA has published (December 23, 1971,
Federal Register) methods for finding the rate of particulate emission
from certain stationary sources. Methods one through four are for
obtaining the number and location of sampling points, stack velocity,
stack gas molecular weight and stack gas moisture. Method five is for
obtaining the mass concentration of particulate emissions. Using these
five methods, the emission rate of the source being tested can be
determined.
These methods were prescribed for compliance testing of sources
covered under the Standards of Performance for New Stationary Sources
as described in the Clean Air Act. The first sources covered were
steam generators, Portland Cement plants and incinerators. Standards
for particulate emissions from a second group of six industries were
promulgated in the FR March 3, 1974.
In addition to use in measuring particulate emissions from new
sources, Method five has been widely used on existing sources and is
the method used by many state and local agencies. When the sampling
conditions are met as described in the Federal Register, it is appli-
cable to most types of stationary sources, although special problems
will require adaptation or use of an alternate procedure.
Evidence shows excellent analyzer repeatability. Some EPA
scientists suggest that an accurately calibrated analyzer can be taken
into the field with an ozone; generator and used as a secondary standard
to obtain a multipoint (typically seven) calibration curves for field
99
-------
analyzers. Returned to the laboratory, the Dasibi analyzer calibration
could be reestablished. It is contended this would be an excellent way
to maintain the accuracy of field CL analyzers.
CAl SAMPU INlfT
Figure 3.11. Ultraviolet absorption ozone analyzer
EPA is working on UV spectrophotometric methods for absolute
determination of ozone generator output and preliminary results sug-
gest it may be possible to make such determinations in the 50 ppb ± 10%
range with relatively inexpensive instruments.
Several other methods are being or have been studied as alternate
methods for stationary source particulate emission measurement. In-
stack filter methods are promising. The current ASTM method uses an
12
in-stack Alundum thimble to collect the particulate matter. As with
most methods, in-stack methods are not suited for all sources.
Several instrumental methods are available also, with instruments
based on beta radiation attenuation (beta gauges) now being produced
commercially. While the beta gauge can yield accurate measurements,
100
-------
its use, like other extractive methods, depends on obtaining a repre-
sentative sample in the measuring zone. This is a difficult problem.
To overcome errors inherent in the use of current sampling equip-
ment, improved in-stack devices and non-extractive optical devices are
being developed. Reviewing the state of the art of instrumental methods
for particulate emissions for combustion sources has been done on an EPA
contract.
Besides determining total mass, measure of particle size distribu-
tion is of major importance for future regulations on fine particle
emission. Most size measurements are now made by inertial sizing
devices. An EPA study of most of the commercially available devices
was made by contract. Other approaches to size distribution measure-
ment including optical methods are being studied. Volume III of
reference four describes many of these methods, as does the NERC/RTP
briefing report on fine suspended particulate matter.
Method eight published In the December 23, 1971, FR measures SO^
and H-SO* from a sulfuric acid plant. Except for this method and
methods published for hazardous air pollutants mercury and beryllium,
no methods to determine composition of particulates emitted from sta-
tionary sources have been prescribed. For data on chemical composition,
particulate matter collected by either Method five or other methods is
generally analyzed by standard methods, as for sulfates and nitrates.
No new approaches for these two not included in the NERC/RTP task force
., ,, 16,17
report are available.
Presently, x-ray fluorescence (XRF) spectroscopy is the most
attractive method for routine multielemental analysis of particulate
101
-------
matter, as shown in a study carried out by EPA interagency agreement
18
with the Naval Research Laboratory. A first phase report described
the laboratory evaluation of XRF for elemental analysis of filter-
19
deposited particulate material. A later report demonstrated the
usefulness of simultaneous multispectrometer instruments available from
x-ray equipment manufacturers, for large-scale routine analysis. More
recently a small, low-powered portable XRF analyzer was designed and a
prototype instrument built to show applicability of the x-ray technique
to on-site use.
3.7 SUSPENDED PARTICULATE
Current research is advancing our understanding of the sources,
properties, chemical composition and dynamics of the atmospheric aero-
sol or total suspended particulate matter. Progress has been made in
establishing multimodal distribution of the atmospheric aerosol, the
typical observation of a biomodal mass-size distribution and differences
in chemical composition, identification of organic material as a major
primary and secondary component of fine particulate matter, and devel-
opment of improved instruments and techniques for measuring aerosol
parameters.
3.7.1 Calibration
For suspended particulate, the concept of calibration needs special
attention. Methods in air pollution are generally designed to be spe-
cific for an element, compounds or class of compounds. All substances
in the air above a certain diameter may be considered, for some purposes,
as suspended particulate. There are many devices for collecting air-
borne particulate. The mass, density, surface area, electrical charge
102
-------
and shape of a particulate are properties that may determine sample
design characteristics and collection efficiency. Operating princi-
ples of remote sensing devices may depend on color, refractive index,
opacity, light scattering, or even attenuation of ultrasonic fre-
quencies. Calibration techniques may vary with the type of instru-
mentation, but will usually involve the generation of aerosols or
participates of known physical and chemical properties. For example,
polystyrene latex spheres of known diameters are available and useful
for calibration of particle size analyzers.
3.7.2 Ambient Air
3.7.2.1 Impactor Measurements of Size Distribution--EPA has completed
tests on impactors showing that apparent size distribution can be
seriously affected by the type collection surface used. A Cascade
impactor with silicone grease-coated-aluminum collection surface was
compared with an identical Cascade impactor with either glass fiber or
clean aluminum collection surfaces. For the greased collection sur-
face, it was seen that apparent mass median diameter is decreased by
factors two to three, when clean aluminum and glass fiber collections
are used respectively. Serious particle-bounce errors can occur in
conventional impactors when a greased collection surface is not used.
3.7.2.2 Virtual Impactor Optimization--To eliminate particle bounce
errors associated with conventional impactors, in a virtual impactor
size fractionation is accomplished aerodynamically without the aid of
an impaction surface. In this device, particles are collected by
filtration downstream of the fractionator. The Lawrence Berkeley
Laboratory has studied loss mechanisms for the virtual impactor and
and developed methods to lower losses for particles smaller than 20 ym.
103
-------
3.7.2.3 Automated Dichotomous Samplers--Twelve automated samplers to
separately collect particles in the 0-2 and 2-20 ym size range have been
fabricated and will be operated in St. Louis beginning November 15,
1974. The devices use a virtual impactor for size fractionation. Auto-
mated sample changing allows collection of up to 35 aerosol samples for
sampling intervals adjustable from two to 48 hours.
3.7.2.4 Manual Dichotomous Samplers--A sampler for collecting parti-
cles for 24-hour intervals in the 0-3.5 and 3.5-20 inn range is being
developed for health effect measurements and routine monitoring. A
virtual impactor is used for size fractionation. A prototype is being
tested. An improved device with flow rate regulators is under develop-
ment and will be tested in 1975.
3.7.2.5 Beta Gauges--A number of beta gauge devices for measuring par-
ticulate mass collected on a filter are presently being tested. These
14
devices consist primarily of a radioactive C source and a solid state
detector. Although the beta gauge method is not necessarily more accu-
rate than a gravimetric method, it is much better adapted for low-cost
automated and semicontinuous measurements. Documentation of field
accuracy will become available in 1975.
3.7.2.6 X-Ray Fluorescence Analysis--X-ray fluorescence spectrometers
are used routinely to determine concentrations of elements with atomic
numbers above 19. Procedures for obtaining quantitative analysis for
elements with atomic numbers between 13 and 19 are being developed.
Good sensitivity for sulfur is obtained using both energy dispersive
and wavelength dispersive techniques. More than ten laboratories were
actively using this technique during 1974.
104
-------
3.7.2.7 Measurement of Visibility Reduction due to Aerosols- -Recent
studies show good agreement between human observer and time averaged
integrating nephelometer measurements of visibility. However, gravi-
metric measurement of total suspended and fine particulates we find
to be only a fair measure of visibility.
3.8 SULFUR OXIDES
3.8.1 Ambient Air
3.8.1.1 Sulfur Dioxide--A new technique to measure SCL has been devel-
oped by the National Bureau of Standards based on measuring the
fluorescence of SCL induced by exposing SCL to ultraviolet monochromatic
radiation. The fluorescence emission is a linear function of S0? con-
LJ
centration. Systems to measure both ambient and source concentrations
of SCL using this principle are now commercially available. This system
is virtually flow insensitive, a major advantage over other S0? monitor-
LJ
ing principles.
Under EPA sponsorship an ultra-sensitive flame photometric detector
has been developed to measure 0.1 ppb of SCL. The system uses a pair of
rare earth optical filters which measures the ratio of emission of the
SCL luminescence to background flame signal. This ratio measurement
improves sensitivity and specificity of the flame photometric detector
for SCL. The detector has applications in areas where background con-
centrations of SCL are needed.
3.8.1.2 Sulfates- -Various classical colorimetric and turbidimetric
methods of measuring sulfates are being compared to measurements of
total sulfur by x-ray fluorescence (XRF). Each study is intended to
establish the limitations of classical methods for sulfates and to
105
-------
yield data on the applicability of XRF methods for sulfate analysis of
particles collected from the atmosphere.
3.8.1.3 Sulfuric Acid--An automated instrument has recently been devel-
oped which collects sulfuric acid from ambient air on an inert teflon
surface. Sulfuric acid is separated from other particulates by volati-
lizing it with warm, dry air. The sulfuric acid volatilized is deter-
mined with a flame photometric detector sensitive to sulfur-containing
compounds. The system is capable of measuring 0.1 ugm/m of H~SO,. EPA
is working to characterize and improve the instrument's performance for
use as a routine field monitor.
3.8.2 Stationary Sources
The SO- compliance test method (Method six, Federal Register,
December 23, 1971) involves gas sample extraction from a source and
measurement by the barium thorin titration procedure.
In addition to compliance testing for SCL the December 23, 1971,
FR requires continuous monitoring for the pollutant gas. Proposed
emission monitoring and performance testing requirements for SO- (NO ,
i* jC
CO and H-S) continuous monitoring systems were published in the Septem-
ber 11, 1974, Federal Register. The approach taken by EPA requires
each source owner or operator to demonstrate the capability of continu-
ous monitoring systems to meet the proposed performance specifications
for his particular source.
Research since 1972 into new analytical techniques for S0? source
measurement has been active. The UV fluorescence technique for source
and ambient SO-? levels has been demonstrated and is now available in a
commercial instrument system.
106
-------
The feasibility of using a PbSnTe-diode laser system for SO- source
measurements has also been shown. A present liability of the technique
for field use is the necessity for liquid helium temperature operation.
o
Under an EPA contract Burch and Gryvnak have applied gas filter
correlation techniques for in-stack measurement of S0_, NO, HF and
HC1. Field evaluation is being done by EPA personnel.
The Barringer correlation spectrometer system for the remote detec-
tion of S0~ was evaluated and appears to have serious drawbacks and
seems unusable for enforcement or research.
The general technique of gas filter correlation appears quite
promising for the remote sensing of gaseous pollutants. An EPA con-
tract has developed a remote sensor for S0~. The sensor is relatively
small (25 Ibs., 1 cu. ft.), simple to operate and field usable with
battery operation. It has undergone field testing at a California oil-
fired power plant and a North Carolina coal-fired power plant. Com-
parative SO- measurements for Method six, a continuous in-stack monitor
and the gfc remote sensor gave relatively good (±15%) agreement among
the three techniques.
3.8.3 Mobile Sources
Sulfur in fuel is normally determined by one of several methods
including x-ray (ASTM P-2622), lamp (ASTM P-1266) , and bomb (ASTM P-129).
MLcro coulimetric and x-ray fluorescence methods have also been employed.
The latter technique yields quick qualitative and quantitative informa-
tion, but is difficult to use with leaded fuels because of peak overlaps
of lead and sulfur. High resolution x-ray fluorescence instrument appears
to have overcome this difficulty.
107
-------
As yet, there is no standard procedure for sampling and analyzing
sulfur oxide emissions from mobile sources. Most earlier data on sul-
fur oxide emissions were obtained using modifications of the Federal
testing procedure for sulfuric acid mist and sulfur dioxide emissions
from stationary sources. In most cases raw exhaust is sampled by
inserting a sample probe anywhere between the exhaust manifold and the
tail pipe. The exhaust sample is drawn from the probe through a heated
sample line at a constant flow rate into a series of impingers contain-
ing 801 isopropanol solution and 31 solution of hydrogen peroxide to
selectively collect SO, and SCL respectively. The S02 is oxidized to
sulfate and is retained as such in the hydrogen peroxide solution. The
sulfate content of the impingers is determined for SO., and S02 by barium
perchlorate or barium chloride titration using thorin as indiator;
variation involves insertion before the impingers of a controlled con-
densation coil with a fritted filter at the end. By controlling the
condenser temperature between 60 to 90°C, sulfuric acid can. be removed
by condensation and filtration.
Another sampling technique which has gained popularity lately is
the dilution tunnel method. Raw exhaust from the tail pipe is fed into
a dilution tunnel where it is mixed with a stream of clean air and drawn
through the tunnel by a pump at the opposite end. A nominal ten-to-one
dilution is normally attained. The temperature of the exhaust drops
during expansion and dilution with clean air in the tunnel, resulting
in formation of sulfuric acid mist. Sample probes protrude into the
air stream at the far end of the tunnel and sample the air-exhaust
mixture in the tube proportionally and isokinetically. Particulates,
108
-------
including H-SO. aerosols, are trapped on the filter in the sample probe.
h T-
SO- is collected by absorption in dilute aqueous sodium tetrachloro-
mercurate and determined by the standard West-Gaeke pararosanilive
method.
In recent months, instrumentation for in-Line monitoring of SCL has
become commercially available. The principle is based on fluorescence
of SO.-, when pulsed irradiated with light of a certain wavelength.
Original application is in-stack monitoring of S0? where the concentra-
tion is in the hundreds ppm range, but adaptation to SO,,-monitoring in
mobile exhaust is a promising possibility.
The analysis of SO- by oxidation to sulfate and titration with
barium ion in the presence of a suitable color indicator or by the West-
Gaeke pararosanilive technique is straightforward. Sampling and collec-
tion techniques appear to be a major source of problems in the overall
analytical scheme. Concentration of S02 in auto exhaust is high enough
that titratable quantities can be collected within a reasonable time
frame.
Analysis of SO, on soluble sulfates in automobile exhaust presents
a much harder problem, especially for cars not equipped with catalyst.
The amount of soluble sulfate varies from almost nothing to a maximum
of about 2% of the total sulfur burned. In a simulated urban driving
cycle, the amount of sulfate particulate sample collected on a 47 mm
filter by sampling tunnel diluted exhaust at flow rate of 1 cfm may
total no more than 10 micrograms as SO.. This amount is much lower than
the sulfate background of some filter collection media, especially glass
fiber filter. Compounding the problem is the fact that this amount is
109
-------
usually beyond the detection limit of most existing analytical methods
of sulfate.
The choice of filter media for particulate sulfate is quite criti-
cal at low concentrations. At high concentrations, the media background
may be reduced to negligible levels by proper dilution. Even supposedly
low sulfate background glass fiber filters were found to contain greater
than 30 micrograms per 47 mm diameter filter.
Fluoropore, a Teflon filter with polyethylene backing, had almost
zero sulfate background. It is available in large enough nominal pore
size to permit sampling at a reasonable flow rate without excessive
pressure drop and is inert to most solvents. One disadvantage, however,
is that only solvents with low surface tension, such as isopropanol or
methyl alcohol, or mixtures of these with water, can be used to extract
the samples from the filter. The filter also curls up after removal
from the holder, thus presenting problems for mounting the filters on
cassettes to do x-ray fluorescence analysis.
Nucleopore, a polycarbonate filter, is another good candidate for
filter medium. It has low sulfate background and remains relatively
smooth after removal from the filter holder. The only drawback for
mobile source sampling is an excessively high pressure drop during
sampling.
Most analytical methods for soluble sulfates involve so much sam-
ple handling--extraction, rinsing, evaporation and concentration, pH
adjustment--that the possibility of sample losses during a handling
process is increased. X-ray fluorescence would obviate most of the
handling steps, but has a rather stringent requirement for filter
110
-------
samples. In most cases only a certain area of the filter is examined
and the deposits are assumed to be uniform. Filters must be reasonably
flat and free from wrinkles. Preliminary analysis of K?SO. aerosols
generated in a bag and collected both on fluoropore and Nucleopore fil-
ters show better correlation between x-ray and gravimetric measurements
with Nucleopore filters than with Fluoroposre filters. At high sulfate
loadings, x-ray fluorescence gave lower values than gravimetric
measurements.
The collection of S0_ in exhaust via absorption in 80% isopropanol
is reported to result in high sulfate values. Apparently, some S0^> is
retained, partly oxidized to SO., or the resulting sulfite responds
quantitatively to barium ion tritration. This problem becomes acute
when baseline data are being generated with low sulfur fuel and non-
catalyst cars.
The Mobile Source Measurements Research Section has developed an
automated version of the barium chloranilate method for sulfates appli-
cable to soluble particulate sulfate from cars run on non-leaded fuels.
It is based on the reaction of sulfate ions with the solid barium salt
of chloranilic acid. The reaction precipitates out BaSO. and releases
highly UV nm and related to sulfate concentration. The sensitivity of
the method is greatly enheinced by conducting the reaction in a medium
less polar than water, such as ethanol-water or isopropanol-water mix-
tures, where the solubilities of both BaSO. and barium chloranilate are
reduced. Sensitivity of the method is better than 0.5 yg SO^/ml of the
extracting solution.
Exhaust samples diluted in a tunnel are collected on Fluoropore
filters, extracted with 60/40 isopropanol/water mixture and injected
111
-------
into a liquid chromatographic system. The sulfate sample is fed through
a strong cation exchange resin in hydrogen form to remove cationic
interference and then through a column of barium chloranilate where
color reaction takes place. BaSO, is retained in the column and the
colored species is measured with a UV-monitor at 310 nm. Through-put
time from initial sample injection is less than four minutes and sensi-
tivity is better than 0.5 yg SO, per ml of extracting solution. Preci-
sion obtained in early work was about ±.05 yg SCK~/ml. There are ionic
interferences from such anions as cl , Br, F and PO.H. For exhause
samples from cars run on non-leaded fuels, ionic interference is minimal.
3.9 ASBESTOS
3.9.1 Stationary Sources
The present NESHAP regulation for asbestos, based on prescribed
control practices rather than maximum permissible emission levels, is
partly the result of inadequate methods for measuring airborne asbestos
concentrations. It is generally agreed that more satifactory asbestos
measurement methods than now exist are need both for (a) rapid routine
screening purposes (to detect sources, measure ambient levels or imple-
ment an emissions standard), and for (b) research applications (e.g.
in support of health effects studies) where maximal information on
fiber count and size distribution are required. A new x-ray diffrac-
tion technique is being developed as a rap d screening procedure though
an interagency agreement with the Naval Research Laboratory. For
research applications, however, and possibly also as a reference pro-
cedure, it is likely that one or more of the current electron micro-
scope (EM) techniques will be used for an indefinite future period.
112
-------
Problems associated with EM methods are well known. The proced-
ures are tedious, expensive (about $250 per sample) and time-consuming
(about one man-day or more per sample). Each of the steps in an EM
assay, e.g. sample acquisition and treatment, specimen preparation,
morphological and confirmatory identification, fiber counting and siz-
ing, etc., can be the source of error. A chaotic situation exists
because there are probably a dozen EM procedures in use that differ
with respect to one or more: of the steps cited above and there has been
no attempt to establish the accuracy or precision of the various methods.
There is an urgent need to establish a single EM procedure as acceptable
standard analytical technique so that valid comparisons can be made of
data obtained by different laboratories.
The feasibility of a new x-ray diffraction (XKD) technique as a
rapid screening method has been shown. The approach consists of
measurements in a specially-designed broad-beam diffractometer after
samples have been pretreated to align asbestos fibers. Measurements
with standard chrysotile samples indicate a detection limit of about
0.3 yg for a 500-second measurement time. During FY 1975 a prototype
XRD asbestos analyzer based on this principle will be built for
delivery to ERC/RTP and testing with field samples by the Chemistry
and Physics Laboratory.
3.10 BERYLLIUM
Methods 103 and 104 for measuring beryllium emissions from sta-
20
tionary sources are described in the Federal Register. These methods
are applicable to all tests for beryllium required under Part 61,
National Emission Standards for Hazardous Air Pollutants. Presently,
113
-------
Part 61, Subparts C and D, requires control of beryllium emissions from
the following beryllium-handling activities: extract and ceramic plants,
foundries, incinerators, propellant plants, machine shops and rocket
21
motor test sites. A background information document issued with the
regulations describes the development of standards. Test methods require
a representative portion of stack gases to be removed from the stack and
filtered, with filtered material analyzed for beryllium content.
Developing instrumentation to continuously monitor particulate mat-
ter, including beryllium, was conducted under contract. The principal
objectives applied to beryllium emissions were two: (1) to prepare a
document on the status of the problem of continuous monitoring and (2)
to produce a prototype monitoring system.
The study identified various methods available, but concluded that
no one method is applicable to the concentrations and types of beryllium
22
particulates existing in stack gases today. The work of Webb, et al.,
suggests a continuous monitor based on emission spectroscopic measure-
ments of a continu us flowing (40 1/min.) sample stream of stack gases.
Such an emission spectroscopic measurement can provide a means of moni-
toring beryllium or most other metallic elements and some non-metals
simultaneously. The technique has the necessary (a) selectivity, (b)
speed (30 sec.) for rapid repeated analyses, and (c) sensitivity
3 9 12
( 1 yg/m ). A method based on the Be(a,n,y) C reaction can be used
for measuring beryllium continuously in atmospheres or stack gas emis-
sions containing principally beryllium or its compounds down to 1 yg/m ,
but cannot be used where other particulate matter is present in large
amounts. A prototype proved not feasible at this time due interfering
particulate matter.
114
-------
3.11 MERCURY
The National Emission Standard for mercury promulgated in April,
1973, is applicable to stationary sources processing mercury ore and
those using mercury chlor-alkali cells to produce chlorine and alkali
metal hydroxide. Reasons for selecting these soures for regulation
and excluding other mercury sources are stated in the background docu-
21
ment issued with the first group of NESHAPS standards. The standards,
stack sampling procedures and reference measurement methods are pre-
20
scribed in the Federal Register. The reference methods are manual
procedures based on isokinetic collection of particulate and gaseous
emissions in acidic iodine monochloride solutions. Method 101 is for
sampling air streams and Method 102, for hydrogen streams encountered in
chlor-alkali plants.
Caution must be exercised in applying this method to other than
regulated sources. For example, if the sampling stream contains high
sulfur dioxide concentr tions, such as are present in fossil fuel power
plants and non-ferrous smelters, the efficiency of collection is ques-
tionable du to reaction of the SO- with IC1 Impinger solution. Sulfur
dioxide scrubbers such as sodium carbonate solutions have been placed
in the front of the sampling train and are reported to overcome this
,, 23,24
problem. '
115
-------
Optical instrumentation has been the most universal approach in
mercury analysis and monitoring. Optical techniques utilize the strong
ultraviolet absorbance of mercury vapor at 253.7 nra. Since mercury is
an atomic species, it absorbs and emits energy at the same frequency.
The extinction coefficient for mercury is approximately 4x10 .0. While
the technique is not specific for mercury, its strong absorpiton allows
mercury to be determined in the presence of 100-1000 times the concen-
tration of weakly-absorbing species. (This is not enough for non-ferrous
smelters and power plants which emit mercury in low concentrations in the
presence of high levels of sulfur dioxide.)
UV analyzers are available commercially in single beam, dual wave-
length and dual beam configurations. To increase the rejection ratio
for interfer nces and make the instruments more specific for mercury,
one approach uses Zeeman splitting of the mercury resonance line while
another makes uses pressure broadening of the 253.7 rnn mercury line. A
third approach to improve specificity involves isolation of mercury by
absorption on a noble metal (silver or gold) collector and subsequent
therman desorption and analysis.
The laboratory phase of the EPA-funded study included development
of a mercury generation system, measurement of interferences and evalua-
tion of pyrolyzers to decompose hydrocarbons and convert mercuric com-
pounds to elemental mercury and evaluation of scrubbers to remove sulfur
dioxide. These data were obtained for each instrument: accuracy, pre-
cision, sensitivity, stability, interferences, reliability and sample
treatment required. The laboratory test program was followed by field
tests at three mercury sources: (1) secondary processing of mercury;
(2) chlor-alkali production; and (3) non-ferrous (zinc) smelting.
116
-------
Precision and accuracy of monitoring instruments is greatly
affected by the sampling interface system and its ability to trans-
port a representative sample. Mercury concentrations sampled in the
field ranged as high as 40 mg per cubic meter, requiring a dilution
system capable of controlling mercury concentration in the range of
the instruments (normally )-1.0 mg per cubic meter). The system used
in the field performed less than adequately to satisfy these require-
ments, but sample treatment for interfering species and conversion of
participate and organomercury compounds to elemental mercury was
accomplished successfully. Laboratory results indicated that a sodium
carbonate scrubber absorbed sulfur dioxide without removing mercury,
All species of mercury tested were converted to elemental mercury
by pyrolysis in a catalytic converter. Hydrogen stream test results
showed that mercury species, other than elemental mercury were present
so that total mercury monitoring of chlor-alkali hydrogen lines would
require some means of gas stream pyrolysis to convert particulate
mercury or organomercury species to elemental form.
It was found that available mercury-measuring instrumentation can
be adapted to measure total, mercury emissions from certain stationary
sources, particularly chlor-alkali plants. The transporting and con-
ditioning of the sample poses difficulties requiring more research.
The need for a dynamic dilution system to condition high-level mercury
emissions sets the requirement for a fairly sophisticated automatic
interfacing subsystem.
3.12 METHODS STANDARDIZATION AND PERFORMANCE EVALUATION
117
-------
3.12.1 General Discussion
EPA is working to standardize reference methods for industrial air
pollutants in ambient air and stationary and mobile sources. The pro-
gram involves in-house methods evaluation and ruggedness testing, method
inter-relativity studies and collaborative testing. All criteria pol-
lutants are under study, and some results have been published for all
six criteria pollutants.
3.12.2 Particulates
After reference method was tested by 12 laboratories, statistical
25
analysis of the resulting data showed:
1) Relative standard deviation for single analyst variation is
3.0%.
2) Relative standard deviation for multilaboratory variation is
3.71.
3) Minimum detectable amount of particulate matter is 3 mg,
equivalent to 1-2 yg/m for a 24-hour sample.
3.12.3 Sulfur Dioxide
The 30-minute sampling option prescribed in the reference method
was collaboratively tested by 14 laboratories and a statistical analysis
o/r
of that data showed:
1) Standard deviation for within-laboratory variation varies
3 3
linearly with concentration from 15 yg/m at zero to 36 yg/m at
1000 yg/m3.
2) Standard deviation for between-laboratory variation varies
linearly with concentration from 29 yg/m" at zero to 70 yg/m at
1000 yg/m3.
118
-------
3) No systematic error, bias or inaccuracy was detected and the
lower limit of detection is 25 yg/m (951 confidence level).
The 24-hour sampling option prescribed in the reference method was
collaboratively tested by four laboratories and a statistical analysis
of the data shows:
1) Standard deviation for within-laboratory variation varies
linearly with concentration from 18 yg/m at 100 yg/m to 51 yg/m at
400 yg/m3.
2) Standard deviation for between-laboratory variation varies
333
linearly with concentration from 37 yg/m at 100 yg/m to 104 yg/m at
400 yg/m3.
3) The 24-hour sampling method apparently has a concentration-
dependent bias which becomes significant at the 95% confidence level
at about 400 yg/m . Observed values tend to be lower than expected
SCL concentration levels.
4) Comparison of results from 30-minute and 24-hour sampling
options shows the 24-hour procedure to be more precise than the 30-
minute procedure. However,, these differences are based on tests of
differing experimental design. Although accepted statistical tech-
niques were used to process the data, the exact degree of improved
precision is uncertain.
3.12.4 Carbon Monoxide
The reference method was collaboratively tested by 16 labora-
28
tories and statistical analysis of the data showed:
1) Standard deviation for within-laboratory variation is
0.57 mg/m3 (1 ppm CO equals 1.15 mg/m CO).
119
-------
2) Standard deviation for between laboratory variation varies non-
linearly with concentration with a minimum of 0.85 mg/m at 20 mg/m
and ranges as high as 1.54 mg/m in the concentration range of zero to
60 mg/m .
3) The minimum detectable limit is 0.3 mg/m .
3.12.5 Nitrogen Dioxide
Evaluation shows the sodium arsenite manual procedure for measure-
ment of ambient NCL has a high, constant collection efficiency of 82%
3 29
over its range of 20 to 750 yg/m NO,,. The method is insensitive to
normal variations in orifice bubbler diameter, temperature of absorber
solution during sampling and concentration of sodium arsenite. No and
CCL were confirmed as positive and negative interferents, respectively.
Evaluation shows the TGS-ANSA manual procedure for ambient NCL has
a high, constant collection efficiency of 931 over its range of 20 to
700 Mg/m NCL. The method is insensitive to normal variations in time
between sample collection and analysis, storage conditions after sample
collection, flow rate, CCL concentration, sampling time and orifice
size. The time interval between addition of the diazonium salt forming
reagent and the diazonium salt coupling reagent is critical and must be
carefully controlled.
Evaluation shows the Triethanolamine (TEA) manual procedure for
ambient NCL has a constant collection efficiency over its range of 20
to 700 yg/m , both with and without added n-butanol using either frits
or restricted-orifice bubblers. The collection efficiency using frits
is about 80%, while the efficiency drops to about 50% using restricted-
bubblers. More work is not contemplated because other methods show more
promise for measurement of ambient NCL.
120
-------
3.12.6 Stationary Sources
EPA test methods promulgated for use in the determination of emis-
sion levels specified in the New Source Performance Standards were
collaboratively tested by four teams simultaneously sampling emissions
from a fossil fuel-fired steam generator. Tested were Method seven
(Oxides of Nitrogen), Method six (Sulfur Dioxide) and Method five
(Particulates). Conclusions derived from the data are:
1) Method seven (NO ). Estimated within-laboratory standard
J\.
deviation is 6.6% of the mean of values measured and estimated between-
laboratory standard deviation is 9.51 of the mean of values measured.
Estimated minimum detectable limit is 5.33 x 10 Ib/scf.
_y
2) Method six (S07). The method is accurate below 300 x 10
Ib/scf, but shows significant negative bias in the range 300-500 x
10 Ib/scf. Estimated within-laboratory standard deviation is 4.0%
of the mean of the values measured and estimated between-laboratory
standard deviation is 5.8% of the mean of values measured. Estimated
minimum detectable limit is 3.16 x 10 Ib/scf.
3) Method five (Particulates). Estimated within-laboratory stan-
dard deviation is 31.1% of the mean of values measured and estimated
between-laboratory standard deviation is 36.7% of the mean of values
measured.
EPA test Method five (Particulates) was collaboratively tested at
a municipal incinerator by having four teams simultaneously sample
particulate emissions. Estimated within-laboratory standard devia-
tion is 25.3% of the mean of the values measured and estimated between
laboratory standard deviation is 38.7% of the mean of measured values.
121
-------
34
In tests at a nitric acid plant estimated within-laboratory
standard deviation is 14.91 of the mean of measured values and esti-
mated between-laboratory standard deviation is 18.5% of the mean of
measured values.
122
-------
REFERENCES
1. Standards of Performance for New Sources, Environmental Protection
Agency, Federal Register, 3(5 (247) 24876-24895, December 23, 1975;
38 (111) 15406-15415, June 11, 1974.
2. Requirements for Preparation, Adoption, and Submittal of Implemen-
tation Plans, Environmental Protection Agency, Federal Register, 36_
(158) 15486-15506, August 14, 1971.
3. Emission Monitoring Requirements and Performance Testing Methods for
Stationary Sources, Environmental Protection Agency, Federal
Register, 39_ (177) 32852-32874, September 11, 1974.
4. Nader, J. S., F. Jaye, W. Conner. Performance Specifications for
Stationary-Source Monitoring Systems for Gases and Visible Emissions.
Environmental Protection Agency, EPA-650/2-72-013, January, 1974.
5. Snyder, A. D., et al. Instrumentation for the Determination of
Nitrogen Oxides Content of Stationary Source Emissions. Vol. I and
Vol. II. Monsanto Contract No. EHSD 71-30. NTIS PB 204-877 and
NTIS PB 209-190. January, 1972.
6. Dee, L. A., et al. An Improved Manual Method for NO Emission
Measurement. Environmental Protection Agency, EPA-Rz-72-067,
October, 1972.
7. McNulty, J. J., J. R. Ehrenfeld, R. L. Goldsmith, J. F. McCoy.
Investigation of Extractive Sampling Interface Parameters. Waiden
Research Assoc., Interim Report Contract No. 68-02-0742, Environ-
mental Protection Agency, EPA-650/2-74-089, October, 1974.
8. Burch, D. E. and D. A. Gryvnak. Infrared Gas Filter Correlation
Instrument for In-Situ Measurement of Gaseous Pollutants. Philco-
Ford Contract No. 68-02-0575. Environmental Protection Agency,
EPA-650/2-74-094, December, 1974.
9. PhiIco-Ford Corp. Contract No. 68-02-0766. Evaluation of a Remote
Passive Technique Based on Gas Filter Correlation Spectroscopy.
10. Federal Register, Vol. 36, No. 247, Thursday, December 23, 1971.
11. Federal Register, Vol. 39, No. 47, Friday, March 8, 1974.
12. ASTM Method D 2928, Annual Book of Standards, Part 23.
13. Instrumentation for Measurement of Particulate Emissions from Com-
bustion Sources. Vols. I, II, III. Reports under EPA Contract
CPA 70-23 by Thermo-Systerns, Inc. Vol. IV EPA Report 650/2-73-022,
1971.
123
-------
REFERENCES (continued)
14. Field Measurement of Particle Size Distribution with Inertial Siz-
ing Devices, EPA Report 650/2-73-035 by Southern Research Institute.
15. Problems Relating to Fine Suspended Particulates. Briefing Report,
Office of Research and Development, Environmental Protection Agency,
September, 1972.
16. Suspended Sulfates and Sulfuric Acid Aerosols. Task Force Report,
Environmental Protection Agency, Research Triangle Park, N.C.,
October, 1972.
17. Nitrates, Task Force Report, Environmental Protection Agency,
Research Triangle Park, N.C., February, 1974.
18. Development of X-Ray Fluorescence Spectroscopy for Elemental Analy-
sis of Particulate Matter in the Atmosphere and in Source Emissions,
EPA Report R2-72-063, submitted by Naval Research Laboratory,
October, 1972.
19. Development of X-Ray Fluorescence Spectroscopy in Elemental Analysis
of Particulate Matter Phase II: Evaluation of Commercial Multiple
Crystal Spectrometer Instruments, EPA Report 650/2-73-006, submitted
by Naval Research Laboratory, June, 1973.
20. Environmental Protection Agency Regulations on National Emission
Standards for Hazardous Air Pollutants. Federal Register, Friday,
April 6, 1973, Washington, D.C., Vol. 38, No. 66, Part II, pages
8845-8850.
21. Background Information on Development of National Emission Standards
for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury,
U.S. Environmental Protection Agency, Research Triangle Park, N.C.
27711, Publication No. APTD-1503.
22. Instrumentation for Monitoring Specific Particulate Substances in
Stationary Source Emissions, U.S. Environmental Protection Agency,
Office of Research and Development, Research Triangle Park, N.C.
27711, Publication No. EPA-R-2-73-252.
23. Statnick, R. M., D. K. Oestreich, and R. Steiber. Sampling and
Analysis of Mercury Vapor in Industrial Streams Containing Sulfur
Dioxide. Research Branch, Control Systems Laboratory, Environmental
Protection Agency, Research Triangle Park, N.C. (Presented at
August, 1973 ACS Meeting in Chicago, Illinois.)
24. Katzman, L, R. Lisk, and J. Ehrenfeld. Evaluation of Instrumen-
tation for Monitoring Total Mercury Emissions from Stationary
Sources, Walden Research, Contract 68-02-0590, Environmental Pro-
tection Agency Report EPA 650 12-74-039, June, 1974.
124
-------
REFERENCES (continued)
25. McKee, H. C., R. E. Childers, and 0. Saenz. Southwest Research
Institute. Collaborative Study of Reference Method for the Deter-
mination of Suspended Participates in the Atmosphere (High-Volume
Method), PB 205-891. June, 1971.
26. McKee, H. C., R. E. Childers, and 0. Saenz. Southwest Research
Institute. Collaborative Study of Reference Method for Deter-
mination of Sulfur Dioxide in the Atmosphere (Pararosaniline
Method), PB 205-893. September, 1971.
27. Camann, D. E. and H. C. McKee. Southwest Research Institute.
Collaborative Study of Reference Method for Determination of
Sulfur Dioxide in the Atmosphere (Parorosaniline Method) . EPA-
650/4/74/027. December, 1973.
28. McKee, H. C. and R. E,. Childers. Southwest Research Institute.
Collaborative Study of Reference Method for the Continuous Mea-
surement of Carbon Monoxide in the Atmosphere (Non-Dispersive
Infrared Spectrometry), PB 211-265. May, 1972.
29. Beard, M. E. and J. H. Margeson. An Evaluation of the Arsenite
Procedure for the Determination of Nitrogen Dioxide in Ambient
Air. March, 1974.
30. Fuerst, R. G. and J. H. Margeson. An Evaluation of the TGS-ANSA
Procedure for the Determination of Nitrogen Dioxide in Ambient
Air. March, 1974.
31. Ellis, E. C. and J. H. Margeson. Evaluation of Triethanolamine
Procedure for Determination of Nitrogen Dioxide in Ambient Air.
EPA-650/4-74-031. July, 1974.
32. Hamil, H. F., D. E. Camann, and R. E. Thomas. Southwest Research
Institute. The Collaborative Study of EPA Methods Five, Six, and
Seven in Fossil Fuel-Fired Steam Generators - Final Report. EPA-
650/4-74-013. May, 1974.
33. Hamil, H. F. and R. E. Thomas. Southwest Research Institute.
Collaborative Study of Method for the Determination of Particulate
Matter Emissions from Stationary Sources (Municipal Incinerators).
EPA-650/4-74-022. July, 1974.
34. Hamil, H. F. and R. E. Thomas. Southwest Research Institute.
Collaborative Study of Method for the Determination of Nitrogen
Oxide Emissions from Stationary Sources (Nitric Acid Plants).
EPA-650/4-74-028. May, 1974.
125
-------
4. ENVIRONMENTAL APPRAISAL
4.1 SOURCES
4.1.1 National Emissions Report
National Emissions Reports estimate source emissions of those air
pollutants for which national standards have been promulgated. The cri-
teria pollutants are particulates, sulfur oxides (SO ), nitrogen oxides
A.
(NO ), hydrocarbons (HC) and carbon monoxide (CO). The reports are
.A.
compiled from source information of the National Emissions Data System
(NEDS). Compilations are presented for the nation, for individual states,
for Air Quality Control Regions (AQCRs) and for interstate AQCR portions.
This first annual report represents the best available indicators of
national emissions for the calendar year 1972. The estimates reported
are based on source records in NEDS prior to receipt of the December 31,
1973, Semiannual Report. Thus, the data are generally comparable to data
collected for the development of State Implementation Plans (SIPs), but
include revisions and updates whenever available.
The NEDS point source file contains data on stationary facilities
that are individual emitters of relatively large quantities of one or
more air contaminants. A point source is often arbitrarily considered
to be a facility in which the total annual emiss on of any one criteria
pollutant equals or exceeds 100 tons per year. In general, individual
emission points are defined as those annually emitting one or more
pollutants equal to or greater than 25 tons per year. In certain instan-
ces, possible significant contributors, although emitters of lesser
annual amounts, are accorded point source designation.
126
-------
At the present time, NEDS contains data for approximately 75,000
emitting points in all states, the District of Columbia and three
territories. Considerable information is stored for each point source;
for example, data on annual operating rates or activity levels (i.e.,
amount of fuel consumed by a boiler, or quantity of product produced by
an industrial process, etc.). Information on existing air pollution
control devices and their efficiencies is also recorded. These and
similar data are subsequently used in determinations of source emission
according to procedures outlined in the following discussion of the
NEDS emission factor file.
Area sources are all other emissions sources not identified as
point sources. For example, the NEDS area source file contains data
on all mobile sources (motor vehicles, aircraft, railroads, etc.),
small stationary sources, gind miscellaneous sources such as forest
fires, agricultural burning and retail gasoline marketing operations.
Unlike point sources, however, individual source data are not stored
in NEDS for each small source. Rather, estimates of total emission
levels for specific categories of area sources are stored for
individual geographic area;;.
A summary of emissions by category is given in Tables 4.1 and
4.2. The data shown in Table 4.2 are not reflected in Table 4.1, so
amounts shown in Table 4.2 should be added to the grand total shown
in Table 4.1.
127
-------
CO
I
LO
WJ
i—i
Di
C "O
0 S
£ o
<3 §
c c
~- \
! * 'j
a. e
I If
3
=3
^ K
W5 O
S
•o
CU
Emission categories
c -o
•£ o
"I
in
. II
— X
z °
"5 x
VI
3
•0
0-
Emission categones
sss
y, ^ s
S8?i
—
2SS
oo t~- r^
VI -V t
Other
Commercial- Institutional
Diesel Fuel
Other
3 23
* r- »
I 35"
sll
s SB
* 06 ae
S 2»"
24,051,2
10,3*0.8
[ 13,670.3
9 ;s
e — oe
S 'R
? £S
£ SE£
- -
Gk\ND TOTAL
-AREA
-POINT
S— OOrJww^* rn f^ ? o — o~ — aciC^^c^S^S^ JQ ae o r^ — 5 — O'^'^Sf^acSSvPj.??'''^^ & S*»
t-^,, m "" PS?S°"'
^5"'- &r'f: ^oao.o^-'* g;5*_ "> * §1^^^^ P; « ? * S SS-^S^^^SS M "s^^.^
*n r* — £2^^*^"" ~
r^ri xsc^^—
5» p.|iSs5S§issllsl5Is5IISslilS<'~<='ls = iSse= RlIIi5il|aqli?fti5g!s=>0°s
S * 5 ~ " S ~ * ' " "
ss siisSlsSss^SsallslsSiKsSslsl?8'!^;!?!^! S53lSI5s?'S5ils5sls=5i°°
s.sss.ss.s;^ - ^2-"" ss-- - - ?*=- = - p?|s-- = '--f!- -" - =:
£ I| ' si ' ' ' i i
1. "•» -IN!' ' it I' 2 • i< g
|r I || j j | i'||_ I'm s^ ,_5 _, _ |i -|
O U tii £ cxi S oi ? W S M (— ^ O Q O QS O — S ^ ™ < > C — ^ y «
S M5 f- £ 7
3§3322!p8P3§3i«*S=S5S SS=l§=5a3Ei*loaS==2lsia£S*SS*§gsSi-E= =5ssigg5=5
S'S°5"SS*S;>"SSRgPSS'cS S|~XK"'S|C5SS2!;-3S'cgSg:~R^a75§ICSS7SS|-*'c SKS'SfSSS??
^O^Sx^^5* S^^S^^"""" SS o* r^S"'*1 ^ -M "^ ~-H r^30 g^S ~ =
r*1rjr^r4 ^^~* — •*•• r4
r.SSsa-^^^-^s-S25^"^ IE "Sl^sssssa SM _ ,^5? -SK'S-S ^- ---
• : " T &
: *''.': c ' u c
,ic ^5 15 ' ^ — . "5 "^'y*_L'"3 tJ~w»p?or E 3_1.'"^ t>~°-o^ ^~c-
I ^ Iili|lil|lliiliiill I I || I | i li'iyi | I |l \ I i ' il^illilf i||
j w ™
u.
§
0)
-------
Table 4.2. EMISSION FROM MISCELLANEOUS SOURCES
Category
New York point
sources
Forest wild
fires
Agricultural
burning
Structural
fires
Coal refuse
fires
Total
L 1 . 2 Nitrogen
Particulates ,
tons/yr
343,000
413,00
300,000
57,000
110,000
1,223,000
Oxides Emissions
SO ,
tons/yr
1,028,000
0
16,000
0
141,000
1,185,000
(NOJ
NO ,
tons/yr
421,000
97,000
32,000
7,000
34,000
591,000
HC,
tons/yr
140,000
583,000
300,000
67,000
68,000
1,188,000
LI. 2.1 Stationary Sources --Stationary sources contribute at
CO,
tons/yr
49,000
3,399,000
1,600,000
220,000
340,000
5,608,000
least half
the total man-made nitrogen oxides (NO ) in the United States. Recent
X
estimates are that over eleven million tons of NO were emitted from
_A.
stationary sources in 1972. Of total stationary source emissions, about
98% result from combustion of fossil fuels. The remainder comes from
industrial processes such as the manufacture of nitric acid, sulfuric
acid and explosives. Figure 4.1, showing major stationary sources of N0
emissions, indicates that utility boilers are the major contributors
(48%). Utility boilers and stationary internal combustion engines and
gas turbines together account for approximately 70% of the total.
Industrial steam-raising boilers account for more than 18% of emissions.
129
-------
As shown in Figure 4.1, the major combustion source categories are
utility boilers, industrial boilers, stationary internal combustion
engines, gas turbines, commercial and residential heating systems,
industrial process furnaces and incinerators. For each of these cate-
gories there are sub-categories based on equipment design and these may
be further subdivided depending on type of fuel. Stationary combustion
processes are fired with a variety of fossil fuels (i.e., coal, heavy
oil, light oil and natural gas) dependent on user and equipment design.
The emission inventory being developed for the Control Systems
Laboratory (ERC/RTP) has subdivided stationary sources into 138 dif-
ferent categories of equipment design and fuel burned. Rank of sources
in general categories appears in Figure 4.1; rank of sources in more
detail is shown in Table 4.3, where those accounting for about 70% of
total emissions are listed. These leading combustion sources show the
very large number of different equipment and fuel combinations and
emphasize the complexity of developing NO control technology for such
A.
varied sources. The two leading sources in Table 4.3 are gas-fired spark
ignition stationary I. C. engines (accounting for over 16% of emissions)
and bituminous coal-burning tangentially fired utility boilers (account-
ing for about 12%).
4.1.2.2 Mobile Sources--Oxides of nitrogen form in mobile sources pri-
marily by the fixation of atmospheric nitrogen in the combination process.
Peak temperatures found are those most significant in dissociating N7
LJ
atoms. Exhaust gas recirculation (EGR) and water injection schemes are
methods to reduce the peak temperature and thus the amount of oxide of
nitrogen produced. EGR methodology was employed in 1973 vehicles. It
130
-------
CO bO
W G
S
>H
O
I—I
H
co
Pi
PH
CO
o
I—1
CO
CO
o
o
u
T) tO
0 G
G-H
3 LO
pq to
0 pq
SrH
03
G O
TO E-H
0 MH
o< o
[-H o\o
o
P r-
8*
MH o
o 5n
PH
rH P 5H
H en O
CO -H '-4H
PJ nJ
03
Csl
CD
0
rH
•8
H
p
rH
03
rH
P 03
G -P
0 H
5n
0 <-w
PH 0
to o
£ rH
O
•rH ^"^
t/)
t/1 >H
._] O
(-3 F"^
•
§ pq
£>^
O 0
bo 3
0 PH
P \
03 p
U G
0
0 S
O O.
5-1 -H
3 3,
0 cH
CO pq
oo
rH
0
rH
to
t"^
oo
rH
to
3
O
G
•H
-a 3
0 P
rH -H
•H PQ
•*
[/) (}£
03 C
CD *H
^H
-•H
ff rr .
0
•H rH
-P 03
•H -H
as
KH 0
W
,-Sr^ C«
^ TO
03 H
CO -
•s ^H
to 0
0 rH
G -H
bQpq
&fr
' -H
U rH
•H
• P
CD
CD
rH
rH
OO
OO
to
rH
rH
03
O
U
rH
to
vO
0
t^-
o
rH
03
O
u
(/)
o
•H
P
•H
PQ
*t
G/3
r-j
'rH
•H
PH
0
O
rH
U
&
to
5n
0
rH
•H
£
X
p
•H
rH
•H
P
rH
CD
,H-
OO
^O
LO
o
-
G
•H
JH
•H
f T !
rH
rH
03
!-£J
0
to
0
QH
£^
rH
rH
P
0
•H
5-i
O
to"
5-c
0
i— 1
•H
P?
£
•H tO
rH 03
•H CD
P
CD
to
to
to
CD
to
O
(fl
o3
•*
bo
p^
• tH
^H
•H
PH
rH
rH
03
P
G
O
rH
t/)
rH
0
rH
•H
O
PQ
fr
•H
rH
•H
P
*
p*
•H
5n
•H
PH
rH
rH
03
12
r-£j
0
to
p*
Ol
c^
i— 1
rH
03
4J
N
•rH
rH
O
a:
C/)
5-i
0
rH
•H
O
PQ
X
P
rH
•H
P
OO
rH
to
oo
^o
to
o
'rt
o
u
'}
C)
•rj
;3
4-J
•v
§
+->
P
O
P;
1*
*3"tO ^HT rH *3-*3- OO OtOCD
rH r — \o \o to to c — vo LO ^-
tOCNI CN! tNl r-JCM I — 1 < — IrHrH
tOvO vO CN1 rHrH VO CN1 t^ f>J
^O rH c ~ i c i £•*-, f^^ [""i CD C — • t""~-
tOtO tO IO CNltNl CO rHrHrH
C3O O O OO O OOO
bO bo 'id
G G rH 0 rH
•> -H - 'H -H bO 0 -H
6 rHgrHOOSbOO
O -H O -H ^ ft
+-> LL,4JpHrHU^rH
P P 03 03 CJ o3
0 rHOrH30H033
PQ I/) rH pq i— IT) PnTJ
I — 1 Cti TO -H T3 .H
X 03: -P ^ to0 nd to
5-i 3 ^ 05-100
CO 0 P-, 'H
•\ rH tO — t/) *> PH "
bO o3 O bO O bOrH bO
Pl3P3 H S -H PQ -H Di -H T) |g T5 CO -H t/)
P *-P P P PC G Po3
T)
rH
0
•H
PH
•\
T3
0
5H
•H
PH
rH
rH
03
r^
P
S
PH
«s
to
5-H
0
rH
•H
O
PQ
rH
03
•H
rH
P
^
T)
G
to
^f
rH
LO
rH
0
rH
•H
O
rH
-3
T3
•H
t/}
Q^
pi
•\
O
rO
3
rH
0
-P
03
T-,
0
•P
O
0
PJ
to
•^f
rH
LO
^O
rH
O
rH
rH -H
rH O
03
S rH
03
T) 3
0 TJ
tO -H
O to
PH.^
^^"
*\
X 0
rH ,0
rH 3
03 P
P 5n
CH 0
0 P
N 03
•H IS
O T3
P
- U
t/) 0
5n 5n
0 UH
rH
•H T)
O rH
pq 0
•H
rH PL,
03
•H tut
rH C
P -H
trt rH
3 -H
CH
^DOr-f
to
t--
OO
to
LO
131
-------
3.6 grains per mile
4.3 grams per mile
5.5 grams per mile
5.1 grams per mile
4.8 grams per mile
2.3 grams per mile
1.6 grams per mile
0.4 grams per mile
reduces oxides of nitrogen emissions from 40 to 60% depending on the
base year chosen.
Oxides of nitrogen emission factors for vehicles are:
Light-duty vehicles
pre 1968
1968
1969
1970
1971-72
1973-75
Diesels pre 1975
Statutory
Heavy-duty vehicles
gasoline pre 1970 9.4 grams per mile
1970-73 9.2 grams per mile
Diesels 34 grams per mile
Aircraft (for example, a
JT-9D jumbo jet)
taxi-idle 2.75 kg/hour
take-off 327 kg/hour
climb-out 208 kg/hour
approach 24.5 kg/hour
per landing-take-off cycle 14.2 killograms
The variations seen for aircraft demonstrate the complexity of
deriving meaningful factors. Fluctuations of two or three orders of
magnitude are common to all mobile source emissions depending upon
operating cycles. In light-duty vehicles these fluctuations may
occur in seconds or less; for this reason, physical integration of the
sample through the constant volume sample is used. This is much more
efficacious than mathematical integration. The latter approach must
132
-------
be used to determine nodal emissions as in design research for control
devices.
In addition to EGR for light-duty vehicles, reduction catalysts
are proposed to attain the statutory limits. These catalysts would
be placed prior to the oxidation catalysts immediately after the exhaust
manifold. The vehicle would be operated in a reducing mode, somewhat
rich. After the reducing catalyst air would be added to the exhaust
stream which would then be passed through an oxidizing catalyst. Single
bed catalysts have also been proposed. These require precise control of
air/fuel ratio in each cylinder and would probably be practical only if
automotive fuel injection were used.
Single bed catalysts may have unique properties depending on operat-
ing conditions. Usually catalyst temperature is controlled, but one
experimental catalyst which showed minimal NO emissions on the Federal
test cycle could be operated under other conditions which (a) showed
almost no NO reduction or (b) almost quantitatively reduced the emis-
.A.
sions to ammonia.
Review of the statutory limit now under way is the key to future
control efforts. The rise of NO in light-duty vehicles seen in 1968-
J\.
69 was one result of cleaner vehicle operation due to hydrocarbon and
carbon monoxide controls. A fuel penalty varies with each engine class.
4.1.2.3 Ambient Air Concentration Summary--In 1972, of 24.1 million
tons of N02 emitted to the atmosphere of the United States, 36% was from
transportation and 57% from stationary point sources.
Figures 4.2 and 4.3 present the seasonal diurnal patterns of NO and
N02, respectively. These curves reveal that NO and NO- have markedly
133
-------
different diurnal patterns; NO concentrations follow traffic patterns
while NCL does not. The quantity of NCL builds up throughout the day-
light hours, reaching a maximum in the early afternoon around 4:00 p.m.
(see Figure 4.3) and then leveling off to a minimum just before sunrise.
Seasonal curves (see Figures 4.3 and 4.4) demonstrate the lower
autumn but higher spring and summer NCL values. Figure 4.4, which is a
composite NCL pattern from all six CAMP sites, shows the more consistent
summer peaks, especially since 1969.
A long-term trend line for NCL is shown in Figure 4.5, which is
again a composite graph of the six CAMP sites. It shows a general
increase in NCL concentration levels at these six sites.
Historically, NCL measurements were made in most of the AQCR's by
now discredited FRM. According to these data, there were 41 AQCR's
which should be classified priority 1. An intense monitoring study was
conducted in these 41 AQCR's using the arsenite and chemiluminescent
methods. After two years of monitoring in these AQCR's, three cities
had yearly averages over 100 yg/m --Chicago, Springfield, Massachusetts
and New York City. In addition, Denver, Cincinnati, Philadelphia and
Salt Lake City should be carefully watched for a future problem.
As part of the NASN gas network, samples were collected once every
12 days since 1972 at approximately 200 urban and nonurban sites across
the U.S. Annual averages from this network show that in 1972 40 sites
had concentrations over 100 yg/m , and in 1973 three sites had over
100 yg/m , all located in one AQCR--Metropolitan Los Angeles Intrastate.
(Note: Continuous data were also collected in L.A. as part of the
CHESS Program. These data also indicate that L.A. is priority 1.)
134
-------
<
DC
O
u
SUMMER:
FALL:
WINTER:
0.00
12 15
TIME OF DAY
Figure 4.2. Seasonal diurnal patterns of nitric oxide concentrations.
0.07
0.06
_ 0.05
2 0.04
<
cc
0.03
o
u
0.02
0.01
0.00
6
9 12 15
TIME OF DAY
18
21
24
Figure 4.3 Seasonal diurnal patterns of nitrogen dioxide concentrations.
135
-------
250
200
~ 150
E
100
A
50
62 63 64 65 66 67 68
YEAR
70 71 72 73
Figure 4.4. Seasonal pattern of nitrogen dioxide concentration levels, composite graph of the
six CAMP sites.
62 63 64 65 66 67 68 69 70 71 72 73
Figure 4.5. Long-term trend line for nitrogen dioxide concentration levels composite graph of
the six CAMP sites.
136
-------
Year
1972
1973
Sites
129
169
<50
79
78
50-74
41
69
Table 4.4. KECLASSIFICATION STUDY - 24-HOUR DATA
Total No. of Sites within Interval-,
No. of Urbai Concentration Interval, yg/m
75-99 >99
5 4
19 3
Total No. of Sites within Interval?
No. of Urbaa Concentration Interval, yg/m
Year Sites <20 >20
1972 22 20 2
1973 29 21 8
We find that NCL is essentially an urban problem influenced by
local sources (primarily mobile in areas of high levels) and meteorologi-
cal factors, as indicated by seasonal patterns. In areas with more than
one monitoring device, data show that local maximum concentrations do
exist where one site is over the standard and others are not in excess
of the standard of 100 yg/m .
Overall, the N0? problem is less severe than we first thought. How-
L*
ever, with four areas above the standard and another four near the stan-
dard, it remains a problem.
4.1.3 Carbon Monoxide
4.1.3.1 Natural Sources--The primary source of CO in urban atmospheres
is incomplete combustion. In the global atmosphere, most carbon monoxide
is produced by natural sources from the oxidation of methane by hydroxyl
radicals. The same OH radicals account for a large amount of global CO
loss. These two competing formation and destruction processes can account
137
-------
for the 0.1 year residence time of CO estimated by radiocarbon measure-
ment. While important in explaining the global balance, oxidation of CO
by OH is insignificant in polluted atmospheres, being unable to compete
with the much faster reactions of hydrocarbons with OH.
Evidence suggests that large amounts of CO are produced through
reaction of methane in the troposphere. The oxidation of CH, by
hydroxyl radicals liberates CH, radicals that can later undergo reaction
to form CO. McConnell, et al. estimated a substantial rate of CO pro-
duction from the reaction of CH. and OH. They concluded that while
combustion sources determine local levels of CO, the natural sources
determine the global level, with the oxidation of methane playing a
leading role. From an analysis of radioactive CO data in the tropo-
2
sphere, Weinstock and Niki concluded that CO is produced at a rate
25 times greater than that of CO formation from combustion. Based on
their measurements, they calculated the rate of CO formation from the
reaction of CH. and OH. The value they derived is of the same order
of magnitude as that extimated by McConnell et al., lending support
to this view.
Levy recently developed a model showing that sufficient global
concentrations of water vapor, CH., 0_ and NO exist to cause chain
HO JC
reactions that can destroy CO as well as produce it. He calculated
that the hydroxyl radicals produced in chain processes can oxidize CO
at a rate in excess of the rate of CO production from the reaction of
CH, with OH. His model predicts that competing formation and destruc-
tion processes result in a residence time of 0.1 year for CO, agreeing
4
with the Weinstock estimate deduced from radiocarbon measurements.
138
-------
4.1.3.2 Mobile Sources--With the start of the 1975 model year, many new
cars were equipped with oxidation catalyst-converters for reducing
emission levels of both CO and hydrocarbons. Preliminary results on
prototype 1975 cars available at this laboratory indicate three to seven
times lower CO emission levels for catalyst-equipped cars over equivalent
1973 and 1975 cars and better than 10-fold improvement over 1972 cars.
The 1975 prototype vehicles evaluated are in some cases considerably
different from production vehicles actually being sold. One manufacturer
uses the catalyst with only one bank of cylinders.
Another method to reduce CO1 and other harmful exhaust gas components
is the new engine design using the stratified charge concept for combus-
tion in a two-chamber engine. Reported values for the only stratified-
charge equipped car commercially available, Nonda CVCC, show it meeting
the 1977 emission standards for CO and hydrocarbons without extra treat-
ment of exhaust gas. CO emission levels are about 2 gm/mi for all com-
mercially available engine sizes of the Honda.
One complication in achieving the ambient air quality standard for
CO by controlling auto exhaust emission is the effect that temperature
has on emission rates for both CO and hydrocarbons. This was first
reported by R. W. Hum of the U.S. Bureau of Mines , who found it to
exist in six 1973-73-74 catalyst-equipped cars. Comparing relative emis-
sions for the three regulated species at ambient temperatures from 20°F
to 110°F, it was found that CO and hydrocarbon emission increased up to
six times at the lowest temperature compared to the normal Federal Test
Procedure temperature range of 70-75°F. There was no significant change
in NO emissions, however.
.A.
139
-------
Other work has also shown higher emission levels at higher tempera-
tures, so that tests under defined ambient conditions represent a mini-
mum case. The importance of these findings is that while new model cars
may meet the emission standard as measured by the Federal Test Procedure,
actual operation of these catalyst-equipped cars at sites of different
temperature or in early morning could produce unexpected higher emission
rates for CO and hydrocarbons. We need to know more about causes and
extent of departures from standard conditions; the Federal Test Pro-
cedure should be capable of detecting emissions in all possible operating
conditions and environments.
4.1.4 Hydrocarbons
4.1.4.1 Mobile Sources--The Clean Air Act specifies a 901 rollback on
hydrocarbon emissions from 1970 levels for light-duty vehicles.
Standards are:
Pre-1968 8.6 estimated grams/mile
1973-74 3.0 grams/mile
1975 US 1.5 grams/mile
1975 Calif. 0.9 grams/mile
1976 0.41 grams/mile
Due to the complexity of hydrocarbon mixture emitted from light-
duty vehicles, control of total emissions was adequate. Minor varia-
tions in composition did not seriously effect the total. Catalytic
controls strongly enhance some materials like methane which is much
more stable and has a profound effect on relative reactivity of total
emissions. Non-catalytically-controlled vehicles produce less than
101 methane. Acetylene is also significant. In catalytic vehicles
higher molecular weight species are preferentially oxidized and methane
140
-------
may account for 30 to 40% of total emissions. In specific operating
modes, as much as 801 of hydrocarbons may be methane.
The N.A.S. report has supports current standards, but only if
other mobile sources conform to similar controls.
As is discussed in Section 3, hydrocarbon emissions vary with
cycle and operating conditions. Emission factors and regulations try
to integrate these emissions into numbers which can be related (through
number of vehicles) to impact of specific source on the environment.
A compilation of these factors by an average use cycle is given in
AP-42 compilation of Emission Feictors, Supplement 2.
Current in-house tests show prototype vehicles* emit:
Ford 0.20 + .02 gm/mi
Plymouth 0.23 ± .03 gm/mi
GM 0.44 ± .03 gm/mi
4.1.5 Sulfur Oxides
4.1.5.1 Mobile Sources--Trace quantities of sulfur mostly in the form
of organic compounds are present in gasoline (0.038%). Combustion of
gasoline in an automotive engine yields principally carbon dioxide
(CO-) and water along with hydrocarbons (HC), carbon monoxide (CO),
nitrogen oxides (NO ), sulfur dioxide (S0?) and sulfur trioxide (SO,).
During normal combustion, sulfur in gasoline is almost all converted to
S02 and SO- with the latter accounting for less than 2% of all sulfur.
The SO, reacts very rapidly with water in the exhaust and is emitted as
sulfuric acid or sulfate particulate.
*These data cover emissions from one vehicle from each manufacturer.
141
-------
The amount of SCL and SO, emissions from mobile sources nationwide
is negligible (less than II) compared with that from stationary sources.
For this reason they are not a significant portion of overall sulfur
oxides air pollution except in a few places like Los Angeles, where
stationary source emissions of sulfur oxides are relatively low.
It was recently reported and has since been confirmed that cars
with catalytic converters emit substantial amounts of sulfuric acid.
Conversion of 401 of the fuel sulfur has been reported. In the absence
of the catalytic converter, conversion is less than 2% of the fuel
sulfur.
Although the contribution of mobile sources to total regional
airborne sulfate is not significant, use of catalytic converters as
presently designed may result in localized buildup of sulfuric acid
and sulfate salts concentrations to dangerous levels, especially near
places like Southern California where the sulfur content of gasoline
is higher than the national average. Because health problems may
result from direct emissions of substantial amounts of sulfuric acid
mist by catalyst-equipped cars, EPA is reassessing the automotive
sulfur oxide emission question.
4.2 TRANSFORMATION AND TRANSPORT
4.2.1 Oxides of Nitrogen
In the laboratory, investigators have been able to account for over
90% of carbon atoms; however, only 50% of the nitrogen could be accounted
for in many cases. Apparently most of the unaccounted nitrogen atoms end
up as nitric acid either on the walls of the reactor or in the gas
phase. ' Recent investigations suggest that nitric acid is also pro-
o
duced in atmospheric photolytic processes.
142
-------
Nitric acid in the gaseous state is colorless and photochemically
stable. It is volatile acid in solution. The equilibrium vapor pres-
sure decreases with increasing dilution and at the parts-per-hundred-
million level—as might be expected in the atmosphere--the solution
would be extremely dilute. In the atmosphere, nitric acid can be pro-
duced by two reactions:
M
(1) OH + N02 y HN03
(2) N205 + H20 -> 2HN03
9
Nicolet has postulated reaction (1) for the presence of nitric acid in
the stratosphere. Gay and Bufalini have postulated reaction (2) for
the formation of nitric acid in smog chambers. It is not known which
reaction is ost important in polluted atmospheres. Good nitrogen bal-
ances have been obtained in smog chamber studies when nitric acid is
included as a product. '
Nitrous acid is not formed in as great quantities as nitric acid.
It can be produced in the atmosphere by the reaction,
NO + N00 + H-0 -*• 2HNO,,
L L i.
This acid is important because of its photochemical reactivity. Nitrous
acid absorbs sunlight and dissociates into OH and NO. Its photolysis
rate is lower than the photolysis of N02, but in photochemical smog pro-
duction it produces very reactive hydroxyl radicals when photodissociated.
Although strategy to alleviate NO -related pollution is in existence,
J\.
the transformation of NO in ambient air is still at issue. The "NO
.X 3C
issue" consists of the following partly answered or unanswered questions:
•What are the products; of atmospheric reactions of NO
.A.
and what determines accumulation levels?
143
-------
•What is the quantitative dependence of NO- on NO and
on other possible precursors?
•What is the role of NO in the atmospheric oxidant
A.
formation process? Should NO control be included in
oxidant control strategy?
EPA is currently conducting laboratory and field studies to develop
analytical methods for measurement of the atmospheric reaction products
of NO .
x
Present NO- control strategy assumed NO to be the sole precursor
Lt .A.
of NO- and to yield NO- on a 1:1 basis. This, however, is only a simpli-
fying approximation. Although NO- is a product of the thermal reaction
of NO with atmospheric oxygen, this reaction alone cannot cause the
observed accumulations of NO- in ambient air. Such accumulation is
caused by a photochemical process that involves both NO and HC as
X
reactants.
