EVALUATION OF
COMMUNITY ODOR EXPOSURE
?port of
re1
A Symposium sponsored by the
ENVIRONMENTAL PROTECTION AGENCY
at
ARTHUR D. LITTLE, INC.
CAMBRIDGE, MASSACHUSETTS
APRIL 26-29, 1971
-------
EVALUATION OF COMMUNITY ODOR EXPOSURE
Report of
A Symposium Sponsored by
ENVIRONMENTAL PROTECTION AGENCY
at
Arthur D. Little, Inc.
Cambridge, Massachusetts
April 26~29, 1971
Edited by
David A. Kendall
Thomas Lindvall, M.D.
"
-------
TABLE OF CONTENTS
Preface
Pllrticipants
1.
Introduction
2.
Human Reactions to Odors
2.1 Adverse
2.2 Human Effects of Unknown
2.3 The Sources of Variation
Importance
in Human Reactions to Odor
3.
Data on Adverse Human Reactions to Odor
3.1 Disease States
3.2 Annoyance Reactions
3.3 Social and Economic Effects of Odor Exposure
3.4 Laboratory Experimental Results with Unknown Effects on
Human Populations
3.5 Interference by Community Odor with Odor-Dependent
Reactions
4.
Empirical Data on Odor Exposure from Different Sources
4.1 Chemical and Physical
4.1.1 Description of the Source
4.1.2 Characterization of Emissions
4.1.3 Analytical Approaches
4.1.4 Analysis Before and After Institution of
Control Methods
4.1.5 Data in the. Ambient Air
4.2
Sensory Analysis
4.2.1 Introduction
4.2.2 Laboratory Studies
4.2.3 Empirical Data at the Source
4.2.4 Data in the Ambient Air
5.
Dose-Response Relationships
6.
Interaction Between Odor Sources
Conditions
6.1 Odorant Interactions
6.2 Environmental Interactions
6.3 Psychological Interactions
and Other Environmental
7.
Temporal Patterns
7.1 Meteorological Considerations
7.2 Sensory Considerations
11
Page
v
vi
1
3
3
3
3
4
4
4
5
6
6
7
7
10
15
17
17
17
17
19
19
20
-------
TABLE OF CONTENTS (Cont'd.)
8.
Recommendations for Basic and Applied Research
21
23
9.
Scientific Basis for Performance Standards
10.
25
Critical Questions
Bibliography
27
iii
-------
LIST OF TABLES
No.
1
Results of a Field Experiment to Test a Method to Predict
Odor Frequencies with the Aid of Sensory Analysis at the
Source and Meteorological Dispersion Ca1cu1~tions
2
Areal Extent of Pe~ceptib1e Odor--Two-Day Technical Field
Investigations
iv
Page
12
14
-------
PREFACE
As a result of the Clean Air Act of 1970, the Environmental Pro-
tection Agency (EPA) of the U.S. Government is committed to a program
of improving the quality of the air from the standpoint of odorous air
pollutants. Both in the U.S.A. and in Sweden research studies have
been carried. out with the objective of defining the dose-response re-
lationships of odorous air pollution. There is particular concern for
the Evaluation of Community Odor ,Exposure.
In June 1970, a symposium sponsored by the Karo1inska Institute
took place in Stockholm to discuss Methods for Measuring and Evaluating
Odorous Air Pollutants at the Source and in the Ambient Air. Prior to
that meeting, the EPA (then the National Air Pollution Control Adminis-
tration) had undertaken to sponsor a complementary symposium in 1971.
Arthur D. Little, Inc., was given the responsibility for organizing
this symposium with the cooperation of the chairman and vice chairmen
from the Karo1inska meeting (Drs. E. R. Hendrickson, Lars Friberg, and
John Goldsmith). Mr. Jerry Romanovsky represented the EPA during the
initial phases and was assisted by Mr. Richard Dickerson. Mr. David
Kendall was in charge of the program for Arthur D. Little.
The meeting was opened with a brief statement of welcome from
Dr. Howard O. McMahon, President of Arthur D. Little, Inc., and an
introductory statement from Dr. John T. Middleton, Acting Commissioner
of the Air Pollution Control Office, EPA. Dr. Lars Friberg was elected
chairman of the conference with Dr. Thomas Lindvall and Mr. Kendall to
act as reporting secretaries. Drs. Trygg Engen, John R. Goldsmith
and E. R. Hendrickson were appointed as panel moderators and group
leaders for the working sessions. A list of. participants in the pro-
gram is included with this report. This report is a result of the
discussions during the meeting and is based on the prepared working
papers. It represents as much as possible the consensus opinion of the
participants.
v
-------
LIST OF PARTICIPANTS
Prof. Dr. Helmut Altner
University of Regensburg,
West Germany
Dr. E. R. Hendrickson
Environmental Engineering, Inc.
Kent Anger
EPA-APCO
Dr. Ulf Hagstrom
The University of Uppsala, Sweden
Dr. Delbert S. Barth
EPA-APCO
Mrs. Ulf Hagstrom
The University of Uppsala, Sweden
Miss Margaret Deane.
California Department of
Public Health
Dr. Robert Horton
EPA-APCO
Dr. Kenneth D. Johnson
Manufacturing Chemists Association
Ido deGroot, M.P.H.
University of Cincinnati
Dr. Erland Jonsson
University of Stockholm, Sweden
Richard Dickerson
EPA-APCO
David A. Kendall
Arthur D. Little, Inc.
Dr. Andrew Dravnieks
lIT Research Institute
Gregory Leonardos
Arthur D. Little, Inc.
Dr. Trygg Engen.
Brown University
Dr. Philip L. Levins
Arthur D. Little, Inc.
R. David Flesh
Copley International Corporation
Thomas Lindvall, M.D.
The Karolinska Institute
Miss Elizabeth Force
NAS-NRC
Karl J. Springer
Southwest Research Institute
Lars Friberg, M.D.
The Karolinska Institute
Dr. Isaiah Gellman
National Council of the
Industry for Air and
Improvement, Inc.
Takeo Suzuki, M.D.
The Institute of Public Health,
Tokyo, Japan
Paper
Stream
Dr. Amos Turk
City College of New York
John R. Goldsmith, M.D.
California Department of
Public Health
vi
-------
EVALUATION OF COMMUNITY ODOR EXPOSURE
1.
INTRODUCTION
Because of growing concern about living standards, annoyance
reactions resulting from exposure to odors have become of increasing
importance to advanced countries. The Air Pollution Control Office (APCO)
of the U.S. Environmental Protection Agency (EPA) has explicitly stated
that airborne odorous substances constitute a threat to the quality of
human life at many locations and considers it imperative that they be
brought under control under the provisions of the Clean Air Act. How-
ever, before an optimum control strategy can be achieved, much more
specific information on dose and response needs to be developed.
In June 1970, the Third Karo1inska Institute Symposium on Environ-
mental Health (Stockholm) contributed valuable information, with primary
emphasis on methodology, and stressed various aspects of the measurement
and evaluation of doses. A second conference, held at Arthur D. Little,
Inc., Cambridge, Massachusetts, in April 1971, was sponsored by the
EPA to deal particularly with the evaluation of the effects of odorous
exposure in the community, as well as to review recent studies of dose,
response, and dose-response relationships. No attempt was made to
replicate the material presented in Stockholm. Therefore, the reader
is advised to consider this report as an extension of the Stockholm
report (35)* and to review that work.
Both while planning this meeting and currently, the Air Pollution
Control Office has been seeking answers to such broad basic questions
(3) as the following:
(1)
What are th~ relative contributions to U.S. odor problems by
motor vehicles, aircraft, and the most important stationary
sources?
(2)
What approximate degree of control will be required to abate
adverse effects in each of these source categories?
(3)
How should emissions standards be stated for each of these
source categories?
(4)
What reference measurement method(s) should be
emissions from each of these source categories
quality ambient air?
adapted for
to attain high
*For this and succeeding references, see the similarly numbered sources
listed in the Bibliography.
