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
                        by

-------
     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




          Under Contract No. CPA 22-69-50




                        by




       COPLEY INTERNATIONAL CORPORATION
       Contributions were made to this report by




           ENGINEERING-SCIENCE, INC.










-------
                            ACKNOWLEDGMENTS
       Copley International Corporation gratefully acknowledges the cooperation of
the many government representatives, civic leaders, businessmen, and private
citizens who have given generously of their time to provide the data needed in this
study.  They have filled out time consuming questionnaires and put aside their own
pressing duties to allow personal interviews .  Without their wholehearted coopera-
tion and assistance, this report would not have been  possible.  Copley International
Corporation is especially grateful to Dr. Donald G. Gillette of the National Air
Pollution Control Administration for his  counsel and  assistance in the development
and conduct of the study.

       The work upon which this report  is based was performed pursuant to
Contract No. CPA 22-69-50 with the National Air Pollution Control Administration,
Public Health Service, Department of Health, Education and Welfare.
                              CONTRIBUTIONS
       The overall responsibility of this study was undertaken by Mr. Stuart H.
Neffeler, Vice President, Copley International Corporation.  Others who contributed
to the success of this study included:

       Dr. Amos Turk, Project Consultant and Professor,  City College of the City
University of New York.

       Dr. Timothy G. Shea, Project Manager,  Dr. Richard M. Males, and Messrs .
James C . Burns, Lawrence P. Kolb, Henry van de Pol, and Donald L. White of the
Engineering-Science, Inc., Research and Development Laboratory.

       Mr. R. David Flesh, Project Manager, Mr. Thomas H. Copeland, Survey
Director, Mrs. Marian O. Doscher, Mrs. Linda E. Hanson, Miss Perry Palmer,
and Messrs .  Roger C .  Brown,  Masashi Saito, and R . Paul Weddell of Copley
International  Corporation.

-------
TEXT
                                 Illl
                NATIONAL SURVEY OF THE ODOR PROBLEM
                               List of Errata
Page No.

   81
Location

Figure 1,
Legend
                 Description

Second and third boxes are reversed.
Description should be:
   $$$;>£  Population of from 150,000 to 327,000

          Population of from 56,000 to 149,000
   83
Figure 2,
Legend
   93

   97

  136

137-144

  153


  166
Line 4

Line 15
Figure 14
Table IV
Second and third boxes are reversed.
Description should be:
          Weighted value of from 246 to 675
          Weighted value of from  75 to 245

"Pump" should be  "pulp."

"Affect" should be "effect."

Should be Page No. 144.

Should be Page Nos. 136-143.

Arrow at left center of page should be labeled,
"To Rendering Plant."

"Kansas City"  should be  175 residents and 25
   businessmen (Test Area), 176 residents and
   25 businessmen (Control Area).
"Buffalo" should be 170 residents and 25 business-
   men (Test Area), 180 residents and 25 busi-
   nessmen (Control Area).
"Tampa" should be 175 residents and 25 business-
   men (Test Area), 175 residents and 26 busi-

-------
                            List of Errata (cont'd)
Page No.

  166



  166
  171
  175
  194
Location

Table IV
(cont'd)
Table V
Table VI
Table VIII
Note 2,
formula
              Description

"San Francisco" should be 177 residents and 25
   businessmen (Test Area), 160 residents and
   22 businessmen (Control Area).

"Kansas City" should be 89  men and 86 women
   (Test Area), 87 men and 89 women (Control
   Area).
"Buffalo" should be 86 men and 84 women (Test
   Area),  86 men and 94 women (Control Area).

"High taxes" should be 62%  (Test Area)  and 58%
   (Control Area).

"Noticeable odors" should be 82% (Test  Area) and
   67% (Control Area).
"Irritation of the eyes"  should be 69% (Test Area)
   and  67% (Control Area).

Replace "r" with "s" above  second summation
   sign.
APPENDIX
Page No.

 Title

  A-9
Location                      Description

Center of page   Contract number should be CPA 22-69-50.
At the end of
Table A-III
Notes to Table A-III should be:
     * Selenium compounds
    ** Acid fumes
   *** Caustic units
Source to Table A-III should be:
   Arthur C. Stern (ed.), Air Pollution, Vol. Ill

-------
                            List of Errata (cont'd)
Page No.
G-4 and
G-5

G-6
G-10
G-13
Location

Appendix G
through M,
the results
of each sur-
vey question
Q4B and
Q4C

Q6B
Q8G
Q14B,
line 1
             Description

Except for base numbers, each number listed
   under the headings "Test, " "Control, " and
   "Total" is a percentage of the number of re-
   sponses obtained for a given category to the
   total number of responses obtained for all
   categories.
Base numbers ("Bases")  indicate the total number
   of responses obtained for all categories.

Bases should be 226 (Test Area), 219 (Control
   Area), and 445 (Total).

Bases should be 394 (Test Area), 351 (Control
   Area), and 745 (Total).

Bases should be 1,000 (Test Area), 811 (Control
   Area), and 1,811 (Total).

"Reduces" should be "reduced."
1-30
Q14C
M-l, M-2
M-5 through
   M-12
M-13, M-17
M-19, M-22
M-25, M-31
M-34, M-43
M-45, M-48
M-50
M-52 through
   M-57
Ql, Q2A
Q3

Q4A, Q5
Q6A, Q7A
Q8, Q9A
Q10, Q16A
Q17A, Q18A
Q18B
Q19 through
   Q23
Bases should be 7 professional/technical (Test
   Area) and 2 professional/technical (Control
   Area)
Occupation categories (column headings) should be
   "Professional/Technical, " "Manager s/Se If-
   Employed," "Clerical/Sales," "Skilled/Service,"
   "Labor," "Student," "Retired," "Housewife,"
   "Unemployed," and "Refused."

Bases for "Buffalo" should be 170 (Test Area) and

-------
                       TABLE OF CONTENTS



                                                            Page

ACKNOWLEDGMENTS AND CONTRIBUTIONS	   i

LIST OF TABLES	  vii

LIST OF FIGURES	  xi

CHAPTER

    I     INTRODUCTION	   1

             OBJECTIVES AND METHODOLOGY	   1
             SUMMARY OF THE PRINCIPAL FINDINGS	   3

    H     THE NATURE OF ODORS, MAJOR ODORANTS AND
             ODORANT SOURCES	   9

             THE NATURE OF ODORS	   9
             DEVELOPMENT OF AGGREGATE ODOR
                 CHARACTERISTIC COEFFICIENTS	  14

    IH     ODOR PRODUCING POTENTIAL OF INDUSTRIAL
             SOURCES	  27

             SELECTION OF NUMBER OF EMPLOYEES AS A
                 MEASURE OF PLANT OUTPUT	  27
             METHOD OF WEIGHTING NATIONWIDE GROUPINGS
                 OF INDUSTRIAL PLANTS	  30

    IV     EFFECTS OF THE NATURAL ENVIRONMENT ON ODOR
             PROBLEMS	  47

    V     CONCENTRATIONS OF POPULATION AND COMBINATION
             OF POTENTIAL ODOR PROBLEM FACTORS	  73

             METHOD OF LOCATING AND RANKING CONCENTRA-
                 TIONS OF POPULATION	  73
             COINCIDENCE OF POTENTIAL ODOR PROBLEM
                 FACTORS	  79

-------
                   TABLE OF CONTENTS (Cont'd)
CHAPTER

   VI     SURVEY OF LOCAL AIR POLLUTION CONTROL AGENCIES ...  89

              DEVELOPMENT OF THE QUESTIONNAIRE AND
                 RESULTS OF THE MAIL SURVEY	  89
              SUMMARY OF DOCUMENTATION OF ODOR PROBLEMS .  .  97

   VII     THE SELECTION OF SEVEN METROPOLITAN AREAS FOR
              INVESTIGATION OF ACTUAL ODOR PROBLEMS	105

  VIII     TECHNICAL FIELD PROGRAM	119

              ANALYTICAL TECHNIQUES FOR ODOR EVALUATION ... 120
              TWO-DAY TECHNICAL FIELD INVESTIGATIONS	124
              SENSORY EVALUATION OF ODORS:  PHILADELPHIA
                 STUDIES	128

   IX     PUBLIC OPINION SURVEY	157

              DEVELOPMENT OF THE SURVEY	158
              SUMMARY OF RESULTS	169

    X     ASSESSMENT  OF THE NATIONAL ODOR PROBLEM	197

              IDENTIFICATION OF POTENTIAL ODOR PROBLEM
                 AREAS	197
              LEGAL RECOURSE TO ODOR PROBLEMS	203

   XI     CONCLUSIONS AND RECOMMENDATIONS	207

              CONCLUSIONS	207
              RECOMMENDATIONS	208

ANNOTATED BIBLIOGRAPHY	                    211

-------
                       TABLE OF CONTENTS (Cont'd)



APPENDIX*

   A     IDENTIFICATION OF ODOR PRODUCING AIR POLLUTANTS

   B     COMBINATIONS OF THREE-DIGIT ZIP CODES INCLUDED IN
              METROPOLITAN AREAS

   C     THREE-DIGIT ZIP CODE AREAS OF THE UNITED STATES

   D     SUMMARY OF SURVEY RESULTS FROM QUESTIONNAIRES
              RECEIVED FROM 100 LOCAL AIR POLLUTION CONTROL
              AGENCIES

   E     LOCATION OF FIELD INVESTIGATION SITES AND PUBLIC
              OPINION SURVEY AREAS

   F     SENSORY EVALUATION OF ODORS DATA

   G     FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS -
              SUMMARY OF TEST AND CONTROL GROUPS FOR ALL
              CITIES COMBINED

   H     FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS —
              SUMMARY, BY INCOME, OF TEST AND CONTROL GROUPS
              FOR ALL CITIES COMBINED

   I      FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS -
              SUMMARY, BY OCCUPATION, OF TEST AND CONTROL
              GROUPS FOR ALL CITIES COMBINED

   J      FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS —
              SUMMARY, BY EDUCATION, OF TEST AND CONTROL
              GROUPS FOR ALL CITIES COMBINED

   K     FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS -
              SUMMARY, BY AGE, OF TEST AND CONTROL GROUPS
              FOR ALL CITIES COMBINED

   L     FINDINGS OF PUBLIC OPINION SURVEY OF BUSINESSMEN —
              SUMMARY OF TEST AND CONTROL GROUPS FOR ALL
              CITIES COMBINED

-------
                        TABLE OF CONTENTS (Cont'd)
APPENDIX*
    M      FINDINGS OF PUBLIC OPINION SURVEY OF RESIDENTS -
              SUMMARY OF TEST AND CONTROL GROUPS BY INDIVIDUAL
              CITIES

    N      PUBLIC OPINION SURVEY QUESTIONNAIRES AND INSTRUCTIONS
              TO INTERVIEWERS
* Separate volume.

-------
                         LIST OF TABLES
Table

          CHAPTER II

 I        LEVELS OF RESPONSE TO ODORANTS	  18

 II        ODORANT RANKINGS	  18

III        INDUSTRIAL ODOR RANKINGS	  20


          CHAPTER III

 I        NUMBER OF PLANTS WITH A PRIMARY LINE OF
          BUSINESS IN A CORRESPONDING SIC CLASSIFICATION	  31

 II        FIFTY METROPOLITAN AREAS IN THE UNITED STATES
          HAVING THE LARGEST NUMBER OF INDUSTRIAL PLANTS
          INCLUDED IN TABLE 1	  34

III        DISTRIBUTION OF PLANTS INTO TWELVE RANGES OF
          NUMBER OF EMPLOYEES - BY SIC CLASSIFICATION	  37

IV        FIFTY METROPOLITAN AREAS IN THE UNITED STATES
          HAVING THE HIGHEST LINEAR WEIGHTED VALUES OF
          INDUSTRIAL PLANTS INCLUDED IN TABLE I	  42

 V        FIFTY METROPOLITAN AREAS IN THE UNITED STATES
          HAVING THE HIGHEST LINEAR-MODIFIED WEIGHTED
          VALUES OF INDUSTRIAL PLANTS INCLUDED IN TABLE I  ...  43


          CHAPTER IV

 I        SUMMARY OF INVERSION FREQUENCY DATA ON
          ANNUAL NATIONWIDE BASIS	  53

 II        SUMMARY OF MONTHLY MEAN MAXIMUM MIXING
          DEPTHS ON NATIONWIDE BASIS	  54

III        INTERPRETATION OF ATMOSPHERIC STABILITY DATA
          WITH RESPECT TO ATMOSPHERIC DILUTION POTENTIAL ...  55

-------
                      LIST OF TABLES (Cont'd)
Table

          CHAPTER V

 I        FIFTY METROPOLITAN AREAS IN THE UNITED STATES
          HAVING THE LARGEST NUMBER OF PERSONS (1969
          Estimates)	  75

 II        PRINCIPAL COUNTIES OF FIFTY STANDARD METRO-
          POLITAN STATISTICAL AREAS IN THE UNITED STATES
          HAVING THE HIGHEST DENSITIES OF POPULATION	  78
          CHAPTER VI

 I        ODOR SOURCES WITH ABOVE AVERAGE NUMBER OF
          COMPLAINTS PER MONTH	  94

 II        NINE CATEGORIES OF ODOR SOURCES RANKED BY
          DEGREE OF SERIOUSNESS AND SITUATION OVER TIME ....  94

III        FILING SYSTEMS FOR AIR POLLUTION COMPLAINTS	  99
          CHAPTER VII

 I        PRELIMINARY LIST OF THIRTY-ONE POTENTIAL ODOR
          PROBLEM AREAS	107

 II        PRELIMINARY LIST OF THIRTY-ONE POTENTIAL ODOR
          PROBLEM AREAS BY CENSUS REGION	109

III        NUMBER OF SIC CLASSIFICATIONS AND INDUSTRIAL
          PLANTS IN EACH OF THE SEVEN METROPOLITAN AREAS
          SELECTED FOR INVESTIGATIONS OF ACTUAL ODOR
          PROBLEMS	115

-------
                      LIST OF TABLES (Cont'd)
Table

          CHAPTER VIII

 I        ENGINEERING-SCIENCE, INC. TECHNICAL FIELD
          PROGRAM ITINERARY	126

 II        AREAL EXTENT OF PERCEPTIBLE ODOR TWO-DAY
          TECHNICAL FIELD INVESTIGATIONS	127

III        ANALYSIS OF TRIANGLE TEST RESULTS	133

IV        ANALYSIS OF ODOR INTENSITY RATING TESTS	  134

 V        UNITED STATES WEATHER BUREAU OBSERVATIONS
          PHILADELPHIA, PENNSYLVANIA, INTERNATIONAL
          AIRPORT OCTOBER 1969	146


          CHAPTER IX

 I        NUMBER OF TEST AND CONTROL AREAS IN EACH CITY. ...  160

 II        BUSINESS AND PROFESSIONAL ESTABLISHMENTS LISTED
          FOR SELECTION OF BUSINESSMEN RESPONDENTS   	162

III        DATES DURING WHICH PUBLIC OPINION SURVEY WAS
          CONDUCTED	164

IV        TOTAL NUMBER OF INTERVIEWS COMPLETED IN EACH
          CITY	166

 V        TOTAL NUMBER OF INTERVIEWS OF RESIDENTS COM-
          PLETED, BY MEN AND WOMEN, IN EACH CITY	166

VI        PERCENT OF RESIDENTS INDICATING THE FOLLOWING
          COMMUNITY PROBLEMS WERE SERIOUS OR SOMEWHAT
          SERIOUS IN THEIR TEST AND CONTROL AREAS	171

-------
                      LIST OF TABLES  (Cont'd)
Table

          CHAPTER IX (Cont'd)

 VII       PERCENT OF RESIDENTS INDICATING THE FOLLOWING
          MAJOR REASONS FOR DESIRING TO MOVE FROM THEIR
          TEST AND CONTROL AREAS	173

VIII       PERCENT OF TEST AND CONTROL AREA RESIDENTS
          INDICATING THE FOLLOWING CONDITIONS WERE
          ASSOCIATED WITH AIR POLLUTION	175

 IX       PERCENT OF RESIDENTS IDENTIFYING ODORS SMELLED
          IN THEIR TEST AND CONTROL AREAS	177

  X       PERCENT OF RESIDENTS DESCRIBING ONE ODOR
          SMELLED MOST OFTEN IN THEIR TEST AND CONTROL
          AREAS	178

 XI       PERCENT OF RESIDENTS INDICATING THEY WERE
          BOTHERED BY ODOR POLLUTION — BY ODOR SOURCE(S)
          IDENTIFIED	180

 XII       PERCENT OF BUSINESSMEN INDICATING THE FOLLOWING
          COMMUNITY PROBLEMS WERE SERIOUS OR SOMEWHAT
          SERIOUS IN THEIR TEST AND CONTROL AREAS	186

XIH       PERCENT OF TEST AND CONTROL AREA BUSINESSMEN
          INDICATING THE FOLLOWING REASONS FOR WANTING
          TO RELOCATE THEIR BUSINESSES	188

XIV       PERCENT OF TEST AND CONTROL AREA BUSINESSMEN
          INDICATING THE FOLLOWING CONDITIONS WERE ASSO-
          CIATED WITH AIR POLLUTION	190

-------
                         LIST OF FIGURES



Figure                                                         Page

         CHAPTER m

  1      WEIGHTS GIVEN TO A CUMULATIVE PERCENTAGE OF A
         TYPICALLY SKEWED DISTRIBUTION OF PLANTS	   39


         CHAPTER IV

  1      TEMPERATURE GRADIENTS AND ATMOSPHERIC STABILITY  . .   49

  2      INVERSION FREQUENCY FOR: (a) Winter, (b) Spring, (c)
         Summer, (d) Fall, (e) Annual	   51

  3      MEAN MAXIMUM MIXING DEPTHS ABOVE THE SURFACE
         DURING JANUARY	   56

  4      MEAN MAXIMUM MIXING DEPTHS, FEBRUARY	   56

  5      MEAN MAXIMUM MIXING DEPTHS, MARCH	   57

  6      MEAN MAXIMUM MIXING DEPTHS, APRIL	   57

  7      MEAN MAXIMUM MIXING DEPTHS, MAY	   58

  8      MEAN MAXIMUM MIXING DEPTHS, JUNE	   58

  9      MEAN MAXIMUM MIXING DEPTHS, JULY	   59

 10      MEAN MAXIMUM MIXING DEPTHS, AUGUST	   59

 11      MEAN MAXIMUM MIXING DEPTHS, SEPTEMBER	   60

 12      MEAN MAXIMUM MIXING DEPTHS, OCTOBER	   60

 13      MEAN MAXIMUM MIXING DEPTHS, NOVEMBER	   61

 14      MEAN MAXIMUM MIXING DEPTHS, DECEMBER	   61

 15      HOURLY MEAN WIND SPEED ESTIMATES FOR
         JANUARY, CORRECTED TO 50 FEET ABOVE  GROUND	   63

-------
                     LIST OF FIGURES (Cont'd)



Figure                                                        Page

         CHAPTER IV (Cont'd)

 16      HOURLY MEAN WIND SPEED ESTIMATES FOR APRIL,
         CORRECTED TO 50 FEET ABOVE GROUND	  63

 17      HOURLY MEAN WIND SPEED ESTIMATES FOR JULY,
         CORRECTED TO 50 FEET ABOVE GROUND	  64

 18      HOURLY MEAN WIND SPEED ESTIMATES FOR OCTOBER,
         CORRECTED TO 50 FEET ABOVE GROUND	  64

 19      PERCENT FREQUENCY OF NIGHTTIME WIND SPEED
         17 Miles Per Hour For: (a) Winter, (b) Spring, (c)
         Summer, (d) Fall, (e) Annual	  65

 20      AIR POLLUTION POTENTIAL WARNING PERIODS FOR
         YEARS INDICATED	  67


         CHAPTER V

  1      A QUARTILE RANKING OF SIZE OF POPULATION IN EACH
         THREE-DIGIT ZIP CODE AREA IN THE UNITED STATES	  81

  2      A QUARTILE RANKING OF LINEAR-MODIFIED WEIGHTED
         VALUE OF EACH THREE-DIGIT ZIP CODE AREA IN THE
         UNITED STATES	  83

  3      AIR POLLUTION POTENTIAL WARNING PERIODS FOR
         YEARS INDICATED	  85


         CHAPTER VII

  1      FIFTY METROPOLITAN AREAS IN THE UNITED STATES
         HAVING THE LARGEST NUMBER OF PERSONS	112

  2      PRINCIPAL COUNTIES OF FIFTY STANDARD METROPOLI-
         TAN STATISTICAL AREAS IN THE UNITED STATES HAVING
         THE HIGHEST DENSITIES OF POPULATION                  112

-------
                     LET OF FIGURES  (Cont'd)



Figure                                                         Page

         CHAPTER VII (Cont'd)

  3      FIFTY METROPOLITAN AREAS IN THE UNITED STATES
         HAVING THE HIGHEST LINEAR WEIGHTED VALUES OF
         INDUSTRIAL PLANTS	113

  4      FIFTY METROPOLITAN AREAS IN THE UNITED STATES
         HAVING THE HIGHEST LINEAR-MODIFIED WEIGHTED
         VALUES OF INDUSTRIAL PLANTS	113

  5      MAP OF THE UNITED STATES SHOWING CENSUS REGIONS
         AND DIVISIONS	114

  6      AIR POLLUTION POTENTIAL WARNING PERIODS	116

  7      ATMOSPHERIC AREAS OF THE UNITED STATES	117


         CHAPTER VIII

  1      MODEL 1-3 SCENTOMETER, BARNEBEY-CHENEY CO	122

  2      SCREENING TESTS FOR SIGNIFICANT DIFFERENCE IN
         PERFORMANCE	135

  3      HISTOGRAM FOR MERCAPTAN CALIBRATION RUN -
         MEAN INTENSITY -	138

  4      HISTOGRAM FOR RENDERING CALIBRATION RUN -
         MEAN INTENSITY -	139

  5      t TEST OF SIGNIFICANCE IN DIFFERENCE IN FIELD
         ODOR CALIBRATION RUNS MEANS OF DATA DISTRIBU-
         TIONS BETWEEN OBSERVER PAIRS	141

  6      COMPARISON OF SCENTOMETER  AND ODOR PANEL
         DATA FOR MERCAPTAN CALIBRATION RUNS  	143

  7      COMPARISON OF SCENTOMETER  AND ODOR PANEL
         DATA FOR RENDERING ODOR CALIBRATION RUN	144

-------
                      LIST OF FIGURES (Cont'd)



Figure                                                          Page

         CHAPTER VIII

  8      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 1  ....  147

  9      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 1  ....  148

 10      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 2  ....  149

 11      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 2  ....  150

 12      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 3  ....  151

 13      REFINERY ODOR EMISSION DISPERSION PATTERN NO. 3  ....  152

 14      RENDERING FACILITIES ODOR EMISSION DISPERSION
         PATTERN	153

 15      RENDERING FACILITIES ODOR EMISSION DISPERSION
         PATTERN	154


         CHAPTER X

  1      COINCIDENCE OF POTENTIAL ODOR PROBLEM FACTORS ....  199

-------
                                 CHAPTER I
                               INTRODUCTION
       There has been an increasing nationwide concern over the general problem of
air pollution in recent years, culminating in the passage of the Air Quality Act of
1967 by the United States Congress.  Within the provisions of this act, the National
Air Pollution Control Administration (NAPCA), an agency of the Public Health Service,
Department of Health,  Education and Welfare, was assigned the task of developing
air quality criteria to serve as guidelines for the local regulatory control of air
pollutants. Primary emphasis in the development of these criteria was placed on
the documentation of technology for the abatement and control of pollutant emissions
and on the assessment of the social,  economic, and general environmental effects of
pollutants.

       In accordance with these plans, NAPCA engaged Copley International
Corporation (CIC) as prime contractor, with Engineering-Science, Inc. (ESI), as
subcontractor,  to conduct a study of the social and economic impact of odors .
Phase I of this study was directed toward the generation of information relative to
the basic definition, identification, and assessment of the national odor problem.
Further research, during Phase II, is to be oriented toward the development of a
method of assessing the social and economic impact of an odor problem on a typical
community in the United States .

       Copley International Corporation was primarily concerned with surveys of
all local air pollution control agencies in the nation and public opinion in seven
metropolitan areas .  Engineering-Science, Inc., was primarily responsible for
a comparison of analytical techniques for odor evaluation.  Both organizations con-
tributed information leading to the identification of odor problem areas throughout
the United States. A statement of the objectives and methodology and a summary
of the principal findings of this study are presented in the following sections of this
chapter.
OBJECTIVES AND METHODOLOGY

       The overall objective of Phase I was an assessment of the national odor
problem based on the projections of the results of investigations in a selected sample
of odor problem areas.  Tasks leading to this objective included:

        •  The identification and description of major odorants and odorant
          sources.

-------
        •  A determination of the odor producing potential and the location of
          groupings of industrial odorant sources .

        •  A determination of the effects of the natural environment on odor
          problems and the identification of odorant supporting atmospheric
          regions.

        «  The location of concentrations of population.

        •  A survey of local air pollution control agencies.

        •  The selection of seven metropolitan areas for investigations of
          actual odor problems.

        •  A series of technical field investigations and sensory evaluation
          of odor studies.

        •  A series of public opinion surveys of residents and businessmen.

        •  A formulation of recommendations for the development of a method
          to assess the social and economic impact of odors (for use in Phase
          II of the study).

       In order to ultimately assess the national odor problem, it was necessary to
identify and evaluate the odor problem potential of areas throughout the United  States.
In doing so, three basic factors — odor producing potential of industrial odorant
sources, atmospheric vulnerability, and size of population — were developed sepa-
rately and then combined for an indication of the odor problem potential of each three-
digit postal zip code area in the nation.  Each of these factors represented a condition
(or conditions) defined as necessary for the presence of an odor problem in an  area:

        •  Industrial odor producing potential was intended to represent the
          number of industrial plants weighted by the aggregate effects of
          odors generated and the relative size of these plants.

        •  Atmospheric vulnerability was intended to represent the major
          causes of impaired atmospheric dilution  capacity.

        •  Size of population was intended to represent the potential extent
          of odor problems.

       The primary objective of the survey of local air pollution control agencies
was to identify areas in the United States where agency officials believe that signi-
ficant odor problems exist. Based on responses received from agency officials and
knowledge of the fifty metropolitan areas having the largest concentrations of persons

-------
and highest industrial odor producing potential, seven metropolitan areas were
selected as representative potential odor problem areas in which actual odor prob-
lems were to be investigated.

       Within the principal cities of each of these metropolitan areas, a technical
field program and public opinion surveys were conducted.  The technical field pro-
gram included two-day field investigations. In addition, sensory evaluation of odor
studies were performed in Philadelphia, Pennsylvania.

       The objectives of the two -day field investigations were to interview air quality
control agency officials in each of the seven metropolitan areas  for the purpose of
determining the manner in which odor problem situations are evaluated and document-
ed .  An additional objective was to identify actual odor problem  areas, the extent of
the areas affected,  and the types of emission causing concern.  A public opinion
survey was then conducted in at least one of these areas in each of the cities inves-
tigated .

       The objectives of the Philadelphia studies  were to apply direct sensory evalu-
ation techniques to the documentation of odorant emissions from two sources,
compare the responsiveness of the scentometer with odor panelists, and develop a
capability for sensory evaluation techniques for use in Phase II.

       The public opinion surveys were conducted as an exploratory approach to
determine the extent to which odors constitute an influence on the lives of people
living and working in actual odor problem areas.  The objectives were to determine
the degree of awareness of the public to odors, whether differences of opinion
regarding odors exist between people living in odor problem areas and people living
in relatively odor-free areas, if people feel they  have suffered socially or economi-
cally due to the presence of odors,  and the degree of interest the public may have in
odor abatement.

       Because of the exploratory nature of the survey (and, indeed, the study as a
whole), a variety of questions and question approaches were used with the expectation
that some would generate usable information while others might not allow for valid
discrimination between the variables .  The usable results were  to be projected to the
nation as a whole.
SUMMARY OF THE PRINCIPAL FINDINGS

       The principal findings of this study were drawn from the results of the tech-
nical field program and the surveys of local air pollution control agencies and public
opinion.  These findings are summarized below.

-------
Results of the Survey of Local Air Pollution Control Agencies

       The responses received from the mail survey of 184 local air pollution con-
trol agencies in the United States indicated that odor problems exist in many cities
and counties throughout the nation. It was further indicated that the odor problems,
in general, affected a very large number of people.

       Four categories of sources were responsible for almost half of the odor prob-
lems reported.  These categories were: rendering/meatpacking/slaughterhouses;
chemical/plastic plants;  sanitary land fill/dumps/open burning/incinerators;
petroleum and natural gas refining/asphalt production.  In contrast, the four cate-
gories reported to have the highest monthly average number of complaints were:
sanitary land fill/dumps/open burning/incinerators; fisheries and fish processing;
rubber/tires production; coffee roasting/spices. The responding agencies viewed
chemical/plastic plants as the most serious odor sources and pulp/kraft/wood mills
as the  sources of the fastest growing problems .

       The identification and assessment of odor problems was accomplished by a
large number of the reporting agencies using the human nose,  common sense,  and
number of complaints as criteria. An equally large number of agencies reported
that there are many inadequacies in current methods of assessment and  expressed
a need for assistance in establishing criteria.

       Over 70 percent of the responding agencies take some form of abatement
action. One of the most popular forms of abatement appears to be persuasion.

       In addition to the mail survey, personal interviews were held with agency
representatives in each of the seven metropolitan areas investigated.  The following
observations are from those interviews.

       The most frequently used legislation relating to odor abatement is found in
local and state nuisance ordinances .  Due to staff limitations in air pollution control
agencies as well as the general requirement that a considerable number of persons
perceive an odor situation as a nuisance, nuisance laws have been ineffective tools
in abating odors. To the extent this study could determine,  none of the metropolitan
areas investigated had a record of substantial fine or other judicial penalty levied
against an odor pollution violator.

Results of the Field Technical Program

       The following results were obtained from the two-day field investigations and
the sensory evaluation of odors studies.

       The scentometer, a vapor dilution technique  device,  was  found to be a utili-
tarian  and effective tool for odor problem identification at a point in time and space.

-------
Using the scentometer in each of the metropolitan areas visited, refinery installa-
tions and chemical facilities were found to be the most significant sources of odor
emissions in terms of areal extent.

       The sensory evaluation technique for atmospheric odor measurement provided
definitive descriptions of the areal extent and relative magnitude of odor emissions.
This technique should have substantive application in those special circumstances
where legal liability or compliance questions require explicit documentation of an
odor emission.

Results of the Public Opinion Survey

       The findings of the public opinion survey indicated that over 60 percent of the
residents and businessmen interviewed in the survey test areas perceived air pollu-
tion as a problem .  In defining air pollution, over 80 percent of these respondents
included "noticeable odors ."

       Residents and businessmen expressed an awareness of odors not only in the
survey test areas, but also in the similar, but relatively odor-free, control areas.
In general, however,  businessmen were  not as concerned as residents about the
presence of odors.

       The type of resident found in both the test and control areas in this study can
perhaps best be  characterized as lower-middle class .  He tended to be blue collar,
middle-aged or older, below average in education and income, and with family size
lower than average.

       Awareness of air pollution problems tended to rise with increases in the
resident's educational attainment, income, occupational level, and youth.  Young
respondents were apt to consider noticeable odors as a basic element of air pollu-
tion to a slightly greater degree than did older residents.   However, the vast
majority of all age groups held this view.

       Although awareness of odors was expressed in test and control areas, test
area residents were more aware of odors, able to identify odor sources,  and discon-
tent with the situation.  Almost 20 percent of the test area residents felt that odor
pollution in their areas of their cities was continuously serious. Only about 5 percent
of control area residents held the same view.  Test area residents stated that indus-
tries were the principal sources of odor  in their areas, while control area residents
believed motor vehicles to be the main source.  Almost 74 percent of test area
residents who stated odor pollution was at least occasionally serious felt that more
money should be spent to control odor pollution in their areas .  About 59 percent of
control area residents felt that more money should be spent for this purpose.

-------
Assessment of the National Odor Problem

       The results of the public opinion survey indicated an awareness of odors
among residents and businessmen in test and control areas .  Difficulty arises in
attempting to extrapolate the awareness level of perception to the problem level.
In this study, the differentiation between these levels was attempted through careful
development of the public opinion survey questionnaires .  It was planned that the
responses to the questionnaires would provide the necessary data for an assessment
of the extent of odor problems in the survey areas .