Smog chamber testing of simulated atmospheric mixtures has shown
that maxiumum NO- concentration during the test is proportional--in fact,
almost equal--to the initial concentration of NO and nearly independent
.A.
of the initial HC concentration (Figure 4.6). However, unlike the maxi-
mum NO-, average NO- concentration during the test was found to depend on
both NO and HC. The average NO- increased with increasing initial NO
but decreased with increasing initial HC (Figure 4.7).
The aerometric data analysis done thus far verified the 1:1 rela-
tionship between the early-in-the-morning (6-9 a.m.) NO and the later-
J\.
in-the-day NO- maximum (Figure 4.8), but failed to establish the depen-
dence of the average NO- on both (the NO and HC) precursors.
Li J\.
144
-------
uidd 'ON -inoq-y
-------
CO
to
OO
o
f--
06
O-
X
o
03
•H
•H
c
LO
o
IO
O
o
o
o
to
o
LO
CM
o
o
o
LO
o
o
o
in
UTUI x mdd
- 146 -
-------
It appears that the daily maximum NCL concentration relates to the
NCL precursors simply; namely, it is proportional to the daily 6-9 a.m.
NO and is independent of HC. Unlike the daily maximum, the daily
X
average NCL is neither simply nor predictably related to the NCL pre
cursors; this makes it difficult to determine control requirements for
NCL, because the NCL-NAQS is defined in terras of an (annual) average
NCL concentration.
To determine control requirements for achievement of NCL standards,
one must consider that the yearly average concentration of NCL in a
region results from combined effects of the following processes:
•Photochemical conversion of the morning-traffic-peak NO
A.
into N07, occurring only during daytime.
•Thermal conversion of the morning- and afternoon-traffic-
peak NO into N0_.
F x 2
•Reaction of the afteinoon-traffic-peak NO with oxidant.
X
The "photochemical" MIL--the largest contributor to the annual
N07--depends on the 6-9 a.m. NO. and possibly on the 6-9 a.m. HC also.
£t Jk.
More aerometric and smog chamber data are needed here for accurately
determining dependence of the average MIL on initial (or early-in-the-
morning) levels of NO and HC.
J\.
The "thermal" N07 depends strongly and predictably on NO so that
Z -X
it could be controlled effectively by NO control. However, thermal
JTV.
N0? does not contribute to annual MIL in a major way.
The "evening" N07, resulting from the 09-NO reaction, appears to
/> L, X.
depend mainly on the oxidant level. Therefore, the most effective
means for controlling such MIL to the 24-hour N07 average has not been
147
-------
•4 4
CTl
00
OH
On
VO X
• O
03
cr>
LO I
vO
rt
CTl
I
X6
O P-i
Z
LO
c
O I
03 ?^ (X
<
O 3 U
M O
I
14-1
O
I
s
o
C 6 fn
O 3 "4-1
•H e
4-> -H Cfl
03 X -M
i—i 03 03
-------
established. However, it has been significant, especially during the
smog season. Analysis of aerometric data is needed here to establish
the importance of such NCL, within and outside the smog season, and to
determine its dependence on the oxidant and NO levels. Since present
X
photochemical pollution control strategies aim at reducing oxidant to
0.08 ppm, one could reasonably assume that evening NCL is decreasing
and will eventually become negligible.
Smog chamber studies have shown that both HC and NO have a role
in the atmospheric oxidant formation process. They have shown the
oxidant to increase with increasing HC reactant, and to show a maximum
dependence on NO . Aerometric data'similarly showed the oxidant to
J\*
increase with increasing concentrations of HC reactant but were not suf-
ficiently abundant to permit delineation of both HC and NO effects.
X
Based on this data, an oxidant control strategy calling for unilateral
control of HC emissions but overlooking the role of NO may not be the
X
most effective.
Present oxidant control strategy is based on empirical correlation
of ambient oxidant and HC concentration data that shows the daily oxi-
dant to decrease with decreasing 6-9 a.m. HC concentration (Figure 4.9).
unfortunately, there is insufficient data to establish the correlative
dependence of oxidant on NO . rfhere is evidence to suggest, however,
-A.
that the NO effect is an inhibitive one, and that as ambient HC levels
x '
decrease with control, the NO -to-HC ratio increases and the inhibitive
.A.
effect of NO is intensified. According to this evidence, ignoring the
X
NO effect unavoidably leads to an overestimation of HC control require
A.
ments.
149
-------
0.30
0.25
0.20 —
X
o
,
0.10
0.05
APPROXIMATE UPPER-LIMIT
OBSERVED OXIDANT
^ j
"WASHINGTON A
LOS ANGELES
LOS ANGELES A
k DENVtR
LOS ANGELES
A A PHILADELPHIA
LOS ANGELES
PHILADELPHI/
PHILADELPHIA
WASHINGTON
WASHINGTON A A A
A .
WASHINGTON *
A> A A A A A
A A A A
A
A At A AA A AA
A A A AA A
4*A* A A A A
A * A / A i A A A AA
A A
0.5 1.0 1.5
NONMETHANE HC, ppm C
2.0
2.5
Figure 4-2. Maximum daily 1-hour-average oxidants as a function of 6-to 9-
a.m. averages of nonmethane hydrocarbons at CAMP stations,
June through September, 1966 through 1968, Los Angeles, May
through October 1967,
150
-------
Hypothetical oxidant control strategy considering the NO role
.A.
should be expected to call for less HC control. Such a hypothetical
strategy is not possible simply because, as yet, no one has been able
to quantitate the role of NO . Aerometric data are not complete
X
enough to permit such quantification. Smog chamber data are more
complete; however, such data do not necessarily represent ambient
atmospheric conditions. This hypothetical strategy may have merit in
an urban region, but its effectiveness in controlling "transported
oxidant" buildup in rural areas downwind from cities is questionable.
4.2.2 Carbon Monoxide
The oxidation of CO by hydroxyl radicals
CO + OH ^ C02 + H
is rapid enough to suggest that this reaction might be a significant
removal process for CO in polluted atmospheres. It was recently
determined, however that the reaction cannot compete effectively with
the much faster reactions of hydroxyl radicals with hydrocarbons. '
In the absence of reactive hydrocarbons, modeling studies predict that
12
OH is an effective sink for CO." Subsequent chain reactions are effec-
tive in converting NO to N02 and small amounts of 0, an be formed,
although under non-urban conditions, not enough to exceed the air
quality standards. No significant oxidation reaction of CO has yet to
be found for urban atmospheres. The reaction of CO with H02
CO + H02 -»• C02 + OH
was though to be an atmospheric reaction of possible importance, but
this reaction is extremely slow.
151
-------
4.2.3 Hydrocarbons
New findings on pollutant transport and hydrocarbon reactivity
suggested that organics formerly classified unreactive actually may be
reactive. Based on these findings, a new classification of organics
was derived and the use of reactivity criteria was reexamined. devel-
opments prompting this reexamination include:
•Reactivity criteria used inconsistently in existing
control regulations.
•Recent smog chamber studies have improved our under-
standing of the photochemical reactivities of organic
compounds and raised questions as to the validity of
the reactivity classification in the widely used Los
Angeles County Rule 66.
•It has become increasingly evident that pollutant
transport does occur, and that such transport enhances
formation of oxidant from low reactivity organics.
•The present emphasis on control of mobile source
emissions may eventually make solvents and other
stationary source emissions the predominant emission
problem.
The Los Angeles County "Rule 66" is lacking in two respects. First,
its reactivity classification of organics is inaccurate, the error being
in the direction of less stringent control. Second, it assumes that
organics of unknown reactivity are non-reactive--an assumption that is
probably wrong and that again leads to less stringent control. EPA
Appendix B is lacking in that it assumes that all except very few
152
-------
organics are equally reactive--an assumption that is known to be wrong
and that causes inflexibility.
Recent findings concerning pollutant transport provide additional
support for the "all-organics" control concept. Thus, under prolonged
irrad ation conditions occurring in transported air masses, it is very
probably that the "less reactive" organics are induced to form as much
oxidant as the "more reactive" ones. Therefore, to alleviate the down-
wind area, nearly universal control of organics probably would be more
effective than the more selective control prescribed by Rule 66.
Control requirements for each source should include consideration
of: (a) the degree of total hydrocarbon control required for the region
or area; (b) the relative strength of the source; (c) the reactivity of
the emissions, and (d) the availability of control technology.
A new reactivity classification of organics based on existing smog
chamber reactivity data has been derived. The new classification cate-
gorizes hydrocarbon and non-hydrocarbon organics into five reactivity
class s and gives a numerical reactivity rating to each class.
4.2.4 Oxidant
Dependence of the atmospheric: oxidant on ambient levels of hydro-
carbon and NO has been explored in laboratory (smog chamber) as well
JC
as field studies. Early laboratory studies had little success--mainly
because atmospheric conditions could not be closely simulated in the
smog chamber. The oxidant - HC - NO relationships from those studies
A.
have only qualitative validity. Aerometric data gave a somewhat more
valid, but considerably less complete, relationship--namely, a corre-
lative dependence of oxidant on HC alone. The dependence on both HC
and NO could not be obtained.
.A.
153
-------
Recent smog chamber studies have used improved methodology and
experimental conditions much closer to those occurring in the real
atmosphere. Their results, therefore, are more valid than those of
earlier studies. The most significant deductions from these recent
studies are that: (1) oxidant concentrations exceeding the standard
(0.08 ppra-0,) can form from only a few parts-per-hundred--million of
hydrocarbon, and (2) the oxidant level within an urban area can be kept
below the standard by maintaining the early-in-the-morning HC-to-NO
J\.
ratio at 2.5:1 or less.
Nevertheless, the question of comparability of smog chamber atmo-
sphere remains open, and the validity of these or any other smog chamber
data cannot be fully defended at this time.
Some recent studies suggest that ozone and ozone precursors can be
transported over long distances. Concentrations of ozone are found in
remote areas that frequently exceed the air quality standard. The
probable cause for these high ozone levels has been hypothesized to be
long-range transport. Present evidence is insufficient to implicate
natural sources as major contributors, although the possibility cannot
be ruled out.
4.2.5 Particulate
Since the original criteria document was issued, major advances
have been made in understanding the sources, properties, chemical com-
position and dynamics of the atmospheric aerosol or total suspended
particulate matter.
Current problems are examined in ERC-RTP Task Force documents:
Summary Report on Suspended Sulfates and Sulfuric Acid Aerosol and
154
-------
Briefing Report on Problems Relating to Fine Suspended Participates.
The National Academy of Sciences has prepared a review, now in draft,
as background for preparation of a new criteria document. Therefore,
in this document only a few key references are discussed. While no
decisions have been made, it is clear that both particle size and par-
ticle composition must be considered and that in addition to primary
aerosols, gaseous precursors of secondary aerosol (aerosol formed by
chemical processes in the atmosphere) must be controlled. These include
S0? (forms of sulfate aerosol), NO (forms nitrate aerosol) and certain
Li .A.
organic vapors (from organic aerosol).
Key areas of progress include:
(1.) Understanding the multimodal distribution of the atmospheric
aerosol, the typical observation cf a bimodal mass-size distribution,
and the differences in chemical composition and effects of fine and
coarse aerosols.
(2.) Use of microscopy and the chemical element balance technique
to determine aerosol source coefficients from measurements of ambient
aerosols. This has led to the realization that secondary sulfate,
nitrate and organic aerosol comprise most fine particulate matter and
that in urban areas, road dust, tire debris and other material generated
by vehicular traffic make a major contribution to coarse particles.
(3.) Identification of organic material as a major primary and
secondary component of fine particulate matter, and determination of the
chemical composition of many individual organic compounds in the atmo-
spheric aerosol.
155
-------
(4.) Realization of the regional nature of sulfate pollution,
identification of several compounds including acid sulfates in the
atmosphere, improved understanding of the coversion of SCL to sulfate
and measurements of transformation rates in power plant plumes and
urban atmospheres.
(5.) Development of improved instruments and techniques for mea-
suring aerosol parameters important in visibility reduction, radiation
balance and weather and climate modification.
4.2.5.1 Distribution of Aerosol and Chemical Composition as a Function
of Size-Distribution--Early studies of ambient particulate matter
used five-stage impactors and found that the size distribution could be
adequately accounted for by a log-normal distribution. However, in
recent years counting techniques and eight-stage impactor studies have
demonstrated a bimodal size distribution in some locations. In the fall
of 1973, three separate studies confirmed a bimodal mass distribution
and review of counting studies indicated a bimodal volume distribution
for almost all measurements.
Most aerosol scientists agree that atmospheric particulate matter
is multimodal with a variety of particles from different sources having
different properties. It is a convenient hypothesis to consider particu-
late matter as being composed of two distinct size classes, fine and
coarse, and such a hypothesis has both an experimental and a theoretical
basis.
There are two mechanisms for creating new particles in the atmo-
sphere. Nucleation and gas-particle conversion produce very small
particles which grow by coagulation and condensation. As these small
156
-------
particles become larger, the growth processes slow down and virtually
cease as the diameter approaches one micron. Comminution or mechanical
processes such as breaking condensed matter into smaller particles by
grinding, rubbing, etc., produce particles with a mass mean diameter
much greater than one micron. The resulting atmospheric distribution
in terms of number, surface and volume is shown in Figure 4.10. The lower
mode may be called the optical mode since it covers the size range most
effective in scattering light, or the accumulation mode since smaller
particles grow into this size range. Particles in the lower mode are
also within the respirable size rajige.
4.2.5.2 Composition and Concentration--Although the concentration
distribution of fine and coarse particles in the ambient air has not
been adequately measured, considerable information about the fine par-
ticulate mode can be obtained by observing composition as a function of
impactor stage. Most comes from the National Air Surveillance Cascade
"IT "I fL
Impactor Network. Other data have been obtained by research groups.
These data indicate that among metals, Pb and V are found primarily in
the fine particle mode; Fe, Mg, Mi, Ba, Cd and Cr are found primarily
in the coarse particle mode. SO^, NO^, Carbon, Cl, NHj and organic
particulate are found in the fine particle mode and P0| in the large
particle mode. Zn, Cu, Ni and Sn do not fall clearly in either mode.
The data are sparse and limited to a few locations.
Where complete analyses of atmospheric particles have been made
(in a limited number of locations), fine particles comprise 25-6-1 of
the total suspended particulate matter, and from 60 to 80% of the fine
14 17
particles are secondary. '
157
-------
rrmrj—.-? iiini|—rnmri—I JTUIIIJ rn
Hgure 4.10. Number, surface area, and volume distributions of a
hypothetical smog. The linear ordinate normalized by total number,
area, or volume is used so that the apparent area under the curves
is proportional to the quantity in that size range.
4.2.5.3 Determination of Aerosol Source Coefficients from Ambient Air
Sampling and Analysis—Combined chemical, physical and microscopic
analysis of collected aerosol is a powerful tool for the identification
and quantification of aerosol sources. Aerosol emission inventories
are source-oriented, giving primary, anthropogenic emissions only.
Inventory information is in mass of particulate material with no size
or chemical composition. Aerosol source coefficients are receptor-
oriented and express various source contributions to the total ambient
aerosol at the point of measurement. Instead of primary anthropogenic
158
-------
emissions only, natural sources, secondary aerosol and unanticipated
sources (such as roadway dust) are included.
Source coefficients are useful for providing quantitative informa-
tion on potential improvement in local aerosol concentrations that would
result from either partial or total elimination of emissions from par-
ticular sources. Also, they emphasize the substantial contributions to
urban aerosols from natural and secondary sources.
The technique of calculating source coefficients from measurements
of tamospheric aerosol concentrations and chemical composition was devel-
18
oped by Friedlander. It has been applied to Pasadena, California, by
19 20 21
Miler et al., Friedlander to Chicago, and St. Louis by Gatz.
This technique can be combined with a simple meteorological model
21
to give areawide emission rates of various individual elements. When
size-differentiated measurements are available, additional information
is obtained for a better estimate of health effects and visibility
22
reduction.
18
Friedlander's results for Pasadena are shown in Table 4.5.
Data from several other cities show that secondary sulfate, nitrate
and organic aerosol contribute 60-80% of fine (< 2 micron diameter) par-
ticles and that the fine particulate component varies from 25-60% of the
total suspended particulate. The contribution of some secondary compon-
ents to total suspended particulate matter is shown in Table 4.6.
159
-------
Table 4.5.
ESTIMATED CONCENTRATIONS OF LOS ANGELES
AEROSOL PARTICLES BY SOURCE18
(Annual Average, 1966)
Source
Natural background
PRIMARY
Dust rise by wind
Sea salt
Spores, pollen, etc.
SECONDARY
Vegetation (organic vapors)
Biological (soil bacterial action,
decay of organics)
Man-made
PRIMARY
Motor vehicles
Organic solvent usage
Petroleum
Aircraft
Combustion of fuels
Other
SECONDARY
Organic*
NH,, NO, , S. .
SO
,12.
Total accounted for
Measured total
Mass Concentration
(iig/m5)
8-20
3
Unknown
3-6
0-7
15
6
1.3
4
5
5
11
13.5
14.4
90-105
119
*Benzene soluble organic material. This is probably low
because some oxidized organic material may be insoluble
in benzene.
160
-------
Table 4.6. PARTICULATE POLLUTION INCLUDING CONSTITUENTS
LARGELY OF SECONDARY ORIGIN
Annual Geometric Mean
vg/m
City
Boston
Chicago Site 01
Detroit
Los Angeles
New York
Philadelphia
Pittsburgh
San Francisco
Total
93
112
134
129
118
112
161
84
SQ2-
6.0
10.7
12.5
10.8
18.6
18.4
14.2
7.6
N03.
1.0
--
3.5
7.7
--
2.1
2.5
2.9
Org*
--
7.4
7.4
15.0
9.3
8.5
9.3
7.7
NH4
1.3
0.8
0.5
0.8
--
1.8
0.8
0.2
Source: EPA
*Benzene soluble organic material. This includes some
primary aerosol but is probably low because some
iodized organic material may be insoluble in benzene.
A substantial fraction of coarse particulates in urban areas can
be attributed to vehicular traffic-tailpipe emissions, rubber tire dust
and mineral fragments. In areas with unpaved roads or parking lots,
23
very large amounts of road dust are generated. Road surfacing and
street cleaning are effective control techniques.
4.2.5.4 Organic Aerosol--It has been discovered that organic particu-
late matter contributes more to total particulate matter than previously
indicated by benzene-soluble measurements. Especially during photochemi-
cal smog are highly oxidized, polar organic compounds formed which are
161
-------
insoluble in benzene. Studies using more polar solvents have shown that
the percentage of organic matter in total suspended particulate in
24
Pasadena, California, averaged 30-40% and varied from 5 to 60%.
25
O'Brien, et al. studied the organic fraction in suspended particle
matter collected at several California sites. Benzene soluble varied
from 1-121 and averaged 4%; additional organic material extracted by an
alcohol mixture varied from 10-42% and averaged 23%. Less extensive
studies in Columbus, Ohio, New York City and Pomona, California, gave an
average of 36% and a range of 16-56%.
Field and laboratory studies have been conducted to determine the
chemical composition of the org nic fraction. Model aerosols have been
generated from individual hydrocarbon precursors-toluene, alpha-pinene,
cyclohexene, and 1-heptene and nitrogen oxides under simulated atmo-
spheric conditions in a 17.3 m environmental chamber. Aerosols were
collected on glass-fiber filters and organic matter was obtained by
solvent extraction. Organic reaction products were fractionated into
acid, neutral, and basic components, and were analyzed by techniques
including gas chromatography and gas chromatography combined with mass
spectrometry. Different polyfunctional reaction products were identi-
fied, including alcohols, aldehydes, ketones, mono- and dicarboxylic
acids, and phenols. Tentative identification was made of nitrogen-
containing products including nitrate esters and aromatic nitro
17
compounds.
Computer-controlled high resolution mass spectroscopy was used to
identify organic compounds in ambient air samples collected in Pasadena
and St. Louis. Several hundred different compounds can be identified
162
-------
based on their mass spectrum and time of volatilization from the heated
27
sample. The diurnal variation in aerosol composition has been studied
for two size ranges (less and greater than 2 wn) and used to postulate
primary and secondary origins of the aerosol. The primary organic aero-
sols included alkyl benzenes, alkaaes, polycyclic aromatics and phenols.
Secondary organic aerosols included acids, aldehydes, alcohols, chlo-
28
rides and nitrates.
Laboratory studies of aerosol generation have shown that aromatic
compounds and cyclic olefins are very effective in producing organic
29
aerosol. The aerosol-forming reactivity of various hydrocarbon com-
pounds is shown in Table 4.7.
It has been suggested that some or all of the reported organic
aerosol may be organic vapors adsorbed on the fiberglass filter. Hanst,
et al.30 measured HCHO, HCOOH and PAN on one of the days that Friedlander
measured organic aerosol. There was as much carbon in these gaseous pol-
lutants as in the organic aerosol. Hanst has suggested that some of the
material reported as organic aerosol may have been absorbed HCHO and
HGOOH and absorbed decomposition products of PAN.
4.2.6 Sulfur Oxides and Aerosols
4.2.6.1 Laboratory Investigations;
4.2.6.1.1 Photochemical and Chemical Oxidation--Chemical mechanisms
transforming SO- in the atmosphere: are still not well known. Oxidation
163
-------
29
Table 4.7. AEROSOL AND OXIDANT REACTIVITY OF HYDROCARBONS
Hydrocarbon
Hexanal
Isooctane
Butane
2-Methylbutane
1-Butene
Ethylene
Cyclohexane
2 -Methyl - 2 - butene
1-Pentene
1-Heptene
Toluene
Mesitylene
Benzene
3-Heptene
n-Butylbenzene
o - Die thy Iben zene
t-Butylbenzene
m-Xylene
Cyclohexene
Max
. 0
0.16
0.37
0.79
0.39
0.68
1.4
1,1
1.1
7.7
11.1
5.7
17.0
4.2
15.8
18.2
20.8
13.7
36.4
Avg
0
0.08
0.18
0.21
0.32
0.34
0.45
0.84
0.89
2.1
2.8
3.0
3.5
3.8
7.2
7.6
7.9
9.5
32.4
Category'*"-'
1
1
1
1
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
Max
0.15
0.08
--
0.05
0.45
0.46
0.10
0.12
0.37
0.17
0.37
0.65
0.06
0.41
0.10
0.22
0.04
0.36
0.41
Category1 J
2
1
1
1
3
3
2
2
2
2
2
3
1
3
2
2
1
2
3
(a) Category Aerosol^ ^ Oxidant
1 <1.3 <0.1
2 1.3-4.0 0.1-0.4
3 >4.0 >0.4
(b) Category based on average aerosol.
by photochemically generated intermediates and catalytic oxidation are
now thought the most likely mechanisms.
Laboratory investigations of the photochemical oxidation of sulfur
dioxide in air--alone and in the presence of nitrogen oxides and
164
-------
hydrocarbons --indicate that the homogeneous reaction of electronically
excited sulfur dioxide ( SO-) cannot account for the observed oxidation
rate of SCL in the atmosphere. Current belief is that reactive imter-
mediates formed in the hydro carbon -NO photochemical system are react-
X
ing with ground state molecules of S02 to form products. These products
are thought to account for the homogeneous formation of SO, and sulfuric
acid mist. Swedish and U.S. workers have identified acid aerosol (H^SO.
and/or NH4HS04) , (NH4)2S04, (NH^^fSO^ and CaS04 in atmospheric sam-
31,32
pies.
The three-body oxidation of S0? + 0 atoms has been postulated to
explain the conversion of S0~. However, the recent evaluation of the
34
rate constant by Davis with N? as the third body has shown the reac-
tion to be almost four times slower than previously thought. Thus,
this reaction is of only minor importance under most conditions.
H0x + S02
Wood, et al. recently conducted a series of photochemical experi-
ments to measure the kinetics of the S0~ reaction with OH. Their data
were interpreted according to the following mechanism:
OH + S02 ->- HS03*
HS03 + M •* HS03
HS03 + OH -> products
2 -1 -1
A rate constant of 5.6 x 10 ppm min for the OH + S02 reaction was
obtained. Depending on the altitude and presence of pollution sources,
OH concentrations between 10 and 10 molecules/cm have been predicted.
Assuming S0? = 1 ppm and OH = 4 x 10 ppm, the rate of S02 oxidation is
165
-------
81 per hour. Thus particle formation from the reaction of S0? with
hydroxyl radicals will be significant under some atmospheric conditions.
The transient species H0? has long been postulated to play an
37
important role in atmospheric chemistry. Davis, et al. have demon-
strated that H0? can readily react with S0?, with the rate constant for
2 -1 -1
this reaction found to be 4.4 x 10 ppm min . This implies that the
H02~SO? reactions may be important in both tropospheric and strato-
spheric chemistry.
Recent modeling of reactions of S02 in the presence of hydrocarbons
and NO mixtures at reactant concentration levels less than 1 ppm indi-
A.
cate that reactions of SO- with OH and HO- can account for up to 10% of
•70
total S02 reacted.
R0x + S02
Indications are that oxidation of SO,, by oxygenated organic radi-
cals (RO, R0?) may be important. Recent studies have been made of
the propylene-S07-0, dark reaction to identify the specific radical
39
responsible for the formation of SO,. Cox, et al." attributed the
oxidation of S02 to a reactive species formed in the 0,-olefin reaction.
40
McNelis studies indicate that the reactive biradical species, RCHOO ,
is responsible for the oxidative consumption of SO,.,. The steps of the
proposed mechanism are
03 + CH3-CH=CH2 -> HCHO + Q^CHOO-
03 + CH3-CH=CH2 + CH3CHO + HCHOO-
RCHOO- + 02 -> -OH + RCOO-
RCHOO- + 07 -> 09 + RCHO
Li Li
CH CHOO -* CH?=C=C + H?0
•3 L* t*
166
-------
•OH + CH3-CH=CH2 + CH3CH-CH OH
RCHOO- + S02 + SO + RCHO
Other Reactions
Other reactions have been postulated as possible contributors to
SO,, oxidation processes. S0~ has been observed to react in the presence
of N0~ and 0_; it has been speculated that this could be due to reaction
of S0~ with N0_ or NJDr which are generated by the reaction of 0, with
N02.
Although the reaction of S02 with 0, has been shown to be slow in
the gas phase, there is some evidence that this reaction occurs rapidly
in the presence of water droplets, and might thus be expected to occur
rapidly on the surfaces of moist particles.
Oxidation Rates
Recent results from smog chamber studies at Battelle, Calspan
and the University of Minnesota show that the SO,., photooxidation rate
differs from chamber to chamber by as much as two to three orders of
magnitude. The highest reported rate for S0? in the "clean" air cham-
ber system (61 per hour, Battelle) is comparable to that in the real,
urban atmosphere and does not seem to be much affected by the presence
of propylene/NO . Experimental study is now under way to find whether
J\.
observed discrepancies in the oxidation rate reflect contaminant effects
or chamber surface effects.
4.2.6.1.2 Aerosol Formation--The marked influence of S02 on aerosol
formation in photochemical smog simulation studies is well documented.
Although the exact mechanism by which S02 is converted into aerosol
remains uncertain, these general conclusions can be drawn: (1) In the
167
-------
absence of SCL, saturated hydrocarbons produce little, if any, aerosol.
In the absence of SCL, aromatic hydrocarbons and olefinic hydrocarbons
with five or more carbon atoms form aerosols; those with four carbons or
less do not. Dienes, terpenes and cyclo-olefins also form aerosols in
the absence of SCL. (2) SCL increases light scattering in studies of
individual hydrocarbons and is particularly noticeable with olefins,
where the increase is enormous. Increase with paraffins is moderate,
and with aromatics, small. (3) SCL increases the condensation nuclei
count in studies of hydrocarbon-NO mixtures. In the case of 1-
dodecene and 1,3-hexadiene, only a few ppb of S0? are required to pro-
duce an order of magnitude effect. With aromatics only small increases
are seen. (4) In general the formation of aerosol in hydrocarbon-NO -
J\.
SCL systems does not occur to any appreciable extent until most of the
NO is converted to N0~ and 0_ begins to form. (5) Experimental evidence
shows that S0~ influences the reaction rate parameters in simulated smog,
the degree of influence depending on such factors as realtive humidity,
initial N09/N0 ratio, hydrocarbon/NO ratio and hydrocarbon type. (6)
L, -X !X
There is controversy about the chemical composition of aerosol formed
from the hydrocarbon-NO -S07 photochemical system. However, there is
-A- L*
compelling evidence that the aerosol formed from C<5 olefins-NO^-SO^ is
predominantly inorganic--mainly H~SO.--whereas the higher olefinic hydro-
carbons produce carbonaceous aerosol as well. Studies on the propylene-
P_-SO,, system indicate that the dominant conversion for SO- results in
formation of H-SO.. In the presence of NH_, NH.SO. aerosol is observed.
Infrared spectra of dried, Cf-LCl- extracts, of atmospheric and smog cham-
ber aerosols, show C-H absorptions around 3 y, C = 0 absorptions around
168
-------
6.5 y, C = C and/or N = 0 absorptions around 6.7 u and a number of peaks
at longer wavelengths. Gas chromatographic-mass-spectrometric studies
indicated organic acids and nitrate esters were also present. Infrared
spectra of aerosols formed from 1-heptene-NO with and without SCL indi-
Jv b
cated that with SCL the aerosol contained a sulfonic acid.
4.2.6.1.3 Catalytic Oxidation--Reaction rates for the system SCL-H-O-
solid surfaces have been made. It was found that fractional collisions
of SCL leading to reaction with AKO, and MnCL is 3x10 , and with fly
ash and charcoal, 5x10 . These values suggest lifetimes of about one
hour in urban plumes and 10 hours in coal-fired power plant plumes for
relative humidities of 50%. Lifetimes increased as realtive humidity
decreased. However, other studies of SCL-H-O-Fe^O, indicate that reac-
tion does not depend on relative humidity. For high SCL concentrations
41
(>700 ppm) and large liquid droplets, Matteson and Giardina measured
a 601 increase in the mass transfer of SCL to the droplet if a surface
2
charge density of 6stat C/cm was applied. The magnitude of the effect
is not known for ambient SCL concentrations charged raindrops.
4.2.6.1.4 Physical--The solubility in water of SCL at low concentrations
42
has been measured at 25.1 C as function of pH. The scheme assumed was:
SCL + H00, . t SCL + H00
2 gas 2 liq 2 aq 2
SCL + H~0 ? H70 + HSO%
2 aq 2 3 3
The second ionization
HSO~ + H20 + H30+ + S032"
was assumed to be negligible. The experimental results agree well with
values extrapolated from measurements at higher SCL concentrations by
43
Johnstone and Leppla.
169
-------
4.2.6.2 Field Studies--Atmospheric data on SCL and particulate sulfates
are available from: 1) 24-hour integrated network measurements of SCL
and suspended water-soluble sulfates at urban and nonurban continental
and maritime sampling sites in the United States, and 2) from field
studies of atmospheric aerosols and gaseous smog components. Interpre-
tation of data has emphasized: (1) the distribution of SCL and sulfate,
(2) the relationship between SCL and sulfate, and (3) chemical forms of
sulfur in particulate matter.
4.2.6.2.1 Distribution of SO,-,--Recent analysis of national and regional
trends in sulfur dioxide show a pronounced decline in sulfur dioxide
44
levels throughout the nation. Figure 4.11 shows that this downward trend
for 1970-1973 occurs throughout the nation with perhaps a degree of
leveling off at stations where concentrations are well below the annual
standard. Based on the number of observation sites (Table 4.8), these
downward trends represent general patterns in urbanized population cen-
ters, but may not adequately represent SCL levels near nonurban point
sources. Data from background sites show that background SCL levels
remained relatively stable in the 1970-1973 period.
Table 4.8. SURVEILLANCE CATEGORIES FOR SULFUR DIOXIDE SITES44
(Number of Sites)
Geographical Area
North
Northeast ^outh Central Midwest West Nation
Population-oriented
Source-oriented
Background
65
1
4
id
1
2
62
2
2
33
0
2
21
0
0
199
4
10
Total 70 21 66 35 21 213
170
-------
T~I i ni r
o
a.
t
3:
o
I I I I I I I I
UJ
£
O
T3
X
O
3
co
C
(LI
S_
(T3
C
O
en
O)
•o
c:
ro
CO
ra CT>
c: i—
o i
•i- O
ro cr»
B,
•H
ft.
171
-------
45
4.2.6.2.2 Relationship between S0? and Sulfates--Altshuller's analysis
of 1964-68 National Air Surveillance Networks data for S02 and sulfate
showed that the relationship between atmospheric SO- and sulfate is not
simple. His results showed that a reduction of SCL may not cause a pro-
portional reduction in sulfate, suggesting in turn that control of other
factors--hydrocarbon, NO , primary aerosol and NH,--in addition to SCL
X O L*
also may be necessary to achieve acceptable low sulfate levels. A labora-
tory program designed to determine the effect of these factors on oxida-
tion of SCL has been initiated. These effects are being studied under
(1) conditions similar to those prevailing in air above an urban area,
(2) conditions simulating transport where the transported pollutants
receive prolonged irradiation, and (3) conditions within air masses in
the vicinity of roadways.
4.2.6.2.3 Chemical Forms of Sulfur in Particulate Matter--In the past
most studies have determined only the sulfate content of ambient
aerosols. Craig et al. have undertaken a study to establish a com-
plete inventory of sulfur compounds present in ambient particulate
matter. Seven separate chemical species of sulfur have been identified
-2 -2 0
in atmospheric samples taken in California: SCL, SCL , SCL, SO, , S , and
two kinds of sulfides. Sulfates were found to be the dominant species,
although reduced forms of sulfur were at times comparable to the sulfate
concentrations.
Novakov et al. determined the chemical states of sulfur in smog
particles collected in two particle size classes as a function of time
of day in Pasadena, California. The sulfur content was found to be
highest at night and consisted of S species and S species. The S
172
-------
species predominated at night on the smaller particles, and was presumed
to be adsorbed S02 which was later oxidized to S (SO. ), and predomi-
nated in larger particles during the day.
48
In laboratory studies, Taarit et al. have shown that S0~ and MgO
powder react in sunlight to form SO,. It is likely that this reaction
occurs in a liquid phase on the surface of the MgO, and may occur on all
basic metal oxides in the atmosphere. The ion SO,., is formed by exposing
V20r on silica gel and MgO to SO,-,. Although SO,, and H2S in the gas phase
49
react alowly, Rao and Lunsford have found the reaction of the S02
surface ion is rapid and yields free sulfur. They also observed the SO.
surface ion.
4.2.6.3 Transport and Removal--SO,, in the atmosphere may undergo reac-
tion with the earth's surface (grass, trees, concrete, etc.). The flux
of SO,, to ground can be expressed as
Flux = K 9c
Z 3i"
where K = the vertical eddy diffusitivy of mass, cm /sec
Li
3c = the vertical concentration gradient of S07 near
3l Z
the ground, gm/cc/cm
The flux is directly proportional to the local SO,, concentration, and
Flux = vc
where v = proportionality constant, commonly called the "deposition
velocity," cm/sec
c = SO,, concentration
Therefore, v = 1 K 3c .
c Z 3z
173
-------
The three terms on the right side of the above equation can be measured.
The magnitude of the deposition velocity, v, is a measure of the
reactivity of the surface. The values reported recently for the depo-
sition velocity, v, of SCL on a variety of surfaces are given in Table 4.9,
Table 4.9. S02 DEPOSITION VELOCITIES
Type of Surface
wet leaves
grass
grass
grass
vegetation
(St. Louis)
iron (90% rela-
tive humidity)
zinc (90% rela-
tive humidity)
Deposition Velocity
1 cm/sec
0.8 cm/sec (summer)
0.3 cm/sec (autumn)
0.55 cm/sec
0.8 cm/sec
.5 to 5 cm/sec
0.015 cm/sec
0.01 cm/sec
0.049 cm/sec
0.063 cm/sec
Reference
50
51
52
53
53
54
55
54
55
If the deposition velocity is as large as more recent studies indicate,
surface removal of S0? in air layers which are well mixed to the ground
may dominate S07 removal and conversion processes.
4.2.7 Mathematical Modeling
The EPA Meteorology Laboratory concentrates most of its efforts on
studies of transport and dispersion. It develops and tests mathemati-
cal models to help to relate rates of source emissions to concentrations
174
-------
at receptors, depending on meteorological conditions. A general discus-
sion of modeling is given in Appendix A, Air Quality Criteria for Sulfur
Oxides. Models have been developed to estimate pollution concentra-
tions over an entire city from a single large source like a power plant
stack, or from a source such as a highway or an airport.
The UNAMAP (Users' Network for Applied Modeling of Air Pollution)
was put into operation in May, 1973. The purpose of UNAMAP is to make
available current air quality simulation models to both EPA and non-EPA
users via a teletype-processing network. The models are all in the form
of computer programs accessible from remote terminals connected to a
central computer by telephone lines.
The six models now on UNAMAP are:
APRAC - urban carbon dioxide model.
HIWAY - hourly concentrations from roadway segments.
CDM - Climatological Dispersion Model.
PTMAX - maximum hourly concentrations from a point source.
PTDIS - hourly concentrations at various downwind distances
from a point source.
PTMTP - hourly concentrations at various receptor locations
from multiple point sources.
The practical application of dispersion models is shown by the fol-
lowing example. Studies concerned with the impact of state implementa-
tion plans on low-sulfur coal requirements predicted a probably deficit
in 1975. Therefore, hourly emissions and meteorological data were used
to model peak 1-hour, 24-hour and annual concentrations of SCL in the
vicinity of 180 power plants. It was determined that 62 plants might
175
-------
be able to operate with full variances, 39 with limited variances and 80
with no variances. The shift in full requirements should result in an
annual saving of 137 million tons of low (less than 1.0%) sulfur coal.
S7 *-iR
A different type of mathematical model ' has also been developed
based on actual air pollution measurements in cities at intervals of
time from one hour to one year. It shows statistical relationships and
is needed for interrelating air pollutant effects, air quality stan-
dards, air quality monitoring results, diffusion calculations, source-
reduction calculations and emission standards.
4.2.7.1 Fluid Modeling Facility--A fluid modeling facility is being
established to study the flow and diffusion of pollutants around build-
ings, highway configurations and complex terrain. It includes large
and small wind tunnels, a model shop, electronics shop, minicomputer,
photographic darkroom, classroom and offices. A contract has been
awarded for the construction of a recirculated water channel/stratified
towing tank. Layered mixtures of salt water will be used in the test
section to produce density stratification. Dyes and shadowgraphs will
be used for flow visualization.
Examples of results that can be obtained with a tow tank are
59
found in similar work performed under EPA contract. One study was
of plume dispersion from an elevated source upwind of an idealized
mountain under two different stability conditions. Highlights of
the study were: (1) within stably stratified flow is upstream blocking
that causes increased plume rise and spreading; (2) the plume tends to
flow around rather than over the ridge, thus greatly enhancing the
plume's lateral spread; (3) concentrations would be larger at stations
on the mountain ridge than at other surface stations.
176
-------
4.2.7.2 Regional Air Pollution Study (RAPS)--There is a need to develop
and validate improved air quality simulation models for applying to
least-cost pollution control strategies, and to other models such as
those for land use, ecological and total cost effectiveness. This need
was recognized in setting the objective for the Regional Air Pollution
Study (RAPS), a five-year study in the St. Louis metropolitan area using
an extensive base of emission, meteorological and air quality data with
research on the processes of pollutant transformations, dispersion and
removal. A 25-station telemetering network will provide measurements of
CD, NO, NCL, 0.,, total hydrocarbons, non-methane hydrocarbons, total
gaseous sulfur, SCU H S, fine particulates, total particulate and meteo-
rological observations. Two large helicopters from EPA's Environmental
Research Center at Las Vegas are outfitted to measure most of the para-
meters measured at the fixed stations. Considerable effort has been
spent setting up data formats and developing software from the RAPS data
bank, which will include emission data and models, all data from the
aerometric network, meteorological upper air sounding data, and most
data from special studies to be performed in conjunction with RAPS.
Studies have already been performed to describe the urban boundary
layer structure with emphasis on morning and evening transition periods
and investigators have made extensive measurements of aerosols in and
around St. Louis. Most special studies are being done cooperatively
with the METRQMEX project, which is concerned with the increased pre-
cipitation amounts observed downwind from large urban or industrialized
areas.
177
-------
4.2.7.3 Large Power Plant Effluent Study (LAPPES)--The Large Power
Plant Effluent Study (LAPPES) was to determine the pattern of ground-
level concentrations of pollutants released to the atmosphere from
large coal-burning mine-mouth power plants. Four volumes of basic
60-63
data obtained from the LAPPES project have been published. Early
analyses indicate that tall stacks help to reduce or sometimes eliminate
significant ground level concentrations under some meteorological condi-
tions, particularly in the stack's vicinity. However, the study also
showed that there are some things tall stacks cannot do. For example,
reasonably increasing stack height will not significantly reduce epi-
sodic ground-level concentrations of a plume trapped within a limited
mixing layer; if emissions increase, so will ground-level concentrations.
Washout of SCL from the plume was also studied in the LAPPES. The effect
of precipitation falling through the plume seems to bring SCL closer to
the ground.
4.2.7.2 Rough Terrain Effects--According to theoretical estimates, high
concentrations of air pollution can occur when a plume from a stack
encounters high ground, and we need to know if such concentrations can
actually occur. Field studies using tracers have been conducted in rough
or mountainous area, one such having been done cooperatively with the AEC
64
at Huntington Canyon, Utah. There it was found that depending on
atmospheric stability conditions, dispersion rates were 4-15 times
greater than might have been estimated using customary procedures. Nfore
rough terrain field studies are planned.
4.2.7.5 Highway Studies--In the past few years more and more governmental
and private groups have used highway (line source) air quality models for
178
-------
environmental impact statement (EIS) preparation and review. Although
several mathematical models have been developed for highway evaluations,
little is known about the initial spread of pollution due to the turbu-
lent wake of vehicles and subsequent dilution within the first hundred
meters of travel from the highway. The Meteorology Laboratory has
sponsored a tracer study of dispersion within the immediate vicinity of
65
a highway. An inert tracer gas (sulfur hexafluoride) was dispersed
from a vehicle moving in the traffic stream of a six-lane highway for
sampling at 12 locations at distances of about 100 meters downwind and
20 meter upwind. Carbon monoxide was measured simultaneously. Early
results indicate that vertical dispersion in the immediate vicinity of
the highway is larger than assumed in the EPA HIWAY model and that the
turbulent wake significantly affects concentrations.
4.2.7.6 Airport Studies--Two field studies have been conducted at air-
ports. The first, at Washington National Airport, was to validate
f\ 7
and further develop an existing mathematical dispersion model. For
six months continuous records of air quality, meteorological and
operational records were compiled. Carbon monoxide, particulates, non-
methane hydrocarbons and nitrogen dioxide were sampled at eight loca-
tions. The final result showed the modified model to consistently
predict mean and median values for CO and particulates (ranging from 16%
under prediction to 36% over prediction). These validation results were
without calibration of any kind. Better results are expected in the
future when the model will be calibrated. The second study, at
Atlanta International Airport, was to determine the effect cf experi-
mental aircraft emissions control operations on air quality. The
179
-------
emissions control procedure consisted primarily of shutting down one
engine by approved aircraft during taxiing from landing runway to
terminal building. A network of eight monitoring stations was used.
While a maximum hourly reduction of about 15% occurred in aircraft
emissions, the instrument network did not detect a statistically sig-
nificant reduction in air quality levels which could be attributed to
emission control procedures.
4.2.7.7 Precipitation Washout of Sulfur Compounds--A model for the
69,70
washout of sulfur dioxide from power plant plumes has been developed
based on reversible gas absorption phenomena. It can be used for pre-
dicting washout of pollutant gases other than sulfur dioxide, with
substitution of appropriate physical properties. The model is based on
concepts of reversible behavior and negligible plume distortion by the
washout. Field experiments showed that sulfur dioxide concentrations in
rain can be calculated within a factor of two of observed values, although
predictability is reduced in the immediate vicinity of the source and
complicated by the occurrence of chemical reactions.
Conclusions from the model, verified by observation, are:
1. Washout of sulfur dioxide from a concentrated, high-elevation
plume can be obscured by the presence of low-elevation, low-concentration
background levels;
2. The acidity of rain strongly influences its sulfur dioxide
scavenging potential and
3. Negative washout occurs as the result of desorption of sulfur
dioxide from the rain below the sulfur dioxide plume.
180
-------
Sulfate washout measurements vrere also made. Results indicate that
the rate of sulfur removal via sulfate washout is up to five times more
rapid than that via sulfur dioxide washout.
5.2.7.8 Precipitation Scavenging of Inorganic Pollutants from Metropoli-
tan Source--In the summers of 1972 and 1973 field experiments were con-
ducted in the St. Louis area to measure the precipitation scavenging of
71
inorganic pollutants. Concentration measurements of collected rain-
water have shown that precipitation, scavenging by convective storms can
be an effective mechanism for removal of urban pollutants from the
atmosphere. Results of the study indicate that scavenging removal rates
can be comparable to urban emission rate estimates.
4.2.7.9 Simulation of the Photochemical Reaction Processes Within a
Power Plant Plume--Based on studies of chemical mechanisms observed in
smog chambers a preliminary plume dispersion model has been developed
to investigate SCL + S0| transformation in power plant plumes. Pre-
liminary results indicate transformation rates of 2-3% per hour. Higher
72
rates of about 20% per hour were reported by TVA investigators.
73
Results from another study agree with the model. Large concentrations
of particulate in the TVA plume may have caused more rapid SCL conversion
because of catalytic reactions. In the other study the plumes had low
particulate concentrations and therefore only homogeneous gas phase con-
version processes were important.
4.2.7.10 High Surface Ozone Concentrations--Recently, high surface
ozone concentrations have been observed in rural areas remote from urban
centers, with values at times exceeding the National Ambient Air Quality
Standard. Consequently, mechanisms for bringing ozone down from the
181
-------
stratosphere, formation during lightning discharges, and near surface
chemical processes were reviewed. High surface values can be caused by
direct transport downward from the stratosphere through the jet stream
region behind strong cold fronts and by convective activity transport-
ing downward ozone previously injected into the upper troposphere.
Lightning contributes only brief small rises in ozone concentration, a
conclusion agreeing with observations of thunderstorm effects made in
74
a study of ozone conditions near 33 cities during summer, 1971.
Chemical reactions make significant contributions over a period of
time, but chemical and physical processes that cause high surface ozone
concentrations are not well understood. Laboratory and field studies
are in progress. In one instance computerized meteorological data from
regularly reporting stations, compiled on a global scale, is being used
to prepare trajectories automatically for tracing the origin of air in
which high ozone concentrations have been observed.
4.2.7.11 Los Angeles Reactive Pollution Project (LARPP)--During the
Los Angeles Reactive Pollution Project (LARPP), a photochemical field
experiment, parcels of air were followed by especially instrumented
helicopters and mobile van in order to determine changes in reactive
pollutants. The helicopters were tracked by radar and guided by bal-
loons that floated at near constant levels. Tracer materials were
also released and sampled. Results of this project are to be reported
in the Santa Barbara, California, workshop-symposium to be conducted
November 12-14.
In connection with LARPP a team from the Meteorology Laboratory
made special solar radiation observations from August 30 through
182
-------
November, 1973. Six stations were instrumented, one of which was on a
mountain at 6000 feet elevation, /is expected, atmospheric pollution
substantially reduced incident ultraviolet energy. For example, when
visibility at El Monte was approximately three to five miles (a common
occurrence during the pollution season), the incident ultraviolet energy
was typically about 30 to 40% less than that received on top of the
mountain, depending on solar elevation. In the absence of pollution
the difference would have been expected only to be about 5% less.
Obviously, the top of a pollution layer receives considerably more
ultraviolet irradiance than the bottom layer. Most photochemical models
today treat ultraviolet energy in a simplistic way and do not account for
its vertical variation through the pollution layer. It has been suggested
that these models improved by allowing for such vertical variation of the
ultraviolet radiation.
183
-------
4.3 REFERENCES
1. McConnell, J. C., M. B. McElroy, S. C. Wolfsy. Natural Sources of
Atmospheric Carbon Monoxide. Nature. 233 (5316):187-8. 1971.
2. Weinstock, B., H. Niki. Carbon Monoxide Balance in Nature. Science.
176:290. 1972.
3. Levy, H. II. The Tropospheric Budgets for Methane, Carbon Monoxide
and Related Species. Presented at the Meeting on Sources, Sinks,
and Concentrations of Carbon Monoxide and Methane in the Earth's
Environment. St. Petersburg Beach, Florida. August 15-17, 1972.
4. Weinstock, B. Carbon Monoxide: Residence Time in the Atmosphere.
Science. 166:224-225. 1969.
5. Hum, R. W. and T. W. Chamberlain. Fuels and Emissions--Update and
Outlook, 1974. Society of Automotive Engineers. Bureau of Mines,
Paper No. 740694. June, 1974.
6. Gay, B. W. Jr. and J. J. Bufalini. Nitric Acid and the Nitrogen
Balance of Irradiated Hydrocarbons in the Presence of Oxides of
Nitrogen. Environ. Sci. Technol. _5:422. 1971.
7, Spicer, C. W. and D. F. Miller. Nitrogen Balance in Smog Chamber
Studies. Paper No. 74-247. Presented at the 67th Annual Air
Pollution Control Association Meeting. Denver, Colorado. June 9-
13, 1974.
8. Miller, D. F. and C. W. Spicer. A continuous Analyzer for Detect-
ing Nitric Acid. Paper No. 74-17. Presented at the 67th Annual
Air Pollution Control Association Meeting. Denver, Colorado.
June 9-13, 1974.
9. Nicolet, M. Nitrogen Oxides in the Chemosphere. J. Geophys., 70:
679. 1965.
10. Dodge, M. C. and J. J. Bufalini. Advances in Chemistry. Series
113:232. 1972.
11. Dimitriades, B. and M. Whismen. Carbon Monoxide in Lower Atmosphere
Reactions. Environ. Sci. Technol. _5:219-222, 1971.
12. Demergian, K. L., J. A. Kerr, and J. G. Calvert. The Predicted
Effect of Carbon Monoxide on the Ozone Levels in Photochemical
Smog Systems. Environ. Letters. 3_:73-80. 1972.
13. Lee, R. E. Jr., S. S. Goranson, R. E. Envione and G. P. Morgan.
National Air Surveillance Cascade Impactor Network, II. Size
Distribution Measurements of Trace Metal Components. Science
178_:567-575. November 10, 1972.
184
-------
REFERENCES (continued
14. Durham, J. L., W. E. Wilson, K. Willeke and K. T. Whitby. Comparison
of Volume and Mass Distribution for Denver Aerosols. Presented
before the Division of Environmental Chemistry, American Chemical
Society, Los Angeles, California. April, 1974.
15. Lundgren, D. A. and H. J. Paulus. Direct Determination of Total
Atmospheric Aerosol Mass Distributing. Presented before the Divi-
sion of Environmental Chemistry. American Chemical Society, Los
Angeles, California. March 30-April 4, 1974.
16. Patterson, R. K. and J. Wagman. Mass and Composition of an Urban
Aerosol as a Function of Size for Several Visibility Levels. Seminar
10.
17, Schwartz, W. E., P. W. Jones, C. J. Riggle and D. F. Miller. Chemi-
cal Characterization of Model Aerosols. Final Report Grant
No. R-801174. EPA. 1974.
18. Friedlander, S. K. Characterization of Aerosol Distribution with
Respect to Size and Chemical Composition. J. Aerosol Sci. 1_:295-
307- 1971.
19. Miller, M. S., S. K. Friedlander and G. M. Hidy. A Chemical Element
Balance for the Pasadena Aerosol. J. Coll. Interface Sci. 59, No. 1.
April, 1972.
20. Friedlander, S. K. Chemical Element Balances and Identification of
Air Pollution Sources. Environ. Sci. Tech. 7^:235-240. 1973.
21. Gatz, C. R. Carbon Monoxide in the Atmosphere. Presented at:
Atmospheric Chemistry Mr Pollution Seminar. Oregon State University,
Corvallis, Oregon. 1972.
22. Heisler, S. R., S. K. Friedlander, R. B. Husar. Relation of Smog
Aerosol Size and Chemical Element Distributions to Source Character-
istics. Atmos. Environ. 7_: 633-649. 1973.
23. Roberts, J. J., E, S. Gobe, A. S. Kennedy. Urban Atmospheric Disper-
sion Model. Chem. Abst. 139828, Vol. 79. 1973.
24. Grosjean, D. and S. K. Friedlander. Gas--Particle Distribution Factors
for Organic Pollutants in the Los Angeles Atmosphere. Presented at
the 67th Annual Meeting of the Air Pollution Control Association.
Denver, Colorado. June 9-13, 1974.
25. O'Brien, E. F., V. P. Gutschick, V. McKoy, J. P. McTague. Polariz-
ability of Interacting Atmospheric Relation to Collison-Induced
Light Scattering and Dielectric Models. Phys. Rev. A. 8:690-696.
1973.
185
-------
REFERENCES (continued)
26. Miller, D. F., W. E. Schwartz, P. W. Jones, D. W. Joseph, C. W.
Spicer, C. J. Riggle, and A. Levy. Haze Formation: Its Nature
and Origin 1973, EPA Contract No. 68-02-0276 with Battelle-
Columbus Laboratories. Columbus, Ohio. June, 1973.
27. Schuetale, D., A. L. Crittenden, and R. J. Charlson. Application
of Computer Controlled High Resolution Mass Spectrometry to the
Analysis of Air Pollution. J. A. Poll. Con. Assoc. 23:704.
1973.
28. Schuetale, D., D. R. Cronn, A. L. Crittenden, and R. J. Charlson.
Molecular Composition of Secondary Aerosol and Its Origin. Pre-
sented at 172nd National American Chemical Society Meeting,
Symposium on Surface and Colloid Chemistry in Air Pollution Con-
trol. To be published in J. Colloid and Interface Science. 1974.
29. Wilson, W. E., D. F. Miller, and A. Levy. The Effect of Fuel
Composition on Atmospheric Aerosol Due to Auto Exhuast. J. Air
Poll. Con. Assoc. 23(11):949. 1973.
30. Hanst, P. L., J. Lehman, H. Reichle, M. Tepper. Remote Detection
of Air and Water Pollution from Satellites and/or Airplanes.
Man's Impact on Climate. W. F. Matthews, ed. MIT Press. Cam-
bridge, Massachusetts. 1974.
31. Brosset, C., K. Andreasson and M. Perm. The Nature and Possible
Origin of Acid Particles Observed at the Swedish Coast. Reported
by the Swedish Water and Air Pollution Research Laboratory. July,
1974.
32. Charlson, R. J. and N. C. Ahlquist. Apparatus and Method for Mak-
ing an Iii Situ Determination of the Chemical Properties of
Atmospheric Aerosols. U.S. 3,700,333. Battelle. Corp. October,
1972.
33. Cadle, R. D. and J. W. Powers. Some Aspects of Atmospheric Chemi-
cal Reactions of Atmoic Oxygen. Tellus 18^:176. 1966.
34. Davis, D. D. New Rate Measurements on the Reaction of 0( P), 0,
and OH. AIAA Paper No. 73-501.
35. Wood, W. P., A. W. Cartleman, Jr., and I. N. Tang. Mechanisms of
Aerosol Formation from SO™. Paper No. 74-153. Presented at the
67th Annual Meeting of the Air Pollution Control Association,
Denver, Colorado. June 9-13, 1974.
186
-------
REFERENCES (continued)
36. Want, C. C. and L. I. Davis, Jr. Measurement of Hydroxyl Concen-
trations in Air Using a Tunable UV Laser Beam. Phys. Rev. Lett.
32;349. 1974.
37. Davis, D. D., W. A. Payne, and L. J. Stief. The Hydroperoxyl
Radical in Atmospheric Chemical Dynamics: Reaction with Carbon
Monoxide. Sci. 17^:280-282. 1973.
38. Dodge, Marcia. Personal Communication. EPA Chemistry and Physics
Laboratory, Research Triangle Park, North Carolina 27711. Septem-
ber, 1974.
39. Cox, R. A. and S. A. Penkett. Aerosol Formation from Sulfur
Dioxide in the Presence of Ozone and Olefinic Hydrocarbons. J.
Chem. Soc. Faraday Transaction I, Part 9, 68_:1735. 1972.
40. McNelis, D. N. Aerosol Formation from Gas Phase Reactions of
Ozone and Olefin in Presence of Sulfur Dioxide. Ph.D. Disserta-
tion. University of North Carolina, Chapel Hill, North Carolina.
1974.
41. Matteson, J. M. and P. J. Giardina. Mass Transfer of Sulfur Dioxide
to Growing Droplets. Environ. Sci. Tech. 8^:50-55. 1974.
42. Hales, J. M. and S. L. Sutter. Solubility of Sulfur Dioxide in
Water at Low Concentrations. Atmos. Environ. 7_: 99 7-1001. 1973.
43. Johnstone, H. F. and P. W. Leppla. Solubility of Sulfur Dioxide at
Low Partial Pressures, lonization Constant and Heat of lonization of
Sulfurous Acid. J. Atmos. Chem. Soc. 56_:2233. 1934.
44. Monitoring and Air Quality Trends Report, 1973. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina 27711,
EPA-450/1-74-007. October, 1974.
45. Altshuller, A. P. Atmospheric Sulfur Dioxide and Sulfate. Environ.
Sci. Tech. 7;709-712. 1973.
46. Craig, N. L., A. B. Harker, and T. Navakov. Determination of
Chemical States of Sulfur in Ambient Pollution Aerosols by X-Ray
Photoelectron Spectroscopy. Atm. Environ. 8^:15-21. 1974.
47. Navokov, T., P. K. Mueller, A. E. Alcocer, and J. W. Otros. Chemical
Composition of Pasadena Aerosol. J. Colloid Interface Sci. 39:225-
235. 1972.
48. Taarit, Y. B. and J. H. Lunsford. Electron Paramagnetic Resonance
Evidence for the Formation of S0~ by the Oxidation of SO" on MgO.
J. Phys. Chem. 77:1365-67. 1973. ^
187
-------
REFERENCES (continued)
49. Rao, K. V. S. and J. H. Lunsford. EPA Grant 801136.
50. Spedding, D. J. Uptake of Sulfur Dioxide by Barley Leaves at Low
Sulfur Dioxide Concentrations. Nature. 224:1229. 1969.
51. Shepard, J. G. Measurements of the Direct Deposition of Sulfur
Dioxide onto Grass and Water by the Profile Method. Atmos.
Environ. 8_:69-74. 1974.
52. Garland, J., D. Atkins, C. Readings, and S. Caughey. Deposition
of Gaseous Sulfur Dioxide to the Ground. Atmos. Environ. 8_:
75-79. 1974.
53. Owers, M. and A. Powell. Deposition-Velocity of Sulfur Dioxide
on Land and Water Surfaces Using a S Tracer Method. Atmos.
Environ. 8_:63-67. 1974.
54. Sydberger, T. and N, Vannerberg. The Influence of the Relative
Humidity and Corrosion Products on the Adsorption of Sulfur
Dioxide on Metal Surfaces. Corros. Sci. 12^:775-780. 1972.
55. Duncan, J. R. and D. J. Spedding. The Effect of Relative Humidity
on Adsorption of Sulfur Dioxide on to Metal Surfaces. Corros.
Sci. 13:993. 1973.
56. U.S. Environmental Protection Agency. Air Quality Criteria for
Sulfur Oxides, Appendix A. (First Working Draft). National
Environmental Research Center, Research Triangle Park, North
Carolina 27711. July, 1972.
57. Larsen, Ralph I. An Air Quality Data Analysis System for Inter-
relating Effects, Standards and Needed Source Reduction. J. Air
Poll. Con. Assoc. 213:933-940. 1973.
58. Larsen, Ralph I. An Air Quality Data Analysis System for Inter-
relating Effects, Standards and Needed Source Reductions, Part 2,
J. Air Poll. Con. Assoc. 24_:551-558. 1974.
59. Lin, J. T., H. T. Liu, and Y. H. Pao. Laboratory Simulation of
Plume Dispersion in Stably Stratified Flows over a Complex Ter-
rain. Final Report under Contract 68-02-0800 with EPA. Flow
Research Report No. 29. Kent, Washington. February, 1974.
60. Schiermeier, F. A. and L. E. Niemeyer. Large Power Plant Effluent
Study (LAPPES). Vol. 1 - Instrumentation, Procedures and Data
Tabulations APTD 70-2, National Air Pollution Control Administra-
tion, Raleigh, North Carolina. 410 p. June, 1970.
188
-------
REFERENCES (continued)
61. Schiermeier, F. A. Large Power Plant Effluent Study (LAPPES).
Vol. 2, Instrumentation, Procedures and Data Tabulations (1967
and 1969), Environmental Protection Agency, APTD 0589. Research
Triangle Park, North Carolina 27711. 337 p. 1970.
62. Schiermeier, F. A. Large Power Plant Effluent Study (LAPPES).
Vol. 3, Instrumentation, Procedures and Data Tabulations (1970),
Environmental Protection Agency, APTD 0735. Research Triangle
Park, North Carolina 27711. 296 p. 1972.
63. Schiermeier, F. A. Large Power Plant Effluent Study (LAPPES),
Vol. 4, Instrumentation, Procedures and Data Tabulations (1971)
and Project Summary. Environmental Protection Agency, Research
Triangle Park, North Carolina. 286 p. November, 1972.
64. Start, G. E., C. R. Dickson, and L. L. Wendell. Diffusion in a
Canyon within Rough Mountainous Terrain, NOAA Tech. Memo. ERLTM-
ARL28, Air Resources Laboratories, Idaho Falls, Idaho. 1973.
65. Clarke, J. F. and K. F. Zeller. Tracer Study of Dispersion from a
Highway, Meterology Laboratory, Research Triangle Park, North
Carolina 27711. AMS Symposium on Atmospheric Diffusion and Air
Pollution, Santa Barbara, California. September 9-13, 1974.
66. Thayer, S. D., D. J. Pelton, G. H. Stradsklev, and B. D. Weaver.
Model Verification - Aircraft Emissions Impact on Air Quality,
Final Report, EPA Contract 68-02-0665. May, 1974.
67. Platt, M., R. C. Baker, E. K.. Bostress, K. M. Chug, and R. D.
Siegel. The Potential Impact of Aircraft Emissions on Air Quality.
Final Report Numbers 1167-1 and 2 (Vols. 1 and 2), Northern
Research and Engineering Corporation, Cambridge, Massachusetts,
NTIS No. PB-208-950. 1971.
68. MacWaters, J. T., R. C. Koch, and S. D. Thayer. Air Quality
Measured at Atlanta International Airport Before and During Experi-
mental Aircraft Taxiing Gperatings, Final Report, EPA Contract 68-
03-0358. June, 1974.
69. Hales, J. M. Fundamentals of the Theory of Gas Scavenging by Rain.
Atmos. Environ. 6_: 635-359. 1972.
70. Dana, M. T., J. M. Hales, W. G. N. Slinn, and M. A. Wolf. Natural
Precipitation Washout of Sulfur Compounds from Plumes. Ecological
Research Series, EPA-R3-73-047, Environmental Protection Agency,
Research Triangle Park, North Carolina 27711, 204 p. 1973.
189
-------
REFERENCES (continued)
71. Dana, M. T., T. M. Hales, C. E. Hane, and J. M. Thorpe. Precipi-
tation Scavenging of Inorganic Pollutants from Metropolitan
Sources. 1974.
72. Gartrell, F. E., F. W. Thomas, and S. P. Carpenter. Atmospheric
Oxidation of S09 in Coal-Burning Power Plant Plumes. Amer. Indus.
Hyg. Assoc. J. ZZ4:113. 1963.
73. Manowitz, B., L. Newman, and W. D. Tucker. The Atmospheric Diag-
nostics Program at Brookhaven National Laboratory: Fourth Status
Report. BNL 50361 (Meteorology - TID-4500). 1972.
74. Fankhauser, R. 1971 Summer Study of Ozone at 33 Cities, Quality
Assurance and Environmental Monitoring Laboratory, Environmental
Protection Agency, Research Triangle Park, North Carolina 27711.
1974.
190
-------
5. EFFECTS
5.1 HEALTH
5.1.1 General Discussion
The Clean Air Act amendments of 1970 mandated that the Administrator
of the U.S. Environmental Protection Agency promulgate ambient air
quality standards for substances that may be emitted into the atmosphere
and thereby represent potential health hazards.
To fulfill the mandate, standards were established which in com-
pliance with the law were designed on the basis of the best information
available to include a margin of safety sufficient to assure protection
of the most sensitive segments of our population. Because the informa-
tion base on which standards were established was incomplete, the agency
committed itself to a continuing research program designed to determine
their adequacy.
As results from this and other research efforts around the world
become available, it is necessary to reappraise and update the health
basis for the standards, as is done in this document.
5.1.2 Nitrogen Oxides
5.1.2.1 Human Studies--The National primary standard for nitrogen
dioxide is 100 yg/m (0.05 ppm) annual arithmetic mean.
5.1.2.1.1 Evidence upon which standard was based--The existing primary
N09 standard of 100 yg/m was based primarily on the Chattanooga School
Children studies ' ' conducted from November, 1968 through April,
1969. Long-term exposure to ambient N0~ concentrations between 117 and
205 yg/m (0.062 and 0.109 ppm) accompanied by suspended nitrate
191
-------
concentrations between 4 and 6 yg/m for six months and longer was
associated with the following effects: (1) an 18.8% excess in acute
respiratory illness in families; (2) significantly decreased lung func-
tion (FEVn _,.) in elementary school females; and (3) significantly
increased bronchitis morbidity in elementary school children exposed for
two and three years.
5.1.2.1.2 Interim evidence (1969 through January 1, 1973)--Estimates of
community exposure to NO- in the initial Chattanooga study were derived
from air monitoring data obtained with gas bubblers employing the 24-hour
Jacobs Hochheiser method. Since the time of the study, the Jacobs Hoch-
heiser method has been found unreliable, creating the need to reevaluate
the relationship between quantitative exposure level and observed adverse
health effects in Chattanooga using data obtained with other methods of
monitoring.
Alternate exposure estimates were derived from the network air
monitoring stations established by the Division of Abatement, National
Air Pollution Control Administration (U.S. Public Health Service) in the
Chattanooga, Tennessee - Rossville, Georgia interstate area. Continuous
recordings of NCL were taken, utilizing the colorimetric Saltzman tech-
nique at ten stations from December, 1967, through November, 1968. A
complete report of data obtained from this network is given in the pub-
lication Chattanooga, Tennessee - Rossville, Georgia Interstate Air
Quality Study 1967-1968 (U.S. Department of Health, Education and Wel-
fare, National Air Pollution Control Administration Publication No. APTD-
0583, 1970).4
Observed NCL concentrations, wind speed and direction and topo-
graphical characteristics were used to obtain NCu exposure. Although
192
-------
health data were collected from November, 1968, through April, 1969, in
the initial Chattanooga study (the six months immediately after the
period of monitoring by the Division of Abatement), NCL exposures of
the affected communities probably did not change substantially from
those measured from December, 1967, through November, 1968, because TNT
production by the Volunteer Army Ammunition Plant remained at peak
capacity throughout these months of 1968 and 1969.