-------
It was obvious at an early stage of the Cambridge meeting that
these questions could be given only very limited answers. This report
does, however, hopefully present some clues to underlyi~g problems
and their potential solutions. In Section 10 of this report, the con-
ferees have presented formal responses to each of the above queries.
These, with the recommendations for research (Section 8) and on the
scientific basis for performance standards (Section 9), summarize the
consensus of the conferees based on the information presently available.
To present a comprehensive catalog of odor sources would have been
beyond the scope of this four-day conference. Even a casual review
would reveal that almost every major industrial operation has at some
time been accused of being an odor nuisance source although the rela-
tive importance of the various odor sources has not been established.
Generally considered especially important, because of the number of
complaints about them or for other reasons, are diesel exhaust, pulp
and paper industries, refineries, meat packing and rendering plants,
certain chemical processes, waste disposal, and animal refuse.
Moreover, a systematic analysis of all such sources would have
been impossible since, in most cases, factual information is virtually
lacking. Thus instead, emphasis was placed on odors from the pulp and
paper industry and diesel engines because most of the available infor-
mation is on these sources. Wherever possible, however, questions con-
cerning dose, response, and dose-response relationships relating to
other sources were considered.
At the Stockholm meeting, odor was defined as "The product of the
activation of the sense of smell, an olfactory experience," as opposed
to the odorous materials (odorants) which activate that sense. This
definition also was considered useful when discussing odors at the
Cambridge symposium. In addition, it was thought essential to make a
further distinction between odorous substances which primarily elicit
responses to odors as such and those odorous materials whose major
adverse effects on humans are not caused by odor response. Many odorous
substances (such as hydrogen cyanide and sulfur dioxide) produce irritant,
toxic, or other undesirable effects at concentrations below their sensory
thresholds. (For example, high ambient air concentrations of hydrogen
sulfide have acute and severe effects on the central nervous and respira-
tory systems.) They are already controlled for reasons other than ,odor.
In general, therefore, for the purposes of the symposium, odorous sub-
stances were defined as those materials whose most common adverse ef-
fect is the annoyance reaction caused by the odor itself. The adverse
effects and dose-response relationships caused by direct toxic action
of particular substances are and, of course, should be considered
separately under other procedures, such as those utilized in 'the Air
Quality Criteria Documents for sulfur oxides and oxidants.
-2-
-------
2.
HUMAN REACTIONS TO ODORS
2.1
Adverse Human Effects
The following types of adverse human reactions to odor in the am-
bient air are known or suspected "on the basis of either laboratory or
community studies:
.
Disease states, including either causation or aggravation of
disease;
.
Annoyance reactions, including action taken to abate perceived
annoyance; and
.
Social and economic reactions.
2.2
Human Effects of Unknown Importance
In addition, several effects of unknown importance have been ob-
served in humans:
.
Physiological responses in human populations observed in labora-
tory studies; and
.
Postulated interference with positive reactions to nonambient
odors, such as those associated with food, sexual behavior, and
avoidance of danger.
2.3
The Sources of Variation in Human Reactions to Odor
Variation in population response to odor exposure is caused by demo-
graphic variables (such as sex, age, marital status, income, and occu-
pation) and differences in attitudes toward the pollution source which
result in tolerance toward the odors or anxiety concerning the effects
on health and property. Annoyance at odor exposure is reported more
frequently among those with a propensity toward neurosis, sensitivity
to aircraft noise, and displeasure with other aspects of community
life. Also, persons who fail to detect or be annoyed by odor could
have deficiencies related to clinical conditions, such as upper respira-
tory infections, as indicated in the Karo1inska report (35). The con-
tribution of these "extraneous" factors is probably least when the ex-
posure is very strong (or so weak as to produce no reactions at all),
but their effect on dose-response relationships must be taken into ac-
count in any report on exposure-reaction relationships.
-3-
-------
3.
DATA ON ADVERSE HUMAN REACTIONS TO ODOR
3.1
Disease States
The evidence presented at this conference could not demonstrate
a relationship between odorous air pollution and the increased incidence
of disease states. The work cited below may be' suggestive of some re-
lationship; however, questions have been indicated in each case.
The Eureka study (29) suggested the association of headache, sinus-
itis, and respiratory conditions with odor exposure. However, according
to the Karolinska report, the study design was inappropriate to test this
possibility.
An Anderson, California survey (7) also showed that the reported
frequency of chronic respiratory, headache, and other symptoms was
highest in the population living in the area with the greatest ,exposure
to odor. However, the study did not define a disease state, and the
reported symptoms could have been related to other atmospheric pollu-
tants or other characteristics of the study populations.
Hyatt has reported that pulmonary emphysema patients have been
found to react to exposure to perfumes. (25) Increased frequency of
asthma attacks in susceptible individuals has also been described as
a result of odor or odorant exposure. (11, 21, 24, 52) However, these
reports were not reviewed d~ring this conference. Thus, although none
of the data presented at Cambridge demonstrate a correlation between
disease states and odorous air pollution, it appears possible that
future research might show an association between odorous a~r pollution
and disease.
3.2
Annoyance Reactions
Although their possible long-term medical implications are not
known, annoyance reactions have been documented in surveys of communi-
ties exposed to odor from pulp mills, manure, oil refineries, and motor
vehicle exhaust. (17, 18, 44) Studies near Swedish pulp mills have
elicited annoyance reports from 12% to 62% of the population, depending
upon distance from the mill. Similar frequencies have been observed
in Eureka, California, and in Clarkston, Washington. (33) Annoyance
reactions due to air pollution have also been studied in urban areas
of the United States, but the distinction between reactions due to
odors and those due to other types of pollution has not always been
clearly defined. (26)
Annoyance reactions are not easily evaluated. For example, one
Swedish study found that the prevalence of annoyance reactions exceeded
20% near an older refinery while fewer than 5% reported annoyance near
several modern refineries. (28) However, it is not known to what
-4-
-------
extent the lower annoyance reaction ratio was due to reduced emission
rates and to what extent to generally more aesthetic plant appearance.
Another form of reaction is petitions and spontaneous complaints.
During 1969, the Japanese government received 40,000 petitions con-
cerning a variety of environmental problems; of these, 8,000 were for
odor, and a slightly smaller number for other forms of air pollution. (50)
(This compares with nearly 18,000 petition complaints received for
noise and vibration.)
However, while often the first sign of an important odor problem,
petitions and spontaneous complaints are in some countries considered
unreliable indices of the importance of the problem. In studies in the
vicinity of a Swedish pulp mill, only about 50% of those persons who
signed petitions reported annoyance in a subsequent survey. (28)
It has become obvious that, because of the influence of factors
unrelated to dose, sociological studies may not yield data which remain
valid over long periods of time. Community attitudes towards the source
of emission seems to be particularly important. In a Swedish field ex-
periment (42), annoyance reaction frequency caused by aircraft noise
was about 50% less in a group preconditioned to have a positive atti-
tude towards the source than in an untreated group. Springer's studies
(43) have shown that, for the same exposure of diesel exhaust odor,
differences in reaction occur not only between residents of different
communities, but also over time. Attitudes towards protection against
adverse effects from the environment may change considerably within the
next 10-20 years. Odor exposures considered acceptable today may thus
become unacceptable in the near future.
Sociologic survey methods thus need to be evaluated to determine
the effects of variations in procedures as well as the comparability
of annoyance responses measured over time. The impacts of different
sampling densities on spontaneous social contacts within the community,
and thus on survey results, should also be studied. Regional annoyance
surveys have already been conducted to study the effects both of other
air pollutants and of noise. The results of community odor annoyance
surveys, physiologic experiments, health surveys, and systematic ob-
servations of behavior should be correlated and compared to determine
the possibility of interaction and common factors.