       A further  difficulty was encountered in attempting to project the  extent of
odor problems beyond the survey areas. To assist in overcoming this problem,
the following assumptions were formulated:

       (1)     The seven metropolitan areas investigated in this study are
              generally representative of most metropolitan areas in the
              United States. (Selection criteria are discussed in Chapter VII.)

       (2)     Two-thirds of the estimated 137 million persons in the  155
              metropolitan areas of the nation are male and female adults
              in  households (residents).

       (3)     Only 10 percent of these residents  live in actual odor prob-
              lem areas (similar to the public opinion survey test areas)
              and that 90 percent live in relatively odor-free areas (similar
              to  the survey control areas).

       From this, it is  projected that 25 million residents in the United States could
be expected to state that air pollution is a problem and that odors are  a  major ele -
ment.  Based on these assumptions, further projections of the extent of  odor problems
can be made:

       (1)     About one and one-half million residents would voluntarily
              state that odors are a disadvantage to living in their areas
              of  their cities.

       (2)     Almost five million residents would admit that odors bothered
              them very much.

       (3)     More than five and one-half  million residents would state that
              odors are a continuously serious problem.

       (4)     About three million residents would feel that odor pollution
              has reduced the value of their home property.

-------
       (5)     Almost five million residents would have seriously considered
              moving away from their areas because of odor pollution.

       (6)     Only one-half million residents would have requested some
              authority or agency to take action concerning air polltuion.
              The number of residents that would have requested action
              concerning odor problems would be a small fraction  of this .

       On the basis of these projections,  a large number of residents perceive odors
as a problem.  Yet only a small percentage  of these residents are motivated to seek
recourse.  The reasons for this apparent apathy cannot be inferred from the public
opinion survey findings.  Instead,  additional study of attitudes is required to develop
meaningful conclusions.

-------
                                 CHAPTER II
                          THE NATURE OF ODORS,
                 MAJOR ODORANTS AND ODORANT SOURCES
       As a preface to the method of locating and assessing potential odor problem
areas across the United States, this chapter begins with a discussion of the nature
of odors.  Particular emphasis is given to the inherent odor characteristics —
intensity,  quality,  acceptability, and pervasiveness .

       The remaining section of the chapter describes the development of a set of
aggregate odor characteristic coefficients, each of which is to represent  an aggre-
gation of the perceptive characteristics of odorants attributable to a specific source.
Engineering-Science, Inc., developed these coefficients by first identifying major
sources of odorant emissions and the physico-chemical nature of the emissions from
these sources, by classifying the sources by Standard Industrial Classifications (SIC),
and then by subjectively assigning a quartile ranking to the sources based on the
characteristics of the odorants emitted.
THE NATURE OF ODORS

       Odors can be classified in two categories: source and ambient.  Source odors
are those existing at the point of origin or at their point of exit to the general atmos-
phere, and ambient odors are those existing in the general atmosphere .   Odors can
be confined or unconfined as to odor source.  Confined sources can generally be char-
acterized in terms of volumetric rate of discharge,  temperature, moisture content,
location, elevation, area, stack height,  and state of aggregation (gas, mist, etc.,
including particle size distribution, etc.).2  Typical confined sources include stacks
and fume vents.  Unconfined sources consist of large sources such as sewage treat-
ment plants and stockyards where the areal extent of the source precludes its
confinement.

       Ambient odors are unconfined in the source context.  However,  a confinement
may exist in the  sense that the atmospheric volume  into which the odor is emitted may
be limited by meteorological considerations such as wind speed, mean mixing depth,
humidity, etc ., and by topographical constraints to  the extent that the level of odorant
may remain steady or even increase rather than decrease.

       Odors can be characterized by intensity,  quality, acceptability, and pervasive-
ness .  Odor intensity is defined as the numerical or verbal indication of the strength
of an odor.  Odor quality is the verbal description of an odor, accomplished within
the vocabulary capabilities of the individual and normally expressed by comparison

-------
with common odors.  Acceptability is an absolute acceptance/rejection of the odor on
the basis of intensity and/or quality.  Pervasiveness is the ability of an odor to spread
throughout a large volume of air and persist at perceptible levels.

Odor Intensity

       Intensity Relationships.  The human nose is the ultimate standard against
which the intensity of an odor must be evaluated. The response of the human nose
to odors is bounded at the lower end of the concentration scale by a threshold value
and at the upper end by an objectionability concentration. The threshold concentra-
tion can be defined in terms of a detection threshold concentration,  at which the
odorant concentration just gives an intensity of zero or as a recognition threshold
concentration, the concentration at which odor quality can be recognized.3  The
detection threshold limit is used most commonly by researchers although it has been
argued that the quality recognition sensation is more reproducible than the detection
sensation. ^

       Individual response for detection of odors varies widely; as  a  result, it is
customary to use a panel to determine odor threshold concentration and to report
threshold values as the "effective dosage" at which 50 percent of the panelists per-
ceive the odor.  Reported threshold concentrations for the same odorant vary widely,
not only because of individual response variability, but also because of lack of stan-
dardization of techniques for managing of samples and presenting them to observers,
and lack of standardization in developing and reporting of threshold odor concentrations
and in describing odors precisely as to their physical and chemical nature.

       The intensity of an odorant has been observed to vary exponentially with its
concentration, as described by the Weber-Fechner Psychochemical Law.^ The Weber
expression is:

                          I =kln C                                (Equation 1)

       Where:   I  =  intensity of odor sensation
                 k  =  constant
                 C  =  odorant concentration

Several odorants are known to follow this law  over very wide ranges.  The range of
intensity from 0 to 5 covers eight log cycles for ethyl mercaptan, six for butyl
thioether, and four for crotonaldehyde .

       The intensity characteristics of a mixture of odorants may be independent,
counteractive, additive, or synergistic. These interactions can be expressed mathe-
matically for two odorants, A and B, as follows.

-------
       Independent:         I^g = k In (C^orCg)                   (Equation 2)

       Counteraction:      I^g < k In (C^orCg)                   (Equations)

       Addition:            IAB = k In (CA + CB)                   (Equation 4)

       Synergism:          I^g > k In (C^ + Cg)                   (Equation 5)
       Human Olfactory Response .  Human odor intensity scales are designed to
measure the perceived intensity of some specified characteristic or attribute of a
material. 8  The descriptor of the odor may be general (e .g. , overall intensity of
odor or flavor) or  specific (e .g. , sweetness of a beverage) .  This method is not to
be used for determining the absolute threshold but may be used with any material
or product and for any attribute which can be understood clearly by the subjects .  In
this type of method trained subjects specifically instructed in the attribute to be eval-
uated are presented a series of samples .  Each sample is rated for intensity with
alternative points anchored as follows:  none,  slight,  moderate, large or strong,
and extreme .

       Olfactory sensitivity is known to be highly variable and is subject to  physical
as well as psychological influences .  Some of the factors which have been reported
to influence the olfactory sensitivity are as follows:"

       ( 1)     Odor sensitivity of the individual observer varies diurnally but
              maintains a reasonable level of constancy over a group of ob-
              servers .

       (2)     The sense of smell is rapidly fatigued, though fatigue for one
              odor does not necessarily affect the perception of dissimilar
              odors .

       (3)     Responses to odors are not totally objective because psycholo-
              gical responses vary in different observers .

       (4)     The sensitivity of observers varies widely in that some persons
              are extremely sensitive while others are almost insensitive to
              odors ; sensitivity decreases with age .

-------
       The above factors have significant impact on the interpretation of subjective
data developed under either in situ or laboratory conditions.  The order of presen-
tation of sample is of paramount importance as is the time of exposure of the
observer to a sample.  Inasmuch as the olfactory sense can become fatigued (or
acclimated) within a time period (or adjustment time), exposure for a duration
greater than adjustment time can lead to diminution or loss of odor perceptiveness .
Because of adjustment time subjective data must be interpreted relative to the ex-
posure context in which it was developed, viz.:

       (1)     Laboratory exposure:  exposure time <  adjustment time.
       (2)     Field exposure:  exposure time > adjustment time.
       (3)     Field exposure:  exposure time < adjustment time.

       Case (2) occurs when one enters and remains in an odorified atmosphere;
Cases (1) and (3) occur only when the subject undergoes a transient exposure to an
odorified atmosphere.

       Recognition of an odor problem  is related functionally to one's perception of
the odor.  However, under circumstances of extended exposure,  such as everyday
life in an area with an odorified atmosphere, perception is minimized due to accli-
mation. It follows that, under field circumstances, odor problems will be perceived
by resident populations primarily under circumstances causing an odorified atmos-
phere to be present on a transient basis.

Odor Quality

       The characteristic of odor quality is normally defined in terms of commonly
perceived odors or by associating unfamiliar with familiar odors. An observer
often does not possess the vocabulary to describe the odors he perceives inasmuch
as humans can distinguish between hundreds of thousands of odors. Often, quality
of an odor may change with dilution of the odor.  In the case of an odorant mixture,
a more pervasive odorant in the mixture may alter the odor quality significantly as
the less pervasive, more concentrated  odorant is diluted.

       Several odor quality classifications exist, three of which are presented
below:10

       (1)     Henning's odor classification

              (a)   Spicy:     conspicuous in cloves, cinnamon, nutmeg, etc.
              (b)   Flowery:  conspicuous in heliotrope, jasmine, etc.
              (c)   Fruity:    conspicuous in apple,  orange oil, vinegar, etc.
              (d)   Resinous:  conspicuous in coniferous oils and turpentine.
              (e)   Foul:      conspicuous in hydrogen sulfide and products
                              of decay.
              (f)   Burnt:     conspicuous in tarry and scorched substances.

-------
       (2)     Crocker-Henderson classification

              (a)    Fragrant or sweet
              (b)    Acid or sour
              (c)    Burnt or empyreumatic
              (d)    Caprylic, goaty or oenanthic

       (3)     Horstman classification of odor descriptors

              (a)    Flowers
              (b)    Pulp mill
              (c)    Smoke, woodsmoke
              (d)    Burning leaves
              (e)    Mustiness
              (f)    Gasoline
              (g)    Rendering plant
              (h)    Rubbish burning
              (i)    Animal odors
              (j)    Miscellaneous odors

Of the above three classifications, the Horstman system represents the most serious
attempt to categorize odor qualities in terms of commonly-understood descriptors.

Odor Acceptability

       An odor may be acceptable or unacceptable depending on its intensity and
quality.  The acceptability of an odor is related to individual perferences which are
known to be related to age, sex, vocation and environment.  Odor acceptability is of
immediate  relevance in evaluation of field odor problems in that odors acceptable to
some individuals on a permanent basis may not be acceptable to others on a transient
basis.  Likewise,  odorants (such as perfumes) acceptable at one concentration may
not be acceptable at another.

Odor Pervasiveness
       Odor pervasiveness is the characteristic by which odors persist at detectable
concentration levels over wide ranges of concentration.  Pervasiveness is often
quantitated in terms of odor potential or threshold dilution ratios.  The odor unit has
been defined to describe the number of dilutions necessary to reduce the odor to the
threshold concentration.  An odor unit is equal to the volumes of air necessary to
dilute the concentration of odorant in one volume of air to the threshold concentra-
tion.     Pervasiveness is indicated mathematically by the k value in the Weber-
Fechner relationship (Equation  1), pervasiveness increasing as the  value of k
decreases.

-------
DEVELOPMENT OF AGGREGATE ODOR CHARACTERISTIC COEFFICIENTS

       At the outset of this study, it was desired to develop a means of estimating
for a specific area the odor producing potential resulting from the presence of odor-
generating activities.  The work was designed so that the information obtained in
pursuing this aim would:

       (1)     Represent the aggregate effects of a specific source, including
              the quality, intensity, acceptability, and pervasiveness of odors
              attributable to the source.

       (2)     Include a measure of the production activity causing the odorant
              emission.

       (3)     Locate specifically a cross-sectional representation of potential
              sources of odorant emissions, in terms of state,  city,  and street
              address.

       (4)     Provide a preliminary estimate of the odor producing potential in
              a given area as a summation of the subproducts of the aggregate
              odor characteristic coefficients and an activity measure for the
              representation of odorant emitting production activities in the area.

       The utility of an estimate of the odor producing potential in each such area in
the United  States would be to aid in selecting cities for the  purpose of conducting tech-
nical field  investigations and public opinion surveys and to  provide a parameter
which could be related, on a neighborhood  scale, to the actual levels of odor pollu-
tion observed during technical and public opinion surveys.

       Difficulty was encountered, however, in attempting to include  odor-generating
activities of natural or unexpected origin,  e .g., the decomposition of sea life or a
neighbor burning trash.  The principal problem was in locating a sufficient number
of these sources within budgetary constraints. Therefore,  attention was directed
principally at the most easily found sources — industrial odorant sources.

       Thus, the procedures used in developing a means of estimating the odor pro-
ducing potential of industrial odorant sources were as follows:

       (1)     A set of aggregate odor characteristic coefficients was developed
              by first identifying major sources of odor emissions and the
              physico-chemical nature of the emissions from these sources, by
              classifying the sources by SIC code, and then by subjectively
              assigning a quartile ranking to the sources based on the character-
              istics of the odorants emitted.  This work was done by Engineering-
              Science, Inc.  It is described in the following pages of this chapter.

-------
       (2)     An indirect measure of production activity (number of employees)
              was  selected by Copley International Corporation, which could be
              tied  readily to a specific potential emitter (an industrial plant) by
              SIC code using information available for purchase from Dun &
              Bradstreet, Inc.

       (3)     Computer tabulation techniques were then used by CIC to develop
              estimates of the odor producing potential of groupings of industrial
              plants throughout the United States . Additional details of the pro -
              cedures used and assumptions made in selecting the activity
              measure and developing these estimates are included in Chapter
              III.

       It is recognized that there are  several assumptions inherent specifically in
the designation of a quartile ranking for an odor emission and tying that ranking
broadly to all sources across the nation falling within an SIC category.  The key
assumptions are that:

       (1)     All industries within an SIC classification are potential sources
              of odorant  emissions .  (The scope of the  study precluded the
              identification of which industries within an SIC category were not
              odorant emitters of concern.)

       (2)     The  level of manufacturing process technology and emission con-
              trol  technology varies widely with age of  plant, specific  mission,
              corporate policy, etc.  (No consideration could be given to this
              factor for the spectrum of facilities found in an SIC code.)

       In addition to the above qualifications, the literature was found to be devoid
of information by which to establish a  proportionality between:

       (1)     Technological progress and a generalized statement of how this
              has attenuated odorant emission.

       (2)     Units of raw material consumption or production and units of
              odorant emission.

       (3)     Units of raw material consumption within an SIC  aggregation of
              industries  and any of the activity information provided by Dun &
              Bradstreet, Inc.

       The  broad scope of these qualifications was necessary for the development
of preliminary estimates  of odor producing potential.  It is emphasized that the
justification for this effort is the utility of these estimates in problem area identi-
fication, which is one of the principal  objectives of this  study.

-------
Identification of Odor Producing Air Pollutants

       Major odor producing air pollutants were identified by class of chemical
compound, class of combustion process, and category of industrial activity from the
basic reference literature. ^  The information is assembled in Appendix A in the
following format:

       (1)     Table A-I: Major Odor Producing Air Pollutants, Olfactory
              Thresholds and Related Data. Pollutants are identified by class
              of chemical compound.

       (2)     Table A-II: Odorous Emissions From Combustion.  Pollutants
              are identified by class of combustion process.

       (3)     Table A-III:  Odorous Emissions From Industrial Operations.
              Pollutants are identified by industrial category and process
              within category.

Industrial Odor Rankings

       As an initial step in developing a means of estimating the odor producing
potential of industrial odorant sources, a tentative set of industrial odor rankings
was devised based on the assumption that those industries emitting the strongest
odorants are the most likely to cause odor problems. These tentative odor rankings
were developed as follows:

       (1)     Those industries which were known to be potential sources  of
              odor emission were identified, along with the major odorants
              associated with each.

       (2)     The odorants were assigned a ranking of from one to four on the
              basis of odor strength, as measured by concentrations required
              to produce threshold perception (using data summarized in
              Appendix A).

       (3)     The industries were assigned a quartile  ranking based on the
              rankings assigned to the industries' characteristic odorants.

       The industrial odor rankings developed in this manner were then reviewed
subjectively to evaluate the correspondence of these rankings to the odor problem
levels of the  industries involved as reported qualitatively in the literature. As a
result of the  subjective review, a set of aggregate odor characteristic coefficients
was established.

-------
       Development of Tentative Rankings Based on Odorant Strength.  The standard
methods for the measurement of odorant strength are based on differing levels of
human response. The threshold, or perception, concept expresses odorant strength
in terms of the minimum concentration required for perception.  Other methods of
odorant measurement are based on the concentration required for some higher level
of response to the odorant involved (beyond mere perception) as expressed in terms
of the perception concentration (Ct) multiplied by the intensity factor (F) raised to
the appropriate power as reported in Table 1.13

       Two levels of response were used as the basis for odorant strength rankings.
One set of rankings was developed on the basis of concentrations required for thresh-
old perception, and another set was based on concentrations required to produce the
most intense effect  defined in the literature, usually stated as "very strong."  These
concentrations were obtained using the expressions shown  in Table I and a compila-
tion of odorant constants Ct and F obtained from the literature.

       In both cases the odorants were arranged in order of decreasing concentra-
tion required to produce a given level of response.  This list was divided into four
roughly equal groups, and the odorants in the first group were assigned a rank of
1, those in the second group a rank of 2, and so on.

       In general, the two sets of odorant rankings thus obtained were similar.
The set based on threshold response was used in developing the tentative industrial
odor ranking because it was based on more complete data (some  F values were not
available for the intense effect rankings), and because they tended to agree more
closely with a consensus of opinion on odorant effects found primarily in the litera-
ture .  These rankings are presented in Table II.

-------
                                  Table I
                   LEVELS OF RESPONSE TO ODORANTS
Level of Response
Perception
Faint (weak, readily perceptible)
Easily noticeable (moderate
Strong (forcible)
Very strong (intense effect)
Concentration
Ct
CtF
°tF3
C'F*
Source: Elmer Kaiser, "Odor and Its Measurement, " Air Pollution, Vol. I., ed.
        Arthur C. Stern (New York: Academic Press, 1962).
                                  Table II
                           ODORANT RANKINGS
            Odorant
        Sulfur Dioxide
        Hydrogen Sulfide
        Sulfides
        Mercaptans
        Ammonia
        Amines
        Selenium Compounds
        Hydrocarbons
        Alcohols
        Phenol
        Esters
        Ethers
        Aldehydes
        Ketones
        Carboxylic Acids
        Hydrochloric Acid
        Unsaturated Alcohols,  Ethers, Ketones, Acids
        Iron Oxides
        Halogens
Ranking
   2
   4
   2
   4
   1
   4
   4
   1
   1
   2
   2
   1
   3
   2
   3
   1
   3
   1
   1
Source: Engineering-Science, Inc.

-------
Odor Characteristic Coefficients

       The principal odorant characteristics of different industries are compiled in
Appendix A.  This list was used to obtain rankings for each industry by using a num-
ber which equalled the sum of the rankings shown in Table  II for each odorant listed
for a particular industry. The values obtained were then ranked on a one to four
scale.

       The industrial odor rankings obtained in this manner were subsequently
reviewed in the light of the problem levels of the industries involved, as reported
qualitatively in the literature.  In general,  the  odor rankings were supported by the
literature in that those industries most often reported as causing odor problems in
the literature tended to have highest rankings,  however, some rankings were
modified.

       The industrial odor rankings developed  on the basis of characteristic odorant
strength are presented in Table III ("Calculated Rank" column),  together with the
odor characteristic coefficients established on  the basis of the subjective review
described above ("Odor Characteristic Coefficients" column). The coefficients were
recommended for use in conjunction with a measure of production activity in esti-
mating the odor producing potential of groupings of industrial plants .

       It is emphasized that the odor characteristic coefficients do not reflect
directly the influence of the numerous variables related to  human odor perception,
as discussed in the first section of this chapter.  In particular,  no account is taken
of the effects of human response to a continuous odor (olfactory  fatigue), the combined
effects of different odorants, e .g.,  masking phenomena, and the importance of odor
variations relative to average odor intensities . However,  in view of the constraints
of data availability noted earlier, it is believed that the set of aggregate odor charac-
teristic coefficients presented here represents a reasonable empirical response to
the industrial odor ranking requirement.

-------
         Table IE
INDUSTRIAL ODOR RANKINGS
Type of Industry
Incinerators
Open Burning
Pulp and Paper


Cement Production
Glass and Ceramics

Asphalt and Bituminous
Iron and Steel Production
Iron and Steel Foundries


Copper Smelting and Refining
Lead Sintering
Titanium

Zinc

Aluminum

Non -Ferrous Foundries


SIC Code
4952
4953
2611
2621
2631
3241
3211
3229
2951
3312
3321
3322
3323
3331
3332
3339
3341
3333
3341
3334
3341
3361
3362
3369
Calculated
Rank
4
4
4
4
4
1
1
1
1
4
3
3
3
3
3
1
1
1
1
1
1
3
3
3
Odor
Characteristic
Coefficient
2
4
4
4
4
1
1
1
3
2
2
2
2
3
3
1
1
1
1
1
1
3
3
3

-------
                              Table in (Cont'd)
Type of Industry
Industrial Inorganic Chemicals
Alkalis
Fertilizers
Petroleum Refineries
Organic Chemicals











Animal Production
Meat Packing

Fish Processing

Fruit and Vegetable Processing
Coffee Roasting

Breweries

SIC Code
2819
2812
2871
2911
2815
2818
2821
2822
2823
2824
2833
2841
2842
2844
2851
2861
4731
2011
2015
2031
2036
2033
2072
2095
2082
2085
Calculated
Rank
2
1
2
3
4
4
4
4
4
4
4
4
4
4
4
4
2
4
4
4
4
4
-
-
-
-
Odor
Characteristic
Coefficient
2
1
3
2
3
3
3
3
2
2
3
2
3
3
4
4
4
4
4
4
4
2
3
3
2
2
Source: Engineering-Science, Inc.

-------
                                    Notes
 1.     Norman A.  Huey,  et al.,  "Objective Odor Pollution Control Investigations, "
        Journal of the Air Pollution Control Association, X, No.  6 (December, 1960),
        pp. 441-446.
 2.     Amos Turk and Stanley Mehlman, "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, 1968),
        p. 27.
 3.     Ralph J. Sullivan,  Air Pollution Aspects of Odors (Bethesda, Maryland:
        Litton Systems, Inc., 1969), draft copy, pp. 6-8.
 4.     G. Leonardos, D. Kendall and N. Bernard, "Odor Threshold Determinations
        of 53 Odorant Chemicals, " Journal of the Air Pollution Control Association,
        XIX, No. 2 (February,  1969), p. 91.
 5.     Sullivan, op. cit., draft copy, p. 5.
 6.     J. F. Byrd and A. H. Phelps, Jr.,  "Odor and Its Measurement, " Air Pollution,
        Vol. II, ed. A. C. Stern (2nd ed.;  New York:  Academic Press, 1968),
        pp. 305-327.
 7.     Sullivan, op. cit., draft copy, p. 6.
 8.      Manual on Sensory Testing Methods, Special Technical Publication No. 434
        (American Society for Testing and Materials, 1968), p. 33.
 9.     S. W. Horstman, R. F. WrombleandA. N. Heller,  "Identification of
        Community Odor Problems by Use of an Observer Corps, " Journal of the
        Air Pollution Control Association, XV, No. 6 (June, 1965), p. 261.
10.     Ibid.,  and Sullivan, op. cit., draft copy, p. 10.

-------
                               Notes (cont'd)
11.    Sullivan, op. cit., draft copy, p. 12.
12.    W. Summer, Methods of Air Deodorization (Amsterdam;  Elsevier, 1963),
       and Arthur C. Stern (ed.), Air Pollution, Vol. Ill (2nd ed.; New York:
       Academic Press,  1968).
13.   J. F. Byrd and A. H. Phelps, Jr., "Odor and Its Measurement, " Air Pollution,
      Vol. I, ed. Arthur C. Stern (New York:  Academic Press,  1962).

-------
                               References
Byrd, J. Floyd and Phelps, Austin H.,  Jr.  "Odor and Its Measurement," Air
       Pollution.  Vol. II. Edited by A. C. Stern, 2nd ed.  New York:
       Academic Press,  1968.
Gould, Robert F. (ed.)  Flavor Chemistry.  Washington, D.C.: American
       Chemical Society Publications,  1966.
Hayashi, T. (ed.)  Olfaction and Taste.  Vol.11.  New York:  Pergamon Press,
       1967.
Horstman, S. W., Wromble, R. F. and Heller, A. N.  "Identification of Commu-
       nity Odor Problems by Use of an Observer Corps, " Journal of the Air
       Pollution Control Association.  XV, No. 6  (June, 1965), 261.
Huey, A., et al.  "Objective Odor Pollution Control Investigations, " Journal of the
       Air Pollution Control Association .  X, No. 6 (December, 1960), 441-446.
Byrd, J. F. and Phelps, A. H.f Jr.  "Odor and Its Measurement," Air Pollution.
       Vol. I.  Edited by A. C. Stern (New York: Academic Press, 1962).
Kerka, William F. and Kaiser,  Elmer P. "An Evaluation of Environmental Odors, "
       Journal of the Air Pollution Control Association.  VII, No. 4 (February, 1958),
Leonardos, G., Kendall, D. and Bernard, N.  "Odor Threshold Determinations of
       53 Odorant Chemicals," Journal of the Air Pollution Control Association.
       XIX, No. 2 (February,  1969), 91.

Manual on Sensory Testing Methods.  Special Technical Publication No. 434 (Ameri-
       can Society for Testing and Materials,  1968).
Stern, Arthur C. (ed.) Air Pollution.  Vol. HI.  2nd ed.; New York: Academic
       Press, 1968.

-------
                             References (Cont'd)
Sullivan, Ralph].  Air Pollution Aspects of Odors.  Bethesda, Maryland:
       Litton Systems, Inc., 1969.  Draft copy.
Sullivan, D. C., Adams, D. F. and Young,  F. A. "Design of an Odor Perception
       and Objectionability Threshold Test Facility, "Atmospheric Environment.
       11(1968), 121-133.
Summer, W. Methods of Air Deodorization.  Amsterdam:  Elsevier, 1963.
Turk, Amos and Mehlman, Stanley.  "Correlations Between Instrumental and
       Sensory Characterization of Atmospheric Odors, " Correlation of Subjec-
       tive-Objective Methods in the Study of Odors and Taste.  Special Technical
       Publication No. 440, American Society for Testing and Materials, 1968.
Zotterman, Y. (ed.)  Olfaction and Taste.  Vol.1.  New York: Pergamon Press,
       1963.

-------
                                CHAPTER III
                      ODOR PRODUCING POTENTIAL OF
                       INDUSTRIAL ODORANT SOURCES
       This chapter describes the process of relating the quartile ranking of odors,
developed by Engineering-Science,  Inc., to nationwide groupings of industrial plants
known to be potential sources of odorant emissions.  Working on the assumption that
odorant emission varies directly with plant output, Copley International Corporation
developed a ranking system for a measure of plant output to be applied as an addi-
tional weighting factor in establishing this relationship.

       The chapter is divided into two sections: The first describes the selection of
number of employees as a measure of plant output; the second indicates the method
of weighting nationwide groupings of industrial  plants by a set of odor characteristic
coefficients and a measure of relative output.
SELECTION OF NUMBER OF EMPLOYEES AS A MEASURE OF PLANT OUTPUT

       Plant output may be basically defined as the quantity of something produced
from scarce resources for consumption or further production. In order to measure
the output of a plant within a single product industry,  it is only necessary to estab-
lish a standard of quantity and a standard of time as a frame of reference.  Once a
frame of reference is established,  it is a simple matter to compare the outputs of
two or more plants in absolute or relative terms .  For example, one crude oil plant
may produce  100 barrels of oil per day, while a second crude  oil plant produces 200
barrels of oil per day or twice as much as the first plant.

       Within a multi-product industry, however,  the measurement of plant output
becomes more complex when different standards of quantity are established for each
product.  Thus,  the crude oil plant may produce 100 barrels of oil per day and 100
tons of coke per day.

       This problem is discussed by Caleb A. Smith in his "Survey of the Empirical
Evidence on Economics of Scale."  He states that:

       "	The measurement of output is unequivocal only if the output is
       homogeneous.  In practice we do not find either plants  or firms which,
       during a period of growth from small-scale to large-scale, produced
       one homogeneous product, nor do we find a group of plants or firms
       of widely different size which produce  a single homogeneous product.

-------
       Output of plants and firms is in fact heterogeneous to a very substan-
       tial extent.

       "... .When we talk about the size of a plant or firm we ordinarily
       mean its capacity to turn out its entire product mix, not its capacity
       to turn out one  specific product."l

       To simplify the comparison of the outputs in multi-product industries, it is
often convenient to  establish a standard of quantity that relates to some single attri-
bute common to the production of each output.  Indeed, when comparing the output
of plants included in an aggregation such as the Standard Industrial Classification
(SIC) system,  it is necessary to establish such a surrogate standard.

       The most meaningful surrogate standard of quantity would be the average cost
per unit of output.  But, unfortunately, this type of cost information is not readily
available. Competitive and other pressures generally restrain plant managers from
publishing such information.  While average cost per unit estimates are available for
a few types of production  — principally for the production of basic raw materials —
there are none available for the  SIC aggregations used in this study.

       Some information that is published by firms,  especially publicly held corpo-
rations, could possibly be used as surrogate standards of quantity. This information
includes such statistics as sales volume, income generated or value added, the value
of assets  and the number of employees.  It is usually published annually, and fre-
quently, for each quarter  year.

       These statistics have been used by economists to represent plant size as a
measure of industrial concentration in any given market — see, for example,
"Measurement of Industrial Concentration, " by Morris A.  Adelman^ and "Measure-
ment of Concentration, " by Gideon Rosenbluth.   Such statistics could also be used
to represent a measure of relative output of plants within the same industry.  In
either usage, however, they are subject to the following criticisms .

Sales Volume

       If  the mix of products sold varies from plant to plant, sales volume may not
be representative of differences in the levels of output. This would be especially
true if there is a substantial difference in the market price for each of the products
produced.

Income Generated Or,  Similarly, Value Added

       It  is  generally agreed that the best single measure of industrial concentration
is income generated within the plant. This measurement is quite similar to the value
added by manufacture measurement used by the Bureau of the Census  in the 1947
Census of Manufactures .   Unfortunately, this statistic is not published by many

-------
firms.  It could be closely estimated by (1) subtracting total purchases from other
firms from total revenue, or (2) summing total wages, transfer payments for inputs
provided by other company-owned operations, and profit.  But, such things as total
purchases from other firms  and transfer payments to other company-owned opera-
tions are rarely reported separately in firms' income statements .

Value of Assets

       As Adelman states:

       "Assets are the result of accumulation over time.  If prices have
       changed to an appreciable degree, comparisons  among firms (or
       among plants)  or groups of firms (or groups of plants)  are impaired
       by variations in the time pattern of accumulation... .The longer the
       time period over which the assets have been accumulated, the greater
       the uncertainty about their valuation on a common basis ."5

Number of Employees

       One problem with using number of employees to represent a measure of out-
put of plants within the same industry is again related to product mix. Even if the
same products are produced at all plants in an industry, these products may be pro-
duced in different relative proportions at each plant.  The  problem with using number
of employees manifests itself when the quantity of labor required to manufacture a
unit of one product is more or less than that required to manufacture a unit of each
of the other products at a given plant. (A similar criticism may be developed against
using the value of assets, since a unit of one product may require more or less asset
intensity as well as  more or less labor intensity.)

       One final obstacle was encountered in the search for a measure of plant output.
While sales volume, value of assets, and number of employees are published at least
annually by publicly held corporations,  these statistics  are typically an aggregate for
all plants owned by each corporation. Statistics for the individual plants are not pub-
lished.

       One statistic — number of employees — is  collected annually for individual
plants in the United States by Dun & Bradstreet,  Inc.  It is  available to clients along
with other aggregate firm data through Dun's Market Identifiers service.  Therefore,
in spite of the criticisms identified above, for the purpose of this study number of
employees was chosen as the measure of plant output.