In the original Chattanooga health studies, the frequency of sam-
pling for NCL exposures using the now discredited Federal Reference
Method allowed dissociations of adverse health effects only with daily
or annual average NCL concentrations. Air monitoring data from the
national abatement air monitoring network using the continuous Saltzman
NCL method provided additional information on hourly and daily frequency
distributions of NCL concentrations. Experimental animal studies sug-
gested that short-term peak exposures lasting for two to three hours
were sufficient to account for excesses in acute respiratory diseases.
Therefore, the 90th percentile of the frequency distribution observed
for hourly concentrations measured by the Saltzman method was selected
as a more appropriate estimate of NCL exposure for relating to the
observed adverse health effect in the original Chattanooga study.
Ideally one would like to look at maximum hourly distributions as well.
An excess of acute respiratory illness among families was observed con-
sistently when the 90th percentile of hourly NCL concentrations was
376 yg/m (.20 ppm) or above. When similar hour exposures were 188 to
282 yg/m (0.10 to 0.15 ppm), the evidence was not consistent. In the
study on bronchitis rates in children, a measure of more severe
193
-------
respiratory disease, children exposed for two or three years to peak
hourly N02 concentrations of 188 to 282 yg/m3 (0.10 to 0.15 ppm) had
significantly higher bronchitis rates than children in the lowest
exposure neighborhood.
5.1.2.1.3 New Evidence (Since January, 1973) Recent Nitrogen Dioxide
Exposure in Chattanooga, Tennessee--Since the initial CHESS study in
Chattanooga, Tennessee, a vigorous pollution control program by the
Department of Defense and a decrease in TNT production has been accom-
panied by a decrease in NCL emissions, evidenced by a decrease in the
mean of the annual average levels of NCL in the most exposed study
neighborhoods from 282 yg/m3 in 1968 to 56 yg/m3 in 1973. (Table 5.1)
Despite these lower annual NCL levels, the population living in the
community in 1972 with the highest annual average exposure to NCL con-
tinued to experience repeated short-term exposure to NCL as measured by
the 90th percentile of average daily concentrations (75 to 220 vg/m ).
(Table 5.2) The 90th percentile of the maximum hourly NCL values ranged
from 228 to 815 yg/m3 in 1972 (Table 5.3). In 1968 and in 1971, the
neighborhoods closest to the plant also were exposed to relatively high
levels of suspended nitrates (5.0-6.0 yg/m ). In the intermediate
exposure community, the mean of the annual average NCL levels also fell
(113 yg/m3 in 1968 to 49 yg/m3 in 1973 - Table 5.1). Short-term exposures
in this community were decreased from former high levels, but still the
90th percentile of the maximum hourly values of about 116 to 164 yg/m
were noted on 36 days each year (Table 5.3). In the low exposure area, the
mean of the annual average NCL exposures were also reduced from 56 down
194
-------
w
[J3
(T!
pTJ
*3"
2 CM
rH t-~-
0>
CO rH
J 1
W CO
,> \o
as
S w
3 W
CO CO
O CO
OH ty
S ip
1^
— ' O
i-3 O
2^
^£
^j^ ^£
1"*") ^H
E^ C3
§r
^ 2
i — i
83
W £4
^> <^
<
>-l
Q e
£P
SS co
r>-l
i — i
H
CO
w
rH
*
LO
CD
rH
•8
H
13
cO 0
n3 bO
g co
X 3 5-i
P P 0
•H CO >
rH rH <
SlPrH
O C3X eg cti
*rH ^ jj
•P rH -H S
^•^£l
o ^)
• H fi
+-> a
•H
^-4
<^
v — '
to
b?
3.
rH «
3 O
§ P
5n
CD P
tiO C-J
CO 0
5n O
0 a
> ^0
t^
CT>
rH
rH
CT)
rH
O
t~^
CJl
rH
CO
vO
a>
rH
•K
bp
pi
X
f^
TO
""d Od
ro 0
^«|
JH ^3
o a
O ?s
CO W
i
3
rH
rH
0
OH
rf_
6
bo
3.
O
o
rH
* * *
* * *
* * *
tO CTl vO
•JC
* *
LO LO CM
^ ^ LO
•K *
* *
LO LO tO
•*•«*•*
*
CM
co
CM
•X -X I
vO to CO
LO rH LO
rH rH
0
•H
*"O
0
1 J,
0 bo
•P -H
> rt x
O rH 1
i-J 1 KH
1 rH rH
h- 1 rH I-H
0 0
bOT3
O -H
5n X
P 0
•H -H
2 Q
0
O
<<<
2,§S
c^-j to vO
M csi ro
C4 tO vO
Q^
I t
cd
•H
rO
(p
g 42
0 bfl
+-> -H
B5K
,-J 1 rH
1 1— 1 rH
r~l rH rH
Is
li
to -p
3 -H
00 2
0
a
0
< < <
§§§
O 0 0
rH rH rH
to to to
rH rH rH
CM O O
rH rH rH
0
P
CO
rO
0
S rC
0 M
P -H
O i — i i
1 rH t— 1
rH 1 — 1 1— 1
1 S
C P
0 cO
P.MH
CO CO
to
bo
p.
LO
CO CM rH
vO \O LO
•^- CM CM
vO vO LO
•=*• tO rH
\iO vO LO
CM CM rH
\o r~> co
0
•H
rO
CD
P rC
0 bO
P -H
O rH 1
_J 1 rH
1 rH rH
1 — II — 1 rH
*
* t/)
* 0
* P
T^ CO
0 rH
T) 3
rH 0 -H
CO P.P
P t/3 M
0 3 cO
H CO PH
cO
0
cO
§
IS)
o
f^
0
f-j
bO
•H
O C
£_<
P TJ
000
5n 0 0
O O
rH JD t/)
O O 0
""d 'H
S f^ S
S 5H rH
NI P -H
p e
co o rd c3
CO 5n O 0
1 1 i c
i i
tJ 0
0 0 CO -H
P P P 5-4
cO cO cO P
Q S O 0
CM -P CM O
O (/) O 0
2 W 2 U
* * * *
* * *
•K *
*
195
-------
CO
CXI
LO
CD
c/)
rH
rt
r
0
X
rH
•H
cd
Q
•H
CD
rH
99th
Percent]
CD
rH
•H
O CD
CTi U
CD
OH
CD
rH bO
Oj rt
g CD
rt
CD
rl
CD
z;
o
rt
p
CO
rH
CD
P rQ
3 E Z
&j CD
Z CTJ 0
03
p
CO
P CD
•H J-i
g 0
Is-
u w
O CTl OO
OO LO O
OO tO CX]
rH
vo to LO
\D to OO
<* tO rH
rH
rH O OO
CX] CX] rH
CTi CX] i — 1
LO rH O
CTi CTl vO
OO CXI IO
vO t^-- ["*"•••
O"} O^ O"i
rH rH rH
rH
to
rH
CX] CXI LO tO \O
OO rH rH rH rH
vO CO t — *^- \O
CX] CX] *3- O CO
LO rH i— 1 rH ,— I
CX] OO rH LO •xt
CO \*O CTi t^^ vO
CX]
CD rH t~~- VO ^j"
LO *x1" LO ^- ^d"
rH
«
*
OO O rH (Nl tO
rH rH rH i— 1 rH
cxi
to
LO
rH
tO vO tO "^ ^D
o to \o i — i r~~-
CX] Cxi rH rH rH
rH
CTl O O 1\ CXI
LO O LO O CX]
CTl CX] rH rH rH
CTl LO -vt ^O tO
OO LO ^Sr CTi CTi
OO LO LO «* tO
rH
•K
•K
OO CD rH CX) tO
\& c^ c^^ r^ r^~
CTi CTl CTi CTl CTl
rH rH rH rH rH
to
to
rH
rH OO O
rH f~- CTl
I-- rH
rH
o to t>-
i~; rH
rH
C~^ O LO
f^- C^*' LO
CX] t-- rH
OO tO tO
CXI
*
*
OO CXI tO
CTi CTi CTi
rH i— 1 i— 1
•xf
to
O
CX]
00
•H
CX] rH rH
LO CTi CTi
*3" Cxi vO
rH OO OO
O •xt CX]
cx] vO O
to
O rH i— I
LO •<*• *3"
rH
*
*
OO CXI tO
CTi CTi CTl
rH rH rH
i— 1
CX]
rH
rH
CX] CX]
1 -=t CX]
rH rH
rH CX]
1 rH rH
rH rH
00 CTl
I t~- 00
^O to
l uo vO
•K
*
OO cx] tO
CTl CTl CTl
rH rH rH
cxi
CX]
1
V CO
p p
c cd
rH -H
196
-------
vO O
LO O
t~- rH
1 vO OO
t-O r--
**• LO
*
•K
00 CXJ l-O
CT»
cr>
.s
p
§
U
(Nl
LO
•s
H
OS 0
CD f-i
S o
P
tfl -H
g g
•H M
o3 rxi
>H . r-
P t) CT>
P! Q I-H
o ^
u p m
ss °
o
0
rH O
r-~ P!
01 S
rH O
t/)
T3 0)
S £
o 3
r-- rH
cr> •H
rH S
"^
oo U
o
01 X
rH &
•K
Crt
I
P
to
o3
X
rH
s
•K
•K
197
-------
to 43 yg/m (Table 5.1). Peak hourly exposures on all but a few days were
much like those of the intermediate exposure community.
This set of circumstances provided the opportunity to assess whether
or not individuals experiencing a decrease in annual average exposure to
NO- levels, but still exposed to intermittent high levels of N0~, would
Li L*
continue to experience significant adverse health effects.
Prevalence of Chronic Respiratory Symptoms in Chattanooga 1971
The frequency of chronic respiratory disease symptoms was increased
during or right after the previously mentioned high exposure period
(1968-1970) only in the most heavily impacted area where increases in
symptom frequency, which are not statistically significant, were observed
three years after the high annual N0~ exposures were reduced 43 to 56
yg/m . However, peak short-term NCL exposures continued. This increase
in symptoms might represent a residual effect from previous excessive
long-term exposures or it might indicate that prior high exposures
coupled with peak current exposures are sufficient to trigger the onset
of chronic respiratory disease symptoms. The presence of elevated
levels of suspended nitrates may also be a contributing factor.
Lower Respiratory Disease and Nitrogen Dioxide Retrospective Survey in
7
Chattanooga, Tennessee, 1971
Acute lower respiratory disease morbidity in the period 1969-71 was
surveyed retrospectively among children one to 12 years of age from the
three original CHESS communities representing an exposure gradient for
NCL (Table 5.4). Lower respiratory disease morbidity rates among children
of residential families residing in areas exposed to high and inter-
mediate levels of nitrogen dioxide continued to be significantly higher
198
-------
g
I-H
H
<
H
CO
<>
^
t— 1 *
£2! t"O
^3 r*-- /•— \
2 1 r<
2 O 0
Br~~ P
CTi ,0
rH r>\
>H
pq « O
W -H
CO W ,£>
S83
i — i H
H § rH
§3 H PH
b ^
pq - S
2: S f-i
O Q W>
u o o
c^S
O EH 'H
^J. r ! ^T1
gb
0
rC
*Sj^
3
1— 1
§
•
to
•
LO
0
rH
•3
H
i — \
to
e
v^_>
CD
• H
^
o
•H
O
0
z
P
•H
^"
^
3
r^
'd
S
0
rH
•rH
X P
P CH
CTi 0)
a> o
^
0
PH
0
rH
•H
^-t I 1
4-J CJ
O 0)
CT) O
0
OH
O
•H 0
rH P bQ
rt 0 cs
§ J 0
S P >
'rl
"^
JH
Oj
0
a
CO O r4
CO -H 0)
rr"! ^j o
S CO g
CO 2
P
P! Pi
0 0 JH
g -H 0)
0 P -Q
rd CO 2
P 0)
•H rH
C 3
S O
S r^4
O ><
\o f^*
CO t —
CTl CTi
LO <*
O vO
a> oo
LO CXI
rH CXI
00 LO
\o o
^ ^
cxi to
r~- t~^-
o o
rH
O}
rH
^* o o\ cxi
O ^ vO LO
LO OS vO C^^
oo o o oo
OO CXI tO CTl
^ LO LO tO
LO o oo r^
CXI t-- CXI LO
CX] CXI (XI t-H
vO CXI O tO
O LO rH O
rH rH rH rH
O rH CXI tO
f^^ f~^. r^. f^*.
< <] 0 O
CXI
to
LO
rH
1
CT^ OO O^ CXI
o cxi cr> LO
rH rH LO t~^
rH rH
O O CT> LO
CT) CT> TJ- LO
fp vO LO V«O
rH
o o •rf LO
LO LO O tO
•53- tO tO CX!
CXI LO •* t-^
CXI \o vO cxi
(XI rH rH rH
O rH CXI tO
<
c^-^ \o
CXI rH
to t~-
CXI OO
i—t
cxi to
t"^- r^ —
0 0
T^-
to
0
(XI
r~*
GO
•H
^- o
LO O
r-H f->
O OO
LO OO
rH rH
\o *^
rH CXI
rH rH
t~*-~ C~"-~
r^ oo
cxi to
0 0
rH
(XI
rH
^O
rH
O OO
r-^ to
CXI N-)
LO OO
CXI CXI
to «*•
^~ vO
rH rH
LO tO
O rH
rH rH
CXI tO
r*~~ r~^
o o
CXI
CXI
1
0
£^
0
.{_>
pj
i— i
t-^ v£)
OO rH
rH CX)
CTl tO
LO OO
rH rH
to o
rH LO
rH rH
LO LO
OO O
rH
CXI K)
0 0
rH
"*
^
O
^
P
c
0
o
t/)
0
y~j
•H
S rt
•H i
•H N
S P
M ^
O CO
CO CO
CO CO
LO LlJ
hn rTJ
QJ ^^
0 <
199
-------
Table 5.4. THREE-YEAR FREQUENCY OF "ANY LOWER RESPIRATORY DISEASE1
AND BRONCHITIS: MODEL PREDICTED RATES FOR CHILDREN
AGED ONE TO 12 YEARS FROM HOUSEHOLDS WITH A HIGH
SCHOOL EDUCATION OR GREATER7
(1969-1971)
Morbidity
Condition
"Any Lower
Respiratory
Disease"
Bronchitis
Number of
Episodes
One
or
More
Two
or
More
One
or
More
Two
or
More
Male
Female
Male
Female
Male
Female
Model
I -Low II
28.
28.
14.
12.
9.
4.
4.
3
0
0
2
8
4
2
Predicted* Rates, \
- intermediate+
28.
30.
17.
13.
16.
8.
9.
9
9
0
9
1
8
8
III -High
38.
36.
20.
19.
20.
10.
9.
5
5
2
8
0
5
6
High/ Low
Community
(p< 0.01)
1.
1.
1.
1.
2.
2.
2.
36
30
44
62
04
39
29
*Age, education head of household adjusted rates for nonasthmatic chil-
dren with three or more years residence duration.
+Intermediate exposure community significantly higher than low exposure
community in 5/8 sex by number of episode categories but significantly
lower than high exposure community in only 2/8 categories.
than those of comparable children living in a low exposure area. Fur-
thermore, lower respiratory illness rates in intermediate exposure
neighborhoods were significantly lower than rates in the high exposure
community.
Because this study was not designed to determine yearly rates, it
was not possible to assess whether the significantly higher rates in the
200
-------
intermediate- and high-exposure communities were consistent over the
three study years or were heavily weighted by the single year, 1969,
when annual mean NO- levels were above the primary standard.
Incidence of Acute Respiratory Disease, Chattanooga, Tennessee, 1972-
19738
A study of acute respiratory disease was conducted in 1972 and 1973
f
in the three Chattanooga CHESS communities. In the 1972 study the excess
risk of acute respiratory disease remained consistently higher for those
living in the high-pollution community compared to those in the low-
pollution community. (Table 5.5) As was noted in an earlier study, the
highest excess risk (50%) was observed in pre-school children. The
intermediate exposure community also showed excess risk of lower
respiratory disease compared to the low-pollution community. Moreover,
excess risk for total respiratory disease was not as great in the inter-
mediate community as in the high-pollution community. Pre-school chil-
dren living in the intermediate exposure area had a lower excess risk
than other age groups in that community. In the second year of study,
the pattern persisted across the three communities, but the excess risks
themselves were lower than the preceding year.
The excess risks of acute respiratory disease in the 1972 study
were similar to those found in the earlier 1968-69 Chattanooga study.
Although the annual average NO- levels in the initial study were sub-
stantially higher than those in the 1972 study, short-term 90th per-
centile hourly NO- exposures associated with an effect were similar in
the two studies (^300 yg/m in the high-exposure area in the 1972 study).
201
-------
LO
LO
•§
H
•°3
CL)
c
rH
rH
1 — 1
X
O
rH
•H
O-
(D
PH
CD
P
•-*
3
<^
U)
0)
c/)
0)
rH
1 j
rH
1—1
O
"rrt
rH
•rH
t/)
r?
P-I
rH
^^
-P
O
CT3
H
*_,
r-*
(P
o
w>
a
•H
rH
O
>
r— 1
P
u
aj
rH
f-H
(D
p<
a
00
•H
rH
0
r— 1
X
I— 1
§
Co
OH
1 — i
I~H
X
-o
3
p
— 1
•d
3
p
'A
rH
1
X
3
P
— i
X
P
CO
rH
—\
X
r^
P
1 — 1
^
P
CO
o
p
u
0>
CO
p
Pi
§
i.
00
CO
rH
•ct
O O
to
I
rH
-------
(Table 5.6) At this time it is not know to what extent those families
living in the neighborhood experiencing the highest short-term exposures
of NC>2 as well as elevated levels of suspended nitrates contributed to
excess risk in the total high-exposure community.
That persistence of excess risk for acute respiratory disease four
years after long-term exposure has decreased may reflect residual
effects from previous prolonged exposure, the effects of intermittent
high peak exposures or both these factors. Findings for pre-school
children in the 1972-73 study suggest the latter may be the case. The
pre-school children in the high-pollution community had limited pro-
longed exposure to high annual average levels of NO- but did have short-
term high exposure, yet these children showed the highest excess risks
in their community. Their counterparts in the intermediate community
had the lowest risk. However, excess risk persisted in the other age
groups living in the intermediate exposure community, suggesting a
residual effect from previous annual average exposure to NO- of 150
yg/m , (Table 5.6) or 90th percentile hourly exposure of ^300 yg/m .
Ventilatory Function in School Children in Chattanooga, Tennessee 1971
Studies showed ventilatory function of children in Chattanooga
was no longer impaired two years after annual exposures were reduced
from vL50-282 yg/m to 43 yg/m even though short-term peak hourly
exposures continued to range from 228 to 815 yg/m for 10% of the days
each year.
q
Asthma Panel Studies
In recent unpublished studies the EPA CHESS program in the New
York-New Jersey metropolitan area, increased asthmatic attacks were
203
-------
OO
•K
CO
2;
o
b
H
CO
C^3
2i
I-H
O
H
tH
2:
i— i
£3
rH
rrj
£>
rH
rH
3
o
re
i
•9
jn
^?
0
rH
•H
nfj ^_J
+-* ci
CTi ^)
CTi O
QJ
OH
0
i
•H
-G 4-^
+-» c
O CD
CTi O
0)
OH
0)
rt nJ
C 0^
1^
s_(
rt
d)
^>H
1
CO §
co 2;
O o
•H
P
rt
CO
0)
e S
< 5 3
u e 2:
tO
t LO VO
^
•X
OO (X] tO
vo t^^ r~^
CTl CTl CTi
rH rH rH
CM
CM
1
l
i
0) 0)
P -p
c! rt
r— 1 -H
204
-------
OO VO
OO rH
rH CXI
§
(/)
CO
t-O (X)
(ft
c
o
•H
P
rt
01
§
g
(J
CTi
CX]
t^~
K)
IS)
*
*
OO CXI
o t-~.
Oi O1
§
O
P
CX]
t^
Oi
§
O
LO
0)
rH
•a
^
^
P
VI
CO
X
205
-------
significantly associated with elevated levels of suspended nitrates when
temperatures were 50°F or above. No such effect was observed with
NCL. Chemiluminescence instruments showed NCL levels met the annual
average standard at a central city station but only the old (now dis-
carded) Federal Reference Method was used for measurement of NCL in
study neighborhoods themselves.
In another study conducted by the same group in to Southeast com-
munities, there was evidence of excess risk of asthmatic attacks asso-
ciated with elevated levels of suspended nitrates, but the findings were
less consistent than those observed in the New York study. Because of
problems in present methods to measure and analyze suspended nitrates,
findings in these studies should be viewed qualitatively rather than
quantitatively. These two studies raise a question as to levels of NCL
which might be necessary to prevent levels of transformation products,
such as nitrates, which in themselves may contribute to adverse health
effects.
Preliminary analyses of California asthma panel studies show no
easily demonstrable increase in attack rates occurs with NCL exposures
ranging up to 400 yg/m for one hour and 200 yg/m for 24 hours. How-
ever, effects of other stronger illness determinants, like aerosols
containing sulfates and nitrates, had not been disentangled at the time
of the present summary.
Indoor No,-, Studies
Preliminary measurements by EPA of NCL indoor air concentrations
show peak hourly levels in the range of 470-940 ug/m (0.25 - 0.50 ppi
in the heating zone of persons cooking with gas stoves. The highest
206
-------
peak indoor NCL concentrations measured was about 1 ppra. EPA's CHESS
studies suggest a greater relative risk of acute lower respiratory
illness among occupants of homes with gas stoves. Mothers were the only
family segment significantly affected. Presumably this excess could be
related to periodic hourly NO,, exposures in the 0.25-1.0 ppm range.
Li
Followup studies in CHESS, however, have not confirmed a relationship
between air pollution due to gas stoves or gas space heaters and lower
respiratory disease in mothers. A study in Ohio also did not find any
respiratory disease health effect related to gas cooking. In all three
health studies, measurements of indoor N07 air pollution were not
LJ
related directly to health. Like EPA, those investigators also found
increased peak indoor NO-., levels, generally about 470 yg/m (0.5 ppm)
but at times as high as 940 Mg/m (1.0 ppm), to be associated with gas
cooking. Thus, there are still questions as to whether or not indoor
hourly NO- air pollution in the range of 0.25-1.0 ppm is related to
increased risk of respiratory disease. Such a relationship would be
qualitatively consistent with data in experimental animals showing an
effect of short-term N02 exposure on resistance to infection.
12
5.1.2.1.4 Other new evidence--Von Nieding and associates studied the
acute effects of experimental NO- exposures on healthy and diseased
subjects using pulmonary ventilation, diffusion and perfusion as end-
points. Exposure of 16 healthy subjects to 9400 Mg/m (5 ppm) No~ for
15 minutes caused a significant decrease in diffusing capacity. Four-
teen subjects with chronic respiratory disease showed a significant
decrease in arterial partial pressure of oxygen and a significant
increase in the alveolar-arterial oxygen pressure gradient upon exposure
207
-------
to 7520 or 9400 yg/m for 60 minutes while no effect was observed at NCL
concentrations of 3760 yg/m (2 ppm). Healthy subjects (n=39) and those
with chronic respiratory disease (n=70) were exposed to 30 breaths of
NCL at concentrations over a range of 1880 to 9400 yg/m (1 to 5 ppm).
These subjects demonstrated increased airway resistance at NCL concen-
trations of 3000 to 3800 yg/m (1.6 to 2.0 ppm). At lower concentra-
tions, NCL exerted no effect, but airway resistance markedly increased
as concentrations exceeded 3800 yg/m (1.6 to 2.0 ppm).
Kennedy studied a group of 100 coal miners with a history of
exposure to nitrous fumes from shotfiring in the mine. Eighty-four of
the mean had emphysema based on a residual volume of more than 150%.
The 16 miners with a residual volume less than 150% of normal had less
exposure to fumes than the remaining 84 men. Kennedy compared the pul-
monary function of the miners exposed to nitrous fumes with non-exposed
miners and those with simple pneumooconesis and found the indirect maxi-
mum breathing capacity, vital capacity, total lung capacity and residual
volumes showed a markedly higher percentage of abnormal values in the
nitrous-exposed group. The primary and secondary charges of the hydrox
shells used in the mines were largely nitrates and fume analysis revealed
(NO and NCL) up to 88 ppm after normal firing and up to 167 ppm after
misfires.
A full report of the study of chronic bronchitis among housewives
aged 30 or more living in six localities throughout Japan conducted dur-
ing the winter of 1970-71 is not yet available in the U.S., but a summary
14
of this study has been made available. This is the study on which the
208
-------
Japanese standard of 0.02 ppm 24-hour annual average is based. The
prevalence rate of chronic bronchitis was found to exceed 5% only in
those areas where the annual NCL concentration was 0.042 ppm (79 yg/m )
or above. However, total suspended particulate concentrations were also
very high, ranging from twice to six times the U.S. primary standard
indicating that the pollutant-disease association in the study cannot be
ascribed to NCL alone.
In a study by Speizer et al., * ' the prevalence of chronic non-
specific respiratory disease was assessed using a modified version of
the Medical Research Council's respiratory disease questionnaire in a
population of 268 policemen exposed to a spectrum of levels of automo-
bile exhaust as traffic officers, patrol car officers and indoor cleri-
cal officers. The over-all prevalence of disease in the policeman
population was 36.9%, and two--thirds of chronic conditions were chronic
bronchitis. There was a clear relationship between smokers and the
prevalence of disease. In nonsmokers and current smokers, men with
greater duration of exposure In traffic tended to have higher disease
prevalence rates, but the difference was not statistically significant.
Measurement for exposure to automobile exhaust was limited, since only
16 work stations out of a total of 58 were chosen for study, and sam-
pling at each station was conducted for only two days during each season
of the year. The NCL values ranged from a minimum of 0. 005 ppm to
0.090 ppm while carbon monoxide ranged from 1 ppm to 44 ppm and hydro-
carbons from 4.4 to 15.8 ppm.
1 8
In a recent preliminary report, Sprey correlated death rates
from hypertensive and arteriosclerotic heart disease and lung cancer
209
-------
with annual NCL concentrations. The author's data source was the
National Air Surveillance Network which operated a single station in
each of 42 metropolitan areas. NCL was measured once each two weeks
by the Jacobs Hochheiser method. The twenty-six values for the year
1962 were averaged and were correlated with cause-specific death rates
for the years 1959-1961.
The results showed a linear relationship with annual NCL concen-
trations for 1962. Higher death rates were found even in areas with
NCL concentrations below the primary national standard, when compared
with the lowest NCL areas (56 yg/m , .03 ppm). Adjustments for area
differences in family income and education, population density, tempera-
ture and sulfate and nitrate concentrations did not affect the basic
association between NO- and death rates. A similar linear relationship
was shown for lung cancer deaths among males and females over 65 years
and NCL concentrations in the 42 areas. The relationship was not
weakened by adjustments for area differences in SCL, sulfates, nitrates,
climate, socioeconomic factors and population density.
The quantitative basis of NCL exposures in Sprey's work is open to
question because of the limited number of air monitoring stations used
as well as the unreliability of the measurement technique. However, the
qualitative nature of the work should draw attention to possible extra
pulmonary systemic effects of ambient NCL.
5.1.2.1.5 General discussion of NO,, Human Studies--Studies of NCL and
human health effects are limited, and many conducted thus far are con-
founded by the presence of high levels of other pollutants. In addition,
measurement data used in these studies were limited either by use of too
210
-------
few sampling stations or the use of an unreliable method for measurement
of N02 or both of these factors.
Five categories of human health effects have been identified in
association with NCL: acute toxic, often fatal, reactions, diminished
lung function, structural and bronchial changes of lung most likely to
promote development of chronic respiratory disease, susceptibility to
infection and systemic effects.
There is sufficient information from both human and animal studies
to determine the levels of N02 which may contribute to toxic, often
fatal, reactions. There is some evidence to show that prolonged
exposure to NCL levels of 117 yg/m -205 yg/m can contribute to increased
prevalence of chronic bronchitis, increased incidence of acute lower
respiratory disease and diminished pulmonary function in school chil-
dren. Unanswered questions still remain with respect to short-term
exposure of NCL and the effect on acute and chronic disease, those
levels of NCL which may contribute to systemic effects and levels of
NCL which may be necessary to control possible adverse health effects
from suspended nitrates.
5.1.2.2 Animal Studies
5.1.2.2.1 Biological effects of N0?--The toxicological aspects of NCL
have been reviewed recently in these documents to which EPA has had
direct input and liaison: (1) NATO Committee on the Challenges of
Modern Society, "Air Pollution: Air Quality Criteria for Nitrogen
19
Oxides," (N. 15) June, 1973; and (2) The National Academy of Sciences,
"Air Quality and Automobile Emission Control," No. 93-24, September,
20
1974. Material contained here treats only a few reports in the
211
-------
literature which are considered significant and which have been pub-
lished since these reviews were compiled, and material which is now
being prepared by various known sources. NO- biological work being
carried out currently in EPA is discussed in some detail. The pos-
sible relevance of such information is discussed in the light of
previous data and predicted future possiblities.
5.1.2.2.1.1 Pathological effects of N0? on the respiratory track--
Action of NO- on the respiratory tract has been reviewed previously.
Its proclivity to elicit necrobiosis and desquamation of Type I alve-
olar epithelial cells with subsequent replacement by Type II cells and
eventual regression to normal is well-known. Such changes are not
21 22
specific for NCL, but occur with CL and hyperbaric oxygen, ' and
are probably indicative of lung reaction to an oxidant or possibly any
irritant at this level. Changes are induced by approximately 2 ppm
NCL and above. Exposure to higher concentrations, 10-25 ppm on a con-
tinuous basis for 3-12 months has induced pulmonary emphysema-like
23_35
disease in rats, mice, dogs and rabbits. * Lesions have been studied
in detail, principally in the rat for a number of years. In the rat the
disease does appear to be emphysema, since the lungs are enlarged and
contain fewer air spaces, presumably from rupture of alveolar septae.
Four recent studies have examined the lesion phenomenon to determine
(a) if it is a problem associated with pulmonary growth, (b) whether it
mimics natural disease, (c) what is the physical mechanism of its pro-
duction and (d) whether interaction with ozone enhances its induction
or severity.
212
-------
Two studies have examined developmental features of the disease in
rats by means of morphometry and post mortem pressure-volume studies.
One, primarily on oxidant (ozone), showed that while exposure to 0.2 ppm
oxidant resulted in significant increase in lung volume and loss of lung
elasticity (lung recoil), no such effects were noted at 0.2 ppm in
N02.36 (Table 5.7)
The other study designed presumably to answer the question of lung
maturation showed that while exposure to 15 ppm NCL from before birth to
65 days of age resulted in slowing of lung growth (interpreted as being
due to slowed general growth for the first 10 days), the animals subse-
quently caught up and had essentially normal alveolar maturation at 65
days.35 (Table 5.8)
•7/r
One of these studies examined the physical mechanism of lung
enlargement associated with both NCL and ozone exposure. The data
indicate that exposure to 0.2 ppm ozone produced a stretching of the
stroma and loss of lung recoil. The implications of this change may
be rather important in exposure to oxidant, NCL and possibly other
factors such as exposure to tobacco smoke, since in earlier studies
structural alteration of colleigen was noted by electron microscopy
examination, and chemical denaturation of collagen and elastin has been
reported in independent studies.
The two studies cited above indicated, to summarize, that exposure
to either NCL or CL does not impair formation of pulmonary alveoli in
the growing animal.
Another study examined the morphologic features of experimentally-
induced emphysema by NCL and papain exposure and compared them with
213
-------
38
natural emphysema in horses and human beings. This study showed that
the NCL-induced lesion had a "stretched" appearance, rather similar to
the natural disease and that thepapain lesion had a flaccid, "motheaten"
appearance. A notable finding of this study was that the physical
appearance of emphysema varied with the species studied and was possibly
influenced by differences in the size and frequency of the inter-alveolar
pores (pores of Kahn) graded as follows:
Horse 27.13
Monkey 8.50
Man 6.14
Hamster 3.64
Mouse 3.08
Rat 1.28
In this study emphysema was shown to be induced in mice after continuous
and intermittent exposure to 100 pg/m (0.5 ppm) NCL.
A study was conducted to look at the possible interaction of NCL
and 0., by the parameter of induced pulmonary lesions. Since these two
gases occur at approximately equal concentrations on a ppm basis in
ambient air, various exposures were carried out with equal parts of the
two gases. The lesions described were qualitatively similar. Desquama-
tive Type I lesions and replacement by Type II occurred at 0.25 ppm 0.,,
but required 10-20 times this amount for NCL to elicit similar lesions.
Thus, the NCL effect was obscured when equal parts of the gases were
employed. Equivalent concentrations needed to produce equal lesions
were on the order of 0.5 ppm to 0.9 ppm 0, and 15-20 ppm NCL. Effect
39
was noted at 0. 5 ppm CL after 24 hours exposure.
5.1.2.2.1.2 Biochemical effects of N0?--The biochemical mechanisms by
which NCL reacts with tissue resemble 0, to some degree by virtue of
their common oxidizing properties. Thus, lipid peroxidation has been
214
-------
Table 5.7. SUMMARY OF DATA FOR RATS IN MORPHOMETRIC STUDY36
Ozone (20) Control (17)
Initial Body Wt. (gm) 87.6 ± 1.2 87.0 ± 1.4 N.S.
Final Body Wt. (gm) 284.8 ± 4.8 291.4 ± 4.5 N.S.
Wet Lung Wt. (gm) 1.80 ± 0.03 1.80 ± 0.03 N.S.
Dry Lung Wt. (gm) 0.438 ± 0.008 0.444 ± 0.011 N.S.
Lung Volume (ml) 9.28 ± 0.37 7.97 ± 0.41 <0.02
Mean Chord Length (y) 62.5 ± 0.7 59.5 ± 0.8 <0.01
Alveolar Surface Area
(on2) 6,786 ± 242 6102 ± 247 <0.05
Alveolar Number
(millions) 39.7 ± 1.4 37.8 ± 2.3 N.S.
Mean values ±1 S.E. for 10 Ozone and 17 Control animals.
N.S. = no significant (P>0.05)
reported both from in vivo and in vitro exposure to NO-. Exposure of
rats to 1 ppm NO- produced evidence of peroxidation as apparent from
the appearance of diene conjugates of polyunsaturated fatty acids in
40
extracts of lung lipids, as described by Thomas et al. Peroxidation
was maximal 24 hours after exposure and persisted for at least 48 hours
after exposure. When exposure consisted of six daily four-hour
exposures, the diene conjugates increased cumulatively.
As with ozone, animals receiving a tocopherol-deficient diet have
been found to be more susceptible to lipid peroxidation. According to
these authors, reactions differ from ozone, at least in their thin
films, in that phenolic antioxidants retarded N02 but had little or no
effect on ozone. The authors speculate that the N0?-catalyzed reaction
proceeds indirectly by the formulation of a nitroso free radical, citing
215
-------
LO
to
CO
^D
pT . *~^
0 5
CO
[H J O
<; C<
r j ^J
U W
Q CM
O
Q Z
^ Q
<>^
o o
LU Cd
i
2 'U
CO U
H Z
g eg
^3 i 1
f— i
f T~] H
i "] ty^)
& H
2 S
7T^ ^ "I
§
S ,_j
^ s
Q
0 >
n
w S
g rH
§ §
]
-Q 52
Ss
T-
o ;
g T
rH T
X> f
4J
•H
f~]
^3
0
crS
o ,
CO
1
•^
O
S
rH
03
P-i
MH
o
SH
0)
g
Z
tn
-p
•rH ^~>
i ^^ i
3 c^
3 dl rH
> & X
} CO^rH
3 ' 6
H r-l v — '
X
O
cuto"^
PL, 0
rH
'S ><
e vJS
^
o
txO
s
-^
r-t O
Q__J
"i
!*£ ^
•^.^
rH
OH
OJ
§
rH
O /— N
J>^
O
4-* rH
•H
.§ ^S
rH
t~- LO CM CD OO tO
OO to OO OO tO \O to
rH CM t-~ O~i CTi CM
rH CM
tO tO O rH LO VO CM
OO LO i-H ^J- VO OO rH
CM CM tO tO CM CM tO
rH tO O O LO LO LO
rH CO tO LO r~-
216
-------
the work of Estefan et al. The nitroso free radical, in turn, initi-
ates auto-oxidation by the abstraction of methyl hydrogen. The alkyl
free radical then reacts with molecular oxygen-forming fatty acid, hydro-
peroxide, leading to peroxidation reaction in a normal manner. Antioxi-
dants thus prevent NCL catalyzed by reaction with the peroxy free
radicals.
The ozone-catalyzed reaction proceeds by the direct addition of CL
across the fatty acid double-bond. These differences may explain the
lower toxicity of NCL as compared to ozone. The autocatalytic reaction
induced by NCL might be expected to produce a low development of the
products of initiation followed by a rapid reaction once critical con-
centration was achieved. Thus, normal cellular protective mechanisms,
such as the antioxidants, may extend this period more or less indefi-
nitely, unless an overwhelming concentration of NCL is employed, whereas
for ozone, direct attack on the fatty acid double-bond would not be
autocatalytic and repair forces would be ineffective. Thus, the
phenolic antioxidants would be ineffective in blocking the reaction.
Studies by Buel, et al. suggest that exposure to NCL may alter
the molecular structure of lung collagen. When animals were subjected
to one exposure, the reaction appeared reversible, but animals exposed
to 0.25 ppm repeatedly showed persistent structural alterations when
viewed by electron microscopy. These changes persisted for at least
seven days after the final exposure. Considerable attention has been
paid to the ability of NCL to cause alterations in blood or tissue
enzymes. Buckley and Balchum " have described the influence of
acute and chronic exposure to NCL on the total activity of lactic
217
-------
dehydrogenases (LDH) and on the proportion of its various isoenzymes.
Briefly, exposure for 10 weeks or longer at 10 ppm or two hours at 40 ppm
caused changes both in tissue oxygen consumption and LDH and aldolase
activity. They noted also that a shift to the anaerobic isoenzyme
46
occurred in the lung. Ehrlich also noted this anaerobic shift in the
heart muscle of exposed hamsters, and Sherwin et al. noted that hyper-
plastic epithelial cells formed in response to N0~ exposure stained
heavily for LDH and suggested the demonstration of these cells by his-
tochemical technique might be a useful index of exposure. This increase
48
in number of LDH positive cells was also noted by Buckley and Loosli
after exposure of germ-free mice to NCL.
The significance of the above findings is unclear, since the high
levels of exposure may have produced artifactual response. However,
recent unpublished and unconfirmed data suggest that a number of blood
and tissue enzymes may be altered in guinea pigs by intermittent expo-
sure for four months at levels as low as 0.5 ppm. These data urgently
need confirmation since their implications for biological effects of
N02 are evident.
The striking difference in methemoglobin generation between in vitro
and in vivo exposure of blood possibly may be accounted for by the
activity of methemoglobin dysmutase, which causes a rapid reversal to
oxyhemoglobin in the intact animal, and that during actual exposure by
inhalation, the conversion to methemoglobin does not exceed the power of
this enzyme to effect the reversal to oxyhemoglobin.
Two biochemical studies currently are being conducted. The first
examines the prospect that the toxicity of NCL and 0, may be an "ultimate"
218
-------
toxicant produced by a reaction between theoxidant and unsaturated fatty
acid. Ozone is known to react with unsaturated fatty acids to produce
fatty acid ozonides. Fatty acid ozonides produce Heinz bodies in human
and mouse erythrocytes. Heinz body formation is inhibited in human red
cells by supplementation with Vitamin E. This property is responsive to
Vitamin E as is the inhalation toxicity of CL in animals. The same
pathways are likely to occur in animal models as in man if fatty acid
ozonides are the true toxicant. Fatty acid ozonides also produce lung
edema. If fatty acid ozonides produce the same anatomical lesions in
the lung as 0 inhalation, one? could suppose that ozonides manufactured
O
on inhalation are the responsible agents. NCL similarly can react with
unsaturated fatty acids to produce nitroxide fatty acids which have an
unpaired electron and are likely to be very reactive. No data are
available on their properties, but nitroxide fatty acids could likewise
49-52
be the "ultimate" toxicant on NCL inhalation.
These studies are important because:
(1) Identification of the chemical mediators of intoxicant ion
allows a rational approach to estimation of health hazards and preven-
tative measures;
(2) Li vitro systems can be used to accelerate studies of long-
term exposure;
(3) Indicators of human exposure may be possible by measuring
blood levels of the "ultimate" toxicant, a by-product or an alteration
of a blood cell population (Heinz bodies in red cells, numbers of poly-
phonuclear cells or modifications of platelets);
219
-------
(4) The occurrence of systemic effects from NCL or 0_ inhalation
(such as carcinogenesis, mutagenesis, impaired liver function or
splenomegaly) may be predicted; and
(5) Animal exposure effects may be related better to human effects,
refining data for criteria and standards.
Progress to Date: Methyl oleate ozonide has been synthesized and
purified. Methods of rat lung fixation for electron microscopy have
been established by either perfusion of the vascular system or by
intratracheal instillation. The former appears most useful for obser-
vations of the alveolar surface while the latter appears best for
general anatomical surveys. A method of imbedding which orients macro-
slices at the light microscopic level has been applied to the lung.
Studies are focused on the terminal bronchiole-alveolus region which
has been shown by Freeman's group to be the most sensitive area of the
lung to NCL or CL inhalation.
Synthesis of the ozonides of all of the remaining common unsatu-
rated fatty acids (linoleic, linolenic and arachidonic acids) is
planned. Studies of methyl oleate ozonide effects on the lung (using
intraperitoneal and intravenous injection) are under way. Lungs will
be examined by the two methods to assess morphological changes at the
alveolar surface and the pneumocyte-capillary interchanges caused by 0,
inhalation. The dose-response relationship for ozonides will be com-
pared with 0, inhalation to estimate the in vivo conversion of fatty
acids to ozonides.
Nitroxide fatty acids will be synthesized and used as above to
determine the similarities and differences between NCL inhalation and
220
-------
nitroxide fatty acid toxicity. The conversion to nitroxide fatty acids
on NCL inhalation will be estimated as for CL by comparative toxicities
using parallel-line bioassay. These observations and those for 0., will
O
be compared to determine if a similar mechanism of action exists.
Because these studies relate to mechanistic observations, future
studies are likely to involve the systemic effects of long-term expo-
sure on cells other than penumocytes. Potential mutagenic and carcino-
genic properties are now known, but likely, since these compounds appear
to act on the cell surface by free radical generating mechanisms. The
co-carcinogenic potential is marked since they are likely to interact
with microsomal oxygenases which are known to produce the ultimate
carcinogens of chemical carcinogenesis. The properties of these ulti-
mate carcinogens, arene oxides;, are similar to fatty acid ozonides.
Co-carcinogenesis studies using the Saffiotti hamster tracheobronchial
tumor induction model and ozone exposure are under way in-house at
this time.
The second biochemical study examines the mechanism of action of
of nitrates and sulfates--The purpose of these studies is to determine
if vasoactive compounds are released from the lung by SO,, or sulfate,
or NCL or nitrate. Since SCL inhalation results in bronchiocon-tric-
tion below the begal reflex, it is likely that direct action on the
bronchial smooth muscle is involved. Histamine aerosols produce
effects very similar to SCL inhalation. Sulfate particulates have been
suggested by the CHESS studies to be better correlated with human
effects than SCL per se. The effects of sulfate as a modulator or
releasing agent of vasoactive compounds from an in vitro model system
were therefore studied.
221
-------
Using guinea pig lung slices because of their supposed similarity
to human lung, histamine was found to be released by sulfate ions in
the presence of ammonium ion. Nitrate ion produced some release also
but ammonium ion is a requisite for release of histamine. Histamine
release was not due to ionic artifacts. By the use of SO. ion uptake
of sulfate was found to occur simultaneously with histamine release.
Sulfate uptake was found to be intracellular. Ammonium ion decreased
sulfate uptake simultaneously with histamine release, suggesting that
histamine release occurred either by cell lysis or by degranulation of
lung mast cells. Sulfate effects were not blocked by pharmacological
agents such as adrenergic or cholinergic drugs nor by cyclic-AMP or
GMP.
These studies suggest that sulfate can act directly upon the lung
mast cell to cause histamine release. Sulfate requires ammonium ion
for effect. (Table 5.9) Sulfate particulates are principally asso-
ciated with ammonia, but ammonium ion could be supplied from the blood.
The potentiation of sulfate effects by ammonia may explain the toxicity
of these aerosols over SO-. Ammonium ion probably exerts this effect
by acting as a proton sink within the cell. Sulfate is probably trans-
ported by the chloride ion carrier. Inhibitors of chloride transport
such as phloritin are being investigated.
An important result of these in vitro studies is that the ammonia
concentration of the inhaled mixture may cause substantial changes in
toxicity studies in animals. Ammonia arises from the action of gut bac-
terial urease on urea in the gut. The second generation catalytic con-
verters deliver ammonia by the incomplete reduction of NO . Ammonium
J\.
222
-------
s
t-OJ
LO
py I
I— lO*l
^-
CO CO
E~* P^i
Z O
Ej i— t
O 5
"^ i—3
PC O
tin CO
CJ H
5 •<
J CO
g tjs
OH
Ei, 0
0
LLj rH
^7
I— 1 PH
S O
i— i z
nn FTI
CO
a,
w
a! 03
C P 0
0 !/) rH
O -H 0
r< PC Pi
0
PM
(r— ^
0 bfl
.^ (/J ^
S 03 rH
o3 0
•P rH c
t/i 0 oh
•H C£ ^^
pr| QO
3.
^ — '
to
p
Pi
m 53
o s
•rl
• 5-1
O 0
r^ Oj
fl
C-M
*"O
g
o
&1
o
o
t"— o^
o o
LO OO
t-O LO
oo «*
o o
bO LO
OO K)
• •
0 0 +| +1
** LO
• •
CTl LO
I— 1
O ^O CTi ^f
a>
T.\ 0
(D -H P
T3 0 5-i o3
•H P O P
5-i Oj rH 0)
O "-H JE; o
rH t— I O
03
T3
0
a
•rH
to
10
o3
X
rH
•H
5-i •
03 to
5-( 0
P 5-i
•H O
<*"i [ ^
5-i 10
03
0
t/) p!
OS -H
o3
0 P
p w
03 -H
t_l _J (-H
rH
3 0
tn ^
G
•rl O
Pi
s °
5 o
*
223
-------
sulfate may therefore result and be more toxic than sulfate or sulfuric
., , 53
acid alone.
Sulfate and possibly nitrate is likely to cause the release of
histamine by competing for histaraine binding sites on heparin in the
histamine granule of the mast cell. Rat peritoneal mast cells will be
used to isolate mast cell granules in studies planned to investigate
this possiblity. Sulfate release of histamine and binding to granule
heparin will be measured using acridine orange metachromasia. The
release of other vacoactive compounds by sulfate from lung slices will
be examined. Prostaglandin synthesis and release will be measured by
radioimraunoassay and smooth muscle bioassay. Slow reacting substance
of anaphylaxis (SRS-A) release will also be measured by guinea pig ileum
assay.
SRS-A has been suggested to be a sulfate ester. In later studies,
the incorporation of SO. into SRS-A and other substances will be
measure to determine if sulfate inhalation stimulates vasoactive hor-
mone synthesis. Sensitized guinea pigs will be used to repeat the sul-
fate studies above to determine the effect of sulfate on atopic-lik
release of vasoactive hormones. Sensitized animals are models of
asthmatic individuals. An important mechanism for sulfate effects
which could influence all hormone action is the sulfation of the enzyme
adenyl cyclase. Adenyl cyclase controls almost all hormone action in
the cell. Phosphorylation of the enzyme has been proposed as a control
mechanism. Sulfate could compete for this reaction leading to a sulfo
enzyme which could be more stable than its phosphorylated counterpart.
224
-------
A reduction in adenyl cyclase levels could have broad effects. The
effect of sulfate on membrane adenyl cyclase will therefore be
determined.
5.1.2.2.1.3 Pulmonary physio1ogy--Few studies on the pulmonary physio-
logic effect of NO- have been reported. The subject was thoroughly
19
reviewed in the NATO document. A small amount of activity on the
pulmonary physiologic effect with nitrate will take place under the EPA
sulfate research program, rationale being the possible similar irritant
action of nitrate and sulfate ions and their respective acids, and the
fact that the chemicals occur generally in the same polluted atmospheres.
Work will proceed on this project commensurate with allocated resources
on the basis of pulmonary physiologic response of nitrates and nitric
acid mist together with sulfate and sulfuric acid, evaluation of the
influence of the ammonium ion. With the present allotment of effort,
work should be completed on the relative irritant action of one nitrate
by July, 1975.
Studies which employ the interaction of pathogenic organisms and
air pollutants to demonstrate increased susceptibility to infectious
pneumonia are unique to the EPA program. This sytem has been effi-
54 55
ciently utilized in environmental toxicological studies of ozone,
56 57
nitrogen dioxide, irradiated auto smog and exposures to trace
metals such as NiO, MnO-. '" It has been shown that the role of
. P . . , , , . - 55,56,58,59 60 , 56
infectivity can be enhanced in mice, rats, hamsters
and squirrel monkeys. Among the infectious agents eliciting this
reaction in the above animals are Streptococcus pyogenes, Klebsiella
225
-------
pneumoniae, Diplococcus pneumoniae, and influenza PR-8 virus. Thus,
this model is an exquisitely sensitive indicator of biological effects
in vivo. In addition, the model is useful in studying the mechanism of
f\*y f\~{
action of the pollutants. *
An important problem in the toxicity of NCL (and ozone) concerns
the rela tive toxicity of the mode of dose since both of these gases are
subject to intermittent diurnal peaks whose frequency and magnitude
reflect the input into the atmosphere of NO and other precursor sub-
stances, and conditions favoring the formation of the secondary pollu-
tants (photochemical smog). Thus, the population in any given area is
not subject to a continuous dose of pollutant, but a series of peaks
and valleys of variable duration and irregular occurrence. Previous
biological studies have never concertedly attacked this problem and
since the import is obvious to heatlh standard-setting and control
measures, EPA is making a serious study of the problem. A portion of
this program is incorporated into the US-USSR Agreement on Cooperation
in the Field of Medical Science and Public Health. Manuscripts are in
64
preparation for presentation in Riga in the USSR in December, 1974,
for the above program and in New Zealand at the Australia/New Zealand
Clean Air Conference in February, 1975. The principal parameter
employed at this state to evaluate the important of mode of dose is
the infectivity system described above. A brief review of this infor-
mation follows.
Figure 5.1 shows greater mortality from pulmonary infection result-
ing from the same concentration X time for exposure to NCL over dif-
ferent time periods. Note that the actual concentration has more
226
-------
influence than duration of exposure. If no interaction existed between
concentration and time (i.e., if they had equal importance) , all the
test groups should show the same mortality in the above experiment --15%.
But the trend of the combined data indicates a gradient response from a
high of approximately 50% to a low mortality equal to controls.
The interaction of a constant concentration with time over a longer
period was examined. Two concentrations of NCL have now been studied.
A level of 3.5 ppm was used for comparison with the study mentioned
above. A lower value of 0.5 ppm was selected since this concentration
represented the lowest figure in the published literature showing posi-
tive results for chronic exposure in the infective system. The data
shown in Figure 5.2 represent 60 individual observations for the 3.5 ppm
level and eight observations for the 0.5 ppm. The simple linear
regression shows that mortality increased with time and concentration
and was statistically significant at the 0.01 probability level. Pre-
dictions that can be generated from these regressions show that sta-
tistically significant difference from controls occurred after one-hour
exposure to 3.5 ppm and after three weeks at the 0.5 ppm concentration.
While the first series of experiments (Figure 5.1) showed that there was
a strong interaction in favor of concentration over time, the latter
study still shows the importance of toxicological information from long-
term exposures for NCL.
The data from the same series of experiments was analyzed for
influences of exposure time to 3.5 ppm on mean survival period calcu-
lated according to the equation:
MST - CA+B) = (d=L)
d n
227
-------
- h
s
T
-*-
s
"T
S
s
Figure 5.1. % difference in mortality
(NCL - control)64
y s 10.42 -f 27.5S LOGio (TIME)
HOURi
TIME
CAYS
MONTHS
Figure 5.2. Difference in % mortality versus
length of exposure to .5 or 3.5
ppm N0
-------
"A" is the last day that any individual mouse was alive; "B" is the num-
ber of mice surviving "A" days; "D" is the last day of the experiment
(in this case 15); "L" is the number of mice that were alive on day "D"
and "n" is the initial number of mice in the experimental group.
A statistically significant (P<.05) regression with a negative
slope was shown for mean survival of mice challenged with Streptococci
versus length of exposure to 3.5 ppm for various periods prior to the
challenge. Figure 5.3 shows that there is an inverse relationship between
survival and length of exposure at constant concentration. With this
parameter, effects were also statistically significant after one hour
of exposure to 3.5 ppm.
Using the same model system (fixed concentration versus time),
other concentrations are now being tested for the purpose of developing
information that can be used to develop regression curves at other con-
centrations. That can then be used for predicting purposes.
Other experiments are also designed to test interaction of concen-
tration and time as above, but they use an intermittent exposure model
in place of continuous exposure. As this report is being prepared,
data from this experiment is still being collected.
5.1.3 Carbon Monoxide
5.1.3.1 Human Studies--The present ambient air quality standard for
3
carbon monoxide [10 yg/m (9 ppm) averaged over eight hours and of
40 yg/m (35 ppm) one hour average] produces carboxyhemoglobin levels
of approximately 1.51. Since the time of establishment of the stan-
dard, new evidence has appeared giving additional support to the cur-
rent standard in three areas. These are: (1) Effects on individuals
229
-------
•10
7.0
5 M
a
3E
^ 5.0
o
X
i 4'°
o 3.0
cc
1 2-0
3C
y LO
l«J
£ o.o
u_
-1.0
-2.0
•»• 2.519
(tine) X
.,
10
TIME IN HOURS
100
1001
Figure 5.3.
The difference in mean survival time between
control of NCL mice versus length of
exposure to 3.5 ppm NCL
with atherosclerotic vascular disease, (2) efffects on exercise capacity
in healthy individuals and (3) additional evidence on alertness effects.
5.1.3.1.1 Effects on individuals with Cardiovascular disease--Three
separate studies have shown that individuals with atherosclerotic
heart disease manifested by angina pectoris experienced chest pain
earlier during exercise when carboxyhemoglobin concentrations are in the
69
2.5-3.01 range than when they are about 1.01. Another study has shown
that men with peripheral atherosclerotic disease develop leg pain after
less walking at carboxyhemoglobin levels of 3.0%. These studies were
well controlled and demonstrate increased morbidity in a segment of the
population with a common disease process.
5.1.3.1.2 Effects on exercise--Young, healthy men at carboxyhemoglobin
concentrations of 5.0% have increased heart rates and decreased maximal
230
-------
70
oxygen consumption when undergoing strenuous exercise. Preliminary
analysis of an EPA study indicated lessened exercise performance at
COHb levels of approximately 5-6%.
Decrements in exercise capability in the healthiest segment of the
population suggest more serious impairments in older persons who
frequently have subclinical cardiovascular disease.
5.1.3.1.3 Effects on alertness--Several recent studies have shown
impairments on vigilance tasks at carboxyhemoglobin concentrations
72-74
ranging from 3.0-6.51. These data tend to support the original
75
neurobehavioral work used to set the standard.
Other evidence has appeared in recent years implicating levels of
carboxyhemoglobin below 10% with increased risk of atherosclerosis,
impaired fetal development and in altered drug metabolism. While
this evidence is certainly inconclusive at present, it still indicates
the range of pathophysiologic functions possibly affected by carbon
monoxide and the need for additional research.
The strongest evidence now supporting the present standard rests
on the harmful effects of CD on individuals with cardiovascular disease.
Cardiovascular disease is widespread among adults and present studies
indicate aggravation of symptoms in a large sub-set of these adults.
The ambient standard is given in Table 5.10. Table 5.11 gives the
judgment threshold estimates for health effects of CO. Values are
given for CO concentrations at light activity and during exercise as
well as rest since equilibration with a given CO concentration occurs
more rapidly with increased ventilation. Table 5.12 gives the estimated
size of the population at risk from adverse effects of carbon monoxide.
231
-------
Table 5.10. AMBIENT AIR QUALITY STANDARD FOR CARBON MONOXIDE
National Primary Standard
3^
Pollutant Concentration (mg/mj)Averaging Time
Carbon Monoxide 10 (9 ppm) Maximum 8-hour
40 (35 ppm) Maximum 1-hour
5.1.3.2 Animal Studies - -Apparently no animal experiments have been con-
ducted since publication of Air Quality Criteria for Carbon Monoxide
(1970) that employed <50 ppm CO exposures. This concentration was the
lowest exposure level reported in that document. The emphasis since
1970, however, has shifted from reporting ambient concentrations to
reporting carboxyhemoglobin (COHb) levels and the relationship of the
parameter with the effects.
New data from animal experiments reported since publication of the
1970 criteria document have shown that COHb levels as low as 2% in
animals decrease pO~ of brain and liver. Dogs were previously reported
to show blood changes at 7.3% COHb, but newer data shows that 5% COHb
increases coronary blood flow. COHb levels of 9-10% maintained in rab-
bits during half of the gestation period resulted in decreased birth-
weight and increased neonatal and infant mortality.
78
Zorn studied the effects of CO inhalation on the brain and liver p0
of animals using Lubber's platinum electrode and surface electrodes. In
both tissues the p02 decreased, even at carboxyhemoglobin of 2% saturation
and the decrease varied approximately linearly with the increase in COHb.
232
-------
0)
bO
•H
,-J
0
in
•H
o i-o
fn t-^
0
P
•H
>
•H
P
U
.3
O
CTl
Ol
LSI
LO
LO
INI
LO
CTi
•*t
(Nl
8
3 B
co •
£3"
rH 10
03 O
%%
3 0
c"
w
0 0
o 0
•H t/1
P 10
03 n3
•H
U
M
0
P
•H
•H
P
(J
P
10
Q
CTi
O
LO
o\°
O
LO
O
LO
o
•
LO
0
o
p]
0
• rH
0
P
P^
0
•H
O
•H
m
fi
•H
0
U
c
0
';>
0
p
Pi
0
•H
U
•H
m
^
P!
•H
0
O
PH
0
•X)
•rH
0
•p
P!
0
•H
O
•H
<4H
m
P!
•H
LO
0
I—I
•s
H
10
P
O
0
w
JS
p
i— i
o3
0
0
10
fH
I
03
P
I
0
O X
JH >
0 -i-l
m P
^ u
0 rt
P
0
I— I
u
10
8
0
03
<-H
O
TJ
0
10
n3
0
(H
U
G
I
O
•8
03
P
0
0
!H
•H
03
233
-------
e
n,
w
Q
|
2:
nJ
O
DH
PH
O
^
CO
I—I
rt
^
w
s
PH
^
§
OH
H,
O
w
tsi
rH
CO
LO
o>
rH
•8
H
*
tfl
p
oj
Pi
5
•H
P
OS
rH
PH
O
o
0
*"O p,
0
MH P
O OS
!H
S^v. /
(_)
Pi P!
0 o
3 -H
O* P
0 OS
£ 3
_ P-
TJ O
0 a,
p
u
0
w
p
u
0
m
m
w
^
rH
as
0
a:
to
0
1
Pi
o
•H
rH
'£
CXI
<*
*
«
;>
t/) O
rH OS
^J
TT? "^
OS p!
as
MH
O t/)
^
o\° as
LO 0
X
oo
rH
v — '
•5 0
0 as
O 0
C^ l/l
Co *H
!H T3
0
. t J ^
O >H
0
C/1
'u ^
§'g
0 10
TJ O
0 10
rH rH
•H 0
aS p*
rH
Pi
O
•H
rH
•H
e
O
iH
cN3
0
o
rH
0
U
TO
tf
^li
O
m
rH
0
^
as
O
•H
10
t
1
as
0
0
a
rj
O
•H
,_)
•H
6
o
rH
d*P
0
0
rH
K^
?***
•H
•H
O
as
rH
as
P
p|
^
^
•5
•H
8
c
§
MH
rH
0
P
Pi
HH
o
•H
P
O
p.
0)
e/)
aj
(U
!H
O
10
OS
S
•S
t/)
CO
CXI CO
234
-------
79
A second study by Rondia found that there was a decreased ability
of rat liver to metabolize 3-OH benzon alpha pyrene in rats exposed to
60 ppm CO.
80
Cobum and Kane studied COHb concentration of 10-12% in anaesthe-
tized dogs and found that this concentration resulted in an inhibition
of loss of hemoglobin-haptoglobin from the plasma. Researchers studying
the relationship of varying concentrations of COHb and increased
coronary blood flow exposed conscious dogs to CO at increasing concen-
trations and observed a 13% increase in coronary blood flow with a 4%
increase in COHb. At 20% COHb coronary blood flow increased by 541.
Coronary flow measurements at 4 COHb concentrations suggested a linear
relation between COHb concentration and flow. A threshold value was not
observed, although measurements were not made below 5% COHb.
COHb levels chronically increased to an average of 12.4% in
cynomolgus monkeys produced polycythemia with an increase in hematocrit
82
from 35-50% volume of packed cells. All animals developed increased
amplitude of P-wave in the EKA, and some developed T-wave inversions
suggestive of myocardial ischemia. Monkeys with experimental myocardial
infarction had higher COHb and more severely abnormal EKG than normal
monkeys or monkeys with myocardial infarction breathing room air.
OT
Wells studied pregnant rats exposed to 15,000 ppm (1.5%) for 5-8
minutes 10 times every other day during 21-day pregnancy and found
maternal unconsciousness and abortion or resorption absorption in most
of the fetuses. The surviving newborns did not grow normally. No
pathologic changes in nervous system were noted. Similar exposure to
5900 ppm affected only a small portion of the animals.
235
-------
84
Williams and Smith exposed rats to GO at 3400 ppm for one hr/d for
three months. Peak COHb = 60 = 70%; of seven exposed females, half the
control number pregnancies occurred. The number of young delivered
decreased and only 2/13 newborns survived to weaning. In five females
exposed for five months, no pregnancies occurred.
77
Astrup et al. studied pregnant rabbits exposed to CD of 90 ppm for
30 days. Exposure during first pregnancy resulted in COHb 9-10%. Birth-
weights decreased 12%. Neonatal mortality increased from 4.5% in controls
to 10% in exposed. Mortality during the following 21 days increased from
13-25%.
In a second experiment, pregnant rabbits were exposed to CO of 180 ppm
for 30 days. Exposure during first pregnancy resulted in COHb 16-18%.
Birthweights decreased 17%, neonatal mortality 35% vs. 1% and mortality in
following 21 days was 27% vs. same in controls.
85
Baker and Tumasonis exposed chicken eggs to various concentrations
of CO 0-650 ppm continuously from laying through 18 days of incubation and
found that hatchability was inversely related to [CO]. 425 ppm reduced
hatching to 75%, embryo weight was normal, no congenital abnormalities,
650 ppm reduced to 46% and developmental anomalies of tibia and metatarsal
bones.
o/r
Baker et al also exposed eggs 12 and 17 days old to CO of 425 ppm
for 24 hours. COHb in 18-day-old chicks averaged 36%. Livers of 18-day-
old chicks had about 50% increase in the activities of two hepatic mixed-
function oxidases, hydroxylase and 0-demethylase. No effects were seen on
13-day-old chicks.
236
-------
87
McGrath and Jaeger found that 50% o£ newly hatched chicks with-
stood exposure to 10,000 ppm for 32 minutes. Survival time to this [CO]
decreased with postnatal age from 10 minutes in one-day-old chicks to
four minutes at ages 8-21 days.
88
Jones exposed rats, guinea pigs, monkeys and dogs to 51, 96 and
200 ppm for 90 days and found no toxic effects of the 00. The only
reported physiological change was increased hemoglobin and hematocrit
values for all species.
89
Ator exposed rats to 100-1000 ppm of CO for 90 minutes and found
no effect on response rate for differential reinforcement of low-rate
schedules at exposures up to 250 ppm.
90
Gage et al. found significantly higher activity in female mice
exposed to 100 ppm CO for four-seven days than in males similarly
exposed or in female controls. No statistically significant effect on
the activity of male mice occurred under these exposure conditions.
5.1.4 Hydrocarbons
The National Ambient Air Quality Standard for hydrocarbons was
based almost entirely on their role as precursors of other compounds
formed in th- atmospheric photochemical system and not upon direct
effects of the hydrocarbons. Particulate hydrocarbons and more specifi-
cally polynuclear hydrocarbons were not considered on the criteria
document. Ambient air data are limited primarily to non-methane gase-
ous hydrocarbons and benzo-a-pyrene.
The National Academy of Sciences published a review on particulate
91
polycyclic organic matter in 1972. A scientific and technical assess-
ment report by EPA on the same subject is in the final stage of
237
-------
publication. EPA reports on polychlorinated biphenyls and vinyl chloride
are also in the final stage of publication.
92
Aksoy et al. has reported on six cases of Hodgkin's disease in
workers with a history of chronic benzene exposure. The subjects were
exposed to benzene for a period ranging from 1 to 28 years with a mean
of 11 years before the onset of Hodgkin's disease. The concentration of
benzene in the working environment ranged between 150 and 210 ppm. Ven-
tilation in the work places was reported as poor. A summary of these
cases is shown in Table 5.13. There was no indication in this report of
the number of workers at risk or any discussion of other possible agents
that might have been present in a poorly ventilated work place.
93
Ahmad and Whitson reported on one incidence of rapid loss of con-
sciousness of 10 new female employees reporting for work in an industrial
garment company. Immediately after the incidence, measured concentrations
of formaldehyde ranged from 2 to 10 ppm. Other workers in the area had no
apparent difficulties. The authors suggest that the response was due to
exposure to formaldehyde although little additonal evidence was given.
94
Thorpe conducted a study of eight affiliates engaged in the manu-
facture, transportation and marketing of petroleum or chemical products
containing benzene for at least 10 years. The incidence of leukemia in
the general population with an age distribution similar to that in the
Thorpe study is about three to eight cases per 100,000 population. Eigh-
teen cases of leukemia were reported for the period 1962 to 1971. These
cases were distributed among four of the affiliates--the other four
reported none. A summary of the 18 cases is shown in Table 5.14.