3.3
Social and Economic Effects of Odor Exposure
It is generally presumed, on the basis of fragmentary information,
that odor field is a factor in determining the social groupings within
a community and the statuses of different communities. (8) However,
there is no conclusive supporting data. Except for differences in
property values, measurements of the social costs of odors have not
been attempted.
-5-
-------
Economic theory suggests that property values may be explained
by neighborhood, occupant, and property characteristics. To the extent
that odors are perceived as neighborhood characteristics, and to the
extent that they are considered objectionable by both buyers and sellers
of property, their presence is negatively capitalized into the value of
lands and of improvements thereon. The results of studies by Ridker
and Henning (38, 39) and Flesh (14) tend to support these premises.
Annoyance surveys have also included measurement of behavior re-
lated to social effects. For example, the Eureka questionnaire (29)
includes questions about interference with daily activities, complaints
to authorities, and various forms of individual action. No validation
has been carried out for these measurements.
3.4
Laboratory Experimental Results with Unknown Effects on Human
Populations
Studies with human subjects using two odorants, amyl acetate and
tetrahydrothiophene, showed consistent changes in respiration rate,
heart rate, and galvanic skin resistance and evidence of changes in
blood flow. (30) Evidence from neurophysiological investigations in
animals indicates that excitation patterns are elicited by the inter-
action of odorous substances with the olfactory receptors and that
these patterns are conveyed to the centers within the central nervous
system from which various physiologic effects as well as psychological
effects can be mediated. (1)
3.5
Interference by Community Odor with Odor-Dependent Reactions
Odor in the ambient air may interfere with the positive effects
of other odors, such as warnings against fire and unwholesome food or
water. For example, mer cap tans , among other agents, are added to
natural and liquefied gases in order to ensure awareness of gas leaks.
-6-
-------
4.
EMPIRICAL DATA ON ODOR EXPOSURE FROM DIFFERENT SOURCES
4.1
Chemical and Physical
Various control agencies, research organizations, and industrial
associations have collected data on odorant emissions at different
sources by physico-chemical analyses. The Air Pollution Control Office,
EPA, probably has the most extensive catalog of data on odorant emis-
sions from the wood pulping industry (20), refineries, coffee roasting,
sulfuric acid plants, nitric acid .plants, diesel engines, and incinera-
tors. State air pollution control agencies, which require submission
of emission data for permit .applications, also have relevant information
on file.
Until recently, techniques for physico-chemical evaluation of
odorants in the ambient air were scarce and primitive. Advances in
trace analysis techniques have now made possible measurement of near-
threshold levels of some odorants in the ambient air.
It is believed, however, that source and emission characterizations,
which would assist in relating dose to response, would provide more
valuable information than simple listings of odorant emissions from
various sources. These descriptions should provide guidance in evaluat-
ing the usefulness of existing data, in acquiring new data, and in de-
signing models of transport phenomena and for odorant source monitor-
ing and abatement strategies.
4.1.1
Description of the Source
Adequate description of both fixed and mobile odorant sources re-
quires attention to details possibly ignored in investigation of other
kinds of air pollution because odor nuisances require only very brief
exposure times and very low concentrations (such as trace impurities)
to evoke adverse reactions. Moreover, there is strong mediation of
receptor reaction to the odor stimulus by attitudes developed or modi-
fied by other factors.
In description of an odor source, the particulars on the points of
emission must be detailed. (27) Confined sources--those emitted from
a specific duct or stack--obviously must be examined with care. The
contribution of unconfined sources of odor, whether from such localized
points as leakages around valve stems, or from such broad and ill-defined
areas as the contaminated soil around tank farms, must also be evaluated.
Otherwise, major expenditures on stack control may result in little or
no reduction in odor nuisance.
Because of the very short time scale of responses to odors, even
the relatively infrequent occurrence of eddies that carry concentrated
stack plumes to ground levels can produce major adverse community
reactions. Thus, the investigator must be alert to aerodynamic factors.
":'7-
-------
Neither manual nor instrumental monitoring methods usually have a time-
resolution capability adequate to detect brief transients reliably.
In addition to the physical features of the source facility,
plant operations and processes must be appropriately described, par-
ticularly factors leading to variability in operating conditions,
such as annual cycles of product scheduling and. the short-term transients
associated with batch cycles. Some wide fluctuations may not be con-
trollable by the operator. Examples are the qualitative and quantita-
tive changes in diesel engine exhausts as a function of engine speed
and load, sulfur recovery in petroleum refineries, and municipal waste
incineration. In the latter cases, the problems created by storing
wastes to enable physical control of feed rates may be far worse than
those that would be generated by operating disposal facilities at rates
higher than their designed capacities.
The extent and nature of interactions between a particular plant's
emissions and those from other community sources is also pertinent to
evaluation of its contribution to a community problem and to determina-
tion of the degree to which control of its emissions would ameliorate
odor nuisances.
4.1.2
Characterization of Emissions
Although odorants may be dispersed as gases or condensed as parti-
culate matter, equal distribution of odorous matter between gases. and
particulates is probably uncommon in hot effluent streams because the
partition coefficient decreases with temperature rise. However, some
odorants do concentrate in a condensed phase. An example is the solu-
tion of phenol in water droplets. In many cases, therefore, gas phase
sampling for odorants is valid, but attention must be given to possible
condensation or adsorption of odorous matter on particulates.
Odorants may appear to be the same although they differ in con-
centration. However, Stone found that, with five odors, two samples of
the same odorant were just noticeably different in perceived odor
intensity if their concentrations differed by at least 20% and in many
cases by at least 50%. (49) Therefore, analytical precision may not be
essential to the instrumental technique, but sensitivity, specificity,
and accuracy must be adequate. Moreover, since the different substances
in a given effluent source will differ very widely in their odor in-
tensities and qualities (53), no qualitative chemical identification
of the components of an odor source can be considered odor-relevant
unless demonstrated by sensory methods.
4.1.3
Analytical Approaches
Complete chemical analysis may be difficult for complex mixtures,
and analytical effort might be conserved by assaying only those com-
ponents that contributed to the perceived odor, a class of components
-8-
. -
-------
that contributed to the perceived odor, or a component or class of
components that served as an indicator of odor, even though odorless.
The assay of the specific odorant compounds is the most direct and,
therefore, the most reliable of these alternatives, but is usually the
most difficult because many components may be involved. Caution must
be exercised, therefore, to avoid false assignments of odor to irrelevant
chemical components by the use of sensory methods.
The second choice, the assay of a class of odorous components in
a given source, presents much less difficulty in analysis, but is an
indirect and, therefore, inherently less accurate method. Probably
the most satisfactory analysis of this kind can be carried out when
the odor source is of the character produced by those sulfur compounds
in which sulfur occurs in its most reduced oxidation state, -2, as in
mercaptans. (48) Another set of odorants which may respond to such a
group assay is the highly odorous nitrogen compounds in which nitrogen
occurs in its most reduced oxidation state, -3, for example, various
amines (RNH2, R2NH, and R3N). Assays of total aldehydes and phenols
have been taken as a measure of odors, but the extreme differences in
the odor intensities and qualities of these classes would make group
assay extremely unreliable in the absence of rigorously demonstrated
correlations.
The third possible choice, the assay of an odor indicator which is
not necessarily odorous in itself, is apt to be the least difficult,
but also the least valid of the analytical choices. When the odor
source is derived from a process that involves oxidation, investiga-
tors have sometimes sought carbon monoxide, but both theoretical con-
siderations and experimental findings militate against its use as an
odor indicator. Another possible indicator, which is particularly easy
to monitor, is total organic matter. An odor indicator for readily
oxidizable odorants which merits some consideration is the reduction
of potassium permanganate. (37) Such an assay might be an appropriate
indication, but would not respond to more stable odorants, such as
esters. Moreover, non-odorous oxidizable components could interfere
and produce spurious results. Any relationship between such an assay
and the odor of the source being examined must be demonstrated in the
most rigorous manner."