-------
METHOD OF WEIGHTING NATIONWIDE GROUPINGS OF INDUSTRIAL PLANTS

       Copley International Corporation obtained a magnetic computer tape from
Dun's Market Identifiers service that contains the names, addresses, five-digit
postal zip codes,  and numbers of employees of approximately thirty thousand indus-
trial plants located throughout the United States.  Each of these plants is included in
at least one of forty-eight SIC classifications of industries known to be potential
sources of odorant emission. Table I lists the number of these plants with a primary
line of business in the corresponding SIC classification.6

-------
                                   Table I
               NUMBER OF PLANTS WITH A PRIMARY LINE OF
            BUSINESS IN A CORRESPONDING SIC CLASSIFICATION
                                                                      No. of
SIC*      	Description	      Plants

2011      Meat packing plants                                           2,434
2015      Poultry and small game dressing and packing, wholesale           518
2031      Canned and cured fish and sea foods                              355
2033      Canned fruits,  vegetables, preserves, jams and jellies          1,321
2036      Fresh or frozen packaged fish and sea foods                       340
2072      Chocolate and cocoa products                                     70
2082      Malt liquors                                                    168
2085      Distilled, rectified, and blended liquors                          151
2095      Roasted coffee                                                  161
2611-    Pulp mills                                                     -108
2621-     Paper mills, except building paper mills                         -481
2631-     Paperboard mills                                              - 279
2812      Alkalies and chlorine                                            114
2815      Cyclic intermediates,  dyes,  organic pigments (lakes and          233
               toners),  and cyclic (coal tar)  crudes
2818      Industrial organic chemicals, not elsewhere classified             747
2819      Industrial inorganic chemicals,  not elsewhere classified           991
2821      Plastics materials, synthetic resins and nonvulcanizable         1,353
               elastomers
2822      Synthetic rubber (vulcanizable elastomers)                       111
2823      Cellulosic man-made fibers                                       52
2824      Synthetic organic fibers,  except cellulosic                         79
2833      Medicinal chemicals and botanical products                       311
2841      Soap and other detergents, except specialty cleaners               787
2842      Specialty cleaning, polishing, and sanitation preparations,        1,822
               except soap and detergents
2844      Perfumes, cosmetics and other toilet preparations               1,078
2851      Paints,  varnishes,  lacquers, enamels and allied products        2,029
2861      Gum and wood  chemicals                                         216
2871      Fertilizers                                                     719
   The forty-eight SIC classifications included in this table do not represent an
   exhaustive list of odorant emitting industries, but rather a cross-section of
   potential sources.

-------
                              Table I  (Cont'd)
                                                                      No. of
SIC*     	Description	      Plants

2911     Petroleum refining                                              593
2951     Paving mixtures and blocks                                       926
3211     Flat glass                                                       87
3229     Pressed and blown and glassware, not elsewhere classified         424
3241     Cement, hydraulic                                              270
3312     Blast furnaces (including coke ovens), steelworks, and            586
               rolling mills
3321     Gray iron foundries                                             972
3322     Malleable iron foundries                                         127
3323     Steel foundries                                                  312
3331     Primary smelting and refining of copper                           36
3332     Primary smelting and refining of lead                             28
3333     Primary smelting and refining of zinc                             39
3334     Primary production of aluminum                                   74
3339     Primary smelting and refining of non-ferrous metals, not          113
               elsewhere classified
3341     Secondary smelting and refining of non-ferrous metals             325
3361     Aluminum castings                                            1,079
3362     Brass, bronze, copper, copper base alloy castings                 687
3369     Nonferrous castings, not elsewhere classified                     512
4731     Stockyards                                                      117
4952     Sewerage systems                                               79
4953     Refuse systems                                               1,442

                                                    Total             25,856
*  The forty-eight SIC classifications included in this table do not represent an
   exhaustive list of odorant emitting industries, but rather a cross-section of
   potential sources.


Source:  Dun's Market Identifiers service.



-------
Groupings of Plants in Postal Zip Code Areas

       As a convenient means of locating nationwide groupings of a cross-section of
industrial odorant sources, the  25,856 industrial plants included in Table I were
sorted by five-digit postal zip code.  It was planned to weight the number of plants
in each of these groupings by a set of odor characteristic coefficients and a measure
of relative output, and then to plot the weighted values on a five-digit zip code area
map of the United States.  This  was to provide a graphical representation of the odor
producing potential of industrial odorant sources in each five-digit area. However,
this plan was necessarily modified.

       In response to an inquiry,  the Post Office Department stated that a five-digit
area map of the United States is not available.  A map of unwieldly proportions would
be required to delineate all of the approximately 38,000 five-digit zip code areas in
the United States. As an  alternative, they recommended use of their three-digit
area map, which outlines the 837 three-digit zip code areas established across the
nation. (A three-digit zip code  area includes all five-digit zip code areas having
the same  first three digits; thus, area 921 would include areas 92101, 92102, 92103,
etc.)  Following this recommendation, the five-digit groupings of industrial odorant
sources were summarized into appropriate three-digit groupings.

       A  ranking of the number of industrial plants in each three -digit zip code area
indicated  that industrial odorant sources are centered principally in the metropolitan
areas of the United States J Table II presents a ranking of the fifty metropolitan
areas having the  largest number of the 25,856 industrial plants included in Table I.
Note that  fifty-eight percent of the plants included in Table I are located in these
fifty metropolitan areas.

       Certain three-digit zip codes were combined to determine the total number
of industrial plants located in each metropolitan area in the nation. The  three -digit
zip codes included in each of the metropolitan areas of Table II are listed in
Appendix B.

-------
                                                   Table II
                             FIFTY METROPOLITAN AREAS IN THE UNITES STATES
                                     HAVING THE LARGEST NUMBER OF
                                  INDUSTRIAL PLANTS INCLUDED IN TABLE I
I
CO
Rank        Metropolitan Area

  1     Los Angeles,  CA
  2     New York, NY
  3     Chicago, IL
  4     Newark, NJ
  5     Philadelphia, PA-Camden, NJ
  6     Detroit, MI
  7     San Francisco-Oakland, CA
  8     Boston, MA
  9     Cleveland, OH
 10     St. Louis, MO-EastSt. Louis,  IL
 11     Houston, TX
 12     Milwaukee, WI
 13     Pittsburgh, PA
 14     Providence, RI
 15     Minneapolis-St. Paul, MN
 16     Portland, OR
 17     Cincinnati, OH
 18     Dallas, TX
 19     Baltimore, MD
 20     Atlanta, GA
 21     Buffalo, NY
 22     Miami,  FL
 23     Seattle, WA
 24     Grand Rapids, MI
 25     Kansas City, KS- Kansas City, MO
No. of Indus-
 trial Plants
      373
      291
      252
      081
      819
      556
      543
      478
      475
      373
      359
      314
      301
      264
      257
      243
      233
      230
      227
      223
      218
      210
      205
      186
      184
Rank    Metropolitan Area

 26     Dayton, OH
 27     Kalamazoo, MI
 28     Indianapolis, IN
 29     Tampa, FL
 30     Louisville, KY
 31     Columbus, OH
 32     New Haven, CT
 33     New Brunswick, NJ
 34     Rochester, NY
 35     Toledo, OH
 36     Santa Ana, CA
 37     White Plains,  NY
 38     Denver, CO
 39     New Orleans, LA
 40     Birmingham, AL
 41     Charlotte, NC
 42     San Jose,  CA
 43     Canton, OH
 44     Syracuse, NY
 45     Fort Worth, TX
 46     Hicksville, NY
 47     Greensboro,  NC
 48     Albany, NY
 49     Memphis, TN
 50     Chattanooga, TN
No. of Indus-
 trial Plants^

   176
   159
   155
   152
   151
   146
   145
   138
   136
   136
   131
   131  -
   129
   119
   117
   111
   111
   109
   108
   106
   104
   101
   100
     98
     93

-------
       The final step in establishing the odor producing potential of industrial odor-
ant sources was to weight the numbers of industrial plants in each three-digit zip
code area by a set of odor characteristic coefficients and a measure of relative out-
put .

Weighting by Odor Characteristic Coefficients

       A set of aggregate odor characteristic coefficients was developed by
Engineering-Science, Inc.,  on the assumption that those industries  emitting the
strongest odorants are most likely to cause odor problems. This effort is described
in Chapter II.

Weighting by Relative Output

       Rather than to assume a single relationship between the levels of odorant
emission and plant output as represented by numbers of employees, two different
weighting functions were developed, which were called "linear" and "linear-modified."

       The linear weighting function is based on the assumption that odorant emis-
sion varies in direct proportion to plant  output. It may be algebraically stated as
foUows.
(linear)
                    =  10 x
       where:   w-  =  the relative output weight applied to plant i of SIC
                  J    classification j.
                 5.  =  the number of employees (representing the output)
                  j    of plant i of SIC classification j.
                E,  =  the number of employees (representing the output)
                  j    of the largest plant of SIC classification j.

For each SIC classification, a minimum weight less than one was given to the plant
with the fewest employees and a maximum weight of ten was given to the plant with
the most employees.

-------
       The linear-modified weighting function is based on the assumption that odor-
ant emission varies in direct proportion to plant output for smaller and medium -
sized plants, but that odorant emission becomes constant for larger plants.  This
reflects the possibility that managers of larger plants are more concerned about
maintaining good relations with nearby residents and businesses; thus, they attempt
to control odorant emission beyond some acceptable level .8   The function may be
algebraically stated as follows:
       (linear-modified) w^  =8, for all plants included in the 20 percent
                           J    having the largest number of employees, and

                                   V
       (linear -modified) w^  v 8 x *—^, for all plants not included in the 20
                           J         1.  percent having the largest number
                                       of employees.
Thus, for each SIC classification, plants were divided into two groups.  The first
group consisted of all plants included in the twenty percent having the largest number
of employees.  These plants were given a weight of eight.  The second group consists
of the remaining plants, with each given a weight ranging linearly from less than one
for the smallest plant to eight for the largest plant in the group.

       Table IE indicates the  distribution of plants within twelve ranges of number
of employees for each SIC classification listed in Table I.  The typical distribution
shows a positive skewness, with a majority of plants having less than 300 employees.
Figure 1 indicates the weights that would be given to a cumulative percentage of a
typically skewed distribution of plants .  Note that the linear weighting function empha-
sizes a small percentage of plants having large numbers of employees,  while the
linear-modified function gives relatively more emphasis to smaller and medium-
sized plants.

-------
CO
                                              Table in
                             DISTRIBUTION OF PLANTS INTO TWELVE RANGES
                           OF NUMBER OF EMPLOYEES - BY SIC CLASSIFICATION
\Range of
Employees
SIC 2011
SIC 2015
SIC 2031
SIC 2033
SIC 2036
SIC 2072
SIC 2082
SIC 2085
SIC 2095
SIC 2611—
SIC 2621 <=
SIC 2631 -
SIC 2812
SIC 2815
SIC 2818
SIC 2819
SIC 2821
SIC 2822
SIC 2823
SIC 2824
SIC 2833
SIC 2841
SIC 2842
SIC 2844
SIC 2851
0-
9
1048
77
74
236
74
23
5
19
40
30
37
23
23
78
233
337
383
35
3
12
119
430
1061
567
755
10-
19
296
35
28
114
43
5
1
19
22
7
21
7
13
35
101
108
196
12
4
3
28
89
230
116
312
20-
39
262
52
33
105
42
3
14
9
19
5
25
18
12
27
86
115
171
11
10
10
20
66
140
74
299
40-
99
267
66
39
131
40
12
20
36
21
3
65
41
7
26
64
122
163
8
4
9
28
44
101
52
215
100-
299
191
136
34
121
26
11
45
20
18
14
113
88
12
18
60
87
148
12
6
12
28
23
36
51
108
300-
599
51
35
7
38
5
0
31
12
8
11
58
31
3
7
32
35
33
4
2
8
7
11
14
19
27
600-
999
22
5
3
11
1
1
11
7
1
11
34
12
7
7
11
17
16
10
2
3
4
5
1
8
4
1000-
2999
20
3
1
8
0
0
11
2
1
9
39
5
6
1
12
9
15
2
8
6
4
4
2
11
4
3000-
5999
5
1
0
1
0
1
2
1
0
1
6
1
0
0
2
1
3
0
4
4
1
0
0
0
0
6000-
9999
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
2
0
0
0
0
1
10,000-
29,999
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
1
30,000
or more
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

-------
                                                    Table III (Cont'd)
OJ
oo
\Range of
Employees
SIC 2861
SIC 2871
SIC 2911
SIC 2951
SIC 3211
SIC 3229
SIC 3241
SIC 3312
SIC 3321
SIC 3322
SIC 3323
SIC 3331
SIC 3332
SIC 3333
SIC 3334
SIC 3339
SIC 3341
SIC 3361
SIC 3362
SIC 3369
SIC 4731
SIC 4952
SIC 4953
0-
9
70
215
116
412
19
190
29
87
140
18
37
4
5
4
13
29
94
393
259
150
45
41
791
10-
19
32
94
53
163
-8
34
17
29
88
13
27
4
0
5
3
10
57
177
133
93
19
10
253
20-
39
22
74
81
89
9
32
15
38
139
15
34
0
4
4
4
13
46
144
102
84
11
2
139
40-
99
12
89
61
48
5
31
32
44
242
13
53
4
7
4
5
15
39
124
90
55
7
1
47
100-
299
14
37
73
7
6
37
107
77
180
29
66
2
5
6
4
11
30
100
38
41
2
1
9
300-
599
5
14
47
3
7
26
12
46
38
16
32
1
3
5
6
6
7
19
3
15
3
0
3
600-
999
1
2
15
0
6
6
1
30
16
6
10
6
0
2
11
5
3
2
1
3
0
0
0
1000-
2999
1
1
16
1
8
11
2
42
9
6
7
S
0
2
12
1
0
7
3
6
0
0
0
3000-
5999
0
0
5
0
0
1
0
25
5
0
0
0
0
0
2
0
0
1
0
0
0
0
0
6000-
9999
0
0
1
0
0
1
0
10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10,000-
29,999
0
0
1
0
0
0
0
12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
30,000
or more
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

-------
                                      Figure 1
                   WEIGHTS GIVEN TO A CUMULATIVE PERCENTAGE
                 OF A TYPICALLY SKEWED DISTRIBUTION OF PLANTS
Applied   10
Weight
                          Linear -Modified
                         Weighting Function
                                                  Linear
                                                 Weighting
                                                  Function
          2 -
          0
                       20
       40         60          80

Cumulative Percentage of Plants
100
             Source:  Copley International Corporation

-------
Application of Weights to Groupings of Industrial Plants

       Two weights were computed for each of the 25, 856 industrial plants included
in the SIC classifications listed in Table I:
                 (linear)  Wj,  = ESI. x  (linear) Wj., and
                            j       J              J
       (linear -modified)  W^.  = ESL x  (linear-modified) w^.
                            J       J                        J
       where:   W;    = the weight applied to plant i of SIC classification j.


                ESL  = the odor characteristic coefficient applied to SIC
                        classification j.
                 j    = the relative output weight applied to plant i of SIC
                  J      classification j.
       Following this,  two weights were computed for the total number of plants of
each SIC classification in a given three"digit zip code area:
                           m
         (linear) W.   -   E   (linear) Wt  , and
                           i =1             J V

                           m
(linear-modified) W.
                           '.    =  E   (linear-modified)  W.
                           Jk    1=i                      'j.
       where:   W.    = the weight applied to the total number of m plants in
                  •*k     SIC classification j,  located in three-digit zip code
                        area k.
                W.    = the relative output weight applied to plant i of SIC
                  i,    classification j, located in three-digit zip code area k.
                   JC

-------
       Finally, two weights were computed for the total number of plants of all SIC
classifications in a given three-digit zip code area:
                                n
                (linear)  Wk =  Z   (linear) W. ,  and
                                             J
                                n
       (linear-modified)  Wj, =  E  (linear-modified) W.
       where:  W,  =  the weight applied to the total number of plants of n
                      SIC classifications, located in three-digit zip code
                      area k.
Weighted Value of Groupings of Plants in Metropolitan Areas

       After computing the weighted values (linear W^ and linear-modified W^) for
each three-digit zip code area in the United States, certain zip codes were again com-
bined to determine the weighted values of industrial plants located in metropolitan
areas.   Table IV presents a ranking of the fifty metropolitan areas having the high-
est linear weighted values . Table V presents a similar ranking of the fifty metro-
politan areas having the highest linear-modified weighted values.

       A comparison of these tables to each other and to the unweighted rankings in
Table II indicates a rather undramatic reordering of the major industrial centers of
the United States.   Of principal note is the difference in the magnitude of weighted
values between Table IV and  Table V. This difference should be anticipated,  however,
because of the typical distribution of plants as indicated in Table III, and because the
linear-modified weighting function was developed to emphasize the smaller to medium-
sized plants.

       Thus, the quartile ranking of odors, coupled with a measure of plant output,
was used to identify those geographical areas having a concentration of industrial
plants known to be potential sources of odorant emissions .  These areas,  as might
be anticipated, are the major industrial centers of the United States .

-------
                                                  Table IV
                            FIFTY METROPOLITAN AREAS IN THE UNITED STATES
                             HAVING THE HIGHEST LINEAR WEIGHTED VALUES OF
                                 INDUSTRIAL PLANTS INCLUDED IN TABLE I
to
 I
                                          Weighted
Rank        Metropolitan Area               Value

  1     Newark, NJ                           873
  2     Chicago, IL                           739
  3     Los Angeles, CA                      723
  4     New York, NY                         673
  5     Philadelphia, PA-Camden, NJ           606
  6     St.  Louis,  MO-EastSt. Louis, IL       398
  7     San Francisco-Oakland, CA             390
  8     Cleveland, OH                         331
  9     Detroit, MI                           316
 10     Boston, MA                           282
 11     Milwaukee, WI                        275
 12     Houston, TX                          254
 13     Portland, OR                          253
 14     Baltimore, MD                        228
 15     Minneapolis-St. Paul, MN              222
 16     Buffalo,  NY                           209
 17     Pittsburgh, PA                         206
 18     Grand Rapids, MI                      198
 19     Cincinnati, OH                        186
 20     Kalamazoo, MI                        182
 21     New Brunswick,  NJ                     181
 22     Kansas City, KS-Kansas City, MO       162
 23     Louisville, KY                        157
 24     Columbus, OH                         156
 25     Toledo,  OH                           153
                             Weighted
Rank    Metropolitan Area     Value

 26     Gary,  IN              153
 27     Atlanta,  GA            150
 28     Tampa,  FL            141
 29     Omaha,  NB            141
 30     Green Bay, WI          139
 31     New Orleans, LA       135
 32     New Haven, CT         134
 33     Anchorage, AK         131
 34     Indianapolis, IN        128
 35     Providence, RI          128
 36     Dayton,  OH            118
 37     Dallas, TX            112
 38     Harrisburg, PA         105
 39     Norfolk, VA            105
 40     San Jose, CA           103
 41     Oshkosh, WI           101
 42     Jacksonville, FL        100
 43     Chattanooga, TN         96
 44     Richmond, VA           95
 45     Seattle,  WA             94
 46     Birmingham, AL         94
 47     Wilmington, DE         89
 48     Miami, FL             87
 49     Springfield, MA         84
 50     Albany,  NY             83

-------
                                                  Table V
                            FIFTY METROPOLITAN AREAS IN THE UNITED STATES
                       HAVING THE HIGHEST LINEAR-MODIFIED WEIGHTED VALUES OF
                                 INDUSTRIAL PLANTS INCLUDED IN TABLE I
CO
 I
                                           Weighted
Rank        Metropolitan Area                Value

  1     Newark, NJ                          30,521
  2     Chicago, IL                          27,539
  3     Pittsburgh, PA                       24,973
  4     Los Angeles,  CA                     23,079
  5     New York, NY                       21,413
  6     Cleveland, OH                       16,961
  7     Philadelphia, PA-Camden,  NJ           15,697
  8     St.  Louis, MO-EastSt. Louis, IL      10,660
  9     San Francisco-Oakland, CA             9,947
 10     Detroit, MI                           9,203
 11     Houston, TX                          7,287
 12     Baltimore, MD                        7,222
 13     Milwaukee, WI                        6,872
 14     Boston, MA                           6,814
 15     Minneapolis-St. Paul, MN               6,741
 16     Louisville, KY                        5,760
 17     Cincinnati, OH                        5,679
 18     Grand Rapids, MI                      5,167
 19     Atlanta, GA                           5,061
 20     Kansas City, KS-Kansas City, MO       4, 895
 21     Buffalo, NY                           4,762
 22     New Brunswick,  NJ                     4,643
 23     New Haven, CT                        4,269
 24     Saginaw, MI                           4,133
 25     DaUas, TX                            4,073
                             Weighted
Rank    Metropolitan Area      Value

 26     Columbus, OH         3,878
 27     Toledo, OH            3,840
 28     Indianapolis, IN        3,732
 29     Gary, IN              3,628
 30     Omaha, NB            3,549
 31     Dayton, OH            3,260
 32     Syracuse, NY          3,026
 33     Madison,  WI           2,963
 34     San Jose,  CA           2,740
 35     Johnson City, TN       2,728
 36     Portland,  OR           2,703
 37     Providence, RI         2,664
 38     Sioux Falls, SD         2,626
 39     Tampa, FL            2,571
 40     Richmond, VA         2,491
 41     Kalamazoo, MI         2,450
 42     Chattanooga, TN       2,417
 43     Albany, NY            2,413
 44     Norfolk, VA           2,312
 45     Memphis, TN          2,291
 46     Miami, FL            2,198
 47     Seattle, WA            2,181
 48     New Orleans, LA       2,124
 49     Canton, OH            2,123
 50     Rochester, MN         2,040

-------
                                    Notes
1.     Business Concentration and Price Policy, National Bureau of Economic
       Research (Princeton:  Princeton University Press, 1955), pp. 215-6.
2.     Review of Economics and Statistics (November,  1951).


3.     Business Concentration and Price Policy, op. cit.


4.     U.S. Bureau of the Census, Census of Manufactures,  1947.


5.     Adelman, Review of Economics and Statistics,  op.  cit.,  p. 9.
6.     The forty-eight SIC classifications included in this study were taken from
       the U.S. Bureau of the Budget, Standard Industrial Classification Manual:
       1967.  They do not represent an exhaustive list of odorant emitting
       industries, but rather a cross-sectipn of potential sources.
7.     For purposes of this study, a "metropolitan area" is defined as a city (or
       contiguous cities) to which the Post Office Department has assigned a
       unique three-digit zip code number(s) plus the surrounding three-digit
       zip code area(s).
8.     It is acknowledged that constant odorant emission may also result from
       production factors or established ordinances.  In accordance with any of
       these reasons for constant odorant emission, "larger" plants were arbi-
       trarily defined as those included in the twenty percent having the largest
       number of employees.
9.     Figure 2 of Chapter V illustrates a quartile ranking of the linear-modified
       weighted values of groupings of industrial plants in each three-digit zip code
       area in the United States .

-------
                                References
Adelman, Morris A.  "Measurement of Industrial Concentration, " Review of
       Economics and Statistics , XXXIII (November,  1951).
Bain, Joe S.  Barriers to New Competition.  Cambridge:  Harvard University
       Press, 1965.
George, Edwin. "Four Comments on "The Measurement of Industrial Concentration':
       III, " Review of Economics and Statistics, XXXIV (November, 1951).
Kaysen, Carl, and Turner,  Donald F.  Antitrust Policy:  An Economic and Legal
       Analysis .  Cambridge:  Harvard University Press, 1959.
Rosenbluth,  Gideon. "Measurement of Concentration, " Business Concentration and
       Price Policy.  National Bureau of Economic Research. Princeton:
       Princeton University Press, 1955.
Smith, Caleb A.  "Survey of the Empirical Evidence on Economics of Scale, "
       Business Concentration and Price Policy. National Bureau of Economic
       Research.  Princeton: Princeton University Press, 1955.
U.S. Bureau of the Budget. Standard Industrial Classification Manual:  1967.


U .S. Bureau of the Census. Census of Manufactures, 1947 .

-------
                                 CHAPTER IV
      EFFECTS OF THE NATURAL ENVIRONMENT ON ODOR PROBLEMS
       The purpose of this chapter is to identify those variables in the natural envi-
ronment which have a significant bearing on the seriousness of an odor problem in a
given area, and to present data relevant to these variables in graphical form for the
contiguous United States.

       As a preliminary qualification to this chapter, it is noted that summary-type
data for the entire United  States may have only limited relevance to the severity of
an odor problem in any given area because of the extremely localized nature of most
odor problems.    The presence of an odor problem in a given area is dependent upon
the existence of an odorant source (or sources), an odor receptor population, and a
transport mechanism to convey the former to the latter.  Variables in the natural
environment act upon this  transport mechanism in two ways which influence the  sever -
ity of the odor problem, by affecting the amount of dilution which takes place and by
affecting the quality characteristics of the odorants emitted.

       The odorant properties of substances emitted to the  atmosphere may be in-
creased or reduced by means of reactions with other components of the atmosphere
such as occurs in photochemical reactions or sorptions onto solid or liquid particu-
lates in the atmosphere.   General meteorological data which are collected routinely
do not provide a basis for  estimating the attenuation of odorant levels.  In reference
to concentration change, however, general meteorologic data provide a basis for
estimating the quantities of dilution air available.  Inasmuch as meteorological data
generated toward assessing the volume of air mass available for dilution and the
exchange rate of the air mass available in that volume provide the only basis for
predicting the fate of odorant emissions in an area, the parameters reflecting those
phenomena (atmospheric stability and wind character) provide the basis for  assess-
ing the effect of meteorology on odorant emissions.

Atmospheric Stability

       As shown in Figure 1, the lapse rate is the  decrease in temperature with
altitude. Normally,  lapse rates are positive,  i.e., air temperature decreases with
altitude, a condition usually associated with "unstable" or vigorously mixed air
masses. However, a negative lapse rate may prevail occasionally over a range of
altitudes. The mass of air which lies within this range is called an inversion layer
and is characterized by a  lack of vertical and horizontal mixing.

       The elevation of the base of an inversion layer above the ground surface is  called
the mixing depth and defines the upper limit of the body of air available for the dilu-

-------
tion of odorants.  It is readily apparent that the lower the base of an inversion layer,
and the longer the period of its persistence, the greater the likelihood of serious
odor problems being  generated within the confined air mass .2

-------
                                Figure 1
                    TEMPERATURE GRADIENTS AND
                       ATMOSPHERIC STABILITY
  IS!
  -a
  3
  o
  i-i
  O
  0)

  I
  <
  0)
  T3
  3
Lapse Rate = A6
             AS
      Negative
       Lapse
       Rate
     Inversion Layer - Stable Air
                                Unstable Air
                               Vigorous Mixing
                                     .t
                                   Mixing
                                   Depth
          Air Temperature  0
Source: Engineering-Science,  Inc.

-------
       Low-level stability conditions for the United States have been described by
C. R. Hosier^ in terms of inversion frequency, and by George C. Holzworth^ in
terms of mean maximum mixing depth.  Inversion frequency is defined as percent
of total hours in which inversions based below 500 feet occur.  Maximum mixing
depth is defined as the maximum diurnal depth of  air column below the inversion
layer, as shown in Figure 1. Both approaches are based on vertical temperature
gradients, information about which is conventionally established on a diurnal basis
by means of radiosonde (balloon-carried transducer)  measurements .

       Both approaches (inversion frequency and  mean maximum mixing depth)
share two limitations. The first limitation is based on the fact that most tempera-
ture soundings are taken at airports and therefore represent conditions at semi-
rural or  suburban areas rather than in built-up city areas.  The heat radiation and
other thermodynamic characteristics of urban areas  differ vastly from those of
rural areas .  As a result, inversions over heavily urbanized areas tend to be based
at somewhat higher altitudes than those found above the surrounding countryside.

       Another limitation to the utility of inversion frequency and mean maximum
mixing depth data is that nothing can be inferred about inversion duration; i.e.,  the
available information is usually not presented in a manner facilitating determination
of inversion duration.  Obviously, seven days of continuous low-level inversion  con-
ditions in a two-week period will have greater impact than inversion conditions which
occur on alternate days over the same period.

       Low-level inversion frequency data for the United States are presented in
Figure 2  for each of the four seasons and for the year as a whole.  Isopleths are in
terms of percent of total hours  in which inversions based below 500 feet are present.

       The data shown in Figure 2 indicate that highest frequencies of low-level in-
versions  generally occur in mountainous regions . For example,  areas having annual
frequencies of 45 percent are located in both the Rocky and Appalachian Ranges . The
lowest levels of inversion frequency are found along the Mid-Atlantic and Gulf Coasts,
and through the Great Lakes.

       Estimates of mean monthly values of the maximum mixing depths attained
during the diurnal cycle are presented in Figures  3 through 14 for each of the
months of the year.  All the values shown are for daylight hours because daytime
mixing depths are consistently higher than those observed at night.

-------
                                  Figure 2

                         INVERSION FREQUENCY FOR:
              (a) Winter,  (b) Spring, (c) Summer,  (d) Fall, (e) Annual
                            (Percent of Total Hours)
 (a)
                                           (b)
(c)
(d)
                       (e)
  Source:  C . R. Hosier, "Low-Level Inversion Frequency in the Contiguous United
          States, " Monthly Weather Review , LXXXIX (November, 1961).

-------
       A summary of inversion frequency data on an annual nationwide basis is re-
ported in Table I.  The range of inversion frequencies is 25-40 percent on an annual
basis along the Pacific Coast, 35-45 percent in the Rockies, 30-35 percent in the
Midwest, 20-45 percent on the Atlantic States, and 5-25 percent in the Gulf States.
(Note that the Atlantic States include Appalachia.)  It is apparent that on a year-
around basis the Rockies and Atlantic States have the greatest frequencies of inver-
sion .

       A summary of monthly mean maximum mixing depth on a nationwide basis is
shown in Table H. Maximum values  of this parameter are found in summer in all
areas of the country, with peak maxima of 3,500 to 4,000 meters occuring in the
Rockies.

       In order to evaluate the atmospheric dilution capacity in a given area as it
relates to atmospheric stability conditions, both mean maximum mixing depth data
and inversion frequency data must be taken into account. The inferences which may
be drawn from these two sets of data are summarized in Table III. It is seen from
Table III that essentially two sets of conditions of mean maximum mixing depth and
inversion frequency occur in the United States:  low/low, typically occuring in  the
Pacific States; and high/high, occuring typically in the Appalachians and Rockies.

       The Pacific Coast presents an interesting anomaly in that low-level inversion
frequencies are among the lowest in the country (see Figure 2) despite the relatively
low mean maximum mixing depth  values indicated in Figures 3 through 14.  This
occurs because the criterion used in  defining low-level inversion conditions (mixing
depth less than 500 feet)  causes ocean-induced inversions based above 500 feet to be
ignored. These inversions are semi-permanent during spring, summer and fall and
are generally based below 2,000 feet."*

-------
                                                    Table I
                                   SUMMARY OF INVERSION FREQUENCY DATA
                                         ON ANNUAL NATIONWIDE BASIS
Season
Winter
Spring
Summer
Fall
Annual
Range of Inversion Frequencies (Percent)
Pacific States
20-50
25-40
20-25
35-40
25-40
Rockies
> 50
35-40
35-40
45-55
35-45
Midwest
20-40
25-30
30-35
25-45
30-35
Atlantic States*
20-45
20-40
10-45
35-45
20-45
Gulf States
5-20
5-25
5-20
5-30
5-25
C/l
CO
       * Atlantic States includes Appalachia.
       Source:  C. R. Hosier,   "Low-Level Inversion Frequency in the Contiguous United States, " Monthly Weather

-------
                                              Table II
                       SUMMARY OF MONTHLY MEAN MAXIMUM MIXING DEPTHS
                                    ON NATIONWIDE BASIS
Month
January
February
March
April
May
June
July
August
September
October
November
December
Range of Mean Maximum Mixing Depths ( 100 meter units)
Pacific States
4-10
6-10
6-16
6-16
6-16
6-16
4-16
4-16
4-12
4-12
4-8
4-8
Rockies
6-10
4-12
14-22
12-26
16-36
18-36
20-40
20-36
12-28
10-18
4-12
4-8
Midwest
4
4
6-10
10-12
10-16
8-18
10-18
10-16
8-12
6-8
2-4
2-4
Atlantic States*
4
4-6
6-8
4-12
4-14
4-14
4-16
4-12
4-12
4-8
4-8
4-6
Gulf States
6-12
4-12
6-12
10-14
10-14
12-16
12-20
12-20
12-14
10-12
8-10
6-10
* Atlantic States include Appalachia.
Source:  George C. Holzworth,  "Estimates of Mean Mixing Depths in the Contiguous United States, " Monthly

-------
                                                      Table HI
                          INTERPRETATION OF ATMOSPHERIC STABILITY DATA WITH RESPECT
                                        TO ATMOSPHERIC DILUTION POTENTIAL
                Relative Value of Parameter
         Mean Maximum
          Mixing Depth
                Frequency of Inversion
                Mixing Depth < 500 Ft.
                       Area Where
                        Conditions
                         Common
         Interpretation
en
en
Low


Low

High
Low


High

High
                                                          Pacific Coast
                                                          Appalachians
                                                          Rockies
             High
                       Low
Continuous low-level inversion based
above 500 ft. low dilution capacity.