238
-------
Thorpe also found that the occurrence of leukemia in a population
of 38,000 working in a variety of petroleum and petrochemical operations,
some of whom were exposed to low levels of benzene over a 10-year period,
was not abnormal compared to the general population in the countries
concerned (see Table 5.15). As pointed out by Thorpe, there were prob-
lems with the data base and a number of important assumptions were
necessary.
239
-------
X
t—I
CO
to
U
z
I—I
Q
l-H HJ
U, CO
<
u ia
i—I W
U M
O Q
OH ^
Q Q
u a:
t—I f-H
CJ I—I
O 2
S w
uj i—i
X H
U
i—i
2
o
CO
LO
.0
rt
E-
u
W
rt
u
^ 3)
0 C
•£•5
/? c
2 a
I' C
V)
C
M
E
0
Cu
E
Cl
V)
d
W
>
>,
§1
rt*
WBC 4200,
lymphopeni
ESR:
149 mm./hr
S^rt
•« •? 60
Cervical, inguin
lymphadenopatl
hepatosplenome
R
?!
^
-c>
Ul t«
""» >
&.£
00
*
OO
in a
* £
O,.vj
32
T-l
>
;»,
"0 ^
_
rt"
WBC 5400,
lymphopeni
ESR:
(A
~'£
'rt *-*
o rt
> °*
u F
Both inguinal, c
axillar lymphadl
moderate
.R
•->
Uj ^
"^ >
£,£
o
"
VI
U
Shoemak
$s
CM
E
CM
OO
"rt
nn
hepatosplenome
»— *
o
*^"i
>>
u c
J.I
E-5
|fr
J -a
0
X ..
o E to .c
OO >.ttj —7.
^"s «• E
1 s §.3
«f
•IS
rt *^
C ™
• - Q. >s
3 o -^;
60 C 3;
Both cervical, ii
axillar lymphadl
moderate
hepatosplenome
.R
Sj
^ m
Ln uT •"•
-J. > 5
£££
vO
^
V
_M
rt
'I
JC
LO
sa
rO
>
Q
O.
u
X
WBC 5800,
lymphopeni
ESR:
w
r-
Cervical, axillar
lymphadenopatl
findings of liver
^
* '
00
14
u
c
1
0
8S
^1
-C
E
E
00
cirrhosis
^
«
o 8
'w C
>^ rt
U C
-§£
D. o
n
J D.
C
It
• r*
Both cervical,
axillar, inguinal
lymphadenopatl
4_,
5P
5*J2
o
U,
**
r-"
c
Televisio
r~-
CM
\o
.
rt*
lymphopeni
ESR:
97 mm./hr.
vt
o
(-"
lymphadenopatl
(-"
repairmai
^
c
'g
T3
rt
•6
"*C
*
240
-------
Table 5.14. LEUKEMIA EPIDEMIOLOGICAL STUDY
Leukemia
Age
Epidemiological Study — Summary of 18 Cases ol Leukemia by Set,
at Death (or 1971), Last Job Assignment and Type of Leukemia.
Sei Age it Death Last Jcb Assignment
F 19
M 26
Mt 33'
M1 36
Mt 40
Mt 41
M 42
M 46
Mt 60
M 61
M 62'
Mt 66
M 68
M 68-
M 69
Mt 70
M 71
Ml 73'
Clerk. Head Office
Engineer, Refinery
Passenijer Car Driver
(Tank Truck Driver 196066)
Mechanic in Refinery
Plant Operator, Refinery
Filler, Refinery
Kerosene Truck Driver, Airpoit
Mathematician, Head Office
Marketing Superintendent
Industrial Assistant Dealer
Carpenter, Refinery
Truck Driver, Marketing
Comme'Cial Clerk. Head Office
Clerk Chauffeur, Refinery
Clerk, Head Office
Process Operator, Refinery
Manage', Head Office
Mechanic. Marketing Repair Shop
Type of Leukemia
Leukemia, Type'
Leukemia, Type'
Chronic Myeloid Leukemia
Leukemia, Type'
Myeloid leukemia
Acute Leukemia
Leukemia, Type'
Leukemia, Type'
Leukemia, Type'
Leukemia, Type'
leukemia, Type'
Leukemia, Type'
Leukemia. Type'
Leukemia, Type'
Chronic Myeloid Leukemia
Leukemia, Type'
Acute Myeloblastic Leukemia
Chronic Lymphatic Leukemia
Living
! In Exposed Group
Table 5.15. EXPECTED VS. OBSERVED DEATHS FROM LEUKEMIA IN
EIGHT EUROPEAN AFFILIATES 1962-1971
Eipected n. Observed Deaths From Leukemia in > European Affiliates 1962 • 1971.
Age it Tim*
of Duth
Estimated
Age Adjusted
Totil
Person Yftir1
Age Specific*
Modality
Rites (WHO-1966)
No of Deaths
Obwrvtd
No. of Deiths
1519
2024
2529
3034
3539
4044
4549
5054
5559
6064
6569
7074
7579
8084
85 And Over
Totil All Apt
3.248
15.592
36.707
49,051
47.102
51,000
48.726
37,681
3S.049
25.960
11.519
B.064
5.610
3,214
1.753
M3.276
22
1 7
27
21
30
34
*6
61
80
124
168
206
285
37 7
300
007
026
099
103
141
1 73
224
230
304
322
194
166
160
1 21
053
23.23
1
0
1
1 (1)"
1
3
1
0
0
3(1)"
4(11"
3(1)"
0
0
0
II
• The Age Specific Mortality Rites shown »ere taken from those ol the counlry (or average ol the countries) reporting the uses
in the Esso population
Cises shown ( )"are still living but for these calculations were assumed to have died it tfmr age in 1971 (see leil)
241
-------
Some polycyclic aromatic hydrocarbons are the most ubiquitous car-
cinogens known. The noxious effects of tobacco smoke in humans and
animals has been attributed to presence of various hydrocarbons. Short-
term exposures of guinea pigs to cigarette smoke decreased the number of
free lung-cells--macrophages and leukocytes--as determined from lung
95
lavage preparations. The decrease in numbers of macrophages was
attributed to the presence of acetaldehyde, tar, nicotine, hydrogen
cyanide and acrolein. And the decrease in the number of leukocytes was
related to acetaldehyde, tar and pH of the particulate phase.
Confirming the previous observations that toxicity of smoke was
part of the vapour phase of the smoke, the above studies also indicate
that biologically significant activities exist in the particulate phase
of the smoke. Necessary long-term biological experiments are warranted
to verify these short-term observations.
The cigarette smoke condensate (CSC) contains aromatic hydrocarbons,
such as pyrene, benzo[e]pyrene and benzo-[g,h,i]perylene, that have been
considered non-carcinogenic up to now. In addition, CSC also contains
phenols, the most important of which is catechol (1,2-dihydroxy-benzene).
Attempts to assess the carcinogenic or co-carcinogenic activity of these
compounds in CSC on female mouse (1 CR/Ha) skin demonstrated remarkable
98
carcinogenic activity of catechol and the three aromatic hydrocarbons.
Each of these produced skin papillomas and squamous cell carcinomas in
mice. Catechol also exhibited co-carcinogenicity in these tests. The
animals received catechol (2,000 yg), benzo[e]pyrene (15 yg), Pyrene
(12 yg) and benzo[g,h,i]perylene (21 yg) per 0.1 ml acetone, three times
a week for 52 weeks.
242
-------
99
Catechol is found in many plants and trees, including tobacco
leaves. Compelling practical problems suggested by this study include
immediate need to develop a less harmful cigarette and to reduce or
eliminate emissions of hazardous material from tobacco.
Certain chemicals such as chloromethyl methyl ether (CMME) and bis
chloromethyl ether (BGME) are widely used as intermediates in organic
synthesis and preparation of ion-exchange resins. An increased incident
of oat-cell carcinomas were seen (13 or 14 men with lung neoplasms) in
men exposed to CMME. Commercial CMME contains 1 to 7% BCME. Both
CMME and BCME are highly carcinogenic for mice and rats.
Mammalian tissue cultures have been used to study induction of
microsomal enzymes because of ease of control of time of exposure and
concentration of inducer. Following this lead, human foreskin obtained
from infants (two to four days old) at time of circumcision was grown
in tissue culture. Such foreskin tissue cultures were exposed to
10 yM of benz-(a)-anthracene (BA) for 16 hours to induce benzo(a)pyrene
hydroxylase. This hydroxylase activity was greater (1.78 to 5.30 times)
in foreskins treated with BA when compared to the foreskins in the con-
trolled medium without BA.
Considering the size of "the skin pool and evidence of inducibility
of benzo(a)pyrene hydroxylase in skin, the role of skin in detoxifica-
tion of various polycyclic hydrocarbons may be significant. The levels
of benzo(a)pyrene hydroxylase (BP hydroxylase) in liver and extraphepatic
tissues differ markedly from strain to strain of mice. This difference
may mean that genetic factors influence the constitutive level and
inducibility of BP hydroxylase. As the extra-hepatic tissues in mice
243
-------
are frequently the targets for polycyclic hydrocarbon carcinogenicity,
the inducibility of BP hydroxylase in extra-hepatic tissues from mouse
strains investigated exhibited 3-methylcholanthrene and 5-6-benzo-
-. 107
flavone.
Results from the liver, lung and skin of mice of various strains
show that tissues vary in their ability to produce BP hydroxylase in a
given strain. The three (3) hydrocarbons above were equally capable of
inducing hepatic BP hydroxylase. However, 3-methylcholanthrene was a
far less effective inducer of BP hydroxylase in extra-hepatic tissues
(lung and skin) than either benz(a) anthracene or 5-6-benzoflavone.
5.1.5 Oxidants
The present U.S. National Primary Standard for photochemical oxi-
:imum one
108,109
dants is 160 yg/m (0.08 ppm) maximum one-hour concentration not to be
exceeded more than once per year.
The original evidence for this standard for people included a
broad range of morbidity such as aggravation of asthma, aggravation of
chronic lung disease, irritation of the respiratory tract in healthy
adults, decreased visual acuity, increased eye irritation and decreased
cardiopulmonary reserve in healthy subjects. These data were completely
reviewed in the original photochemical oxidant criteria document.
Three important papers on pulmonary function after short-term ozone
exposure were published during the 1969-1974 period. All three studies
were of healthy adults. Bates et al. observed substernal soreness,
cough and some phoryngitis and dyspnea along with decreased pulmonary
function among healthy, lightly-exercising volunteers experimentally
exposed to 1500 yg/m of ozone for two hours. Hazucha et al. observed
244
-------
3 3
similar effects in volunteers exposed to either 1500 yg/m or 740 yg/m
of ozone for two hours.
112
Bates reported a marked decrease in pulmonary function among
healthy subjects doing light exercise and exposed simultaneously to
3 3
740 yg/m of ozone and 1000 yg./m of sulfur dioxide for two hours. This
deleterious effect may be synergistic and a deleterious synergism of 0,
and SO- upon plants has been described in the original criteria docu-
«. 108
ment.
Hammer et al. ' studied student nurses in Los Angeles prospectively
using daily symptom diaries during three years of their training (6).
Headache, eye discomfort, cough and chest discomfort were found to be
related to ambient daily photochemical-oxidant exposure. Dose-response
thresholds could be calculated for each symptom in relation to the daily
maximum hourly photochemical-oxidant concentration. They were as fol-
3 3
lows: Headache - 100 wg/m , eye discomfort - 300 yg/m , cough - 510
yg/m , chest discomfort - 580 yg/m .
The threshold for eye discomfort was comparable to that observed in
several earlier experimental and panel studies. Thresholds for cough
and chest discomfort were in the same range found in two studies of
^- j i 114,115
occupationally exposed workers.
On highest oxidant days (800-1000 yg/m ), students reported 481
more cough and 100% more chest, discomfort than on days when ambient
oxidant levels were at or below the present U.S. National Primary Stan-
dard (160 yg/m3).
Two recent reports summarize new studies on experimental ozone
exposure. Kerr et al. studied 20 lightly-exercising subjects exposed
245
-------
to 0.5 ppm ozone for six hours. Symptoms most commonly noted were dry
cough and chest discomfort. These symptoms were experienced more fre-
quently by nonsmokers. Significant changes in specific airway conduc-
tance (S Gaw) , pulmonary resistance (Rj), FVC and FEV,,, but not diffusing
capacity (DLGO), were also found. In addition, nonsmokers showed a sig-
nificant decrease in dynamic compliance during exposure. In general,
there was a relationship between experiencing symptoms and alterations
in pulmonary function among the study subjects.
117
Hackney exposed adult male volunteers doing intermittent light
exercise to 0.37 or 0.50 ppm of ozone for two hours. These subjects
not only developed measurable physiological and biochemical changes, but
felt physically ill and were unable to do their normal jobs during expo-
sure and for several hours afterward. All subjects experienced biochemi-
cal changes. The least sensitive subjects experienced no respiratory
symptoms or physiological changes after five-hour exposures to 0.5 ppm
ozone on two successive days. The most sensitive subjects experienced
respiratory symptoms after a single two-hour exposure to 0.37 ppm of
ozone. The more reactive subjects were generally had a history of asthma,
allergy or previous adverse symptom reactions to smog exposure.
No additional effects were observed when 0.30 ppm nitrogen dioxide
was added to ozone or after adding 30 ppm of carbon monoxide to 0,-NCL
mixture. There were no significant changes among the most sensitive
subjects during ozone exposures of 0.25 ppm for two hours suggesting
that this concentration was a lower limit for these effects in healthy
adults.
Three comments about the new evidence are pertinent. First,
experimental, occupational and epidemiologic studies of healthy males
246
-------
and females exposed to ozone per se or ambient photochemical oxidants
3
indicates that as little as one to two hours of exposure to 500-600 yg/m
of ozone or photochemical oxidants causes cough, chest discomfort, sub-
sternal soreness and impaired pulmonary function in these people.
Hence, it would seem that these thresholds should generally hold true
for healthy adults and would probably represent an upper limit for
especially sensitive persons, e.g., those with asthma or pre-existing
cardiovascular disease.
Second, the synergistic effect of ozone and sulfur dioxide on human
pulmonary function has been reported in the past two years. This syner-
gism has been described in plants for several years and was summarized
in the original criteria document. Also, there is concern that inter-
action of these two pollutants on lung function may actually be additive
and not synergistic. Since lung function is impaired in a non-linear
fashion to any stimulus, "extra" decrement attributed to the combina-
tion of SCL and ozone could be an additive effect in the non-linear
part of the lung-function curve.
Third, the terms "radiomimetic" and "mutagenic" should be used
with considerable caution when applied to ozone. Animals that showed
alterations in red blood-cell fragility had been exposed to 10,000 rads
(a large radiation dose) as well as ozone. Chromosomal breaks in peri-
pheral lymphocytes of hamsters following ozone exposure have been
observed. But this may just be a peripheral and not a stem cell effect.
If so, it is not properly described as a mutagenic effect.
Photochemical oxidants, ozone in particular, are primary irritants
aggravating asthma, impairing athletic performance, causing severe
247
-------
respiratory tract irritation and altering pulmonary function in healthy
adults under experimental, occupational and ambient exposure conditions.
In animals, ozone alters host defenses against infection, causes various
structural changes to lung and hematopoietic tissue and produces chromo-
some breaks in peripheral lymphocytes even at ambient concentrations.
Old and new evidence to date suggests that the present primary
National Air Quality Standard for photochemical oxidants of 160 yg/m
(9.08 ppm) maximum concentration for one hour is reasonable. The evi-
dence does not warrant a different standard.
Air quality standards should be based on complete and reliable
dose-response relationships. Most quantitative health intelligence to
date was obtained from animal models. Direct extrapolation of this
research to humans is questionable. Until comprehensive data are accumu-
lated about acute and chronic human exposures, margins of safety will
remain obscure.
Exposures of other vulnerable people such as infants, pregnant
women and persons with heart and lung disease, have not been investi-
gated. These persons may require more stringent air pollution controls
than healthy young adults.
Finally, photochemical oxidants occur with other atmospheric pollu-
tants. Nearly all epidemiologic findings to date may reflect interac-
tions between oxidants and other environmental factors such as ionizing
radiation, sulfur oxides or nitrogen oxides. These interactions remain
to be discovered in the laboratory and field. Because interactions have
not been characterized in laboratory or epidemiologic studies, existing
evidence can be applied only with reservations to community exposures.
248
-------
Considerable human dose-response data under controlled laboratory condi-
tions and in natural community settings are urgently needed to establish
protective air quality standards for photochemical oxidants with adequate
but no excessive margins of sa.fety.
5.1.5.1 Animal studies--There is much documentation about exposure to
ozone at concentrations
-------
macrophage) was also deleteriously affected by low level 0_ exposures
[981.5 yg/m3 C-5 ppm)].122'124
Ozone-induced tolerance to pulmonary edema is a well-recognized
phenomenon. " However, new investigations indicate that short-
term exposures to 1963 yg/m (1.0 ppm) 0, produced some tolerance to X-
irradiation and allergic responses. Several bronchemical changes
are also produced by ozone. Many experiments confirmed the anti-oxidant
character of Vitamin E that protected animals from some of the harmful
alterations produced by ozone. " This oxidant causes alterations
131-133
in lecithin information and in activities of several pulmonary
134-143
and serum enzymes. Short-term low level 0, exposures also pro-
duced changes in pulmonary function. ' '
Some interesting systemic 0, effects are also known. It is possi-
ble that this oxidant produced pulmonary artery hypertension in animals
that were genetically susceptible. The parathyroid gland was also
affected by exposure to 1472.3 yg/m (.75 ppm) 0,. Additionally, mice
that received pentobarbital after exposures of 1963 yg/m (1 ppm) for
three hours/day for three days showed an increased sleeping time. All
of the foregoing information is summarized in detail in charts follow-
ing this section.
While reviewing the 0, literature, several gaps are noticeable.
Most research used healthy, young-adult animals. Attempts were then
made to extrapolate this data to people. Since many people are very
young, elderly or impaired in some way by various debilitating ill-
nesses, more research should be done with specialized animal models
capable of representing these groups.
250
-------
Also, exposures to 0_ are usually of a long-term intermittent nature
suggesting the need for longer-term experiments. Some more recent work
also examines the question: Are some pollutant effects naturally
reversible? Perhaps more research should attempt to answer this impor-
tant question. Also, since 0, induces tolerance to pulmonary edema, it
would be worthwhile to examine this matter further.
Some researchers have tried to delineate the actual mechanism of
action of CL, but much work still needs to be done. Furthermore, since
most research concentrates on 0,-induced pulmonary damage, it would be
of great interest to examine extra-pulmonary effects. The existing
data on ozone's systemic effects justify further delving.
5.1.5.1.1 Pathological effects--The pathological effects of 0^ on pul-
monary tissue are well documented ' and additional study would
119 3
only be of interest. Pan et al. exposed rabbits to 785.2 yg/m
(.4 ppm) 0, for six hours/day,, five days/week for 10 months. The
authors observed emphysematous and vascular lesions in the lung. The
small pulmonary arteries were thicker than controls because of the
increased size of the tunica media vasorum that in turn appeared to be
a result of tissue edema or muscular hyperplasia in some instances.
5.1.5.1.2 Infectivity--Experiments reported in the initial criteria
document showed that three hours of 1570.4 yg/m (.08 ppm) 0,
enhanced mouse mortality because of subsequent exposures to Group C
, 3 , 147,148 , .
Streptococcus pyogenes. Goldstein et al. exposed mice to an
32
aerosol of P labeled Staphylococcus aureus either after 17 hours
of CL or before four hours of 0.,. The concentrations used were 1177.8,
1374.1, 1570.4, 1963 or higher yg/m3 (.6, .7, .8 or 1 ppm).
251
-------
Mechanical clearance and bactericidal capabilities of the lung were
then measured. Pollutant exposure for 17 hours before infection caused
a decrease in tidal volume and a subsequent decrease in pulmonary deposi-
tion of bacteria (67% at 1963 yg/m ). There were also significant
2
decreases in bactericidal activities beginning at 1963 yg/m (1 ppm).
When
the mice were exposed to four hours of 0^ after being infected,
slightly different effects were noted. Bactericidal activity was
decreased at a lower level--1177.8 yg/m (.6 ppm) and there was no
effect on pulmonary deposition or clearance of the Staphylococci. The
proposition was advanced that since mucociliary clearance is unaffected
by subsequent 0, exposure, bactericidal defects must be due to dysfunc-
tion of another line of pulmonary defense, most probably the alveolar
macrophage.
Others working in the same field found that 981.5 yg/m (.5 ppm)
of 0,, 16 hours/day for seven months had no effect on the mechanical
clearance of polystyrene and iron particles in rabbits. There was only
a small decrease in bacterial clearance in guinea pigs.
121 3
In other studies, mice were exposed to either 1963 yg/m (.1 ppm)
03 and 3762 yg/m3 (2 ppm) N02 or 785.2 yg/m3 (.4 ppm) 03 and 13731 yg/m3
(7.27 ppm) NO £ for 17 hours before infection or four hours after infec-
tion. The lung was capable of functioning until the higher pollutant
level was reached. Physical clearance was not affected at either level.
The injuries demonstrated were those that would be expected from each
individual oxidant, so no synergism was noted.
In most of the foregoing studies, the authors recognized the pos-
sibility that damage to the alveolar macrophage was primarily responsible
252
-------
for enhanced infectivity and lowered bactericidal response. This func-
tional damage to the macrophage could occur through various mechanisms.
First, when rabbits are exposed to 981.5 yg/m (.5 ppm) for eight hours/
day for seven days, they exhibit increased membrane fragility, even
122
though lipid peroxidation was undetectable at this level. Since
phagocytosis begins with particle attachment to the cell membrane, this
fragility could be involved.
In normal animals, alveolar macrophages are a renewable population,
so studies were conducted to see if this process continues after a six-
hour exposure to 981.5 yg/m (.5 ppm), 2159.3 yg/m (1.2 ppm) and higher
levels of 0 . Aging mice (18-20 month-old males) were given injections
of tritiated thymidine to mark alveolar cells undergoing DMA synthesis.
Inhibi-ion of DMA synthesis occurred at all CL levels about equally,
representing miximum inhibition. This depression was visible up to 24
hours after ozone mice were transferred to clean air. ENA synthesis
124
did not completely return to normal until 72 hours later.
150
In a recent study, Fairchild reported on the respiratory depo-
sition of vesicular stomatitis virus in mice after exposure to 1766
yg/m (.9 ppm) 0,. This exposure produced a 70% increase in respiratory
viral deposition. The increase was also due to increased deposition in
the nasal cavity. No changes occurred in the lung or trachea. This
effect was particularly important since the respiratory volume in 0.,-
exposed mice was approximately 201 less than in controls. The author
suggested that 0, may increase the severity of infections caused by
organisms that primarily exert their effect on or through the nasal
mucosa.
253
-------
5.1.5.1.3 Ozone induced tolerance- -Ozone induced tolerance to pulmonary
edema has been reviewed in prior documents. ' However, other types
of tolerance have been recently investigated. It is interesting to note
that when mice were exposed to 1963 yg/m (1 ppm) 0, for one hour and
then X- irradiated to 800R after 10 days, 60% of these pre-exposed ani-
mals survived even though all non-ozonized mice died. This tolerance
125
lasted at least 30 days.
126
Matsumura " investigated the effects of oxidants on induced
allergic respiratory disease in guinea pigs. When the animals received
a one-hour pretreatment of 1963 yg/m (1 ppm) 0,, no allergic responses
to inhaled acetylcholine were visible after a later exposure to 3926
(2 ppm) 03.
The mechanism of tolerance development to 0,- induced edema is not
well understood, but the thymus appears to be involved. Male mice,
thymetomized at birth, were unable to develop tolerance to a two-hour
dose of 39260 yg/m (20 ppm) 0., even though they had been pre-exposed
to 588.9 yg/m (.3 ppm) of 0, for one hour. Sham-operated animals
148
exhibited tolerance under the same oxidant conditions.
5.1.5.1.4 Pulmonary function effects- -Ozone- induced changes have been
the subject of many recent reports.
1 OO
Moorman et al. investigated ozone decomposition in the naso-
pharynx of beagles by exposing the animals to the gas and measuring
mean tracheal concentrations of ozone. One group was chronically
exposed to 1963 yg/m (1 ppm) for three hours/day for 18 months.
Another group received an acute one-hour exposure to the same gas con-
centration. The chronically treated dogs had higher tracheal levels
254
-------
of O^ due probably to a reduced ability of the upper respiratory system
to remove the pollutant.
144
In another investigation, the right lung of rabbits was exposed
to 1963 yg/m (1 ppm) 0., for three hours. The left pneumatotic lung
was used as a control after the air pressure was returned to normal.
The exposed lungs of rabbits sacrificed one and three days after exposure
showed reduced vital capacity. Those studied seven days after exposure
had only a slight, non-significant, reduced capacity.
145
Yokiyama also studied the ventilatory functions of guinea pigs
before, during and after a two-hour exposure to 981.5 yg/m (.5 ppm) CL.
The following results were noted: air current resistance increased,
frequency of respiration increased and tidal volumes decreased. Com-
parison studies were made with SCL and NCL and differences were found.
S0? produced a decrease in respiratory rates, while N09 and CL had the
i_l L* O
opposite effect.
5.1.5.1.5 Extra-pulmonary effects--Most 0, studies have involved pul-
monary effects. But some interesting investigations have shown that
0, can effect other organs and physiological functions as well.
Gardner et al. exposed mice to 1963 pg/m (1 ppm) 0, for three hours
up to seven days and injected pento-arbital sodium immediately after
exposure. There was a significant increase in the sleeping time of the
exposed animals after two and three days of 0, exposure. After one day
and after four or more successive exposures, there was no effect. Pos-
sibly, this resulted from development of tolerance.
The authors hypothesize that increases in sleeping time might
result from formation of ozone reaction-products that could damage
255
-------
hepatic microsomal enzymes that detoxify pentobarbital. The "tolerance"
could result from an increase in induction of these enzymes. The actual
mechanism of 0, action is now under investigation.
Circulating blood is also affected by ozone exposure. Female mice
inhaled 1688.2 yg/m (.86 ppm) for eight hours/day, five days/week for
six months. At this time, the animals were infected with Plasmodium
berghei and exposed mice exhibited increased parasite infestation and
increased mortality. Additionally, up to the sixth week of exposure,
there was a significant increment in acid resistance of the erythro-
153
cytes. Other investigators showed an increased resistance to hemo-
lysis in erythrocytes in mice exposed to 1963 yg/m (1 ppm) for 30
- «. 154
minutes.
The effects of CL on pulmonary arterial pressure was investigated
in dogs. Eight animals were exposed to 1963 yg/m (1 ppm) for 17
months and for varying hours per day. Three dogs developed pulmonary
artery hypertension and nine dogs developed excessive systolic pressure.
Since there was no proportional relationship between pulmonary artery
hypertension and oxidant exposure, it was suggested that the results
seen were caused by genetic susceptability.
Matsumura exposed albumin sensitized guinea pigs to 1963 yg/m
(1 ppm) 0., and found no increases in anaphylactic attacks after subse-
quent antigen inhalation.
1 £*f-\ X
Roth exposed rats to 490.8 yg/m (.25 ppm) 0, for two hours and
found no differences in gastric secreto-motor activities, even though a
temporary effect was present at higher levels.
Atwall, using light and electron microscopy, found histological
and ultrastructural changes after a four-eight-hour dose of 1472.3 yg/m
256
-------
(.75 ppm) 03 in the parathyroid glands of rabbits. The major changes
found in exposed animals were: large number of secretion granules,
hyperplasia of chief cells, proliferation and hypertrophy of the rough
endoplasmic reticulum, free ribosomes and mitochondria, lipid bodies
and an accumulation of secretion granules inside the vascular endo-
thelium of the blood stream.
This increased activity does not abate until 66 hours after expo-
sure. The damage could produce alteration of parathormone production
and storage.
Certain radiomimetic properties have already been documented for
0, and further studies have been made. A review article by Veninga ~
presents evidence for the radiomimetic properties of CL and includes
some of his own research in the presentation.
C. Black Mice were treated with 392.6 yg/m (.2 ppm) 0, for seven
hours/day, five days/week during gestation and then for the first 13
weeks of life. 0.9% of the normal newborn mice had unlimited incisor
growth while 0, exposed animals had an incidence of 5.4%. In the same
species, neo-natal death rose from 9% to 34% after treatment. Addi-
tionally, blepharophimosis in the normal mice occurred in .6% of the
animals while it was 9.6% in ozone groups.
Similar, but less dramatic, results were seen in another species
of mouse where the rate of blepharophimosis rose from 4.5% to 9.2%.
142
In another study by the same researcher, binucleated lymphocytes in
murine blood doubled in number after exposure to 392.6 yg/m (.2 ppm)
for two hours.
5.1.5.1.6 Smog--In the environment, 0, co-exists with many other pol-
lutants that can modify the actual 0, effect. Some well-defined studies
257
-------
on pollutant combinations have been reported elsewhere in this manu-
script. This section will deal with the effect of synthetic and actual
smog on animals.
123
Emik exposed various laboratory animals to ambient California
air over 2.5 years. The average ambient concentrations were .057 ppm
oxidant, 1.7 ppm CO, 2.4 ppm hydrocarbons (as carbon), .019 ppm NCL,
.015 ppm NO and 4.2 ppb PAN. The following results were obtained:
reduced pulmonary alkaline phosphatase in rats, reduced serum glutamic
oxaloacetic transasminase in rabbits, increased pneumonitis in mice,
increased mortality in male mice, reduced body weights in mice and
decreased running activity of male mice. When older guinea pigs that
lived in smog for two years were exposed to 981.5 yg/m (.5 ppm) 0_
for 10 minutes, they had lowered deleterious responses to 0, compared
o
to animals that existed in clean air prior to the same 0, exposure.
The authors suggest that a possible adaption occurred. Pulmonary
tumors were found in the lungs of some mice, but there was no signifi-
can induction of lung adenomas in the two strains of mice exposed to
ambient, smog-containing air. And it was thought that this exposure
was probably near the threshold of effect.
In further attempts to produce synthetic atmospheres, Nettesheim
158
et al. exposed mice to ozonized, un-leaded gasoline. The maximum
concentrations delivered to the animals were 1963 yg/m (1 ppm) 0., and
24-30 ppm gasoline for 5.5 hours/day, five days/week for 127 weeks.
Some groups of animals were also treated with X-radiation (100R) and
infected with influenza virus. Several results were found: irradiation
raised the occurrence of lung tumors in females, but not in males;
258
-------
chronic ozonized gasoline exposure increased the lung tumor incidence in
both sexes (p<.05) and the viral infection appeared to retard lung tumor
growth. The only lung tumors found were adenomas and adenocarcinomas.
Furthermore, the investigators state that these experimental conditions
had no significant additive or potentiating effects.
159
Schlipkoeter et al. also investigated the possibility of syner-
gism using a variety of pollutant exposures to mice and rats. They
found that 1688.4 yg/m (.86 ppm) 03 and (50 and 100 ppm) CO had antago-
nistic effects on erythrograms.
5.1.5.1.7 Biochemical Effects--The actual toxicity of ozone at the
subcellular level is still debatable. However, several theses have been
advanced about the biochemical damage caused by 0,. They include the
following: (1) non-specific stress with the release of histamine,
(2) oxidation of sulfhydryl groups or their precursors, (3) oxidation
of polyunsaturated lipids contained mainly in cell membranes, (4) for-
mation of toxic compounds (ozonides) through reaction with polyunsatu-
rated lipid, (5) formation of free radicals, (6) injury mediated by
a pharmacologic action, i.e., via a neurohormonal mechanism. These
mechanisms have been discussed in the prior document. ~ ' Only new
or substantiating material will be presented here.
Many experiments have been conducted to find the effect of 0, on
-i i?/- *7
lipid perioxidation. Chow et al. exposed rats to 1472.3 yg/m
(.75 ppm) Ox and found an initial decrease and subsequent increase in
O
the activities of glutathione peroxidase, GSH reductase, G-6-PD, 6-P-GD
and pyruvate kinase. They reasoned that this pattern was probably due
to increased needs for reductive detoxification of lipid peroxide.
259
-------
Cortes prepared a methyl linoleate ozonide and gave it to rats
by IV injection or oral dosage. This treatment was lethal because of
acute lung congestion from edema. Significant changes also occurred in
the fatty-acid composition of serum and lung lipids. Other experiments
where rabbits were exposed to 1963 yg/m (1 ppm) 0,, for four hours
showed a decrease in the rabbits of incorporation of fatty-acids into
lecithin. Pulmonary lavage from thse animals indicated increased
activity of lecithins.
The authors observed that 0, may affect the lung by decreasing
lecithin formation while simultaneously stimulating the release of sur-
factant lecthins. Since surfactant serves to protect the host, this
132
lipid damage could have harmful consequences. The formation of pul-
monary phospholipid was investigated in another series of experiments.
When rabbits inhaled 981.5 and 1963 yg/m (.5 and 1 ppm) 0, for one-two
weeks, the amount of C taken into sphingomyelin, lecithin, phospatidyl
pthanolamine was significantly reduced. The same researchers further
studied these phenomena. Under exposure to 981.5 yg/m (.5 ppm) 0,,
14
rabbit lungs showed decreased incorporation of C into lecithin until
three hours after exposure when this parameter returned to normal. The
reduction in phospholipid formation was more pronounced at two than one
week exposure.
That lipid oxidation is a significant cause of toxicity is further
supported by many experiments into the role of Vitamin E, an antioxidant.
129 3
Roehm et al. exposed rats to 1963 yg/m (1 ppm) 0., and this resulted
in 100% mortality because of pulmonary edema. Fifty percent of the
alpha-tocopherol depleted group died in 8.2 days while the vitamin-
260
-------
supplemented rats did not reach a 50% lethal time for 18.5 days. Con-
current studies were carried out with 18810 yg/m (10 ppm) NCL for four
weeks and comparisons indicated that 0, was more toxic and less sensi-
tive to protection by antioxiclants. Later investigations by the same
group revealed that rats continuously exposed to 1963 yg/m (1 ppm)
0., for nine days show a significantly shorter 50% lethal time and greater
symptoms of respiratory distress when they were Vitamin E depleted.
When exposure was lowered to 981.5 yg/m (.5 ppm), pulmonary edema
and higher mortalities were noted after six weeks in animals with diets
depleted in alpha-tocopherol. The lung weights of the alpha-tocopherol
starved rats were also different from those of the diet-supplemented
groups. After six weeks of 07 treatment, significant changes were seen
in the fatty-acid compositions of total lung lipid. The most extensive
change was an increase in arachidonic acid in both diet groups. But the
increase was higher in Vitamin E depleted rats. This effect did not
occur until late in the exposure when it was associated with the onset
of pulmonary edema and death.
Unlike lung-tissue lipids previously described, fatty acids from
endobrochial saline lavage were affected little by 0, or Vitamin E. In
both exposures, the onset of edema was associated with elevated levels
of arachidonic acid and another fatty acid (22:6). These two compounds
are important constituents of cell membranes, so that could be the site
of ozone damage.
1 f-\ "7
A review by Shakman seems to support the idea that ozonolysis
causes peroxidation of unsaturated fatty acids of the cell membrane.
128
Additional data from the ssme group indicated that increases in
261
-------
arachidonic acid and decreases in oleic and linoleic acid after 65 days
of 981.5 yg/m (.5 ppm) 0, exposure may be responses of the lung to
oxidant stress because there was also a depression of serum-reduced
glutathione and tissue sulfhydryl compound contents. However, the
changed lipid levels could also reflect other metabolic roles of
Vitamin E in the bio-synthesis of polyunsaturated fatty acids.
These authors also found concomminent depletion of Vitamin E from
erythrocytes of CU exposed rats in 23 days vs. 36 days for clean air
127
animals. When others studied the continuous effects on rats of CL
from 1570.4 yg/m (.8 ppm) to higher concentrations for seven days,
they too found that alpha-tocopheral protected the animals from severe
toxicity (mortality) and weight change. At low levels, Vitamin E pro-
tection against lung peroxidation was a reciprocal function of the
logarithm of dietary alpha-tocopheral.
In a review, Shakman also considered the protective anti-oxidant
character of Vitamin E. Since varying dietary conditions exist in the
population, perhaps some malnutrition states would represent an impaired
group in terms of ozone exposure.
Ozone may affect not only the cell-membrance unsaturated fatty-
acids , but may damage acetycholinesterase (ACHE), an enzyme bound to
141
the cell membrane. This was demonstrated in vitro by obtaining ACHE
from bovine erythrocytes and exposing it to various concentrations of
CL. The threshold level was 588.9 yg/m (.3 ppm). Above this level,
the enzymatic activity was inhibited.
A whole variety of enzyme activities have been investigated after
0, exposure. Rats exposed to 1570.4 yg/m (.8 ppm) 0, for seven days
262
-------
exhibited increased activities of the pentose shunt and glycolytic
* *u , 136
enzymes of the lungs.
Marine blood and organs were assayed after exposure to 392.6 yg/m
(.2 ppm) 03 for periods up to eight hours/day, five days/week for three
weeks. The following changes were noted: increase in serum glutamic
pyruvic transaminase (SGPT) after two hours of treatment indicating
more fat deposits in the liver and increase in hepatic ascorbic acid.
142
There was no change in blood catalase, a strong reducing agentc
A slow, but steady rise in serum protein esterase was also found
after rabbits were exposed to 785.2 yg/m (.4 ppm) for six hours/day,
five days/week for 10 months. " For 10 months, rabbits inhaled
785.2 yg/m (.4 ppm) 0, and concentration of serum trypsin protein
esterase trippled by the end of treatment. This rise may reflect
140
thickening of small pulmonary arteries that were also seen.
139
Mustafa et al. studied mitochondrial oxygen consumption in rats
exposed to 1570.4 yg/m (.8 ppm) for 10-20 days. A 451 increase in
lung cellular oxygen utilization was found by the 20th day of the
experiment. There was also a 20-30% increase in the oxidation of 2-
oxyglutarate and glycerol-1-phosphate. And there was a simultaneous
increase (3 fold) in the number of Type II alveolar cells that are
rich in mitochondria, suggesting the possibility that increase in CL
LJ
consumption was produced by proliferation of these cells. At 392.6 yg/m
(.2 ppm) 0, for seven days, there was a 17% increase in oxygen consumption.
143
Veninga's work, where rabbits were exposed for 60 minutes to 392.6
yg/m (.2 ppm), showed a small but significant drop in total blood sero-
tonin immediately after termination of CL. This reduction was probably
263
-------
due to a loss of platelet-bound amine since there was no measurable
free circulation plasma serotonin.
In still another study, rats were continuously exposed to 1374.1
3 3
yg/m (.7 ppm) CL for five days or 1570.4 yg/m (.8 ppm) for seven days.
The activities of various lysosomal hydrolases were measured in whole-
lung homogenates and found to increase significantly. That was attri-
buted to an infiltration of phagocytic cells into the lung in response
to CL. The higher enzymatic levels may also reflect greater cell-
membrane lability. And the increase in protease and peptidase activi-
129
ties could be related to chronic obstructive pulmonary disease.
Damage to lysosomal membranes was also suggested by others after
histo and cytochemical studies into effects of 1374.1-1570.4 yg/m
134
(.7-.8 ppm) 0, continuously for seven days. These researchers saw
increased acid phosphatase activity with no change in B-glucuronidase
reactivity in alveolar macrophages of rats. The terminal-airway
epithlium and adjacent structures also showed increased acid phosphatase
activity. This particular enzyme had a nonvacuolar and an intracellular
distribution, indicating lysosomal membrane damage.
In another publication, the same authors reported on an examina-
tion of similarily exposed rats for enzyme reactivities. In areas of
bronchiolar epithelium infiltrated by mononuclear cells, there was a
lower MATH- and NADPH- diaphorase activity and increased ATPase activity.
However, intra-alveolar septa in the centriacinar region had increased
diaphorase, LIH, ATPase and cytochrome oxidase activities. These
increases were also seen in septa that were not thickened or infiltrated
with increased numbers of cells. It is suggested that these enzymatic
264
-------
changes are parts of protective-adaptive mechanisms that operate mainly
in the centriacinar region of lungs.
Trams et al. found no lipid peroxidation in brain tissue of dogs
chronically exposed to 1963 yg/m (1 ppm) 0,.
5.1.5.1.8 Behavior Effects--Several studies have shown the effect of
ozone on behavior patterns in animals. When rats were intermittently
exposed to 0., between 588.9-981.5 yg/m (.3-.5 ppm) and clean air over
a two-month period, they required a longer time to learn specific tasks
and their learning was quite unstable compared to controls. And they
exhibited reduced ability to discriminate temporal events such as judg-
ing lengths of time.
Xintaris et al. studied the evoked response to flash in the
specific visual cortex and in the superior colliculus of rats in response
to exposures of 981.5-1963 yg/m (.5-1 ppm) 0., for one hour and found
depression of responses. He suggested that altered evoked responses
reflect possible morphological or functional change in the central
nervous system. Other researchers exposed rats to 98.11, 196.3,
392.6, 981.5 and 1963 yg/m3 (.05, .1, .2, .5 and 1 ppm) 03 and found
that the greater the concentration of gas, the greater the depression
in gross motor activity.
The pollutant-exposed animals also had prolonged periods of inac-
tivity and an increase in urination rates. In another series of
experiments, sub-human primates showed an elevation of both simple
and choice reaction time in atmospheres of 981.5 yg/m (.5 ppm) 0_
after only a 30-minute pre-exposure to this concentration. When
127 3
Fletcher continuously exposed rats to 1963 yg/m (1 ppm) for seven
days to Ov he found an 84% reduction in voluntary activity.
•O
265
-------
to
0)
•H
O
o
ft
•p
•H
n}
o^1
to
8
CD
10
§X
•H P
co co
CO
O
£
Q
W
1
CO
B
55
to
p
o
<1J
w
13
g
0)
to
CO
rH
3
^8
T3
t/) -
O fi -H
X rH H
O oJ
\D
i—I
LO
0>
rH
•8
m
O ]
3 \
DO
§
W
O
VO LO rH
o"1
LO
oo
266
-------
b
iz:
o
tsi
O
UH
O
e
&
w
t~^
rH
LO
0
rH
•8
ed Effect(s) Species Ref
1
in
O
-4H
O 0
rH
jq 3
4-> m
DO O
CD r
1
P
U
0
m
m
CD
o
^_j
0
p
Lj_l
ctf
•
(/)
£_J
^f^
"^
0)
V)
Co
CD
o
>s
•H
•H
P
£
rH
•H
O
•H
rH
0
P
O
cd
OQ
CD
(_J
f~*
cd
•H
^1
0
HU>
U
cd
x>
ical clearance of rabbits 146
i ron parti cl es ,
c3T3
•y i
i p
p 0
O rH
0 >,
MH P
m in
0 X
rH
O O
^^
crS
rc3
-"x.^
t/) •
?H C/)
J^ O
£
0
1 — 1 t-~
rH
CN1
rH
bO
•H 0
bO-H
cd g
•H
in
•H
I
MH
O •
in
in I-H
0 rH
•P 0
Cd U
rH
T3 Cd
0 rH
rH O
0 0
f >
O rH
J cd
•
t/1
rH
t~J
vO
Oi
rH
rH
P
•H
XI
1
rH
cd
rH
0
£
rH
cd
m
o
•H
rH
•H
tl i ^
CD
"^j bo
0 cd
a^i
CO OH
0 O
rH rH
O O
HH 1
in
X
c?
[*-^
X
.
in
rH
<~J
OO
ivity inhibited by 0,, mice 117
ilicotic condition
p in
u
cd T3
3 3
"TD "^3
•H PI
0 -H
•H
£c3
p
O P
cd 3
PQ X
.
t/)
JH
r^
^j-
d to 7.27 ppm NCL mice 118
Physical clearance
t bactericidal activity
sm noted.
^
0
rH
O
rH
O
Pi
0
cd
rH
OH
in
0
0
o
in
rH
cd
0
u
.p-J
*
rH
O
'£
OH
.
t/1
rH
XI
r~-^
rH
3 -H
X^ rH
rH •> 0
3 0 X
O P W
0 O
g 0 O
P ' I 1
rH Cd
3
S +->
oo
oo
01
oo
CTI
LO
rH
OO
LO
oo
LO
00
267
-------
oo
fM
to
w
u
a
£
Q
8
3
g
I-H
a
o
CO
rH
*
LO
CD
r—t
g
t/)
0)
•H
0
CD
PH
CD <1)
0 O
•g -a
00
^-^
t/)
^ — /
P
o
CD
<4H
w
•8
CD
I/)
,£>
0
o 1 o'S e
w o o
O *T^J Qt
rCJ ^1) Qi {> -H rt i
rH ,£ • -H 0) «*P
TO ,5 O "o M O
•H CD !H iH
cB 'rt '"T ° .
CD CD -H O CD
N rH C W > •
•H O CD i— 1 o\<= -H T3
P Oi CD 'rl ^ "Tj
OOP g •
fi CD JH CD rH
X >
? 1 i * '•rt
•£3 VI CD 8 -H
•P o P P f-t
.J W
S rH fi'o
CX CD CD 'H
P P r<
O rt »-J DH
.
1
o x1
w
^S i d
^^ cu tn
OO r< O
^
0
s
to
. .
rC X
I— 1 1— 1
_J
r'n
en
•
CO (O
oo vo
rH O)
rt
CD
C tn
•H GO
Boex
o
•H
CD rH
rH CO
"rt q CD
P O i-H
P O
8 CD'O
cB c ?•?
r< O .C
CD D,P
rH t/> CD
O
-------
m
en
to
•et
LO
o
LO
LO
10
LO
LO
CTl
rH
LO
0
rH
•a
tn
0
•H
O
0
PW
CO
in
o
0
m
w
JH
0
in
3
m
o g
p in
bo o •
g£
I
Pu
to
o
to
to
o
0
o
•H
S
Cfl
cfl
-§
p
0
B
0
^
•H
&
0
rH
tn
0
JH
U
d
Td
43
to
to
CTi
bO
3
d in
to
\o
CTi
0
O
in
•H
in
X
0
0
O
cB
p
in
•H
tn
%
0
(J
• tn X tn
tn d jo d
O O TJ -H
e -H \ e
JH tn
f^- Cfl JH O
rH > 43 tO
to
\o
CTi
cO
0
•H t>0
3 -H
DO PL,
tn
£
in
0
0
JH
U
0
&
2 g
43 ,-£3
p tn
^
P
o •
C in
o 0
P 'Td DO
aJ
cO 0
bO
'Td -H
0 -p
s*
•H JH
10 0
f3 P
0 <4H
'fO
'-O
CTl
0
U
tn
0)
•H O
P ^
O x!
0) P
0)
£ °
cO U
x§
P 4->
•H (/)
rH -H
cfl tn
p 0
JH JH
Pi 0
o in
•H Cfl
P 0
Cfl JH
P O
0 rH
Pi
•H
T3 0
0 43.
in po
cO
0 0
b-~
in
in
OO LO
vD
OO
OO
OO
1
o
in
0
Pi
rt)
DO
CO
rH
> <
rO
0
£
6
43
in
g
•H
P
cO
Pi
£5
X
0
nj
£^
Co
i— I
00
Td
•H
O
JH
J^|
CO
JH
CO
CM
Is, prolifera-
ticulum, free
rH 0
0 JH
(J
O
m -H
0 6
•H 10
U rH
Pin
o ro
f4
cO 0
•H
tn 43
CO bO
rH 3
&2
0
PL, "4-1
X 0
accumulation
T3
g
«t
(/)
0
13
O
Td
•H
Pin
•H
rH
•H
Td
Pi
43 0
X43
"43^
Q.) O
i-H M
£5 +•*
C^ ^H
c3 0
JH CLj
t>o X
cJ
o Td
p 3
2d
U 0
0 -H
in P
o
^J
•H
E3
Td
Co
t/)
i
o
in
o
•H
JH
0
in
0
•
•H
rH
0
i_J^
P
O
Pi
0
JH
CO
rH
rj
(_J
in
cO
j>
0
•H
tn
p|
•H
tn
0
rH
^3
rH
CO
JH
M
CJ
Q
•H
•P
JH
O
0
in
o
0
tn
7^
0
J_<
0
P
MH
cO
tn
JH
j3
Q
^ <~|
\o
rH
g
^4
O
d
0
*"^
0
IP
(D
^
0
JH
in
0
CO
-5
P
cfl
JH
&
•H
•H
P
U
CO
JH
O
•P
Q
pJ
I
Q
p
0
JH
O
0
tn
o
•H
JH
P
tn
cfl
DO
Pi
•H
tn
0
o
O
J-(
MH
MH
•H
O
z
0
o
rH
O
d
1
0
0\°
CTi
d
•H
d
0
0
10
l/J
Co
£
43
p
^J
2
DC
JH
O
in
•H
U
d
•H
Td
0
p
•H
3
ncreased from
•H
43
P
cfl
0
iH
cO
p
cO
Pi
O
0
.
0
O
•rH
^
ITS
0
p
cO
0
JH
P
0
'4_i
MH
0
^
ccurrence to
o
oNQ
\o
,
g
JH
MH
0
in
in
•H
tn
o
G
•H
43
&
JH
CO
43
PL,
0
tH
pT)
•
o\=>
O
p
0\°
CTi
XrX
tn
JH
OO
i
^
LO
r--
to
•
cxi
r^~
*^j-
TH
f
in
JH
43
CXI
LO
(XI
•
OO
•
o
CTl
^f
Co
*"O
**»>^
t/J
J_4
43
^
cxi
•
^•O
*
CXI
CTi
to
^
's-v^
in
M
•^3
LO
another
^
•H
o\°
CS1
•
CTi
O
P
cX0
LO
•
*^"
6
o
^
r^
f^
cd
tn
0
•H
U
0
PU
10
0
Pl
0
Pl tn
•H 0
•H
o\° U
0 0
• PH
CTi 10
Pi CO JH
O rX -H
•rH [2 ,Q
•H
1
JH
8
m
cO
JH
0
d
d
•H
Td
CD
rH
^3
Q
"nJ
*x3
o
o
rH
43
d
•H
in
0
p
t_)
o
43
pq
£»^
rH
T3
0
P
CO
0
rH
^
•H
m
•
2
^3
0
$
0
DOP '
. Pi
•H
JH
^J
Td
CO tO JH
-P 0
10 Td P
0 d '•H
DO CO CO
in
JH
43
cxi
rxi
vO
•
cxi
CTl
to
269
-------
LO
LO
LT>
ipecies
0
o
•d
82
•H CO
S
tfl
co rt
co
0 bo
0 Pi .H
U -H p,
CO
U
c/5
§
£
HH
W
o
CV]
LO
0
i—i
•8
H
co
V - /
•M
U
,0
M-i
m
13
0
0
CD
^CO +3
0 N O
a PI^CI
•H 0
rH 3 C
O rH -H
CO 4-1
cd p| 0
bO-H O
SX 0
Pu-P 13
PVH.H
^O 13 -H
• 0
«* P fn
CM U Q
^£ i
0 P! P
bO
•H
rH CO
O ^
bO+->
TJ°
^
g^B
SP"H
co 0
id -H p;
0 'd -H
> CO iH
•H ^H O
0 I 10
o X nj
0 bo
f-i 0
?-i 13
to 0 0
rH |2 Nl .
g -H CO
•H § g X
Pj O Nl 0
<; co o to
CD
L>
8
I
O
o
13
O bO
LO O
s;
• HO
Si3
•H O
r£} 0)
S 'M
o m
o 0
Cj U
•H -H
4->
to to
0 -H
S g
8 cf
P! PI to
•H -H <-»
PI to bo as
O -H -H 0
•H 4-> 0 Pi
LO O fi
0 3 O
!H 0 4D
-s-^'e-g
•H 13 0 3
X 0 to 13 Pi
O 4-> ro 0 -H
0 ^ _
•H tO P! tO >H +J
*Q P; -H CO O PL,
SO 0 0 cO
co JD ?H 13 i3
!H CO O CO
O cS 0 C ^
+•> o S -H e 13
ox co 0
0) fn
fi
. _ p| 0 0 5
3 XS U U&
to j3 -H .H to
O| «\ fi ^jQ
PT o o c c -H
WO to -H -H Ct,
tn
I
•p to
bo o
Sx^
-3 W
S
a
PL,
K>
0
& trtr^
^! X £
LO Tj t~^
(NJ
LO LO rH
rH
PI
•H
•P
c
o
o
o t^
OO LO
• o
V — 1
.
to
!-.
X
CXI
/ — \
e
Pv
S
bO
vO
oo
00
g
•H bO
270
-------
LO
CXI
CD
•H
O
CD
CO
bo
-§
2
O
P
Cj
§
pTj
/— ^
^ rH
0 10
£^
^, rH
•tn §
0
• o
p *"O
•H 0
U rH
S
PH r«(
•H Oi
a
t
0
rH
'-H .
0
'"CJ £
fH ^
CO rH
O
0 >
0
p 'S
V) -H
•H -P
co
0 Pi
rH -H
P 0
C t/3
0 co
rH 0
rH rH
3 0
'O
!H
•H Cti
cd
r^
s s
•H CO
0 g
(/) O
CO -H
0 -P
rH CO
U rH
G-H
•H PH
W
rH
rO ^H
0 O
l3 ^-«
tS 0
o rt
CD cr
00 O
f.
2
rG
00
OO
• V)
rH rH
XI X
I/I
I
LO
"bo
to
to
\Q
01
to
\Q
CTi
LO
rH
00
271
-------
CNl
CN1
LO
0
4-1
0
CT> O 00
CM to LO
&t rH rH rH
0
•H
(J
0
P,
P
•H
t*,
•s
CO rH
to
u
0
4-1
4-1
W
'd
0
rH
0
tfl
rQ
O
•p
P
o
CO
4-1
rH
3
4-1
0
0
cS
0
rH
s
T3
i
PI
o
•H
p
cO
,0
4-1
licithin
.
**O tfl
0 C
ro f*^
0 -P
rH -H
u o
0 0
H rH
4-1
O 0
A 3
4-J V)
bO O
X$
G
a
to
o
to
E
bO
3.
to
O
3
B to
* P« r-^
tfl -H ^
X PJ fM
•3- U i— 1
LO
rH
LO
tO rH
vO 00
CT> C%
i— 1
oo r^
CM CM
A A
\O LO OO
CM rsl CM
rH rH rH
to
ro
rH
0
go
(fl -H
pg
tfl ^O
S P
S-bVS
13 -H -H
'x^ o
O - rH
•H X 0
+-1 P PH
c3 rH 13
ro -H
Cj ^"** O-i
« JH -H
O rH
Co ^^-^ ' *^3
t/) (/) Co
•^ c
0 5 P
CO -H P|
PJ CO P
P: 0 o
•H P U
4___1 ,__i
•H tOU
> O rt
S
|
Pi tfl tfl
•H X X
U 01 r--
LO OO
rH
LO ^f
tO rH 0
\Q CO t-~
cji o~i LO
rH i— 1
O
vO
i— 1
bO
•Q
•
0
3
tfl
(0
•H
P
P!
•a
rH
P!
•H
P!
O
•H
peroxi dat
•H
PU
•H
i— 1
0
^
0
•H
Pj
g
6
t— i
to
^o
CTl
1— 1
to
to
I— 1
(fl
co
rH
o
•H
p
x
rH
0
o
x
rH
bO
1
P
1
to
0
(fl
3
PI
0
o
to
0
activiti
f lungs.
Tj O
0
to to
co 0
0 S
(J N
Pi P!
i— i 0
tfl
X
•3
00
*
^J-
o
t —
LO
I— 1
vO
to
rH
tfl
4~J
C\J
rH
13 1
1 -P
O O rH Pi
P 4J CO 0
•H CO rH O
S rH O Pi
CO 0 O
bO-P > U
Pi 3 rH
3 I— 1 CO tO
iH 00 O
0 X^rH rH
& O 0
P 1 0 S
. CNJ cL, o
Pi XrH
-H 4-1 i?
O 0
0 4-1 ^
tfl pj O -P
ro o
D .H rH 4-J
?H .p 0 ro •
e r§ •§ r;-) g
'H '* 1 §, T>
O Q,
Pi P! **"! • 5
O -H +3 cO tfl
•H -H p|
•P PIT5 rH 5
ro o c TJ u
rH -H f3 Pi
•P -P O (Nl
f0 ^ C^
•p (J
co q PI
PI tfl & -P -rH
O Pi Pi-r-l
0 0 tfl S 0
rH°^fir3
0 (Nlp^.H 0
bO i— 1 -£i O
•H rH l TJ Pi
,C CtJ rH -H -H
•H O rH
0 rH rH
,£2 TJ 0 tfl ••
1-3 P! O rH Pi
O XrH O
< U bOO +->
tfl
t
o
1
o
I— 1
oo CM
*
** *°.
o f^J
t — . o^
LO tO
rH
to
•S
I— 1
to
I—I
tfl
ro
rH
%
tfl 0
.as
P TJ
•H
T3 -P
rH p>
ro 0
p)
tfl
0 T3 •
•p g tfl
co ro 0
PI OO
CD CD p;
bO (fl co
S 0 O
O P
fi 0 0
rH -H
bo p< +-1
Sal
r- 1 -H X
0 0 P;
rH tfl 0
o ro
•H h— 1 O
tfl _ r^
0 • -P
hydrolas
increased
along wi
rH PJ
ro tfl 0
S Pi 0
o o to
tfl -H
0 -P CD
tfl L) rH
X 0 0
H-3 tfl |£
to tfl
'S ^
LO t~~
t^
*
«3-
!*-»
to
1— 1
CNI
to
rH
tfl
ro
rH
0
(/)
Co
rH
O
-a
Sir
•H
rrt
-C
R
g
1
^i
|£
•H
rH >
0 -H
S t)
rH CO
rrj g)
ro to
f— | CQ
B <
•H
rH rH
•H -H
CO §
"o x
•H 4->
,d -r-l
rH U
PQ rt
B
§
PH t/)
•rH ^%
C 'S
CJ t^--
f-- oo
* *
rH ^f
**• o
^» t^.
to \o
rH f^^-
to
to
I— 1
tfl
-p
ro
rH
4-1 T3
°S
to
0 /-N
•P 1
r* 1
•rH CJ
P
O *
ro p
•H vO
0 O
tfl
cO "
0 0
rH tfl
O Cd
Pi P
•H 0
P 13
Pi 0
0 rH
CT* IE
0 CO
tfl Cj
-8 .
i j j_j *•*
•H 3 S
C rH X
rH 00 P,
LO
r~-
*
to
CM
f^^
«£^»
rH
272
-------
LO
IO
rH
to
p
•H
1
t*--
K)
rH
I/)
P
•H
X>
•a
oo at
to to
.— i i— i
o
f-,
4_)
•H
>
CD
c~j
•H
f-i
O
i— 1
in
p
•H
rO
•8
S
*
TO
LO
ase
»
CD
{/J
TO
(H
CD
P
to
CD
c
•H
CD
J_J
T^
8
PU
€
3
rH
0
w
Pi
•H
CD
W
s
tl.
CD
to
rt
JM
CD
P
to
CD
cj
•H
CD
O
p)t
g
CD
to
PI
•H
CD
(/)
TO
CD
0
r^
1 — i
.
rrj
CD
•H
X
•rH
^c^
"{~^
• H
~>
CD
O
trans aminase ,
in blood catal,
o
•H CD
> £30
*3 C
rH CO
^NrCj
p>^ O
u o
E
Co
P •
^3 rT3
iH -H
t30 U
s rt
1 0
M -H
CD ,0
t/) f-H
o
P! o
•H (/)
Co
CD
to O
Oj -H
cp P
0 PL,
PI CD
It
tonin.
o
CD
to
rr3
0
o
t-H
JO
i— 1
rt
p
o
P
•H
CD
(/I
TO
CD
f-i
U
4)
i— 1
1
p^
£j
CO
VI
to
to
O
'O
CD
^-1
I
LO
OO
(XI
LO
OO
oo
oo
LO
CM
'•O
1X1
^
IO
273
-------
M-t
&
0)
•H
O
0)
ex.
CO
*d-
CNJ
to
I
CM
LO
to
to
CO
b
^
w
r "1
§
o
£
^^
w
«
^
to
CM
LO
s
rH
0
^
p*
•H
fi
O
•rH
^_j
o
•s
8
o\o
OO
3
s
.3
4-*
g
o
rH
to
E
"S8
3.
to
0
to
\D
CTi
•
C/l
t \^
c§
^j
o
•H
•H
O
il
e
cO
rH
O
i
•H
4_)
f-!
0)
0
rH
CO
<1>
!H
•H
3
Cf
£
s
•H
|
0)
4-)
c
rH
to
en
oo
OO
LO
•
JC
(/J
C^J
o
^
w
s
3
5
f~
o
to
LO
LO
rH
oo
Cf^
cO
fH
-1
CD
18
(Li
rH
O
rH
^
4-)
•H
•H
4_)
O
CO
rH
O
^_)
g
w
to
o
rH
bO
M-l
O
f^ *
O C!
•H 0
to -H
tO P
JH f^
gi'H
(H 3
LO
O
rH
i— 1
OO
CPJ
274
-------
5.1.6 Total Suspended Particulates
5.1.6.1 Human Studies--The national primary air-quality standard for
total suspended particulates is 75 yg/m (geometric mean) when measured
as an annual average and 260 yg/m when measured on a daily basis.
5.1.6.1.1 Evidence for the standard--The current national primary air-
quality standards forparticulate matter are shown in Table 5.24. These
standards are based on evidence of adverse health effects ranging from
symptoms of increased irritation to excess mortality associated with
various concentrations and durations of particulate exposure (Table 5.25)
Although it was not possible in these studies to entirely separate
effects of particulate and sulfur dioxide, setting separate standards
for both was wholly consistent with the Administrator's legislated
obligation to provide adequate safety margins in setting ambient air-
quality standards.
Although these epidemiologic studies were the best available at
the time, their limitations were recognized. Chapter 11 of the Par-
ticulate Criteria Document fully and completely discusses interpretative
cautions.
In brief, measurement methods were not often directly comparable.
The physics of particle size and chemistry of particulate matter had
almost always been ignored in field studies and pollution and health
indices were not always measured over the same time periods. Few
studies identified or adjusted for all other known or possible morbidity
determinants such as cigarette smoking, occupational and other past
exposures, past medical history, socioeconomic status, residential
mobility and indoor pollution exposure.
275
-------
Table 5.24. U.S. NATIONAL PRIMARY STANDARD FOR TOTAL
SUSPENDED PARTICULATES
U.S. National Primary Standard
Pollutant Concentration (pg/m ) Averaging Time
Total 75* Annual geometric mean
Suspended
Particulates
260 24 hours
*Geometric mean
Also, almost no epidemiologic studies had been possible in areas
characterized solely by particulate matter. And almost no studies of
highly susceptible individuals had been done.
5.1.6.1.2 Interim Evidence--Subsequent evidence on health effects of
particulate pollution is sunmarized in Table 5.26. These studies were
also complicated somewhat by presence of considerable amounts of sul-
fur dioxide. Despite that complication, best judgment estimates for
the various health effects that have been associated with particulate
exposures are given in Table 5.27.
These data strengthened and extended the initial observations on
health effects. The CHESS studies were designed to indentify or adjust
for effects of other possible disease determinants in addition to
ambient air pollutants. The limitations of the CHESS studies differ
for the short-term and long-term health effects studies.
For the studies of short-term health effects, estimates of expo-
sure are more reliable. But the indicators used to measure health
effects require further quantitative refinement. For the studies of
276
-------
to
CD
O
0
rH
CD
»j
S
g
J < -H T3
Z Z |S >
rH H 4-> 0
H CO 03 to
p O
CO 2 4-J c3
* L* O3 ^
, — 3 CJ rH
h— ( 4-) 4-)
ril t— i rjj (J
H pi 00
f-H «^ (_) <4H
pi j r^ LJ^
oi 00
SQ U
w
p— •)
Is
M 3
s w
•H
H
DO
PJ
•H
r?
0
"^
-r
-^_
DO
3.
y -*
<~*
O
•H
4->
03
4^
§
§
o
w^
rH D
PH H
rH
g
TTT
1 1
O P
CO O
0
LO fU
LO 4->
rH
0 oj
rH 0
^Q H
03
to
0
vO
\o
r 1
r~ i
LO
vD
rH
co
O
oo
1
«^J-
CXI
0
CO
•H
03
rH O
O 0
CXI
o
LO <4-l
t-- O
^
4J
•H
rH
03
4J
rH
0
(/)
CQ
0
o
i — i
.
rH
t~~-
v£>
rH
CO
T3
CXJ
O
O
CX]
X
• H
rH
ctf
1 1
rH
^ to'
*~* •$
c3 crJ
m 0
Cj Tj
•H
•gS
o]§3
0 U
rH
& £
rH 03
.
CXI
O"l
\^^
r J
r~i
C30
rH
co
rH
|
^*
CXI
*
LO
r^~
K)
r7
Q
P
03
•rH
to
rH
^
0
13
I
JB
o
PJ
rH
«
to
.
£*
•H
!Q
s^
o
s
(J
•H
rH
03
g
•H
4->
O
0
MH
PI
•H
CD
t—t
to
j2
«^i"
rxi
O
o
CXI
B
>^
•rH
rH
ctS
4J
J-4
|
CO
•H
4->
•rH
•6
|
CO
co
g
.
*3-
rH
rH
>f
•H
03
Q
O
O
p!
•3
i
o
4->
E^
?*•»
to
o
CJ
orseni;
£
CD
P
%
m
LO
to
P!
0
4_)
03
CO
•H
4->
•H
'S
g
(X)
rH
to
0
X
CXI
0
o
rH
O
'p)
5
•63
03
g
OT3
(_j_) CD
X 0
rH 0
4-> -H
S fr
TJ O
0 4->
to 03 •
03 rH CO
0 -H 0
rH D-, CO
U CO 3
rH !H O
t
vO
LO
^
r^
to
rH
rH
3
O
O
CXI
1
0
o
rH
t>> • •* "^ rjj
rH CO CO
4-» rH 4->
c\3 ^Zl ?"^ 0^ CJ
^•H rt3 CJ t/) C«
•H o3 !2 o3 3
P^ Q (J_) C(— j
t/) (3 p~l C/)
0 -H -H X
Sp; os
c3 P pj
'S 0 fH 03 §
O rH 0 rH rH
rH H ^ 0
T3 fttJ O 0 rH
tO 0 tO 03 10 rH
O3 tO O3 rH -H -H
^-1 0 rH PVH U
o to y to s
B
I>-
•
/. — ^
rO
E
M
p-
LO
r-j
rH
II
r-
s
r— 1
.p
0
rH
V__^
c/>
4—*
. i— (
g
33
CJ
§
rH
rrj
4^
rH
0
CJ
*
277
-------
long-term health effects, previous exposures had to be estimated because
of incomplete exposure monitoring data. Hence, estimates of exposure
were somewhat less accurate. But the observed health effects were quite
consistent.