4.1.4
Analysis Before and After Institution of Control Methods
Odor control is sometimes effected by a change that produces a
uniform decrease in quantity or concentration of all of the components
of the odor source. Examples of such odor control methods currently
in use are partial shutdown of operations, adsorption by activated
carbon, and incineration. In such instances, analytical methods may
need only to be modified to take such quantitative changes into ac-
count. However, various new odorants, or new ratios of odorant con-
centrations may be produced by selective adsorption, partial oxidation,
or other processes incidental to the odor control effort.
...9-
-------
Other control methods, such as process modification, the use of
ozone, chlorine, potassium permanganate, or hypochlorites, or the
application of nonreactive agents, such as odor maskants, do not pur-
port to be nonselective. In such cases, the analytical problem after
the control method is instituted must be regarded as a new situation,
and all of the above considerations as to what to assay must be reeval-
uated.
4.1.5
Data in the Ambient Air
Except for some compounds studied for other than odorant reasons,
relatively few valid observations of odorant concentration in the am-
bient air have been made until recently. Limited physico-chemical
data are, however, available for kraft pulp mills (41), rendering
plants and synthetic resin plants. (9, 10)
4.2
Sensory Analysis*
4.2.1
Introduction
In general, sensory analysis is concerned with the hedonic charac-
teristics, quality, and intensity of odorants. Intensity involves ,so-
called thresholds,detectability indices, and suprathreshold intensity
and psycho-physical scaling methodology. For purposes of illustration,
this report concentrates on demonstrating use of sensory analysis.
methods related to pulp mills, diesel exhaust, and other odorant sources
with which members of this symposium were particularly familiar.
4.2.2
Laboratory Studies
A number of studies by various laboratories on sensory analysis
of single odorants have been useful in defining minimal detection
levels as well as suprathreshold response of the human nose to a
variety of odorous materials. Response to odor also involves quality
and hedonics as well as the complex interactions of all three factors.
Further research in each area is required to develop laboratory models
for possible extrapolation to the air pollution situation.
4.2.3
Empirical Data at the Source
Pulp Mills. Sensory data on odor problems associated with pulp
mills are scanty and are only available from a few laboratories. .(31,
41) Investigations in Sweden include mapping out odor generating
processes, study of various control measures, prediction of odor in-
tensity, and forecast of odor distribution and frequency. To meet
these needs, mobile odor laboratories have been constructed for field
*The definitions of sensory analysis and recommended strategies for
sensory odor measurements presented at the Third Karolinska Institute
Symposium form the basis on which related findings at this conference
were built.
-10-
-------
studies under controlled and standardized conditions. Use of these
air-conditioned, odor-free units has made it possible to limit varia-
tions resulting from the ambient background, and their dynamic dosing
systems have minimized losses due to adsorption and other factors.
These standardized experimental conditions are not, however, necessarily
comparable to normal conditions of exposure in the ambient air. Exten-
sive studies on the reliability of the entire experimental procedure
have been undertaken. (40) .
Diesel Exhaust. Few studies have been made on odorous exhaust
products from gasoli~e-powered cars and aircraft or turbojet-powered
aircraft, but a numbe~ of ways of measuring and expressing odor ~esponse
to diesel exhaust have been described in published literature. (References
2, 5, 6, 19, 40, 44, 46, 47, and 54 describe a few.) A Swedish study
on gas engines using fuel comparable with that used in the United'
States shows that this exhaust is detectable at the same concentrations
as diesel exhaust. Unfortunately, no data have yet been obtained on
the qualitative characteristics of the odor or its objectionability. (16)
Though threshold methods gained popularity for a while in diesel
odor research, emphasis now is on the evaluation of 'odor intensity
and quality when diluted to the suprathreshold levels more typically
encountered in urban areas. The U.S. Public Health Service Quality-
Intensity kit developed for use in diesel odor research in the United
States requires a trained panel of observers, housed in a special odor
representation facility, to relate the particular exhaust odor to the
kit's standard diesel-like odors. (15, 45, 51)
Other. There is little available data on odor exposure at the
source for other possible odor contributors to the environment. Studies
using the ASTM syringe dilution technique as a measure of odor control
have been conducted prior to and after incineration on stack samples
from a variety of industrial odor sources, such as glass fiber curing,
wire enameling, and auto paint baking ovens and a pulp mill recovery
system. (4)
4.2.4
Data in the Ambient Air
Pulp Mills. The odor coverage in the neighborhood of Swedish
pulp mills is calculated by a prognosis method which utilizes known
source strength of the gas, expected distributio~ of the gas mass over
the investigation area, and observations of its dilution at ground
level at different distances from the point source. The source
strength has until now been evaluated by absolute odor threshold
methods (with some measures possibly obtained by scaling techniques)
to determine the frequency with which odors will be discernible at
various distances from the plants. (22)
. -11-
-------
Although odor frequencies may be predicted by dispersion calcu-
lations basically like those used for other air pollutants, the
sampling time is more critical since olfactory reaction times are fairly
fast (at the most half a minute). Studies with the absolute odor
threshold dispersion model with very short sampling times show that
the concentration of odorous gas may exceed the odor threshold several
times during an hour even though the hourly mean. concentration is below
the threshold. A method based on these principles is currently being
used in Sweden to predict ambient air odor frequencies for pulp mills
and other odorous sources. (23) Table 1 shows typical results obtained
with a panel of trained observers. Near the source, the result of the
theoretical computations is in very good agreement with the observa-
tions, but with increasing distance there is an increasing discrepancy.
This may be due to an inadequate dispersion model, psychological factors,
chemical reactions, and/or physical separations within the odorous
plume, or it may even be explained as a pure statistical effect. (22,
31) (See also Section 7.1 for discussion of meteorological factors.)
TABLE 1
RESULTS OF A FIELD EXPERIMENT TO TEST A METHOD TO PREDICT
ODOR FREQUENCIES WITH THE AID OF SENSORY ANALYSIS AT THE
SOURCE AND METEOROLOGICAL DISPERSION CALCULATIONS
Distance from the source 2 km 5 kIn 10 kIn 20 kIn
Total number of observations 6426 7490 5528 6976
Numbers of positive observations 696 736 470 360
Observed odor frequency (%)* 10.8 9.8 8.5 5.1
Predicted odor frequency (%)* 9.1 5.7 3.2 1.7
*"Predicted odor frequency" (number of observations of odor expected
per hour) is determined on the basis of dispersion model calculations.
"Observed odor frequency" is based on actual observations.
Source:
U. Hagstrom, "A Method for Predicting Odor Frequencies from a
Point Source," manuscript submitted to Atmos. Environment, 1971.
Moreover, the prognosis of the expected odor coverage, expressed as a
mean frequency value over a considerable length of time, does not neces-
sarily reflect the subjective experience of the population. One reason
may be that, in addition to frequency and duration of odor exposure,
contextual factors, such as adaptation and expectation, affect per-
ception of odors. (32)
-12-
-------
Systematic determination of occurrence and magnitude of supra-
threshold odor has been reported for the vicinity of two pulp mills
in California. (7, 41) A panel of two to four "trained" individuals
visits predetermined locations at predetermined times and compares
odors with continuously variable dilutions which range to nondetect-
ability. Results are expressed in proportion of observations at
which odor is detected and the dilution ratio when odor is present.
Odor threshold determinations with methyl mercaptan are given to each
subject before and after each day's series of tests. Similar supr&-
threshold methods may be applicable to other sources, provided a charac-
teristic or single odorant may be chosen.
Diesel Exhaust. The extent to which exhaust is diluted behind and
beside a diesel-powered truck or bus before being experienced by the
public has been established experimentally. Addition of a tracer gas
to the exhaust of a vertical stack-equipped truck and a city bus with
a horizontal, below-the-bumper, exhaust pipe made it possible to com-
pare diluted concentrations in the exhaust with atmospheric concentra-
tions. These tracer gas studies and later work (6) indicate a seven-
fold increase in dilution of the vertical'stack in contrast to the
horizontal pipe. (Dosage was also further defined by a small panel
of observers in an open sedan.)