Low dilution capacity.

Frequent low-level inversions at
night, with mixing depths usually
high during daytime.

Dilution capacity depends on fre-
quency and duration of low daytime
mixing depths.

High dilution capacity.
        *  These conditions do not commonly occur anywhere in the United States; in most areas normal values are neither
          exceptionally high nor low.

-------
                                 Figure 3

  MEAN MAXIMUM MIXING DEPTHS ABOVE THE SURFACE DURING JANUARY
                (Isopleths Are Labeled in Hundreds of Meters)
                                 Figure 4

                MEAN MAXIMUM MIXING DEPTHS, FEBRUARY
Source:  George C. Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review, XCII (May, 1964).


-------
                                 Figure 5

                 MEAN MAXIMUM MIXING DEPTHS, MARCH
                                 Figure 6

                 MEAN MAXIMUM MIXING DEPTHS, APRIL
Source:  George C. Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review, XCII (May, 1964),

-------
                              Figure 7

                MEAN MAXIMUM MIXING DEPTHS, MAY
                               Figure 8

                MEAN MAXIMUM MIXING DEPTHS, JUNE
Source:  George C. Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review, XCII (May, 1964).

-------
                                 Figure 9

                  MEAN MAXIMUM MIXING DEPTHS,  JULY
                MEAN MAXIMUM MIXING DEPTHS, AUGUST
                                   16*14
Source: George C . Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review,  XCII (May, 1964).

-------
                               Figure 11

             MEAN MAXIMUM MIXING DEPTHS, SEPTEMBER
               MEAN MAXIMUM MIXING DEPTHS, OCTOBER
                                                        •12
Source:  George C. Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review,  XCII (May, 1964).

-------
                               Figure 13

              MEAN MAXIMUM MIXING DEPTHS, NOVEMBER
                               Figure 14

              MEAN MAXIMUM MIXING DEPTHS, DECEMBER
Source:  George C. Holzworth,  "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review,  XCII (May, 1964).

-------
Wind Characteristics

       Estimates of mean hourly wind velocity at an elevation of 50 feet are present-
ed in Figures 15 through 18 for representative months for each of the four seasons .
Most values fall within the 5 to 15 miles per hour range, with highest velocities
occurring along the coasts and across the Great Plains, and lowest velocities occur-
ring in the general vicinity of the Appalachian and Rocky Mountain  Ranges .  The
persistence of surface wind speeds of less than about 7 miles per hour have often
been found detrimental to the processes by which pollutants are diluted in the atmos-
phere ."  Percent frequency of nighttime winds of less than 7 miles per hour is
indicated in Figure 19 for each season and for the year as a whole. Low nighttime
winds are observed to be most common in Rocky Mountain and Appalachian areas,
and least common along the coasts and across the Great Plains.

       As was the case for inversion data, the wind speeds shown  are mostly the
results of airport measurements . However,  studies indicate that such data can be
quantitatively useful to estimate air flow over adjacent cities . For example, studies
conducted at Nashville, Tennessee, indicate that urban wind velocities average about
60 to 70 percent of those measured at a nearby airport.

-------
                               Figure 15

         HOURLY MEAN WIND SPEED ESTIMATES FOR JANUARY,
        CORRECTED TO 50 FEET ABOVE GROUND (Miles Per Hour)
                             Figure 16
           HOURLY MEAN WIND SPEED  ESTIMATES FOR APRIL,
                CORRECTED TO 50 FEET ABOVE GROUND
                           (Miles Per Hour)
Source:  U.S. Air Force, Handbook of Geophysics and Space Environments, ed,
        Shea L. Valley (Cambridge, Massachusetts: Air Force Cambridge
        Research Laboratories, Office of Aerospace Research, 1965).

-------
                             Figure 17

          HOURLY MEAN WIND SPEED ESTIMATES FOR JULY,
      CORRECTED TO 50 FEET ABOVE GROUND (Miles Per Hour)
                             Figure 18
        HOURLY MEAN WIND SPEED ESTIMATES FOR OCTOBER
              CORRECTED TO 50 FEET ABOVE GROUND
                          (Miles Per Hour)
           10
10
Source:  U.S. Air Force, Handbook of Geophysics and Space Environments, ed,
        Shea L. Valley (Cambridge, Massachusetts: Air Force Cambridge
        Research Laboratories, Office of Aerospace Research, 1965).

-------
                                     Figure 19

                  PERCENT FREQUENCY OF NIGHTTIME WIND SPEED
                                17 Miles Per Hour For:
                 (a) Winter, (b) Spring, (c) Summer, (d) Fall, (e) Annual
   (a)
                                              (b)
(c)
(d)
                       (e)
    Source: C. R. Hosier, "Low-Level Inversion Frequency in the Contiguous
            United States, " Monthly Weather Review, LXXXIX (November,  1961)

-------
Weather Bureau Air Pollution Potential Warnings

       The atmospheric variables identified in preceding sections as being major
causes of impaired atmospheric dilution capacity form the basis of the U.S. Weather
Bureau Air Pollution Potential warning system. The criterion used for the issuance
of warnings is a prediction of 36 hours or more continuous duration of simultaneous
low-level inversion conditions, and low mean wind velocities within the mixing
depth.   These warnings represent the synthesis of the effects of all the atmospheric
variables found to have bearing on the dilution capacity of the atmosphere and,  thus,
have unique value to this study.

       Unfortunately,  records of air pollution potential warnings are not a part of
the routinely tabulated data available from the Weather Bureau.  However, a graph-
ical summary has been prepared which, although covering only a limited span of
time, provides a useful picture of the relative atmospheric vulnerability of areas
of the United States to  serious odor problems.  This summary is presented  in
Figure 20.

       The pattern which is shown in Figure 20 is similar to that observed for in-
version and wind data considered separately and reinforces the observation  that the
Pacific Coast and the Rocky and Appalachian Mountain areas are the parts of the
United States most vulnerable to odor problems from a meteorological standpoint.

-------
                                     Figure 20

                  AIR POLLUTION POTENTIAL WARNING PERIODS
                             FOR YEARS INDICATED
  Western United States:
  Five warnings issued from
  October,  1963 to June,  1964
Eastern United States:
Forty-one warnings issued from
August, 1960 to June,  1964
       12
                                     Legend

            n ••• Total number of periods with predicted 36 hours or more
                 continuous duration of simultaneous low-level inversion
                 conditions and low mean wind velocities within the mixing
                 depth, where n is 2, 4, 8, 12, 14,  16,  24, 32 or 40.
Source: George C.  Holzworth,  "Estimates of Mean Mixing Depths in the Contiguous
        United States, " Monthly Weather Review, XCII (May, 1964).


-------
Topography

       Local topography can modify significantly the patterns of atmospheric stabil-
ity and wind characteristics presented above. Because of the localized nature of
such effects, no graphical summary for the entire country is feasible; however,
some principles may be stated for application to specific areas:

       (1)    Inversion frequencies are normally lower for mountain slopes
              and summits than for adjacent valleys .9

       (2)    Valleys tend to channel winds parallel to their axes.l®

       (3)    Differential heating phenomena can cause diurnal variations in
              wind direction, with up-slope winds in daytime and down-slope
              down-valley winds at night.

       (4)    Wind velocities are often higher on ridgelines and passes be-
              cause of volumetric  constriction of the moving body of air.

       (5)    Differential rates of heating and cooling of adjacent land and
              water surfaces can lead to daytime landward winds  and sea-
                                    11
              ward breezes at night.

       Atmospheric stability, wind characteristics, and topography are all impor-
tant factors affecting the perception of odors as a problem. Geographical areas in
the United States that exhibit the least favorable characteristics for odor dispersion
are the Appalachians, the Rocky Mountains, and the Pacific Coast. The Pacific
Coast is included since ocean-induced inversions above 500 feet are semi-perman-
ent during spring, summer,  and fall.

-------
                                    Notes
 1.     Interview with Mr . Benjamin R. Ford, California State Department of Public
       Health, Bureau of Air Sanitation.
 2.     R. A. McCormick,  "Air Pollution Climatology, " Air Pollution, Vol. I, ed.
        A. C. Stern (2nd ed.; New York: Academic Press, 1968), 275-320.
 3.     C. R. Hosier, "Low-Level Inversion Frequency in the Contiguous United
       States, " Monthly Weather Review, LXXXIX (November,  1961), 319-339.
 4.     George C .  Holzworth, "Estimates of Mean Mixing Depths in the Contiguous
       United States, " Monthly Weather Review,  XCII (May,  1964), 235-242.
 5.     Hosier, op. cit.


 6.     McCormick, op. cit., 288-294.


 7.     Ibid., p. 277.
 8.     U.S. Weather Bureau, Research Progress and Plans of the U.S. Weather
        Bureau Fiscal Year 1964, 1965, pp. 93-95.
 9.     McCormick,  op. cit., 288-294.
10.     H. A. Panofsky and B. Prasad, "The Effect of Meteorological Factors On
        Air Pollution in a Narrow Valley, " Journal of Applied Meteorology , VI
        (June, 1967), pp. 493-499.
11.     R. C. Wanta,  "Meteorology and Air Pollution, " Air Pollution, Vol. I, ed.
        A. C. Stern (2nd ed.; New York: Academic Press,  1968), pp.  187-226.

-------
                                 References
Boettger, C . M.  "Air Pollution Potential East of the Rocky Mountains:  Fall 1959, "
       American Meteorological Society Bulletin, XLII (September, 1961),  615-630.
California, Department of Public Health.  Interview with Mr. Benjamin R.  Ford,
       Bureau of Air Sanitation.
Environmental Data Service, Silver Springs, Maryland.  Correspondence with Mr .
       R. W. Schloemer,  Director, Office of Data Information.
Holzworth, George C.  "Estimates of Mean Mixing Depths in the Contiguous United
       States, " Monthly Weather Review, XCII (May, 1964), 235-242.
Hosier, C. R.  "Low-Level Inversion Frequency in the Contiguous United States, "
       Monthly Weather Review, LXXXIX (November,  1964), 319-339.
McCormick, R. A.  "Air Pollution Climatology," Air Pollution, Vol. I.  Edited by
       Arthur C . Stern.  2nd ed.  New York:  Academic Press, 1968.
Niemeyer, L. O .  "Forecasting Air Pollution Potential, " Monthly Weather Review,
       XXXVIII (March, 1960), 88-96.
Office of Civil Defense, Santa Rosa, California.  Interview with Mr. David Heiman.
Panofsky, H. A. and Prasad, B.  "The Effect of Meteorological Factors on Air
       Pollution in a Narrow Valley, " Journal of Applied Meteorology. VI (June,
       1967), pp. 493-499.
San Francisco Bay Area Air Pollution Control District.  Interview with Mr. James
       Sandberg, Chief of Meteorology.

-------
                             References (cont'd)
U.S. Air Force.  Handbook of Geophysics and Space Environments .  Edited by
       Shea L. Valley.  Cambridge, Massachusetts:  Air Force Cambridge
       Research Laboratories, Office of Aerospace Research, 1965.
U.S. Weather Bureau.  Research Progress and Plans of the U .S . Weather Bureau
       Fiscal Year 1964: 1965.
U.S. Weather Bureau.  Interview with Mr. G. Brown, Redwood City, California.
U.S. Weather Bureau,  Interview with Mr. C. R. Elford, California State Climato-
       legist, San Francisco, California.
Wanta, R. C.  "Meteorology and Air Pollution, " Air Pollution.  Vol. I.  Edited by
       Arthur C . Stern.  2nded.  New York: Academic Press,  1968.

-------
                                 CHAPTER V
                   CONCENTRATIONS OF POPULATION AND
           COMBINATION OF POTENTIAL ODOR PROBLEM FACTORS
       As stated in the previous chapter, the presence of an odor problem is depen-
dent upon the existence of an odorant source (or sources) and a transport mechanism
that would convey these odorants to an odor receptor population.  But the problem,
itself, emerges from the reaction of the population affected.  Thus, the extent of the
problem is dependent upon the size of the population affected.

       Based on the above, the initial step toward an assessment of the national odor
problem was undertaken.  The distribution of population in the United States was
examined and areas of concentrated population were located.  Results of the preced-
ing chapters were then used to determine which of these areas were coincident with
areas of industrial odor producing potential and within regions of odorant supporting
atmosphere . This chapter describes these activities .
METHOD OF LOCATING AND RANKING CONCENTRATIONS OF POPULATION

       To achieve graphical consistency with the industrial odorant source informa-
tion described in Chapter HI, it was necessary to locate a source of population data
based on three-digit postal zip code areas so that areas with both large population
and high industrial odor producing potential could be easily identified.  An exhaustive
search proved that recent governmental sub-state population estimates are available
only by Standard Metropolitan Statistical Area and, in some cases, by county and
city.  Few, if any, of these regions closely represent the areas of the three-digit
zip code system. The search did, however, result in the discovery of a useful
commercial source — the Rand McNally Commercial Atlas & Marketing Guide. *

       Aside from a multitude of maps, the Commercial Atlas & Marketing Guide
contains economic, demographic, and marketing information on several types of
geographic subdivisions. Most important for this investigation, population esti-
mates for 1969 were provided on a three-digit zip code basis .2

Size of Population in Postal Zip Code Areas

       For purposes of this  study, a "metropolitan area" is defined as a city (or
contiguous cities) to which the Post Office Department has assigned a unique three-
digit zip code number (or numbers) plus the surrounding three-digit zip code area
(or areas).  This definition reflects the combinations of three -digit zip code areas
for which population data is available in the Commercial Atlas & Marketing Guide.

-------
       Applying this definition to a ranking of the number of persons in each three-
digit zip code area supports common knowledge that the population is concentrated
in the metropolitan areas of the  United States. Table I presents a ranking of the
fifty metropolitan areas having the largest number of persons. It was determined
that forty-nine percent of estimated national population of 201 million persons (as
of January 1, 1969) are located in these fifty metropolitan areas .3  The three-digit
zip codes included in each of the metropolitan areas of Table I are listed in
Appendix B.

-------
                                                Table I

                          FIFTY METROPOLITAN AREAS IN THE UNITED STATES
                       HAVING THE LARGEST NUMBER OF PERSONS (1969 Estimates)
     Rank         Metropolitan Area

       1     New York, NY
       2     Los Angeles,  CA
       3     Chicago, IL
       4     Philadelphia, PA-Camden, NJ
       5     Detroit, MI
       6     Newark, NJ
       7     San Franc is co-Oakland, CA
       8     Boston, MA
 1      9     St. Louis,  MO-EastSt. Louis, IL
c^     10     Cleveland, OH
 1     11     Houston, TX
      12     Minneapolis-St. Paul, MN
      13     Milwaukee, WI
      14     Pittsburgh, PA
      15     Baltimore, MD
      16     Atlanta, GA
      17     Washington, DC-Arlington, VA
      18     Miami, FL
      19     Kansas City, KS-Kansas City, MO
      20     Dallas, TX
      21     Buffalo, NY
      22     Cincinnati, OH
      23     Hicksville, NY
      24     Providence, RI
      25     San Diego, CA
                                              No. of Persons
                                               (in thousands)

                                                  8,946
                                                  7,
                                                  7,
                                                  5,
                                                  4,
                                                  4,
                                                  2,
                                                  2,
                                                  2,
                                                  1,
                                                  1,
                                                  1,
                                                  1,
                                                  1,
                                                  1,
                                                  1,
                                                  1.
 ,468
 ,057
 ,177
 ,693
 ,005
3,281
3,260
 ,490
 ,368
 ,154
 ,998
 ,957
 ,939
 ,924
 ,893
 ,743
 ,703
 ,547
                                                  1,535
                                                    490
                                                    441
                                                    402
                                                    327
                                                  1,300
Rank      Metropolitan Area

 26     Indianapolis, IN
 27     Columbus, OH
 28     Seattle, WA
 29     Tampa, FL
 30     Prince Georges, MD
 31     New Orleans, LA
 32     San Antonio, TX
 33     Denver, CO
 34     Santa Ana, CA
 35     Dayton, OH
 36     New Haven, CT
 37     Greensboro, NC
 38     Phoenix, AZ
 39     Norfolk, VA
 40     San Jose,  CA
 41     Portland,  OR
 42     Rochester, NY
 43     White Plains,  NY
 44     Charlotte, NC
 45     Birmingham, AL
 46     Nashville, TN
 47     Memphis, TN
 48     Louisville, KY
 49     Sacramento, CA
 50     Hartford, CT
                                                 No. of Persons
                                                 (in thousands)
 ,295
 ,246
 ,186
 ,171
 ,137
 ,117
 ,116
 ,115
 ,100
 ,073
 ,067
 ,066
1,058
 ,051
 ,004
 ,001
  968
  934
  934
  932
  918
  907
  898
  884
  869

-------
       It is acknowledged that the size of population in each three-digit zip code
area is not an optimum measure of concentration because of vast differences in the
number of square miles and a lack of information as to the distribution of population
within each three-digit area.  These deficiencies extend to the metropolitan areas,
which are defined above as combinations of certain three-digit areas.  As an example
of the differences in land area, the number of square miles in each of the fifty metro-
politan areas of Table I ranges from 469 square miles for New York, NY, to 28,903
square miles for Phoenix, AZ. (That the size of the Phoenix metropolitan area
approaches an extreme can be seen by noting that the median of this range is 4,214
square miles and that only three of 155 metropolitan areas in the Unites States are
larger.)

Density of Population in Postal Zip Code Areas

       To compensate for these deficiencies,  it was decided to compute the density
of population of each metropolitan area in the United States .  The decision was based
on the premise that odor problems would be more extensive in a metropolitan area
supporting a large and compact, rather than smaller or less dense,  population.
However, an additional difficulty  was encountered using the three-digit zip code
system.  An examination of the geographic features of the metropolitan areas re-
vealed that some are extremely urban in character, while many others are only
partly urban.  Because of the presence of rural and vacant land outside a metropoli-
tan area's primary population center, its overall density could be greatly diluted.
Such a dilution effect may result in an unwarranted conclusion that a more urbanized
area contains greater concentration of population than the partly urbanized areas.
As an example of the dilution effect, note that Phoenix, AZ, with a density of popula-
tion of 1,058 thousand persons/28,903 square miles,  or 37 persons per square mile,
was estimated to be at least seventy percent open range, desert, and mountains.

       For increased comparability of population density, it was decided to abandon
the use of three-digit zip code data for population and land area information. An
ideal basis for such data would be combinations of the smaller five -digit zip code
areas.  But it is anticipated that five-digit zip code data will not be available until
1972, at which time the Bureau of the Census will publish 1970 population statistics
for each of 38,000 possible five-digit zip code areas in the United States.  It is
expected that this information, together with currently available five-digit area data
on the odor producing potential of industrial odorant sources, will provide consider-
able accuracy in locating potential odor problems within metropolitan areas.
Density of Population in SMSA Principal Counties

       In lieu of three -digit zip code areas, Bureau of the Census information was
found to provide 1968 estimates of population based on Standard Metropolitan
Statistical Areas (SMSA's) .5  Although SMSA's also differ greatly in land area,
each SMSA contains a principal county (or, in a few cases, an independent city)

-------
which is urban in character. Estimates of the land area included in each principal
county were obtained from the Rand McNally Commercial Atlas & Marketing Guide .
Using this data,  the population density of each SMSA principal county in the United
States was calculated. Table II presents a ranking of the principal counties of fifty
Standard Metropolitan Statistical Areas having the highest densities of population.

-------
                                                                   Table II
                                 PRINCIPAL COUNTIES OF FIFTY STANDARD METROPOLITAN STATISTICAL AREAS
                                     IN THE UNITED STATES HAVING THE HIGHEST DENSITIES OF POPULATION
          Standard Metropolitan
Rank        Statistical Area
  1     New York, NY
  2     Philadelphia, PA
  3     Washington, DC
  4     Jersey City, NJ
  5     Boston, MA
  6     Baltimore, MD
  7     St. Louis, MO
  8     Newark, NJ
  9     Denver,  CO
 10     Norfolk-Portsmouth, VA
 11     Richmond, VA
 12     Chicago,  IL
 13     Detroit,  MI
 14     Milwaukee, WI
 15     Cleveland, OH
 16     New Orleans,  LA
 17     Paterson-Clifton-Passaic, NJ
 18     Cincinnati, OH
 19     San Francisco-Oakland, CA
 20     Pittsburgh, PA
 21     Indianapolis, IN
 22     Minneapolis-St. Paul, MN
 23     Louisville, KY
 24     Los Angeles,  CA
 25     Anaheim-Santa Ana-Garden
         Grove, CA



Principal County( s)
Bronx, Kings, Manhattan,
Queens, Richmond
Philadelphia
District of Columbia
Hudson
Suffolk
*
*
Essex
Denver
*
*
Cook
Wayne
Milwaukee
Cuyahoga
Orleans
Pas sale
Hamilton
San Francis co -Alameda
Allegheny
Marion
Hennepin- Rams ey
Jefferson
Los Angeles
Orange
No. of
Persons
Per Sq .
Mile

26, 547
15, 755
13, 245
13,078
12,278
11,533
11,247
7,498
7,251
5,930
5,784
5,683
4,430
4,355
3,820
3,095
2,407
2,235
2,233
2,198
1,917
1,862
1,800
1,686
1,612



Rank

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
   Standard Metropolitan
     Statistical Area
Columbus, OH
Dallas, TX
Toledo, OH
Akron, OH
Dayton, OH
Portland,  OR
Atlanta, GA
Bridgeport, CT
New Haven, CT
Providence-Pawtucket-Warwick, RI
Omaha, NB
Hartford, CT
Honolulu, HI
Buffalo, NY
Gary-Hammond-East Chicago,  IN
Rochester, NY
Memphis,  TN
Houston, TX
Wilmington, DE
Nashville, TN
San Jose,  CA
Kansas City, KS - Kansas City, MO
Fort Worth, TX
Tampa, FL
Oklahoma City, OK
Principal County( s)
Franklin
Dallas
Lucas
Summit
Montgomery
Multnomah
DeKalb- Fulton
Fairfield
New Haven
Kent-Providence
Douglas
Hartford
Honolulu
Erie
Lake
Monroe
Shelby
Harris
New Castle
Davidson
Santa Clara
Clay-Jackson
Tarrant
Hillsborough-Pinellas
Oklahoma
No. of
Persons
Per Sq.
Mile
1,467
1,417
1,399
1,356
1,288
1,270
1,255
1,498
1,202
1,197
1,135
1,085
1,055
1,022
1,005
994
950
946
852
846
758
742
734
709
705
 * City( s) independent of any county.

 Source:  U.S. Bureau of the Census, Current Population Reports, Series P-25, No. 432, October 3, 1969,  and

-------
COINCIDENCE OF POTENTIAL ODOR PROBLEM FACTORS

       As an initial  step toward an assessment of the national odor problem, it was
necessary to determine which three-digit zip code areas ranked high in both size of
population and industrial odor producing potential and were located within regions
of odorant supporting atmosphere.  This step also provided basic criteria for the
selection of seven metropolitan areas in which the extent of actual odor problems
was investigated.

       Figure 1 (fold-out) illustrates a quartile ranking of size of population in all
three-digit zip code  areas (termed  "Sectional Areas" by the Post Office Department)
in the United States .6   Note that each three-digit area is uniformly color-coded to
relate to four ranges of population.  The uniformity of color is not meant to indicate
uniformity in concentration  of population throughout each area.  Rather, it reflects
the fact that presently there are no more accurate means of locating concentrations
of population using the zip code system.

       As could be expected, the three-digit areas of the upper quartile (328,000
persons or more)  include many state capitals and other principal cities.

       Figure 2 (fold-out) gives a quartile ranking  of the odor producing potential
of industrial odorant sources in all three-digit zip code areas .  In this figure,  each
three-digit area is uniformly color-coded to relate to four ranges of linear-modified
weighted values.  (As described in Chapter III, the linear-modified weighted value
in each three -digit zip code area is the summation  of the products of the numbers of
industrial plants in any of forty-eight SIC classifications, the applied odor character-
istic coefficient, and relative output.)    Again, the uniformity of color is not meant
to indicate uniformity in concentration of groupings of potential odorant sources
throughout each area.  Instead,  it is used for comparison with the size of population
rankings of Figure 1.

       The coincidence of the upper quartiles of both figures, especially in the
industrial centers  of the United States,  is the result of two basic factors: Since
industrial centers  typically  contain a broad representation of industries known to
be potential sources of odorant emission, the overall odor producing potential is
high; and since industrial centers support a large labor force, the total population
is high.

       For comparison with the preceding figures, the graphical summary of the
relative atmospheric vulnerability of areas in the United States to serious odor
problems (Chapter IV, Figure 20) is reproduced in Figure  3 (fold-out) on a three-
digit zip code map of the United States.  A composite map of this summary of air
pollution warning periods and the upper quartiles of Figures 1 and 2 is presented in
Chapter X with an assessment of the national odor problem. A listing of the num-
ber and name of the  principal transportation hub (termed "Sectional Center") in each
three -digit zip code  area appearing in these figures is provided in Appendix C.

-------
       It is emphasized that the development of this set of potential odor problem
factors was for the purpose of problem identification on a metropolitan area scale.
The availability of additional data (especially population data on a five-digit zip code
basis) is required before potential odor problems can be accurately identified in
terms of neighborhoods.

       Seven metropolitan areas were selected from those judged as odor problem
areas to determine the extent of actual odor problems in different parts of the nation,
The criteria used in selecting these metropolitan areas and the investigations con-
ducted in each are discussed in the following four chapters .

-------
                                                                      Figure 1

                                          A QUARTILE RANKING OF SIZE OF  POPULATION  IN EACH
                                             THREE-DIGIT ZIP CODE AREA IN  THE  UNITED STATES
                                                                                                                                                                                               LEGEND

                                                                                                                                                                                  Population of 328,000 or more
                                                                                                                                                                                  Population of from 150,000 to 327,000


                                                                                                                                                                                  Population of from 56,000 to 149,000


                                                                                                                                                                                  Population of 55,000 or less


                                                                                                                                                                                  State boundaries


                                                                                                                                                                                  Sectional Center boundaries


                                                                                                                                                                                  National Zip Code Area boundaries
      STATUTE MUS

-------
                                                                          Figure 2

                                                 A OUARTILE  RANKING OF LINEAR-MODIFIED WEIGHTED
                                                      VALUE OF EACH THREE-DIGIT ZIP CODE AREA
                                                                  IN THE UNITED STATES

                                                                                                                                                                                         LEGEND


                                                                                                                                                                            Weighted value of 676 or more


                                                                                                                                                                            Weighted value of from 246 to 675


                                                                                                                                                                            Weighted value of from 75 to 245


                                                                                                                                                                            Weighted value of 74 or less


                                                                                                                                                                            State boundaries


                                                                                                                                                                            Sectional Center boundaries


                                                                                                                                                                            National Zip Code Area boundaries

-------
                                                                                                                                              Figure 3
                                                                                                                       AIR  POLLUTION POTENTIAL  WARNING PERIODS
                                                                                                                                     FOR YEARS INDICATED
                                                            Western United States:
                                                            Five warnings issued from
                                                            October 1963 to June 1964
Eastern United States:
Forty-one warnings issued from
August 1960 to June 1964
12
                                                                                                                                                                                                                                                           LEGEND
                                                                                                                                                                                                                                      nBBi^^KTotal number of periods (n = 2, 4, 8, etc.)
                                                                                                                                                                                                                                              with  predicted  36  hours or more con-
                                                                                                                                                                                                                                              tinuous duration of simultaneous low-
                                                                                                                                                                                                                                              level inversion conditions and low mean
                                                                                                                                                                                                                                              wind velocities within the mixing  depth
                                                                                                                                                                                                                                              State boundaries


                                                                                                                                                                                                                                              Sectional Center boundaries


                                                                                                                                                                                                                                              National Zip Code Area boundaries
                                                             STATUTE M1US
                                                                                                                                                                                                      STATUTE MU£S

-------
                                    Notes
1.     "Zip Code Sectional Areas:  Population and Sales Data, " Commercial Atlas
       Marketing Guide (100th ed.; Chicago:  Rand McNally & Company,  1969).
2.     Ibid., pp. 53-6.
3.     Ibid., p. 56.  Because a public opinion survey of households was to
       be conducted as a part of this study, a  similar ranking of number of house-
       holds was developed.  The same metropolitan areas appeared in this ranking
       as in Table I; thus,  to minimize any confusion in the use of population data,
       no further reference to number of households was considered.
4.     Ibid., pp. 53 and 56.


5.     Current Population Reports, Series P-25, No. 432, October 3, 1969.
6.     The Post Office Department refers to a three-digit zip code area as a
       "Sectional Area."  Mail is distributed throughout a Sectional Area from a
       town or city, referred to as a "Sectional Center, " which is the principal
       transportation hub of the Sectional Area.
7.     It was not in the scope of this study to determine whether the linear or
       linear-modified weighted values, as defined in Chapter III, best represent
       the odor producing potential of groupings of industrial plants in three -digit
       zip code areas .  However,  a review of a number of business and industrial
       journals provided some indication that the linear-modified values might be
       the best choice.

-------
                                References
Commercial Atlas and Marketing Guide.  100th ed. Chicago: Rand McNally
       & Company,  1969.
Baier, Martin.  "Zip Code — New Tool for Marketers, " Harvard Business Review
       (January-February, 1967) .
Jones,  Richard A.  "New Developments in Zip Code Marketing, " Sales Management
       (June 10, 1969) .
Post Office Department. Correspondence with Mr. Richard W. Jones,  Customer
       Relations Officer, Bureau of Planning and Marketing, Washington, D.C.
Post Office Department. Interview with Mr. Edward P. White, Postal Service
       Officer, San Diego Sectional Center, San Diego, California.
U.S. Bureau of the Census. Current Population Reports.  Series P-25, No. 432,
       Octobers,  1969.

-------
                                CHAPTER VI
          SURVEY OF LOCAL AIR POLLUTION CONTROL AGENCIES
       This chapter is divided into two major sections.  The first provides a
summary of the development and results of the survey of local air pollution control
agencies conducted by Copley International Corporation. The second provides a
summary of the documentation of odor problems prepared by Engineering-Science,
Inc.,  from discussions with representatives of air pollution control agencies in
seven metropolitan areas.

       Frequent references are made to Appendix D.  It contains a summary of the
survey results; a copy of the questionnaire, the cover letter,  and the follow-up
letter sent to each agency; and several examples of complaint forms used by state
and local air quality control organizations .
DEVELOPMENT OF THE QUESTIONNAIRE AND RESULTS OF THE MAIL SURVEY

       The primary objective of this survey was to identify areas in the United States
where local air pollution control agencies believe that significant odor problems
exist. A secondary objective was to gather professional opinion and information
from  local air pollution control agencies concerning:

       .  The extent and seriousness of existing odor problems .
       •  The means by which they define and assess these problems .
       °  Their opinions of public  reaction to the problems.

Development of the Questionnaire

       In view of the nature of the  information desired and the scope of this investi-
gation, a mail survey was considered the most feasible means of obtaining the data.
Accordingly,  a rough draft questionnaire was developed.  This draft was critiqued
by officials of the County of San Diego,  California,  Department of Public  Health.
Further refinement was attained through discussion and correspondence with the
National Air Pollution Control Administration.

       A final pretest was conducted by mailing a preliminary form to a  sample of
ten local air pollution control agencies  scattered across the nation.  The responses
of these agencies were evaluated for clarity and completeness .  As a result, a  few
minor revisions were made.  The final form was then submitted to  NAPCA and  the
Bureau of the Budget for approval.  A copy of the approved questionnaire  is included
in Appendix D.