5.1.6.1.3 New Evidence--Few epidemiologic studies on long-term exposure
to total suspended participates have appeared recently. Those that have
appeared come from the CHESS program of EPA. Responses related to long-
term effects include chronic respiratory disease and lower acute respira-
tory disease. Recent findings from CHESS relevant to long-term health
effects of particulate pollution have come from studies in New York and
Birmingham-Charlotte.
Of the two sets of findings, those from Birmingham-Charlotte are
more directly relevant to particulate health effects since the SCL pol-
lution there is minimal. For that reason, the Birmingham-Charlotte
findings will be presented first and in more depth than the New York
observations.
* Birmingham-Charlotte Chronic Respiratory Disease
During November, 1971 in Birmingham, Alabama and Charlotte, North
Carolina, the rate of chronic respiratory disease symptoms for parents
of elementary school children was determined by self-administered
questionnaires. Survey results noted small, but not statistically sig-
nificant differences, in prevalence between the cities. Observed dif-
ference included the direction of pollution exposure. Birmingham was
higher. The equivocal nature of prevalence difference in this study
made it difficult to form firm conclusions about the relationship of
air pollution to chronic respiratory disease. It appeared that long-
278
-------
CXI
uo
0
rH
rQ
H
CD
O
PI
P
CD
Q)
H
00
Pi
•rH
00
03
rC rH
O CD
•H rO J>
P CD
03 to
tj ~S S
rH "~ 00
•P P 3.
PI U ^
CD 0
O MH p|
£H MH O
o w -H
U -P
03
rH
CD
O
Pi
0
o
P
U
CD
m
M
4J
CD
CD
to
rH
1
\&
rH
in
$~i
r-5
0
*^-
CXI
CTi
CXI
1
O
o
rH
*"O
CD
P
P U
t/) *H
CD rH
rC P
U (/)
CD
r^
?!
8 a
. p
t3 rH •
CD O X
rH O -H
U to 4->
PI -H 0
rH "O OS
rH
t^
r_^
•ft
^4
3
O
r
«^f-
CXI
(0
o
tH
I
O
<3O
10
Pi • ^
I -H fi CD U •
O -H fi V) CtJ OS
•H t/j P oS >P E
'•d E T3 -H CD P S
rH O pi f
rH
3
O
•^
^-^.
txi
o
vO
1
r^-.
i
O
t[f*
'4^
•H
O to > OO
• H C
T3 CD t/) 3
rn tn P rH
03 crj pi
U CD CD U
to -H -H
m -H P d
O T3 oS O
PI f*^ rH
O H X O
•H O rH
p p f_ VH .
03 03 CD O CD
> rH T3 (/)
03 -H rH P 03
rH P< 0 rH 0
OO t/) OS t/)
00 CD P{ 0 'H
<£ rH 'rH f~^ rrj
to
oo
F~~
rH
rH
^
3
^3
<
0
rH
rH
O
•H
P
(J
fj
M-l
OS pi
C 0
§ T3
rH -rH
d -H
p-s
creased
school
a .s
^
Ol
t^-
rH
rH
03
3
g
<
O
o
rH
TJ rH
pj CD
m ^ rH
5 -H
XtH
|0|
^3 £3 t/) •
a4 u -H p|
0 Cfl T) 0
rn rH
m m x^
O H rH
T3 0 -H
0 X+-1 ra
to P 03 U
O3 -H rH rH
0 rH -H O
rH 0 fO, O
(J > (/> J^
pi 0) 0 O
1— 1 tO rH t/)
uo
0
oo
rH
rH
aS
3
g
<
o
o
rH
1
a3
rH
•H
P<
0 S
rH 0
P
Pi ?^
O 10
rH
rC| 0
U t/1
o3
T3 0
0 to
CO -H
03 T3
0
rH X
0 rH
Pi 0
1— t P
\£>
•
tfl
•P
rH
03
1=1
•H
0
evalenc
rH
a,
279
-------
p
co rt
U -H
W P
K (/>
PH W
W
t-H O
3 CO W
is £
o p
p., o p
rH rt
CO H CD
S
tu
Q> J>
r0
rt rt
0 -H
•H
•P W
rt P
rH &
t*o S»%
•< en
r^
§
o
^
CXI
o
CXI
A
O
• H
8
•§
O •
0)
grt1
•H CD
P to
rt -H
rt
rH bO
GO S
< rH
^
to
rH
§
r~*
^
CXI
0
o
rH
rt
i
43
%
o
rt
o
•H
P
1
rH
W
<
^
^-
10
rH
O
O
O
o
rH -P
rt -H
> X
Ci) U
rH rt
4)
C^.^H
rH O
H^
o
o
280
-------
term exposure to total suspended particulates at 150 yg/m or less in
the presence of low sulfur dioxide was not strongly related to chronic
• * j- * 184
respiratory disease symptoms.
• Birmingham-Charlotte Lower Acute Respiratory Disease
The acute lower respiratory disease experience of 10,000 children
aged 1-12 years was determined in a retrospective survey of the chil-
dren's parents. During the 12 years prior to the survey, estimated total
suspended particulate exposure ranged from 103 to 168 yg/m in Birmingham
and from 72 to 107 yg/m in Charlotte while SO- was less than 25 yg/m .
Significant increases in pneumonia and croup among blacks and "any lower
respiratory disease," bronchitis and croup among whites, were noted in
Birmingham residents. Since increases were noted in children aged one
to four, the authors' best judgment as to the level of total suspended
particulate capable of producing an effect is between 103 and 139 yg/m
185
in the presence of low sulfur dioxide.
* Birmingham-Charlotte Acute Respiratory Disease
Panels of families in both Birmingham and Charlotte sent bi-weekly
reports of their morbidity status for 16 week periods in 1970-71 and
1971-72. Incidence of acute respiratory disease did not show a con-
sistent relationship with air pollution, although data on severity of
illness among residentially stable families did indicate that exposure
to high air pollution was associated with restricted activity and
physician visits. Best judgment estimates of air pollution effect are
333 186
90 yg/m TSP, 12 yg/m SS, and 3 yg/m SN for three years' exposure.
• New York Studies
Chronic respiratory disease rates among parents of elementary
school children was not elevated in areas of higher particulate and
281
-------
sulfur dioxide exposure. Complicating the interpretation of a unique
particulate effect was the presence of a previous study that did
observe an effect coupled with a long-term decline in levels of both
TSP and SCL and a short-term dramatic decrease in S02 between the two
187
studies. Lower acute respiratory disease was, however, shown to be
higher in areas of higher particulate and SCL exposure. The authors'
best judgment of the exposures associated with increased lower respira-
tory disease were 175-250 pg/m S02, 85-110 pg/m3 TSP and 13-14 pg/m
A A ir * 188
suspended sulfates.
Similar to the long-term health effect of particulates, few new
studies have investigated short-term effect. Studies involving panels
of asthmatics, elderly and families have been used to obtain data
relating to any short-term effect. Again, studies done in Birmingham
and Charlotte are more likely to be relevant to such effects than
studies done in New York.
" Birmingham-Charlotte Episodes
Acute irritation symptoms reported by 1500 families of elementary
school children were studied in Birmingham, Alabama and Charlotte,
North Carolina during 1971 and 1972. Following three designated air
pollution episodes (two only for Charlotte) and one control period,
telephone interviews were conducted and responses compared across all
time periods for each participating family member. Although Birmingham
residents did experience an elevated symptom rate in the designated air
pollution episodes, especially the first, Charlotte residents did not.
The rate of symptoms even in Birmingham did not vary with level of
189
pollution experienced by Birmingham residents across all episodes.
282
-------
* Birmingham-Charlotte Asthma
Panels of asthmatics; located in Birmingham and Charlotte were
observed from the fall to the spring of 1971-72. Aggravation of asth-
matic attacks was not shown to be related to TSP or RSP. Some rela-
tionship with suspended nitrates was noted. No reliable estimate of the
level of nitrates necessary to produce an attack could be made because
190
of analytic problems associated with air sampling.
" New York Studies
The New York studies add relatively little information to the short-
term TSP exposure data base. Adult panels of well and cardiopulmonary
patients show a TSP effect of between 70 and 120 yg/m 24 hour average.
These effects, however, occur in the presence of significant quantities
191
of sulfur dioxide. Studies of asthmatics showed no association with
192
TSP, but did with suspended sulfates and nitrates. The observed
3 3
threshold levels were 8 yg/m and 3 yg/m 24 hour values respectively.
However, there is a question about the accuracy of the analytic tech-
nique used for suspended nitrates so that the 3 yg/m level should be
considered with caution.
5.1.6.1.4 Discussion--Results from the previously discussed studies
have not significantly altered the state of the art knowledge of health
effects of total suspended particulates. Results from long-term studies
(chronic respiratory disease, lower respiratory disease and acute
respiratory disease) were unable, because of their design, to distinguish
between total particulates and respirable particulates. Similarly, these
studies were not able to explain the possible effect of chemical composi-
tion of particulates because of the nature of the study design. The
283
-------
studies, however, did further document health effects associated with
long-term exposure of about 100 yg/m (lower respiratory disease and
acute respiratory disease).
Short-term health studies were somewhat more effective in provid-
ing insight into the relative importance of respirable to total particu-
lates and possible importance of chemical composition. Studies of
asthmatics did not show any relationship between asthmatic attacks and
either total suspended particulates or respirable particulates, but did
with suspended sulfates and suspended nitrates. Aggravation of symptoms
in panels of adults was most often associated with suspended nitrates,
respirable particulates and total suspended particulates. There was
little difference in the strength of the relationship between health
effects associated with respirable particulates as compared to total
particulates.
Table 5.28 presents the best judgment estimates of effects from
particulate exposure based on the new evidence summarized above. In
general, few modifications are noted from the estimates presented in
Table 5.27 and most modifications are relatively minor. These results
seem to indicate that the annual average TSP standard of 75 yg/m (geo-
metric mean) is still quite acceptable, but some modification in the
24-hour standard may need to be considered because of the evidence from
the adult panels. Since the question of respirable vs. total and chemi-
cal composition of particulate matter has existed for a number of years,
definitive studies designed to find the answers to these questions are
needed and should be conducted in the near future.
284
-------
1
2 ^
rrt
^ ^
O -H
^_>
et/)
w
W O T3
PH H *o
M 10
§ ™ £3
2s
S 2 •*-*
oi z S
So g.
P-* DO
P T3
CO PQ |j
p ^
<; PQ -p
i-1 CO 0)
S3g «
KH
g
•H
f-H
(M
Pi
• H
bO
CO
f-<
§
^6
bo
3.
* '
•H
j_)
g
P
rt
CD
O
CJ
96
O I-H
to S
§ §
s
f-H f^ -fJ
pt"] ^_)
f— i C^ Q^
^^ t^ MH
FTJ y^ LJ_ (
s cu w
r^ i"^ ^p|
^ rH
CO
H CD
CO X
w
pq CD
to
00 >
LO
CD
!Q
10 (/)
fi fn
0 0
^1 rJ5
^t1 ^"
E?
f), ^.^
& V)
CO • PI
(/) O
MH rH 'H
O CD P
Pi CO
Pi CO -P
0 PH -H
•H fn
•P P fn
CO rH «H
CO rO CD
fn CO +J
bO 3
bO P! 0
<-H <
rH (NJ
CO
H
t/)
^
<™j
Q
(~|
•<*
csj
O
CV)
(A
Cu
m
0
•H •
•P y
cO -P
<£, in
KJ
rH
rt
3
pj
•5
o
o
rH
m
o
CD
O •
cd t/n
CD *H
rH -P
03 *H
pH
0
CO 0
CO (/>
0 ro
!H 0
O 10
P! -H
M T3
LO
rH
CO
=!
pj
•<
O
CTi
P
O
P
cfl
•H
t/1
O
^
.^1
!8r^
r^8
a -8
Q to
t-C
285
-------
I
1I I ^
o in
O r-~,
•H -P O
P cd o
rH rH
8,0 rH
O -H CD
rH -P &
PH C3
rH CD
13 3 P
0 CU Cd
P O &
U p i v—'
W -P
O O
O O
O O
O
cxi
cxj
O
CXI
CXI
o
o
o
o
o
o
s
o
o
o
o
o
o
oo
ex)
o
o
o
o
o
o
o o
o o
o o
o
o
o
o o
P O
o
o * «*
o o oo
O ^ rxl
- t—t
o
o
cxi
o
o
o
o
o
o
oo
o
o
o
o
o
o
K)
cxi
*
o
CD
cxj
in
3
K)
o
CX]
CD
O
in
oj
s
rt
CXJ LO
%
W
cn
cxi
LO
0
O
CD
m
<4H
w
ai
CD
IS
I
0)
^
•H
rH
CS
P
O
•H
g
-6
MH
rt
rH DO
bO g
^rH
CD
LO
g
•H
•P
O
"O
CD
0 -H
m
o
0
o
0
rH
Ctf
in
•H
in O
cd -H
O rH
r^-6
«J
rH
to
CD
0
s
rt
0 in
tn 0
ro in
0 rt
rH 0
O in
000
bo bO bo
oj cd cd
(/) 4-) 4-» -M
0 rH !H rH
g O O O
286
-------
5.1.6.2 Toxicology--At this point in time, theoretical definition of
suspended fine particles as they relate to health must necessarily be
broad and somewhat arbitrary. The deposition of particles anywhere in
the respiratory tract, from the nose to the alveoli, may produce sig-
nificantly harmful effects on health. There is some reason to believe,
however, that non-viable particles deposited below the trachea may be
more harmful than those deposited in the nasal cavity.
Whether particles must reach the alveoli to harm is a more con-
troversial issue. It is reasonable to assume that particles deposited
in the bronchi or bronchioles are more important in producing bronchitis
than those deposited in the alveoli. At present, there seems to be
little value in confining the category of suspended fine particles only
to those particles that reach the alveoli.
The Task Group on Lung Dynamics has calculated that up to 10% of
inhaled particles of diameter 2.5 ym are deposited in the tracheo-
bronchial compartment and that up to 30% are deposited in the alveolar
compartment. Most particles of diameters above 5 ym are deposited in
the nasal cavity. That means a size range of fine particulate up to
205 ym diameter, measured aerodynamically, is of principal interest from
the standpoint of health effects.
Toxicological properties of inhaled particles are related to a com-
plex of factors that include specific physical-chemical properties of
the particles themselves and host conditions that affect response to
exposure to these particles. As for toxicity of certain specific par-
ticulates that may exist; in the atmosphere, moderate amounts of infor-
mation are known. But such information is often unavailable in a form
287
-------
suitable for accurate assessment of minimum time-concentration for
exposure that will adversely affect health.
Even less information is known about how physical environmental
variables (weather conditions) may interact with suspended particulate
matter and alter its chemical composition and affect its toxicity.
Weather conditions, per se, may affect man in such a way as to alter
his exposure to inhaled particulates. The information that is avail-
able, however, provides a basis for estimating relative toxicity of
certain substances and provides insights into problem areas where our
information is inadequate.
5.1.6.2.1 Dose-Response Relationships--The principal way air parti-
culates exert an effect on health is presumed to be through inhalation
and consequent effects on the respiratory system. This presumption is
acceptable in determination of short-term effects of irritant aerosols.
Such an assumption may be less tenable in consideration of long-term
exposure effects where such responses as carcinogenesis, mutagenesis
and subtle metabolic effects result from whole body burdens. In these
cases, other routes of entry besides inhalation may be important in
establishing dose-response relationships.
Inhalation, however, may be a significant route of entry for non-
respiratory toxicants that are first deposited in the respiratory
system, but then translocated to the gastro-intestinal system by muco-
ciliary transport and swallowing. There they may exert a primary
toxic effect or be absorbed and translocated to other tissues where an
adverse health effect might occur.
5.1.6.2.2 Physical Factors--Most calculated and experimental data on
respiratory deposits of particles uses the term "aerodynamic diameter."
288
-------
Effective aerodynamic diameter is a function of particle size, density
and shape. The distribution and mean of aerodynamic diameters are
required to predict patterns of deposit in the respiratory tract.
In terms of mass, many aerosols fit a log-normal distribution,
that is, the logarithms of particle size are normally distributed.
Thus, two indices are used to describe particle distribution: (1) the
mass median aerodynamic diameter (MMAD) and the distribution parameter
(erg) that represents geometric standard deviation.
The MMAD is the particle size in a log-normal distribution below
which 50% of total mass resides. The og is the ratio of particle sizes
at 50.0% and 84.13% probability. These parameters, then, are used to
determine the mass of material that may be deposited in a given lung
compartment.
In some cases, especially with fine suspended-particle materials,
numbers of particles may be more relevant than mass. To relate number
of particles to mass size distribution requires determination of fre-
quency distribution. Particle number can be toxicologically important
since the large surface area may promote gas-particle interaction or
volatile substance-particle interaction. Coupled with small size,
these interactions may result in deeper penetration in the respiratory
tract and more rapid dissolution. Surface area is also important for
highly insoluble particles that show their toxic potential only at the
surface.
Another important factor in regional deposition of inhaled par-
ticles is hygroscopic!ty. Particles as usually measured externally
are not saturated with water. The growth of sodium chloride particles
289
-------
is a function of relative humidity as demonstrated in an experimental
194
study by McJilton, Frank and Charlson. Particle size increased when
relative humidity exceeded 75%. The respiratory tract, with its satu-
rated atmosphere, may increase effective particle size considerably
over and above that measured in ambient air. However, hygroscopicity
is of less significance with particles of high density.
A moderate amount of information is available on respiratory par-
ticle deposition in people and lower animals. These data have been
195 196 197
critically reviewed by Morrow and others. ' These data are
based on mathmetical calculations and to a limited extent, experimental
data obtained from man. The calculated data agree in general with
experimental observations of people.
A figure depicting respiratory deposition of particles as a func-
tion of particle diameter is shown in Figure 5.4. These data indicate
that most particles larger than five ym are deposited in the nasal
cavity or nasopharynx. As particle size becomes smaller, an increasing
number of particles are deposited in lungs. This is a possible explana-
tion of experimental results by Amdur who reported that sulfate aerosols
of 1 ym diameter are greater respiratory irritants for guinea pigs than
2-3 1 ym-sized aerosols.
Only a few comparisons to people have been made on regional respira-
tory particle deposition in laboratory animals. (Figure 5.5)
Although there have been some differences in particle deposition in
rodents, especially in efficiency of nasal deposition, patterns are suf-
ficiently similar to justify use of laboratory animal experimentation in
inhalation toxicology.
290
-------
Each of the shaded areas (envelopes) indicates
variability of deposition for a given mass
median (aerodynamic) diameter in each compart-
ment when distribution parameter ag varies
from 1.2 to 4.5 and tidal volume is 1450 ml.
0.05 0.1 0.5 1.0 5 10
MASS MEDIAN DIAMETER, microns
50 100
Figure 5.4. Effect of particle size on regional
respiratory particle deposition in man
291
-------
100
80
oo
^60
n. .
012
40
CCin
uio
0-0.
30
„
20
U C
- Won
Q- Monkey
o- Guinea Pig
I- 100
z
UJ
o
2"
2 60
O
fc40
o
tr
< 20
O
UJ
_ 3 o-
8 < 0
-Man
0- Monkey
o-Guineo Pig
SIZE OF UNIT DENSITY SPHERES, MICRONS
AP
7 8
Figure 5.5. Deposition vs. particle size of inhaled particles
in the upper respiratory tract and lungs of guinea
pigs and monkeys compared to people
292
-------
A major question exists, however, about applicability of respira-
tory deposition data from normal individuals to predictions of how
exposure to suspended particulate matter may affect particle deposition
rates and regional deposition. To the extent that suspended particu-
lates may act as respiratory irritants, consequent effects on mechanics
of breathing may affect both the rate and site of particle deposi-
205,206
tion.
Under these circumstances, total inhaled dose may be increased.
However, this hypothesis lacks experimental verification. Also, many
diseases and symptoms associated with elevated levels of particulate
matter are airway related. In these diseases, the contribution of
inhaled particulate matter to their pathogenesis may not require pene-
tration into the alveolar region to produce a toxic effect. This point
cannot be fully evaluated, however, since pathogenesis of chronic
obstructive lung disease is not adequately understood.
207
Finally, as noted recently by Casarett the varied pulmonary
responses to particles can readily be seen as interacting with effects
produced by gases. In any case, it is clear that effects of complex
mixtures of gases and particles are not likely to be accounted for by
simple summation of individual components.
5.1.6.2.3 Host-related Factors--The state of health, especially of the
cardiorespiratory system and the amount of activity of an individual
affect the total amount and site of deposition of particulates that an
individual may inhale at any given time. The pattern of breathing
affects the depth of particle deposition, i.e., deep breathing increases
alveolar deposition of particles. During exercise, people commonly
293
-------
shift from nose to mouth breathing and probably increase particle pene-
201
tration into lungs as a result.
At the same time, exercise increases respiratory volume that then
increases the inhaled dose of any toxicant. This effect of exercise
208
has been demonstrated in people experimentally exposed to ozone, and
would presumably occur with inhalation of particulates as well.
Pre-existing disease of the airways or alveoli is though to pre-
dispose affected individuals to more severe response when subsequently
exposed to atmospheres that contain high amounts of suspended particu-
late matter. Exacerbation of symptoms of disease that occur in such
individuals usually occur in conjunction with weather conditions, i.e.,
cold and high relative humidity, that in themselves may have adverse
209 212
health effects. ' Just how interaction between physical weather
factors, particulate matter and gaseous pollutants interact to affect
health is not well understood.
5.1.6.2.4 Retention Factors--The deposition rate and retention time of
inhaled particles are separate phenomena with overlapping biologic
effects. Particle retention time depends on (1) site of initial deposi-
tion in the respiratory tract and (2) chemical composition and properties
of the particles.
If particles are deposited in ciliated epithelium of airways, par-
213
tide clearance is reasonably rapid. In this event, the toxicity
depends on solubility of the particle in mucus while in transit up the
muco-ciliary escalator. If the particle is highly soluble, toxic
inflammation in the airway epithelium may occur.
294
-------
Particles deposited in alveoli of lungs may only very slowly clear
from this location. ' ' The extent of clearing reflects solu-
bility and subsequent translocation via lymphatic drainage or macrophage
217
phagocytosis and clearance via muco-ciliary transport mechanisms.
The preponderance of evidence, however, is that non-soluble par-
ticles remain in the deep lung for long periods of time (weeks, months
and even years). As a result of this slow clearance, carcinogenic
hazard of long-lived radioactive metals, i.e., plutonium and uranium
and airborne chemicals, especially hydrocarbons, is of special concern.
218
A recent hypothesis by Kilburn of a two-zone distal lung merits
some consideration of deep lung clearance of particles of limited solu-
bility . The first zone consists of cul-de-sac alveolar ducts with
"resident" macrophages that assist in enzymatic digestion of particles
for clearance only to lymphatics or blood vessels. The second zone is
represented by respiratory and terminal bronchioles permeable to macro-
phages and polymorphonuclear leukocytes. These cells are subject to
recruitment when dusts contact "sentinel" macrophages in the area. This
hypothesis is consistent with a considerable body of circumstantial
evidence and deserves further investigation.
207
A second hypothesis expressed by Casarett, not exclusive of the
one just presented, is that phagocytesis by macrophages or, to a lesser
extent, by alveolar epithelial cells, is preliminary to removal of par-
207
tides whatever their subsequent destination. So Casarett supports a
concept of a rate-limiting biological process in lung clearance composed
of phagocytosis and cellular movement. This concept does not deny the
219
importance of physicochemical indices, but views them in terms of
their influence on interaction between particle and cell.
295
-------
In support of this view, it can be said that the central role of
the alveolar macrophage in defense of the lung against insult by par-
220 221
ticulate material is now firmly established. ' Investigations to
examine influence of particulate materials and soluble analogues on
structural and functional integrity of alveolar macrophage have been
222 223
reported recently by Bingham et al., Burrell and Anderson and
224
Waters, Gardner and Coffin.
There is much evidence that particles that penetrate into fixed
tissues show clearance half-times ranging from a few days to thousands
of days. The probability, then, of particle penetration is critical in
determining clearance of particles from non-ciliated regions of the
lungs. Phagocytosis by alveolar macrophages decreases the probability
of particle penetration and aids clearance.
The crucial issue in a discussion of clearance rates is whether
they are sufficiently great to balance deposition rates. It is the
equilibrium level, or retention integrated over time and the physi-
cochemical nature of the particle itself, that presumably determine
probability of a pathological response.
In a recent report, Brain^J applied the ICRP lung model
described in the Air Quality Criteria Document for Particulate Matter
to predict retention of particles in lung compartments at different
times. Three exposure patterns were considered in the predictions:
acute pulse, continuous chronic and 40-hour work week exposures.
The results of the study underscore the importance of both size
distribution and physicochemical composition in quantitation of par-
ticle retention. They clearly demonstrate that changes in particle
296
-------
size and solubility can alter the magnitude of retention and integrated
exposure in some compartments even when identical masses of particulate
are inspired. Although simplifying assumptions were made in preparation
of the ICRP lung model, results of Brain's study reinforce the growing
conviction that air quality standards must take into account size distri-
bution and solubility of particulate material.
As for certain metallic air contaminants, there is more awareness
of concentration gradients that exist in particulate samples as a func-
. , , 226,227 3 ^ 228-230 ,
tion of particle size. Lee et al. and others have
established that Pb, Cd, Zn, Cr, V, Ni, Mn and Cu are found at highest
concentrations in the smallest particles collected from ambient air. To
this list have been added As, Sb, Se and Tl, found by Natusch and
231
Wallace to be most concentrated in the smallest particles emitted
from coal-fired power plants.
Unfortunately, little is known about matrix composition of small
particles with which Fovic metals are preferentially associated. It
is thought, however, that surface absorption or condensation of
volatilized elements and organic compounds on to small particles is
responsible for increased efficiency of respiratory deposition of these
232
substances. That this view has considerable significance in terms
of environmental health is suggested by the fact that 1 ym MMAD par-
ticles are deposited predominantly in alveolar regions of the lung
231
where absorption efficiency for most trace metals is 50 to 801.
Similar conditions may prevail in the case of potentially carcinogenic
organic vapors.
Because certain metals may be soluble in respiratory secretions,
toxic properties of these substances may be manifested in lung
297
-------
parenchyma or airway epithelium or translocated so as to induce lesions
in other sensitive tissues. Vanadium is an example of such a metal
whose effects may show up through this mechanism. However, minimum
time-concentrations of vanadium and other metal exposure that produce
toxicity are not well understood. Also, toxicological assessment of
metals that accumulate in the lung or other tissues present different
problems than many other respiratory irritants because the relationship
of tissue burdens to carcinogenesis, metabolic defects and mutagenesis
or tetrogenesis is more difficult and time-consuming to assess.
5.1.6.2.5 Biological Respirable Particles--There are species of pollens,
fungi, bacteria and viruses within respirable range in particle size.
These biological particles are included among particles collected for
measurement as respirable particles. However, it is difficult to
characterize the concentration of these particles because of complexi-
ties involved in precise species identification. In the case of viruses
and bacteria, concentrations may go low as to cause problems in
measurement.
Respirable biological particles have received low priority by EPA
because of lack of appropriate technology for their control that can be
applied to the general population. However, biological respirable aero-
allergens, microorganisms and certain other substances that may interact
with biological tissues can be important in the etiology of immunologi-
cal respiratory disease.
Certain paratoxic effects of gaseous atmospheric pollutants that
involve increased susceptibility to airborne microorganisms have been
233
reviewed recently by Coffin and Gardner. These effects may be related
298
-------
to impaired pulmonary defensive capabilities and altered dynamics of
respiratory air flow that change the pattern of regional deposition
of inhaled particles. Vfrth regard to the latter, see Fairchild224
O?r
and Lippmann, Albert and Peterson.
5.1.7 Sulfur Oxides
5.1.7.1 Human Studies--The standards for SCL are shown in Table 5.30
and studies on which they were based when set on April 30, 1971 are
presented in Table 5.31. These studies represented the best informa-
tion available. Their limitations and shortcomings were recognized,
but the Clean Air Act mandated that a standard be established. Included
among this best available evidence for establishing an ambient air
quality standard for SCL in 1970 were the following reports:
7 "^ft
(1) Brasser, et al. reviewed major episodes of air pollution
in London and Rotterdam between 1952 and 1962. They indicated that
excess mortality resulted when 24 hour mean levels of SCL exceeded
500 yg/m (0.19 ppm) for a few days. Also, hospital admissions for
respiratory illness as well as absenteeism increased whenever 24 hour
mean levels of SCL reached 300-400 yg/m for three-four consecutive
days.
737
(2) Joosting studied data from London episodes and determined
that significant correlations between CL pollution levels and deaths
or disease occur with 24 hour mean SCL concentrations of 400-500 yg/m
(0.15-0.19 ppm) when there is a high soot content.
2 38
(3) Carnow, et al. studied a group of patients with chronic
bronchitis. Substantial increases in illness were reported by persons
55 years and older with more severe bronchitis when concurrent or
299
-------
proceeding daily SCL levels ranged from 119 to 249 yg/m (.05-.09 ppm).
The relationship was not observed among persons with less severe bron-
chitis or patients less than 55 years of age. During the period of
study, TSP levels exceeded the 24 hour standard on about 15% of the days.
So the effects observed may have been related to the combination of
materials that the patients were exposed to.
241
(4) Wicken and Buck studied mortality in two communities in
England with contrasting pollution exposure levels. They found increases
in lung cancer mortality in men and increases in bronchitis mortality in
men and women associated with long-term mean S02 of 115 yg/m (0.040 ppm)
3
when compared with SCL exposure levels of 75 yg/m (0.026 ppm). The high
exposure community also experienced TSP levels of 160 yg/m compared with
80 yg/m in the low exposure area.
239
(5) Douglas and Waller found increases in lower respiratory
disease in London children to be directly related to a pollutant gradient
that they were exposed to. Upper respiratory disease did not show a
similar relationship. LRD was increased in areas where annual mean SCL
concentrations were 130-148 yg/m (.05-.06 ppm) or higher together with
TSP levels of 91-138 yg/m3 or higher.
243
6) Lunn et al. observed increases in lower respiratory illness
in England children associated with exposure to annual mean SCL levels
of 194 yg/m (.074 ppm) or higher and TSP levels of 177 yg/m or higher
compared with S02 and TSP levels of 109-134 and 70-78 yg/m respectively.
To fill gaps in the information base as rapidly as possible, the
Environmental Protection Agency maintains a continuing program of research
to specifically assess adequacy of maximum permissible long-term or short-
term exposure levels.
300
-------
Table 5.30. PRIMARY NATIONAL AMBIENT AIR QUALITY STANDARDS
FOR SULFUR OXIDE
National Primary Standard
Pollutant Concentration (yg/m ) Averaging Time
Sulfur dioxide 80 ug/m (0.03 ppm) Annual arithmetic mean.
365 yg/m (0.14 ppm) Maximum 24-hour concentration
not to be exceeded more than
once per year.
301
-------
Table 5.31. INFORMATION BASE FOR ESTABLISHMENT OF NATIONAL
PRIMARY AMBIENT AIR QUALITY STANDARD FOR SULFUR
DIOXIDE
Adverse Health Effect
Concentration at Which Effect
Was Observed
Concentration Averaging Time References
(yg/m3)
1. Increased mortality.
2. Increased hospital
admissions with
respiratory illness.
3. Increased frequency and
severity of respiratory
illness.
4. Increased specific
mortality.
500 yg/m
(0.19 ppm)
300-500 yg/m3
24 hours
24 hours
224,225
226
(0.11-0,19 ppm)
130 yg/nT
(0.06 ppm)
130 yg/in
(0.06 ppm)
Annual mean
Annual mean
227,228
229
302
-------
One of the primary problems in determining health effects of S02
continues to be development of an understanding of the manner in which
this gas interacts with other substances in the atmosphere. Laboratory
studies had demonstrated that the levels of S02 found in the ambient air
are innocuous in an experimental situation until combined with other
substances. Similar studies with animals demonstrated the high toxi-
cities of sulfuric acid mists and various sulfate compounds that are
products of the atmospheric transformation of SCL.
Observed associations of adverse health effects in epidemiologic
studies with ambient levels of suspended sulfates have supported these
laboratory studies. Table 5,35 presents recent study results in a man-
ner that relates adverse health effects to S07 exposure and suspended
sulfate exposure. The studies referenced in Table 5.32 relate to a
variety of health parameters.
An analysis of daily mortality in the New York metropolitan
region between 1962 and 1966 indicated that S02 measurements contributed
significantly to prediction of deaths. Mortality was 1.51 less than
expected on 232 days with levels below 30 yg S02/m and 2% greater than
expected on 260 days with levels above 500 yg S09/m after correcting
Z<
for other factors.
In each of three New York neighborhoods, cold temperatures were
directly related to increased symptom rates in persons with combined
heart and lung disease. Elevated suspended sulfates were the only
pollutant consistently aggravating symptoms. Daily sulfur dioxide and
total suspended particulate concentrations could not be associated with
symptom aggravation in the "heart and lung" panel that was most sensi-
tive to variations in daily pollutant concentrations.
303
-------
tO
to
CM
•\
CM
to
CM
LO
tO
CM
to
CM
oo
to
CM
to
CM
to
CM
O
**
CM
•*
01
to
CM
OO
to
CM
to
CM
e
FH
Q
i — i
W
U
I-H
P-,
t-H
H
w
1 — 1
£_}
en
d
r^H
O
H
i— t
Q
r_
.
CM
tO
LO
CD
i — 1
•8
H
'•rj
P
CD
to
fi
0
1
3:
P
U
CD
4H
4-1
W
f|
O
•H
r^
§:
P
CO
Pi
O
•H
P
CO
£~t
4->
d
CD
O
r*j
5
U
1
H
•H
bo
rt
CD
^
to
e
bo
pj
O
•H
4_>
cO
P
CD
O
O
to
S
bO
~1
CO
CO
/B— ^
6
a
K1
C*
•^v^
bO
3.
CN
8
+-*
o
CD
4H
4H
W
r*|
4_j
rH
CO
CD
2C
CD
trt
rH
T3
SH
O
X!
^3-
CM
<£
>^i
LO
0 rH
O •
«* O
1 1
O rH
O rH
to •
o
•
fr
•H
rH
CO
P
%
P«
T3
CD
8
O
VI
o
X
r^j-
CM
O
t— 1
l
OO
f — ^
LO rf
\O rH
to •
o
^ — /
Pi
•H
j-
G
O
p
^^
(/}
4H
0
Pi ^N
•H r<
P CD
rt T)
CO CD
f-H
bO CD
£
o
^j-
CM
o
1— 1
1
OO
f — \
01
0 0
LO •
CM O
1 1
o r-
oo o
rH •
0
,
^
\f\
Co
4H
0
Pi
0
•H
P
1
bO
g
Q
£
rH
^j
p*
p
^^
rH
rH
f — ^
LO
o r~-
o o
CM
O
Pi
•H
_.
0
•H
P
O
3
4H
bO
rH
T)
CD Pi
3 £
CD TJ
>H rH
O -iH
g
s
rH
c8
3
pj
c
^^
o>
o
0
rH
O
f-l
g
rH
CD
P
O
cO
Tj
rease
o
^— ^
to
0
•
0
1
T^
to
o
o
V 1
£
•H
to
CO
CD
tfl
•H
T3
S^
0 •
P 0 0
rH
O
o
p
4-1 cO
0 t-t
•H
CD O,
0 • t«
p| tn CD
CD -H ?-i
rH -P
CO -H CD
^> f* j-i
CD O 3
r< Pi 0
P< O cO
"^ o rcj
CD CD
tO O (/)
CO -H CO
CD pi CD
U rH O
•
CD
•H
rH
ip-j
ri
4H
.s
CD
S
CD
c/)
fi-S ££43
t-H TJ
to
304
-------
Best judgment estimates based on preliminary studies of 24-hour
mean threshold levels for aggravation of cardiopulmonary symptoms in
the elderly are 8-10 yg/m of suspended sulfates.
Studies of daily asthma attack rates in the Salt Lake Basin showed
that attacks were more consistently correlated with colder outdoor
temperature than with any measured pollutant. An analysis of asthma
attack rates compared to daily pollutant concentrations was carried out
within two ranges of minimum temperature, 30 to 50°F and greater than
50°F. Virtually no relationship between sulfur dioxide and attack rates
appeared.
Total suspended particulates and suspended sulfates were positively
correlated with daily asthma attack rates. A higher frequency of asthma
attacks was observed at the same daily total suspended-particulate con-
centration when a high sulfate fraction was present in the atmosphere.
It then appears that sulfate levels were a stronger determinant of asthma
attack rates than total suspended particulates.
In New York, also, asthma attack rates were more consistently asso-
ciated with daily suspended sulfate levels than with either sulfur
dioxide or total suspended particulates.
In the opinion of the authors, best judgment estimates of the 24-
hour mean levels of SCL that might serve as precursors of significant
ambient SO, concentrations were 180-250 yg/m .
Ventilatory function of elementary school children diminished as
measured by 0.75-second forced expiratory volume in areas of elevated
exposure to sulfur oxides. In New York only older children (ages nine
to 13 years) who had been exposed to substantially elevated pollutant
305
-------
concentrations for the first five to 10 years of life showed evidence of
reduced ventilatory function. The best available estimates of these
previous annual average exposures were as follows: sulfur dioxide, 131
to 435 yg/m ; total suspended particulates, 75 to 200 yg/m ; suspended
sulfates, 18 to 28 yg/m . The best judgment estimate of a threshold
level for SCL was approximately 200 yg/m .
In a Cincinnati study of ventilatory function in children, suspended
sulfates emerged as a pollutant of particular concern. Sulfur dioxide
concentrations were at or below the moderate level of 57 yg/m during the
time of testing permitting suspended sulfates to be assessed in the
relative absence of sulfur dioxide. Ventilatory function in children
exposed to suspended sulfate concentrations of about 9.5 yg/m was lower
than that of children exposed to concentrations of about 8.3 yg/m . It
was the authors' best judgment that eight or nine years of exposure to
about 10 to 13 yg/m of suspended sulfates might reduce ventilatory
function. If these suspended sulfate exposures were accompanied by
exposures to about 200 to 250 yg/m of sulfur dioxide and about 100 to
150 yg/m of total suspended particulates, further reductions in FEV
0.75 might be expected.
Lower respiratory disease studies conducted in the Salt Lake Basin
and Rocky Mountain areas found no significant association between total
lower respiratory disease and pollution for children whose parents had
been residents of their communities for less than three years. For
single and repeated episodes of croup and repeated episodes of any lower
respiratory disease, families of children who had lived three or more
years in the high exposure communities reported more illness among all
306
-------
ages of children from 0 to 12 years than did their counterparts in less
polluted communities.
Best judgment estimates of the threshold levels for increases in
acute lower respiratory tract infections in children were three years of
3 3
exposure to 95 yg/m of sulfur dioxide, 102 yg/m of total suspended
participates and 15 yg/m of suspended sulfates.
Chronic bronchitis rates in CHESS areas showed a consistent pattern
of excess among residents of more polluted communities.
In the Salt Lake Basin, male smokers and non-smokers showed an
excess of chronic respiratory disease whether defined in terms of
bronchitis prevalence or severity scores. Chronic bronchitis rates and
severity scores were higher among male and female smokers, non-smokers
and ex-smokers from more polluted areas than they were in comparable
groups from less polluted areas.
Among young white military recruits in the Chicago area, the con-
tribution of air pollution was relatively much less than that of
cigarette smoking. However, there was evidence that even among these
young inductees, respiratory symptoms were more prevalent among persons
from more polluted communities.
Best judgment estimates of the threshold levels for increase in
prevalence of chronic bronchitis in adults were six years of exposure to
95 yg/m of sulfur dioxide, 100 yg/m of total suspended particulates,
and 14-15 yg/m of suspended sulfates.
Studies of total acute respiratory disease among family members in
Chicago and New York showed that, with the exception of fathers, who
often have greater occupational exposures and daily changes of exposure
307
-------
due to place of work, a consistent excess of acute respiratory disease
was reported among family members living in more polluted neighborhoods.
Best judgment estimates of the threshold levels for increase in
frequency or severity of acute respiratory illness in families are three
years of exposure to 106 yg/m of sulfur dioxide, 151 yg/m of total
suspended particulates and 15 yg/m of suspended sulfates.
Best judgment estimates of the exposure threshold for adverse
health effects are given in Table 5.33. The margins of safety are
included in the air quality standards in Table 5.34. And estimates of
the population at risk are presented in Table 5.35.
5.1.7.2 Animal Studies--The health effects of sulfur oxides have recently
been critically reviewed in four major reports:
1. Summary report on adverse health effects related to sulfur
oxides exposure, Environmental Protection Agency, Research
Triangle Park, N. C., September, 1973. Unpublished.
2. ERC/RTP Task Force report on sulfates and sulfuric acid
aerosols, October, 1973. Unpublished.
253
3. Toxicology of sulfur dioxide decay products by Lewis et al.
4. A review of the toxicology of oxidation products of sulfur
dioxide by M. 0. Amdur.254
Information in these documents help provide the basis for Air Qual-
ity Criteria for Sulfur Dioxide. For information published before 1972,
readers should refer to the above referenced documents. In the two
years since these documents were compiled, only a limited amount of new
information has been obtained that contributes significantly to under-
standing of toxicologic properties of sulfur oxides.
Recent research on sulfur oxides has continued to demonstrate that
the respiratory tract is the principal organ system affected. The
308
-------
Table 5.33, BEST JUDGMENT THRESHOLD ESTIMATES FOR HEALTH
EFFECTS OF SULFUR OXIDES
Best Judgment Threshold Estimate
1.
Adverse Health Effect
Aggravation of cardio-
Concentration yg/m Averaging Time
so2 ss
365 10 24 hours
respiratory symptoms or
lung symptoms in elderly
patients with related
illness.
2. Increased mortality.
3. Aggravation of asthma.
4. Reduced lung function.
5. Increased frequency or
severity of lower
respiratory illness.
6. Increased prevalence of
chronic bronchitis.
7. Acute respiratory disease.
350
200
200
91
95
106
NA 24 hours
8-10 24 hours
11 Annual mean
9-12 Annual mean
14 Annual mean
12 Annual mean
309
-------
Table 5.34. SAFETY MARGIN INCLUDED IN SULFUR OXIDE STANDARD
Threshold Primary Safety*
Adverse Effect Concentration Standard Margin
1.
2.
3.
4.
5.
6.
Aggravation of heart or
respiratory symtoms in the
elderly.
Increased mortality.
Aggravation of asthma.
Reduced lung function.
Increased lower respiratory
illness.
Increased chronic
bronchitis .
365 365
350 365
200 365
230 80
91 80
95 80
0
0
0
(-45)
187
14
19
*cnff^ M-^e-in - Threshold Concentration - Standard „ inn
Standard
310
-------
Table 5.35. SIZE OF POPULATION AT POTENTIAL RISK
OF POLLUTANT-INDUCED EFFECT
Adverse Effect
Expected Frequency of
Adverse Effect in the Population
(Rate per 100)
Potential
Size* of
Population
at Risk
Chronic bronchitis.
Asthma.
Chronic heart illness.
Susceptible to respiratory
infection.
10% of adult population
i&
5% of adult population
100
8,400,000
4,200,000
4,200,000
140,000,000
*Assumes a U.S. population of 200,000,000 and 70% of total population living
in urban areas of increased pollution.
311
-------
available data on comparative toxicity of sulfur oxides indicate that
sulfur dioxide is only a mild respiratory irritant, while specific
sulphate and sulfuric acid aerosols are more potent respiratory irri-
tants. The importance of particle size as well as mass concentration
to toxicity of sulfur dioxide aerosols is of paramount importance.
254
Experiments demonstrating this point have been reviewed recently.
Most recent studies of sulfur-oxide toxicity have been concerned
with sulfur-oxide aerosols or sulfur dioxide in combination with an
inert particulate aerosol. The importance of high relative humidity
to enhancement of the effect of sulfur dioxide and sodium chloride
aerosol has been demonstrated through sensitive repsiratory physiologic
255 3
responses. These compounds (S0~- 1 ppm and sodium chloride 1 yg/m )
alone or in combination caused no effect on pulmonary airflow resist-
ance of guinea pigs at low relative humidity (40%). However when the
relative humidity was increased to > 80%, a 50% increase in pulmonary
airflow resistance developed. The increase in relative humidity
resulted in an increase in aerosol size and production of an acid
aerosol that are thought to be important events in the production of an
irritant aerosol leading to increased pulmonary airflow resistance.
A concentration of3020ygm/m H-SO. (1.8 u CMD size) that has been
shown to increase pulmonary airflow resistance in guinea pigs also pro-
duces an increase in total respiratory particle deposition in guinea
7 c A
pigs. This increased particle deposition occurred primarily in the
nasopharynx, i.e., a proximal shift in the normal pattern of regional
deposition of radio-labelled bacterial aerosol occurred. Dose-response
experiments revealed that even at a concentration of 30 ygm/m (0.25 u
CMD size) a proximal shift in the pattern of respiratory particle
312
-------
deposition occurred. But in this instance the shift was to the trachea.
These results suggest where the site of action is of H-SO, aerosols of
this mass concentration and particle size.
Particle deposition in the respiratory tract is correlated to air-
flow dynamics in this system. The suggestion has been made that these
results reflect a subtle patho-physiologic consequence to altered
respiratory airflow resistance. So these results provide added signifi-
cance to respiratory physiologic responses that are commonly used in air
pollution health effects studies.
A concentration of 15 yg/m of H-SO, (2.5 u CMD size) for 90 minutes
per day for four consecutive days induced a reduction in the physical
clearance of radio-labelled streptococcous particles deposited in the
245
nose and lungs of mice. No effect was demonstrated on the clearance
of viable streptococci from these same tissues. However, no effect on
physical clearance of streptococci was observed at a mass concentration
of 1.5 yg/m H2S04 (0.8 u CMD size). These data suggest that altera-
tions in respiratory particle deposition and path/physiologic conse-
quences of these alterations are a more sensitive index of response to
H^SO. inhalation than mucociliary clearance of particles.
258
Busey et al. have demonstrated that chronic exposure (78 weeks)
of guinea pigs to a combination of fly ash (0.55 yg/m ) and H2S04 aerosol
produced inflammatory lesions in terminal bronchioles of the lung
characterized as moderate hyperplasia of the epithelium.
The only alteration in respiratory physiologic measurements was a
small increase in respiratory airflow resistance that might be associated
313
-------
with histopathologic lesions. The results of this experiment are
important because a much higher concentration of sulfur dioxide did not
-^ rn
produce any health effect. A combination of SCL (up to 5 ppm)
and fly ash (0.5 yg/m - size 2.6 - 3.4 u MMD) exposure for 78 weeks
did not induce any respiratory physiologic or histopathologic effects
, 262
in monkeys or guinea pigs.
9 A T
Zakower has observed small changes in immunologic responsive-
ness of rats chronically exposed to sulfur dioxide and carbon dust.
These observations have not been confirmed or the pathophysiologic
consequence of this effect demonstrated.
5.1.8 Mercury
Several recent animal exposure studies using methyl mercury given
in repeated oral doses over relatively long time periods show that a
"no-effect" level for development of neurological signs and/or histo-
pathological evidence of mercury poisoning is from 0.03 mg/kg body
weight (monkeys, cats) to 0.1 mg/kg body weight (calves).
Daily oral doses of alkyl mercury compounds as small as 0.4 mg/kg
body weight have produced neurological disorders in rabbits, cats and
dogs. Daily oral doses of 0.25 mg/kg body weight of methyl mercuric
chloride given throughout most of the gestational period caused increased
abortions and fetal abnormalities in cats. Similarly, levels above 0.3
mg/kg given during gestation increased fetal abnormalities in rat fetuses.
Most animal experiments were carried out with daily dose levels
much higher than the foregoing experiments and the most chronic exposure
found was that of monkeys that were exposed for two years. Low level,
long-term exposures were notably absent.
314
-------
Behavioral changes; have been produced in rodents where no histo-
logical lesions were found from single oral doses of methylmercuric
chloride. The relatively high 25 mg/kg dose level was required since
doses at 20 mg/kg level did not produce significant changes.
264
Ikeda fed methyl mercuric chloride (MMC) to four groups of
five male monkeys each for two years. Groups were made up of three
Macaca Mulatta and two Macaca Irus each. Levels used were 0.01, 0.03,
0.1, 0.3 mg/kg/day plus a control level. In the two highest levels
(0.1 and 0.3 mg/kg/day), toxic neurological signs appeared at 62 and
182 days respectively when total amount of MVE ingested was 15.3 and
15.9 mg/kg. No neurological signs were noted in the other groups
although animals fed at the 0.03 mg/kg level ingested a total of 18.4
mg/kg MC by the end of two years. No significant alterations occurred
in hematological or biochemical parameters tested in the two lower dose
groups. The mercury content of hair showed good correlation with the
daily dose of mmc. This study is still in progress.
*? f\ Q
Takeuchi reported on total amounts of methyl mercuric chloride
(MV1C) necessary to produce poisoning similar to Minamata disease. In
cats, the total dose was 20.6 to 25.7 mg/kg at a daily level of 1.5 to
2.0 mg/kg. In rats, total dose was 100 to 200 mg/kg at a daily level
of 5 to 10 mg/kg. Extrapolating from these figures to man is a total
dose of 20 mg/kg.
266
Khera fed pregnant cats orally with doses of methyl mercuric
chloride (MC) from the tenth day to the 58th day of induced (gonado-
trophin) pregnancy to determine teratogenic effects. The results were:
315
-------
Dose Level Number Results
0 7
0.03 mg/kg 5 No signs of poisonii^ detected in dams.
0.83 mg/kg 6 No signs of poisoning detected in dams.
0.25 mg/kg 9 Increased abortions and fetal anomalies.
0.75 mg/kg 4 Vomiting, ototoxia and death of dams
within 32 days post exposure.
Minimal or no embryopathic effects were noted at the two lowest levels.
The 0.03 mg/kg dose was considered the "no-effect" level. This corre-
264
lates well with the IDEDA experiment using monkeys where the feeding
period was two years.
*7 f-\ "7
Magos studied cumulative effects of oral methyl mercury
dicyandiamide (MMD) on rats. Female albino rats were divided into the
following three groups:
I. 3.36 mg/kg for five days/week for 16 doses N=14
II. 1.68 mg/kg for five days/week for 34 doses N=12
III. 0.84 mg/kg for five days/week for 59 doses N=16
Group I showed weight loss after the fourth dose and neurological
abnormalities after the twelfth dose.
Group II lost weight after the 24th dose when four of the six ani-
mals remaining were ototoxic.
Group III increased in weight throughout the experiment but two of
these animals showed slight ototoxia.
All groups exhibited histopathological changes in the kidney and
cerebellum.
268
Irukayama reported on short-term (10-90 days) multiple oral-
exposures to methyl mercury that produced neurological disorders in
rabbits, cats and dogs. The daily intake of MM was 0.4 to 1 mg/kg and
the total intake in these species between 10 and 20 mg/kg. To produce
316
-------
similar signs in rats required a total dose of approximately 100 mg/kg
indicating the higher degree of resistance of this species.
269
Post recorded behavioral changes in the following three groups
of young rats (force-fed a single dose of methyl mercuric chloride
[JMC]):
Group I 15 days old 20 mg/kg N=15
Group II 21 days old 20 mg/kg N=15
Group III 60 days old 25 mg/kg N=45
The differences in behavioral measurements in Groups I and II vs.
controls were small or statistically insignificant. Behavioral changes
in Group III were significant, however. The authors state that signifi-
cant behavioral changes in Group III could result from higher doses of
MC or age differences. Since no histopathological changes were found
in any group, the authors conclude that behavioral testing was the more
sensitive indicator of mercury poisoning.
270
In a study by Fukuda, two of six rabbits that were exposed to
an average concentration of 4 mg/m of mercury vapor for six hours a
day, four days a week, for 13 weeks, developed a fine tremor and clonus
in the fore- and hind-legs. (This level is 4000 yg/m , a relatively
high exposure.) Analysis of CNS tissue showed mercury levels of 0.8
to 3.9 yg/g wet tissue.
Beliles found fine tremors in the fourth week in rats repeatedly
exposed to mercury vapor at: 17 mg/m for three hrs/day, five days/week
for four to five successive weeks. Histological changes in the CNS were
also present.
272
Charbonneau exposed two groups of cats to 0.25 mg/kg/day methyl
mercury for 12 to 14 weeks. One exposed group received methyl mercuric
317
-------
chloride in gelatin capsules. The other group received methyl mercury-
contaminated fish. A control group received uncontaminated fish. Clini-
cal signs of neurological abnormality developed in both exposed groups
between 55 and 96 days. At that time, total methyl mercury ingestion was
between 14 and 24 mg/kg.
Tissue mercury content and pathologic changes were similar in both
exposed groups of treated animals. Lesions were found in the cerebellar
vermis and the cerebral cortex.
Pre-test blood mercury concentrations in the cats were considerably
higher than background value for humans, possibly reflecting the dietary
history of the cats. Cats seem similar to other species, including man,
that develop clinical signs of methyl-mercury toxicity when brain concen-
trations are in the range of 10 ppm.
273
Gale injected golden hamsters with mercuric acetate (inorganic
mercury) and phenylmercuric acetate (organic mercury) on the eighth day
of pregnancy and sacrificed them on the 12th or 14th day of pregnancy
to study potential teratogenic effects in this species. Mercuric acetate
levels used were 2, 3 and 4 mg/kg, while phenylmercuric acetate was given
at 5, 7.5, 8 and 10 mg/kg. All levels of mercuric acetate and all but the
lowest level of phenylmercuric acetate produced resorption rates signifi-
cantly above control values. Both compounds produced the same changes in
embryos.
Embryonic growth rate was retarded one to two days and abnormal
fetuses had sub-cutaneous edema. Finding of similar results from exposure
to organic and inorganic mercurials may be due to rapid metabolism of
organic mercury into inorganic form in both liver and kidneys.
318
-------
274
Gale injected pregnant hamsters intravenously with mercuric
acetate in combination with either cadmium sulfate or zinc sulfate on
the eighth day of gestation. (Dose levels of each compound were either
2 or 4 mg/kg.) On the 12th or 14th day of gestation, embryos and
implantation sites were examined for malformations and resorptions. The
detrimental effects of Hg + Cd were much worse than those of the Hg + Zn
combinations. Manifestations of toxicity in dams were directly propor-
tional to the mercuric acetate dose. Animals that received the higher
mercury dose (4 mg/kg) showed more severe toxic effects than any of the
Hg + Cd or Hg + Zn combinations.
275
In a study by Spyker, methyl mercury dicyandiamide (MMD) was
injected into pregnant mice on the seventh or ninth day of gestation at
levels of either 4 or 8 mg/kg. The number of apparently normal neonates
killed by the dam was significantly increased if offspring were pre-
natally exposed to WE). Additionally, open-field testing at 30 days of
age revealed significant differences in latency, backings, defecations
and urinations.
77ft
In another study by Spyker, a total of 667 female mice of two
strains (129/Sr-Sl and A/J) were given I.P. injections of Methyl mercury
elicyandiamide (MMD) on one of days six to thirteen of gestation and
2,280 implantation sites were examined on day 15 or 18. Dose levels
used were 0, 2, 4 or 8 mg MMD/kg. The 8-mg/kg dose produced growth
retardation in both strains and was 88% embryo-lethal when given on
days nine to 13 in strain 129.
This dose level was not embryolethal in strain A. But it did cause
cleft palate in this strain when injected on days nine to 13. Examina-
tions of fetuses from strain 129 on day 15 revealed 52% of embryos had
319
-------
brain, palate or facial abnormalities. Off-spring that appeared normal
showed significant behavioral differences when submitted to battery
tests.
277
Klein exposed 25-day-old rats to sub-lethal oral doses of mer-
curic chloride at either 5 or 7 rag/kg daily for two weeks prior to
active-avoidance training. Klein's results indicate that mercuric
chloride administration significantly impaired active-avoidance acquisi-
tion without influencing either escape latencies or general physiological
condition.
278
Kawabe gave methyl mercury chloride orally to pregnant mice on
the tenth day of gestation at dose levels of 0.1, 0.2, 0.3, 0.5, 1.0,
1.5 and 2.0 mg/kg. Fetuses were removed and examined on the 15th ges-
tational day. At levels of 0.5, 1.0, 1.5 and 2.0 mg/kg the number of
dead fetuses and resorptions were increased. Levels of mercury that
produced abnormalities in this experiment substantially agree with the
experiments of Spyker (mice) and Khera (cats).
279
Munro gave groups of eight to 10 male Wistar rats daily oral
doses of methyl mercuric chloride (MMC) . Clinical signs of mercury
intoxication appeared in all groups but the lowest dose level. Patho-
logical changes occurred in the nervous system of all treated rats
involving Wallerian degeneration of sensory nerves, swelling with loss
of myelin in the posterior white column of the spinal cord, secondary
loss of nerve cells in dorsal root ganglia and atrophy of the cerebel-
lar vermis from loss of granular cells and some Purkinje cells. The
results were:
320
-------
Dose Appearance of Clinical Signs
10 mg MC/kg/day 7 days
5 mg MC/kg/day 14 days
1 mg MC/kg/day 65 days
0.5 mg MMC/kg/day 120 days
0.05 mg MC/kg/day No signs at 280 days
280
Herigstad fed six four-week-old male Holstein-Friesian calves
methyl mercury that produced neurological signs and histopathologic
changes of mercury poisoning at levels as low as 0.2 mg/kg/day at 90
days exposure.
281
Iverson induced overt clinical signs of mercury intoxication
in guinea pigs after daily dosing for nine days at 5 mg methyl mercury/
kg.
28?
Tatetsu gave 5 mg/kg dimethyl mercuric sulfide orally daily
to pregnant rats during two to three weeks of pregnancy producing
degenerative changes in the brains of their newborn.
283
Suzuki produced neurological signs in mice at dietary levels of
10 yg of methyl mercury per gram of feed. This group did not show any
loss of body weight or neurological symptoms until the 41st day of
feeding. The most sensitive neurological sign was noted at a methyl-
mercury concentration of 10 yg per gram of brain.
284
Iverson found the LD5Q (14) for methyl mercuric chloride in
female guinea pigs to be 5.5 mg/kg I.P. and 16.5 mg/kg P.O. The tissue
distribution of a single orally-given dose was studied with the follow-
ing results: Kidney>Liver>Spleen>Heart>Brain. A biphasic excretion
pattern was common to most organs except the kidney. Half-times were
t, = two to three days and ^ = 15-16 days for most tissues. Kidney
turnover rates were much slower.
321
-------
Newberne et al. fed rats fish-protein concentrate (FPC) as part
of their diet to provide 20% protein for five generations. Average
total mercury content of FPC samples was 0.22 ppm with more than 901 in
the form of methyl mercury. Over the two-year feeding period, there
was no evidence of effects on growth, development, reproduction or
behavior throughout the five generations. Histologic examination
revealed no lesions of the brain, kidney or liver caused by the diet.
The average weekly intake of the animals was 12 yg/kg per week.
286
Singhal et al. injected young male rats with 0.4 or 4.0 mg
methyl mercury/kg/day for 45 days and produced a significant decrease
in liver weights at both doses. Gluconeogenic enzyme activities
increased in nearly all enzymes examined at both dose levels. Elevated
blood glucose and urea along with reduced liver glycogen content occur-
red at both dose levels. Withdrawal of mercury exposure to 4 mg/kg/day
for 28 days post-exposure generally failed to reverse observed metabolic
changes in hepatic and renal carbohydrate metabolism.
287
Lucier et al. found that subcutaneous injection of male rats
with methyl mercury hydroxide at 2.5-10.0 mg/kg/day for two days
decreased hepatic mixed-function oxidase activity and microsomal cyto-
chrome P-450 content by approximately 50%. Oral administration with
drinking water of M£l of approximately 0.7/mg/kg/day for 30 days or 1.0
mg/kg/day for 17 days cytochrome P-450 and aminopyrine demethylase were
decreased by only 20% approximately.
288
Klein et al. injected rats daily for five days with 2-10 mg/kg
of methyl mercury hydroxide (NMI). Only the highest dose group had
weight loss. No CNS damage was noticed. Focal necrosis of kidney
322
-------
tubular epithelial cells were noted by day six. These studies identify
the kidney as an early target in MM intoxication and suggest that bio-
transformation of MMH into inorganic mercury may produce part of the
pathological changes noted.
289
Fowler et al. gave rats 55 ppm Hg (as methyl mercury hydroxide)
in drinking water for four weeks. Proximal tubule cells of kidneys
were found to contain swollen mitochondria with dense granual cytosomes.
Sodium phenobarbital, a microsomal enzyme inducer, given weekly during
MMH administration, prevented these histologic changes (possibly by
stimulating the metabolism of MM and reducing renal mercury levels and
attendant nephrotoxic effects).
290
Hastings et al. found that injection of 10 mg MMH/kg for two
days reduced serum cholinesterase activity by 501 in male and female
rats, but had no effect on brain acetylcholinesterase levels.
291
Khera et al. intubated pregnant mice with CH,, Hg Cl at doses
of 0.001 to 10 mg Hg/kg/day from day six to 17 of gestation. Doses
of 5 mg/kg/day caused 100% stillbirths or neo-natal death but no
maternal toxicity. Doses of 1 mg/kg/day caused a low incidence of
delayed cerebellar differentiation of progeny and depressed activity
of cerebellar enzymes. Young female rats were fed methyl mercuric
chloride in the diet at levels of 0.002 to 0.25 mg Hg/kg/day until
sexually mature and then mated to untreated males. Treatment con-
tinued throughout pregnancy, delivery and weaning. At weaning, half
of the progeny remained on the experimental diet and others reverted
to a control diet. All animals were subsequently mated and sacrificed
on day 20 of gestation.
323
-------
Long-term feeding of parent female rats with diets containing up
to 0.25 mg/Hg/kg/day for 122 days had no discernible effect on growth,
behavior or pregnancy. No effects were seen at any dose level on the
F7 fetal weight, abnormalities, gross or microscopic histology of brain
LJ
and spinal chord or viability. A variable incidence of ocular lesions
was dose-related.
5.1.9 Beryllium
Significant animal studies conducted since the classification of
beryllium as a hazardous pollutant are reviewed briefly.
A soluble tissue extract prepared from animal and human lung con-
nective tissue sensitized normal guinea pigs to subsequent exposure of
beryllium as assayed by the skin reaction, macrophage migration inhibi-
tion factor and lymphotoxin production. Greater than 50% of animals
with tuberculosis, berylliosis or coal pneumoconiosis exhibited some
evidence of cell-mediated sensitivity to appropriate soluble tissue
* ^292
extract.
Mice and guinea pigs treated with metyrapone to alter adrenal func-
tion had a more severe reaction to beryllium (BeSO. or BeO via trans-
thoracic or intratracheal injection) than control animals as indicated
by weight loss, metal-ion shift and elevation of serum enzymes, all
signs of systemic disease. Response to altered adrenal function has
293
been seen in human cases, but never before demonstrated in animals.
Injection (I.p) of 0.8 mg BeSCL/kg into rats induced a maximum
elevation of liver acid phosphatase after eight hours, and elevation
of the serum isocitric dehydrogenase, glutamic-pyruvic transaminase and
324
-------
glutamic oxalacetic transaminase activities 48 hours after injection.
The serum lactic dehydrogenase was not elevated at any time.
295
Clary et al. studied the long latent period, sometimes 15 or
more years, between onset of the disease and last exposure to beryllium.
The latent period may be explained on the basis of beryllium accumula-
tion in a latent form in the lyosome. This latent period could persist
for years until the general decline in adrenal output due to aging was
combined with adrenal stress. Under these conditions, especially if
the adrenal reserve was also low, adrenal output may not be adequate to
meet all needs. The lyosomal instability could result, initiating on-
set of beryllium disease.
7
After i.v. injection of Be the biological half-times were 1210,
890, 1770 and 1270 days in mice, rats, monkeys and dogs respectively,
296
compared with the 540 day value cited by the ICRP.
Partially hepatectomized rats given 15, 30 or 60 y moles BeSO./kg
showed liver necrosis when examined 48 hours after i.v. injection.
The abolition of mitoses was observed after 20 or 16 hours in animals
treated with 30 v moles BeSO./kg 8, 12 and 22 hours after partial
297
hepatectomy.
298
Jacques and Witschi found that beryllium did not prevent induc-
tion of aryl hydrocarbon hydroxylase by methylcholanthrene. The
authors concluded that it is unlikely that beryllium could act as a
pulmonary cocarcinogen by interfacing with the metabolis of carcinogenic
polycyclic hydrocarbons in the lung.
Delayed-type hypersensitivity bo beryllium was demonstrated in
guinea pigs sensitized with BeSO. by the skin test using BeSO. and Be-
325
-------
guinea pig albuminate, but not with beryllium hydrgencitrate or beryllium-
299
aurintricar-boxylate.
Guinea pigs sensitized to beryllium by i.d. injections had skin
reaction to beryllium salts. Sensitivity to beryllium was passively
transferable by peritoneal exudate and spleen cells. Macrophage migration-
inhibitory factor (MIF) was produced in approximately 74% of guinea pigs
with positive skin tests. Six out of seven chronic berylliosis patients
showed MIF production. There was systemic involvement of the alveoli,
bronchiole muscles and spleen due to berylloisis.
Schneider et al. found that a nine-fold increase in the rate of
bone formation occurred in the beryllium-injected tibia during the first
four weeks. The rate declined to normal during the last four weeks.
Pathological defects were prominent in newly-formed bone. Some areas
were deficiently mineralized and had an increased number of resorption
spaces.
302 303
Slonim et al. ' found that the amount of beryllium concentrated
in guppies and fry was influenced by concentration of beryllium sulfate
in the surrounding solution, by fish size (inversely) and, to a lesser
extent, by exposure period; but not by fish age, solution buffering or
water hardness. Beryllium uptake was highest in the gastrointestinal
tract and kidneys, but varied greatly among guppies exposed to the same
conditions. Beryllium lethality in fish may not depend on amounts con-
centrated in fish, but on its effects on a particular target organ or
cellular or subcellular component.
The variability in beryllium uptake observed in expired fish was
great, indicating that amounts absorbed in the body are not important
determinants of beryllium lethality.