Recent pilot studies in Sweden indicate that signal detection
methodology seems to be a realistic alternative in dose description
when dealing with the contribution of diffuse sources to the ambient
air, particularly such generalized pollutant sources as traffic arteries,
where the question of frequency is less important. (32) The index of
detectability, measured for odor levels in the ambient air in one of the
main streets of Stockholm city during rush hours varied in an expected
fashion, from 1.0 to 3.0, in relation to variations in the traffic
load during rush hours. No significant correlation between odor index
and concentration of carbon monoxide was found.
Other Sources. Field studies utilizing a vapor dilution technique
(Scentometer) and a sensory evaluation technique ("trained" panelists--
reference standard as described by Turk and Mehlman) (55) on odor in
the ambient air from a variety of sources have been described. (36)
The extent of perceptible odors in seven American Cities (Table 2)
was observed according to type of industry or major product. According
to this study, refinery installation and chemical facilities were
found to be the most significant sources of odor emissions from the
viewpoint of range. Odorous areas in the selected cities were mapped
to indicate their extent and intensity.
Sensory evaluations were carried out by a panel in the Philadelphia
area on ambient odors from an oil refinery complex and adjacent meat
rendering plants. The panel, consisting of 18 women, was given in-
tensity training with reference standards prior to the observations.
Results of this field survey are expressed in terms of dispersion
-13-
-------
patterns for each observation period, utilizing two types of contours:
(1) the main intensities of the observations of the individual panels;
and (2) the percent of the total observations of each individual in
which odor was perceived. (36)
TABLE 2
AREAL EXTENT OF PERCEPTIBLE ODOR
TWO-DAY TECHNICAL FIELD INVESTIGATIONS
Summary of Observations in the Principal Cities
of Seven Metropolitan Areas
Type of Industry Number of Areal Extent (square miles)
or Major Product Observations Average Range
Refinery 3 10 1.5 - 23
Agricultural Chemical 1 6
General Chemical 3 5 0.6 - 8.4
Polluted Bay 1 4
Sewage Treatment 1 2.5
Granary 1 2
Electrical 1 1.7
Tall Oil 1 1.25
Meat Packing 1 1
Rendering 4. 0.8 0.2 - 1.3
Paint and Varnish 1 0.4
Tanning 1 0.1
Source: Engineering-Science, Inc. (36)
-14-
-------
5.
DOSE-RESPONSE RELATIONSHIPS
Dose is here defined as a concentration of odorant substance(s)
as measured by chemical, sensory, or other analyses; and response, as
adverse effects within an exposed population. Only a few points are
known on the dose-response curve for certain odors--for annoyance
around pulp mills and experimental response to diesel exhaust. As
noted above, in a Swedish survey, 12% to 62% of the population
(depending on distance from the plant, within a radius of about 10
to 15 miles) reported annoyance reactions. Odor threshold studies
were not carried out in this community at the same time, but studies
at another time indicate that odor strength, as measured by dilution
of stack gases, would correspond to a dilution factor of about 105
times above threshold.
In the Eureka studies (29), annoyance reaction and odor threshold
determinations were made in addition to chemical measurements in three
areas at different distances from the plant. The chemical measurements
did not show distinctive differences in odorant levels among the areas,
but the frequency of annoyance reactions and the frequency and magnitude
of suprathresho1d odor detected by the method used (Sanders) did
discriminate among the three areas with presumptively different exposure.
However, it is not possible to use these data for a quantitatively exact
statement of a dose-response relationship. (41)
Various levels of diesel odor, defined in terms of the U.S. Public
Health Service Quality-Intensity kit, were presented to over 5,000
quota-sampled Americans in a mobile odor testing laboratory during 1969
and 1970. Participants indicated their response via identification with
five cartoon figures that depicted different stages of odor objection-
ability from "pleasant" to "unbearable." This survey with controlled
odor exposure resulted in a type of dose-response curve that could be
used as a base for estimating what effects various levels of odor control
would have on relative annoyance as expressed by a large population. (19, 44)
The data now available from long-term community studies do not
show how much any adverse reaction within the population would change
if the exposure were doubled or halved. Moreover, epidemiological data
refer only to particular situations, and emitted gases can differ
widely in composition from one plant to another and from time to time.
It would seem possible to carry out epidemiologic studies which
could give more or less complete dose-response data for pulp mills or
any other well-defined point source. Then expected magnitude of changes
in response could be predicted for alterations of dose by certain factors
for that source. The dose could be evaluated either by chemical or
sensory analysis of stationary source emissions in combination with
meteorological dispersion models, or by direct chemical or sensory
analysis of the dose in the ambient air. Because of the difficulties
. -15-
-------
in obtaining valid measurements in the ambient air, the first approach
to dosage estimation seems to be the more promising one for guidance of
control strategies at present. The lack of valid dose-response data
does not prevent use of the best practicable means for the guidance of
certain air pollution control strategies while awaiting more exact
procedures.
-16-
-------
6.
INTERACTION BETWEEN ODOR SOURCES AND OTHER ENVIRONMENTAL CONDITIONS
Two important problems are the classic effect of mixing odorants
and related environmental conditions, particularly the impact of aes-
thetic factors.
6.1
Odorant Interactions
If one mixes odorants A and B, several alternative odor effects
are possible: (a) the odor intensity of the mixture may be perceptu-
ally stronger than the odor intensity of either component; (b) the
odors may cancel each other so that the mixture is odorless; (c) one
of the odors may mask the other, so that its odorant provides the
dominant quality of the perceived odor; (d) synergistic effects may
result in one of the odors being stronger in the mixture than by it-
self; and (e) if the odorants are mixed successively, the odor inten-
sity and quality of the second odor may be decreased or altered due to
cross-adaptation or conversely, it may be facilitated (that is, seem
stronger) through addition of the second odorant to the one already
present in the olfactory system. The importance of being able to pre-
dict such effects for different mixtures in different environments is
obvious. Unfortunately, almost no data are available on any of these
forms of interaction beyond the findings from a few debatable laboratory
studies. The problem of interaction thus represents a most important
area for future research. (12)
6.2
Environmental Interactions
The effect of other environmental situations on the perception of
odors is similarly an important research target. The potential effects
of temperature, humidity, atmospheric pressure, and individual and
social factors remain largely unexplored.
6.3
Psychological Interactions
The most important consideration might be the interaction of
learning or experience and odor perception. Psychologists have re-
peatedly stressed the extent to which prior bias, either for or against
an alleged odor source, can influence the emotional response to an odor
dosage. Thus any other aesthetic insult from the source, whether in
the form of visible emissions, noise, vibration, or glare, or even such
non-specific factors as disorderliness or distasteful plant architecture,
may largely negate any potential decrease in community annoyance from
reduction in odorant dosage, however, achieved.
--17-
-------
It has been established that the presence of visual cues, such as
smoke, may increase the frequency of reports of the perception of odor.
(13) It is not clear at this time whether such results can be explained
as evidence of biased responding or of keener odor perception because
of sensory input from another modality and thus increased attention and
awareness (similar to facilitatory effects). (56) In any case, a visible
emission, because it is viewed as an air pollution contribution, does
appear to be a particularly potent evoker of negative emotional response.
If a visual plume has been identified by the public as an odor source,
a major reduction in odorant emissions unaccompanied by a corresponding
reduction in plume visibility may fail to diminish the community resent-
ment which may still be expressed as an "odor" annoyance. Awareness of
these aesthetic factors by control agencies and by industrial manage-
ments can do much to minimize the frequency with which constructive
abatement programs fail to secure the community appreciation and support
that they deserve.
-18-
-------
7.