-------
       The questionnaire was designed to obtain a broad view of each local air pollu-
tion control agency.  It covered the geographic and budgetary scope of the agency,
the extent of odor problems in its district (nature of source, area and population
affected, degree of seriousness, etc.), the identification and assessment of problems,
the use of odor abatement measures, and the degree of recognition of odor problems
by the local mass media.

       Specifically,  Questions 1 through  3 were designed to determine the geographic
area and number of people in each local air pollution control district.  Question 4
dealt with the local agency's total annual  budget and that portion applied to odor prob-
lems .

       Question 5 asked  for a descriptive list of individual odor problems from the
agency's complaint records.  For each specific odor problem listed, information
was requested on the source of the odor,  the area affected, the number of people
affected, the rate of complaints generated,  the degree of seriousness of the problem,
and its development over time.

       Question 6 focused on the means by which the agency assessed odor problems .
It asked how the public was contacted, what types of data were collected from the
public, what tasks were performed by the technical staff in assessing odor problems,
and what types of equipment were used.  Question 7 asked for information on how
each agency handled the abatement of odor problems .

       Question 8 requested estimates of the number of news items on air pollution
carried by the local mass media. And, Question 9 inquired whether or not any odor
pollution studies had been conducted in the agency's district. The final question
(No. 10)  asked for any additional information that might assist in defining and assess-
ing the nature and seriousness of odor problems in general.

       Questionnaires were mailed to the chief officers of 184 local agencies
selected  from the Air Pollution Control Association 1969 Directory (of) Governmental
Air Pollution Agencies .  All questionnaires  were mailed with a personally addressed
and signed covering letter and a stamped, self-addressed envelope to facilitate the
return.

       Most of the questionnaires were mailed on September 3, 1969. A  follow-up
letter was mailed on September 30 to the  agencies that had not responded  by that
date.  Personal telephone calls were placed about mid-October to those who still  had
not responded.  See Appendix D for copies of the covering and follow-up letters .

       Of the 184 agencies to which questionnaires were mailed, 109 responded — a
response rate of 59 percent.  Of the 109 responses, nine were not usable, primarily
because of a lack of sufficient information.  Of the 75 making no response at all,  the
impression received from the telephone discussions was that they did not  have the
personnel available to complete the questionnaire within the  established time limit.

-------
       The following analysis is thus based on the 100 usable questionnaires which
were completed, wholly or in part, and returned.  These responses represented
almost 55 percent of the total sample .

Results of the Mail Survey

       While there was no way of knowing whether the information received from
the responding agencies is typical of all agencies, the distribution of the responding
agencies by type and size  of district and by geographic distribution suggested that
a highly representative sampling of all agencies was obtained.

       Responses indicated that some of the agencies have a legal jurisdiction con-
fined to corporate city limits, some to single counties, and some to clusters of
counties . However, the most common finding was that of an agency responsible
for a single county, with well over half of the agencies falling into this category
(Appendix D, Table 1. Note that the table numbers are comparable to the question
numbers on the questionnaires).

       The size of the air pollution control districts ranged from a relatively small
2.5 square miles up to a substantial 37, 500 square miles (reported by the Anchorage,
Alaska, agency). The average size of all reporting agencies was computed at about
1,935 square miles per district (Appendix D, Table 2).

       The total population contained in the districts represented by the reporting
agencies was approximately 51 million or about one-fourth of the population of the
United States .  The population of these districts ranged from a low of 30,000 to a
maximum of 4,150,000 with an average population figure per district  of approximately
510,000 (Appendix D, Table 3).

       There was also a very wide range in the total annual budget reported by the
agencies (Appendix D, Table 4A),  with a reported low of $11,000 and  a maximum of
approximately $1,835,000.  The annual budget most agencies reported was around
$200,000.  It is noted that while some agencies reported their budgets for air pollu-
tion control,  as requested, an unknown number of others reported the total budget
of the environmental and health organizations of which they are only a part.  Due to
the lack of consistency in  reporting, the summaries of budget data must be viewed
with reserve.

       Reporting agencies estimated the proportion of their total annual agency bud-
get applied to odor problems in the last fiscal year as exceedingly small.  Almost
half of the agencies stated they devoted less than one percent of their  total annual
agency budget to odor problems, while 70 percent spent less than 5 percent.  Only
four of the 100 responding agencies spent over 25 percent of their budget on odor
problems.  For all agencies, the average proportion of the total agency budget applied
to odor problems in the last fiscal year was approximately 6 percent (Appendix D,
Table 4B).

-------
       Over half of the reporting agencies (52) had a population density in their dis-
tricts that could be considered as light — less than 1,000 persons per  square mile.
About a third of the agencies (31) had a moderate density, with from 1,001 to 5,000
persons per square mile, while  13 indicated a heavier density with more than 5,000
persons per square mile (Appendix D, Table 4C).

       Almost three-fourths of the reporting agencies (72) had a per capita annual
budget of less than $0.50 for air pollution control in their districts .  Only seven
agencies reported spending more than $1.00 per capita.  The average expenditure
for all reporting agencies was estimated at approximately $0 .38 per person.  That
proportion of the budget devoted to odor problems was estimated at an  average of
only $0.02 per person.

       Eighty-five of the agencies reported having one or more odor problems in
their districts.  The average number of complaints received by these agencies over
the last 12 months was approximately 148, although the number of complaints ranged
from two for one agency to 1,600 for another (Appendix D, Tables 5A and 5B).

       However, odor complaints were a relatively small percentage of all air pollu-
tion complaints received by the agencies  in the past 12 months.  Only seven agencies
reported that their proportion of odor complaints (as a percentage of all complaints)
ran over 50 percent.  The average proportion for all reporting agencies was approxi-
mately 24 percent (Appendix D, Table 5C).

       Almost half (40)  of the agencies experiencing odor problems in  their districts
indicated they had no criteria for defining the problems, while most of the other
reporting agencies used criteria involving a greater or lesser degree of subjectivity.
Only six agencies reported using scentometers or other scientific instruments in
defining odor problems (Appendix D, Table 5D).  These findings strongly support
the reported need for improved air quality criteria relating to odors.

       Agencies with odor problems were requested to:

       •  List each specific problem.
       •  Identify its source .
       •  Estimate the number of square miles affected.
       •  Estimate the number of people affected.
       •  Indicate the average number of complaints received each month.
       •  Indicate the degree of seriousness of each problem.
       •  Indicate each  problem's progression over time.

       The agencies responding to this question listed a total of 452 specific odor
problems.  The sources responsible for these problems were classified into 23
categories.  Four of these categories accounted for almost half (48 percent) of the
total number of sources.  These four categories were: rendering/meat packing/
slaughterhouses; chemical/plastic plants;  sanitary land fill/dumps/open burning/

-------
incinerators;  petroleum and natural gas refining/asphalt production (Appendix D,
Table 5E-1).

       Considering the average number of square miles affected per odor source,
motor vehicles were by far the worst offenders, followed by pump/kr aft/wood mills .
Next in rank were sanitary land fill/dumps/open burning/incinerators, followed by
fruit and vegetable processing plants .  Petroleum  and natural gas refining/asphalt
production were  next, followed by animal farms/feed lots/stockyards (Appendix D,
Table 5E-2).

       A somewhat different ranking was obtained by considering the average popu-
lation per odor source in the affected area.  Motor vehicles were again ranked first,
followed by the category which included sanitary land fill/dumps/open burning/
incinerators .  Next in line were pulp/kraft/wood mills;  then steel mills/foundries/
metal works, followed by fisheries and fish processing, and then chemical/plastic
plants.

       Another means of examining these data was to consider the monthly average
number of complaints per odor source (Appendix D, Table 5E-3).  The monthly
average number  of complaints per odor source for all sources  listed was 4.9.  Six
of the odor source categories had higher averages as indicated in Table I.

       As a part of this examination of odor sources, the agencies were asked to
rate each source as to its degree of seriousness on a scale of:  critical, very serious,
serious,  moderately serious, not serious.  They were also asked to note if they felt
each source was: more serious each year, continuously serious, occasionally serious,
less serious each year, not serious.

       Had the odor source categories  contained enough individual sources in each
category (at least 20 listed sources per category)  to permit percentaging over the
five-point scale, the information developed would  have been more useful.  As re-
ported, only nine categories produced large enough samples to be meaningful.
Table II presents a  summarization of these nine categories .

       The nine  categories of Table II are listed by rank in terms of most serious to
least serious. The column labeled "Situation Over Time" ranks each category in
terms of its progressive seriousness (Rank 1 relates to the fastest growing problem).
For example,  pulp/kraft/wood mills are ranked third in terms of overall seriousness,
but rank first in  terms of becoming more serious  each year.

-------
                                  Table I
              ODOR SOURCES WITH ABOVE AVERAGE NUMBER
                       OF COMPLAINTS PER MONTH
               Source Category

Sanitary land fill/dumps/open burning/incinerators
Fisheries and fish processing
Rubber/tires production
Coffee roasting/spices
Chemical/plastic plants
Pulp/kraft/wood mills

Source:  Copley International Corporation
    Monthly Average
 Number of Complaints
    Per Odor Source

        15.4
         9.0
         8.3
         5.6
         5.3
         5.2
                                  Table II
             NINE CATEGORIES OF ODOR SOURCES RANKED BY
           DEGREE OF SERIOUSNESS AND SITUATION OVER TIME
                Source Category

Chemical/plastic plants
Rendering/meat packing/slaughterhouses
Pulp/kraft/wood mills
Steel mills/foundries/metal works
Sewage/human wastes
Animal farms/feed lots/stockyards
Sanitary land fill/dumps/open burning/incinerators
Petroleum and natural gas refining/asphalt production
Commercial: restaurants/dry cleaning plants/etc .

Source:  Copley International Corporation.
  Rank By
 Degree of
Seriousness

    1
    2
    3
    4
    5
    6
    7
    8
    9
  Rank By
 Situation
Over Time

    6
    5
    1
    8
    2
    4
    3
    7
    9

-------
       Another series of questions examined the use of agency staff in handling odor
problems.  About half of the reporting agencies indicated they established contact
with the general public (home owners,  residents, complainants,  etc .) as odor prob-
lems were  brought to their attention. However, almost a third of the agencies said
they established contact with representatives of the offending sources, generally
industrial and business management. Only three agencies said they made random
samplings of the public in order to develop systematic data (Appendix D, Table 6A-1),

       In contacting people concerning odor pollution problems,  35 of the agencies
reported they generally talked to less than 25 individuals, while  only six agencies
reported generally talking to more than 100 . The number of people contacted aver -
aged about 59 for all agencies,  with a wide range of from one to  750 (Appendix D,
Table 6A-2).

       Over half of the responses  indicated  that air pollution control agency person-
nel initiated contact with the general public .  Half of the agencies also reported
having formal procedures for collecting data, while the other half had very informal
procedures (Appendix D, Tables 6A-3 and 6A-4).

       Those agencies with a technical staff available to use in assessing odor prob-
lems indicated a wide range in the type of work performed by the staff. The most
frequently mentioned activities  were routine surveillance and investigation of com-
plaints .  A substantial number of agencies engaged in various forms  of meetings
with representatives of the offending sources wherein corrective recommendations
could be made.  A smaller number of agencies had their staffs engage in instru-
mented field investigations with greater or lesser degrees of sophistication (Appen-
dix D,  Table 6B).

       In a question on the type(s) of test and/or measurement equipment used in
assessing odor pollution problems, the scentometer was mentioned most often —
by 12 agencies (Appendix D, Table 6C).  About half of the agencies listed some type
of test and/or measurement equipment, the  other half stated they had no such equip-
ment.

       Question 6D was open-ended, asking for additional comments on the agencies'
methodologies in identifying and assessing odor problems .  Appendix D,  Table 6D,
presents the verbatim comments made by the responding agencies.

       One observation to be readily made from studying these comments is that
there are two distinct schools of thought, about equally divided in the number of
adherents. One group feels the identification and measurement of odor problems
is a relatively simple one, dependent upon the nose, common sense,  and the number
of complaints.  The other group realizes that there are many inadequacies in the
current methods used to assess odor pollution and express a need for assistance in
this area.

-------
       The great majority of those agencies reporting odor problems takes some
sort of abatement action (Appendix D, Table 7A) .  One of the most popular approaches
to abatement appears to be that of "friendly persuasion."  About a third of the agen-
cies fall into this category, citing the use of conferences,  meetings, requests for
cooperation, and other forms of persuasion.

       Another third of the agencies takes a harder line on abatement.  This group
reports having established requirements for abatement equipment, including after-
burners, filters, scrubbers,  incinerators, aerators, and  condensers. Others
require a change in operating time or changes in production procedures.

       Only a fifth of the agencies takes an enforcement approach to abatement,
resorting to ordinances, regulations, legal actions, and prosecution. In contrast,
a few agencies reported that they investigate and evaluate  odor problems but take
no follow-up action.

       The means currently available to agencies to enforce abatement of odor pollu-
tion problems are few in number, but are used to some extent.  The most frequently
mentioned was the use of air pollution regulations and ordinances, followed by the
use of local nuisance regulations.  A few agencies mentioned recourse to sanitation
and health ordinances.  Well over a fourth of the agencies appeared to have no means
of enforcement available (Appendix D, Table 7B).

       A series  of questions was asked to determine the amounts of attention air and
odor pollution have received in the local mass media (Appendix D, Tables 8A, 8B,
8C and 8D).  Responding agencies estimated an average of eight  stories per month
were run in the major local newspapers in their districts.  They also estimated that
approximately seven stories were aired each month by the local  radio and television
stations in their districts . While the local print and broadcast media tend to give
equal treatment  to air pollution problems on the average,  there is a fairly wide range
of attention when individual districts are considered.

       Responding agencies also estimated that on the average approximately 13
percent of the air pollution stories in print dealt in whole or in part with odor prob-
lems . The comparable figure for radio and television was 8 percent. Again, there
was a very wide range between individual districts.

       Putting these averages into a  time frame, it can be said that newspapers in
the average district run about one story per month on odor problems, while radio
and television carry about one story  every two months .

       The majority of responding agencies indicates that no odor studies have been
conducted in their districts (Appendix D, Table 9),  Of the few agencies which indi-
cated studies had been done, only seven had formal, printed study reports which
they were able to enclose with their questionnaires.

-------
       The last question was open ended, asking for additional comment and informa-
tion (not already covered in the questionnaire) which agency personnel felt NAPCA
should be aware of in relation to defining and assessing the nature and seriousness
of odor pollution problems . Appendix D, Table 10, contains the verbatim comments
of those who responded.

       The contents  of most of these verbatim comments can be easily summarized
into one or more of the following categories:

       •  More work and attention needs to be given to odor measurement and
          testing.

       •  No local air and/or odor pollution laws are available to use in abate-
          ment.

       •  Most odors are of a temporary and minor nature,  are under control,
          and warrant little attention.

       •  More basic study of odors is needed, such as their complex chemical
          makeup, their affect on people,  etc .

       •  Abatement of odors is best accomplished by utilizing nuisance ordi-
          nances .

       •  Most odors arise outside the district boundaries and are beyond
          control.
SUMMARY OF DOCUMENTATION OF ODOR PROBLEMS

       With the assistance of the National Air Pollution Control Administration,
arrangements were made for a team of engineers from Engineering-Science, Inc.,
to meet with air quality control agency representatives in each of the seven metro-
politan areas selected for the investigation of actual odor problems .

       The procedural format of these meetings required a two-day visit to each
area and accomplishment of the following:

        •  Introductory meeting with the local administrators .

        •  Orientation by field investigators to establish a common context to
          facilitate discussion of the situations  where odors were found in that
          area.

        •  A technical level meeting to develop descriptions of the technical in-
          formation available in each agency for documenting odpr situations.

-------
       •  A one-day field investigation of industrial and other sources of odor-
          ants in each area,  at which time additional informal interviews were
          conducted with available agency representatives.

       The orientation sessions with the local administrators were conducted at a
level of informality designed to circumvent the formalized responses and elicit
expressions of the informal feelings.  Structure was limited to the consideration of
certain empirically selected  issues deemed basic to the operation of an air pollution
control agency.

       Each of the two-day studies  served to develop a reservoir of both technical
and non-technical information.  The technical information is discussed in detail in
Chapter VIII.  The non-technical observations of the ESI team, relating to the docu-
mentation of odor problems,  are presented in the balance of this chapter .  This
information was generated within the context of the major functional areas of activity
of the air pollution control agencies visited:  receipts of complaints and investigations,
abatement activities,  technical studies, and public relations.

Complaints and Investigations

       All of the agencies visited maintained records of citizen odor complaints on
some type of standardized form. The vast majority of all complaints are received
by telephone and handled by inspectors or clerical personnel. Information requested
by agencies receiving odor complaints typically includes:  name and address of com-
plainant,  nature and strength of the odor, time and duration of the odor, suspected
source (often obvious in chronic problem areas), and, occasionally, wind direction.
Typical complaint forms are presented in Appendix D.

       Complaint records are filed by a variety of systems, which are summarized
in Table III.

-------
                           Table III
      FILING SYSTEMS FOR AIR POLLUTION COMPLAINTS
          System*
Number of Agencies Using"
By source category
  (industrial, commercial, etc.)

By violation category
  (particulates,  odors, etc.)

By geographical area

By name of violator
           4

           2

           6
 * Systems above refer to primary categorization schemes; within
   each category other systems may prevail.

** Some agencies retain multiple copies and file according to differing
   methods. These agencies are counted more  than once in the above
   table.
Source: Engineering-Science, Inc.

-------
       None of the air pollution control agencies visited maintained separate odor
complaint files.  Those agencies which categorized air pollution complaints accord-
ing to type included odors with other types of complaints such as  "irritation" or
open burning. Moreover, many air pollution complaints involving odors are filed
according to associated phenomena, such as visible emissions.

       Investigations of odor problems by air pollution inspectors are initiated on
the basis of citizen complaints, routine air pollution patrols,  and site visits to
known odor sources to verify compliance with odor abatement arrangements.  The
alacrity with which investigations of odor problems are initiated depends upon the
nature and intensity of the odor reported, the  size of the area affected, the size of
the agency involved, and the time of day in which the odor was first reported.

       Certain odors, such as those associated with natural gas leaks, receive
immediate response in all agencies.  Most large agencies attempt to  dispatch an
inspector to the site immediately upon receipt of an odor complaint, primarily in
order to allow personal assessment of the odor by the  inspector. Radio contact
with inspectors in the field, provided in about half the  agencies visited, greatly in-
creases the ease with which this can be accomplished.  Smaller agencies with
limited staffs investigate odor complaints according to a routine schedule of visits
to various geographical sectors of the area under the jurisdiction of the agency.

       An odor investigation by an inspector usually begins with a visit to the affected
area in an attempt to personally assess the odor.  Personal contacts with the com-
plainant may follow, especially if no odor is perceived by the  inspector. Interviews
by air pollution inspectors of odor complainants require considerable skill in order
to obtain relevant information.  Inspectors noted the following problems:  (a) "cranks"
or citizens registering persistent complaints of odors  not perceptible to neighbors;
(b) descriptions of odors in terms meaningless to anyone but the complainant; and
(c) attribution of odors to sources guilty of problem odor emissions in the past,
whether odors were of similar quality or not.

       The specific information developed by  inspectors in the course of odor inves-
tigations is virtually identical to that cited above for citizen complaints. In most
agencies an inspector's report is either included on the same  form used for a
citizen complaint (Appendix D, Pages D-36 and D-37),  or is recorded on another
copy of the same form. The State of New Jersey is unique in that a special form is
used for odor investigations (Appendix D, Page D-34).   In no agency visited did
inspectors routinely record odor intensities in terms other than verbal, although
scentometers were on hand in two agencies .

       In general, records of odor investigations are filed either with complaints,
as described previously, or by the name of the odor source in files maintained in
conjunction with abatement proceedings, as described  below.

-------
Abatement Activities

       A detailed consideration of the administrative and legal methodology by which
the air pollution agencies effected odor abatement is outside the scope of this study.
However, a few observations relative to the documentation of odor problems are
provided below.

       The first step in the abatement process of odor problems is the identification
of the source.  In most odor situations this is readily accomplished by experienced
enforcement personnel. Source identification is usually followed immediately by
direct contact with the source.  The air pollution inspectors interviewed in the course
of the study often knew plant operating staffs personally for those industries with a
history of odor problems and, in many cases, were able to  identify the specific
reason for the odor emission (certain product runs, chronic process upsets, etc.)
even before contacting the sources .

       The documents generated within the context of abatement activities fall into
four categories:  correspondence between the agency and the violator, records of
hearings, compliance agreements,  and reports of inspectors verifying compliance.
In the agencies visited, these documents were available in a separate file maintained
for the violator involved.  The following types of information were found in the abate -
ment files  inspected in the field technical program:

       (1)     Information on the intensity and areal extent of odor episodes,
              as contained  in complaint investigation reports prepared by
              inspectors.

       (2)     Technical data relative to operating conditions causing odors
              and to process modifications or equipment specified for odor
              abatement.

       (3)     Records of procurement, delivery, and installation of process
              modification  or specialized pollution abatement equipment
              obtained in compliance with abatement orders, together with
              reports of inspectors on the efficacy of such  arrangements .

Technical Studies
       Technical studies may be grouped into the following broad categories: ambi-
ent air quality characterization, air pollution source inventories, meteorological
studies,  and special studies relating to specific types of industries or pollutants.
In general, the agencies visited accorded technical study priorities in the order of
these categories .  Of the eleven agencies visited, all conducted ambient air testing
(although the number of analyses varied) and three conducted source testing.  About
half the agencies had accomplished or participated in meteorological studies related
to air pollution.

-------
       In the context of odor problems, ambient air quality, air quality characteri-
zation, and source testing are of little value except for data compiled relative to
hydrogen sulfide or sulfur dioxide. Meteorological studies, on the other hand,
proved very useful in understanding the nature of odor problems because of the
direct relationships  observed between odor levels in a given area and wind direc-
tion and/or atmospheric mixing conditions.

       Special studies relating to the general  problem of odors had been conducted
in three of the seven metropolitan areas visited. All of the studies consisted of
surveys of odor violations by geographical area, together with data regarding the
nature and source of the odor.  Other factors  considered included meteorology and
measurement methodology.

Access to Odor Documentation
       The air pollution agencies interviewed in the context of this study indicated
similar policies and attitudes toward public access by agency documents .  Citizen
complaints, investigations, records of informal hearings, and source testing data
are almost universally regarded as privileged information, either to protect private
citizens from possible retaliation by neighbors or employers, or to protect indus-
trial secrets .  In general, any information gained by inspectors via privileged  access
to sources of air pollution was regarded as privileged information.

       In all agencies interviewed, ambient air quality data, meteorological studies,
and public relations publications were open to the public .  Official policies toward
access to records of abatement actions  varied, but in practice all of the agencies
were reluctant  to make  such information available to the public, usually citing  a
policy of "trying to work with industry without making headlines . "

       In all but two of the agencies visited, the team of engineers from ESI was
given full access to all records based on the NAPCA support of the  project and  con-
tingent upon assurances that no data regarded as privileged would be divulged in the
study report  or by any other means . Of the two agencies imposing restrictions,
only one specifically denied access to documents desired by the study staff.

Public Relations

       All eleven of the agencies interviewed conducted public relations programs .
The objectives  of public relations efforts were  cited as development of public aware-
ness with regard to air pollution and informing the public of progress being made by
the agency.  Public relations efforts were carried out principally by means of public
speaking engagements before various civic groups, and the distribution of publica-
tions prepared  either by NAPCA or the local agency.

-------
       All of the publications prepared by the agencies interviewed paid scant atten-
tion to odors,  despite the tendency of odor problems to generate intense citizen
concern.  This was apparently due to the tenuous position of the agencies regarding
odor measurement and abatement.

-------
                                CHAPTER VII
           THE SELECTION OF SEVEN METROPOLITAN AREAS FOR
                INVESTIGATION OF ACTUAL ODOR PROBLEMS
       One of the primary objectives of this study was to develop a means of iden-
tifying potential odor problem areas throughout the United States.  This was
accomplished as reported in Chapters II through V.  The most likely odor problem
areas were found to be metropolitan areas with a broad representation of industries
known to be potential sources of odorant emissions. A second objective was to
investigate the extent of actual odor problems in seven of these areas selected from
different parts of the nation.  The results of the investigations were to be projected
to the nation as a whole.  The criteria used in the selection process are outlined in
this chapter.

Initial Selection of Thirty-One Metropolitan Areas

       The initial task in the selection process was to develop a list of potential odor
problem areas from which the seven metropolitan areas  could be selected. Only
those areas included in each of the following rankings were given further considera-
tion:

       ( 1)    The fifty metropolitan areas having the largest number of persons
              (Chapter V,  Table I) .

       ( 2)    The principal counties of fifty Standard Metropolitan Statistical
              Areas having the highest densities of population (Chapter V,
              Table II) .

       ( 3)    The fifty metropolitan areas having the highest linear weighted
              values of industrial plants (Chapter III, Table IV) .

       (4)    The fifty metropolitan areas having the highest linear-modified
              weighted values of industrial plants (Chapter III,  Table V).

       The assumptions which formed the basis of this initial task were:

       (1)    That odor problems would be more extensive in areas support-
              ing large and compact,  rather than smaller or less dense,
              populations.

-------
       (2)    That the linear and the linear-modified weighting functions
              represent relationships between the level of odorant emis-
              sion and plant output, but that insufficient evidence exists to
              permit a best choice.  (It was not within the scope of this
              study to obtain such evidence.)

       (3)    That the areas included on the list would provide adequate geo-
              graphic representation to permit a selection of seven metro-
              politan areas from different parts of the nation.

       (4)    That each of the areas listed would contain a large number of
              industrial plants representing many of the SIC classifications
              discussed in Chapter in,  and that,  consequently, a variety
              of odor problems would be encountered.

       Thus, the initial screening of metropolitan areas eliminated those ranking in
the top fifty in terms of population size but with relatively low population density.
Others were eliminated because they did not appear in both odor producing  potential
rankings.  In this manner, a preliminary list of thirty-one potential odor problem
areas was developed.  It is presented in Table I.

-------
                                Table I
                   PRELIMINARY LIST OF THIRTY-ONE
                   POTENTIAL ODOR PROBLEM AREAS
       Metropolitan Area

New York, NY
Los Angeles,  CA
Chicago, IL
Philadelphia, PA - Camden, NJ
Detroit, MI
Newark,  NJ
San Francisco-Oakland, CA
Boston, MA
St.  Louis, MO - East St. Louis, IL
Cleveland, OH
Houston, TX
Minneapolis-St. Paul, MN
Milwaukee, WI
Pittsburgh, PA
Baltimore, MD
Atlanta, GA
Kansas City, KS - Kansas City, MO
Dallas, TX
Buffalo, NY
Cincinnati, OH
Providence, RI
Indianapolis, IN
Columbus, OH
Tampa, FL
New Orleans, LA
Dayton, OH
New Haven, CT
Norfolk,  VA
San Jose, CA
Portland, OR
Louisville, KY
 Size of
Population
  Rank  .

    1
    2
    3
    4
    5
    6
    7
    8
    9
   10
   11
   12
   13
   14
   15
   16
   18
   19
   20
   21
   23
   25
   26
   28
   30
   34
   35
   38
   39
   40
   48
Linear-Modified
 Weighted Rank

      5
      4
      2
      7
     10
      1
      9
     14
      8
      6
     11
     15
     13
      3
     12
     19
     20
     25
     21
     17
     37
     28
     26
     39
     48
     31
     23
     44
     34
     36
     16
Source: Copley International Corporation.

-------
Final Selection of Seven Metropolitan Areas

       Two additional criteria were used in the final selection of the seven metro-
politan areas: geographic representation and the availability of local air pollution
control agency odor complaint records.

       Geographic representation was required to project the results of the investi-
gations of actual odor problems in the seven metropolitan areas to an assessment of
odor problems in the nation as a whole.  To assure this, the thirty-one metropolitan
areas were categorized by Census Region as indicated in Table II.  Two metropoli-
tan areas were to be selected from each of three  regions and one area from the
remaining region.

       It was planned to locate actual odor problems in each of the areas finally
selected with the assistance of representatives of the local air quality control
agencies in those areas.  Therefore, it was necessary to screen the thirty-one
areas further to determine the degree of interest in the study on the part of agency
administrators and the availability of odor complaint records.

       Odor studies of a more specialized nature were previously conducted in
three of the thirty-one areas of Table I.  Telephone discussions with the agency
directors in those areas confirmed their interest in odor problem research and the
availability of odor problem data.  Early results  of the survey of local air pollution
control agencies (discussed in Chapter VI), followed by additional telephone con-
tacts, indicated interest and the existence of information in four other areas.

-------
                               Table H
           PRELIMINARY LIST OF THIRTY-ONE POTENTIAL
              ODOR PROBLEM AREAS BY CENSUS REGION
Northeast

Boston, MA
Buffalo,  NY
Newark,  NJ
New Haven, CT
New York, NY
Philadelphia, PA - Camden, NJ
Pittsburgh, PA
Providence, RI
North Central

Chicago, IL
Cincinnati, OH
Cleveland,  OH
Columbus,  OH
Dayton, OH
Detroit, MI
Indianapolis, IN
Kansas City, KS - Kansas City, MO
Milwaukee, WI
Minneapolis-St. Paul, MN
St.  Louis, MO - East St. Louis, IL
South

Atlanta, GA
Baltimore, MD
Dallas, TX
Houston,  TX
Louisville,  KY
New Orleans, LA
Norfolk, VA
Tampa,  FL
West

Los Angeles,  CA
Portland, OR
San Francisco-Oakland, CA
San Jose, CA
Source: Copley International Corporation.

-------
       Based on the results of the selection process and the approval of the National
Air Pollution Control Administration, actual odor problems in the following metro-
politan areas were investigated:

       .  Portland, OR
       •  Kansas City, KS - Kansas City, MO
       •  Cincinnati, OH
       .  Buffalo, NY
       •  Philadelphia, PA - Camden, NJ
       •  Tampa, FL
       •  San Francisco, CA

Characteristics of the Seven Metropolitan Areas

       Figures  1,2,3 and 4 indicate the positions of the seven metropolitan areas
(by the arrows) in the distributions of the fifty areas having the largest number of
persons, highest densities of population, highest linear weighted values, and highest
linear-modified weighted values .  As indicated, the seven areas provide a good
representation of each of the distributions .

       Figure 5 shows the locations  of the seven metropolitan areas on  a Census
Region map of the United States .  Geographic representation was achieved by select-
ing two areas in the Northeast, North Central, and West regions and one area from
the South region.

       Table III indicates the number of SIC classifications discussed in Chapter in
that are represented in each of the seven areas .  The number of industrial plants
with a primary line of business corresponding to these SIC classifications is also
indicated.

       Meteorological information was not used in the selection of the seven metro-
politan areas because of the expected short duration of the  investigations in each
area (one to three days for the technical field studies and five to eight days for the
public opinion surveys). It was felt that investigations of longer duration or greater
frequency would be required to relate atmospheric influences to the results obtained.
However, for completeness,  meteorological information was gathered for the days
of the investigations (see Appendix E).  Figures 6 and 7 show the locations of the
seven areas in terms  of the graphical summary presented in Chapter IV, Figure 20,
and the atmospheric areas of the  United States as identified in the Air Quality Act
of 1967.

       The investigations of actual odor problems in seven metropolitan areas were
of two types — technical field studies and public opinion surveys.  Chapter VIII deals
with the technical field program conducted by a team of engineers from  Engineering-
Science,  Inc.  Chapter DC discusses  the public opinion surveys of residents and
businessmen directed by Copley International Corporation.

-------
       Because the investigations took place almost entirely in the principal cities
of the seven metropolitan areas, the next two chapters include many references to
the cities rather than the metropolitan areas themselves.  Appendix E provides data
on the locations of the most serious sources of odors, the neighborhoods affected,
and the public opinion survey test and control areas in each of the cities investigated,

-------
                                  Figure 1
              FIFTY METROPOLITAN AREAS IN THE UNITED STATES
                   HAVING THE LARGEST NUMBER OF PERSONS
   12,000r
   10,000
ra
§   8,000
•5
a
o   6,000
to    '
H
"8   4,000
Ja
|   2,000
z
        0
li
           2    6    10    14    18    22   26    30   34    38    42    46    50
                                                                      Rank
                                   Figure 2
   PRINCIPAL COUNTIES OF FIFTY STANDARD METROPOLITAN STATISTICAL AREAS
      IN THE UNITED STATES HAVING THE HIGHEST DENSITIES OF POPULATION

  30, 000

i 25,000
§
cf 20,000
0)
m 15,000
o
CO
& 10,000
M-l
O
JQ 5, 000
.0





•

»



<


























!',
1 1 1 1 1 1 I 1 1 1 1 1 ( 1 1 1 1 1 1 1 M 1 1 1 I I I i I i i
2 6 10 14 18 22 26 30 34 38 42 46 50
Rank
 Source:  Copley International Corporation.