326
-------
Vacher found that contact of beryllium with skin produces
a delayed-type hyperserisitivity reaction in guinea pigs, while paren-
teral administration of beryllium produces tolerance. The immune
response to beryllium is associated with different forms of beryllium,
i.e., immunogenic (bound to skin constituents) and tolerogenic (freely
diffusible). Low dose tolerance was achieved by i.p. injection of a
very low dose of beryllium (4.78 yg/kg) and high dose tolerance by i.v.
injection of 400 pg/kg. BeSO.-induced blockage of the reticuloendo-
thelial system (RES) is due to "non-diffusible" phosphate fraction
formed after i.v. injection of sulfate. This blockage does not appear
to result from a cytotoxic effect since injected beryllium phosphate,
five times less toxic than BeSO., is active at doses 100 times lower
than toxic doses. Blockage and cytotoxicity of beryllium do not
appear to be linked to the same mechanism. RES blockage is thought to
occur as a result of saturation of absorption sites on phagocytic cells.
Intravenous injection of beryllium phosphate and zymosan, before or
during development of adjuvant-induced polyarthritis in the rat,
decreased inflammatory reactions in the paw injected with adjuvant.
The observed effects appear to be related to a nonspecific activation of
the phagocytic cells with possible lyosomal changes that lead to produc-
tion of substances demonstrating the observed properties.
307
Vasil found that the nucleo-protein fraction of lungs of rats
with experimental berylliosis contains six new antigens (as revealed
by immunoelectrophoresis). Some of these antigens contain beryllium.
5.1.10 Asbestos
The association of impaired human health with industrial exposure
to asbestos is well known. Many hundreds of cases of asbestosis
327
-------
(fibresis of the lung) and pulmonary cancer associated with mining,
milling and manufacturing of asbestos and asbestos products have been
documented. The subject has been reviewed many times recently.
An attempt will be made here to give a cogent story on the biologi-
cal experimentation, to discuss possible implication of air pollution of
general or point source exposure to commercial asbestos and the possible
hazard of exposure of human beings to asbestos-like fibrils that are
generated inadvertently from mining and quarrying.
There is little evidence that light exposures such as might be
encountered in ambient air produce asbestosis or pulmonary carcinoma in
human beings. These conditions have been consistently reported only
from heavy industrial contact with asbestos. However, the discovery of
association of a unique tumor, mesiothelioma of the pleura of peritoneum,
with very low exposure, or after casual contact with chrysotile asbestos,
usually after a long period of latency, has raised a question about risk
due to air pollution.
Many of these exposures had come from such things as living in
homes occupied by asbestos workers or living in vicinity of asbestos
factories and mills. In some cases mesiothelioma developed as long as
30-40 years after any known contact. While most of the South African
314
cases were attributable to chrysotile asbestos, mesothelioma was
315
found in the USA in workmen working solely with amosite. In Finland
no mesothelioma is associated with the prevalent anthophyllite
316
asbestos.
The following summary points have been listed for the epidemiology
r v *. 317
of asbestos:
328
-------
(1) All major types of asbestos can cause lung cancer, although
there are clear differences in risk with type of fiber and nature of
exposure. Since exposure and response to asbestos are related, there is
no excess risk when occupational exposure has been low.
(2) All commercial types of asbestos except anthophyllite may
cause induction of mesothelioma. The risk is greatest with crocidolite,
less with amosite and apparently less with chrysotile. With the last
two, there seems to be a greater risk in manufacturing than in milling.
(3) There is evidence of an association of development of meso-
theliomas with air pollution near crocidolite mines and factories using
mixed fibers. There is no excess risk from air pollution near chrysotile
and amosite mines.
(4) There is, a± present, no evidence of any cancer risk to the
general public from asbestos in air, water, beverages, food or in fluids
used for administration of drugs.
(5) Cigarette smoking enhances the risk of developing bronchogenic
cancer in workers exposed to asbestos. No association between cigarette
smoking and development of mesothelioma has been demonstrated.
(6) Pleural plaques are associated with past exposure to all com-
mercial types of asbestos although not all pleural plaques are related
to asbestos.
The amphiboles are an important group of rock-forming minerals
estimated to comprise as much as 31 of the earth's crust and associated
with both igneous and metamorphic rock. They are divided geologically
into three sub-groups: (1) iron-magnesium amphiboles, (2) calcium
amphiboles and (3) alkali amphiboles.
329
-------
The first sub-group contains grunerite, the commercial asbestos
form of that is termed "amosite." Massive amounts of grunerite and
other amphiboles of non-commercial asbestos quality occur in the United
States and are mined inadvertently because of association with other
minerals that have commercial value, e.g., the Reserve Mining Company's
taconite ores from the Peter Mitchell Pit. As a result since amphibole
ore is estimated to comprise about 3% of the earth crust, they are pre-
sent as matrices in many mining, quarrying, tunneling or other hard rock
activities where inadvertent exposure may occur both occupationally and
318
environmentally in the surrounding community.
It is possible that even agricultural processes in areas where
surface outcrops of asbestiform rock occur may pose a risk to human
319
health.
There is a growing body of evidence suggesting that health effects
associated with exposure to asbestos may not be unique to commercial
forms, but may also be associated with asbestos-like fibers found in
other rocks and even man-made fibers. It is now known that the
external gross appearance of the mineral may not be the determinant.
It may be its microfibrillar nature after crushing or grinding.
The toxicity potential of the silicate mineral may then be related
to their proclivity to form micro-crystals of a certain configuration
when crushed and ground. It is likely, then, that a mode of toxicity
formerly thought to be associated only with asbestos may be present to
a greater or lesser degree in fibrous amphiboles of non-asbestos com-
,.325
mercial quality.
330
-------
Recent reports on exposure of rats to various forms of asbestos
indicate that, while there appears to be a time-dependent dose response
in tumor incidence, no significant difference was noted among various
amphibole asbestos when compared to chrysotile from both South African
326
and Canadian sources.
It would appear possible that pleural and peritoneal mesothelioma
and possibly other disease of man can result from exposure to amphibole
of non-commercial asbestos quality.
"Asbestos bodies" are prevalent in the general population of city
dwellers as is well know. But the question of association of these
bodies with asbestos fibers is a moot point since it is known that simi-
lar fibers of other kinds can develop into these structures after resi-
dence in the tissue. However, critical examination of 28 human lungs
from autopsy material obtained in N.Y.C. by means of appropriate E.M.
327
technique has revealed chrysotile fibers in all lungs. Similar
studies in London have shown that chrysotile fibers are found in 80%
of the specimens of lung included in recent surveys of post-morten
material. (See Tables 5.36 and 5.37.)
The discovery of asbestos fibers in lungs of human beings not occu-
pationally exposed indicates potential risk to the general public. More
surveys are needed to properly to assess this risk. However, since
asbestos is a chemical now common only because of widespread commercial
use in the past 100 years or so and there has been a thousand-fold
increase in its use in the 20th century, the risk appears to be growing.
Studies under way on occurrence of asbestos fibers in the ice cap
in Greenland suggest that there has been an increase in fiber deposition
331
-------
since the 20th century began. As such, the latency associated with prime
low- level lesion (mesotheliomi) and apparent increase of potential expo-
sure from air pollutants cannot be disregarded.
5.1.10.1 Biological effects of Asbestos --The principal effects of
asbestos exposure in people that are reproducible in experimental animals
include pulmonary fibrosis and carcinogenicity.
5.1.10.1.1 Fibrogenesis—Chrysotile asbestos was shown to produce
314
fibrosis as early as 1925. Amosite and crocidolite have also been
shown to have this property. It now appears that most types of asbes-
tos are fibrogenic. Long fiber length is considered important to fibro-
genesis according to some investigators. But others have found short
fibers to be fibrogenic.
The exact mechanisms that initiate various asbestos lesions are
still unknown. It is possible that the mechanisms for different lesions
(fibrosis, pulmonary, carcinoma and mesothelioma) are all different.
Fibrosis (asbestosis) is usually related to massive dose exposure and
has the shortest period of latency so that asbestos is associated with
all forms of asbestos fibers in industrial exposure. The subject was
329
reviewed in 1972.
There is some evidence that smoking increases the risk of asbesto-
sis. Another variety of fibrosis (parietal pleural calcification)
appears to be strongly associated with industrial exposure to asbestos
and to residents in regions where asbestos is produced. Such situations
314
have been reported in Finland from exposure to anthophyllite' and in
330
Czechoslovakia from exposure to chrysotile.
332
-------
Table 5.36. CORRELATION OF INORGANIC FIBERS WITH ASBESTOS
BODIES IN LUNGS OF 3,000 CONSECUTIVE
AUTOPSIES IN NEW YORK CITY, 1966 TO 1968*
Percentage of cases showing asbestos bodies in
relation to occurrence of thin inorganic fibers.
Thin Inorganic Fibers
rtSUCSUUS
Bodies
0
1-4
5-14
15+
Total
No.
1,168
705
76
13
1,962
%
59.5
35.9
3.8
0.66
65 .,4
No.
332
424
89
20
865
%
38.3
49.0
10.3
2.3
28.8
No.
35
52
21
17
125
I
28.0
41.6
16.8
13.6
4.1
No.
16
13
6
13
48
%
33.3
27.1
12.5
27.1
1.6
Percentage of cases showing thin inorganic fibers
in relation to occurrence of asbestos bodies.
Asbestos Bodies
Thin
Inorganic
Fibers
0
1-4
5-14
15+
Total
* Correlation
0
No.
1,168
332
35
16
1,551
was done
1-4
% No.
75.3 705
21.4 424
2.3 52
1.0 13
51.7 1,194
%
59.0
35.5
4.3
1.1
39.8
by optical microscopic
5-14 15+
No.
76
89
21
6
192
study,
I No.
39.6 13
46.3 20
10.9 17
3.1 13
6.4 63
from Langer et
\
20.6
31.7
27.0
20.6
2.1
i 13
al.
Table 5.37. CHRYSOTILE IN 28 CASES STUDIED BY ELECTRON MICROSCOPY
No. of Chrysotile
Group Fibers + Fibrils Cases Male Female
1
2
3
4
5
"Blank grids"
<9
10-50
51-99
100-200
>201
<9
4/28
11/28
6/28
4/28
3/28
--
3
5
1
4
3
--
1
6
5
0
0
--
333
-------
Most types of asbestos are fibrogenic. The exact mechanism of
action is unclear at this time. But there is evidence that lysosomal
331
injury is involved. Long fibers stimulate more and quicker fibresis
332
than short ones, possibly because they are less well phagocytized
since transport begins with fibers of 20 y and increases as the fibers
become shorter. Short fibers are engulfed by macrophage, long ones
333
are not.
5.1.10.1.2 Pulmonary Carcinogenes is - - Exposure of animals by both inha-
lation and intra-tracheal instillation has produced pulmonary cancer in
animals. In one study of rats given chrysotile, one-third of animals
surviving 16 months or longer had malignancies of the lung divided about
evenly between adenocarcinoma and other tumors consisting of squamous
*ZO A T 7 C
cell carcinoma and fibrosarcoma. One mesothelioma was seen. ' Of
interest is the authors' attribution of their tumor yield to concomitant
presence of trace metals from grinding methods since they contaminated
their asbestos with nickel, cobalt and chromium. But other investi-
gators have produced tumors of the lung and related them to the influence
of asbestos alone.
In a recent study of inhalation, S.P.F. rats were exposed to three
types of amphibole asbestos (amosite, anthophyllite and crocidolite)
•TO £
and two types of chrysotiles (Canadian and Rhodesian). The rate of
deposition and clearance of asbestos dust was studied. Retention was
cumulative with length of exposure. When animals were removed from
exposure, the amount of dust declined. There appeared to be a better
retention of amphibole asbestos (amosite, crocidolite, anthophyllite)
than of chrysotiles (Figure 5.6 and 5.7).
334
-------
Weight of dutt
In lunge (mg)
After removal
from exposure
24 Time (month•)
r
10000 20000
Cumulative dose (mg/m'hours)
—Mean weight of dust in lungs of rat* in relation to dose and time.
314
Figure 5.6. Effects of inhalation of asbestos in rats
335
-------
3 MONTHS
6 MONTHS
12 MONTHS
24 MONTHS
y>0 ?K> IOOO
ANTHOPHYLLITE
5OO 750 IOOO
CROCIDOUTE
_n
XX) ""f 'OOO
CHRYSOTILE Conodion
MO /50 IOOO
CHXYSOTIIE Rhodeiion
500 fM 000
Squomou! (orcinoma
nn
7iO IOOO
750 WOO
500 /yj KOO 500 750 K)00 500 75O KXX)
woo
500 750 IOOO 5OO 750 WOO 500 750 KXX)
li n M>
500 750 IOOO 500 750 KXX) 500 750 KXX)
5OO 75O IOOO 500 75O KXX) 500 750 KXX)
ffl AoVoocarcinoma H Adenoma / odenomatotij D Other causes
'—Distribution of survival times in days after first exposure.
Figure 5.7. Effects of inhalation of asbestos in rats
314
336
-------
Tumor induction was found in all forms of asbestos. No tumors
occurred until 300 days or more after exposure began. More tumors
occurred with 12 months; exposure than with six months exposure. But
there was no difference; between 12 months and two years of exposure.
Of interest is the indication in this report that mesothelioma occurs
in two amphiboles (amosite and crocidolite) with only one day's expo-
sure to asbestos underscoring potential danger for brief exposure by
people.
Tumor frequency relative to treatment is summarized in Table 5.36.
Significant findings in the above study besides successful tumor
induction by inhalation routes include: (1) difference in retention
rate for amphibole compared to chrysotile and (2) were the approxi-
mate equality in the ability of all asbestos tested to induce pulmonary
and mesothelial tumors.
5.1.10.1.3 Mesothelial Tumor Induction--Mesothelial tumor induction in
experimental animals has been previously mentioned. Mesothelial tumors
are readily induced by all forms of asbestos via the inhalation route
(see Table 5.38) even though most experin. ats use intrapleural injection .
A report in 1962 said that crocidolite and chrysotile produced
mensothelial tumors in mammals with intrapleural innoculation of the
fibers. Since then, many reports have been published on other forms
of asbestos and their effects on mice, rats, fowl and hamsters. Experi-
ments suggest that the fibrous nature of the material may have a bearing
on the tumorogenicity and a hypothesis could be advanced that carcino-
genicity of the material is realated to the ability of its fibers to
react with cell membranes without destroying the cells. Also, the
337
-------
presence of trace metals or organics on fibers were not determinates.
Particle size and a relationship between the smallness of the fiber and
the number of mesotheliomas was more obvious. '
Other experiments promoted tumors (sarcomas) of mesothelial tissues
through use of non-asbestos fibers such as fiber glass and aluminum
oxide (AL?0,) when placed on a fiberglass mesh and implanted on the
I* O
pleural surfaces of rats. On the basis of this work, the theory was
advanced that a particular configuration of the crystal or fiber was
the determinant for tumorogenesis, i.e., diameters of 5 ym or less and
lengths of 20 ym or more. However, other studies made with Canadian
chrysotile having extremely small fibers produced many mesotheliomas in
* 320
rats.
But the significance of these reports is somewhat in question
because the tumors induced were sarcomas rather than classical meso-
theliomas that are tumors of mesenchymal and epithelial characteristics.
The histogenesis of pleural and peritoneal tumors is complex. The
earliest neo-plastic response often consists of small pedunculated
nodules with a central core of reticulin surrounded by a surface of
non-differentiate connective tissue. These may spread diffusely over
the mesothelial surface as a sheet of connective tissue cells with
epithelial cells on the surface. It is assumed that these tumors come
from undifferentiated mesenchymal cells in the submesothelial tissue
that become differentiated, giving rise to epithelial cells on the
surface and spindle-shaped connective tissue cells in the depths.
5.1.10.1.4 Intratracheal Studies--Intra-tracheal studies have been
promoted as a convenient method to test the potential of a chemical to
338
-------
•p I
•H O
5 CO
6 o>
r-l O O O O i— 1 OOOi-HrH(Nl
rH i— ( O CM O
CD O O
WH
O
&.
oo
to
LO
0)
&
4-1
O
0)
S3
u
rt
u
o o
c s
4) T-I
d
U
co
o
I
s
bO
4H oj *
O r^
CO CO
• P -H
O nj 5*
2 rH
OOOOvOO
OOi— li— IvOOO
rH rH (XI
O rH O \O (XI CTl
O O rH rH
to
vO
O O i— 1 i— 1 tO LO
O O r- 1 •* K) OO
i— 1 i— 1 O to CNI
oo o o i— 1 \o LO rsi ofvirsi^)-LOto
r— 1 i— 1
to r-- i— i LO
tO
CMrHLO
"3-tOrHrslr-lr}-
CVJ
to
tO O
O O CM tO tO OO
O LO (Nl rH
O
(X]
rH OO LO rH CD LO
(Nl -^ 1^-- tO rH t^
•^f- tO rH (Nl (XI tO
e
3
CO
o
UU
co to oo to
^£r
efl
p
•H
(/)
O
tO
p
r-*
^7
rH tO
CO CO CO
0)
4-i
O cd O O O
T3 T3 S S E
•H
(J i—I tO v£> 1X1
O rH
nj
O
H
10 CO CO 10
rt
crj
+-'
O
!/)
X
rH
339
-------
o o o o o o
o o o o
o o o vo LO
0000
rH rH r-J (N) \O
rH K> t'O r-. LT) CT>
o o o o
O tNl K) *d- rH O
CD O O O
§
0
i— 1 (/)(/)
p) pj p;
03
\O
U ^
10
rH CO Q O O
O "3 E E H
P rH fO ^t
g 7
(A 0
fre
O -iH
O |S
tO
•P t/5
Co f!
0 +3
0
-P M
CO CO1
0 p
3 rH
(/) .Q
-6.3
0
* +
340
-------
induce pulmonary neoplasms and, in particular, to show response to a
co-factor in conjunction with a chemical carcinogen. Comparative
studies were made using chrysotile and intratrachel inhalation and
337
intrapleural injection.
Intra-tracheal Injection gives a dose-response according to the
frequency of injections. The induction of BaP was required to elicit
tumors. (Twenty-nine percent of the rats had tumors.)
Inhalation for 13.5 months on a schedule of two hours a day, five
days a week, at 230 mg/m produced pre-cancerous lesions in 20% of the
rats on chrysotile alone, 59% with BaP added and 38% with tobacco smoke
added. A total of 29% of the rats had tumors. Carcinoma of the lungs
occurred only in the chrysotile BaP group. Previous experiments
338
demonstrated a slowing of the excretion of BaP with asbestos.
Intrapleural Administ r ation was done by injection of chrysotile
three times per month into the right pleural cavity (20 mg fibers in
0.5 ml saline). Almost all the rats experienced local proliferation
of the mesothelium that was presumably pre-neoplastic. Mesothelioma
appeared after eight months and increased with time: 13 of 47 rats
at 8-17 months and 15 of 21 between 18 and 24 months. Forty-six per-
cent of all rats showed mesotheliomas.
As mentioned previously, qualitative differences in fiber size and
shape appear to determine induction of mesotheliomas. Carcinogenesis
appears to be unrelated to trace metal content or presence of organic
matter on fibers. The possible induction of tumors by material other
than commercial asbestos has also been considered. And experimental
data suggest that a number of inorganic chrystals below 5 urn in diameter
341
-------
may be equally effective in inducing these tumors. The implication of
these findings that small, fine particles are the effective component
for these tumors is important for industrial and community air pollution
control.
5.1.10.1.5 Synergistic Effects--Reference has been made earlier to the
remarkable association of asbestos and smoking in promoting bronchio-
genic carcinoma in exposed workers in the asbestos industry. Animal
exposure by inhalation to chrysotile for 16 months has produced up to
33% pulmonary tumor adenocarcinomos, fibro-sarcomas, squamous cell car-
334
cinoma and one mesothelioma. The carcinogens responsible for these
were thought to be contaminated by metallic substances from hammer
milling.
The same author reported similar findings in a second inhalation
335
experiment. The role of BaP in promoting lung tumors with exposure
of animals to various forms of asbestos is known. But the evidence for
337,338
such effect is at present mainly for chrystotile. The other
forms of asbestos have not been sufficiently studied in connection with
333
lung tumors.
Contamination by organics including polycyctic hydrocarbons because
of geological and industrial factors has also been suggested as a causa-
tive factor in pulmonary carcinoma. But the use of fibers after
extensive extraction and purification of the material have shown no
339
difference in this regard.
Evidence now available suggests that either: (1) some direct action
of the asbestos fiber or interaction with naturally occurring co-factors
is responsible for tumorigenesis of the lung in experimental exposure of
342
-------
animals; (2) the addition of BaP or tobacco smoke enhances tumorigenesis
with chiystotile; or (3) all forms of commercial asbestos are more or
less equally carcinogenic.
5.1.10.1.6 In-Vitro Effects- -Much about the basic biological effect of
asbestos can be learned from studies in vitro. This work has been
reviewed in detail in reference and will only be briefly summarized
here.
Among actions reported for asbestos are many similar to silica.
These may be listed as follows:
(1.) Lysis of erythrocytes.
(2.) Cytotoxicity on pulmonary macrophages.
(3.) Stimulation of collagen synthesis.
The mechanisms involved are known in many cases and this information may
eventually have great value for basic understanding of lesions produced
by asbestos, asbestos-like minerals and airborne particles in general.
However, their detailed consideration is beyond the scope of this report.
5.2 MATERIALS
5.2.1 Nitrogen Oxide
The damaging effects of atmospheric nitrogen oxides on materials
are mainly of concern to the textile industry. Nitrogen dioxide (NO,,)
is the troublesome pollutant; nitric oxide (NO) appears to be harmless.
Atmospheric N0? can react with a number of important textile dyes
and produce noticeable color change or fading. This fading problem has
been a problem for the textile industry for decades. A more recent
problem concerns yellow discoloration of un-dyed-white and pastel-colored
fabrics. Chemists have traced this yellowing to the action of NO on
343
-------
certain additives applied to fabrics to enhance their marketing proper-
ties. In the United States, annual gross damage to materials by NCL is
estimated to be three-quarters-of-a-billion dollars.
These cause-effect relationships have been recognized for years.
What is lacking, however, is comprehensive dose-response information.
Such information would help to set secondary air quality standards and
make cost-benefit studies.
5.2.1.1 Field Studies--Sixty-seven dye-fabric combinations were exposed
to different types of urban and rural environments with no sunlight
involved. Fading was observed in 43 fabrics. The magnitude of fading
was greater in urban areas than in corresponding rural areas, but varied
among urban areas. It was impossible to specifically attribute fading
to NCL since fading might have been caused by other pollutants as well.
340
However, NCL was a significant factor for seven of the fabrics.
5.2.1.2 Laboratory Studies--Controlled environment (no light) chamber
studies were conducted to assess individual effects of SCL, NCL and
ozone on 20 selected dye-fabric combinations. NCL was the most potent
pollutant, producing the greatest amount of fading on 13 fabrics.
341
Relative humidity was a significant factor in accelerating fading.
Three randomly selected drapery fabrics were exposed to controlled
environments (with simulated sunlight) containing three pollutants
(SCL, NCL and ozone). Direct and interaction effects of NCL caused
significant fading in one of the fabrics. A relationship was developed
expressing the amount of fading as a function of time, NCL concentra-
tion and moisture content of the air. A relationship also was developed
for the percentage life lost as a function of NCL concentration and
.334
moisture content of the air.
344
-------
Exposures in a controlled environment chamber showed that NO- had
no damaging effect on weathering steel, galvanized (zinc coated) steel,
stressed high-strength aluminum alloy, oil base house paint, acrylic or
vinyl coil coatings, rubber tire sidewalls, marble and concrete.
5.2.2 Sulfur Dioxide (S0?)
S02 accelerates the deterioration of a wide variety of materials
as has been recognized for many decades. Both field and laboratory
studies have shown that SCL-polluted environments (1) increase corrosion
rates of most metals, especially iron, steel and zinc; (2) damage a num-
ber of building materials such as limestone, marble, roofing slate and
mortar; (3) discolor and physically deteriorate statuary and other works
of art; (4) increase the erosion rates of oil-base paints; (5) degrade
textile fibers, especially cotton and nylon and fade dyed fabrics; (6)
embrittle and discolor paper and (7) weaken leather.
Although these cause-effect relationships are well known, the
important need today is for comprehensive information on dose-response
relationships. Such information is necessary to set secondary air-
quality standards and for economists to make cost-benefit studies.
With this in mind, EPA has conducted both field and laboratory
studies. The laboratory studies have included controlled environment
exposures using specially designed chambers that were scheduled to be
used in FY-75 to develop dose-response relationships for economically
vulnerable meterials.
5.2.2.1 Field Studies--Ambient levels of airborne S02 were major fac-
tors in control of rates of corrosion for zinc (galvanized steel) . A
345
-------
simple linear function of average S02 concentration and averaged rela-
tive humidity account for 92% of the variability in average zinc cor-
rosion rates. The useful life of galvanized steel at various average
342
annual S02 concentrations and relative humidities can be predicted.
Ambient levels of airborne S02 were a major factor in accelerating
corrosion rates for steel. Equations were developed for four types of
steel relating depth of corrosion to mean SCL levels and other fac-
343,344
tors.
Paints containing calcium carbonate extender pigments were found to
345
be particularly susceptible to attack by SO- environments.
5.2.2.2 Laboratory Studies--Direct and interaction effects of airborne
SCL accelerate corrosion rates of weathering steel, galvanized (zinc
coated) steel, stressed high strength aluminum alloy and erosion rates
346
of oil-base house paints. Preliminary damage-predictive equations
were developed for weathering steel, galvanized steel, stressed aluminum
alloy and oil-base house paint, based on laboratory data and known physi-
346
cal and chemical relationships.
Airborne SO- does not damage acrylic or vinyl coil (factory applied)
coatings, weaken rayon sidewall reinforcing cord in radial rubber tires
346
or produce fading in three randomly selected drapery fabrics.
The effects of airborne SO- on marble, latex house paint and con-
346
crete were inconclusive.
5.::. 3 Ozone (07)
Damage by atmospheric ozone occurs mainly to two classes of mate-
rials: elastomers and textile dyes. Not all elastomers, however, are
subject to ozone damage. Synthetic elastomers with saturated chemical
346
-------
structures have inherent ozone resistance. These elastomers are rela-
tively expensive and account for only a fraction of the market. The
great bulk of the elastomers, including natural rubber and synthetic
polymers of styrere-butadiene, polybutadiene and polyisoprene, contain
double bonds that are prone to oxidation and therefore sensitive to
ozone attack.
Researchers have developed anti-ozonant additives capable of pro-
tecting elastomers from ozone attack. The protection, however, is not
permanent. But for certain items, such as tires, it usually last for
the life of the item.
Some important textile dyes fade when exposed to ozone. Certain
blue dyes used on acetate and polyester/cotton fabrics and nylon car-
pets have been particularly troublesome. To prevent or mitigate fad-
ing by ozone, the textile industry must use more resistant dyes and/or
inhibitors. But these increase the cost of finished goods.
5.2.3.1 Field Studies--Sixty-seven dye-fabric combinations were exposed
to different types of urban and rural environments in the absence of
sunlight. Two-thirds of the fabrics showed appreciable fading. The
fabrics generally faded significantly more at urban sites than at cor-
responding rural sites. Apparently, air pollutants were responsible for
the observed fading. It was impossible, however, to separate the effects
of individual pollutants because they were mixed with effects of other
347
pollutants.
A number of paint types were exposed to different environments.
Oil-base house paint was the most sensitive. The high smog environment
of Los Angeles produced the most damage. It was not possible, however,
347
-------
to identify what pollutant or combination of pollutants was respon-
sible.
5.2.3.2 Laboratory Studies--In controlled environment (no light) cham-
ber studies, low ozone levels (0.05 ppm) produced visible fading in
half of 20 dye-fabric combinations and high ozone levels (0.5 ppm) in
348
two-thirds of the fabrics.
In controlled environment (including simulated sunlight) chamber
studies of different types of paint, high ozone levels (1.0 ppm) damaged
oil-base house paint, but low ozone levels (0.1 ppm) did not.
In more sophisticated environmental (including simulated sunlight)
chamber studies of different types of paint, ozone levels up to 0.5 ppm
did not significantly damage any of the paint types. Ozone produced
small cracks in rubber tire sidewalls. But their rate of penetration
349
was not enough to significantly reduce the life of the tire.
5.3 WEATHER, VISIBILITY AND CLIMATE
5.3.1 Air Qualtiy Trends Analysis
The importance of meteorology in interpretation of trends has been
recognized in the first annual report of EPA's Monitoring and Data
Analysis Division. An a-typical pattern of significant upward trends in
total suspended particulate (TSP) was found over the Western U.S. for
1968-71. This trend is apparently related to a decrease in rainfall
over the same period.
TSP monitors are affected by contributions of particulate matter
from all sources, including natural dustiness. There appears to be an
increase in fugitive dust emission (i.e., agricultural, upward roads,
construction, etc.) during drier periods.
348
-------
5.3.2 Atmospheric Turbidity
Turbidity observations are made only when the line of sight to the
sun is free of cloud. The measuring instrument used is called a sun-
photometer. It is designed to measure the amount of light received
at wavelengths practically unaffected by water vapor absorption. Mea-
surements are regularly made' at 80 stations, 50 of which are in the
United States.
The Meteorology Laboratory provides the necessary number of instru-
ments and periodically recalibrates them. Although the longest turbidity
record is only from 1961, it. appears that certain urban areas (Youngstown
and New York City) may have atmospheres that are less turbid now than in
the mid-60's. The opposite trend may be occurring in some rural areas.
5.3.3 Visibility
Reduction in visibility because of air pollution is, in most cases,
caused by scattering and absorption by aerosols. Major progress has
been made in development of techniques to determine light scattering and
light absorption and to relate these parameters to meteorological range
and prevailing visibility or visual range.
Light scattering, as measured by the integrating nephelometer, is
highly correlated with sub-micron aerosol volume measured by in-situ
techniques and reasonably well-correlated with sub-micron and total
aerosol mass. Extensive tests in three California cities of relation-
ships between prevailing visibility, light scattering and aerosol mass
concentration indicate that light scattering, as measured by the inte-
grating nephelometer, is a good predictor of prevailing visibility.
349
-------
Progress has been made in determining the contribution of various
aerosol sources to visibility reduction. Because of their hygroscopic
or deliquescent nature, secondary aerosols such as (NH.)?SO., NH.HSO.,
M-LNCL and oxidized organic compounds grow as the relative humidity
increases and cause more visibility reduction than their dry weight
• j- <- 352
indicates.
5.4 ECOLOGICAL
5.4.1 Introduction
The over all effect of man's activities on terrestrial vegetation
and ecosystems and cumulative effects of his activities on estuaries,
coastal areas and the biology of oceans are major ecological concerns.
Studies dealing with effects of toxic substances in the biosphere have
generally focused on toxic substance effects on individual plants and
animals rather than total ecosystems.
The studies usually deal with specific pollutants rather than
interactions between and among pollutants. This approach is used with
the hope that effective control of specific pollutants and determination
of costs of controlling them would be possible, particularly in agri-
culture. This approach is also used because it is more difficult despite
its importance to determine the impact of pollutants on ecosystems in
particular and the biosphere in general.
Dr. G. M. Woodwell, however, has shown that disturbances to
natural (i.e., non-anthropogenic) ecosystems from any cause produce
changes that are similar and predictable. The parameters most changed
by pollution are diversity, structure and ratio of photosynthesis to
repiration.
350
-------
Pollution first affects an ecosystem by changing its diversity.
Sensitive flora and fauna are eliminated. This results in shorter
food chains, fewer symbiotic relationships and less stability. The
species affected are those with low reproductive rates, carnivores at
353
the top of food chains and species with the most complex life cycles.
The structure of an ecosystem is affected next. The trend is toward
plants of lesser stature. Taller plants are eliminated first. Trees are
replaced by shrubs, shrubs by herbs, herbs by low-growing plants and in
some instances by mosses and lichens.
The ratio of photosynthesis to respiration is involved in structural
change. Chronic disturbance often damages the capability of plants to
carry on photosynthesis while not appreciably reducing the total amount
of respiration. That makes plants use more energy than they produce.
Large plants with large respiration requirements are more vulnerable
than small plants.
A reduction in nutrients is also a result of structural change in
an ecosystem. A pool of nutrients is released by the decay of larger
plants. These nutrients cannot be retained in the soil or recycled.
They are lost to aquatic ecosystems where they promote cultural eutro-
phication of rivers and lakes. This process also points out the link
354
between terrestrial and aquatic ecosystems.
Crop agriculture involves complete replacement of a natural eco-
system with a highly simplified ecosystem composed of plants and ani-
mals wholly dependent on care and protection by people and requiring
energy inputs. Historically, the continuing development of agriculture
has decreased diversity and increased artificiality.
351
-------
Since World War II, traditional farming with small fields growing
a variety of crops and interspersed with pasture and hedgerows or fence-
rows has been replaced by monocultural systems that are highly vulnerable
to pollution, pests and diseases because of low genetic diversity.
Standards set for various substances are based primarily on effects
these pollutants have on man and man-made ecosystems. But this is only
one aspect of the pollution problem. Not considered is the threat to
maintenance of a biosphere suitable for life as we know it. And it does
not underscore man's dependence of the integrity of diverse earth eco-
systems: those living systems that probably built and now maintain the
biosphere.
While toxic substances that are mutagenic to man, either through
inhalation or through the food chain, are of great importance, those
pollutants that affect components of earth's biota other than man present
a greater problem. Effects of pollutants are often chronic and cumula-
tive over time rather than short-lived. And the effects such pollutants
have on structure, biological diversity and mineral cycling in ecosystems
353
is not known.
Standards that were set have only superficially attempted to assess
effects of those levels on ecosystems. In addition, they have only
dealt with pollutant interactions in a cursory way. Little is known
about effects of pollutant interactions except for substances involved
in photochemical oxidant formation and with nitrogen oxides, ozone and
hydrocarbons.
The National Primary and Secondary Standards are the same except for
sulfur oxides and particulate matter.
352
-------
5.4.1.1 Carbon Monoxide--The effects of carbon monoxide in the soil and
357
in aquatic ecosystems still need clarification. Inman et al. found
that soil takes up CO in large quantities. But the changes brought
about in soil by CO up-take by microorganisms and subsequent production
358
of methane have only been touched upon.
Plants in general are relatively insensitive to carbon monoxide at
low concentrations toxic to animals.
5.4.1.2 Nitrogen Oxides--The primary importance of nitrogen oxides
involves their participation in photochemical reactions with formation
of peroxyacyl nitrates (PAN). The photochemical reactions that produce
359
oxidants have not yet been thoroughly examined.
Evidence of damage to plants from direct exposure to NO in the
J\.
atmosphere is usually confined to thproximity of specific industrial
sources. For example, damage from high ambient levels of nitrogen
dioxide (NO-) are observed near nitric acid plants. Direct damage by
nitric oxide (NO) and other NO components or derivatives have not been
.A.
delineated in the field because effects on vegetation from the oxides
359
of nitrogen per se are difficult to assess.
Most research delineating the effects of NO en vegetation was done
J\.
in the laboratory.
The biochemical-enzymatic mechanisms by which oxides of nitrogen
directly injure plants is not understood.
No attempts have been made to determine effect of nitrogen oxide
emissions in formation of acid rainfall.
5.4.1.3 Photochemical Oxidants--Photochemical oxidants are products of
sunlight-initiated chemical reactions in the atmosphere that involve
353
-------
hydrocarbons and oxides of nitrogen. Those that have been measured in the
atmosphere include ozone, peroxyacyl nitrates and nitrogen dioxide. The
foregoing shows that photochemical oxidants are a variable mixture of com-
pounds. Each component of this mixture may have a different degree of
phytotoxicity so that the composition of the mixture determines the extent
of the hazard.356
Concentration of total oxidants may vary during the day. While the
one-hour maximum may not be exceeded, chronic injury to vegetation may
occur because of long-term average concentrations.
The presence of other pollutants and changes in environmental con-
ditions may affect tolerance of plants to photochemical oxidants.
Little is known about the tolerance of plants under field condi-
tions. Selection of a genetically uniform susceptible crop and plant-
ing it over large areas could end in large crop losses.
Studies are currently under way to determine effects of photochemi-
cal oxidants on the San Fernardino Forest ecosystem.
The primary standard of 160 yg/m (0.08 ppm) for a one-hour average
may not protect all vegetation from adverse effects. A secondary stan-
dard should recognize that both concentrations and duration of exposure
are important in determining effects on vegetation and ecosystems.
Various factors, e.g., soil and humidity, that affect plant sensitivity
have not been well-documented.
5.4.1.4 Hydrocarbons--Ethylene is the chief hydrocarbon gas that pro-
duces adverse effects at known ambient concentrations. Plants are more
361
sensitive to ethylene than to other hydrocarbon gases. Acetylene,
propylene and vinyl chloride all cause phyto-toxicity symptoms resembl-
ing those of ethylene. But 60 to 500 times the ethylene concentration
354
-------
is required to produce comparable effects. Ranking according to toxi-
city is: ethylene>acetylene>propylene>vinyl chloride.
The effects of poly-nuclear aromatic hydrocarbons on higher plants
have not been adequately assessed. There is some indication that the
aromatic hydrocarbons act as plant hormones.
Studies have shown that gaseous hydrocarbons can be metabolized by
micro-organisms. High traffic areas with high amounts of auto
exhaust have a soil microflora different from that in low traffic areas.
Studies have also shown that natural gas leaking into the soil, because
of its utilization by micro-organisms, can stimulate growth in that
364
soil. Gaseous hydrocarbons, then, can affect soil fertility.
Effects of these changes on the growth of higher plants and eco-
systems in general have not been assessed. In addition, the effects of
hydrocarbons on the microflora of plant surfaces have not been assessed.
Changes in microflora of leaf, bark or stem could increase disease
susceptibility.
Reactions between atmospheric hydrocarbons and sulfur oxides have
not been adequately assessed.
5.4.1.5 Sulfur Oxides--The Secondary Standard of 60 yg/m (0.1 ppm)
annual arithmetic mean and the 24-hour guideline of 260 yg/m (0.10 ppm)
for a 24-hour average were withdrawn in September, 1973. The present
3
standard of 1300 pg/m (0.5 ppm) for a three-hour average may not pro-
tect all vegetation against adverse effects as the standard is close to
exposure levels that pose a risk of detrimental economic effects and
discernible aesthetic injury to susceptible species of plants."
355
-------
Chronic effects of sulfur dioxide and their interaction with other
pollutants such as ozone, nitrogen oxides and hydrocarbons require
further study to consider for the standard. Since environmental condi-
tions not only affect the susceptibility of plants, but effect their
distribution, standards that are adequate for one geographic region may
not be adequate for another.
The present standard is probably not adequate to control long-term
low-level exposure to sulfur dioxide, particularly as it relates to
chronic toxicity of sulfate accumulation in plant foliage. The annual
average was an index of the concentration of sulfur dioxide and the
fraction of time that any measurable concentration was present.
Of increasing concern, but not dealt with by the standards, is the
role of sulfur dioxide in acidic precipitation. Long-term average
annual concentrations play an important role in the nature and severity
of this problem.
The role of sulfur dioxide in changing the pH of plant surfaces has
not been studied in the U.S. Studies in Sweden have indicated reduction
in growth of pine and spruce trees in areas of high S0~ emissions. The
increased susceptibility to disease of plants so affected or the role of
SCL in changing ecosystem structure have not been studied to any
extent.
5.4.1.6 Sulfates/Sulfuric Acid Aerosols--The sulfates that affect vege-
tation are formed by chemical reactivity of SCL as far as is known.
Sulfur dioxide reacting with moisture on the leaf, bark or stem surfaces
produces sulfuric acid and causes leaf damage and pH change in the bark.
Sulfuric acid mist, acid aerosols, acid rain or sulfate-bearing particu-
366
lates that dissolve in the dew produce similar effects.
356
-------
Sulfates are usually considered beneficial to soil and soil orga-
nisms except in those cases where they are attached to specific metallic
ions. Changes in soil pH because of acidic precipitation may result in
soil reactions detrimental to plants and soil-living decomposer orga-
nisms. Soils that are acidic or poorly buffered are particularly
susceptible.
Sulfate that contributes to formation of acid rain is usually attri-
buted to S07 in the atmosphere. The interaction of S07 and nitrogen
tit Lt
oxides, however, may also be involved.
Studies by Likens, Borman and their co-workers deal with effects of
"ZfL *T
acid rainfall on deciduous forest ecosystems. Neither nitrogen oxide
nor sulfur oxide standards have dealt with this problem.
Much overlap exists between the effects of SCL and sulfates.
5.4.1.7 Nitrates--The movement of nitrogen through the atmosphere uses
three separate cycles: the N20 cycle, the NH, cycle, the NH, cycle and
the NO cycle. The NO (nitrogen oxides) cycle has the greatest effect
X X
on plant communities and accordingly, other organisms that reside in
these communities.
NO and N0~ from urban sources may move into the atmosphere where
reactions with 0, and water produce HNO, vapor and eventually nitrate-
salt aerosols. Rain or dry particulate deposition eventually brings
'TfiO
nitrate down to the earth's surface. Robinson and Robbins estimate
that the amount of nitrates brought down is 462 x 10 tons per year.
The greatest amount of nitrate fallout occurs over oceans.
Nitrate fallout of the above-mentioned type has not been directly asso-
ciated with increasing acidity of rainfall in the Northwestern United
States.367
357
-------
The most significant and widespread effects of nitrogen oxides are
indirect effects that occur because of involvement of nitrogen oxides in
the formation of photochemical oxidants. The nitrogen oxides from com-
bustion processes may enter into photochemical reactions and form
peroxyacyl nitrates instead of moving off into the atmosphere in urban
areas. The formation of these compounds is discussed in greater detail
369
in the publication, "Air Quality Criteria for Photochemical Oxidants."
The nitrogen oxide standard should help cut down formation of NCL
and subsequent formation of nitrate aerosols through oxidation. But
reduction of atmospheric nitrates will not influence the problem of ter-
restrial nitrate runoff and euthrophication that results from excess
fertilizer application and barnyard runoff.
5.4.1.8 Beryllium--Beryllium, like many trace elements, produces a varied
reaction in plants. Both beneficial and detrimental effects of beryllium
370
on plant growth processes have been reported.
Studies show that be-yllium in the soil can be taken in and accumu-
lated by plant roots. Severe stunting of bush beans resulted when they
were grown in nutrient concentrations of 0.5, 1.0, 2.0, 3.0 or 5.0 ppm.
Abnormal symptoms were progressively more severe as beryllium concentra-
370
tions increased.
Injury to higher plants from beryllium might occur if high levels
of beryllium were dispersed in soil or ground water where plant roots
could make contact. Beryllium does not seem to inhibit formation and
370
functions of chlorophyll.
The present standard that prevents emission of high levels of beryl-
lium should be adequate for protecting plants.
358
-------
5.4.1.9 Mercury--In the Federal Register, Vol. 38, No. 66 - Friday,
April 6, 1973 p. 8825, the following statement is made: "Current data
on the environmental transport of mercury do not permit clear assessment
of the effect of mercury emissions into the atmosphere on the mercury
content in the aquatic and terrestrial environments." This statement
still holds true. The standard set was intended to protect people from
effects of inhaled mercury.
5.4.1.10 Asbestos--Exposure of wild animals to asbestos is not likely
unless they are in the vicinity of a man-made source. Then, the stan-
dards that protect man should protect them.
Damage to plants is not likely unless the leaves of plants become
coated with asbestos to the point where photosynthesis cannot occur.
359
-------
5.5 REFERENCES
1. Shy, C. M. et al. The Chattanooga School Children Study: Effects
of Community Exposure to Nitrogen Dioxides. Incidence of Acute
Respiratory Illness. J. Air Pollution Control Assoc. 20(9):582-
588, September, 1970.
2. Shy, C. M. et al. The Chattanooga School Children Study: Effects
of Community Exposure to Nitrogen Dioxides. I. Methods, Descrip-
tion of Pollutant Exposure and Results of Ventilatory Function
Testing. J. Air Pollution Control Assoc. 2£(8):539-545, August,
1970.
3. Pearlman, M. E. et al. Nitrogen Dioxide and Lower Respiratory
Illness. Pediatrics £7(2):391-398, February, 1971.
4. Chattanooga, Tennessee - Rossville, Georgia Interstate Air Quality
Study 1967-1968. U.S. Public Health Service National Air Pollu-
tion Control Administration Publication No. APTD-0583, Durham,
N.C., October, 1970.
5. Shy, C. M., L. Niemeyer, L. Truppi and T. English. Re-evaluation
of the Chattanooga School Children Studies and the Health Criteria
for N0? Exposure. In-house technical report, Environmental
Research Center, EPA. Research Triangle Park, N.C. March, 1973.
6. Galke, W. and D. House. Prevalence of Chronic Respiratory Disease
Symptoms in Chattanooga, Tennessee. 1971. (Unpublished Data).
7. Hammer, D., and F. Miller. Lower Respiratory Disease and Nitrogen
Dioxide Retrospective Survey in Chattanooga, Tennessee. 1971.
(Unpublished Data).
8. Riggan, W. et al. Acute Respiratory Disease in Chattanooga, Tenne-
ssee. 1972-1973. (Unpublished Data).
9. Prevalence of Chronic Respiratory Disease Symptoms in the Los
Angeles Basin. 1972. (Unpublished Data).
10. French, J. G., V. Hasselblad, R. Johnson. Aggravation of Asthma by
Air Pollutants. 1971-72 New York-New Jersey Metropolitan Communi-
ties . (Unpublished Data).
11. French, J. G., V. Hasselblad, R. J. Johnson. A Study of Asthmatics
in Two Southeast Communities 1971-72. (Unpublished Data).
12, Von Nieding, G. D., H. Krekler, R. Tuchs, H. M. Wagner, and K.
Koppenhagen. Studies of the Effects of NCL on Lung Function:
Influence on Diffusion, Perfusion and Ventilation in the Lungs.
Int. Arch. Arbeitsmed. 31:61-72. 1973.
360
-------
REFERENCES (continued)
13. Kennedy, M. C. S. Nitrous Fume and Coal-Miners with Emphysema.
Ann. Occup. Hyg. 15:285-300. Pergamon Press, 1972.
14. Report of the Expert Committee on Air Quality Criteria for Oxides
of Nitrogen and Photochemical Oxidants (Japan). Central Council
for Control of Environmental Pollution, Sub Council for Air Pollu-
tion Control. June 20, 1972.
15. Speizer, F. E. and B. G. Ferris, Jr. Exposure to Automobile
Exhaust. I. Prevalence of Respiratory Symptoms and Disease.
Arch. Environ. Health. 26_:313-318. June, 1973.
16. Speizer, F. E. and B. G. Ferris, Jr. Exposure to Automobile Exhaust.
II. Pulmonary Function Measurements. Arch. Environ. Health. 26:
325-329. June, 1973.
17. Burgess, W., L. DiBerardinis and F. E. Speizer. Exposure to Automo-
bile Exhaust. III. An Environmental Assessment. Arch. Environ.
Health. 26;325-329.
18. Sprey, P. Health Effects of Air Pollutants and Their Interrela-
tionships Contract No. 68-01-0471, Submitted to the Environmental
Protection Agency, Washington, D.C. September, 1973.
19. Air Quality Criteria for Nitrogen Oxides. NATO/CCMS Document No.
15, June, 1973.
20. Air Quality and Automobile Emissions Control. MAS Documents Nos.
93-23 and 24. September, 1974.
21. Adamson, I.A.R., and D. H. Bowden. The Type II Cells as Progenitor
of Alveolar, Epithelial Regeneration. Laboratory Investigation.
30:35, 1974.
22. Weibel, E. R. Oxygen Effects on Lung Cells. Arch. Int. Med. 128:
54, 1971.
23. Freeman, G. and G. B. Haydon. Emphysema After Low-Level Exposure
to N02. Arch. Environ. Health. 8_:125, 1964.
24. Riddick, J. H., K. I. Campbell, and D. L. Coffin. Emphysematous
Changes Secondary to Nitrogen Dioxide Exposure to Dog Lungs. Pre-
sented at the Annual Scientific Assembly Am. Soc. Clin. Patholo-
gists, Chicago, September 22-30, 1967.
25. Blair, W. H., M. C. Henry and R. Ehrlich. Chronic Toxicity of
Nitrogen Dioxide: II. Effects of Histopathology of Lung Tissue.
Arch. Environ. Health. 18:186, 1969.
361
-------
REFERENCES (continued)
26. Freeman, G., R. J. Stephens, S. C. Crane and N. J. Furiosi. Lesions
of the Lung in Rats Continuously Exposed to Two Parts Per Million of
Nitrogen Oxide. Arch. Env. Health. 17:181, 1968.
27. Haydon, G. B., G. Freeman, and N. J. Furiosi. Covert Pathogenesis
of NCL-Induced Emphysema in the Rat. Arch. Env. Health. 11:776,
1965.L
28. Freeman, G., S. C. Crane, R. J. Stephens, and N. J. Furiosi.
Environmental Factors in Emphysema and a Model System with N0?.
Yale J. Biol. Med. 4£:566, 1968. L
29. Freeman, G., S. C. Crane, and R. J. Stephens. Pathogenesis of
Nitrogen Dioxide-Induced Lesion in the Rat. A review and presen-
tation of new observations. Am. Rev. Resp. Dis. 9_8_:429, 1968.
30. Freeman, G., R. J. Stephens, S. C. Crane, and N. J. Furiosi.
Lesions of the Lung of Rats Continuously Exposed to 2 ppm of NCL.
Arch. Env. Health. 17:181, 1968.
31. Stephens, R. J., G. Freeman, S. C. Crane, and N. J. Furiosi. Ultra-
structural Change in the Terminal Bronchiole of the Rat Continuously
Exposed During Continuous Low-Level Exposure to Nitrogen Dioxide.
Exp. Mol. Pathol. 14:1, 1971.
32. Stephens, R. J., G. Freeman, S. C. Crane, and N. J. Furiosi.
Embedding
33. Stephens, R. J., G. Freeman, and M. J. Evans. Ultrastructural
Changes in Connective Tissue in the Lungs of Rats Exposed to NO
Arch. Int. Med. 127:873, 1971.
2'
34. Freeman, G., S. C. Crane, and N. J. Furiosi. Healing in Rat Lung
After Subacute Exposure to NO-. Am. Rev. Resp. Dis. 100:662,
1969. L
35. Freeman, G., L. T. Juhos, N. J. Furiosi, R. Mussenden, and T. H. Weiss.
Delayed Maturation of Rat Lung in NO--Containing Environment. (In
Preparation).
36. Bartlet, D. Jr., C. S. Faulkner, and K. Cook. Effect of Chronic Expos-
ure (to 0, and N0~) on Young Rats.
37. Buel, G. C., Y. Tokiwa, and P. K. Mueller. Lung Collagen and Elastin
Denaturation in vivo Following Inhalation of Nitrogen Dioxide. Pre-
sented at the 59th Annual Air Pollution Control Assn. Meeting, San
Francisco, June, 1966. APCA Paper No. 66-7.
362
-------
REFERENCES (continued)
38. Port, C., D. L. Coffin, P. B. Kane, and K. V. Ketals. A Comparative
Study of Experimental and Spontaneous Emphysema. (In Preparation).
39. Freeman, G., L. T. Juhos, N. J. Furiosi, R. Mussenden, R. J. Stephens,
and M. J. Evans. Pathology of Pulmonary Disease from Exposure to
Interdependent Ambient Gases (Nitrogen Dioxide and Ozone). Arch.
Env. Health. 29_:203-210, 1974.
40. Thomas, H. V., P. K. Mueller, and R. L. Lyman. Lipoperoxidation of Lung
Lipids in Rats Exposed to Nitrogen Dioxide. Science. 159_:532, 1968.
41. Estefan, R. M., E. M. Cause, J. R. Rowlands. Electron Spin Resonance
and Optical Studies of the Interaction between N0? and Unsaturated Lipid
Compounds. Environ. Research. 3^:62, 1970.
42. Buel, G. C. Y. Tokiwa, P. K. Mueller. Lung Collagen and Elastin Denatura-
tion in vivo Following Inhalation of Nitrogen Dioxide. Presented at the
59th Annual Air Pollution Control Assn. Meeting, San Francisco, June,
1966. APCA Paper No. 66-7.
43. Buckley, R. D. and 0. J, Balchum. Acute and Chronic Exposure to N0?.
Arch. Env. Health. 14_:424, 1967.
44. Buckley, R. D and 0. J. Balchum. Effects of N07 on LDH Isoenzymes.
Arch. Env. Health. 10:220-223. L
45. Buckley, R. D. and 0. J. Balchum. Enzyme Alteration Following N02
Exposure.
46. Ehrlich, R. Effects of Nitrogen Dioxide on Resistance to Infection.
Bacteriol. Review. 3£:604, 1966.
47. Sherwin, R. P., S. Winnick, and R. D. Buckley. Response to LEH-Positive
Cells in the Lungs of Guinea Pigs Exposed to Nitrogen Dioxide. Am. Rev.
Resp. Dis. 96:319, 1967.
48. Buckley, R. D. and C. G. Loosli. Effects of N02 Inhalation on Germ-
free Mouse Lung. Arch. Env. Health. 18_:251, 1965.
49. Menzel, D. B., J. N. Rhoem, S. D. Lee. Vitamin E., the Environmental
Antioxidant. J. Ag. Food and Chem. 20_:481-486, 1972.
50. Menzel, D. B., R. Slaughter, A. M. Bryant, and H. 0. Jauregeii.
Heinz Bodies Formed in Erythrocytes by Fatty Acid Ozonides. Arch.
Env. Health. (In Press).
51. Menzel, D. B., R. Slaughter, A. M. Bryant, and H. 0. Jauregeii. Pre-
vention of Ozonide-Induced Heinz Bodies in Human Erythrocytes by
Vitamin E. Arch. Env. Health. (In Press).
363
-------
REFERENCES (continued)
52. Msnzel, D. B. The Role of Free Radicals in the Toxicity of
Air Pollutants (Nitrogen Oxides and Ozone). In: Free
Radicals in Molecular Biology. W. A. Prior (ed.), Academic
Press. (In Press).
53. Charles, J. M. and D. B. Menzel. Sulfate Stimulated Release
of Histamine from Lung Fragments. Environmental Research.
(In Press).
54. Coffin, C. L. and D. E. Gardner. Interaction of Biological
Agents and Chemical Air Pollutants. Ann. Occup. Hyg. 15:
219-234, 1972.
55. Coffin, D. L., E. J. Blommer, D. E. Gardner, and R. S.
Holzman. Effect of Air Pollution on Alteration of Sus-
ceptibility to Pulmonary Infection. Proceedings of 3rd
Annual Conference on Atmospheric Contamination in Confined
Space. Aerospace Medical Research Lab. 71-80, 1968.
56. Ehrlich, R. Effects of Nitrogen Dioxide on Resistance to
Respiratory Infection. Bact. Rev. 30_:604-614, 1966.
57. Coffin, D. L. and E. J. Blommer. Acute Toxicity of Irra-
diated Auto Exhaust. Its Indication by Enhancement of
Mortality from Streptoccal pneumonia. Arch. Env. Health.
15:36-37, 1967.
58. Port, C. D., J. D. Fenters, R. Ehrlich, D. L. Coffin, and
D. E. Gardner. Interaction of Nickel Oxide and Influenza
Infection in the Hamster. Abs. Env. Proceedings Conference
on Heavy Metals in the Environment. (In Press) . Env.
Perspectives.
59. Maigetter, R. Z., J. Findlay, J. Fenters, and R. Ehrlich.
Effect of Manganese Dioxide on Resistance to Respiratory
Infections. Abs. Am. Soc. of Microbiology, Chicago, Illinois,
1974.
60. Blommer, E. J. (Unpublished Data) .
364
-------
REFERENCES (continued)
61. Henry, M. C., R. Ehrlich, and W. H. Blair. Effect of Nitrogen Dioxide
on Resistance of Squirrel Monkeys to K. pneumoniae infection. Arch.
Env. Health. 18:580-587, 1969.
62. Coffin, D. L., D. E. Gardner, and R. S. Holzman. Influence of Ozone
on Pulmonary Cells. Arch. Env. Health. 1^:633-636, 1968.
63. Gardner, E. D., R. S. Holzman, and D. L. Coffin. Effects of Nitrogen
Dioxide on Pulmonary Cell Population. J. Bact. 9£: 1041-1043, 1969.
64. Coffin, D. L., D. E. Gardner, and E. J. Blommeir. Time/Dose Response
for NO- Exposure in an Infectivity Model. To be presented at Soviet-
American Symposium on the results of the joint scientific work for
1973-74, Riga Lalvian SSR, December, 1974.
65. Coffin, D. L. and D. E. Gardner. Influence of Mode of Dose on Toxi-
city of NO-. Australia-New Zealand Clean Air Conference, Rotorua,
New Zealana, February, 1975.
66. Aronow and Isbell: Carbon Monoxide Effects on Exercise-Induced
Angina Pectoris. Ann. Intern. Med. 79;392-395, 1973.
67. Anderson, Andelman, Strauch, Fortuin and Knelson: Effect of Low-
Level Carbon Monoxide Exposure on Onset and Duration of Angina
Pectoris. A Study of Ten Patients with Ischemic Heart Disease. Ann.
Intern. Med. 719:46-50, 1973.
68. Aronow, Harris, Isbell, Rohaw and Imperato: Effect of Freeway
Travel on Angina Pectoris. Ann. Intern. Med. 77:669-676, 1972.
69. Aronow, Stemmer and Isbell. Effect of Carbon Monoxide Exposure on
Intermittent Claudication. Circulation 49;415-417, 1974.
70. Horvath, Dahms and Gray. The Determination of Maximal Aerobic Power
Consequent to Carbon Monoxide Exposure. J. Appl. Physiol. (In
Press).
71. Haak, Griggs. Effects of Low-Level Carbon Monoxide on Exercise Per-
formance, Blood Lipids and Clotting Factors. (In Progress).
72. Fodor and Winneke. Effect of Low CO Concentrations on Resistance
to Monotony and on Psychomotor Capacity. Staub Reinhart. Luft.
(English ed.) 32 (4) 46-54, 1972.
73. Groll-Knapp, Wagner, Hauck and Haider. Effects of Low Carbon Monox-
ide Concentrations on Vigilance and Computer Analyzed Brain Poten-
tials. Staub Reinhart. Luft. 32:64-68, 1972.
365
-------
REFERENCES (continued)
74. Horvath, Dahms and O'Honlon. Carbon Monoxide and Human Vigilance.
A Deleterious Effect of Present Urban Concentrations. Arch.
Environ. Health. 23^:342-347, 1971.
75. Beard, R. R. and N. Grandstaff. CO Exposure and Cerebral Function.
Arch. Environ. Health. 2^:154-164, 1970.
76. Wald, Howard, Smith and Kjeldsen: Association Between Atheroscle-
rotic Disease and Carboxyhemoglobin Levels in Tobacco Smokers.
Brit. Med. J. 3^:761-765, 1973.
77. Astrup et al.: Effect of Moderate CO Exposure on Fetal Development.
Lancet 2_: 1220-1222, 1972.
78. Zorn, H. The Partial Oxygen Pressure in the Brain and Liver at
Subtoxic Concentrations of Carbon Monoxide. Staub-Reinhalt. Luft
(English ed.) 32_:24-29, 1972.
79. Rondia, D. Abaissement de 1 activte de la benzopyrine-hydroxylase
hepatique in vivo apres inhalation d'ozyde de carbone. C. R. Acad.
Sci. (d) 771T6T7-619, 1970.
80. Coburn, R. F. and P. B. Kane. Maximal Erythrocyte and Hemoglobin
Actabolism. J. Clin. Invest. £7:1435-1446, 1968.
81. Adams. J. D., H. H. Erickson and H. L. Stone. Myocardial Metab-
olism During Exposure to Carbon Monoxide in the Conscious Dog.
J. Appl. Physiol. 34:238-242, 1973.
82. Debias, D. A., C. M. Banerjee, N. C. Birkhead, W; V. Harrer and
L. A. Kazal. Carbon Monoxide Inhalation Effects Following Myo-
cardial Infarctions in Monkeys. Arch. Environ. Health 27:161-
167, 1973.
83. Wells, L. L. The Prenatal Effect of Carbon Monoxide on Albino
Rats and Resulting Neuropathology. Biologist. L5_:80-81, 1933.
84. Williams, I. R. and E. Smith. Blood Picture, Reproduction and
General Condition During Daily Exposure to Illuminating Gas.
Amer. J. Physiol. 110;611-615, 1935.
85. Baker, F. D. and C. F. Tumasonis. Carbon Monoxide and Avian
Embryogenesis. Arch. Environ. Health. 24^:53-61, 1972.
86. Baker, F. D., C. F. Tumasonis and J. Barren. The Effect of Carbon
Monoxide Inhalation on the Mixed-Function Oxidase Activity in the
Chick Embryo and the Adult Mouse. Bull. Environ. Contam. Toxicol.
9_: 329-336, 1973.
366
-------
REFERENCES (continued)
87. McGrath, J. J. and J. Jaeger. Effect of lodoacetate on the Carbon
Monoxide Tolerance of the Chick. Resp. Physiol. 12_:46-53, 1971.
88. Jones, R. A., J. A. Strickland, J. A. Stunkard and J. Siegel.
Effects on Experimental Animals of Long-Term Inhalation to Carbon
Monoxide. Toxicol. Appl. Pharmacol. 19_:46-53, 1971.
89. Ator, N. A. Effects of Carbon Monoxide on DRL Performance in Rats.
(Atstract of a paper to be presented at American Psychological
Association Meetings, New Orleans, September, 1974).
90. Gage, M. I., Y. Y. Yang., A. L. Cohen and J. F. Stara. Alterations
of Wheel Running Behavior of Mice by Automotive Fuel Emissions.
Paper presented at American Psychological Association Meeting, 1972.
91. Particulate Polycylic Organic Matter. National Academy of Sciences,
Washington, B.C., 1972.
92. Aksoy, M., S. Erdem, K. Dincol, T. Hepyuksel, and G. Dincol.
Chronic Exposure to Benzene as a Possible Contributary Etiologic
Factor in Hodgkin's Disease. Blut Band XXVIII, Seite 293-298,
1974.
93. Ahmad and T. C. Whitson. Formaldehyde: How Much of a Hazard.
Industrial Medicine and Surgery, September, 1973.
94. Thorpe, J. J. Epidemiologic Survey of Leukemia in Persons Poten-
tially Exposed to Benzene. Journal of Occupational Medicine, Vol.
16, No. 6. June, 1974.
95. Rylander, R. Toxicity of Cigarette Smoke Components: Free Lung
Cell Response in Acute Exposures. American Review of Resp. Dis.
108:1279-1282, 1973.
96. Stedman, R. L. The Chemical Composition of Tobacco and Tobacco
Smoke. Chem. Rev. .68:153-207, 1968.
97. Waltz, P. et al. MeThodeu der Quantitativen Bestimmung des
Brenzcatechins in Cigrettenrauch. Beiter Tabakforsch. _3:263-277,
1965.
98. Van Duuren, B. L. et al. Carcinogenic Agents in Tobacco Carcino-
genesis. J. Natl. Cancer Inst. .51:703-705, 1973.
99. Chemistry of Carbon Compounds (Rodd, E. H., ed.), Vol. 3A. Amster-
dam, Elsevier Publishing Co., 1956, p. 465.
367
-------
REFERENCES (continued)
100. Johnstone, R. A., J. R. Plimmer. The Chemical Constituents of
Tobacco and Tobacco Smoke. Chem. Rev. 59^885-936, 1959.
101. Figueroa, W. G. et al. Lung Cancer in Chloromethyl Methyl Ether
Workers. The New Eng. J. Med. 288_:1096-1097, 1973.
102. Van Duuren, B. L., B. M. Goldschmidt, B. S. Catz, et al. Alpha-
haloethers: A New Type of Alkylating Carcinogen. Arch. Environ.
Health 16^:472-476, 1968.
103. Leong, B. K. J., H. N. MacFarland, W, H. Reese, Jr. Induction of
Lung Adenomas by Chronic Inhalation of Bis (Chloromethyl) ether.
Arch. Environ. Health 22^663-666, 1971.
104. Laskin, S. M. Kuschner, R. T. Drew, et al. Tumors of the Respira-
tory Tract Induced by Inhalation of Bis (Chloromethyl) Ether.
Arch. Environ. Health 23:135-136, 1971.
105. Van Duuren, B. L., C. Katz, B. M. Goldschmidt, et al. Carcino-
genicity of Haloethers. II. Structure-Activity Relationships
of Analogs of Bis (Schloromethyl) Ether. J. Natl. Cancer Inst.
48^1431-1439, 1972.
106. Alvares, A. P. et al. Inducibility of Benzo(a)Pyrene Hydroxylase
in Human Skin by Polycyclic Hydrocarbons. Chemical Pharm. and
Thera. 1£: 30-40, 1973.
107. Wiebell, F. J. et al. Aryl Hydrocarbon (Benzo(a)Pyrene) Hydro-
lase: Inducible in Extraphepatic Tissues of Mouse Strains Not
Inducible in Livei. Archives of Biochem. and Biophy. 154:292-
294, 1973.
108. Federal Register, Vol. 36, No. 84, Part II, pp. 8186-8190,
April 30, 1971.
109. Air Quality Criteria for Photochemical Oxidants. USDHEW, PHS,
EHS, NAPCA, Washington, D.C., March, 1970, No. AP-63.
110. Bates, D. V. et al. Short-Term Effects of Ozone on the Lung.
J. Appl. Physiol. 32(2), February, 1972.
111. Hazucha, M., F. Silverman, C. Parent, S. Fields and D. V. Bates.
Pulmonary Function in Man after Short-Term Exposure to Ozone.
Arch. Env. Health 27^183-188, September, 1973.
112. Bates, D. Hydrocarbons and Oxidants: Clinical Studies Paper
Presented at Conference on Health Effects of Air Pollutants,
National Academy of Sciences, Washington, D.C., October 4, 1973.
368
-------
REFERENCES (continued)
113. Hammer, D. I., V. Hasselblad, B. Portnoy and P. F. Wehrle. Los
Angeles Student Nurse Study. Daily Symptom Reporting and Photo-
chemical Oxidants. Arch. Environ. Health 28;255-260, May, 1974.
114. Challen, P. et al. An Investigation of Some Health Hazards in an
Inert-Gas Tungsten-Arc Welding Shop. Brit. J. Indus. Med. 15:
276, 1958.
115. Young, W. A. et al. Pulmonary Function in Welders Exposed to Ozone.
Arch. Environ. Health ^:737, 1963.
116. Kerr, H. D., T. J. Kulle, M. L. Mcllhany and P. Swidersky. Effects
of Ozone Function in Normal Subjects. An Environmental Chamber
Study. (In Press).
117. Hackney, J. D. Physiological Effects of Air Pollutants in Humans
Subjected to Secondary Stress. Final Report, January 1, 1973,
through June 30, 1974. State of California Air Resources Board
Contract No. ARB 2-372.
118. Air Quality Criteria for Photochemical Oxidants and Related Hydro-
carbons. Committee on Challenges of Modern Society. NATO, G.