TEMPORAL PATTERNS
7.1
Meteorological Considerations
When discussing temporal patterns of odor in the surroundings of
a source, it is necessary to distinguish between point sources and area
sources -- the variation in odorous impact at a fixed observation point
being very much larger for a point source than for an area source (pro-
vided its output of odorous compounds is constant over time). Because
temporal variations in odorous impact are due to corresponding varia-
tions in meteorological factors it is, in principle, possible to pre-
dict the variations in odor if there is enough information available
regarding (1) the relevant dispersion mechanism and (2) the local dis-
persion climatology. Concerning the dispersion mechanism, an extensive
literature has developed, particularly over the last two decades. A
comprehensive summary of the present status of the art is given in
Meteorology and Atomic Energy, 1968 (34).
In typical temporal odorous patterns in the surroundings of a
point source, odors are experienced in episodes, usually about one to
several hours long. During such an episode, the odor may be continuous
if the odorous source strength is high enough, but it is more likely to
be discontinuous, which means short-term variations in the concentration
of odorous compounds around the odor threshold. The odor threshold may
be exceeded only during a minor portion of the time. Such rapid varia-
tions which are inherent in the atmospheric dispersion process are well
borne out by a field experiment in the surroundings of a pulp mill in
Sweden. (31) It is reasonable to believe that the perception of odor
for such an episode within a population is somehow related to the de-
tailed structure of the episode, but this has not been investigated so
far.
The feasibility of predicting the absolute frequency of perceptible
odor cycles in the surroundings of a point source on the basis of senso-
ry analysis of the chimney gases and dispersion calculations has been
demonstrated. (See Section 4.2.4.) The success of the calculations
depends largely on the -effect of local wind flow patterns on the fluc-
tuating plume model. A dynamic flow pattern model like one being
developed for Oslo, Norway, for air pollution in general might provide
a practicable solution for some cases. (23) The absolute perceptible
odor frequency concept may not be ideally related to the perception of
odor within a population living in the surroundings of a point source.
But if it is possible in the future to establish a description of
instantaneous changes in dose concentration, this could be included in
a dispersion model. Such a dose description could, for instance, be
in terms of the percentage of time in which the concentration exceeded
the odor threshold and could be generally experienced as "an odor
episode".
-19-
-------
7.2
Sensory Considerations
Perception of temporal patterns has up to now mainly considered
problems of temporal integration, adaptation, and recovery. Important
data are still lacking on how frequency and duration of odor exposure
are perceived in the ambient air. Thus field research, as well as
laboratory studies in this area, are strongly recommended. The deter-
mination of temporal patterns obviously requires sampling over time on
a systematic basis. In the Eureka and Anderson studies (7, 29) obser-
vations by olfactometry were made systematically over a period of
several weeks and at intervals of from one-half hour to one hour. In
both cases, measurements were limited to the hours of the day when
pollution was expected to be greatest on the basis of meteorological
and topographical variables and to several weeks of the year when
pollution was greatest.
-20-
-------
8.
RECOMMENDATIONS FOR BASIC AND APPLIED RESEARCH
The following recommendations related to the need for basic and
applied research were developed through discussions at the symposium:
8.1
Of the many effects postulated to result from odorous air pollution,
two--t~e perception or detection of odor and the degree of annoyance
. response as a result of this odor perception--cannot be refuted.
The major weakness of current knowledge is 'that the bulk of
systematically obtained information deals only with exposure from
two sources, pulp and paper plants and diesel vehicles, using a
standard odor measurement technique.
Odor perception of emissions from a variety of other sources, such
as petroleum refineries, rendering plants, petrochemical plants,
paint and varnish plants, and most importantly gasoline and aircraft
engines, should be evaluated. In addition, possibly concurrently,
methods should be developed to measure the degree of annoyan~e'pro-
duced by these various types of odor exposure. The difficulty of ac-
complishing this scientifically shouid be recognized, and such studies
should be carefully designed. '
8.2
As a potential alternative to measuring annoyance, physiological
changes should be studied to determine the dose-response relation-
ships associated with odor perception so that alternative criteria
may be developed as indicators of response to odors.
8.3
Further work should be carried out to explore the possible effects
of odor exposure in aggravating certain pathological effects,
particularly respiratory ailments.
8.4
A systematic comparison of methodological approaches to sensory
measurement should be carried out to develop a preferred method(s)
for measurement of odor in ambient air and at the source.
8.5
Recent developments in physico-chemical methods for analysis of
odorants at the source and in the ambient air have resulted in
greater confidence in their usefulness. Their application,
however, can only be successful to the extent that they can be
correlated with detectabi1ity or perception of odor and with
the resulting annoyance reaction. There is a need for development
of instruments of improved specifi~ity and sensitivity which can
be correlated with observations of odor detection and annoyance
reactions. In addition, it would be desirable to investigate
the possibility of developing nonspecific indicators of odor. It
would be helpful to determine whether there are acceptable indices
of odorant mixtures, such as the concentration of total reduced
sulfur which has shown some promise.
-21-
-------
8.6
It would be useful to define and elaborate the mechanisms of
odorant transport and emission interactions which occur in the
ambient air and result in changed odorant form or composition
and, thus, a different odor exposure from that experienced at.
the source.
8.7
Further work on the economic costs of odor pollution would be of
value in determining pollution control strategies. The cost of
control must be given serious consideration, and the cost of odor
pollution is not well documented.
-22-
-------
9.
SCIENTIFIC BASIS FOR PERFORMANCE STANDARDS
.
A firm scientific basis for establishing performance standards
requires that each odor problem is clearly defined in terms of degree
of annoyance produced, the size of the population exposed, frequency
and duration of the events, the extent and nature of emissions, and the
relationship of cost of control to its cost benefits. In order to .
establish emissions limits on a firm scientific basis, one must have
substantially more information on the specific quantities of odorants
and the types of odorant chemicals involved, as well as knowledge of
how much reduction in concentration is required to produce signiricant
changes in the degree of annoyance, in the number of people affected
within an exposed population, and the total size of the exposed popu-
lation. Even th~ugh it is not possible to establish standard controls
now for the whole range of emissions that produce significant annoyance
reactions, specific controls could be established for the most highly
odorous, objectionable, and persistent odorant chemicals which can be
detected at very low concentrations.
Data necessary for an adequate description of levels of odor in
ambient air can be obtained through the use of human sensors (many
untrained or a few trained individuals) to conduct consecutive obser-
vations of odor occurrence within a given area for a fixed period of
time. Possibly such panels could be aided by a simple dilution device,
such as a scentometer, which, if suitably calibrated, might serve as a
supplement to more rigorous sensory studies. Used alone, however, re-
sults from such devices lack the validity needed in odor research.
Generally a sophisticated study design is necessary to obtain data of
value from the viewpoint of public welfare.
Another approach is to interview the resident population to find
out their retrospective opinions concerning the average odor ~requency
and duration of a smell. One difficulty of this method is ensuring
that answers reflect actual experience during the period under study.
Also, the degree to which respondents may unconsciously exaggerate or
depreciate the frequency of exposure as a result of negative or posi-
tive attitudes is not-known.
For well-defined point sources, indirect evaluation of ambient air
dosage based on a predictive type of method can be useful. Sensory
intensity analyses at the source, supported by physico-chemical measure-
ments and calculations, may be used to predict the frequency with which
odors should be discernible in the ambient air at various distances
from the source. In the dose desecription step required before estab-
lishing dose-response relationships in a single-source case, predicted
odor frequency may be considered an adequate alternative to survey
methods and direct measurement in the ambient air.
For ambient air measurements of odor caused by surface sources,
signal detection methods are suggested to obtain an idea of the intensity
with which odors are experienced. Detectability indices of this kind
-23-
-------
do not measure perceived intensity above the barely perceptible level,
but do provide valuable information regarding source control and the
distribution of odorant emanations in the atmosphere.
Practical experience has also shown the value of modern psycho-
physical scaling, based on standard references suitable for the speci-
fic odor, for ambient air measurement of surface source odors. No
general odorant reference standard can be recommended for all types
of odor studies. In each case, the reference should be chosen on the
basis of a pilot project.