-------
                                     Figure 3

           FIFTY METROPOLITAN AREAS IN THE UNITED STATES HAVING THE
              HIGHEST LINEAR WEIGHTED VALUES OF INDUSTRIAL PLANTS
1,200
1,000
0)
tl 800
T3
a
§ 600
'53
§j 400
.§
200
0

•
•
•
•








i

\
' « < ,
u in
2 6 10 14 18 22 26 30 34 38 42 46 5
Rank
                                     Figure 4

           FIFTY METROPOLITAN AREAS IN THE UNITED STATES HAVING THE
         HIGHEST LINEAR-MODIFIED WEIGHTED VALUES OF INDUSTRIAL PLANTS
ou, uuu

-------
                                                         Figure 5
                               MAP OF THE UNITED STATES, SHOWING CENSUS REGIONS AND DIVISIONS
          [Alaska and Hawaii are drawn at different scales from conterminous United States and are not shown in their correct relative geographic positions]

-------
                                Table IE
        NUMBER OF SIC CLASSIFICATIONS AND INDUSTRIAL PLANTS
      IN EACH OF THE SEVEN METROPOLITAN AREAS SELECTED FOR
              INVESTIGATIONS OF ACTUAL ODOR PROBLEMS
                                      No . of SIC          No . of Industrial
     Metropolitan Area               Classifications*         Plants*	

Philadelphia, PA - Camden, NJ              43                   819
San Francisco-Oakland, CA                 42                   543
Portland, OR                              37                   243
Cincinnati, OH                            34                   233
Buffalo,  NY                               34                   218
Kansas City, KS - Kansas City, MO          32                   184
Tampa,  FL                               30                   152
  From Chapter III, Table I.
Source:  Copley International Corporation.

-------
                                                  Figure 6

                                AIR POLLUTION POTENTIAL WARNING PERIODS
                                       (Refer to Chapter IV, Figure 20)
        8
Source:  George C. Holzworth, "Estimates of Mean Mixing Depths in the Contiguous

-------
                                             Figure 7
                            ATMOSPHERIC AREAS OF THE UNITED STATES
    WASHINGTO
      COASTAL
        AREA
CALIFORNIA
  OREGON
  COASTAL (,
   AREA
                ROCKY IWOUNTAI
                      AREA
                                            GREAT OA!NS
                                                ARE
                                                                APPA^ACHIA
                                                                       A
MID-ATLANTIC
   COASTAL
    AREA
                                                                          SOUTH FLORIDA
                                                                              AREA
                    Source: U.S. Congress, Air Quality Act of 1967, Public Law 90-148, 90th

-------
                                 References
Holzworth, George C.  "Estimates of Mean Mixing Depths in the Contiguous United
       States, " Monthly Weather Review, XCII (May, 1964), 235-242.
U .S. Bureau of the Census . Statistical Abstract of the United States: 1969.
U.S. Congress.  Air Quality Act of 1967.  Public Law 90-148,  90th Congress,
       S 780, 1967.

-------
                                CHAPTER VIE
                        TECHNICAL FIELD PROGRAM
       This chapter describes the development, activities, and results of the tech-
nical field program conducted by an Engineering-Science, Inc ., team of engineers
in the principal cities of seven metropolitan areas across the United States.  The
chapter is divided into three sections .  The first section presents a detailed discus-
sion of the odor analysis techniques used in the field activities.  Sections two and
three deal with,  respectively, the two-day technical field investigations in each of
the cities and the sensory evaluation of odor studies performed in Philadelphia,
Pennsylvania.

       The objectives of the two-day investigations were to:

       (1)     Develop, at the administrative level, a context for evaluation
              of technical information available in the agency relative to odor
              problem situations.  (The results of this effort are presented
              in Chapter VI along with the results of the survey of local air
              pollution control agencies, conducted by Copley International
              Corporation.)

       (2)     Review,  at the technical level, the technical documentation of
              odor problems available in the agency relative to specific odor
              problem observations as well as to meteorological and other
              conditions conducive to odor problems .  (Documentation of odor
              problem areas in each of the seven metropolitan areas and
              meteorological conditions at the time of the field investigations
              are included in Appendix E.)

       (3)     Identify, through evaluation of information obtained in meeting
              objectives (1) and (2) and by direct field investigations,  areas
              where odor problems exist, the extent of the areas affected, and
              the types of emission causing the concern. (Surveys of public
              opinion on the extent of odor problems were conducted in at
              least one of these areas in each of the cities investigated. These
              surveys are discussed in Chapters IX and X.)

       Midway in the development of the technical field program, it became apparent
that there was a lack of information demonstrating the utility of sensory evaluation
techniques for documenting the time and intensity of odorant emissions .  Under
authorization by the National Air Pollution Control Administration, the technical
program in the Philadelphia area was extended to include a study of the efficacy of

-------
sensory evaluation using odor quality reference standards. The objectives of this
study were to:

       (1)     Apply direct sensory evaluation techniques to document odorant
              emissions from two sources, a rendering plant and a refinery,
              located in Philadelphia, and identified during the two-day study.

       (2)     Compare the responsiveness of the scentometer with the sensory
              evaluation technique.

       (3)     Develop and utilize a capability for sensory evaluation techniques
              as will be necessary for subsequent studies.
ANALYTICAL TECHNIQUES FOR ODOR EVALUATION

       Three generic categories of odor analysis techniques are in use: sensory,
chemical, and instrumental techniques.  Sensory techniques are  subjective methods
relying completely on the human olfactory capacity.  These methods are of primary
interest to this study and are examined in detail in the following pages.  Chemical
and instrumental techniques are multitudinous in number and highly specific for
individual classes or species of odorants within a defined range of odorant concen-
trations .  The limitations in the utility of the latter methods are  that odorified
atmospheres are usually complex mixtures and that ambient concentration ranges
often are very low. In contrast, the human nose is extremely sensitive to literally
thousands of different odor qualities over a very  broad range of concentrations.
Sensitive noses are capable of detecting odors in quantities impossible to discern
and track with commercially available instrumentation or chemical methods.  Because
of these reasons, as well as budgetary constraints, all consideration of analytical
methods for use in the technical field program was directed to  an evaluation of sen-
sory methods.

Sensory Techniques

       Sensory techniques can be classified broadly as fie Id-oriented or as laboratory-
oriented and in terms of whether or not a sampling step is required before the analysis
is done.  Field sensory techniques can be utilized either directly (obviating the need
for sampling) or in conjunction with an intermediate sampling step.  Direct field
sensory measurements are used commonly in evaluating ambient odor situations.
However, direct field sensory measurements are frequently impossible because of
extremes of temperature or if concentrations of noxious  components make the source
intolerable for human exposure. Under such circumstances a sampling step is re-
quired . It is generally implicit that a sampling step is required  in conjunction with
laboratory sensory techniques.

-------
       The requirement of an intermediate sampling step is a major concern in the
accurate determination of odorant levels.  Sampling is required when the odorant
must be diluted, concentrated, warmed, cooled, or otherwise modified, when a large
uniform sample is required,  or when transportation to analytical facilities is required.
Amos Turk^ has pointed out some of the problems associated with sampling:

       (1)     If the odorified sample is warm, then condensation on cooling
              may result in the selective removal of odorant from the vapor
              phase.

       (2)     Between the time of sampling and evaluation,  odorous material
              may be sorbed on the walls of the container or on particulate
              matter in the sample.

       (3)     Chemical changes after sampling may alter the odorant.

These problem areas point out a key disadvantage of sampling and indicate the need
for detailed evaluation of all sampling efforts to avoid alteration of the odor charac-
teristics of the sample.

       Vapor Dilution Technique.  Vapor dilution is the most commonly used form
of sensory analysis. The normal scope of procedure entails the sampling of an
odorified atmosphere,  adjustment of odorant concentrations in the sample by dilution
or concentration, and presentation of the adjusted sample to a single observer or
panel of observers for evaluation of the  threshold odor concentration. Vapor dilution
techniques can be static, continuous, or entail volatilization, depending upon the
manner in which the sample is presented to the observer.

       The vapor dilution technique of foremost interest to this study utilized the
scentometer, a continuous technique device.   The scentometer is of interest
because it combines the sampling and analysis function into a single utilitarian unit
(shown in Figure 1). The device was developed by Norman A. Huey and his colleagues
in the late 1950's and was subjected to considerable evaluation before its introduction
into the commercial market .2  The fundamental purpose of the scentometer was to
provide a ready tool for the measurement of ambient odor intensity to be used by
field air pollution control inspectors as  a means of documenting odor problems.

       The scentometer functions by diluting ambient air samples with purified air
and delivering the mixture to the olfactory tract for appraisal through two nose
pieces,  air being drawn through the instrument by the action of the observer's lungs .
The various dilutions are supplied by a  combination of critical size orifices for ad-
mission of ambient air to the mixing chambers of the scentometer.  Purified air for
dilution purposes is provided by the passage of ambient air through beds of activated
carbon contained in the unit.  The size of the unit is 5"x6"x2.5".  The activated
carbon unit used in the unit is reportedly serviceable for at  least three months when
the unit is subjected to daily use.

-------
        Figure 1
MODEL 1-3 SCENTOMETER
 BARNEBEY-CHENEY CO.

-------
       Under normal conditions of use, the observer breathes through a scentometer
for a minute with all odorous air inlets closed.  This period of time is to refresh or
defatigue the observer.  Then the small odorous orifice which gives the highest dilu-
tion is opened,  and if no stimulus is obtained the next larger orifice is tried until a
positive stimulus is obtained. The strength of the odor is designated as the dilution
at which the observer can first perceive the odor.

       The utility of the scentometer is that it provides a rapid observation at a
specific point in time and location.  The observation so obtained is a function of
three factors: the individual response to an odor stimulus,  the effectiveness of the
unit at the time of observation, and the ambient variability of  the odorified atmos -
phere at the time of observation.  The individual response is related to the observer's
knowledge of odors and level of experience in interpreting the  stimuli.  His receptivity
can vary on an hourly and diurnal basis from day to day and in response to physical
and psychological variables in his daily life. The effectiveness of the unit is related
to the degree of dilution achieved (in part,  a function of consistent operator practice),
the degree of odorant sorption/particle filtration on the activated carbon bed, and
the degree of aggregation of the odorant.  Ambient variability of odorant levels is
related to the degree of instability of dispersion and circulation patterns in the atmos-
phere, to source variability, and in a broader sense to topographic and changing
meteorology conditions. Within this framework of qualifications,  the scentometer
has found its most useful application under the following set of conditions:

       (1)     Day-to-day point evaluation of odor sources, with the instant
              readout providing an opportunity for rapid feedback to effect
              abatement, enforcement, etc., on the odor emitter where
              identifiable.

       (2)     General odor problem identification.

The scentometer was used in the technical field program of this study to accomplish
general odor problem identification in areas having odor problem situations.

       Sensory Evaluation Techniques.  The principal developmental work in the
area of sensory evaluation of atmospheric odors has been done by Amos Turk and
Stanley Mehlman.    Because no system of universally understood odor quality
descriptors or primary odors has yet been developed to describe the large variety
of specific odor types encountered in polluted atmospheres, Turk devised an empir-
ical approach to sensory evaluation that does not entail any assumptions concerning
primary odor types.  The key to Turk's approach is the use of quality descriptors
that are represented by chemically defined odor quality reference standards.  Quoting
from Turk and Mehlman:^

              "In this method, the group of odors to be judged is defined in
       terms of a few (usually three to eight) qualities that seem reasonable
       in the light of subjective associations and chemical analysis.  The

-------
       selections are made by people who are familiar both with the odors in
       question and with the analytical findings, even though the latter may be
       incomplete. Then an odor quality reference standard is made  up to
       represent each quality description. The chemicals used in a reference
       standard represent the best choice available on the basis of odor, sta-
       bility, lack of toxicity, and correspondence with constituents that are
       known or suspected to exist in the odorant.  Each reference standard
       may then be expanded into a dilution scale using a suitable odorless
       diluent.  For convenience and in order not to overload the judges'
       capacity for yielding informative responses, the number of points on
       the dilution scale should correspond to the following relationship:
       number of quality standard x number of intensity standards per quality =
       12 to  36.  Then an odor to be appraised may be described in terms of
       intensities of the various qualities by proper matching with the refer -
       ence standards.

              "The odor standards may be taken from the emission sources
       themselves.  If the odor sources undergo chemical modification before
       they are released to the atmosphere, then the chemical modification
       may be  simulated in the laboratory in making up the standard.  Alter -
       natively, an actual sample of the atmospheric discharge may be taken
       for the standard. The primary requirement for the odor standards is
       that the panel members must be able to distinguish them from  each
       other and to relate each standard to the emission source which it repre-
       sents ."

       The principal advantages of the foregoing method are that the program of
sensory measurements  can be related to the objectives to be met by appropriate
selection of quality reference standards .  The quality reference standards relate
directly to the problem  of concern and can be presented to the observer in a natural
state for his  training and memorization of intensity.  The method has  a broad mobi-
lity, a foremost utility of the method being in field applications  oriented to the docu-
mentation of  dispersion patterns from confined and unconfined source  emissions.
Odor intensity data correlative to the source of concern can be obtained by placing
trained panelists on arcs increasingly distant from the source in the path of  the
dispersion pattern. Direct evaluations of two major odorant emitters  in the
Philadelphia area were carried out using precisely this approach.
TWO-DAY TECHNICAL FIELD INVESTIGATIONS

       Two-day technical field investigations were planned and conducted, chrono-
logically, in Portland,  Kansas City, Cincinnati, Buffalo, Philadelphia, Tampa, and
San Francisco.  Three areas of activity were carried out in these surveys: adminis-
trative-level discussions, technical information documentation, and identification of
specific neighborhood areas and districts where major odor problems occur.

-------
       The approach used in setting up the field investigations was to establish con-
tact with state, regional, and local air pollution control agencies in each of these
cities, using the liaison of NAPCA Regional Directors.  The itinerary developed in
this initial effort is presented in Table I.  After an itinerary was established,  a
framework document outlining the objectives of the two-day visits was sent to  each
agency.  During the meetings held in each two-day visit, the relevance of odor prob-
lems and extent of odor problem areas was identified through direct discussion with
agency technical personnel immediately familiar with the problems . Information
flows were documented with respect to agency response and to management of odor
complaints, odor problem areas were visited,  and street maps were sketched  to
indicate the extent and (using the  scentometer) intensity of the odor  problems.

       A summary of the administrative discussions and technical documentation on
odor problems available in the files of air pollution control agencies in the seven
metropolitan areas investigated is presented in Chapter VI. Maps illustrating odor
problem  areas in each of the principal cities in these areas and meteorological data
for the monthly period in which the scentometer readings were made are presented
in Appendix E . The information descriptive of the odor problem situations in  each
of the cities is summarized in terms of the extent of areas affected  by specific odor
sources.

Areal Extent of Odors

       Observations on the areal extent of perceptible odors in the cities listed
above are summarized in Table II. Refinery installation and chemical facilities
were found to be the most significant sources of odor emissions from the viewpoint
of areal extent. The average area affected by odor emissions from a single refinery
was 10 square miles, and that from chemical facilities, in excess of 5 square miles .
The least significant sources, in  terms of areal extent of odor impact, were render-
ing, paint and varnish, and tanning facilities.  There is a correlation between the
size of facility, category of source, and areal extent of odor impact.  For example,
refinery  and  chemical facilities are generally large complexes with a significant
number of stacks and confined sources of odors, whereas rendering and the other
facilities generally having the least areal impact are usually small facilities with
unconfined or general odor sources .

-------
                                           Table I
                                ENGINEERING-SCIENCE, INC.
                          TECHNICAL FIELD PROGRAM ITINERARY
    Dates
  Metropolitan Area
          Agencies Involved
September 18-19

September 22-23
September 24-26
October  1-3
October  6-7
October  8-10
October 13-22
October 23
Portland, Oregon

Kansas City,  Kansas-
Kansas City,  Missouri
Cincinnati, Ohio
Buffalo, New York
Philadelphia,  Pennsylvania-

Camden, New Jersey
Tampa, Florida
Philadelphia,  Pennsylvania
(sensory evaluation of
odors studies)

San Francisco, California
Columbia-Willamette Air Pollution Authority

NAPCA Region VI
Division of Environmental Sanitation,
   Kansas City, Missouri, Health Department
Air Pollution Control Division, Kansas City,
   Kansas -Wyandotte County Health Department

Cincinnati Division of Heating Inspection and
   Air Pollution Control

Division of Air Resources, New York State
   Department of Health
Air Pollution Control Division, Erie  County
   Health Department
Bureau of Air Pollution  Control, Niagara County
   Health Department

Division of Air Pollution Control, Pennsylvania
   Department of Health
Division of Clean Air and Water, New Jersey
   State Department of  Health
Air Management Services, Philadelphia Depart-
   ment of Public Health

Division of Environmental Engineering,
   Hillsborough County Health Department
Manatee County Health Department

NAPCA Division of Economic Effects Research
NAPCA Region IX
Bay Area Air Pollution Control District

-------
                                Table II
                 AREAL EXTENT OF PERCEPTIBLE ODOR
              TWO-DAY TECHNICAL FIELD INVESTIGATIONS

               Summary of Observations in the Principal Cities
                        of Seven Metropolitan Areas
Type of Industry
or Major Product
Refinery
Agricultural Chemical
General Chemical
Polluted Bay
Sewage Treatment
Granary
Electrical
Tall Oil
Meat Packing
Rendering
Paint and Varnish
Tanning
Number of
Observations
3
1
3
1
1
1
1
1
1
4
1
1
Areal Extent (square miles)
Average
10
6
5
4
2.5
2
1.7
1.25
1
0.8
0.4
0.1
Range
1.5 - 23
-
0.6 - 8.4
-
-
-
-
-
-
0.2 - 1.3
-

Source:  Engineering-Science, Inc.

-------
SENSORY EVALUATION OF ODORS:  PHILADELPHIA STUDIES

       The sensory evaluation studies in Philadelphia were carried out over a week-
long period from October 14-20, 1969.  The initial steps in the studies entailed the
arrangement of laboratory and training facilities for preliminary screening and sub-
sequent training of panelists to be deployed in the field sensory evaluations.  Training
facilities were leased from the University City Science Center, and laboratory facili-
ties were made available in the Monel Institute, both institutions being affiliated with
the University of Pennsylvania and located in Philadelphia.  The training facilities
room was of sufficient size to provide approximately fifty square feet of working
space per trainee, which was adequate for purposes of this effort. The sequence and
scope of activities were as follows:

       (1)     Triangle odor testing with food flavorings (almond,  anise and
              lemon) to discern differences in odor sensitivity between panel
              members under laboratory conditions.

       (2)     Odor panel training sessions with odor quality reference standards,
              followed by field tests to ascertain odor emission dispersion pat-
              terns .   Three field tests were done in the vicinity of two large
              refineries.  Hexadecylmercaptan in mineral oil was used as a
              quality reference standard.  An additional field test was done in
              the vicinity of a rendering plant.  In this case, linseed oil diluted
              in mineral oil was used as the quality reference standard.

       (3)     Calibration runs were conducted to discern differences in odor
              sensitivity between panel members under field odor conditions.
              All panelists were located at the same sites during each run.
              One site was near two large refineries and the second site was
              near a rendering plant.

       (4)     Odor intensity rating tests were carried out as a cross check
              against the triangle odor tests to discern differences between
              panel members under laboratory conditions.

Methodology

       Triangle Odor Tests.   Eighteen candidates, recruited from the League of
Women Voters, City of Philadelphia Chapter, were screened in a series of seven
triangle odor tests on the first day of  the sensory evaluation study. These tests
were conducted according to procedures described by Janet Wittes and Amos Turk.^
Three odorous solution samples were  presented to a candidate at the same time;
two were identical and one was different.  The candidate was instructed to select
the sample which was different from the other two and to record on a score sheet
the code number of  the sample chosen as odd. After the series of triangle tests
were completed, the candidates were  scored by the following procedure:

-------
       (1)    The number of correct answers given by each candidate was
              listed in order, highest score first.

       (2)    The difference between two candidates in the number of correct
              answers was correlated with the probability that at least the
              difference could have been obtained by chance.

       (3)    A decision was made as to the number of panel members re-
              quired from the candidates available.

       (4)    A criterion was selected at which the difference between candi-
              dates was deemed statistically significant (rather than due to
              chance alone). Statistical significance was defined in terms of
              confidence limits.

       Aqueous solutions of almond and lemon extracts and anise were used in the
triangle tests.  The samples were prepared by diluting each odorant approximately
to its threshold level.

       Odor Panel Training Sessions.  Odor panel training sessions were conducted
in the mornings prior to each of the four field tests. Two sets of odor quality stan-
dards were prepared in four dilutions,  each to provide the panelists with a means of
familiarizing themselves with the odors prior to the field activities. Concentrations
were selected empirically to approximate the range of  odor intensities encountered
in the field.  Hexadecylmercaptan diluted into mineral  oil at 0.05, 0.0125, 0.00313,
and 0.0078 percent by volume was used to prepare the  standards simulating refinery
odorant emissions. Undiluted linseed oil and linseed oil diluted into mineral oil at
concentrations of 25,  6.25,  and  1.56 percent were used to simulate rendering plant
odorant emissions. The quality reference standards were prepared in 1-oz. amounts
in 4-oz. flexible plastic bottles. Panelists were instructed to shake each bottle,
remove the cap, squeeze the sides, and sniff the resulting vapor.  By repetition and
concentration, the panelists were able to memorize the intensities of each of the
four dilutions for later reference under field conditions.

       Field Sensory Evaluations.  The field sensory  evaluations were conducted at
two locations using all eighteen of the original candidates as panelists.  (As explained
below, none of the candidates were rejected on the basis of insensitivity to odors.)
On each of the four occasions the candidates were placed randomly in the field.  No
effort was made to locate any of the panelists at the same site twice.

       On each of the field odor  evaluations, the panelists were instructed to com -
mence recording odorant intensities at the same time and to record their observations
at thirty-second intervals for 20 to 30 minute periods.

       The two areas selected for sampling were:

-------
       (1)     An area bounded by Vare Street, Dickinson Street, and 29th Street,
              located northeasterly one -eighth to one mile from the refinery com -
              plexes of Gulf Oil Company and Atlantic-Richfield Oil Company in
              southeastern Philadelphia.  This neighborhood contains primarily
              lower middle-class two-story row houses in a predominately
              black neighborhood.

       (2)     An area bounded by Interstate Highway 1-95 and the North Sewage
              Treatment Plant,  located generally easterly of Keystone Rendering,
              Inc.,  and Mutual Rendering,  Inc., in northeastern Philadelphia.
              The neighborhood can be classified as primarily lower-class
              two-.story row houses and frame houses, many of which have been
              abandoned, and is mixed racially.

       The two areas are shown on the locational map of Appendix E, Figure E-5.
The neighborhood adjacent to the refineries  is located at the northeasterly edge of
the Plain of Confluence of the Schuykill and Delaware Rivers, and is about four miles
northeast of Philadelphia International Airport.  The neighborhood in which the ren-
dering facilities are situated is located on the west bank of the Delaware River about
12 miles northeast of the Philadelphia International Airport.

       Calibration Runs.  One calibration run was carried out in each of the areas
studied in the field sensory evaluations.  The calibration runs were conducted by
stationing all panelists in the same area of the dispersion pattern of odorant emis -
sions from the refineries and the rendering  plants. The panelists began recording
odorant intensities simultaneously and recorded their observations on half-minute
intervals for 20 minutes. Scentometer readings were taken by an ESI engineering
team  member at recorded time intervals during the 20 -minute periods.

       Odor Intensity Rating Tests.  Odor intensity rating tests were conducted at
the end of the field sensory evaluations and  calibration runs to provide a check of
panelist proficiency in odor intensity memorization which could be compared with
the triangle tests. Intensity rating tests were conducted according to procedures
described in Wittes and Turk.?  The test consisted of setting up a series of four
dilutions of quality reference standards.  The panelist was given one of the four
standards and asked to identify its intensity.

       Two series of odor intensity rating tests were given, one with the refinery
odor standards and the second with the rendering plant standards.  In each test the
previously described dilutions were used for each standard.  Each series consisted
of six individual tests in which the panelist was requested to replace the dilution of
standard removed from the series back into the sequence of dilutions.  The panelists
were  scored by the same procedure discussed previously for evaluation of results of
the triangle odor tests.

-------
Results

       Triangle Odor Tests. The data developed in the seven triangle odor tests
are presented in Appendix F.  The results of statistical analysis of these data are
presented in Table in and illustrated in Figure 2. As reported in Table in, six of
the eighteen candidates had  six of seven answers correct, three had five correct
answers, four had four correct answers,  three had three correct answers, and no
results were obtained for two candidates (Numbers 9  and 16). The number of
acceptable candidates can be obtained by selecting the confidence limits at which the
difference in scores between candidates is deemed insignificant. Figure 2 shows a
plot of number of candidates accepted versus confidence  limits,  where the confidence
limits were taken as the  probability that at least the difference could have been ob-
tained by chance.  Based on this criterion, fifteen panelists would have been accepted
at the 10 percentile confidence limits,  nine at 50 percent, three at 90 percent,  and
two at 95 percent.

       For rough screening purposes, Wittes and Turk recommend that confidence
limits be set at about 25  or  30 percent in order to effect  the initial rejection of a
large proportion of the candidates that are likely to be unsuitable.^  At confidence
limits of 30 percent, only twelve of the eighteen candidates would have been accepted.
Because of the difficult circumstances under which the candidates were obtained, no
candidates were rejected on the basis of the triangle tests.  On the basis of the odor
intensity rating tests discussed below, it was observed that no correlation existed
between poorly performing candidates in one test and in the other.  In fact, several
candidates who performed poorly in one test did well  in the other.

       Odor Intensity Rating Tests.   The data developed in the odor intensity rating
tests  are presented in Appendix F. The results of statistical  analysis of these  data
are summarized in Table IV and in Figure 2.  The results of these tests reflect the
experience of several days of training sessions and field surveys. There is a basic
similarity in the results  obtained from tests with both of the quality reference stan-
dards .  Panelist No. 10 performed poorly in both intensity rating tests (Table IV),
whereas  Panelist No. 13  performed well in the rendering standard testing but poorest
in testing with the  refinery standard. Panelist No.  13 was one of the best panelists
participating in the triangle tests (Table III), whereas Panelist No. 10 performed
with mediocrity  in the triangle tests. These results indicate the rather random
performance by  candidates, depending on the type of test.  Only one  panelist, No. 10,
was a consistently poor performer.  Because of the general lack of correlation  be-
tween performance levels of candidates by either of the tests, all candidates were
accepted as panelists and used randomly in the field sensory evaluations.

       The correlation between number of acceptable panelists and confidence limits
at which  the difference in scores  between panelists  is deemed insignificant is shown
in Figure 2.  At the 30 -percent confidence limit fourteen of the panelists were
acceptable by either of the tests.  Generally, the curves for the  refinery odor test-
ing paralleled that for the rendering odor, but both  of these curves indicated that

-------
more candidates were acceptable in the 30 to 70 percent confidence limit range by
the odor intensity rating tests than by the odor triangle tests.  It is believed that the
experience gained by the panelists in the six-day time interval between the triangle
tests and the intensity rating tests is a significant factor explaining the better per-
formance of the panelists in the latter tests.

-------
                                                         Table IE
                                         ANALYSIS OF TRIANGLE TEST RESULTS




Candidates
-
4, 18, 1, 11, 8, 13
6, 7, 12
2, 10, 14, 15
3, 5, 17



Number of
Correct Answers
7
6
5
4
3


Difference In
Number of Correct
Answers
-
0
1
2
3
Probability That
At Least Difference
Could Have Been
Obtained By Chance
(percent)
100
78
39
15
4


Panelists
Accepted
(Total)
0
6
11
15
18
00
co
         Source of procedure for statistical analysis:  Amos Turk and Janet Wittes, "The Selection of Judges for Odor
                Discrimination Panels, " Correlation of Subjective-Objective Methods in the Study of Odors and Taste,

-------
                                   Table IV
               ANALYSIS OF ODOR INTENSITY RATING TESTS




Candidate

Score
(Highest Score
Reflects Worst
Performance)
Probability That
At Least Difference
Could Have Been
Obtained By Chance
(percent)


Panelists
Accepted
(total)
Refinery Quality Reference Standard
5
1, 2, 11, 14
3, 15, 18
6, 8
16
7, 9
4
-
-
-
-
10
-
13
0
1
2
3
4
5
6
7
8
9
10
11
12
13
100
87
79
71
63
56
-
-
-
-
-
22
-
15
1
5
8
10
11
13
14
-
-
-
-
15
-
16
Rendering Quality Reference Standard
_
-
7, 9, 11, 13, 18
2, 5, 6, 16
3, 4, 8
14, 15
1
-
-
-
10
0
1
2
3
4
5
6
7
8
9
10
100
86
77
68
59
30
25
20
16
13
10
_
-
5
9
12
14
15
-
-
-
16
Source of procedure for statistical analysis:  Amos Turk and Janet Wittes, "The
       Selection of Judges for Odor Discrimination Panels, " Correlation of Subjective•
       Objective Methods in the Study of Odors and Taste, Special Technical Report
       No. 440 (American Society for Testing Materials,  1968).

-------
                                   Figure 2
                    SCREENING TESTS FOR SIGNIFICANT

                       DIFFERENCE IN PERFORMANCE
•d

-------
       Field Sensory Evaluation of Refinery Odor Emissions.   Dispersion patterns
for odorant emissions in the neighborhood of the refineries are plotted in Figures 8
through 13 for each of the three field sensory evaluations done in the area. The
three runs were carried out on October 16,  1969 at noon and 3:00 p.m. and on
October 17 at 1:00 p .m. Weather observations during each of these monitoring
periods are reported  in Table V,  as recorded at the Philadelphia International
Airport, located about four miles  southwest of the Vare Street study area across
the broad flat plain at the confluence of the Schuykill and Delaware Rivers (see
Appendix E, Figure E-5).  It is assumed that airport data were  applicable in describ-
ing weather conditions at the monitoring site.

       Dispersion patterns for odorant emissions in the vicinity of the rendering
plants are plotted in Figures 14 and 15 for the field sensory evaluation done on
October 17, 1969 at 1:00 p.m.  Weather observations from the Philadelphia Interna-
tional Airport during  this monitoring period are also presented in Table V.  The
Philadelphia International Airport  is located about 12 miles southeast and on the
same (west) bank of the Delaware  River as the area  in which the rendering facilities
are located.  It is assumed, because of the topographical continuity  between the air-
port and the neighborhood of the rendering facilities, that airport weather data were
indicative of weather  conditions at the monitoring site.

       Dispersion patterns are plotted in Figures 8  through 15 using two types of
contours: those based on the mean intensities of the observations of the individual
panelists, and those based simply on the percent of the total observations of each
individual in which odor was perceived.

       Figures 8 and 9 illustrate clearly the first dispersion pattern developed for
refinery odor emissions.  The contours point to the  location of a major emission
source westerly of Morris Street and a generally unconfined source  paralleling Vare
Street. The dispersion pattern in Figure 9 indicates generally the presence of a
second source southwesterly of New Hope Street.  The 100 percent contour parallels
the 3 MIU (mean intensity unit) contour on Morris Street,  and the 80 percent contour
is similarly paralleled by the  2 MIU contour.  The 0.50 MIU contour extending along
Tasker and 30th Streets is somewhat correlative with the  40 percent contour.  Based
on this observation, it appears that the MIU contours and the percentile contours both
provide a basis for describing the extent, but not the magnitude, of the dispersion
pattern.