Hueter, Chairman, 1974, pp. 6-13.
119. Pan, A., J. Belaud and Z Jegier. Ozone-Induced Arterial Lesions.
Arch. Environ. Health 2_4:229-232, 1972.
120. Goldstein, E. and P. Hoeprich. Influence of Ozone on Pulmonary
Defense Mechanisms of Silicotic Mice. Arch. Environ. Health
24_: 444-448, 1972.
121. Goldstein, E., D. Warshaver, W. Lippert and B. Tarkington. 0., and
N09 Exposure. Arch. Environ. Health 28:85-90, 1974.
Lt '—L—L
122. Dowell, A., L. Lohrbauer, D. Hurst, S. Lee. Rabbit Alveolar
Macrophage Damage Caused by .m vivo 0, Inhalation. Arch. Environ.
Health 21:121-127, 1970.
123. Emik, L, R. Plata, K. I. Campbell, and G. L. Clarke. Biologi-
cal Effects of Urban Air Pollution. Arch. Environ. Health 25:
335-342, 1971.
124. Evans, M., W. Mayr, T. Bils and C. Loosli. Effects of Ozone on
Cell Renewal in Pulmonary Alveoli of Aging Mice. Arch. Environ.
Health 22:450-453, 1971.
369
-------
REFERENCES (continued)
125. Hattori, K., N. Kato, M. Kinoshitan, S. Kinoshita and T. Sunada.
Protective Effect of Ozone in Mice Against Whole-Body X-Irradia-
tion. Nature (London), Vol. 198:1220, 1963.
126. Matsumura, Y., K. Mizuno, T. Miyamoto, T. Suzuki and Y. Oshima.
The Effects of Ozone, Nitrogen Dioxide and Sulfur Dioxide on
Experimentally Induced Allergic Respiratory Disorder in Guinea
Pigs, IV Effects of Respiratory Sensitivity to Inhaled Acetyl-
choline. Am. Rev. Respirat. Dis. 1015:262-267, 1972.
127. Fletcher, B. L. and A. L. Tappel. Protective Effects of Dietary
Alpha-Tocopherol in Rats Exposed to Toxic Levels of Ozone and
Nitrogen Dioxide. Environ. Res. 6_: 165-175, 1973.
128. Moorman, W. J., J. J. Chmiel, J. F. Stara and T. R. Lewis. Com-
parative Decomposition of Ozone in the Nasopharynx of Beagles.
Actue vs. Chronic Exposure. Arch. Environ. Health 26:153-155,
1973.
129. Rochm, N., J. G. Hadley and D. B. Menzel. Antioxidants vs. Lung
Disease. Arch. Internal. Med. 1_28:88-93, 1971.
130. Roehm, J., J. Hadley and D. Menzel. Influence of Vitamin E on
the Lung Fatty Acids of Rats Exposed to Ov Arch. Environ.
Health 24;237-242, 1972.
131. Shakman, R. Nutritional Influences on the Toxicity of Environ-
mental Pollutants. Arch. Environ. Health 2,8:105-113, 1974.
132. Kyei-Aboagye, K., M. Hazucha, I. Wyszogrodski, D. Rubinstein and
M. E. Avery. The Effect of Ozone Exposure in vivo on the Appear-
ance of Lung Tissue Lipids in the EndobroncHTal Lavage of Rabbits.
Biochem. Biophys. Res. Commun. 54_:907-913, 1973.
133. Seto, K., M. Kawakami, M. Takeshima, M. Kon, K. Sugita and M.
Shishido. Influence of Short-Period Exposure to Ozone and
Nitrogen Dioxide on the Formation of Phospholipids in the Lungs
of Rabbits. Igaku to Seibutsugaku. 8^7:57-60, 1973.
134. Castleman, W. L., D. L. Dungworth and W. S. Tyler. Cytochemi-
cally Detected Alterations of Lung Acid Phosphatase Reactivity
Following Ozone Exposure. Lab. Invest. 29^310-319, 1973.
135. Castleman, W. L., D. L. Dungworth and W. S. Tyler. Histochemically
Detected Enzymatic Alterations in Rat Lung Exposed to Ozone. Exp.
Mol. Pathol. 19:402-421, 1973.
370
-------
REFERENCES (continued)
136. Chow, and Tappel. Activities of Pentose Shunt and Glycolytic
Enzymes in Lungs of Ozone Exposed Rats. Arch. Environ. Health
26:205-208, 1973.
137. Dillard, C. J., N. Urribarri, K. Reddy, B. Fletcher, S. Taylor, B.
Delumen, S. Langberg, A. L. Tappel. Increased Lysosomal Enzymes
in Lungs of Ozone-Exposed Rats. Arch. Environ. Health 25:426-
431, 1972. ~
138. Jegier, Z. Ozone as an: Air Pollutant. Can. J. Public Health
(Toronto). 64_: 161-166, 1973.
139. Mustafa, M. G., A. J. Delucia, G. K. York, C. Arth and C. E. Cross.
Ozone Interaction with Rodent Lung. II. Effects on Oxygen Consump-
tion of Mitochondria. J. Lab. and Clin. Med. 82^:557-565, 1973.
140. Pan, A. S. and A. Jegier. Trypsin Protein Esterase in Relation
to Ozone-Induced Vascular Damage. Arch. Environ. Health 24:255-
256, 1972.
141. Pan, A. and Z. Jegier. The Effect of Sulfur Dioxide and Ozone on
Acetylcholinesterase. Arch. Environ. Health 21_: 498-501, 1970.
142. Veninga, T. S. Ozone-Induced Alterations in Murine Blood and
Liver. Preprint, International Union of Air Pollution Prevention
Associations, 10P, 1970, 6 Refs. (Presented at the International
Clean Air Congress, 2nd, Washington, D.C., December 6-11, 1970,
Paper MB-15E.)
143. Veninga, T. S. Toxicity of Ozone in Comparison with Ionizing
Radiation. Strahlentherapie (Munich). 134_:469-477, 1967.
144. Yokoyama, Eiji. The Effects of Low-Concentrations of Ozone on
the Lung Pressure of Rabbits and Rats. Preprint, Japan Society
of Industrial Hygiene, pp. 134-135, 1973. (Presented at the
Japan Society of Industrial Hygiene Annual Meeting, 46th, April,
1975, Paper 225.)
145. Yokoyama, Eigi. A Comparison of the Effects of SO-, NO- and 0,
on the Pulmonary Ventilation of Guinea Pigs. Sangyo Igaku (Jap.
J. Ind. Health). 11^:563-568, 1969.
146. Bloch, W. N., F. J. Miller and T. R. Lewis. Pulmonary Hyperten-
sion in Dogs Exposed to Ozone. Air Pollution Control Office,
Cincinnati, Ohio, Biological Research Br., 10P., 1971.
147. Goldstein, E., W. Tyler, P. Hoeprich and C. Eagle. Ozone and the
Antibacterial Defense Mechanisms of the Murine Lung. Arch. Internal
Med. 128:1099-1102, 1971.
371
-------
REFERENCES (continued)
148. Goldstein, E., W. S. Tyler, P. D. Hoeprich and C. Eagle. Adverse
Influence of Ozone on Pulmonary Bactericidal Activity of Murine
Lung. Nature 229:262-263, 1971.
149. Friberg, L., B. Holma, M. Piscator and R. Rylander. Effects of
Prolonged Exposure to Low Concentrations of Lead and Ozone in
Rabbits and Guinea Pigs--with Special Reference to Lung Clearance
and Alveolar Macrophages. Karolinska Inst., Stockholm (Sweden),
Dept. of Hygiene; National Inst. of Public Health, Stockholm
(Sweden), Dept. of Environmental Health; Copenhagen Univ. (Den-
mark), Inst. of Hygiene, PHS Grant AP 00212, Swedish Medical
Research Council Proj. B69-14x-18, 26P., 1970.
150. Fairchild, G. A. Ozone Effect on Respiratory Deposition of Vesi-
cular Stomatitis Virus Aerosols. Am. Rev. Resp. Dis. 109:446-
451, 1974.
151. Gregory, A., L. Ripperton and B. Miller. Effect of Neonatal
Thymectomy on the Development of Ozone Tolerance in Mice. Am.
Ind. Hyg. Assoc. J. 28^:278-283, 1967.
152. Gardner, D. E., J. Illing, F. Miller and D. Coffin. 1974. The
Effect of Ozone on Pentobarbital Sleeping Time in Mice. EPA,
NERC/RTP, EBL, PER, (in preparation).
153. Schlipkoeter, H. and J. Bruch. Functional and Morphological
Alterations Caused by Exposure to Ozone Zentralbl. Bakteriol,
Parasitenk. Infektionskr. Hyg. Abt. l:orig, Reihe B, 156:486-
499, 1973.
154. Mizoguchi, I., M. Osawa, Y. Sato, K. Makino and H. Yagyu.
Studies on Erythrocyte and Photochemical Smog, Part I: Effects
of Air Pollutants on Erythrocyte Resistance. Taiki Osen Kenkyu
J. Japan. Soc. Air Pollution. 8^:414, 1973.
155. Matsumura, Y. The Effects of Ozone, Nitrogen Dioxide, and Sulfur
Dioxide on the Experimentally Induced Allergic Respiratory Dis-
order in Guinea Pigs. I. The Effect on Sensitization with
Albumin Through the Airway. AM Rev. Respirat. Diseases. 102:
430-437, 1970.
156. Roth, R. P. and M. F. Tansy, Effects of Gaseous Air Pollutants
on Gastric Secreto-Motor Activities in The Rat. J. Air Pollu-
tion Control Assoc. 22;706-709, 1972.
157. Atwal, Onkar and T. Wilson. Parathyroid Gland Changes Following
Ozone Inhalation. Arch. Environ. Health. 28:91-100, 1974.
372
-------
REFERENCES (continued)
158. Nettesheim, P., M. G. Hanna, Jr., D. G. Doherty, R. F. Newell and
A. Hellman. Effects of Chronic Exposure to Artificial Smog and
Chromium Oxide Dust on the Incidence of Lung Tumors in Mice.
National Cancer Inst., Bethesda, Md., and Atomic Energy Commission,
Oak Ridge, Tennessee, Proc. Conf. Inhalation Carcinogenesis,
Gatlinburg, Tennessee, 1969, pp. 305-320.
159. Schlipoeter, H. W., G. G. Fodor, L. Ghelerter and R. Dolgner.
Findings from Animal Tests Concerning Synergism. Proc. Int. Clean
Air Cong., 3rd, Duesseldorf, West Germany, 1973, P. A26-A30.
160. Cortes, R. and 0. Privett. Toxicity of Fatty Ozonides and Per-
oxides. Lipids. 7_: 715-721, 1972.
161. Seto, K, M. Kawakami, K. Sugita, M. Shishido, M. Yamaji and T.
Tsuda. Influence of Inhalation of Ozone and Nitrogen Dioxide
on the Formation of Phospholipids on the Lung. Igaku to
Seibutsugaku 86:317-320, 1973.
162. Shakman. Nutritional Influences on Toxicity of Environmental
Pollutants. Arch. Environ. Health. 28.: 105-113, 1974.
163, Trams, E., C. Lauter, E. Brown and 0. Young. Cerebral Cortical
Metabolism after Chronic Exposure to Ov Arch. Environ. Health.
24_:153-159, 1972. *
164. Litt, R. S., S. Vaughan, P. Birkinshaw, H. Coit and B. Sanders.
Effects of Low Concentrations of Ozone on Temporal Discrimination.
In: Air Pollution Project: An Educational Experiment in Self-
Directed Research, 1968 Associated Students of the California
Inst. of Tech., Pasadena, pp. 51-64, 1968.
165. Xintaras, C., B. L. Johnson, C. E. Ulrich, R. E. Terrill and M. F.
Sosecki. Application of the Evoked Response Technique in Air
Pollution Toxicology. Toxicol. Appl. Pharmacol. 8^77-87, 1966.
166. Konigsbergs, A. A Study of the Measurement of Ozone and Some
Effects of Ozone and Nitrogen Oxides on the Activity of Rats.
Syracuse Univ., N.Y., Eept. of Physics, Thesis (Ph.D.), January,
1969, Ann Arbor, Michigan, Univ. Microfilms, Inc., 1969.
167. Reynolds, R. W. and R. R. J. Chaffee. Studies on the Combined
Effects of Ozone and a Hot Environment on Reaction Time in Sub-
human Primates. In: Project Clean Air. California University,
Santa Barbara, California, Univ. Res. Proj. 5-6 8P., September 1,
1970.
373
-------
REFERENCES (continued)
168. Martin, A. E. and W. Bradley. Mortality, Fog and Atmospheric
Pollution. Monthly Bull. Ministry Health 19;.56-59, 1960.
169. Lawther, P. J. Compliance with the Clean Air Act: Medical Aspects,
J. Inst. Fuels (London) 36:341-344, 1963.
170. Report of the International Joint Commission, United States and
Canada, on the Pollution of the Atmosphere in the Detroist River
Area. International Joint Commission (United States and Canada).
1960. p. 115.
171. Greenburg, L., F. Field, J. I. Reed and C. L. Erhardt. Air Pollu-
tion and Morbidity in New York City. J. Amer. Med. Assoc.
182.: 161-164, 1962.
172. Greenburg, L., M. B. Jacobs, B. M. Drolette, F. Field and M. M.
Braverman. Report on an Air Pollution Incident in New York City,
November, 1953. Public Health Reports. 77:7-16, 1962.
173. Buck, S. F. and D. A. Brown. Mortality from Lung Cancer and
Bronchitis in Relations to Smoke and Sulfur Dioxide Concentration,
Population Density and Social Index. Research Paper No. 7.
Tobacco Research Council, London, Fjigland. 1964.
174. Lawther, P. J. Climate, Air Pollution and Chronic Bronchitis.
Proc. Roy. Soc. Med. 5.1:262-264, 1958.
175. Winkelstein, W. The Relationship of Air Pollution and Economic
Status to Total Mortality and Selected Respiratory System Mor-
tality in Man. Arch. Environ. Health. 14_:162-169, 1967.
176. Holland, W. W., D. D. Reid, R. Seltser and R. W. Stone. Respira-
tory Disease in England and the United States. Studies of
Comparative Prevalence. Arch. Environ. Health. 10:338-545,
1965.
177. Douglas, J. W. B. and R. E. Waller. Air Pollution and Respiratory
Infection in Children. Brit. J. Prevent. Soc. Med. 20^:1-8, 1966.
178. Lunn, J. E., J. Knowelden and A. J. Handyside. Patterns of
Respiratory Illness in Sheffield Infant School Children. Brit.
J. Prevent. Soc. Med. 21^:7-16, 1967.
179. Nelson, C. J. C. M. Shy, T. English, C. R. Sharp, R. Andleman,
L. Truppi and J. Van Bruggen. Family Surveys of Irritation Symp-
toms During Acute Air Pollution Exposures. J. Air Pol. Control
Assoc. 23_(2):81-86, 1973.
374
-------
REFERENCES (continued)
180. Health Consequences of Sulfur Oxides: A Report from CHESS, U.S.
Environmental Protection Agency (in press).
181. Shy, C. M., V. HasselbLad, R. M. Burton, C. J. Nelson and A. A.
Cohen. Air Pollution Effects on Ventilatory Function of U.S.
School Children. Arch. Environ. Health. 27:124-128, 1973.
182. French, J. G., G. Lowrimore, W. C. Nelson, J. F. Finklea, T.
English and M. Hertz. The Effect of Sulfur Dioxide and Suspended
Sulfates on Acute Respiratory Disease. Arch. Environ. Health.
27:129-133, 1973.
183. Chapman, R. S., C. M. Shy, J. F. Finklea, D. E. House, H. E.
Goldberg and C. G. Hayes. Chronic Respiratory Disease in Military
Inductees and Parents of School Children. Arch. Environ. Health.
27:138-142, 1973.
184. Galke, W., D. House, et al. Prevalence of Chronic Respiratory
Disease Symptoms in Adults: 1971-1972 Survey of Two Southeastern
United States Communities. In-house Technical Report. October 18,
1974.
185. Hammer, D. I., F. J. Miller, D. E. House, K. E. McClain, E.
Tompkins and C. G. Hayes. Frequency of Acute Lower Respiratory
Disease in Children; Retrospective Survey of Two Southeastern
Communities, 1968-1971. In-house Technical Report, May 8, 1974.
186. Love, G. et al. Acute Respiratory Disease in Families in Two
Southeastern Communities.
187. Galke, W. A., D. E. House, et al. Prevalence of Chronic Respira-
tory Disease Symptoms in New York Area Adults--1972. In-house
Technical Report, October 31, 1974.
188. Hammer, D. I., F. J. Miller, A. G. Stead and C. G. Hayes. Acute
Lower Respiratory Disease in Children in Relation to Sulfur
Dioxide and Particulate Air Pollution: Retrospective Survey in
New York City, 1972. In-house Technical Report, September 13,
1974.
189. Fogleman, E. et al. Acute Air Pollution Exposures in Two South-
eastern Communities, 1971-1972. Draft Report, March 11, 1974.
190. French, J. et al. Aggravation of Asthma Symptoms in Two South-
eastern Communities, 1971-1972. Draft Report.
191. Stebbings, J. Frequency and Severity of Cardiopulmonary Symptoms
in Adult Panels, 1971-1972 New York Studies. Draft Report.
375
-------
REFERENCES (continued)
192. French, J. et al. Aggravation of Asthma Symptoms in the New York-
New Jersey Area, 1971-1972. Draft Report.
193. Love, G. J. et al. Summary and Conclusions. Respirable Particu-
late Monograph. To be published.
194. McJilton, C., R. Frank and R. Charlson. Role of Relative Humidity
in the Synergistic Effect of a Sulfur Dioxide-Aerosol Mixture on
the Lung. Science. 182:505-504, November, 1973.
195. Morrow, P. E. Experimental Studies on Inhaled Materials. A Basis
for Respiratory Models. Arch. Internal Med. 126_:466-470, 1970.
196. Task Group on Lung Dynamics. Deposition and Retention Models for
Internal Dosimetry of the Human Respiratory Tract. Health Physics.
n:173-207, 1966.
197. Air Quality Criteria for Particulate Matter, Chapter 9. National
Air Pollution Control Administration Publication No. AP-49, 1969.
198. Amdur, M. 0. and M. Corn. The Irritant Potency of Zinc Ammonium
Sulfate of Different Particle Sizes. Am. Indust. Hyg. Assoc. J.
24_:326-333, July-August, 1963.
199. Amdur, M. 0. and D. W. Underbill. Response of Guinea Pigs to
Sulfur Dioxide. Arch. Environ. Health. 16_:461-468, April, 1968.
200. Amdur, M. 0. Toxicologic Appraisal of Particulate Matter, Oxides
of Sulfur and Sulfuric Acid. J. Air Poll. Cont. Assn. 19_:639-
646, September, 1969.
201. Hatch, T. and P. Gross. Pulmonary Deposition and Retention of
Inhaled Aerosols. Academic Press, New York, 1964.
202. Dimmick, R. L. and A. B. Akers. An Introduction to Experimental
Aerobiology. Chapter 16. Wiley-Interscience, New York, 1969.
203. Palm, P. E., J. M. McNerney and T. Hatch. Respiratory Dust Reten-
tion in Small Animals. A Comparison with Man. A.M.A. Arch. Ind.
Health. 13;355-365, 1956.
204. Fairchild, G. A. Measurement of Respiratory Volume for Virus
Retention Studies in Mice. Appl. Microbiology. 24:812-818,
November, 1972.
205. Lippmann, M., R. E. Albert and H. T. Peterson, Jr. The Regional
Deposition of Inhaled Aerosols in Man. In: Inhaled Particles
III. Vol. 1. Ed. W. H. Walton, Unwin, London, 1971.
376
-------
REFERENCES (continued)
206• Lippmann, M. Deposition and Clearance of Inhaled Particles in the
Human Nose. Ann. Otology, Rhin. and Laryng. 79:519-529, June,
1970, ~
207, Casarett, L. J. The Vital Sacs: Alveolar Clearance Mechanisms
in Inhalation Toxicology. In: Essays in Toxicology. Ed. Wayland
J. Hayes, Jr., Academic Press, New York, 1972.
208. Bates, D. V., G. M. Bell, C. D. Burnham, M. Hazucha, J. Mantha,
L. D. Pengelly, F. Silverman. Short-Term Effects of Ozone on the
Lung. J. Appl. Physiol. 32^:176-181, 1972.
209. McCarroll, J. Measurements of Morbidity and Mortality Related to
Air Pollution. APCA Journal. 17:203-209, April, 1967.
210. Thompson, D. J., M. Lebowitz, E. J. Cassell, D. Wolter and J.
McCarroll. Health and the Urban Environment. VIII--Air Pollution,
Weather and the Common Cold. Am. J. Pub. Health. 6£:731-739,
April, 1970.
211. Ipsen, J., M. Deane and F. E. Ingenito. Relationships of Acute
Respiratory Disease to Atmospheric Pollution and Meteorological
Conditions. Arch. Environ. Health. 18^462-472, April, 1969.
212. Finklea, J. F., D. C. Calafiore, C. J. Nelson, W. B. Riggan and
C. B. Hayes. Aggravation of Asthma by Air Pollution: 1971 Salt
Lake Basic Studies. Health Consequences of Air Pollution: A
Report of the CHESS Program, 1970-71, EPA Technical Report.
213. Lippmann, M. and R. E. Albert. The Effect of Particle Size on
the Regional Deposition of Inhaled Aerosols in the Human Respira-
tory Tract. Am. Indust. Hyg. Assn. J. 30:257-275, May-June,
1969.
214. Morrow, P. E., F. R. Gibb and L. Johnson. Clearance of Insoluble
Dust from the Lower Respiratory Tract. Health Physics. 10:543-
555, 1964.
215. Morrow, P. E., F. R. Gibb and L. J. Leach. The Clearance of
Uranium Dioxide Dust from the Lungs Following Single and Multiple
Inhalation Exposures. Health Physics. 12^:1217-1223, 1966.
216. Gross, P. The Processes Involved in the Biologic Aspects of Pul-
monary Deposition, Clearance and Retention of Insoluble Aerosols.
Health Physics. 10_:995-1002, 1964.
217. Mercer, T. T. On the Role of Particle Size in the Dissolution of
Lung Burdens. Health Physics. 13:1211-1221, 1967.
377
-------
REFERENCES (continued)
218. Kilburn, K. H. Clearance Zones in the Distal Lung. Ann. N.Y.
Acad. Sci. 22^:276-281, February 28, 1974.
219. Kanapilly, G. M. and C. H. Goh. Some Factors Affecting the in
vitro Rates of Dissolution of Respirable Particles of Relatively
Low Solubility. Health Physics. 2_5:225-237, September, 1973.
220. Bowden, D. H. The Alveolar Macrophage. Current Topics in
Pathology. 5!5_:l-35, Springer-Berlag Pub., New York, 1971.
221. Van Furth, R. (Ed.) Mononuclear Phagocytes. Blackwell Sci.
Pub., Oxford, 1971.
222. Bingham, E., W. Barkley, M. Zerwas, K. Stemmer and D. Taylor.
Responses of Alveolar Macrophages to Metals. Arch. Environ.
Health. 25^206-414, December, 1972.
223. Burrell, R. and M. Anderson. The Induction of Fibrogenesis by
Silica-treated Alveolar Macrophages. Environmental Research.
6_: 389-394, 1973.
224. Waters, M. D., D. E. Gardner and D. L. Coffin. Cytotoxic Effects
of Vanadium on Rabbit Alveolar Macrophages in vitro. Toxicol.
Appl. Phamacol. 2^:253-263, 1974.
225. Brain, J. D. and P. A. Valberg. Models of Lung Retention Based
on ICRP Task Group Report. Arch. Environ. Health. 28_:1-11,
January, 1974.
226. Lee, R. E. Jr., and D. J. Von Lehmden. Trace Metal Pollution in
the Environment. APCA Journal. 23_: 853-857, October, 1973.
227. Lee, R. E. Jr., R. E. Goranson and J. B. Moran. National Air
Surveillance Cascade Impactor Network II Size Distribution
Measurements of Trace Metal Components.
228. Kertesz-Saringer, M. E. Meszaros and T. Varkonyi. On the Size
Distribution of Benzo(a)Pyrene-Containing Particles in Urban Air
Atmos. Environ. ,3:429-431, 1971.
229. Rann, K. A., R. Dams, J. A. Robbins and J. W. Winchester.
Diurnal Variations of Aerosol Trace Element Concentrations as
Determined by Nondestructive Neutron Activation Analysis.
Atmos. Environ. 5:413-422, 1971.
378
-------
REFERENCES (continued)
230. Harrison, P. R., W. R. Matson and J. W. Winchester. Time Varia-
tions of Lead, Copper and Cadmium Concentrations in Aerosols in
Ann Arbor, Michigan. Atmos. Environ. 5^613-619, 1971.
231. Natusch, D. F. S. and J. R. Wallace. Toxic Trace Elements:
Preferential Concentration inRespirable Particles. Science. 183:
202-204, January, 1974.
232. Natusch, D. F. S. and J. R. Wallace. Urban Aerosol Toxicity: The
Influence of Particle Size. Science. 186_: 695-699, November, 1974.
233. Coffin, D. L. and D. E. Gardner. Interaction of Biological Agents
and Chemical Air Pollutants. Ann. Occup. Hyg. Vol. 15, pp, 219-
234, 1972.
234. Fairchild, G. A. Ozone Effect on Respiratory Deposition of Vesi-
cular Stomatitis Virus Aerosols. Am. Rev. Resp. Dis. 109:446-451,
1974.
235. Lippman, M., R. E. Albert, and H. J. Peterson, Jr. The Regional
Deposition of Inhaled Aerosols in Man. Inhaled Particles III,
Vol. 1. W. H. Walton, ed. Unwin Brothers Ltd., London, 1971.
236. Brasser, L. J., P. E. Joosting and D. Von Zuilen. Sulfur Dioxide--
To What Level Is It Acceptable. Research Institute for Public
Health Engineering, Delft, Netherlands, Report G-300, July, 1967.
(Originally published in Dutch, September, 1966.)
237. Joosting, P. E. Air Pollution Permissibility Standards Approached
from the Hygienic Viewpoint. Ingenieur. 79_(50) :A739-A747, 1967.
238. Carnow, B. W., R. M. Senior, R. Karsh, S. Wesler and L. V. Alioli.
The Role of Air Pollution in Chronic Obstructive Pulmonary Disease.
J. Amer. Med. Assoc. 214(5):894-899, November 2, 1970.
239. Douglas, J. W. B. and R. E. Waller. Air Pollution and Respiratory
Infection in Children. Brit. J. Prev. Soc. Med. 20_:l-8, 1966.
240. H. E. Goldberg, A. A. Cohen, J. F. Finklea, J. H. Farmer, F. B.
Benson and G. J. Love. Frequency and Severity of Cardiopulmonary
Symptoms in Adult Panels: 1970-1971 New York Studies. In:
Health Consequences of Sulfur Oxides: A Report from CHESS, 1970-
1971. U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina. Publication No. EPA-650/1-74-004, 1974.
241. Wicken, A. J. and S. F. Buck. Report On a Study of Environmental
Factors Associated with Lung Cancer and Bronchitis Mortality in
379
-------
REFERENCES (continued)
Areas of North East England. Tobacco Research Council, London,
Research Paper 8, 1964.
242. Buechley, R. W., W. B. Riggan, V. Hasselblad and J. B. Van Bruggen.
SO- Levels and Perturbations in Mortality. A Study in the New
York-New Jersey Metropolis. Arch. Env. Health. 2^:134-137, 1973.
243. Lunn, J. E., J. Knowelden and A. J. Handyside. Patterns of
Respiratory Illness in Sheffield Infant School Children. Brit. J.
Prev. Soc. Med. 21.: 7-16, 1967.
244. Finklea, J. F., D. C. Calafiore, C. J. Nelson, W. B. Riggan and
C. G. Hayes. Aggravation of Asthma by Air Pollutants: 1971
Salt Lake Basin Studies. In: Health Consequences of Sulfur
Oxides: A Report from CHESS, 1970-1971. U.S. Environmental Pro-
tection Agency. Research Triangle Park, North Carolina, Publica-
tion No. EPA-650/1-74-004, 1974.
245. Finklea, J. F., J. H. Farmer, G. J. Love, D. C. Calafiore and
G. W. Sovocool. Aggravation of Asthma by Air Pollutants: 1970-
1971 New York Studies. In: Health Consequences of Sulfur Oxides:
A Report from CHESS, 1970-1971. U.S. Environmental Protection
Agency. Research Triangle Park, North Carolina, Publication
No. EPA-650/1-74-004, 1974.
246. Shy, C. M., V. Hasselblad, J. F. Finklea, R. M. Burton, M. Pravda,
R. S. Chapman and A. A. Cohen. Ventilatory Function in School
Children: 1970-1971 New York Studies. In: Health Consequences
of Sulfur Oxides: A Report from CHESS, 1970-1971. U.S. Environ-
mental Protection Agency. Research Triangle Park, North Carolina,
Publication No. EPA-650/1-74-004, 1974.
247. shy, C. M., C. J. Nelson, F. B. Benson, W. B. Riggan, V. A. Newill
and R. S. Chapman. Ventilatory Functions in School Children:
1967-1968 Testing in Cincinnati Neighborhoods. In: Health Conse-
quences of Sulfur Oxides: A Report From CHESS, 1970-1971. U.S.
Environmental Protection Agency. Research Triangle Park, North
Carolina, Publication No. EPA-650/1-74-004, 1974.
248. Nelson, W. C., J. F. Finklea, D. E. House, D. C. Calafiore, M. B.
Hertz and D. H. Swanson. Frequency of Acute Lower Respiratory
Disease in Children: Retrospective Survey of Salt Lake Basin
Communities, 1967-1970. In: Health Consequences of Sulfur
Oxides: A Report from CHESS, 1970-1971. U.S. Environmental Pro-
tection Agency, Research Triangle Park, North Carolina, Publica-
tion No. EPA-650/1-74-004, 1974.
380
-------
REFERENCES (continued)
249. Finklea, J. F., J. G. French, G. R. Lowrimore, J. Goldberg, C. M.
Shy and W. C. Nelson. Prospective Surveys of Acute Respiratory
Disease in Volunteer Families, 1969-1970 Chicago Nursery School
Study. In: Health Consequences of Sulfur Oxides: A Report from
CHESS, 1970-1971. U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, Publication No. EPA-650/1-74-004,
1974.
250. Love, G. J., A. A. Cohen, J. F. Finklea, J. G. French, G. R.
Lowrimore, W. C. Nelson and P. B. Ramsey. Prospective Surveys of
Acute Respiratory Disease in Volunteer Families, 1970-1971 New
York Studies. In: Health Consequences of Sulfur Oxides: A
Report from CHESS, 1970-1971. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, Publication No. EPA-650/
1-74-004, 1974.
' \
251. House, D. E., J. F. Finklea, C. M. Shy, D. C. Calafiore, W. B.
Riggan, J. W. Southwick and L. J. Olsen. Prevalence of Chronic
Respiratory Disease Symptoms in Adults: 1970 Survey of Salt Lake
Basin Communities. In: Health Consequences of Sulfur Oxides: A
Report from CHESS, 1970-1971. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, Publication No. EPA-
650/1-74-004, 1974.
252. Finklea, J. F., J. Goldberg, V. Hasselblad, C. M. Shy, C. G. Hayes.
Prevalence of Chronic Respiratory Disease Symptoms in Military
Recruits, 1969-1970. In: Health Consequences of Sulfur Oxides:
A Report from CHESS, 1970-1971. U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, Publication No. EPA-
650/1/74-004, 1974.
253. Lewis, T. R., M. 0. Amdur, M. D. Fritzhand and K. I. Campbell.
Toxicology of Atmospheric Sulfur Dioxide Decay Products. Environ-
mental Protection Agency Publication No. AP-111, National Environ-
mental Research Center, Research Triangle Park, North Carolina,
July, 1972.
254. Amdur, M. 0. A Review of the Toxicology of the Oxidation Products
of Sulfur Dioxide, Arch.. Environ. Health. 2^:459-468, 1971.
255. McJilton, C., N. R. Frank and R. Charlson. Role of Relative
Humidity in the Synergistic Effect of a Sulfur Dioxide-Aerosol
Mixture on the Lung. Science. 18^:503-504, 1973.
256. Fairchild, G. A. and S. Stultz. Sulfuric Acid Effect on Deposi-
tion of Radioactive Aerosol in the Respiratory Tract of Guinea
Pigs. Submitted to Arch. Environ. Health. November 15, 1974.
381
-------
REFERENCES (continued)
257. Fairchild, G. A., P. Kane, B. Adams and D. L. Coffin. Effect of
Sulfuric Acid Inhalation on Clearance of Radioactive Aerosol from
the Respiratory Tract of Mice. In manuscript undergoing NERC
clearance prior to submission for publication, October, 1974.
258. Busey, W. M. Chronic Exposure of Cynomolgus Monkeys to Sulfuric
Acid Mist and Fly Ash Mixtures, Project RP-74, Hazleton Labora-
tories, Inc. report to Edison Electric Institute, December, 1972.
259. Alarie, Y., C. E. Ulrich, W. M. Busey, A. A. Krumm and H. N. Mac
Farland. Effects of Long-Term Continuous Exposure to Sulfur
Dioxide in Cynomolgus Monkeys, Arch. Environmental Health. 24:
115-128, 1972.
260. Alarie, Y., C. E. Ulrich, W. M. Busey, H. E. Swann and H. N. Mac
Farland. Effects of Long-Term Continuous Exposure of Guinea Pigs
to Sulfur Dioxide. Arch. Environmental Health. 2.1:769-777, 1970.
261. Alarie, Y., W. M. Busey, A. A. Krumm and C. E. Ulrich. Long-Term
Continuous Exposure to Sulfuric Acid Mist in Cynomolgus Monkeys
and Guinea Pigs. Arch. Environmental Health. 27^:16-24, 1973.
262. Alarie, Y., R. J. Kantz II, Charles E. Ulrich, A. A. Krumm and
W. M. Busey. Long-Term Continuous Exposure of Sulfur Dioxide and
Fly Ash Mixtures in Cynomolgus Monkeys and Guinea Pigs. Arch.
Environmental Health. 27^251-253, 1973.
263. Zabower, A. Alterations in Antibody Response Induced by Chronic
Inhalation of S09 and Carbon. Arch. Environmental Health. 25:
45-50, 1972. L
264. Ikeda, Y. et al. Long-Term Toxicity Study of Methylmurcuric
Chloride in Monkeys. Toxicology lj61-375, 1973.
265. Takeuchi, T. Environmental Mercury Contamination. Eds. R.
Harting, B. D. Dinman. Ann. Arbor. Sci. Pub. 1972, p. 247.
266. Khera, K. A. Teratogenic Effects of Methylmercury in the Cat.
Teratology 8;293-304, 1973.
267. Magos, L. et al. Cumulative Effects of Methylmercury Dicyandia-
mide Given Orally to Rats. Fd. Cosmet. Toxicol. 1£: 513-17, 1972.
268. Irukayama, K. Cited by G. Lofroth in Methylmercury Bull. No. 4,
Ecological Res. Comm. Swedish Natural Sciences Res. Council, 1970.
269. Post. E. M. et al. Behavioral Changes in Young Rats Force-Fed
Methylmercury Chloride. Proc. Soc. Exp. Biol. Med. 143:1115-16,
1973.
382
-------
REFERENCES (continued)
270. Fukuda, K. Metallic Mercury-Induced Tremor in Rabbits and Mercury
Content of the Central Nervous System. Brit. J. Ind. Med. 28:
308-311, 1971. ~~
271. Beliles, R. P. et al. Tox. Appl. Pharmacol. .12:15-21, 1968.
272. Charbonneau, S. M. Subacute Toxicity of Methylmercury in the
Adult Cat. Tox. Appl. Pharmacol. 27_: 569-581, 1974.
273. Gale, T. F. et al. Embryopathic Effects of Mercury Salts. Life
Sci. 10:1341-1347, 1971.
274. Gale, T. F. The Interaction of Mercury with Cadmium and Zinc in
Mammalian Embryonic Development. Environ. Res. 6_:95-105, 1973.
275. Spyker, J. M. et al. Effects of Methylmercury on Prenatal and
Postnatal Development in Mice. Teratology £:242, 1971.
276. Spyker, J. M. et al. Behavioral Teratology of Methylmercury in
the Mouse. Pharmacol. 1.3:275, 1971.
277. Klein, S. B. Mercury Chloride Influence on Active-Avoidance
Acquisition in Rats. Bull. Psychon. Soc. 1_:437-438, 1973.
278. Kawabe, H. et al. The Effect of Prenatal Methylmercury Poisoning
on the Mouse Fetus. Teratology 6:11, 1972.
279. Munro, J. C. et al. Toxic Effects of Methylmercury in the Rat.
Tox. Appl. Pharmacol. 2_2:29, 1972, Abs. No. 53.
280. Herigstad, R. R. et al. Chronic Mercury Toxicosis in Calves.
J.A.V.M.A. 16(3:173-182, 1972.
281. Iverson, F. Accumulation and Tissue Distribution of Mercury in
the Guinea Pig During Subacute Administration of Methylmercury.
Tox. Appl. Pharmacol. 27_:60-69, 1974.
282. Tatetsu, M. et al. Psychiat. Neurol. Jap. 7£:162, 1968.
283. Suzaki, T. et al. Neurological Symptoms and Mercury Concentration
in the Brain of Mice Fed with Methylmercury Salt. Ind. Hlth. 9j
51-58, 1971.
284. Iverson, F. Methylmercury: Acute Toxicity, Tissue Distribution
and Decay Profiles in the Guinea Pig. Tox. Appl. Pharmacol. 24:
545-554, 1973.
383
-------
REFERENCES (continued)
285. Newberne, P. M. et al. Chronic Exposure of Rats to Methyl Mercury
in Fish Protein. Nature 237;40-41, 1972.
286. Singhal, R. L. et al. Metabolic Alterations in Liver and Kidney
Following Chronic Msthyl Mercury Treatment and Withdrawal. Environ.
Res. 7; 220-229, 1974.
287. Lucier, G. W. et al. Methyl Mercury-Induced Changes in Rat Liver
Microsomes. Environ, Hlth. Perspect. 4:102, 1973.
288. Klein, R. et al. Methyl Mercury Intoxication: Early Functional
and Morphological Changes in the Rat Kidney. Environ. Hlth.
Perspect. 1_:100, 1973.
289. Fowler, B. A. et al. Phenobarbital Protection Against Methyl
Mercury Nephrotoxicity. Environ. Hlth. Perspect. 4_:100, 1973.
290. Hastings, F. L. Methyl Mercury-Cholinesterase Interactions in
Rats. Environ. Hlth. Perspect. 4_:101, 1973.
291. Khera, K. S. et al. Effects of Methylmercuric Chloride on the
Progency of Mice and Rats Treated Before or During Gestation. Fd.
Cosmet. Toxicol. 11_:245-254, 1973.
292. Gate, C. C. and R. Burrell. Lung Antigen Induced Cell-Mediated
Immune Injury in Chronic Respiratory Diseases. Amer. Rev. Resp.
Dis. 109_: 114-23, 1974.
293. Clary, J. J., C. R. Hopper and H. E. Stokinger. Altered Adrenal
Function as an Inducer of Latent Chronic Beryllium Disease.
Tox. Appl. and Pharmacol. 23_:365-375, 1974.
294. Clary, J. J. and D. H. Groth. Comparative Changes in Serum Enzyme
Levels in Beryllium- or Carbon Tetrachloride-Induced Liver
Necrosis. Proc. Soc. Exp. Biol. Med. 143:1207-1210.
295. Clary, J. J. and H. E. Stokinger. The Mechanism of Delayed
Biologic Response Following Beryllium Exposure. J. of Occ.
Med. 15^:255-259.
296. Furchner, J. E., C. R. Richmond and J. E. London. Comparative
Metabolism of Radionuclides in Mammals. VIII. Retention of
Beryllium in the Mouse, Rat, Monkey and Dog. Health Physics
24_: 293-300, 1973.
297. Goldblatt, P. J., M. W. Lieberman and H. Witschi. Beryllium-
Induced Ultrastructural Changes in Intact and Regenerating Liver.
Arch. Environ. Health 26^:48-56.
384
-------
REFERENCES (continued)
298. Jacques, A. and H. R. Witschi. Beryllium Effects on Aryl Hydro-
carbon Hydroxylase in JRat Lung. Arch Environ. Health 27:243-
247, 1973. ~~
299. Krivanek, N. and A. L. Reeves. The Effect of Chemical Forms of
Beryllium on the Production of the Immunologic Response. J. Amer.
Indust. Hyg. Assoc. 33:45-52, 1972.
300. Marx, James J. and R. Burrell. Delayed Hypersensitivity to
Beryllium Compounds. J. Immunol. 11^:590-598, 1973.
301. Schneider, M., M. I. Resnick and K. F. Wellman. Increased Bone
Formation in Rabbits Following Intravenous and Intra-Osseous
Injection of Zinc Beryllium Silicate. Clin. Orthop. Relat. Res.
92^251-259, 1973.
302. sionim, A. R. and F. C.. Dann. Beryllium Uptake by the Common
Guppy Using Radioactive Beryllium-7. II. Beryllium Concentration
in Fish. U.S. Nat. Teca. Inform. Service, Iss. No. 758815, p. 29,
1972.
303. sionim, A. R. Acute Toxicity of Beryllium Sulfate to the Common
Guppy. J. Water Pollut. Contr. Fed. 4_5:2110-2122, 1973.
304. Vacher, J. Immunological Responses of Guinea Pigs to Beryllium
Salts. J. Med. Microbiol. 5>:91-108, 1972.
305. Vacher, J., R. Deraedt and J. Benzoni. Compared Effects of Two
Beryllium Salts (Soluble and Insoluble): Toxicity and Blockade
of the Reticuloendothelial System. Toxicology and Applied
Pharmacology 24_:497-506, 1973.
306. Vacher, J., R. Deraedt and J. Benzoni. Influence of a Change in
the Activity of the Reticuloendothelial System on the Development
of Adjuvant-Induced Polyarthritis in the Rat. J. Reticuloendothel.
Soc. 13:579-588, 1973.
307. Vasil, E. Changes in the Antigenic Composition in the Lungs in
Experimental Berylliosis. Byull. Eksp. Biol. Med. 73:76-80, 1972.
308. Criteria for a recommended standard. Occupational Exposure to
Asbestos, NIOSH, HEW, 1972.
309. EPA National Emission Standards for Hazardous Air Pollutants,
Asbestos, Beryllium and Mercury. Federal Register, Vol. 38,
No. 66, April, 1973.
385
-------
REFERENCES (continued)
310. Proceedings of the Second International Conference on Biological
Effects of Asbestos, Dresden, 1968 Proceedings.
311. International Conference on Pneumonoconiosis, Johannesburg, 1969,
Oxford University Press, 1970.
312. Proceedings of the Working Group Meeting on Biological Effects of
Asbestos. Lyon, 1972. Preprint 30 Int. Agen. Res. Cancer, Lyon,
France.
313. Williams, S. Asbestos. A Bibliography of the World's Literature
Abstracted and Indexed (a thesis). Pneumonconiosis Research Unit
CSIR, 1969. P.O. Box 1038, Johannesburg.
314. Wagner, J. C. Brit. J. Ind. Med. 20, 1963.
315. Selikoff, I. J., E. C. Hammond and M. Seedman. Working Group
Meeting on Biological Effects of Asbestos. Preprint 33, Int.
Agency Res. Cancer, Lyon, France, 1972.
316, Meurman, L. 0., R. Kiviluoto and M. Hakama. Int. Conf. Bio.
Effects Asbestos. Preprint 31.
317. Cilson, J. C. Asbestos Cancer Report of the Advisory Committee
to Director Int. Agency Res. Cancer, Lyon, France, 1972.
318. Bragg, L. W. and G. F. Claringbull. Crystal Structure of Miner-
als. Bell, London.
319. Kiviluoto, R. Pleural Calcification as Roentgenologic Sign of
Nonoccupational Endemic Anthophyllite - Asbestos. Radial Supp.
194, pp. 1-67. 1960.
320. Wagner, J. C. and G. Berry. Brit. Jour. Cancer 23^:567, 1969.
321. Harrington, J. S. In Working Group on Biological Effects of
Asbestos. Preprint 45, Int. Agency Res. Cancer, Lyon, France,
1972.
322. Stanton, M. F., R. Blackwell and E. Miller. Amer. Ind. Hyg.
Assoc. J. 30_:236, 1969.
323. Stanton, M. F. and C. Wrench. J. Ntl. Cancer Inst. 48_:797,
1972.
324. Stanton, M. F. In Working Group Meeting on Biological Effects
of Asbestos. Preprint 43A. Int. Agency Res. Cancer, Lyon, 1972.
386
-------
REFERENCES (continued)
325. Davis, J. M. G. Histogenesis and Fine Structure of Peritoneal
Tumors Produced in Animals by Injection of Asbestos. J. Ntl.
Cancer Int. 52^1823, 1974.
326. Wagner, J. C., G. Berry, J. W. Skidmore and V. Timbrell. The
Effect of the Inhalation of Asbestos in Rats. Brit, of Cancer
2£:252, 1974.
327. Selikoff, I. J., W. J. Nicholson and A. M. Langer. Arch. Env.
Hlth. 2,5:1, 1972.
328. Vorworld, A. J., T. M. Durkan and P. C. Pratt. Arch. Ind. Hyg.
Occup. Med. _3:1, 1951.
329. Smither, W. J. and H. C. Ldwinsohn. In Working Group Meeting on
Biological Effects of Asbestos. Preprint 26, Int. Agency Cancer,
Lyon, France, 1972.
330. Naviatil, M. and F. Trippe. Prevalence of Pleural Calcification
in Persons Exposed to Asbestos Dust and in the General Population
in the Same District. Env. Res. 5_:210, 1972.
331. Miller, K. and J. S. Harrington. Brit. J. Exp. Pathol. 55:397,
1972.
332. Davis, J. M. G. Brit. J. Exp. Pathol. 5_3:190, 1972.
333. Harrington, J. S., A. C. Allison and D. V. Badomi. Mineral Fibers
Chemical, Physiochemical and Biological Properties Advances in
Pharmacology. (In Press.)
334. Gross, P., R. T. P. de Treville, E. B. Talker, M. Kaschak and M. A.
Babyah. Arch. Env. Health. 15:343, 1967.
335. Gross, P, R. T. P. de Treville and L. J. Cralley. In Pneumono-
coniosis (H.A. Shapiro ed.). Oxford University Press, London and
New York, 1970.
336. Wagner, J. C. Nature (London). 196:180, 1962.
337. Shabad, L. M., L. N. Phlev, L. V. Krivosheeva, T. F. Kulagina and
B. A. Nemenko. Expermental Studies on Asbestos Carcinogenecity.
J. Natl. Cancer Inst. 52;1175, 1974.
338. Pylev, L. M., F. J. Roe and G. P. Warwick. Elimination of Radio-
activity after Intratracheal Instillation of 3.4-BP in Hamsters.
Br. J. Cancer 23:103, 1969.
387
-------
REFERENCES (continued)
339. Wagner, J. C. Recent Results. Cancer Research 39_:37, 1972.
340. Beloin, N. J. Fading of Dyed Fabrics by Air Pollution: A Field
Study. Text. Chern. Color. 4^43-48, March, 1972.
341. Beloin, N. J. Fading of Dyed Fabrics by Air Pollution: A Cham-
ber Study. Text. Chem. Color. 5^:128-133, July, 1973.
342. Haynie, F. H. and J. B. Upham. Effects of Atmospheric Sulfur
Dioxide on the Corrosion of Zinc. Materials Protection and
Performance. 9;35-40, August, 1970.
343. Haynie, F. H. and J. B. Upham. Effects of Atmospheric Pollutants
on Corrosion Behavior of Steels. Materials Protection and Per-
formance. 10_: 18-21, November, 1971.
344. Haynie, F. H. and J. B. Upham. Correlation Between Corrosion
Behavior of Steel and Atmospheric Pollution Data. In: Corrosion
in Natural Environments, ASTMSTP 558, American Society for Testing
and Materials. Philadelphia, 1974. pp. 33-51.
345. Campbell, G. G., G. G. Schurr, D. E. Slawikowski and J. W. Spence.
Assessing Air Pollution Damage to Coatings. J. Paint Technology.
46_(593): 59-71, June, 1974.
346. In Preparation, to be published.
347. Beloin, N. J. Fading of Dyed Fabrics by Air Pollution: A Field
Study. Text. Chem. Color. 4_: 43-48, March, 1972.
348. Beloin, N. J. Fading of Dyed Fabrics by Air Pollution: A Chamber
Study. Text. Chem. Color. 5_: 128-133, July, 1973.
349. In Preparation, to be published.
350. Charlson, R. J., A. H. VanderPol, D. S. Covert, A. P. Waggoner,
N. C. Ahlquist, Sulfuric Acid-Ammonium Sulfate Aerosol: Opti-
cal Detection in the St. Louis Region. Science. 184:156, 1974.
351. Samuels, H. J., S. Twiss, E. W. Wong. Visibility, Light Scatter-
ing and Mass Concentration of Particulate Matter. Report of the
California Tri-City Aerosol Sampling Project, State of California
Air Resources Board, 1973.
352. Covert, D. S. A Study of the Relationship of Chemical Composition
and Humidity to Light Scattering by Aerosols. Phd. Thesis, Uni-
versity of Washington, Seattle, Washington.
388
-------
REFERENCES (continued)
353. Woodwell, G. M. Effects of Pollution on the Structure and Physi-
ology of Ecosystems. Science 168_:429-433, 1970.
354. Likens, G. E. and F. H. Borman. Linkages between Terrestrial and
Aquatic Ecosystems. Bio Science. 24_:447-456, 1974.
355. Smith, R. L. Ecology and Field Biology. 2nd. ed., Hoper § Row,
New York, New York. 1974.
356. McCune, D. C. and W. W. Heck. Health Effects of Air Pollutants.
In: Air Quality and Automobile Emission Control - Vol. 2. Com-
mittee of Public Works, United States Senate. Pp. 503-511. 1974.
357. Inman, R. E., R. B. Ingersol and E. A. Levy. Soil: A Natural Sink
for Carbon Monoxide. Science. 172:1229-1231, June 18, 1971.
358. Air Quality for Carbon Monoxide, N. 10. A Report by the Expert
Panel on Air Quality Criteria, NATO Committee on the Challenges
of Modern Society. Research Triangle Park, North Carolina, pp. 1-9
to 1-11. 1972.
359. Air Quality Criteria for Nitrogen Oxides, N. 15. A Report by the
Expert Panel on Air Quality Criteria. NATO Committee on the Chal-
lenges of Modern Society. Research Triangle Park, North Carolina,
Chapter 7. 1973.
360. Miller, P. L. Oxidant-Induced Community Change in a Mixed Conifer
Forest. Adv. in Chem. Series No. 122. Air Pollution Damage to
Vegetation, pp. 101-117. 1973.
361. Air Quality Criteria for Photochemical Oxidants and Related Hydro-
carbons. N. 29. NATO Committee on the Challenges of Modern
Society. Research Triangle Park, North Carolina. Chapter 4. 1974.
362. Heck, W. W. and E. G. Pires. Growth of Plants Fumigated with
Saturated and Unsaturated Hydrocarbon Gases and Their Derivative.
The Agricultural and Mechanical College of Texas, Texas Agr. Expt.
Station. MP603. p. 12. 1962.
363. Andelman. J. B. and M. J. Suess. Polvnuclear Aromatic Hydrocarbons
in Water Environment. Bull. World Health Organization. 4J5:479-
508, 1970.
364. Garner, J. H. B. The Death of Woody Ornamentals Associated with
Leaking Natural Gas. Proc. j9 Intl. Shade Tree Conf., Boston,
Massachusetts, p. 13-17. 1973.
389
-------
REFERENCES (continued)
365. Westman, Lars. Air Pollution Indications and Growth of Spruce
and Pine Near a Sulfite Plant. Arnbio. 3_:189-193. 1974.
366. ERC/RTP Task Force Report on Suspended Sulfates and Sulfuric
Acid Aerosols. Research Triangle Park, North Carolina. March,
1974.
367. Likens, G. E. and F. H. Borman. Acid: A Serious Regional
Environmental Problem. Science. 1M: 1176-1179, 1974.
368. Robinson, E. and R. C. Robbins. Gaseous Nitrogen Compound Pol-
lutants from Urban and Natural Sources. J.A.P.C.A. 20:303-
306, 1970.
369. Air Quality Criteria for Photochemical Oxidants. AP-63.
Washington, B.C. 1970.
370. Romney, E. M., J. D. Childress and G. V. Alexander. Beryllium
and the Growth of Bush Beans. Science. 135:786-787, 1962.
390
-------
6. CONTROL TECHNOLOGY
6.1 STATIONARY SOURCES VS. POLLUTANTS
Stationary combustion sources contribute more to ambient atmospheric
loading of SO , NO and total suspended particulates (TSP) than any other
J\. X
anthropogenic source category. These sources are responsible for (1)
approximately 80% of the SO emissions, 50% of the NO emissions and
.X «X.
(2) over 501 of ambient fine particulates.
Industrial processes and open sources (e.g., open burning, mining,
incineration, fugitive emissions) are responsible for (1) approximately
161 of the SO emissions, 441 of the hydrocarbon emissions, 22% of the
A.
CO emissions and (2) over 25?. of ambient fine particulates. They are
also responsible for most of the hazardous air pollutant emissions (i.e.,
asbestos, beryllium and mercury), vinyl chloride and other potentially
hazardous carcinogenic pollutants.
6.1.1 Particulate Control
The most pressing need in particulate control technology is for
development of commercially available and widely applicable fine particle
(less than 3 microns) control technology to support the establishment of
Fine Particulate standards to provide control of secondary particulate
precursors, catalysts and hazardous trace elements. Such technology is
inadequate and only just undergoing development under EPA sponsorship.
Scrubbers, electrostatic precipitators and fabric-filter particle
collectors are principal particulate collection devices. They are being
upgraded by projects concerned with decreasing costs, increasing appli-
cability of devices to a wider range of sources and extending their range
391
-------
of particle collection capability to smaller particle sizes. In addi-
tion, entirely new concepts of particle collection are being investi-
gated.
System studies have been completed for each of the conventional
devices. These define potential areas of improvement and new applica-
tions. Long-range projects include development of generalized models
for existing collectors, indentification and laboratory scale testing
of totally new concepts in particle control technology and pioneering
work in small-particle detection and small-particle generation equip-
ment.
6.1.2 Sulfur Oxides Control
For Sulfur oxides control, achievement of Ambient Air Quality
Standards require control of industrial combustion, industrial pro-
cesses and area sources, including utility combustion. There is only
enough natural low-sulfur fuels to satisfy about 40% of the demand of
State Implementation Plans. The remainder must be supplied by con-
trol technology in the form of effluent desulfurization, fuel cleaning
or process modification.
Although the currently defined sulfur oxides control technology
program contains developmental projects of sufficient diversity to
ensure achievement of Ambient Air Quality Standards, a substantial EPA
research and development investment will be required to bring this
technology to commercial fruition. Only in the case of effluent de-
sulfurization technology for electric utility applications should con-
trol technology achieve widespread commercialization the next few years.
The program has emphasized stack-gas cleaning systems with five or
392
-------
six such processes to be demonstrated before 1976. It has also included
work in such areas as fuel modification and desulfurization, industrial
process modification and advanced combustion processes.
Most of EPA's early coaL de-sulfurization work involved improving
and evaluating physical cleaning methods for coal. With the over all
potential of physical de-sulfurization and de-ashing of coal reasonably
well established, this approach should have moderate applicability in
reducing sulfur by 30 to 70% in specific coals. On-going work in this
area includes washability testing, coal de-sulfurization and cleaning
and pollutant control evaluation and optimization. The most recent work
is aimed at chemical de-sulfurization of coal and problems of hazardous
pollutants in fossil fuel.
Laboratory-scale technology developments are under way on processes
to remove sulfur from coal, producing a clean fuel for combustion. The
two processes under development are unique since one is completely
selective to inorganic sulfur in coal, the other is being developed for
preferential organic sulfur removal and both produce a solid fuel that
retains the basic physical characteristics of coal.
The technical feasibility of the inorganic sulfur-removal system
has been proven in the laboratory and on the bench scale. It is under-
going scale-up design to a pilot plant size operation. Construction of
this plant is scheduled to begin in early 1974. The degree of sulfur
removal that can be obtained from the organic sulfur-removal process is
currently being quantified in a laboratory apparatus.
This system could be well enough developed by mid-1974 to initiate
scale-up design to continuous pilot operation by September, 1974.
393
-------
The two systems will be examined in detail for their capatibility for
integration into one process scheme for complete removal of sulfur from
coal. After successful pilot-plant operation, plans call for scale-up
to a demonstration size plant.
Two factors--the major contribution of small coal and oil-burning
sources (area sources) to degradation of ambient air quality and the
severely limited supplies of low-sulfur fuels--indicate a need for clean
fuel or coal conversion processes. CSL has provided funds to both the
Bureau of Mines and the Office of Coal Research to assist in development
of conversion and gasification processes that can provide clean fuels
and to determine what environmental controls will be required for these
techniques.
The work on advanced processes primarily involves development of
fluidized-bed combustion. Fluidi^ed-bed combustion process options
being investigated include: pressurized coal combustion, coal gasifi-
cation for production of low-sulfur and low-Btu fuel gas and oil gasi-
fication. System studies also have been done on development of advanced
power cycles.
Work is being conducted to control emissions from smelters, iron
and steel operations, refineries, iron foundries, secondary aluminum
furnaces and other industrial sources.
6.1.3 Nitrogen Oxides Control
On a national basis, 98% of the nitrogen oxides (NO ) from sta-
X
tionary sources are from combustion processes. For this reason, the
NO control program is oriented heavily toward combustion control. Con-
J\,
trol technologies under investigation to provide near-term solutions to
394
-------
the problem include: combustion modifications such as flue gas recircu-
lation, off-stoichiometric combustion and low excess air operation;
identification of alternate or novel combustion systems with inherently
low NO emission levels; changes in burner design and the use of alter-
A.
nate fuels.
Although flue-gas cleaning processes and advanced power cycles are
also being investigated, they may offer longer-term solutions to the
problem. Major emphasis in this program has been on coal combustion in
utility and industrial boilers since that poses the most difficult
problem.
The main approach has been modification of the combustion process
to reduce NO formation. Bench-scale work and field testing indicate
that control for the immediate future can be achieved by this method.
In addition, promising stack-gas cleaning processes are being evaluated
and investigated on a small scale.- Results of this research, however,
have not been encouraging so far.
A complicating factor in developing effective control technology
is that NO are formed during combustion through two separate mechanisms -
the fixation of nitrogen in combustion air and the conversion of fuel-
bound nitrogen contained in fuels. Although some control of both types
of NO has been achieved by combustion modifications, it is not known
if combustion modification alone will adequately control the NO from
fuel nitrogen.
6.1.4 Hazardous and Other Control
As EPA regulatory programs develop, the number of pollutants requir-
ing control has expanded to include more than 30 potentially hazardous
395
-------
and annoying air pollutants. Since a majority of the hazardous pollu-
tants are likely to be emitted as particulate matter, there is strong
interaction between these two programs.
Examples of specific hazardous and other pollutants under con-
sideration include asbestos, beryllium, cadmium, carbon monoxide,
fluorides, hydrocarbons, lead and mercury and odorous materials from
rendering plants, pulp mills, sewage treatment plants and chemical pro-
cess industries. Because of the paucity of data currently available,
many problem definition and source characterization studies are required
to deal with these pollutants on a rational basis.
6.2 STATIONARY SOURCES VS. CONTROL
Significant advances in development and demonstration of technology
to control SO , NO and primary particulates have already been achieved.
X X.
Additional improvements in the state of the art are projected for coming
years. A brief summary of proven and promising technology under develop-
ment to control various combustion emissions are described below.
6.2.1 Combustion
6.2.1.1 Fuel Gas De-Sulfurization--Consistent with Congressional guide-
lines, major concern has focused on control of sulfur oxides from fuel
combustion. Expenditures for this have been about 80% of the NERC/RTP
Control Systems Laboratory's expenditures to date. And flue-gas de-
sulfurization (FGD) technology development has comprised a major portion
of this work.
Flue-gas de-sulfurization systems have been installed on several
major commercial power plants in the U.S.A. At least 80 more such
systems will be installed by 1980. With the exception of use of clean
396
-------
fuels, flue-gas de-sulfuriza.tion is the only sulfur oxides control
technology that will have widespread application in the next 15 years.
With the expected shortage of clean fuels (amounting to about 20% of
the coal demand), flue-gas de-sulfurization will be needed to off-set
that shortage.
The need to develop techniques that permit use of eastern and
midwestern coal is urgent since only about 7% of this coal is now
usable under New Source Performance Standards. And a domestic resource
about equivalent in value to the GNP will not be available for use
without FGD technology.
Flue-gas de-sulfurization is a technically sound solution that is
also economically acceptable. The cost of flue-gas de-sulfurization
for coal-fired power plants should be 2 to 3 mills/Kwbr^ ' compared to
2 to 4 mills/Kwhr^- ' for naturally-occurring clean fuels delivered to
the power plant site and even higher costs of approximately 8-13 mills/
Kwhr^ ' associated with coal gasification technologies.
6.2.1.1.1 Sodium hydroxide scrubbing with electrolytic regeneration
(Stone § Webster/Ionics)--In July, 1972, EPA and Wisconsin Electric
Power Company (WEPCO) initialed a 3-1/2 year, three-phase program involv-
ing the Stone $ Webster (SWI) sodium hydroxide scrubbing process.
In Phase I, recently completed, an integrated pilot plant was
built. Operating tests were performed for about a one-year period and
a prototype-scale electrolytic cell system was designed, fabricated and
tested. Preliminary design of a 75 MW prototype system and development
of detailed test programs and operating schedules for the prototype
system were also completed. The Phase I final report was scheduled for
publication in late 1974.
397
-------
Based on favorable assessment of Phase I results and continued
technical and economic viability of the process, a 16-month-long Phase
II effort is being considered for the detailed design, procurement and
installation of the 75-MW prototype. This would be followed by Phase
III, a 12-month start-up and operational period for the 75-MV prototype.
Assuming a decision to proceed, EPA and WEPCO would co-fund the $7 mil-
lion program.
Chief advantages of the process, expected to apply to both exist-
ing and new power plants over a broad range of sizes, are: highly
efficient removal of SO ; production of easily handled non-slurry flow
Jv.
streams; no solid waste; and recovery of SCL for subsequent processing
into liquified SO-, sulfuric acid or elemental sulfur.
Potential disadvantages of the process include: power require-
ments for electrolytic regeneration, adverse influence of particulate
and flue-gas trace constituents on reliability of the electrolytic cell
and need to remove from the system any sulfates produced by oxidation
in the scrubber. The Phase I final report will evaluate and assess
these potential disadvantages.
6.2.1.1.2 Catalytic oxidation (Monsanto Cat-Ox)--The catalytic oxida-
tion (Cat-Ox) process is an adaptation of the contact sulfuric acid
process. Monsanto Enviro-Chem Systems, Inc. has developed this adap-
tation through work on a pilot-scale unit and a 15-MW prototype. EPA
and Illinois Power Company (sharing a $7 million total funding require-
ment) prepared to demonstrate the process on a 100-MV coal-fired boiler
at Illinois Power's Wood River Station by late 1974 with accompanying
commencement of the one-year test program.
398
-------
The Cat-Ox system is available in two configurations: the Reheat
system for retrofitting existing plants and the Integrated system for
incorporation into new power generating facilities.
The Reheat system operates with flue gases passing first through a
high efficiency (99.6%) electrostatic precipitator. Then, in prepara-
tion for catalytic conversion of the S0? to SO,, the temperature is
raised to about 850°F. This heat is added in three increments by: a
primary input from an external combustor, a gas heat exchanger and a
secondary input from the external combustor.
After heating, S02 contained in flue gases is converted to SO, by
the catalyst. Leaving the converter, the hot flue gases pass through
the high-temperature side of the gas heat-exchanger to the acid produc-
tion section. The sulfuric acid is formed by the standard SO,-FLO
absorbing tower contact process.
The acid product is cooled and sent to storage while the flue gases
pass through a fiber-packed mist eliminator (where residual traces of
sulfuric acid mist are removed) and on to the stack where clean gas
exits to the atmosphere. At this point, essentially all particulate
matter and 85% of the SO- have been removed from the gas stream.
Trace and hazardous element analyses account for an important
portion of the over all Cat-Ox test program. A complete characteriza-
tion of Wood River Unit No. 4 (prior to Cat-Ox equipment tie-in)
included analyses for some 30 trace elements in the coal, hopper ash
and slag. Other analyses included fly ash, where elemental analysis
has been done for a complete range of size fractions. These tests will
be repeated during the one-year test program (after the Cat-Ox system
399
-------
becomes operational) to determine effects of the system on concentration
and distribution of trace elements.
6.2.1.1.3 Sodium ion scrubbing (Wellman-Lord)--EPA has high confidence
for success of this first coal-fired boiler demonstration system in
meeting guarantees for pollution control, product quality and material
and utility requirements. This confidence is based on the already
large amounts of successful operating experience to date for W-L Systems
on various applications, including acid plants, Glaus plants and oil-
fired boilers. Seven systems are now in operation in the U.S. and
Japan.
The knowledge gained from operating these systems has produced a
series of process improvements (reducing costs and purge requirements)
that are incorporated in the EPA/NIPSCO demonstration (115 MW coal-fired
power plant for 90-95% SO- removal efficiency). CSL is considering
expansion of the Wellman Power Gas Demonstration to achieve 98-99%
removal efficiency. This could be done with little interruption of the
current project and at an incremental cost of approximately $400,000-
$2,000,000, depending on the length of the high efficiency trials.
As an on-going project with high efficiency potential, the Well-
man Power Gas System (Sodium base scrubbing) could be the first tech-
nique available to achieve eventual SCL-sulfate standards. Data from
the Wellman Demonstration could be available prior to 1978.
The S0~ product from the W-L Process is suitable for recovery in
three forms: liquid SO-, sulfuric acid and elemental sulfur. For pur-
poses of the EPA/NIPSCO demonstration, the Allied Chemical S02 Reduction
Process will be applied to generate a most salable and environmentally
400
-------
sound product elemental sulfur. The Allied Process uses natural gas as
a reductant in a novel catalytic reactor system. The process has been
demonstrated on a large scale with treatment of a 12% SCL gas stream
from a nickel ore roaster at Sudbury, Ontario.