In technological evaluations, suprathreshold perceived intensity
and quality are of interest since they concern changes in perception
associated with reduction in the physical odor intensity. In studying
intensity of odor, one should preferably use modern psycho-physical
scaling methods.
-24-
-------
10.
CRITICAL QUESTIONS
In light of the proposed method of approaching the control of
odorous air pollutants, Dr. Barth submitted for the conferees' con-
sideration four critical questions. (See Section 1.) These were dis-
cussed by the working group in plenary session on Thursday, April 29,
and the consensus responses appear in the following paragraphs.
10.1
On the basis of data available to the conferees, it is
sible to develop the requested specific ranking of the
contributions to U.S. odor problems by motor vehicles,
and the most important stationary sources because:
not pos-
relative
aircraft
.
Considerable variations in relative odor contribution can be
expected in different local or regional jurisdictions. (In
large cities, transportation odors may be deemed the major
source of odorous emissions, while in a small mill town the
carbon disulfide associated with rayon manufacture may con-
stitute the overwhelming problem.)
.
Certain process operations, regardless of manufacturing
category, may constitute a problem where other factors, such
as the local topography and meteorological conditions, act
to augment the situation.
Thus, it would appear more practical to develop rank orderings
of importance of various sources for specific local or limited
regional areas.
However, to provide a rational basis for establishing such
priorities, APCa should promulgate guidelines' to assure uniform
evaluation. The weighting of such factors as the numbers of
people exposed, the frequency and duration of exposure, the
relative strength and severity of probable exposure, and the
objectionability or frequency of annoyance indicated by the
affected population should be considered in measuring the odor
contribution. Through application of such guidelines at local
or regional levels, it may be possible to integrate the resultant
local priorities to develop a national priority statement if this
is desired.
10.2
It is not possible to state categorically the degree of control
required for abatement for a given type of source since there
are considerable variations within groups, both in odorant con-
centration and complexity. Because of the nonlinear relation-
ship of odorant concentration to perceived odor intensity and
objectionability, it is probable that control measures will re-
quire measurement of reductions in exponential terms rather than
in terms of percentages.
-25-
-------
10.3
10.4
Ultimately statement of standards of performance or emission
standards for each source category will require an administrative
decision formulated on the basis of definitions resulting from the
best available technology. Ideally, standards for a maximum emis-
sion should be expressed in terms of odorant or odorant class
per unit volume for some percent of time (i.e., mg/m3 for
hrs/month). The maximum emission rate should be based on
'the dilution required to attain a perceived odor intensity de-
fined in terms of standard sensory measurement procedures.
Examples of such standards are available from Sweden (1, 44) and
similar principles have been followed in California.
Reference measurement methods, including present technology, were
reviewed by the Cambridge conference report, and the recommendations
summarized in Section 9 of this report represent the consensus
of the working groups on applicable methods. The needs for
further basic and applied research on odors where adequate infor-
mation is not available are discussed in Section 8.
-26-
-------
(10)
BIBLIOGRAPHY
(1)
A1tner, H. "Neurological Responses to Odorants." Unpublished paper
presented at the Conference on the Dose-Response Relationships
Affecting Human Reactions to Odorous Compounds, Cambridge, Massa-
chusetts, April 1971.
(2)
Barnes, G. J. "Relation of Lean Combustion Limits in Diesel Engines
to Exhaust Odor Intensity." SAE Paper No. 680445, Mid-Year Meeting,
Detroit, Michigan, May 20-24, 1968.
(3)
Barth, D. S. "Administrative Mechanisms Available for Control of
Odorous Compounds under the U.S. Clean Air Act as Amended." Un-
pub1ished,paper presented at the Conference on the Dose-Response
Relationships Affecting Human Reactions to Odorous Compounds,
Cambridge, Massachusetts, April 1971.
(4)
Benforado, D. M., Rotella, W. J., and Horton, D. L. "Development
of an Odor Panel for Evaluation of Odor Control Equipment,"
J.A.P.C.A., 1969.
(5)
Colucci, J. M., and Barnes, G. J. "Evaluation of Vehicle Exhaust
Gas Odor Intensity Using Natural Dilution." SAE Paper No. 700105,
Automotive Engineering Congress, Detroit, Michigan, January 12-16,
1970.
(6)
Colucci, J. M., and Barnes, G. J. "The Effect of Exhaust System
Geometry on Exhaust Dilution and Odor Intensity." SAE Paper No.
710219, Automotive Engineering Congress, Detroit, Michigan,
January 11-15, 1971. .
(7)
Deane, M., and Goldsmith, J. R. "Health Effects of Pulp Mill Odor
in Anderson, California." Unpublished paper presented at the
Conference on the Dose-Response Relationships Affecting Human
Reactions to Odorous Compounds, Cambridge, Massachusetts, April
1971.
(8)
deGroot, 1. "Reactions of Human Populations to Odor Exposures."
Unpublished position paper presented at the Conference on the
Dose-Response Relationships Affecting Human Reactions to Odorous
Compounds, Cambridge, Massachusettss, April 1971.
(9)
Dravnieks, A., et a1. "High Speed Collection of Organic Vapors
from the Atmospher~" to appear as a Technical Note in Environmental
Science and Technology, 1971.
Dravnieks, A., and O'Donnell, A. "Principles and Some Techniques
for High Resolution Headspace Analysis," to be published in
Journal of Food and Agricultural Chemistry, 1971.
-27-
-------
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
Engebretson, G. R. "Allergies and Odors Arising from Indoor
Environments," Amer. Jour. Pub1. H1th., 61:2, Feb. 1971, pp. 366-
375.
Engen, T.
presented
Affecting
chusetts,
"Method and Th~ory of Odor Preference." Unpublished paper
at the Conference on the Dose-Response Relationships
Human Reactions to Odorous Compounds, Cambridge, Massa-
April 1971.
Engen, T. The Effect of Expectation on Judgments of Odor, to be
published by the U.S. Public Health Service, 1971.
Flesh, R. D. "Property Value Differentials as a Measure of Economic
Costs Due to Odors."Unpub1ished paper presented at the Conference
on the Dose-Response Relationships Affecting Human Reactions to
Odorous Compounds, Cambridge, Massachusetts, April 1971.
Ford, H. S., Merrion, D. F., and Hames, R. J. "Reducing Hydro-
carbons and Odor in Diesel Exhaust by Fuel Injector Design."
SAE Paper No. 700734, Combined National Farm, Construction, and
Industrial Machinery and Powerp1ant Meetings, Milwaukee, Wisconsin,
September 14-17, 1970. .
Friberg, L. "Summary of Technical Problems." Unpublished paper
presented at the Conference on the Dose-Response Relationships
Affecting Human Reactions to Odorous Compounds, Cambridge, Massa-
chusetts, April 1971.
Friberg, L., Jonsson, E., and Ceder1of, R. "Studies of Sanitary
Nuisance Caused by Flue Gases from a Sulphate Cellulose Factory,"
Nord. Hyg. Tidskr., Vol. XLI, 1960.
Friberg, L., Jonsson, E., Rylander, R., and v.Ubisch, H. "Reactions
to Environmental Disturbance Factors Caused by Oil Refineries."
Unpublished paper of the Institute of Hygiene, Karo1inska Institute
Department of General Hygiene, National Institute of Public
Health, Stockholm; and Institute of Hygiene, Unea University,
Unea, Sweden (no date).
Hare, C. T., and Springer, K. J. "Public Response to Diesel Engine
Exhaust Odors." Final Report to the Environmental Protection Agency,
Air Pollution Control Office, Contract CPA 70-44, April, 1971.
Hendrickson, E. R., Roberson, J. E., and Koogler, J. B. Control of
Atmospheric Emissions in the Wood Pulping Industry (3 volumes).
NAPCA Contract 22-69-18, Gainesville, Florida, March 15, 1970.