       Figures 10 and 11  illustrate dispersion patterns for a second run at the refin-
ery area done about three hours after the first runs (on October 20,  1969).  The
wind direction had changed from 200 degrees previously to 190 degrees, as was
evidenced by the different dispersion patterns observed in the second run. The
contours in both Figures 10 and 11 point to two distinct sources  of odor emissions,
one due westerly of Morris Street and the second southwesterly  of New Hope Street.
The greatest MIU contours are 3 at Morris Street and 2 at New Hope and Vare
Streets .  The MIU contours decrease rapidly to less than  1 within 1,000 feet of Vare

-------
       Calibration Runs.   Data developed in the calibration runs represented an
opportunity to evaluate the variability of response of the panel members under con-
ditions of an ambient odor environment. The observation sets developed by each
individual in each run were analyzed for mean value. These sets of information have
been analyzed as a basis for characterizing the performance variability of the panel-
ists .  A histogram for the mean intensities developed in the mercaptan calibration
run is presented in Figure 3.  A similar histogram for mean intensities developed
in the rendering calibration run is presented in Figure 4. Each figure contains the
following information on the histogram depicted: mean and median value of distribu-
tion, maximum frequency (number of panelists occurring in the range of mean
observations or sample size), and the intensity at which the maximum frequencies
occurred.

       The mean and median intensities observed by the panelists as a group were,
respectively,  2.34 and 2.88 for the mercaptan calibration run (Figure 3). Three
individuals had mean intensity observations of 3.25.  The mean intensities of the
individuals were grouped into two sets, one set varying between mean intensities of
0.50 to 1.25,  the second between 2 .25 to 3.75. The mean intensity of all panelists
for the rendering calibration run was 2.23 and the median intensity was 2.50
(Figure 4).  Three individuals had mean intensity observations of 2.00,  and three
additional individuals had mean intensity observations of 2.75. It is apparent from
comparison of the histograms in Figures 3 and 4 that the distribution of mean inten-
sities in the rendering run was much more proximate to a normal Gaussian distribu-
tion than was the distribution of mean intensities observed in the mercaptan. On the
basis of these histograms, it is apparent that the five panelists in the mercaptan
runs with mean intensities varying between 0.50 and 1.25 responded at significantly
lower levels than did their eleven counterparts with mean intensities of 2.25 to 3.75.
There is no immediate explanation for this unexpected division other than that a
similar but less defined spread was observed in the laboratory odor intensity rating
tests reported in Table IV.

       Based on the  above analysis of the calibration runs and the previous discus-
sion of the laboratory odor intensity rating tests, it is apparent that both the
laboratory and field  testing with the mercaptan provided a more stringent evaluation
of individual capabilities than did the rendering odor tests .

-------
                                     Figure 3
                  HISTOGRAM FOR MERCAPTAN CALIBRATION RUN
                               - MEAN INTENSITY -
       Total Data Points  =  16

       Mean =  2.34

       Maximum Frequency  = 3

       Maximum Frequencies at: 3.25
                Sample Size  = .25

                Median  =2.88
  0
  0.50
  i.oo  H
•55 i.so  H
8 2.00  -

| 2.50

  3.00
  3.50  -
  4.00  -1
                     20
                      i
   Number of Persons
 1                      2

   Percent of Maximum
     40          60
	I	I	
80
 I
100
       Source: Engineering-Science, Inc.

-------
                                     Figure 4
                  HISTOGRAM FOR RENDERING CALIBRATION RUN
                               - MEAN INTENSITY -
       Total Data Points = 16

       Mean = 2.23

       Maximum Frequency = 3

       Maximum Frequencies at:  2 and 2.75
                          Sample Size = .25

                          Median = 2.50
         0

         0
  0

  0.50   -

>> 1.00
•*-4
I 1.50
  2.50
  3.00

  3.50
20
 I
     Number of Persons
1                      2
     Percent of Maximum
     40          60         80
       I	I	    i
 3

100
       Source: Engineering-Science, Inc

-------
       Additional statistical analysis of the calibration runs was done to determine
the number of panelists performing at the same level of intensity observation over
a range of confidence limits.  A matrix of "t" statistics was developed to compare
the statistical significance of the difference between the data distributions of all
pairs of observers.  The matrix contained 105 pairs of observations.  Figure 5
shows the relationships between the number of observer pairs and the confidence
limits at which the difference  in performance levels of the individual would be
statistically insignificant.  At a confidence limit of 90 percent, only two percent of
the observer pairs tested with the rendering odor and three percent of the pairs
tested with the mercaptan odor could be accepted by this criterion.  At confidence
limits of 50 percent the acceptance levels increased to 14 percent with the rendering
odor and eight percent with the mercaptan odor. At confidence limits of 30 percent,
the acceptance levels increased to 23 percent with the rendering odor and 14 percent
for the mercaptan odor. It is apparent from a comparison of relationships presented
in Figures 2 and  5 that the calibration runs were a much more severe test than the
laboratory tests in that a much lower level of acceptance is obtained at a given con-
fidence  limit when observer pairs are evaluated under field conditions rather than
individual panelists under laboratory conditions. It is also apparent from Figure  5
that the panelists participating in the calibration runs had considerably more diffi-
culty with the mercaptan odors than with the rendering odor.   The opposite effect
was observed with the intensity rating tests (Figure 2) in which the individual
panelists performed slightly better with the mercaptan odors than with the render-
ing odor.

-------
                                   Figure 5
<1
      t TEST OF SIGNIFICANCE IN DIFFERENCE IN FIELD ODOR CALIBRATION
        RUNS MEANS OF DATA DISTRIBUTIONS BETWEEN OBSERVER PAIRS
                   I      I      I     I
2     5    10     20
  Confidence Level at Which Difference in Means of Data
 Distributions of Observer Pairs is Insignificant (percent)
                                                                       90
>90

-------
       Scentometer readings were also made during the calibration runs.  These
readings are plotted against mean panel odor intensities for the mercaptan odor in
Figure 6 and for the rendering odor in Figure 7.  The best-fit lines (least-squares
method) correlating the scentometer and mean panel readings are described by the
following equations:

       Mercaptan odor:     S  =  1.69 P  -  2.62                     (Equation 1)
       Rendering odor:     S  =  2.63 P  -  4.87                     (Equation 2)

where:

       S = scentometer reading at time t
       P = mean panel odor intensity at time t

       The spread in mean odor panel observations at each of the scentometer read-
ings is caused by two factors:

       (1)     The sensitivity of the  scentometer differs significantly from the
              aggregate direct sensory response of the panelists as intensity
              of odor is varied.

       (2)     Single  scentometer readings were made for each set of panelist
              responses used to  calculate mean intensities. The single scent-
              ometer reading is  subject to individual response variabilities as
              discussed in Chapter II. Within these qualifications Equations  1
              and 2 provide a tool for conversion of intensity readings made by
              the panelists (trained  with the above-described quality reference
              standards) into equivalent scentometer readings.

       Equations 1 and 2 indicate that at low panelist readings the scentometer lacks
the sensitivity of readout that can be provided by panels (as  evidenced by the negative
y-intercepts of these  equations).  From the  viewpoint of ambient odor intensity
monitoring, an aggregate responsiveness to low level odor intensities is a foremost
requisite,  and it is apparent from the calibration runs that direct sensory evaluation
with a team of panelists can provide  a useful level of sensitivity in those instances
where it is required.

-------
Co
              1.50
                                                   Figure 6
                            COMPARISON OF SCENTOMETER AND ODOR PANEL DATA
                                     FOR MERCAPTAN CALIBRATION RUNS
                     I    I     I    I     I     I     I     I     I     I
2.00
2.50
3.00
                           Mean Panel Odor Intensities for Mercaptan Calibration Run (P)
                                     (All Panel Members In Same Location)

-------
                                                 Figure 7
                           COMPARISON OF SCENTOMETER AND ODOR PANEL DATA
                                  FOR RENDERING ODOR CALIBRATION RUN
W
9
1
0)
 P!
 
-------
Street (Figure 10).  The percentile contours (Figure 11) were maximal at 80 percent
in the vicinity of Mifflin and Vare Streets and along Morris Street.  The percentile
contours do not parallel the MIU contours as was observed in the first run; the 60
and 80 percent contours extend easterly along Morris Street to 29th Street.  It
appears from these results that the intensity level and frequency of occurrence of
odor perception did not correspond in the second run as occurred in the first run.

       Figures 12 and 13 show dispersion patterns for the third  run at the refinery
area done four days after the first and second runs.  The wind direction on the day
of the third run was to the northeast (240 degrees), and the wind speed was 12 knots
or twice that of the day of the other runs. The contours of Figures 12 and 13 identify
two discrete  sources as noted previously. The maximum MIU contours were 2 along
Vare Street near Morris Street as well as at New Hope and Vare Streets. the MIU
contours decreased to 0.5 MIU within 1,000 feet of Vare Street.  As was the case in
the first run, the percentile contours  of Figures 12 and 13 paralleled generally the
MIU contours. The maximum percentile contour was 100 percent at Vare Street
near Morris  Street areas and 80 percent along Vare Street between New Hope and
29th Streets.  In this run, the percentile and MIU contours both provided a  similar
basis for describing the extent of the dispersion pattern.

       Figures 14 and 15 illustrate the dispersion pattern observed in the neighbor-
hood of the rendering facilities on October 17, 1969.  The wind direction was from
the northwest and the wind  speed was  14 knots.  The dispersion patterns evidence
clearly the locations of the sources.  The maximum MIU contour of the easterly
plant (Figure 14) was 2; the maximum MIU contour relative to the westerly facility
(located outside of the survey area) was 1. The maximum percentile contour was
100 percent at the easterly plant (Figure 15) and 80 percent relative to the westerly
plant.  The MIU contours and percentile contours generally paralleled each other
and both provided a basis for establishing the extent of the dispersion pattern.

       The field sensory evaluations described above provide weighty evidence for
the utility of the sensory evaluation technique as a means of identifying the extent
and magnitude of the dispersion pattern of an odor emission when expressed in MIU
units.  The use of percentile contours in lieu of MIU units obviates the need for
extensively training panelists in memorizing odor intensities. Although panelists
are still required, no measure of odor intensity can be obtained  when the parameter
indicating the percent of time that an odor was perceived is used.  The above study
results indicate that the full-scale sensory evaluation technique for atmospheric  odor
measurement should have significant application in those special circumstances
where legal liability or compliance questions require that an odor emission problem
be documented in an explicit and defensible manner.

-------
                                            Table V
                         UNITED STATES WEATHER BUREAU OBSERVATIONS
                      PHILADELPHIA, PENNSYLVANIA, INTERNATIONAL AIRPORT
                                          OCTOBER 1969
Date
16
16
17
20
Hour
1 pm
4 pm
1 pm
10 am
Ceiling
(feet)
8,000
Unlimited
10,000
Unlimited
Visibility
(miles)
8
10
10
5
Wind
Direction*
(degrees)
200
190
290
240
Wind
Speed
(knots)
6
6
14
12
Dry Bulb
Temperature
(degrees F)
61
64
60
66
Relative
Humidity
(percent)
63
61
46
63
Notes
-
-
-
Hazy,
Smoky
* Wind originates from direction indicated, which is clockwise from true north.
Source:  U.S. Department of Commerce.  Local Climatological Data.  A shville, North Carolina:  Environmental

-------
                      Figure 8

REFINERY ODOR EMISSION DISPERSION PATTERN NO. 1
       (Contours based on mean odor intensities)
       ^^ To Oil Refineries

   Source: Engineering-Science,  Inc.

-------
                     Figure 9

REFINERY ODOR EMISSION DISPERSION PATTERN NO.  1
  (Contours based on percent of time odor was perceived)
 Source: Engineering-Science, Inc.

-------
                         Figure 10

   REFINERY ODOR EMISSION DISPERSION PATTERN NO. 2
          (Contours based on mean odor intensities)
                                                        •0.19
                                                         0.28
                                                 0.50
Source: Engineering-Science,  Inc.

-------
                      Figure 11

REFINERY ODOR EMISSION DISPERSION PATTERN NO. 2
 (Contours based on percent of time odor was perceived)
   Source: Engineering-Science, Inc.

-------
                      Figure 12

 REFINERY ODOR EMISSION DISPERSION PATTERN NO.  3
        (Contours based on mean odor intensities)
                  0.50
                    1.00
 0
500'
     2.02
 1000'  2.00
"
             To Oil Refineries
Source: Engineering-Science, Inc.
                        - 151 -

-------
                     Figure 13

REFINERY ODOR EMISSION DISPERSION PATTERN NO/ 3
 (Contours based on percent of time odor was perceived)
 Source: Engineering-Science,  Inc.

-------
                                        Figure 14
                         RENDERING FACILITIES ODOR EMISSION
                                   DISPERSION PATTERN
                         (Contours based on mean odor intensities)
                               Richmond
0.50
                                                               0.28
                            0
500
                    Source: Engineering-Science,  Inc.

-------
                        Figure 15


RENDERING FACILITIES ODOR EMISSION DISPERSION PATTERN
     (Contours based on percent of time odor was perceived)
  Source: Engineering-Science,  Inc.


-------
                                    Notes
1.     Amos Turk and Stanley Mehlman, "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, 1968),
       p. 29.
2.     Norman A. Huey, et al., "Objective Odor Pollution Control Investigations, "
       Journal of the Air Pollution Control Association, X, No. 6 (December, 1960),
       pp. 441-446.
3.     Ibid.,  p. 442 and 443.


4.     Turk, op. cit., pp. 31-33.


5.     Ibid., pp. 32 and 33.
6.     Amos Turk and Janet Wittes,  "The Selection of Judges for Odor Discrimina-
       tion Panels, " Correlation of Subjective-Objective Methods in the Study of
       Odors and Taste, Special Technical Publication No. 440 (American Society
       for Testing and Materials,  1968).
7.     Ibid., pp. 53-57.


8.     Ibid., p. 52.

-------
                                 References
Turk, Amos and Mehlman, Stanley.  "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,  1968.
Huey, Norman A., et al.  "Objective Odor Pollution Control Investigations, " Journal
       of the Air Pollution Control Association, X, No. 6 (December,  1960), 441-
       446.
Turk, Amos and Wittes, Janet. "The Selection of Judges for Odor Discrimination
       Panels, " Correlation of Subjective-Objective Methods in the Study of Odors
       and Taste. Special Technical Publication No. 440, American Society for
       Testing and Materials, 1968.

-------
                                 CHAPTER IX
                           PUBLIC OPINION SURVEY
       The public opinion survey was conducted by Copley International Corporation
as an exploratory approach to determine the degree to which odors constitute a sig-
nificant influence on the lives of people living and working in actual odor problem
areas.

       Within this broad framework four objectives were established and research
was designed to explore and/or develop data on each. These objectives were to
determine:

       (1)     How aware people are of odors .

       (2)     If there are differences in opinions between people living in
              odor problem areas and people living  in relatively odor-free
              areas.

       (3)     If people living in odor problem areas feel they have suffered,
              either socially or economically,  due to odors.

       (4)     The degree of interest the public may have in solving odor
              problems.

       Since no prior research has been conducted in the area of public opinions on
odors, this survey was designed as a broad, exploratory approach toward establish-
ing benchmarks from which more definitive research thrusts can be made .  Essen-
tially, it constitutes the basic research cornerstone necessary for a planned
exploration into an area with many unknowns.  As is  the case in  any research of an
exploratory nature, the data generated should be used to develop hypotheses and
direct further research into more effective channels  . Therefore, the findings should
not be considered as definitive or as requiring no further investigation.

       Because of the exploratory nature of this study, a variety of questions and
question approaches were used with the expectation that some would generate usable
data while others might not allow for valid discrimination between the variables.
The survey was conducted in the principal cities of seven metropolitan areas to ob-
tain a representation of opinion from individuals in different regions  of the United
States as well as  different stations in life.

-------
       This chapter provides a description of the development of the public opinion
survey and a summary of the survey results.  The findings are divided into three
major sections:

       (1)     An analysis of the opinions of male and female adults in house-
              holds (hereafter referred to as residents).

       (2)     An analysis of the opinions of businessmen.

       (3)     A comparison of results obtained from each  of the cities sur-
              veyed .
DEVELOPMENT OF THE SURVEY

Description of Survey Technique

       Within each city investigated, the team of engineers from Engineering-
Science, Inc., in conjunction with officials of the local air pollution control agency
in that city, selected one or more actual odor problem areas for investigation.  The
engineering team determined the boundaries of each of the odor problem areas,
sketched the boundaries on a city map (see Appendix E), and conducted technical
field studies within the described boundaries.

       Upon completing their studies in each city, the team of engineers deposited
the map at the office of the local air pollution control agency, for later retrieval by
the CIC survey director, and then traveled to the next city scheduled for investiga-
tion .  Within a day or two the survey director retrieved the map.

       In preparation for selecting a sample of residents and businessmen to inter-
view , the survey director then called upon the marketing research staff of the local
newspaper to obtain information concerning the demographic composition of the odor
problem areas described on the city map.  (Most newspapers in  large metropolitan
areas maintain a wealth of recent demographic information on the potential audience
in their market area.)

       Utilizing the acquired information, survey test and control areas were
established. The control area in each city was selected to match as nearly as
possible the test area in terms of socio-economic characteristics, yet be as free
as possible of odor problems . Appendix E presents detailed information on the
area boundaries  as well as an analysis of the socio-economic characteristics of the
matched areas.

-------
       A single test area was established in four of the cities surveyed.  In contrast,
it was necessary to establish two test areas in the other three cities to provide ade-
quate representation.  With the exception of Tampa, all of the cities had a single
control area.  Table I indicates the number of test and control areas established in
each city.

-------
                            Table I
    NUMBER OF TEST AND CONTROL AREAS IN EACH CITY
                               Number of           Number of
      City                     Test Areas         Control Areas

Portland, Oregon                    1                    1
Cincinnati, Ohio                    2                    1
Kansas City, Missouri               1                    1
Buffalo, New York                   1                    1
Philadelphia, Pennsylvania           2                    1
Tampa, Florida                    2                    2
San Francisco,  California           1                    1
Source: Copley International Corporation.

-------
       Prior to the sampling and interviewing conducted in each city, the survey
director contacted a local interviewing service company to complete final prepara-
tions for the survey work. Printed questionnaires and interviewer instructions were
sent to each company in advance to allow as much time as possible for staff familiar-
ization.  Meetings were held with the interviewing company supervisor in each city
to discuss and complete details of sampling, briefing, execution,  and verification of
the survey design.

       Sampling was conducted in the following manner .  Within each test and control
area, all blocks were numbered in a serpentine order.  From each area, approximately
35 blocks were randomly selected in an "every-nth" fashion to provide geographic dis-
persion of the households  drawn for the public opinion sampling of the residents.

       Using a street address (reverse-order) telephone directory, all telephone
numbers were listed for each block selected for the sample.  Again, using the
"every-nth" random selection procedure, five primary and two alternative phone
numbers were selected for each block.  Whenever a block did not have the required
number of telephones listed, the remainder were  selected from an adjacent block in
the same random fashion.

       Purposive sampling was used to select the retail businessmen from each area.
Available business and professional establishments within each area were listed.
From this listing 25 were judgmentally selected to provide a wide variety of retail
outlets.  It was planned to include types of retail outlets which could conceivably
suffer loss of patronage because of odor pollution in the vicinity.  Table II provides
a list of the types of retail outlets which were considered for  this  sampling.

-------
                        Table H
    BUSINESS AND PROFESSIONAL ESTABLISHMENTS
LISTED FOR SELECTION OF BUSINESSMEN RESPONDENTS
         Appliance stores
         Auto dealers
         Bars, taverns, and nightclubs
         Beauty and barber shops
         Book and stationery stores
         Camera and photographic shops
         Clothing stores
         Department stores
         Drive -in eating places
         Drive-in movies
         Drug stores
         Florist and garden shops
         Food and grocery stores; supermarkets
         Furniture stores
         Hardware and building supplies stores
         Jewelry stores
         Laundry and dry  cleaning shops
         Liquor stores
         Motels
         Music and record stores
         Pet shops
         Physicians and dentists offices
         Restaurants and cafes
         Service stations (gas  and oil)
         Shoe stores
         Sporting goods stores
         Tire and auto accessory shops
         Variety,  novelty, and gift shops
Source: Copley International Corporation.

-------
       Only professionally trained interviewers were used.  They were selected,
briefed, and supervised by the local interviewing company supervisor who was
assigned the responsibility for the execution of the survey.

       All interviewers were briefed verbally and given two detailed sets of
"Instructions to Interviewers, " one set for use in interviewing residents of house-
holds, the other set to use with businessmen (see Appendix N).  These instructions
gave detailed advice on how the interviewer should proceed in selecting individual
respondents and in administering the questionnaires .

       The interviewing procedure required an initial call plus up to five follow-up
calls,  as necessary, to contact each respondent. Only  after  six unsuccessful
attempts to contact a respondent was an alternate telephone number to be used.
This callback technique provided for a high rate of completion of the sampling
units originally selected,  and thus generated a more representative sample than
otherwise could have been obtained.

       The interviews were scheduled to begin as soon  as possible after the ESI
engineering team had delineated the  test areas.  Table III indicates the dates during
which the public opinion survey was  conducted in each city.

-------
                          Table IE
DATES DURING WHICH PUBLIC OPINION SURVEY WAS CONDUCTED
       City                          Dates

   Portland                September 24 - October 1, 1969
   Cincinnati               October  3 - October 10, 1969
   Kansas City             October  6 - October 10, 1969
   Buffalo                 October 14 - October 21, 1969
   Philadelphia             October 20 - October 25, 1969
   Tampa                 October 20 - October 25, 1969
   San Francisco           October 27 - November 1, 1969
Source: Copley International Corporation.

-------
       When two test or control areas were designated in a city, the sample was
divided and approximately equal numbers of interviews were  completed in each area,
The survey design called for a total of 400 completed interviews in each city.  Of
this total, 200 were to be conducted in the test area and 200 in the control area.

       In both the test and control areas, the sample was further divided.  Of the
200 interviews to be conducted in an area, 175 were to be with residents (male and
female adults in households) with an approximate 50-50 split  between males and
females, while 25 were to be with small retail businessmen.  Table IV indicates
the total number of interviews  completed in each of the cities surveyed, while Table
V provides the breakdown of the total number of interviews of residents completed,
by men and women.

       The survey technique employed in each city was identical.  In compliance
with the project budget and specifications, all interviews were made by telephone.
The exception was in San Francisco, where the available supply of telephone num-
bers in both the test and control areas was exhausted and personal interviews had to
be conducted to supply the required number of responses.

-------
                                Table IV
        TOTAL NUMBER OF INTERVIEWS COMPLETED IN EACH CITY
                            Test                        Control
                  Residents     Businessmen     Residents     Businessmen

Portland              173            25              177           25
Cincinnati            175            25              175           25
Kansas City           175            25              175           25
Buffalo               175            25              175           25
Philadelphia           175            25              176           25
Tampa               175            25              175           25
San Francisco         177            25              160           25
                                Table V
       TOTAL NUMBER OF INTERVIEWS OF RESIDENTS COMPLETED,
                   BY MEN AND WOMEN, IN EACH CITY
                              Test                   Control
       City                Men     Women       Men     Women

       Portland              72        101           61        116
       Cincinnati            89         86           82         93
       Kansas City          89         86           87         88
       Buffalo               89         86           83         92
       Philadelphia          90         85           86         90
       Tampa               87         88           88         87
       San Francisco        69        108           79         81
Source:  Copley International Corporation.

-------
Description of the Questionnaires

       The questionnaires were designed to generate a variety of data which could
be used in obtaining a measurement of opinions concerning the effects of odor prob-
lems .  The opinions of householders were sought about the areas in which they
lived and the opinions of businessmen about the areas where they operated retail
businesses.

       The CIC professional staff involved in this study provided many ideas and
suggestions on the composition of the questionnaires to be administered to residents
and businessmen. In addition, a number of questions included in a recent air quality
opinion survey of Clarkston,  Washington, were adapted for use.1  After pretesting
a preliminary form and incorporating the suggestions of the  National Air Pollution
Control Administration, the final forms of the questionnaires were developed.

       The questionnaire  constructed for administration to residents (see Appendix
N) was designed to shed light on a number of objectives .  Questions 1 through 6 were
structured to determine the degree of satisfaction or dissatisfaction people had with
the area in which they reside .  These questions provided respondents the opportunity
to volunteer any belief that odor problems existed in their neighborhoods .

       Thus, one purpose of these six questions was to obtain an unaided recall
measurement of odor problems .  If respondents did not volunteer information con-
cerning odor problems, the subject was then approached through subsequent questions
which aided recall. The rationale was that greater confidence could be placed in
awareness of odor problems when the respondent volunteered the information than
if the subject had to be introduced through prompting. In addition, Question 3
attempted to obtain a measurement of how the respondents ranked some of the
common social and economic problems in their areas relative to the general prob-
lems of air pollution.

       Question 7 was designed to determine,  through direct (aided) questioning,
whether or not respondents were aware of air pollution in their area and, if so, in
what time periods it was most noticeable.  At this point, the interview was focused
on air pollution; odor problems had not yet been mentioned by the interviewer.

       Question 8 asked respondents to define the meaning of air pollution, thus
allowing them an additional opportunity to volunteer an environmental association
with odors .  If odors were mentioned, respondents were then asked to describe
what the odors s me lied like.

       Question 9 sought to determine if respondents were aware of odor problems
in their area, even if the respondent had not mentioned such problems in the inter-
view up to this point.  If an awareness of odors was revealed, respondents were then
asked if they knew the sources of these odors.  Question 10 provided a slightly differ-

-------
ent approach to determining awareness of odor pollution by asking if the respondents
felt the problem had been increasing,  decreasing, or was constant.

       Questions 11 through 15 were designed to be answered only by those individuals
who indicated in Question 10 that they were aware of odor problems in their neighbor-
hoods .

       Questions 11 and 12 were structured so that respondents could indicate how
they felt about the efforts of the local government to control odor pollution, and
whether or not they felt more money should be spent in attacking the odor problem.
These questions also provided for some measurement of the intensity of their feel-
ings toward odor problems. Similarly, Question 13 attempted to note the degree of
their feelings and reactions to odor problems by assuming those most troubled by
odors would attempt greater efforts to reduce or eliminate the odors from their
homes.

       Questions 14 and 15 were designed to obtain a very rough measurement of
whether or not the respondents felt odor pollution had harmed them economically
or socially. Questions 16 and 17 were designed to develop additional data on the
possible effects  of odor pollution in areas removed from their immediate residences,
i.e.,  where they worked or shopped.

       Questions 18 through 28 were included to collect data to provide a comparison
of the socio-economic characteristics of the respondents.  These questions also
enabled detailed analyses to be made by allowing for examination of economic and
demographic factors as they relate to  answers given to other questions.

       The questionnaire form developed and directed to the businessmen (see
Appendix N) was very similar  to the form used with residents, with two exceptions.
First, the businessmen were questioned concerning the relationship between odor
problems (among other factors) and their business establishments rather than their
residential areas .  The other major difference was that economic and demographic
data was not collected from businesses because of restrictions established by the
Bureau of the Budget.  As a consequence of this last difference, no relationships
could be drawn between economic and  demographic factors and data on the  other
variables gathered in the survey of businessmen.

Description of the Statistical Analysis Applied

       In the following summary of findings, frequent reference is made to the
tables in Appendices G through M. These tables present the survey findings .
Responses were tabulated and  a percentage distribution was prepared for each
question.  Findings for both the test and control areas, as well as total responses,
are presented in these appendices .

-------
       The tables in the appendices are numbered in a fashion that permits easy
reference to the original question in the questionnaire.  For example, Table Q2A
in Appendix G presents the findings of Question 2A on the "General Public" question-
naire .

       Each table in Appendices G and L also contain significance test designations.
Chi-square tests were conducted to determine if there were significant differences
between the test and control areas.2  in addition to the five percent levels, one
percent significance levels were also used.  Keys on Appendices G and L indicate
the symbols used to signify:  significant to the one percent level,  significant to the
five percent level, not significant at the five percent level, and chi-square tests not
applicable.
SUMMARY OF RESULTS

Residents - Summary of All Cities

       This summary is based on the results of all cities combined (Appendix G)
and, as such, is representative of findings which might be obtained if identical
studies were to be conducted in other comparable cities elsewhere in the nation.

       The type of resident found in both the test and control areas in this study can
perhaps best be characterized as lower-middle class.  The resident tended to be a
blue collar worker, middle-aged or older, below average in education and income,
and with family size lower than average.  An excellent division of the sample be-
tween men and women was obtained (approximately 47-53 percent). The occupational
category of sales/white collar is sizeable because of the number of women clerical
workers obtained in the sample.  Men were typically employed as factory and ser-
vice industry workers, truck drivers, or construction workers.

       Most residents  interviewed liked their test and control areas as a place to
live (Ql). Approximately 60 percent of the respondents rated their area of the city
as an excellent or good place to live. Only  10 percent rated it as poor or very poor.
Residents of the control areas tended to rate their areas somewhat higher than did
residents of test areas .

       A sizeable proportion of the residents of both test and control areas, approxi-
mately 54 percent and 45 percent, respectively, could think of disadvantages about
living in their particular area of the city (Q2A). It is significant to note that a larger
proportion of test than control area residents could think of disadvantages .

       When asked what some of these disadvantages were,  approximately 16 per-
cent of the test group and  2 percent of the control group mentioned air pollution (Q2B).
About 6 percent of the test group and 2 percent of the control group also mentioned

-------
odors.  These unaided responses indicate a highly significant difference between the
two groups in terms of air quality perception.

       Residents of both groups were also dissatisfied with:  street disrepair,
deteriorated areas, traffic,  and dirt/litter. To a lesser degree, they were dis-
satisfied with:  transportation,  crime, noise, and undesirable people.

       Residents of both groups were also asked to rate their areas of the city on a
number of community problems . Air pollution was considered the most severe prob-
lem in the test areas, followed closely by high taxes (Q3,  "Serious" and "Somewhat
Serious" responses combined).  The residents of the control areas did not view air
pollution as being as severe a problem (see Table VI).

-------
                         Table VI
   PERCENT OF RESIDENTS INDICATING THE FOLLOWING
        COMMUNITY PROBLEMS WERE SERIOUS OR
 SOMEWHAT SERIOUS IN THEIR TEST AND CONTROL AREAS
       Community Problem               Test        Control

Air pollution      .                        65%         36%
High taxes                                62           89
Deterioration of streets, parks, etc.         42           34
Crime                                    38           33
Water pollution                            33           22
Inadequate public transportation             30           32
Deterioration of homes, stores, etc.         28           23
Racial tensions                            18           11
Source:  Copley International Corporation.

-------
       With regard to taking action and complaining to a local agency about some
problem in their areas of the city, about 22 percent of the test area residents had
done so compared to 18 percent for the control area (Q4A).  When asked what the
problems were about which they had complained, 10 percent of the test area com-
plainants indicated air pollution compared to only 2 percent of the control area
complainants (Q4B). Air pollution was the only problem mentioned that showed
such a marked distinction between the two groups . This provided an additional
indication of the similarity of the  test and control areas in all variables except the
degree of air and odor pollution.

       Other problems ranking high as the subject of complaints were: juvenile
delinquency and vandalism;  street disrepair; mosquitoes, rats,  and snakes; heavy,
noisy, and speeding traffic;  poor utility service;  derelict buildings; theft; and  sewers .
In response to this last item, poor sewerage was cited as the main problem.  How-
ever, the attendant  odors may have been a major factor in prompting the demands
for action.

       Many of the  people who made complaints had little knowledge of where to
address them.  For example, many complaints about air pollution were made to the
police or to city hall rather  than to the local air pollution control agency or health
department.  This is not readily apparent from looking at the summary data (Q4C)
but is revealed when individual questionnaires are studied.

       Residential mobility  was low for both areas (Q5).  The residents of the test
areas appeared to be slightly more stable in terms of  long-time residence than
people in the control areas.  However, roughly 30 percent of each group had thought
seriously about moving to another area (Q6A). Table  YE presents some of the major
reasons given by respondents for  desiring to move from their areas (Q6B).

-------
                       Table VII.
       PERCENT OF RESIDENTS INDICATING THE
     FOLLOWING MAJOR REASONS FOR DESIRING TO
     MOVE FROM THEIR TEST AND CONTROL AREAS
Reasons for Desiring to Move         Test        Control

Get bigger and/or better house         18%         17%
Move into suburbs                     16           11
This area deteriorated                14           17
Personal reasons                     14           15
Pollution                             8            1
Better school district                   7            4
Newer and cleaner area                6           11
Get smaller house  or apartment         5            7
Source:  Copley International Corporation.

-------
       While the proportion of people in the test areas who mentioned pollution is
not high (about 8 percent), it is significant that only one percent of the control group
cited this reason.  The significance of this difference is enhanced by the fact that
these responses were not aided or prompted, but were completely voluntary.