6.2.1.1.4 Citrate process--EPA and the U.S. Bureau of Mines are coop-
erating and pool ing funds and technical talents to demonstrate the
Citrate Process developed through pilot scale by the Bureau of Mines.
A concurrent development program by an industrial consortium headed by
Pfizer Chemical Company also led to a successful pilot operation using
the process. Based on the success of these two pilot programs, EPA and
the Bureau of Mines decided to demonstrate this technology at the 30 to
60 MW scale.
Current plans call for two separate demonstration plants: one to
treat oil-fired flue gas and the other, coal-fired flue gas. Negotia-
tions are currently under way with prospective host sites and process
vendors. Contracts were to be signed in the Spring of 1975 for these
two projects that will include substantial cost sharing by the host site.
The Citrate Process consists of five steps: gas cleaning and cool-
ing, sulfur dioxide absorption, sulfur precipitation and solution regen-
eration, sulfur separation and hydrogen sulfide generation. Although a
number of reactions take place in the reactors, the net reaction is com-
bining of sulfur dioxide and hydrogen sulfide to form elemental sulfur.
Based on favorable projected economics for the Citrate Process and
encouraging results at the Pfizer and Bureau of Mines pilot plants, the
process is considered a candidate for commercial application of a second
generation flue-gas de-sulfurization process.
401
-------
6.2.1.1.5 Ammonia scrubbing with bisulfate re generation--Stack gases
have been commercially de-sulfurized with solutions of ammonium sulfite
and bisulfite since the mid 1930's. The early processes recovered SCL
in pure form by acidifying the scrubbing liquor with such acids as
sulfuric, nitric and phosphoric. The resulting ammonium salt of the
acid was further processed for use as a fertilizer.
Because of the enormous tonnages of SCL involved in de-sulfurizing
power-plant stack gases, fertilizer markets will not support wide-scale
use of fertilizer-producing ammonia processes. So CSL, in a joint
venture with TVA, is developing a completely cyclic ammonia scrubbing/
bisulfate regeneration process that has a concentrated stream of SCL
as its major product.
The process (shown below) removes SO from stack gases by absorp-
tion in a solution of ammonium sulfite and bisulfite. Scrubber product
liquor is acidified with ammonium bisulfate to produce SCL and form
ammonium sulfate. After condensing out most of the water, a pure stream
of SCL can be further processed to sulfuric acid, elemental sulfur or
liquid SCL.
Three purge methods can be used. If a fertilizer market exists
for ammonium sulfate, ammonium sulfate crystals can be purged prior to
decomposition. It there is no fertilizer market, ammonium bisulfate
can be reacted with lime to form gypsum and regenerate the ammonia.
And ammonium bisulfate can be injected into the utility boiler where it
will be decomposed into nitrogen and SCL.
In 1973, attention focused on eliminating an objectionable plume
of ammonia-based salts in scrubbed gas leaving the stack. Based on
402
-------
visual observation, this plume was eliminated or reduced to an accept-
able opacity by quenching the flue gas with water prior to introducing
the gas into the ammoniacal scrubber. Other explanations can be offered
for elimination of the plume, but they are considered less likely.
During FY 74, the pilot plant was modified. A venturi-type pre-
scrubber was added to humidify and cool the incoming gas and remove
particulates and chlorides. Two marble-bed stages were replaced with
Koch valve trays and another valve tray was added. An evaporator/
crystallizer was installed to produce solid ammonium sulfate in the
regenerator section. The pilot plant was operated for two periods
totalling about four months of operating time.
The pre-scrubber reduced particulates and chlorides by about 93%
(average) and essentially 1001 respectively. Typical sulfur dioxide
removal efficiencies in the scrubber were 90% or greater. Observed
plume opacity was about 5% or less and ammonia losses were in the 10-
30 ppm range.
CSL's efforts during FY 75 focus on improved sulfur dioxide recovery
from rich scrubber liquor, improved scrubber performance through reduc-
tion of inter-stage mixing and better control and improved mist elimina-
tion. In addition, an electrical decomposer is being designed and
fabricated. The decomposer regenerates ammonium bisulfate and ammonia
and is expected to be operational early in FY 76, allowing integrated
operation of the pilot plant.
6.2.1.1.6 Magnesium oxide (Chemico Mag-Ox) scrubbing--A two-year stack-
gas scrubbing demonstration funded by EPA/CSL using a magnesium oxide
slurry has been successfully completed on Boston Edison's oil-fired,
403
-------
155 MW steam electric plant, located at Mystic Station, Massachusetts.
The chief advantage of the process is its wide applicability to both
existing and new steam-electric plants and its highly efficient removal
of both sulfur oxides and particulates with no interference to boiler
operation. Sulfur removal efficiencies in excess of 93% have been
achieved.
During final months of the demonstration phase, the control system
availability approached 90%. This level of availability is significantly
better than in previous periods of operation. Removal efficiencies cited
above are adequate to meet existing sulfur oxide emission standards. The
ability of the process to produce high-quality sulfuric acid for marketing
was clearly established.
This control technology is of special importance because it is
environmentally clean. There is no disposal of waste products after
processing the magnesium oxide slurry for sulfur recovery and marketing
and recycling of recoverable magnesium oxide. Industrial support is
evidenced by new construction planned by other utility companies imple-
menting this process.
6.2.1.1.7 Activated carbon--The use of multi-stage, dry fluidized beds
of recycling activated carbon appears attractive both for sorption of
SO from flue gases and for converting removed SO to elemental sulfur.
J\. J\.
Under an EPA contract, development of the activated-carbon-based flue-
gas de-sulfurization process has advanced to a stage where three major
process units--sorber, sulfur generator and carbon regenerator--are
integrated for continuous and cyclic operation.
Integrated pilot plant operation, a culminating point in efforts
to determine over all technical feasibility of the process, is complete.
404
-------
Cyclic operation of the approximately 300 scfm capacity pilot plant has
produced encouraging preliminary results. Estimated process economics
based on such results compare favorably with those of other flue-gas
de-sulfurization processes. Westvaco, EPA's contractor, was to provide
a detailed final report early in 1975 on this process development.
6.2.1.1.8 Lime/Limestone wet scrubbing--Lime/Limestone scrubbing is the
prevalent FGD technology in the U.S. today. Eleven of 16 operational
facilities are either lime or limestone scrubbing systems. And 13 of 18
FGD facilities currently (November, 1974) under construction are lime or
limestone systems.
In 1972, EPA initiated a program to test a prototype lime and lime-
stone wet-scrubbing system to remove sulfur dioxide and particulates from
stack gases. Three scrubbers (10 MW equivalent each) of different design
were connected to stack-gas ducting of a coal-fired boiler at the TVA
Shawnee Power Station, Paducah, Kentucky. Bechtel Corporation was the
major contractor.
Major goals of the program are to: (a) characterize effects of
important process variables on sulfur dioxide and particulate removal;
(b) develop mathematical models for design of full-scale systems for
installation at other power plants; (c) perform long-term reliability
testing; and (d) conduct a sludge disposal evaluation program.
Reliability of both the limestone and lime scrubbers has been
improved. The scrubbers can be operated continuously for three months
before shut-down is required to clean the mist eliminator. Present
efforts focus on solving the mist eliminator pluggage problem. That
would allow reliable operation for periods of six months and longer.
405
-------
During system testing, over all particulate removal efficiencies
of 98.4 to 99.4% and SO removal efficiencies of 77 to 99% were obtained.
X.
System tests continue with modification of various operating parameters
to determine effects that such changes have on system reliability and
system performance.
The sludge disposal evaluation program has recently been initiated.
This program will evaluate disposal of both untreated and treated
(fixated) lime and limestone sludge materials.
6.2.1.1.9 Double alkali--The double-alkali process, like lime/limestone
wet-scrubbing processes, produce a throwaway product consisting of fly
ash and calcium sulfite/sulfate. The process was developed in various
forms in an effort to avoid problems associated with use of absorbent
slurries in lime/limestone processes. The Double Alkali Process uses
a clear scrubbing liquor containing soluble alkali species rather than
absorbent slurries.
Although less thoroughly researched and developed than lime/lime-
stone wet-scrubbing processes, double-alkali systems have the potential
to attain high sulfur-oxide removal efficiency and good reliability at
competitive costs. One problem associated with these systems is the
potential for ground and surface-water pollution by solubles in the
waste product. However, there are ways to reduce or eliminate this
potential secondary problem.
6.2.1.2 Combust ion modificat ion--Combus t ion sources contribute about
98% of NO emissions from stationary sources. Studies to date indicate
A
that combustion modification is the principal method available in the
near future to control NO emissions from combustion of fossil fuels.
J\.
406
-------
The most promising combustion modification techniques include combustion
with low excess air, recirculation of flue gas in the fuel/air mixture,
staged combustion and burner/combustor re-design.
So far, flue gas recirculation has been the most effective combus-
tion control technique for NO emission from nitrogen fixation of atmos-
J^
pheric nitrogen. Staged combustion is the most effective control method
for NO emissions from fuel nitrogen conversion. The development of
X
this technology by EPA has allowed control of NO emissions from gas-
X.
and oil-fired utility boilers to a level of 150-250 ppm. In short-term
testing, combustion modifications have also resulted in NO reductions
jC
of up to 50% in commercial coal-fired utility boilers.
Also, burner and furnace design variables cause widely varying NO
Jk.
emission levels in all boiler categories. Current efforts are directed
at continued development and demonstration of modified operating pro-
cedures for NO control in pilot- and commercial-scale demonstrations of
X
combustion control of NO for existing and new combustion sources in all
J\.
size categories.
In addition, R§D efforts will be extended to develop control tech-
nology applicable to combustion of additional fuel types (including
coal-derived fuels such as low Btu gas, mixed fuels, waste fuels,
mathanol, etc.) and covering a wider range of combustion equipment (e.g.,
stationary gas turbines, internal combustion engines, industrial process
furnaces, etc.).
EPA/CSL has recently completed a field test program to study method
of reducing NO formation in coal-fired steam generating units. Intro-
J\.
duction of biased firing has developed as a product of this study. Under
407
-------
these conditions, fuel-rich first-stage firing is followed by delayed
introduction of the total combustion air supply. In actual measurements,
the NO emission level was 1.04 gms NO /10 , or 17.5% lower than the
New Source Performance Standard.
6.2.1.2.1 Utility boilers--
Control Limitations--Although general techniques are well known,
results of application of these techniques to practical combustion
systems are still uncertain. Most applications to now have been retro-
fits to existing equipment and they do not approach optimum levels of
NO control that should be achieved.
Application of techniques is difficult because they must be applied
in a manner that prevents increases of other pollutants (i.e., - CO and
smoke) and losses of system efficiencies. Further development work is
needed, particularly for heavy oil and coal-fired systems. The impact
of various alternate fuels on the control techniques must also be
assessed.
Control Application Experience--Implementation of combustion modi-
fication techniques to date have been almost exclusively used with
utility boilers. The Los Angeles APCD has had standards for existing
gas- and oil-fired utility boilers for some time (until December 31,
1974, 225 ppm for gas and 325 ppm for oil). These standards are being
met using combinations of staged combustion, low excess air and flue
gas recirculation.
In addition, the Los Angeles APCD has a regulation limiting new
units, regardless of size, to 140 Ibs. per hour. That means that a
315 MW boiler can emit about 40 ppm NO (@ 31 07). A unit (Scattergood
b
408
-------
3) has been installed that is expected to meet this limit firing natural
gas. This unit uses the full range of available combustion modification
techniques and if successful, will significantly advance the state of
the art of application to gas-fired utility boilers.
Federal standards have been set for new (ordered after August, 1972)
utility boilers greater than 25 MW for all fuels. The approximate levels
are: 168 ppm NO for natural gas, 230 ppm for oil and 500 ppm for coal
(all concentrations are at 31 CL dry basis).
Boiler manufacturers are currently selling units guaranteed to meet
these standards. Staged combustion, low excess air and burner design
changes are being combined to meet these regulations. The main uncer-
tainty is how to achieve the standards on coal wall-fired boilers.
As previously mentioned, combustion modification techniques used
to control NO may adversely affect emissions of other pollutants (i.e.,
A,
- CO, HC and particulates) and system efficiency. However, since much
of the data from existing field equipment is not designed for applica-
tion of these techniques, the significance of these adverse effects
must be carefully assessed. There is good evidence that a system
initially engineered for NO control can also achieve emission standards
J\
for other pollutants at equal or greater efficiency.
6.2.1.2.2 Industrial/residential/commercial boilers--For other sources--
industrial and commercial boilers, residential heaters, industrial pro-
cess furnaces and stationary internal combustion engines--no standards
have been proposed. The Rfp for each of these sources is in various
stages of progress and will influence both standards and implementation
schedules.
409
-------
The residential and commercial boiler field tests were initiated
by Battelle-Columbus Laboratories in October, 1970 under sponsorship of
the American Petroleum Institute (API). To expand EPA's in-house stud-
ies in this area, a follow-up program jointly sponsored by API and EPA
was awarded to Battelle and completed in September, 1973. The tests
were primarily with oil-fired equipment, but some gas-fired boilers
were included.
The equipment used ranged from 80,000 to about 20,000,000 Btu's
per hour. The major conclusion reached was that combustible emissions
from residential equipment can be greatly reduced by proper servicing.
By replacing old, worn-out units, CD emissions were reduced by more
than 65%, HC emissions were reduced by 87%, filterable participate by
17% and total particulate (filterable and condensable) by 33%. Tuning
the burner by proper service methods further reduced CO emissions for
a total reduction of more than 81%.
The Bacharach smoke number was reduced by 59%. Even these reduc-
tions are conservative since Battelle only included burners that were
under a burner service contract. The findings strongly indicate that
emissions from residential sources could be reduced to insignificant
levels simply by proper maintenance that identifies units in need of
replacement and by tuning the remainder. Significant NO reductions
J\.
should result from burner re-design programs described elsewhere in
this Report.
The major results from commercial boiler operations were that
emissions are mainly affected by boiler and burner design and fuel
type. Also, for commercial boilers operating at steady state
410
-------
conditions, Bacharach smoke numbers showed consistent trends in rela-
tionship to filterable participates when firing different oil in the
same boiler.
Earlier studies have indicated that the most important source of
polycyclic organic matter, particularly benzo(a)pyrene, was inefficient
combustion of coal in small hand-fired residential furnaces. At present,
a study is under way to evaluate emissions from combustion of coals of
various types in a stoker-fired system and to assess combustion of smoke-
less coal. The ultimate aim of this work is to provide improved low-
emission technology for operation of this class of residential and small
commercial coal-burning furnaces.
6.2.1.2.3 Gas turbines--Publication of New Source Performance Standards
for stationary gas turbines was planned for December, 1974. Standard
setting was planned for the first half of 1975. It appears that dif-
ferent standards may be established for gas and oil-fired units. Cur-
rent target levels for stationary gas turbines of over 1.5 million Btu/
hr. (about 125 HP) are 77 ppm NO (15% excess oxygen) for oil-fired
.A.
units and 55 ppm NO (15% excess oxygen) for gas-fired units.
J\.
Combustion modification experience has centered chiefly on methods
to reduce peak temperature. Although interim methods such as water or
steam injection and exhaust gas recirculation offer potential for sig-
nificant reductions in NO emissions, advanced combustor concepts will
Jv.
be needed to achieve optimum control of all pollutants. Turbine manu-
facturers are developing low NO combustors since they think that "dry"
Jv
controls will be more economical and attractive to users. At present
EPA is conducting surveys and field tests of gas turbines to establish
411
-------
the state of the art of control technology and to identify R§D require-
ments for these sources.
6.2.1.3 Primary Partidilate Control Devices--Initial work has empha-
sized characterization of conventional control equipment. Prior to
January, 1975, sufficient data should have been available to learn about
the actual control capability in terms of fractional efficiency of
electrostatic precipitators, fabric filters and wet scrubbers. A series
of tests are under way to determine relative collection difficulty asso-
ciated with specific industrial dusts. Plans to accomplish this call
for use of standardized control equipment (pilot scale) with accompany-
ing measurement systems available on mobile mounting. This will facili-
tate comparison of numerous sources. The standardized equipment include
a fabric filter, two scrubbers and an electrostatic precipitator.
In the past year, CSL has been involved in a program designed to
find acceptable methods to control emitted fine particulate matter.
Program results so far indicate that several conventional devices, such
as fabric filters, high efficiency electrostatic precipitators and cer-
tain high-energy scrubbers, are able to control fine particulate emis-
sions on particulate matter as small as 0.01 microns in diameter. These
devices, to be effective, must be designed, operated and maintained to a
very strict standard. Much additional work is necessary to reduce capi-
tal and operating costs associated with these devices when they are used
as fine particulate control devices.
CSL is conducting field tests of industrial ESP's to determine fine
particle collection and economics of ESP systems. Recent results indi-
cate that properly designed and operated ESP's can control fly ash from
412
-------
coal-fired utility boilers at an over all mass efficiency in excess of
991. These same efficiency measurements indicate that these ESP's are
more than 90% efficient on a mass basis in collecting all particle size
fractions as small as 0.01 microns.
Last year, full-scale fabric filter systems on combustion sources
were characterized as to their ability to remove particulates as small
as 0.01 ym. Test results show that the filter systems typically oper-
ate at 99.8% over all collection efficiency and at greater than 99%
efficiency even for sizes as small as 0.1 ym.
Work is also being done on particulate collection at elevated
temperatures and pressures as required to meet the needs of coal-
combustor-gas turbines and coal gasification processes. Development
has started on high-temperature electrostatic precipitation. Granu-
lar filters, special inertial separators and other high temperature
devices are being evaluated for particulate clean-up on a fluidized
bed coal reactor.
6.2.1.4 Novel Combustion Systems
6.2.1.4.1 Fluid bed combustion--Under contract to EPA/CSL, Exxon R§E
is investigating sustained operation parameters of a coal-fired pres-
surized (10 atmos) 630 KW mini-plant at their Linden, N. J. headquar-
ters. Sustained operation attainment has been delayed for about six
weeks to correct minor technical problems.
Significant increases in combustion efficiency and improved pol-
lution control are attainable by fluidized bed combustion. This offers
the greatest potential for producing electrical energy from fossil
fuels within present environmental constraints and at a lower cost than
conventional power plants using low-sulfur fuel or stack gas cleaning.
413
-------
6.2.1.4.2 Refuse firing for steam generating--A CSL study to evaluate
technical aspects, cost effectiveness and environmental benefits of
burning municipal refuse in utility grade boilers was completed by the
Envirogenics Company in 1971. Cost analyses indicated that, in most
cases, solid waste disposal costs would be significantly less than the
cost of disposal in a municipal incinerator. In some instances, sav-
ings were $2 to $5 per ton of refuse burned. Analyses also indicated
that if all available refuse in a selected large metropolis were to
replace coal in power plants not currently having SO- control, SCL
burdens in that air pollution region would be reduced by about 7%.
There would also be a moderate reduction in atmospheric particulates.
The feasibility of burning shredded municipal wastes as a sup-
plementary fuel in coal-fired utility boilers is being demonstrated in
a program sponsored by EPA, the City of St. Louis and the Union Electric
Utility Company. A waste processing plant has been constructed and fir-
ing facilities have been constructed at a power plant 18 miles away.
Initial tests were conducted by Union Electric of St. Louis,
Missouri, to determine effects of combined firing of shredded refuse
and coal on pollutants emitted from a boiler, boiler operating
characteristics, efficiency of the electrostatic precipitator and com-
bustion efficiency of solid waste as fuel. Tentative conclusions
indicated that increased refuse firing rates produce increased combus-
tion efficiencies. No significant changes were noted in gaseous
pollutant levels when refuse and coal are fired together under test
conditions. However, there is an apparent decrease in electrostatic
precipitator efficiency.
414
-------
This drop in ESP efficiency is probably the result of increased
gas flow-rates and changes in ESP operating characteristics. Additional
tests are being undertaken to further investigate effects of refuse pro-
cessing and firing on emissions and energy recovery. Successful comple-
tion of this project will help solve solid waste disposal problems and
add an increment of fuel to meet industrial needs.
6.2.1.5 Catalytic reduction--In a contract with Environics. Inc.,
selective noble metal reduction of NO with ammonia is being explored
on a pilot scale. The experimental system uses a side stream equiva-
lent to 2 MV from a gas-fired utility boiler as a flue-gas source. The
flue gas, containing around 225 ppm of NO, is mixed with ammonia and
then contacted with a catalyst bed at elevated temperature. NO reduc-
tions of 90% have been achieved. However, equal or greater amounts of
NH7 (e-g-> 25-30 ppm) are also present in the exhaust.
The pilot scale unit has accumulated about 2000 hours of testing
with natural gas firing with essentially no performance degradation of
the catalyst. Current plans, made necessary by recent natural gas cur-
tailment for utility use in the Los Angeles area, are to test with sul-
fur-containing flue-gas beginning in the Winter of 1974. In preparation
for testing with flue-gas from oil combustion, an electrical pre-heater
has been installed to raise flue gas temperature prior to entering the
catalyst bed. A report on pilot plant results with gas firing and oil
firing should be available in Spring, 1975.
TRW Systems has done a technical and economic assessment of vari-
ous catalytic schemes for NO control for stationary power plants. The
purpose was to review catalyst technology for NO abatement, assess
J\.
415
-------
technical and economic feasibility of various schemes, evaluate appli-
cability to power plants and recommend the most promising schemes. A
comprehensive data bank was developed with about 250 articles and patents
on pertinent NO catalysts and catalytic processes.
-A.
On a lab scale with simulated flue-gas (without fly ash), cata-
lytic reduction of NO was investigated over space velocities ranging
from 5,000 to 20,000 hr STP and at temperatures from 200 to 450°C.
Approximately 45 catalysts were screened to determine nonselective NO
A.
reduction with H~ and CD and selective reduction with NH,, H2 and CO.
Identified as being most promising (because of their activity) for
selective reduction with NH, are iron-chromium, venadium, copper-lead,
molybdenum and platinum catalysts.
Parametric investigations of best catalyst systems with major
emphasis on selective NO - NH, non-noble systems have been completed.
A. O
These parametric studies indicated NO conversions of about 60-95% with
NH
_ stoichiometry about 1.0, temperature of 400°C, space velocities of
5000 to 20,000 hr"1 and inlet NO concentrations of 250-1000 ppm.
Preliminary capital and operating cost estimates for application
of this process on an 800 MW power plant are $6.15/Kw and .88 mills/Kwh
respectively. These estimates assumed fly ash removal prior to NO
removal and that the SO present did not affect the catalyst. This
A.
latter assumption was verified in preliminary screening of Fe-Cr and
vanadia catalysts.
6.2.1.6 Fuel Cleaning
6.2.1.6.1 Coal--The goal of the coal-cleaning program is to determine
the impact that coal cleaning could have on control of sulfur oxide by
416
-------
increasing supplies of low sulfur coal. The approach taken here has
been to identify qualities of U.S. coals, potential sulfur reduction
attainable by physical methods and improvements in coal-cleaning
techniques that would maximize sulfur reduction.
So far, this program has achieved many of its objectives. The
vast coal float-sink testing program initiated in 1965 is 75% com-
pleted. Evaluations and applicability of existing physical cleaning
techniques are complete. And modification of froth-flotation to
obtain sulfur reductions not achievable before have been piloted. The
emphasis placed on R/D in coal and coal-cleaning programs by EPA has
stimulated interest in additional efforts by other organizations.
The work supported at TRW on chemical extraction of pyrite involves
an approach that would achieve 95 to 100% pyrite removal. The full
development of this approach would enable coal processors to guarantee
sulfur reduction to levels that would meet new performance standards.
In the Appalachian coal states (Alabama, Maryland, Ohio, Pennsyl-
vania, Tennessee, Virginia and West Virginia), much of the coal cur-
rently mined could be reduced to a 0.75% sulfur level or less by
application of the Meyers' process. The combination of mechanical and
chemical cleaning techniques would enhance the over all applicability
of coal-cleaning technology and should improve coal-cleaning economics.
A shift in production to specific coal beds could make an even
more significant impact. The coal-cleaning plant in the near future,
then, would most likely involve an integrated physical and chemical
removal process that would produce a coal of lowest sulfur and lowest
ash with a minimum of trace metallic constituents.
417
-------
CSL has recommended that a prototype coal cleaning demonstration
plant not be built. They think that EPA can have the greatest impact
through technical support in engineering systems for coal cleaning (in
conjunction with other control techniques where required) that would
result in the most appropriate control strategies.
Such an approach has been initiated with the state of Iowa. CSL
will provide a technical data base for coal cleaning (mechanical and
chemical) and other control methods that will allow for development of
strategies to meet Iowa's short-and long-range programs. In this study,
it is evident that coal cleaning will not only be an initial technique
to meet current state standards, but may be a necessary step to meet
long-range sulfur emission levels.
An integrated coal cleaning development program would result in
the earliest availability of a process that could have maximum impact
on all users of coal. In the Appalachian coal region, use of this com-
bined approach on coals now being mined for combustion purposes would
result in 50 to 60% of the coals having a sulfur level less than 0.75%
and 70 to 80% of the coals having a sulfur level less than 1.0%. The
use of these lower sulfur-level coals could achieve the following
results:
1. Residential users (normally the lowest income group) would
not have to switch to other fuels. In many cases (such as in Phila-
delphia, Detroit, Chicago, etc.), this user group could not financially
afford to switch from coal to gas or oil.
2. Industrial and commercial users (using about 15% of the com-
bustion tonnage) could meet existing standards and avoid requirements
to switch to alternate fuels.
418
-------
3. Large utility users, in the majority of cases, should be able
to meet state standards for existing units and many new or planned
stations could use coal instead of gas or oil. The continued use of
coals would aid in and prevent increased use of gas and oil by the
utility industry.
So over all, there would be an increase in the use of coal by the
utility industry and an increase in availability of gas and oil to area
source users who have few alternative control strategies available to
them.
EPA/CSL is currently negotiating for construction of a chemical
coal de-sulfurization pilot plant. This technology will not be avail-
able for commercial application until the mid-1980's.
6.2.1.6.2 Liquids--Much of the oil currently available in the U.S. for
fuel is high sulfur, high metal content residual oil. Even if these
oils could be economically de-sulfurized, high vanadium and nickel con-
tent would probably present environmental problems and would certainly
cause equipment problems in combined cycle operations. Nitrogen contents
of these oils can also be a problem especially with combined cycle
operation.
Because of the inter-related and dependent nature of these con-
taminates, EPA/CSL has programs to study separate and combined effects.
For example, both enhancing and inhibiting effects of simultaneous
hydro-de-sulfurization and hydro-de-nitrogenation reactions have been
observed. Additional discussions of liquid fuel cleaning are presented
in the section on Coal Conversion.
6.2.1.7 Fuel Gas Production
419
-------
6.2.1.7.1 CAFE--The Chemically Active Fluid Bed (CAFB) process for
gasifying and de-sulfurizing heavy oil has been demonstrated in a 0.75-
Mtf continuous pilot unit at Esso Research Center, Abingdon, England
(Esso, England). The CAFB process involves injection of heavy fuel
oil into a shallow (about two feet deep) fluidized bed of lime/lime-
stone particles to partially oxidize, gasify and de-sulfurize the oil.
The continuous pilot unit fires a commercial boiler rated at 10 million
Btu/hour and demonstrated a service factor of 95% in the latest run of
about 400 hours duration.
In addition to 90% sulfur removal at a calcium-to-sulfur feed
ratio of 1.5 to 1, the CAFB has completely removed vanadium and has
removed 75% of the nickel and 36% of sodium. A reduction in the NO
A.
emissions from 263 ppm (when the boiler was oil-fired) to 160 ppm
(when fired on the CAFB product) has been demonstrated.
Economic studies continue to show that CAFB is a viable com-
mercial process. The effluent from the limestone regenerator is
5-10% SCL and can be reduced to sulfur using existing technology.
Demonstration of a 50-MW unit is planned at Providence, R. I. New
England Electric System (NEES, a private utility) is participating in
a process design effort with Stone and Webster Engineering Company
(SWEC), Esso England and Westinghouse Research Laboratories, under EPA
sponsorship, to design the demonstration unit. The sulfur removed by
CAFB will be reduced to elemental sulfur using a modified Glaus process.
An atmospheric pressure CAFB gasifier demonstration plant will be
constructed and ready for integrated operation by 1977. Simultaneously
with demonstration of the atmospheric pressure gasifier, a pressurized
420
-------
CAFB facility will be constructed and operated in conjunction with a gas
turbine. The thinking is that a pressurized CAFB can reduce costs of
power generation by gas turbines by 30%.
6.2.1.7.2 Fuel pollutant characterization--As a first step in any
program involving fuels, a certain minimum amount of data on fuel is
necessary. Today, this must be broadened to include all potential pol-
luting constituents. Recognizing this, EPA/CSL has issued a report on
an extensive literature survey that presents data on potential pollu-
tants (e.g., lead, cadmium, beryllium, mercury, sulfur, nitrogen, etc.)
in fuels.
Since coal is to be a major energy source and existing data are
inadequate for many constituents, CSL's interest is both in developing
analytical techniques for trace elements and in the analyses themselves.
Techniques used include neutron activation, x-ray fluorescence, atomic
absorption, ion-specific electrode, direct reading optical emission and
others.
Since the form of a constituent in coal may effect its pollution
potential, an effort is also under way to better define the nature and
composition of coal's mineral matter. These combined studies are pro-
viding information that will be helpful in determining the quantity and
fate of potential pollutants in coal's use, processing or conversion to
clean fuel.
6.2.1.7.3 Coal conversion
Liquid Fuels
Technology for conversion of coal to liquid fuels is being developed
in the United States at an accelerated pace. This technology represents
421
-------
the major alternative to increased dependence on imported petroleum
fuels because of present domestic demand and inability of the domestic
supply to meet this demand.
Various methods are under development to produce liquid fuels from
coal. Fuel products include all ranges of low-sulfur oil that can be
burned directly in utility and industrial boilers, heavy oils that can
be processed through a conventional refinery and naphtha. The heavier
liquid fuels produced from coal retain significant (up to 2%) quanti-
ties of coal nitrogen and usually less than 1% sulfur. Control of 98+%
of the inorganic coal constituents (trace metals, ash) is projected.
Technologies to produce liquid fuels from coal will not be commercially
available before 1980.
EPA/CSL has several activities under way to obtain data on liquid
fuel from coal processes. One of these activities involves study of
catalyst materials for selective removal of contaminates such as
vanadium, nickel, nitrogen, sulfur and others from coal-derived oils.
Such a ctalyst material must be developed to control the relatively
high nitrogen content of the oil. And it must perform satisfactorily
in the presence of the sulfur content of the oil.
Study of competing reactions and removal of these contaminates
will increase the usefulness of these liquids (e.g., combined cycles)
and result in a cleaner environment. Another data acquisition task
that CSL has under way is directed toward obtaining quantification of
the pollutant potential of the Solvent Refined Coal (SRC) process.
Determination of material balances for sulfur and nitrogen in the SRC
process was included and is currently in the process of being reported.
422
-------
Work is also in progress on preparation of engineering designs of over
all coal-processing complexes based on the SRC process that will pro-
duce optimally clean gaseous and liquid fuels.
A program for total environment assessment of liquid fuel from
coal technologies is currently under way and is being expanded by EPA/
CSL. This type of environmental assessment activity is required because
of the importance of insuring that coal liquefaction technologies will
include adequate environmental safeguards. Their use otherwise to meet
domestic energy needs may be severely restricted because of potential
environmental impact.
Liquid fuel from coal complexes of 30,000 and more tons of coal
per day are projected. Obviously, technologies that treat such large
quantities of coal for liquid-fuel production will also handle and pro-
cess enormous quantities of sulfur, nitrogen, trace constituents and
polynuclear aromatics under conditions that could lead to significant
pollutant discharges without adequate process control.
This environmental assessment activity will determine those con-
trols required for liquid fuels from coal technologies or will identify
additional information that must be obtained to identify control needs.
After defining process pollutant control requirements, appropriate
solutions can be pursued,
Gasification
Coal conversion to a gaseous fuel is an important way to provide
environmentally clean energy when done properly. A significant
increase in coal gasification activity by the private sector has not
taken total environmental impact into account. HPA/CSL has done studies
that address this need.
423
-------
Using currently available information and engineering estimates,
coal conversion processes are being evaluated to determine associated
environmental problems. Reports have already been issued on Koppers-
Totzek, Synthane and Lurgi processes. And reports on other processes
will be issued soon.
Although there are many potential problems with elements and com-
pounds of sulfur, nitrogen, mercury, lead, cadmium, vanadium, particu-
lates, etc., there is an urgent need for quantifying constituents in
all output streams. As a result, use is being made of new environmental
data obtained or planned for from existing commercial gasification com-
plexes overseas and pilot plant operations in this country.
Through a planned and integrated effort to obtain data, do environ-
mental assessment, identify needed environmental safeguards and research,
develop and demonstrate environmental controls, energy processes that will
be efficient and environmentally sound can be commercialized.
The preceding basically applies regardless of final gas Btu con-
tent. However, use of low or intermediate Btu gas presents additional
problems. EPA/CSL is studying this in relation to environmental aspects
of combined cycle power generation and retrofit of existing commercial
and industrial facilities to use this alternate fuel source. Results to
date indicate that over all gasification and power generation efficiencies
will be improved with proper integration of environmental controls while
developing environmentally sound systems.
These studies also help to identify alternatives and environmental
needs. And some users of clean premium fuels could switch to an alterna-
tive gasified fuel with proper environmental planning and free premium
fuels for premium needs.
424
-------
6.2.2 Industrial Processes
6.2.2.1 Coke Making--Production of coke in conventional coke oven bat-
teries has traditionally been a major source of particulates, sulfur
oxides, odors, carcinogens and hazardous trace metals. Until recently,
emissions from this source have been relatively uncontrolled.
The majority of existing coke-making facilities will be operable
for the next 25 years. And most facilities built in the next 10 years
are expect to feature conventional batteries. As such, CSL has studied
developing control technology for conventional batteries to meet pollu-
tion control guidelines until the end of the century.
Indications are that pollution control technology developed for
conventional batteries, whatever it is, will probably be expensive,
have high-maintenance cost arid still not control all pollutants. The
expected long-term solution to the coking problem is a continuous cok-
ing process that emits pollutants in a more controllable form. CSL's
involvement in coking should ultimately include continuous coke-making
when it's more feasible.
Specific coke oven emission-control projects with CSL involvement
includes smokeless coke charging, enclosed coking pushing and quenching,
smokeless coke pushing and improved coke oven door seals.
Coke oven projects obviously concern pollutants other than hydro-
carbons. Nevertheless, these projects will have a significant impact
on stationary source hydrocarbon emissions.
6.2.2.1.1 Coke oven charging emission control program--EPA/CSL jointly
funded an emissions control program with the American Iron and Steel
Institute to improve coke oven charging techniques and abate the large
425
-------
volume of particulates and gases released during charging operations.
The program operation was managed by Jones and Laughlin Steel Corpora-
tion at their Pittsburgh Works. The program culminated in design and
construction of a new type of larry (coal charging) car that was both
functional and smokeless. It also incorporated an environmentally
controlled cab for safety and comfort of the operator.
The Mitre Corporation concurrently did an emissions test and
evaluation program. Measurements of major gases emitted and reductions
attained using the new larry car are tabulated below:
TSP (total suspended 120 gms/charge 85% reduction
particulate)
THC (total hydrocarbon) 20.43 scf/charge 40% reduction
CO 6.28 scf/charge 64% reduction
CCL 3.82 scf/charge 87% reduction
NO .021 scf/charge 82% reduction
J\.
Favorable response of the coke industry to development of this
improved technology is demonstrated by their commitment of $10,000,000
for new installations incorporating these and other features.
6.2.2.1.2 Enclosed coke pushing and quenching--EPA and the National
Steel Corporation are funding a demonstration of an enclosed coke push-
ing and quenching system on National's new Weirton Steel Division
Brown's Island coke plant. Construction was completed and the system
placed into operation in May, 1973. The demonstration unit is presently
being brought up to full production status. This will be followed by a
one-year emission testing and system evaluation phase to be completed
in 1975.
426
-------
In this system, the coke is completely enclosed from the moment it
leaves the oven until after it is quenched. Emissions produced in the
push and transfer to the quench station are drawn off and removed with
a high-energy scrubber on the gas cleaning car. Emissions produced by
hot coke in the underground track hoppers are also controlled by a high-
energy scrubber. The relatively low-volume continuous steam plume
generated in the continuous quenching operation is contained by hoods
and controlled by a vapor suppressor in the stack.
Current information indicates that this system will apply to nearly
all new coke batteries. This is particularly significant because half
of the existing coke batteries are at least 20 years old. Based on an
average life of 30 years, nearly half of the 250 existing batteries will
have to be replaced in the next 10 years.
Since continuous coke-making processes may not be available until
the end of that period, most new batteries will be conventional slot
ovens with useful lives extending well into the next century. Demon-
stration of this system will provide proven emission control technology
that can be integrated into the initial plant design.
6.2.2.1.3 Smokeless coke pushing--CSL contracted with Ford Motor Company
to test and evaluate the coke pushing emission control system developed
by Koppers Company and installed at Ford's River Rouge Plant. Principal
features of the system, now in operation at Ford's River Rouge Plant,
Dearborn, Michigan, include a fume-collecting hood, a fume main-gas
cleaning unit and a modified quench car.
Since the control system will apparently fit most, if not all, exist-
ing installations, demonstration of the Ford system will make available to
427
-------
industry a relatively low-cost device that can significantly reduce coke
oven emissions. It will serve as an interim solution to existing bat-
teries (more than 15 to 20 years old) until these batteries can be
replaced (average battery life is 30 years). At that time, a more com-
plete control system can be installed.
6.2.2.1.4 Coke oven door seals--Leakage of gases and condensible hydro-
carbons from coke oven doors contribute significantly to typical poor
air quality in the vicinity of coke batteries. These emissions contain
high levels of carcinogenic materials such as benzo(a)pyrene. They are
potential health hazards to workers and surrounding communities. The
solution to this problem has a relatively high priority within EPA and
the steel industry.
The major technical objective of this research effort is to create
and develop practical solutions to sealing coke oven doors to eliminate
or reduce emissions from this source to the lowest extent practical.
Another objective is development of methods to measure emissions from
coke oven door seals as installed on an operating commercial battery.
Work on this project is being done by Battelle Columbus Labora-
tories. It is co-funded by the American Iron and Steel Institute (AISI).
The work was initiated in July, 1974. The primary work done to date
includes problem definition, review of technology and technology transfer.
About 17 coke-producing facilities have been visited. That includes
most U.S. batteries with door seal development programs or new sealing
techniques being tried. New systems have been cataloged and appriased.
Arrangements have been made to conduct tests that will define the
operating conditions that a door seal will encounter. These measurements
428
-------
include temperature, fluid pressures, warpuge, material stress and com-
position. Other difficult sealing problems in such areas as nuclear
reactor, power boilers and spacecraft are being analyzed for possible
technology transfer.
6.2.2.2 Basic oxygen process (BOP) charging emission control--Primary
operations to produce steel by the basic oxygen process (BOP) include
charging scrap, charging hot metal, oxygen blow, chemical tests and
tapping. In charging, particularly with hot metal and during oxygen
blow, large quantities of emissions are produced. There is no proven
technology at present to control charging emissions.
A project was initiated in June, 1973, with National Steel Corpora-
tion's Research and Development Center to develop technology for control
of particulates, gases and fumes emitted in charging of basic oxygen
steel-making process vessels. This development program focuses on pro-
cess modifications that will allow emissions to be collected. Considera-
tion will also be given to suppressing emissions in the vessel.
Specifications will be developed for a prototype emission-collection
system. Additional specifications will be developed for a gas-cleaning
system to be added to the prototype collection system. The development
program is structured so that results will be applicable to the total
industry.
6.2.2.3 Iron foundry--The iron foundry cupola system is the predominant
U.S. process for melting and refining iron. The principal drawback of
the conventional system is that effluent must be cooled by air dilution
to remove particles from exhaust gases so that it can be treated by
lower temperature gas-cleaning devices. This requirement produces more
429
-------
gas to be treated, higher gas cleaning costs and loss of heat that other-
wise could be a source of non-polluting energy.
The control system that is being demonstrated to overcome the above
problems involves a new water-cooled un-lined cupola coupled to a com-
plete emission control system consisting of an afterburner, a dry media
heat exchanger, a pair of cyclones and a fabric filter (the filter to be
installed later). The system has been operating in the Flynn and Emrich
Foundry at Baltimore.
All carbon monoxide is converted to carbon dioxide in the natural-
gas-fired afterburner. The heat exchanger extracts heat from the after-
burner exhaust gases, cools the gases so that they are compatible with
the dust-collecting unit and does this with minimal air or water dilu-
tion. Transferring this heat to the blast air-stream reduces the amount
of coke used that, in turn, reduces both coke particulate blown from the
cupola and over all emissions from the coke manufacturing operation.
The dry media heat exchanger minimizes such problems as fouling of
the heat transfer surfaces by contaminated cupola exhaust gases. This
problem results in frequent and costly maintenance with conventional
tube-type heat exchangers. The expected slight fouling of the dry media
in the heat exchanger will produce only negligible performance deterio-
ration.
Dust particles in the cooled heat exchanger exhaust-stream are
removed by the cyclones and fabric filter. Since only about two-thirds
of heat exchanger capacity is required for blast air heating, the
remaining one-third is available as a non-polluting heat source for
other foundry functions.
430
-------
6.2.2.4 Sinter plant windbox emission control--The sinter plant takes
ore fines, mill scale, iron-bearing particulate matter from control
devices, etc. and agglomerates them into large pieces suitable for
blast-furnace feed. In that way, their iron values can be recovered.
This is done in a sinter machine where iron-containing material is
mixed with limestone and coke breeze and fed in a thick layer on a long
slow-moving continuous belt.
Coke at the top of this layer is ignited at the front end of the
machine. Ambient air is drawn down through this slow-moving bed to
provide oxygen for maintenance of coke combustion. The products of
combustion are collected in a windbox and exhausted to the atmosphere.
As the mixture proceeds,, the flame front goes deeper and deeper
into the layer until it readies the belt at the discharge end of the
machine. The heat given off by the burning coke is sufficient to make
other materials tacky and stick to each other. But the heat is not
great enough to completely liquify it. The fused material being dis-
charged from the machine is then broken into large pieces in a sinter
breaker and screened. The screened product is used as blast-furnace
feed.
The emissions contain particulates, S02, other gases and hydrocar-
bons derived from oils that are driven off by distillation on some com-
ponents of the charge material.
CSL's current contract with National Steel Corporation calls for
identification of windbox emissions versus feed characteristics on a
conventional sinter machine and design of a windbox gas-recirculation
system. The second phase, now in final stages of negotiations, calls
431
-------
for an extensive test and evaluation of a sinter machine fitted with a
windbox gas-recirculation system.
In this approach, part of the windbox gases are returned to a hood
located above the sinter bed instead of being discharged to the atmos-
phere. Enough ambient air is mixed with recirculated gases above the
bed to provide oxygen for the process.
The recirculation approach promises several advantages, including:
1. Reduced volume of gases to be cleaned.
2. Reduced hydrocarbon emissions by using the sinter bed as an
afterburner.
3. Reduced particulate emissions by using the sinter bed as a
filter.
6.2.2.5 Ferro-alloy furnaces
Characterization. The Office of Air Quality Planning and Standards
(OAQPS) has conducted extensive source emission tests over the past
three years on a number of ferro-alloy furnaces here and abroad to
establish the basis for a Standard of Performance for new ferro-alloy
plants. Facilities tested included all principal ferro-alloy products
along with calcium carbide that are smelted in submerged arc-furnaces.
Since many of these samples still exist, CSL's program obtains
these samples and results of the analysis performed by OAQPS and does
more detailed and exhaustive analysis where indicated. And in those
cases where there is enough sample, the samples will undergo toxicologi-
cal screening tests.
This effort will result in maximum use of available samples to
detect those smelting operations posing the greatest danger of hazardous
432
-------
element emissions. The samples would also be used for any further emis-
sion testing and R§D on improvement of control technology.
So far, samples from five different furnaces have been analyzed for
60 elements and screened for polynuclear aromatic hydrocarbons (PNA's).
The elemental analyses agree well with the limited analyses performed by
OAQPS.
The PNA screening that was not done previously on these samples
showed high levels in some samples. The next step will be to collect new
samples with recently developed procedures that guarantee preservation
of the true character of samples (PNA's are susceptable to light and
heat).
The submerged arc-electric furnace is the principle source of emis-
sions in the ferro-alloy industry. Smelting of ferro-alloys is usually
accomplished by reducing ore with carbon that produces desired metallics
as well as substantial quantities of CO. In some cases, the quantity of
CD produced actually exceeds the production of metallics. The gases
evolved from ferro-alloy furnaces carry large quantities of particulates
that, because of high temperatures in the reaction zone, are primarily
in the submicron size range.
Conventional practice has been to leave the furnace top open. This
allows the CO to mix with large volumes of air and burn above the furn-
ace. The mixture is then collected and treated with conventional par-
ticulate control equipment before being vented to the atmosphere.
More recently, furnaces have been used in Europe and Japan that
are hooded tightly so that no excess air is carried in the furnace off-
gas. And combustion does not take place above the furnace. When this
433
-------
type of system is used, the volume of gas is decreased by 20 to 50 times
since excess air is not included.
This decreased amount of emissions can be cleaned to the same
degree as emissions from an open furnace. So total particulates emitted
are decreased by approximately the same factor of 20 to 50 times. Also,
the gas that is no longer burned over the furnace can be used as a low
Btu fuel after cleaning.
U.S. ferro-alloy producers hesitate to install totally enclosed
furnaces because they think it may reduce their product flexibility.
To solve this problem is the project goal. Specific objectives of this
study include:
1. To compare flexibility of open and totally enclosed ferro-
alloy furnace to learn about the true significance of the flexibility
problem.
2. To identify current practices that minimize any flexibility
problem that may arise in totally enclosed furnaces.
3. To develop, where warranted, recommendations for an additional
program that could be undertaken by EPA to develop or demonstrate tech-
niques and systems to minimize the problem.
6.2.2.6 Smelters--Because of substantial SO- emissions from non-ferrous
smelters and lack of available control technology, CSL has initiated a
set of development projects. Two new projects currently in operation
have as their respective objectives:
a. Providing information to EPA on a smelter-by-smelter basis
about optimum application of currently demonstrated technology to con-
trol all smelter streams with SCL and particulates at the same time.
434
-------
The output will be in a form currently acceptable to the courts as
demonstration by means of technology transfer.
b. The assessment of environmental considerations of emerging
technology for metal-winning in the primary non-ferrous industry. In
this project, consideration will be given to hydrometallurgy as new
smelting processes.
CSL and Phelps Dodge Corporation started a cost-sharing study in
July, 1971, of a new sulfur-oxide control method based on reacting SO-
with H^O and recovering the sulfur involved as elemental sulfur. Ini-
tial attempts to generate needed H~S by roasting available copper ore
with steam at 800°C showed that it was technically feasible but the
economics were prohibitive.
Emphasis has now shifted to leaching the concentrate with hydro-
chloric acid that decomposes pyrite and pychloric in the concentrate
and releases ample quantities of H2S. Preliminary results are promis-
ing. Tests are required, however, to understand the observed variations
in H2S generation as a function of the concentrate source.
EPA has entered into two separate Public Law 480 agreements (excess
foreign currency programs) with Yugoslavia. One program is with the
Institute for Copper, in Bor and the other program is with the Lead and
Zinc Institute, in Trepca. The Bor p ject is a three-year investigation
(begun in May, 1971) at an American-engineered copper smelter (the
largest in Europe). The Trepca project is a similar three-year effort
(started in October, 1971) at a lead and zinc smelter.
Objectives of both projects are:
a. Qualitative and quantitative characterization of all air pol-
lutant emissions from this type of smelter.
435
-------
b. Relating air pollutant emissions to the feed and to feed changes
so that data will apply to other smelters.
c. Relating air pollutant emissions to operations changes (changes
in smelter configuration) so that data will apply to other smelters.
d. Determining the effects of controlling sulfur dioxide and par-
ticulates on emission of all other pollutants, especially "hazardous"
heavy metals.
The relationships being developed in these two PL-480 projects
should be applicable to the domestic non-ferrous smelting industry.
Additionally, two smaller studies on potentially hazardous materials
have been done.
The first was to develop methods to remove mercury from smelter
gases. The results show promise.
The second was to determine the fate of various pollutants (such
as As, Cd, Se, Hg and Pb) when a sulfuric acid plant is used as a S0_
control process. Those pollutants ending up in the acid pose a poten-
tial hazard. About half of the acid manufactured domestically is used
for acidulation of phosphate rock. It could be transferred to foods
when used as a fertilizer.
6.2.2.7 Kraft pulping--The Pulp Mill Indirect-Contact Evaporator
Conversion project at Hoerner-Waldorf Corporation is a first-of-a-kind
demonstration of retro-fitting an existing furnace to eliminate the
standard practice of direct contact between furnaces gases and incom-
ing liquor. The direct contact portion of the furnace is a major
source of furnace emissions.
EPA is demonstrating viability of the retrofit concept by support-
ing a detailed emission-testing program for both the retrofit and the
436
-------
new boiler. Also, capital and operating cost requirements, startup
experiences and operating capabilities are being documented. And the
testing program is designed to establish optimum boiler conditions for
low-odor SCU and particulate emissions.
Preliminary indications are that a retrofitted boiler can be
operated successfully. EPA's complete documentation of retrofitted
furnace capability to control noxious emissions efficiently and eco-
nomically will provide existing mills with a viable option for meeting
local as well as anticipated Federal emissions standards.
6.2.2.8 Alfalfa dehydration--In the alfalfa industry, dehydration of
harvested "raw" alfalfa is a necessary step in processing the crop to
the standard final product of pelletized alfalfa. This dehydration is
done in a continuous system rotary-drum dryer where gas temperatures
may reach 2000°F. Because of the non-homogeneity of moisture content,
however, some of the alfalfa may be so dry when introduced into the drum
that it burns, producing significant particulate emissions.
The American Dehydrators Association (ADA) investigated this serious
problem in 1971. EPA joined in the effort in July, 1972, by providing a
grant to help the ADA further its studies. The final report on the pro-
gram, finished in December, 1973, is now available.
Results of the first test phase (1971) indicate that the drying
operation causes 75% of total emissions. Results of the second test
phase (1972) indicate that medium-efficiency wet scrubbers have the
potential to bring alfalfa dryer emissions into compliance with process
weight and rate standards. Results of the third phase (1973) indicate
that it is possible for each dehydrator to plan and schedule a technically
437
-------
and economically sound strategy to control particulate emissions so as
to meet the Bay Area Process Weight Rate Standard. The report includes
capital and operating costs applied to a representative control model
for control schemes investigated.
6.2.2.9 Nitric acid plants--CSL is involved in a joint demonstration
with the U.S. Army of the capability of molecular sieves to abate NO
A.
from nitric acid plants. This effort includes testing and evaluation
of a molecular sieve system retrofitted to a 50-ton per day ammonia
oxidation process (AOP) unit.
The objective of this program is to determine economics and oper-
ating characteristics of molecular sieves for this specific application.
The program will include a thorough evaluation of the efficacy of this
pollution control method. This method has the potential to allow the
present NO performance standard for new nitric acid plants (3 pounds/
J\.
ton of acid produced, expressed as NO-) to be reduced, perhaps by as
much as a factor of 10 to 20. System start-up was planned for March,
1974.
6.2.2.10 Refineries--Petroleum refinery catalytic-cracking feedstocks
contain varying amounts of sulfur that are distributed between the
cracked products and carbonaceous matter laid down on the catalyst dur-
ing the cracking reaction. Although distribution depends on many fac-
tors, about 5% of the total sulfur generally ends up with the catalyst
to be regenerated.
In 1970, estimated refinery SO emissions (containing 15-60% S0»)
J\. *J
from catalytic cracker regenerators amounted to 359,000 tons (1% of the
total SO emitted in the U.S.). The regenerator is recognized as an
1\.
438
-------
important source and should be more significant as refiners process
more higher sulfur crude oils.
CSL has developed a program to identify and evaluate techniques for
reducing SO emissions from regenerators by 901. The program will
j\.
include laboratory and pilot plant studies as needed to confirm the most
advantageous process identified.
6.2.2.11 Plastics and petrochemicals--The ethylene oxyhydrochlorination
vent is the main source of hydrocarbon emissions associated with manu-
facture of ethylene dichloride, a five-billion pounds per year business.
At present, there is no practical way to eliminate the oxyhydrochlorina-
tion vent from existing processes. The gases are too dilute for direct
incinceration and addition of natural gas to make the gases burn is an
extravagance. Furthermore, incineration will produce hydrochloric acid
that must be controlled by scrubbing.
It is believed that hydrocarbon control can be done with only a
modest increase in cost because anticipated benefits of increased raw
material yields will off-set control costs to a great extent. Unit cost
increases of 0.05<£ per pound of EDC are projected.
The objective of this 12-month project is to demonstrate technical
and economic feasibility of a process to produce ethylene dichloride by
ethylene oxyhydrochlorination characterized by a hydrocarbon emissions
rate-reduction of 90% or greater. Current processes use air as the
source of oxygen and vent the resultant inerts along with about 0.03 ton
hydrocarbons per ton of EDC to the atmosphere.
The proposed process, feeding oxygen with recycle vent-gas, will be
applicable to existing facilities to the extent that it will be compatible
439
-------
with other units of the over all process to produce vinyl chloride from
ethylene and chlorine. The proposed process will probably be economi-
cally competitive with existing ethylene oxyhydrochlorination processes.
6.2.2.12 Asbestos mining--Phase I of this program involved problem
definition. One asbestos-milling tailings pile and two asbestos cement-
manufacturing waste piles were sampled with hi-vol samplers and the fil-
ters analyzed for asbestos. Ambient asbestos concentrations in air were
evaluated using dispersion models. It was concluded that asbestos min-
ing -and- mi 11 ing waste probably did not have serious impact on public
health because of remoteness of sites. It seems that asbestos cement
waste probably did have high public health impact because of ambient
levels calculated as present in populated areas.
A competitive procurement now in process is entitled "Field Test-
ing of Control Concepts for Abatement of Emissions from Asbestos
Manufacturing Waste." This project will involve selection of control
options to reduce emissions from manufacturing waste-piles and field
testing these options for technical and economic viability.
6.2.2.13 Enclosed asbestos operations--Phase I of this program
involved problem definition. Five sites processing asbestos were
visited and baghouses were typically 99% efficient, baghouse effluent
7 9
gases contained between 10 and 10 fibers/cubic meter. It was con-
cluded that further improvements in control efficiency may be required
to protect public health.
Phase II of the program is a laboratory study to optimize baghouse
performance for asbestos fibers. Various options will be tested for
technical feasibility and a cost-benefit relationship established for
options.
440
-------
6.2.2.14 Asphalt roofing manufacturing--The primary objectives of this
program are to characterize atmospheric emissions of particulate poly-
cyclic organic matter (PPOM) from asphalt roofing manufacturing pro-
cesses and to describe general demographic parameters of this industry
such as size, location and growth patterns. This characterization of
PPOM includes: quantitative measurements of effluent gas streams to
determine concentration and. emission rates of total particulate, concen-
trations of selected polycyclic organic compounds and particle size
distribution.
Emission rates are related to various process variables to provide
a basis for estimating emissions from other plants in this industry.
Measurements are made before and after use of emission control systems.
Specific polycyclic compounds identified are listed below:
7,12-Dimethylbenz (a) anthracene
Dibenz (a,h) anthracene
Benzo (c) phenanthrene
3-Methylcholanthrene
Benzo (a) pyrene (C-OH,-)
Dibenzo (a,h) pyrene
Dibenzo (a,i) pyrene
Dibenzo (c,g) carbazole
The secondary objectiAre of this project is to characterize other
atmospheric emissions from these manufacturing processes such as alde-
hydes, gaseous hydrocarbons and inorganic particulate.
The following specific processes were sampled to determine emission
rates:
441
-------
Asphalt blowing
Felt saturating
Mineral coating
The blowing and saturating operation are, by far, the largest
potential sources of PPOM. Mineral handling and coating operations are
potential sources of particulate, but not PPOM. Where possible, each
process was sampled while saturating felt for shingles and then roll
roofing and under two operating conditions with two different asphalts.
When the PPOM and other pollutants have been characterized, a pro-
gram to develop and demonstrate a control technique will be initiated
if needed.
6.2.3 Particulate Control Technology
For the most part, fine particulate control is in a very early
stage of development. Recently, however, actual source tests have
shown that both electrostatic precipitators (ESPs) and baghouses
should be capable of controlling fine particulate from a limited number
of sources emitting fly ash.
An ESP has been shown to be effective on a utility boiler emitting
fly ash with a resistivity conducive to electrostatic collection. On
this utility boiler, burning a relatively high sulfur coal, a high
efficiency ESP (99+% over all mass efficiency) was more than 90% effi-
cient in mass removal of all particulate fractions as small as 0.02
microns.
Preliminary tests indicate that a baghouse is effective on a
utility boiler if close operating limits are maintained on a number of
critical variables.
442
-------
It is quite possible that applicability of conventional ESPs over
a broad rnage of sources to fine particulate control can be extended by
developing dust-conditioning techniques and by modifying the design of
charging sections and collecting electrodes. Considerable research
and development will be needed before either method can be put to prac-
tice. The development of special operating techniques and filter fabrics
seems to offer the best hope of extending the useful operating range of
baghouses.
6.2.3.1 Electrostatic precipitation--In an electrostatic precipitator
(ESP), dust suspended in gas streams are electrically charged and passed
through an electric field where electrical forces cause particles to
move toward a collection electiode. The dust, separated from the gas
by being retained on the collection electrode, is later removed from the
device. In a conventional ESP, the dust is removed mechanically. In a
wet ESP, the dust is removed by continuous washing of the collection
electrode.
A systems study of conventional ESPs has been completed. Its
results are published in an ESP handbook. A systems study of wet ESPs
was scheduled for completion in late 1974. A mathematical model that
quantitatively describes ESP operation under all conditions and indi-
cates potential performance limitation- and design improvements has been
developed and is being up-graded to improve agreement between predicted
and measured performance--especially for fine particles.
CSL is conducting field tests of industrial ESPs to determine fine-
particle collection and economics of ESP systems. Recent results indi-
cate that properly designed and operated ESPs can control fly ash from
443
-------
coal-fired utility boilers at over all mass efficiency in excess of 99%.
These same efficiency measurements indicate that these ESPs are more
than 90% efficient on a mass basis in collecting all particle size
fractions as small as 0.01 microns.
A theoretically-sound method of sizing ESPs has been developed
based on the mathematical model and field test results. A report
describing the design method and presenting equations and figures neces-
sary to use the method was scheduled for publication in late 1974.
ESPs are being re-designed to improve their particulate collection
efficiency, especially foi fine particles and high resistivity dusts.
This work includes development of dust-conditioning methods and special
particle-charging and collection section designs. Reports on fly ash
conditioning have been published. The work on conditioning has resulted
in discovery of the current carrying mechanism in fly ash at high tem-
peratures and demonstration of a new conditioning agent.
6.2.3.2 Fabric filtration--A typical fabric filter consists of a flexi-
ble layer of porous cloth through which a dusty gas is passed to sepa-
rate particles from the gas stream. Because accumulating particles
increase the filter's resistance to gas flow, deposits are removed
periodically by cleaning the cloth to maintain pressure drop across the
filter within practical operating limits. The ability to clean the
fabric without removing it is a distinguishing characteristic of this
class of gas filter.
In the last year, full-scale fabric-filter systems on combustion
sources have been characterized as to their ability to remove particu-
late as small as 0.01 ym. Test results show the filter systems to
444
-------
operate typically at a 99.81 over all collection efficiency and at
greater than 991 efficiency even for sizes as small as 0.1 ym. Experi-
ments have shown that fiber and fiber characteristics can be manipulated
to improve filter performance. Manufacturers are producing fabrics
that filter at higher efficiencies and lower pressure drops than before.
Laboratory work both in-house and by contract is continuing to
determine parameters and interactions that most affect filtration per-
formance (dust, gas, fabric, process variations and baghouse operation).
Specific programs to be pursued over the next year include: development
of filtration theory for needle-punched fabircs, testing of filter bags
made with various special fibers, optimization of cleaning parameters,
investigation and use of electrostatic phenomena, high-velocity filtra-
tion and continued characterization of existing industrial filter-
systems.
6.2.3.3 Wet scrubbing--In a wet scrubber, both gaseous and particulate
pollutants are transferred from the gas to a scrubbing liquid. Avail-
ability of a wide variety of scrubber types and scrubbing liquids gives
wet scrubbers potential for satisfactorily removing many pollutants.
(Details of many gaseous scrubbing systems are described elsewhere in
this Report: See Section 6.3.1.)
Results of a wet scrubber system study have been published as a
handbook. It is an authoritative and comprehensive reference source on
wet scrubber technology. A system study of wet-scrubber entrainment
separators is underway. The results of this study will also be published
as a handbook. The first section of the handbook was scheduled for publi-
cation in late 1974.
445
-------
The Control Systems Laboratory is conducting field tests of indus-
trial wet-scrubber systems to better define fine-particle performance
and economics of conventional scrubber systems. The data from these
tests will also be used to improve and develop scrubber design equations.
A feasibility study of using condensation effects to improve fine-
particle collection has been completed. Based on results of this study,
CSL is developing a new class of scrubber systems called Flux Force/
Condensation (FF/C) scrubbers for collecting fine particles. A pilot
scale demonstration of FF/C scrubbers is under way.
Study of liquid and energy use in conventional scrubbers is also
sponsored by CSL. Results of the first phase of this work have already
been published.
6.2.3.4 Novel particulate collection devices--All known novel devices
and new collection concepts are being reviewed by EPA. Those with poten-
tial for fine-particle collection are being evaluated for performance
and related cost. Systems showing promise of high-efficiency fine-
particle collection at reasonable cost and meriting further development
will, if necessary, be further supported by CSL.
Evaluation of four novel devices has been completed. The Lone Star
Steel Company Steam-Hydro scrubber installed on an open hearth furnace
had 99.9% collection efficiency, the best efficiency of the devices
tested. This scrubber has greatest potential where waste heat is avail-
able for steam generation since energy requirements range from 8250 to
12,750 Btu/1000 scf.
The Aeronetics two-phase scrubber has similar power requirements.
Tests on an Aeronetics scrubber installed on a ferro-alloy plant gave
collection efficiencies of 95.1 to 96.7%.
446
-------
The RP Industries Dynactor scrubber was tested in the laboratory
using re-suspended fly ash and iron oxide. It did not give high effi-
ciency collection of submicron particulate, but in applications that a
venturi scrubber might be suitable for, the Dynactor scrubber may be
an economical alternative.
The Braxton Corporation sonic agglomerator was tested in the labora-
tory using re-suspended cupola dust and fly ash. The collection effi-
ciency with this device was not adequate to justify further EPA
development of sonic agglomeration.
Test programs are now under way to evaluate four additional novel
devices, i.e., the Purity Pentapure Impinger, the Entoleter Centrifield
Scrubber, the Johns-Manville CHEAP Filter and the Rexnord Granular Bed
Filter. Annual evaluations of four to six devices are expected through
FY 80. Laboratory pilot-scale development of one-to-three novel collec-
tion systems will be initiated in FY 75.
A feasibility study on use of charged droplets to collect fine
particulate matter has been completed. Research in critical areas of
charged-droplet technology, identified in the feasibility study has also
been completed by M.I.T. Problems related to scale-up of the charged-
droplet scrubber and eventual commercialization are to be solved under
a program with TRW. Pilot demonstration of the charged droplet scrubber
is scheduled for completion in 1975.
Feasibility of collection of charged particles in fiber beds has
been demonstrated by Battelle Northwest. They will continue with efforts
to model the collection mechanism and scale-up the device.
447
-------
Other new concepts using electrostatics are being assessed. The
University of Illinois is studying the electric curtain and Carnegie-
Msllon University is studying electrostatic effects in fabric filtration,
both under grants.
The feasibility of foam scrubbing to collect fine particles will be
assessed under a contract with Monsanto Research Corporation. Demonstra-
tion of these concepts is expected in 1976.
448
-------
6.3 REFERENCES
1. OAWP Data Files of Nationwide Emissions. 1971.
2. Contribution of Source Categories to Man-Made Ambient Fine Particu-
lates. T. Ripberger: Internal Report. Industrial Environmental
Research Laboratory. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. October 26, 1974.
3. Sulfur Oxide Throwaway Sludge Evaluation Panel Vol. 1: Executive
Summary, Unpublished EPA Report, p. 18. September, 1974.
4. Ponder, W. H. Status of FGD Technology for Power Plant Pollution
Control. Thermal Power Conference, Washington State University,
p. 12. October, 1974.
449
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-600/1-76-023
4. TITLE AND SUBTITLE
SCIENTIFIC AND TECHNICAL DATA BASE FOR CRITERIA AND
HAZARDOUS POLLUTANTS - 1975 ERC/RTP Review
3. RECIPIENT'S ACCESSIOf*NO.
5. REPORT DATE
January 1976
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Health Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
10. PROGRAM ELEMENT NO.
1AA601
11. CONTRACT/GRANT NO.
27711
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Yearly Review 1976
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report is a review and evaluation of the current knowledge of the criteria and
hazardous pollutants: oxides of nitrogen; carbon monoxide; hydrocarbons; photo-
chemical oxidants; suspended particulates; sulfur oxides; asbestos; beryllium; and
mercury. Primary emphasis is on results achieved from the ERC/RTP research program
since publication of the criteria documents, which were used as the data bases for
the National Ambient Air Quality Standards. This document is the first in a series
of annual reviews to be conducted by the Environmental Research Center in the
Research Triangle Park, U.C. on the results of on-going research.
The document attempts to assess the adverse effects of the criteria and hazardous
pollutants on health, vegetation, climate, visibility, and materials. Information
on instrumentation, measurement, pollutant characterization, transformation and
transport, and control technology for the specific pollutants is also given.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution
Chemical Analysis
Abatement
Distribution
Lnviron. Biology
Oxides of Nitrogen
I'l. Ol^TRIHUl 11'N SI AftMt NT
Release Unlimited
Carbon Monoxide
Hydrocarbons
Photochem. Oxidant;
Particulates
Sulfur Oxides
Asbestos
Beryl!iun
Mercury
b.IDENTIFIERS/OPEN ENDED TERMS
Environmental Pollution
Air Pollution
COS AT i Field/Group
13B
06A
13B
04B
0GB
07B
07C
07C
07D
07D
07B
07B
07B
19. SECURITY CLASS (fins Report/
Unclassified
21 NO. OF PAGES
449
20. SECURITY CLASS (Tliispafc)
Unclassified
22. PRICE
EPA Form 22JO-' (9-73)
------- |