Herbert, M., et a1. "Olfactory Precipitants of Bronchial Asthma,"
Jour. Psychosom.~es., 11, August 1967, pp. 195-202.
-28-
-------
(22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
(33)
Hogstrom, U. "A Method for Predicting Odour Frequencies from a
Point Source," manuscript submitted to Atmos. Environment, 1971.
Hogstrom, U. "Predicting Odor Frequencies by Dispersion Calcula-
tions." Unpublished paper presented at the Conference on the
Dose-Response Relationships Affecting Human Reactions to Odorous
Compounds, Cambridge, Massachusetts, April 1971
Horesh, A. J. "The Role of Odors and Vapors in Allergic Disease,"
Jour. Asthma Res. 4:2. pp. 125-136, Dec. 1966.
Hyatt, R. Personal communication to John R. Goldsmith.
Interstate Air Pollution Study, Phase II Project Report, Vo1s. I,
II, and IV. U.S. Dept. of Health, Education, and Welfare, Public
Health Service, Division of Air Pollution, May 1965 and May 1966.
Johnson, K. D. "Odorant Sources in Air Pollution." Unpublished
paper presented at the Conference on the Dose-Response Relation-
ships Affecting Human Reactions to Odorous Compounds, Cambridge,
Massachusetts, April 1971.
Jonsson, E. "On Annoyance Reactions Observed within Human Popula-
tions." Unpublished paper presented at the Conference on the Dose- \
Response Relationships Affecting Human Reactions to Odorous Com-
pounds, Cambridge, Massachusetts, April 1971.
Jonsson, E., Deane, M., and Sanders, G. "Community Reactions to
Odors from Pulp Mills--A Pilot Study in Eureka~ California."
Unpublished paper presented at the Third Karo1inska Institute
Symposium on Environmental Health, Stockholm, June 1970.
Kendall, D. A., and Jacobs, ~. P. "Physiological Responses to
Odorants." Unpublished paper presented at the Conference on the
Dose-Response Relationships Affecting Human Reactions to Odorous
Compounds, Cambridge, Massachusetts, April 1971.
Lindvall, T. "On Sensory Evaluation of Odorous Air Pollutant
Intensities," Nord. Hyg. Tidskr., 1970, supp1. 2, pp. 1-181.
Lindvall, t. "Sensory Evaluation of Odors in the Ambient Air."
Unpublished paper presented at the Conference on the Dose-Response
Relationships Affecting Human Reactions to Odorous Compounds,
Cambridge, Massachusetts, April 1971.
Meda1ia, N. Z. Community Perception of Air Quality: An Opinion
Survey in Clarkston. Washington. Public Health Service Publica-
tion No. 999-AP-10, U.S. Dept. of Health, Education, and Welfare
Cincinnati, Ohio, 1965.
-29-
-------
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
Meteorology and Atomic Energy 1968. U.S. Atomic Energy Commission
Div. Techn. Inform., Washington, D.C., 1969.
Methods for Measuring and Evaluating Odorous Air Pollutants at
the Source and in the Ambient Air. ,Report of the Third Karolinska
Institute Symposium on Environmental Health, Stockholm, June 1970.
National Survey of the Odor Problem, Phase I of a Study of the
Social and Economic Impact of Odors, prepared for the National
Air Pollution Control Administration. Copley International Corpora-
tion, January, 1970.
Posselt, H. S., and Reidies, A. H. "Odor Abatement with Potassium
Permanganate Solution," 1.& E.C. Product Research and Development,
Vol. 4, March 1965, pp.48-50.
Ridker, R. G. Economic Costs of Air Pollution. Frederick O. Praeger,
New York, 1967, Chapters 6 and 7.
Ridker, R. G., and Henning, J. A. "The Determinants of Residential
Property Values with Special Reference to Air Pollution," Review
of Economics and Statistics, XLIX, ,May 1967, pp. 246-257.
Rounds, F. G., and Pearsall, H. W. "Diesel Exhaust Odor--Its
Evaluation and Relation to Exhaust Gas Composition." SAE Paper
No. 863, SAE National Diesel Engine Meeting, November 1956 (also
SAE Transactions, 1957).
Sanders, G. R., Umbraco, R. A., Twiss, S.,
"The Measurement of Malodor in a Connnunity
metry." Unpublished paper presented at the
stitute ~ymposium on Environmental Health,
and Mueller, P. K.
by Dynamic Olfacto-
Third Karolinska In-
Stockholm, June 1970.
Sorenson, S. "On the Possibilities of Changing the Annoyance
Reactions to Noise by Changing the Attitudes to the Source of
Annoyance," Nord. Hyg. Tidskr., 1970, Supplementum 1.
Springer, K. J. "Studies of Public Opinion of a Traffic Odor."
Unpublished paper presented at the Conference on the Dose-Response
Relationships Affecting Human Reactions to Odorous Compounds,
Cambridge, Massachusetts, April 1971.
Springer, K. J., and Hare, C. T. A Field Survey to Determine
Public Opinion of Diesel Engine Exhaust Odor. Final Report No.
AR-718, to the Environmental Protection Agency, Air Pollution
Control Office, Contract PH-22-68-36. Southwest Research Institute,
San Antonio, February 1970.
Springer, K. J., and Hare, C. T. "Four Years of Diesel Odor and
Smoke Control Technology Evaluations--A Summary." ASME Paper No.
68-WA/APC-3, Winter Annual Meeting, Los Angeles, California,
November 16-20, 1969.
-30-
-------
(46)
(47)
(48)
(49)
(50)
(51)
(52)
(53)
(54)
(55)
(56)
Stahman, R. C., Kittredge, G. D., and Springer, K. J. "Smoke and
Odor Control for Diesel-Powered Trucks and Buses." SAE Paper No.
680443, Mid-Year Meeting, Detroit, Michigan, May 20-24, 1968 (also
SAE Transactions, Vol. 10, pp, 1657-94).
"Standard Method for Measurement of Odor in Atmospheres (Dilution
Method)," 1966 Book of ASTM Standards, Part 23, October 1966,
pp. 728-731.
Stevens, R. K., O'Keeffe, A. E., Mulik, J. D., and Krost, K. J.
"1. Direct Chromatographic Analysis," Gas Chromatography of
Reactive Sulfur Gases in Air at the Parts per Billion Level. Pre-
print from National Air Pollution Control Administration, Cincinnati,
Ohio, 1969.
Stone, H., Ough, C. S., and Pangborn, R. M. "Determination of Odor
Difference Thresholds," Journal of Food Science, 1962, 27, p. 197.
Suzuki, T. "Sununary of the Problems in Odorous Air Pollution in
Japan." Unpublished paper presented at the Conference on the
Dose-Response Relationships Affecting Human Reactions to Odorous
Compounds, Cambridge, Massachusetts, April 1971.
Symposium on Odor. Special Technical Publication No. 164, American
Society for Testing and Materials, Philadelphia, 1954.
Taub, S. J. "The Role of Odors and Vapors in Allergic Disorders,"
E.E.N.T. Monthly, 46, June 1970, pp. 779-780.
Turk, A. "Properties of Odorants." Unpublished paper presented at
the Conference on the Dose-Response Relationships Affecting Human
Reactions to Odorous Compounds, Cambridge, Massachusetts, April 1971.
Turk, A. Selection and Training of Judges for Sensory Evaluation
of the Intensity and Character of Diesel Exhaust Odors, U.S. Dept.
of Health, Education, and Welfare, Public Health Service, 1967.
Turk, A., and Mehlman, S. "Correlations between Instrumental and
Sensory Characterization of Atmospheric Odors," Correlation of
Subjective-Objective Methods in the Study of Odors and Taste.
Special Technical Publication No. 440, American Society for Testing
and Materials, Philadelphia, 1968.
Waller, R. A. "Environmental Quality, Its Measurement and Control,"
Regional Studies, Vol. 4, 1970, pp. 177-191.
-31-
------- |