       When residents were asked specifically if there was air pollution in their
areas of the city at any time during the year, approximately 70 percent of respond-
ents in the test areas replied affirmatively. Only 36 percent in the control areas
responded in this way (Q7A).

       Residents who stated they did have air pollution in their areas were asked
when they noticed it the most (Q7B).  Summer drew the largest number of responses
from both test and control areas.  Many of these respondents commented that this
was the season of year when they spent more time outside the house and thus were
more apt to notice air pollution.

       In both test and control areas a sizeable proportion of responses mentioned
the occurrence of air pollution during the morning or during the day.  This time
period was of greater concern in the control areas than in test areas, however.
By contrast, mention of air pollution occurring in the evening, or when it was dark,
was of major concern in test areas but not in control areas.  Approximately 19 per-
cent of the test group mentioned this time period compared to only 9 percent of the
control group.  In the test areas, respondents  remarked to the effect that factories
waited until it was dark so that people would not see the pollution pouring from
their smokestacks.

       Further questioning was conducted to determine what the term "air pollution"
meant to the respondents (Q8A-Q8E). Of the five conditions examined only two
elicited a significant difference between the test and control groups, with higher
proportions of the test groups stating that they considered odors, and haze or fog,
as part of what they meant by air pollution.  Table VIII indicates the proportion of
respondents in the test and control groups who felt the indicated conditions meant
air pollution to them.

       However, within the test area, two conditions mentioned most were smoke
or dust and noticeable odors.  Smoke or dust was the only similarly high-ranked
condition in the control groups.  Thus, in odor problem areas, odors are as highly
equated with  "air pollution" as are other conditions.

-------
                     Table VEI
  PERCENT OF TEST AND CONTROL AREA RESIDENTS
   INDICATING THE FOLLOWING CONDITIONS WERE
          ASSOCIATED WITH AIR POLLUTION
    Condition                   Test          Control

Smoke or dust                    84%            82%
Noticeable odors                  82             66
Irritation of the eyes              69             66
Nose or throat  irritation           65             66
Haze or fog                      58             50
Source:  Copley International Corporation.

-------
       Residents who said that odor was associated with air pollution were then
asked what some of these odors smelled like (Q8F).  Two particular responses
are notably different for the test and control groups .  "Chemical" was highest for
the test group, and "auto exhaust" was highest for the control group.  Table DC
summarizes the  items most frequently mentioned by residents.  Note the relatively
high percentage of test area residents identifying odors smelled as "rotten." This
is interpreted as more of a catchall of undefined replies than a specific category of
smell.

       When residents who had described one or more odors were asked which one
they smelled most often in their area of the city (Q8G),  the same major items were
singled out as were mentioned in response to Question 8F and shown in Table IX.
Table X summarizes those  items most frequently mentioned by the residents .

-------
                     Table IX
   PERCENT OF RESIDENTS IDENTIFYING ODORS
  SMELLED IN THEIR TEST AND CONTROL AREAS
  Odors Smelled            Test           Control

Chemical                    19%            14%
Rotten                      13              4
Eggs/Sulphur                13             11
Sewage                      12             10
Oil/Petroleum                11             12
Burning                       8             12
Burning hides/flesh            5              *
Trash/Garbage                4              8
Auto exhaust                  3             16
  Less than 0.5 percent
Source:  Copley International Corporation.

-------
                         Table X
      PERCENT OF RESIDENTS DESCRIBING ONE ODOR
SMELLED MOST OFTEN IN THEIR TEST AND CONTROL AREAS
       Odor Smelled             Test          Control

     Rotten        .              13%             2%
     Chemical                    12              7
     Eggs/Sulphur                 8              5
     Sewage                      7              6
     Oil/Petroleum                7              5
     Burning                      6              9
     Burning hides/flesh            6              1
     Auto exhaust                  3             13
     Garbage/Trash               2              6
Source:  Copley International Corporation,

-------
       To investigate another dimension of the involvement of residents with odor
pollution,  they were asked directly how much odor pollution in their area of the city
bothered them (Q9A). Again, distinct differences between test and control areas
were obtained. Approximately 20 percent of the test area residents said "very much1
compared to about 4 percent for the control area. And,  46 percent of the test area
residents said "some" compared to only 30 percent for the control area.

       Residents who stated that odor pollution  in their area of the city bothered
them "very much" or  "some" were asked to identify the source or cause of the odors
(Q9B).  Many of the responses were vague. Even when residents were quite specific
as to the type of source, very few could name specific companies as the odor origi-
nators .

       Table XI shows some of the more frequently mentioned sources .  Again,
there are marked differences between the test and control areas .  Almost 30 per-
cent of the control area respondents viewed "traffic" as the dominant odor pollution
source compared to only 7 percent in the test group. On the other hand, the test
group focused on the more traditional odor pollution sources.

-------
                   Table XI
PERCENT OF RESIDENTS INDICATING THEY WERE
       BOTHERED BY ODOR POLLUTION —
        BY ODOR SOURCE(S) IDENTIFIED
  Odor Source             Test           Control

Chemical plant             28%             4%
Packing house              20              1
Factories                  13             10
Steel plant                 13              *
Oil refinery                 9              2
Traffic                     7             30
Swamp/River/Bay            6              3
Dump/Incinerator            5             13
Sewage                     4              7
  Less than 0.5 percent
Source:  Copley International Corporation.

-------
       Viewed against the dimension of time, residents in the test area again indi-
cated their awareness of odor problems (Q10).  About 20 percent of test area resi-
dents felt odor pollution was "continuously serious" compared with less than 5
percent in the control areas. Also, roughly 22 percent of the test area residents
and 20 percent of the control area residents felt odor pollution was "becoming more
serious."  On the other hand, about 60 percent of the control area residents stated
odor pollution was "not serious" in their area of the city, while only 26 percent in
the test areas took this position.  These individuals were not asked Questions 11
through 15.

       In terms of the efforts that are being made to control odor pollution in their
area of the city, 50 percent of the test area residents and 40 percent of the control
area residents felt "some" or a "great deal of effort to control" odor pollution was
being made.  However,  36 percent of  the test area and 40 percent of the control
area residents felt "very little" or "no effort to control" odor pollution was being
made(Qll).

       Responding to another question directed at odor pollution control efforts,
approximately 74 percent of the test area and 59 percent of the  control area resi-
dents felt more money should be spent to control odor pollution in their areas of
the city (Q12). However, in all cities in which this survey work was conducted,
voluntary asides stressed the fact the money should come from the offenders and
not from the taxpayers .

       When asked what they do to reduce or eliminate offensive odors from outside
sources that invade the home, over one-third replied  "nothing" or  "there's nothing
you can do."  Many residents felt there were few  courses of action available and
seemed resigned to living with the odors . About 25 percent mentioned closing doors
and windows,  and almost another one-third mentioned spraying or deodorizing (Q13).

       An attempt was made to determine if residents experienced economic effects
attributable to odor pollution. Residents were asked if they owned or were purchas-
ing their home (apartment or townhouse). About 75 percent of the test area and 69
percent of the control area residents were classified as owners (Q14A).  Those who
owned homes were then asked if they felt odor pollution had reduced the value of
their home property in any way. About 21 percent of the test area and 9 percent of
the control area residents felt odor pollution had reduced the value of their property
(Q14B).

       When asked to relate how odor pollution had reduced their property values,
almost half of the test area residents  and over a third of the control area residents
said potential buyers would change their minds if  they came on  a day when offensive
odors were present.  About 17 percent of the residents of both test and control
groups showed some degree of confusion between  air pollution and odor pollution
by answering that the odor pollution ruined the paint and/or left a film on their
houses (Q14C).

-------
       At this point a comparison question — "Have you ever seriously considered
moving away from here because of odor pollution in the air?" — was asked so that
the response might be compared with the unaided answers received from Question 6.
With prompting, about 17 percent of the test area and 11 percent of the control area
residents agreed that they had (Q15).

       As noted previously in Table VII, 8 percent of the test and one percent of the
control area residents  volunteered they had thought seriously about moving because of
pollution (Q6B).

       A very small proportion — around 2 percent — of both groups indicated
there were stores or shopping areas in the city which they did not patronize because
of odor pollution (Q16A).

       About half of the respondents were employed away from home.  Of these
employed residents,  almost a third reported odor pollution bothered them "some"
or "very much" at the place where they worked (Q17A).  The majority of these
reported it merely "smelled bad" but roughly 10 percent each reported eye irritation,
a choking feeling, and dirt/dust problems.  Responses were very similar between
test and control group workers (Q17B).

       In summary,  many of the people living in the areas affected by odor problems
are very much aware of these problems .  They are also aware of the offending
source(s).  They seem to be disturbed and want something done about it. However,
many of them appear to have a feeling of hopelessness,  believing they have little
power to change the situation.

       Many residents see the abatement problem as belonging to the industries that
create the odors .  Respondents felt that those producing the odors should pay to
eradicate them.

       There is a high degree of agreement among residents of the particular areas
with regard to the factors  that constitute air pollution.  The people, especially in
polluted areas, put great emphasis upon odors as integral elements of air pollution.

Residents - Summary of All Cities by Demographic Breakdown

       Cross tabulations by income,  education,  occupation, and age were developed
for Questions 1, 3, 8,  14 and 15 for both the test and control groups.  Appendix H
contains these tables, for income,  Appendix I for occupation,  Appendix J for educa-
tion, and Appendix K for age.

       A complete individual summary of every table in these appendices is beyond
the scope of this report due to the magnitude of the data.  In addition, many of the
tables fail to reveal any especially meaningful information.  Therefore, significant
highlights have been chosen and summarized briefly.

-------
       Overall, control area residents were more satisfied with their area of the
city as a place to live than were test area residents .  In both groups, highest ratings
were given by those individuals reporting incomes in the upper ranges . The profes-
sional, technical,  and managerial occupational types were decidedly more satisfied
with their area than were the blue collar workers .  Higher levels of educational
attainment also seemed to be related to a better outlook on their area of the city.
Residents below 24 years of age appeared more dissatisfied than did the other age
groups.

       The better educated, upper income,  and older residents tended to  view crime
in their area of the city as being not quite as serious as did the younger, lower in-
come, and least educated residents.  The test areas were viewed as more serious
than the control areas among those who admitted some degree of seriousness in the
crime problem. This was especially true of those residents with incomes over
$10,000 and holding white collar jobs .

       Awareness of the water pollution problem in the respondents' area of the city
was only slightly heightened by increased educational attainment, income, occupational
level, and youth.  By contrast, awareness of the air pollution problem in their area
of the city rose markedly with increases in these same characteristics.

       Deterioration of streets, parks, sewerage, utilities, public buildings, and
so on, in their area was  seen as more serious  than pollution by the younger residents
in the lower income groups who hold blue collar jobs.  These same residents also
viewed deterioration of shopping centers, stores, homes, and industrial buildings
as being more serious than pollution.

       Older residents viewed racial tensions as a more serious problem than
pollution in their area of the city. Little difference resulted from variations in
education, income, or occupation.

       High taxes in their area of the city was  viewed as decidedly more serious by
the residents with lower  incomes (including the retired and unemployed),  lower
educational achievements, and lower occupational categories .

       Among the residents in both the test and control areas who felt their area of
the city suffered from air pollution, no demographic differences were found relating
to smoke or dust and haze or fog. However, irritation of the eyes, nose, and throat
are considered important elements of air pollution by residents with better educa-
tions and higher incomes .  A vast majority of all residents consider noticeable odors
as a basic element of air pollution,  with a slightly greater percentage of younger
residents holding this view.

-------
       As might be expected, ownership of homes tended to rise with age, better
educations,  higher incomes, and white collar occupations .  However, the younger
residents in the test areas tended to feel more strongly that odor pollution had re-
duced the value of their property.

       Age appeared to be the only demographic variable having an effect on a
serious consideration to move from the area because of odor pollution. Younger
residents gave greater consideration to moving.

Businessmen - Summary of All Cities

       This summary is based on the findings reported in greater detail in Appendix
L.  It is a report of all seven cities combined and, as such, is representative of
findings which might be obtained were identical studies conducted in other comparable
cities across the  nation.

       The primary target of this portion of the overall survey was the small busi-
nessman who might suffer economic loss because of odor pollution in the area of his
business . Most respondents were owners/managers . Their establishments were
generally small.  In terms of total number of people working in the establishment,
including the owner/manager, the modal group was three. Approximately twice as
many automotive  establishments were sampled in the test areas as in the control
areas .  However, other types of businesses were well distributed between test and
control areas (Q16B).

       In general, most businessmen in both  the test and control areas liked their
area of the city for business. Approximately 65 percent of both test and control
groups felt their area was "excellent" or "good" for business (Ql). Only 6 percent
felt it was "poor " or "very poor."

       However,  40 percent of the test group and 36 percent of the control group
said they could think of disadvantages related to the area (Q2A). Of those mention-
ing disadvantages, theft was mentioned by 22  percent of the test group and  12 per-
cent of the control group.  The other two outstanding items indicating a marked
difference between the two groups were parking, with 13 percent in the test and 22
percent in control group, and traffic,  with 9 percent in the test and 17 percent in
the control group  (Q2B).  Identification of traffic problems in control areas was
principally focused on one-way streets. Odors were mentioned by 7 percent of the
test and 3 percent of the control groups, roughly equivalent to the unaided response
given by residents to the same question.

       Crime in their area of the  city was viewed with some degree of seriousness
by approximately half of both test  and control  groups (Q3A).  Businessmen, on the
whole, tended to view the problem of crime as being a little more serious than did
the residents (see Table XII).

-------
       Businessmen considered air pollution to be as serious as did residents.  The
same was true for the deterioration of streets, parks, sewerage, utilities, and
public buildings (Q3C and Q3D). When the problem was identified as deterioration
of stores, shopping centers, homes, and industrial buildings, both test and control
area businessmen felt the problem was more  serious than did the residents (Q3E).

       On two particular problems  businessmen in control areas appeared more
concerned than businessmen in test areas (Q3G and Q3H). Approximately 51 percent
of control area businessmen felt some degree of concern  about high insurance premi-
ums, compared to only 44 percent of test area businessmen. About 48 percent of
control area businessmen were concerned about inadequate parking facilities, com-
pared to 41 percent of test area businessmen.

-------
                           Table XII
    PERCENT OF BUSINESSMEN INDICATING THE FOLLOWING
           COMMUNITY PROBLEMS WERE SERIOUS OR
    SOMEWHAT SERIOUS IN THEIR TEST AND CONTROL AREAS
            Community Problem                 Test      Control

Air pollution                                     60%       34%
Crime                                          49        47
High insurance premiums                          44        51
Deterioration of streets, parks,  etc.               42        33
Inadequate parking facilities                       41        48
Deterioration of shopping centers, homes, etc.      36        32
Inadequate public transportation                    25        26
Water pollution                                  25        22
Racial tensions                                  17        17
Source: Copley International Corporation.

-------
       Businessmen were more active than residents in requesting some authority
or agency to take action concerning some problem in their area of the city (Q4A).
In addition, test area businessmen were more active than control area businessmen.

       Theft was the primary complaint of about 25 percent of all businessmen while
being mentioned by only 5 percent of all residents (Q4B). Test area businessmen
were more concerned about vandalism and derelict buildings while control area busi-
nessmen complained about parking and street repair.  Also, control area businessmen
tended to complain more to the police while test area businessmen favored city hall
by a wide margin (Q4C).

       With both the test and control groups, the average businessman has operated
his business in his area of the city for about ten years (Q5). Approximately 25 per-
cent of test area and 20 percent of control area businessmen have thought seriously
about moving to another area (Q6A).  These percentages for businessmen are lower
than responses given by residents of both test and control areas,  perhaps reflecting
the difficulty of relocating established business operations .

       Reasons why businessmen would want to  relocate could be expected to be
different from those of the private citizen, at least in most respects.  However, no
mentions were made of air pollution by test area businessmen while 8 percent of
test area residents had mentioned it.  Some interesting differences were also dis-
covered between test and control area businessmen (Q6B).  Table XIII summarizes
some of these differences .

-------
                         Table XHI
    PERCENT OF TEST AND CONTROL AREA BUSINESSMEN
         INDICATING THE FOLLOWING REASONS FOR
         WANTING TO RELOCATE THEIR BUSINESSES
   Reasons for Wanting to Relocate           Test        Control

Area deteriorating                         26%          34%
Crime                                    20           3
No money here/Move to wealthier area        14           19
Need to enlarge                            11           3
Personal                                   9           9
No market here for product/service           6           19
Source:  Copley International Corporation.

-------
       Two-thirds of the test area businessmen, compared to 26 percent of the con-
trol area businessmen, said there was air pollution in their area of the city at some
time during the year (Q7A).  While in test areas there was no significant difference
in the level of response to this question between businessmen and residents,  in con-
trol areas more residents than businessmen reported air pollution.

       Businessmen who stated there was air pollution tended to notice it more during
the day, and especially in the morning, than during the evening. Approximately 21
percent of the control area businessmen mentioned "summer" while 30 percent of
the test area businessmen said "most of the time ." (Q7B)

       When asked what the words "air pollution" meant to them, "noticeable odors"
was the only element which elicited a significant difference between the test and con-
trol groups of businessmen,  with more test area businessmen mentioning odors (Q8).
Table XIV summarizes the responses to the five conditions tested.

-------
                      Table XIV
PERCENT OF TEST AND CONTROL AREA BUSINESSMEN
   INDICATING THE FOLLOWING CONDITIONS WERE
          ASSOCIATED WITH AIR POLLUTION
     Condition                  Test           Control

Smoke or dust                    89%            93%
Noticeable odors                 87             72
Nose or throat irritation           73             69
Irritation of the  eyes              69             70
Haze or fog                      47             47
Source:  Copley International Corporation,

-------
       In describing the smell of some of the odors,  test area businessmen stressed
chemical, sewage, rotten flesh, and garbage smells while control area businessmen
stressed exhaust and burning smells (Q8F).  These responses were quite similar in
proportion to those given by residents, with the exception that control area business-
men stressed "exhaust" much more than did control area residents .

       Approximately 10 percent of the test area and 2 percent of the control area
businessmen felt odor pollution had some effect upon their businesses (Q9A). This
may be compared with 66 percent of test area and 34 percent of control area resi-
dents who indicated odor pollution bothered them to some degree. Apparently, odor
pollution is not felt to be as demanding a problem in the business community as it is
in the nearby residential neighborhood.

       A further indication of this  fact is the test area finding that  37 percent of
businessmen and only 26 percent of residents felt that odor pollution in their area
of the city was  "not serious." However, 21 percent of the test area businessmen
said it was "continuously serious"  and 20 percent said it was "becoming more
serious. "(Q10)

       Roughly half of both test and control area businessmen felt some degree of
effort was being made to control odor  pollution in their area, while about one-third
of both groups felt little or no effort was being made (Qll). Also,  68 percent of test
area and 58 percent of control area businessmen felt more money should be spent to
control odor pollution in their area (Q12). As with residents, businessmen felt the
offenders should be required to provide funds for control.

       When offensive odors from  outside sources invade their business establish-
ments, 53 percent of test and 31 percent of control area businessmen said there was
nothing they could do to reduce or eliminate them (Q13), while 21 percent of  the test
and 35 percent of the control area businessmen said there was no problem.   How-
ever, approximately 35 percent of  test and 44 percent of control area businessmen
mentioned such actions as: closing doors, spraying, opening doors, air condition-
ing,  and ventilating.

       Approximately 9 percent of test and 2 percent of control area businessmen
felt odor pollution had had a bad effect upon their business operation (Q14A).  Only
6 percent of the test area businessmen, and none of the control area businessmen,
said they had seriously considered moving their businesses from the area because
of odor pollution (Q15).

Residents - Summary of Individual Cities

       A detailed city-by-city analysis of the responses by residents to  survey
questions is presented in Appendix M.  While the sample size in any one city was
adequate for analyzing the total sample base for that city,  caution should be used

-------
when interpreting the findings based on sub-samples .  In some cases,  the sub-sample
sizes were too small to allow definitive conclusions to be drawn.

       The summary of aggregate results from all cities has already covered the
typical responses received in the survey. Therefore, this discussion will focus
primarily on the exceptional findings for individual cities.

       In San Francisco over 30 percent of both test and control area residents rated
their area as "poor" or "very poor."  The city that was judged most favorably by its
residents was Cincinnati.  Nearly 74 percent of Cincinnati  control area residents
rated their area of the city as "excellent" as did 35 percent of test area residents
(Ql).

       When asked if there were  disadvantages to living in their area of the city,
respondents in San Francisco were the most dissatisfied.   More than four-fifths of
test  area and two-thirds of control area residents in San Francisco complained of
disadvantages (Q2) .  Both areas are  ghettos . More of the Portland test area resi-
dents complained of odors than in any other city or group.  Nearly 27 percent
expressed this complaint without  aided recall (Q2B).

       In San Francisco relatively large proportions of both test and control area
residents rated crime,  racial tensions, and high taxes as  serious problems .  In the
Buffalo test area, 69 percent of the residents felt that air pollution was a serious
problem, and almost 44 percent rated  water pollution the same (Q3).

       In Kansas City,  Philadelphia, and San Francisco more control area residents
had thought seriously about moving than had test area residents .  Slightly more than
half  of the San Francisco control  area  residents had contemplated moving compared
with almost 46 percent of the test area residents  (Q6A).

       For all cities  except Portland,  there was a significant difference between
test  and control area responses with regard to the presence of air pollution.  In
most cities the percentage of respondents citing air pollution ranged from nearly 30
to 50 percent greater in test areas .  In Portland the difference was only about 11
percent. This may be attributed  to the fact that both Portland areas are at times
affected by air and odor pollution from a paper mill at Camas,  Washington. Other
than this, however, the control area is largely free of local industrial air pollutants
(Q7).

       In describing the nature of odors present in test areas,  the largest propor-
tion  of Portland and San Francisco residents mentioned smelling "rotten" odors
most often.   "Burning hides/flesh" was mentioned by only slightly fewer Portland
test  area residents .  Odors described  as "chemical" were  identified by a significant
proportion of Cincinnati test area respondents, while "oil/petroleum" was mentioned
most frequently by Philadelphia test  area residents.  A sizeable number of test area

-------
respondents in both Philadelphia and Kansas City were unable to describe the odors
perceived most often or did not know which odors they were (Q8G).

       Odor pollution bothered more test area residents in Philadelphia and San
Francisco than in other cities. More than one-quarter of the respondents in both
cities indicated that odor pollution bothered them 'Very much." (Q9A)

Summary of Survey Findings

       In summary, a large proportion of the residents and businessmen surveyed
were aware of the problem of air pollution.  In defining air pollution,  most of the
respondents included "noticeable odors."

       Residents in test areas identified air pollution as their area's most serious
problem, followed closely by high taxes.  In control areas the order was reversed.
High taxes  received the most mention, with air pollution a distant second.

       Roughly half of all resident respondents could think of disadvantages associ-
ated with their areas, yet the population was not a very mobile one .  About 79 per-
cent had lived in the area six years or more.  Approximately 30 percent indicated
that they had seriously considered moving from the area, but only a small percentage
because of air pollution.

       Both residents and businessmen indicated that air pollution was most evident
during the daytime and in the summer.  However,  unlike control area residents,  a
sizeable number of test area residents felt that air pollution was most noticeable
during evening hours.

       Odor, smoke, and dust conditions were those most usually associated with
air pollution by both residents and businessmen.  About 20 percent of test area
residents felt odor pollution was a continuously serious problem,  but only about  10
percent of the businessmen felt that business had been affected by it.  Approximately
21 percent  of the test area residents felt that property values in their area had been
adversely affected by odor pollution.

       Although awareness of  air pollution problems tended to rise with increases
in the residents' educational attainment, income, occupational level, and youth,
survey area residents were very much aware of the problem.  Young respondents
were apt to consider noticeable odors as a basic element of air pollution to a slightly
greater degree than did older residents. However, the vast majority of all age
groups held this view.  Most respondents also felt that more money should be spent
on controlling odor pollution in their areas .

-------
                                    Notes
1.     Nahum Z. Medalia, Community Perception of Air Quality:  An Opinion Survey
       in Clarkston, Washington (Cincinnati,  Ohio:  U.S. Public Health Service, 1965).
       For each question asked, the several response categories were cross tabulated
       against the test and control areas of all cities combined.  If no difference
       between the two areas was found to exist, the percentage distributions should
       be identical.  Since samples and not censuses were taken, a certain amount
       of difference between the two distributions was expected due to randomness.
       Therefore, a test of statistical significance was applied to determine whether
       the responses were independent of the area  in which the question was asked.

       Formulated in this manner, the appropriate form of statistical analysis is
       that of contingency tables.  Because the total number of interviews in the test
       and control areas was designed to be identical, both margin totals were not
       randomly determined.  To simulate a randomness in the determination of
       both margins,  the rejects, no-responses, and don't-know answers were
       eliminated.  Thus, instead of comparing the distribution of all responses,
       only those that fitted into the remaining categories were compared.

       The determination of independence using contingency tables is accomplished
       by a chi-square ( x2) test.  The sample chi-square value for each table was
       calculated using the following expression:
                                                 }-)2
                                           n

             r  = number of rows
             s  = number of columns (always = 2 in this analysis)

       The two distributions were said to be different if the observed sample chi-
       square value or a larger one had less than a 5 percent chance of occurring
       if indeed there was no difference. Put another way,  if the two distributions
       were identical,  then 95 percent of the time the calculated chi-square value
       should be between zero and some upper limit (determined from a table of
       chi-square values). If the observed chi-square value was above this limit,
       then it was concluded that there probably was a significant difference in the
       test and control area distributions.

-------
                                  References
Breslow, L.  Air Pollution:  Effects Reported by California Residents.  Berkeley,
       California:  State of California, Department of Public Health,  1956.
Crowe, M. Jay.  "Toward a 'Definitional Model' of Public Perception of Air Pollu-
       tion, " Journal of the Air Pollution Control Association ,  XVIII (March,  1968),
       154-158.
Medalia, Nahum Z.  Community Perception of Air Quality: An Opinion Survey in
       Clarkston, Washington.  Cincinnati, Ohio: U.S. Public Health Service,
       1965.
Stalker, W. W.  "A Method for Using Air Pollution Measurement and Public Opinion
       to  Establish Ambient Air Quality Standards, " Journal of the Air Pollution
       Control Association, XVII (March, 1967), 142-147.

-------
                                 CHAPTER X
               ASSESSMENT OF THE NATIONAL ODOR PROBLEM
       The purpose of this chapter is to project the principal findings of this study
to an assessment of the national odor problem.  The initial section is concerned
with the identification of potential odor problem areas throughout the United States.
The application of the set of factors developed in Chapters II through V in locating
these areas is described.   The value of this method is discussed in terms of the
correlation of data obtained from the survey of local air pollution control agencies .

       Then, the results of the public opinion surveys in the seven metropolitan
areas are projected to the nation as a whole.  The final section presents a few ob-
servations concerning the legal recourse available as a solution to odor problems.
IDENTIFICATION OF POTENTIAL ODOR PROBLEM AREAS

Definition of Potential Odor Problem Areas

       Potential odor problem areas may be defined as populated areas each having
an odorant source (or sources) and an odorant supporting atmosphere . Odor prob-
lem potential may be expressed as a function of the population, odor producing
potential, and atmospheric vulnerability of an area.

       It is recalled that this study gave principal attention to the odor producing
potential of industrial odorant sources .  In fact, only industrial odorant sources
are considered in the identification of odor problem areas by the method described
in the following paragraphs.

       A set of potential odor problem factors was developed not only to locate,
but also to compare potential odor problem areas throughout the nation.  The set
of factors included:

       (1)     A quartile ranking of size of population in all three-digit zip
              code areas in the United States (Chapter V,  Figure 1).

       (2)     A quartile ranking of the odor producing potential of industrial
              odorant sources in all three-digit zip code areas in the United
              States (Chapter V, Figure 2).

-------
       (3)     A summary of the relative atmospheric vulnerability to serious
              odor problems of areas in the United States (Chapter V, Figure 3).

       In applying these factors, one of sixteen possible combinations of the size of
population and odor producing potential quartiles could be coincident in each three-
digit zip code area.  These combinations, along with the summary of atmospheric
influence, are the parameters of problem potential.  However, without a precise
statement relating the three factors to an indication of overall influence, the rela-
tive problem potential of each area cannot be compared.  The development of such
a statement is beyond the scope of this study, however, a few generalizations can
be made.

       First,  the least  likely odor problem areas in the United States are among
those having a coincidence of the lower quartiles of size of population and industrial
odor producing potential and which are located within regions having the lowest fre -
quency of air pollution warning periods.  Second, within regions of equal atmospheric
influence, areas having a coincidence of higher quartile rankings are more likely to
be odor problem areas than areas having a coincidence of lower quartile rankings.
Third, the most likely odor problem areas are among those having a coincidence of
the upper quartile rankings and which are located within regions having the highest
frequency of air pollution warning periods.

       The relative odor potential of various zip code areas is shown on the follow -
ing map.  The purple-shaded areas within the higher numbered isopleths of Figure  1
(fold-out) represent the most likely odor problem areas in the United States . As
indicated, they are found principally in the Appalachian and California Coastal areas .
Because of occasional atmospheric vulnerability to serious odor problems in almost
all areas of the United States, it is estimated that most of the purple-shaded areas
of Figure 1 have experienced odor problems of a "serious" to "critical" degree .

-------
                                                                                                                                         Figure 1

                                                                                                            COINCIDENCE Of POTENTIAL ODOR PROBLEM FACTORS
12
                                                                                                                                                                                                                                                                       LEGEND

                                                                                                                                                                                                                                                         Three-digit  zip  code  areas  with  the
                                                                                                                                                                                                                                                         highest  quartile ranking  of  population
                                                                                                                                                                                                                                                                    ZIO  COQG  QrOOv  WITH  Tn9
                                                                                                                                                                                                                                                         highest  quartile  ranking  of  linear-
                                                                                                                                                                                                                                                         modified weighted values
                                                                                                                                                                                                                                                         Three-digit zip code areas with coinci-
                                                                                                                                                                                                                                                         dence of highest quartile ranking of size
                                                                                                                                                                                                                                                         of  population and linear-modified
                                                                                                                                                                                                                                                         weighted values
                                                                                                                                                                                                                                                         Total number of periods (n = 2,4,8, etc.)
                                                                                                                                                                                                                                                         with predicted  3 hours of more continu-
                                                                                                                                                                                                                                                         ous duration of simultaneous low-level
                                                                                                                                                                                                                                                         inversion conditions and low mean wind
                                                                                                                                                                                                                                                         velocities within the mixing depth


                                                                                                                                                                                                                                                         State boundaries


                                                                                                                                                                                                                                                         Sectional Center boundaries


                                                                                                                                                                                                                                                         National Zip Code Area boundaries

-------
                                                                           !   I
Correlation of Potential and Actual Odor Problem Areas

       If the assumptions incorporated in the development of the set of potential
odor problem factors are valid, then the odor problem areas identified in this study
should show positive correlation with actual odor problem areas as identified by
local air pollution control agencies . To test this possibility, partial results of the
survey of local air pollution control agencies were utilized.  Specifically, the re-
sponses to Question 5B ("How many complaints about odors has your agency received
in the past twelve months?") were compared to the three potential odor problem fac-
tors as follows:

       (1)    Because local air quality agencies typically serve individual
              cities or counties,  it was necessary to combine the responses
              of certain agencies in an attempt to construct an equivalence
              to the areas of the three-digit zip code system .  In only a few
              cases was this entirely successful.

       (2)    Despite this difficulty, the responses  from 56 approximately
              equivalent three-digit zip code areas were ranked in order of
              highest to least number of complaints .

       (3)    Each odor problem factor was ranked in order of highest to
              lowest value for the same 56 three-digit zip code areas.  On
              the assumption that the  factors are of equal influence, an
              average rank was computed for each of the 56 three -digit  areas.

       (4)    The number of complaint rankings representing actual odor
              problem areas and the average rankings representing the
              problem potential of the same areas were correlated using
              the Spearman rank correlation technique .  The correlation
              coefficient was found to be positive (+0.50).

       It is emphasized that the Spearman analysis was performed only as a gross
indication of the value of the method of potential problem area identification described
above.  A more refined analysis of the value of each potential odor problem factor
as a contributor to the location of potential odor problem areas must await knowledge
of how these factors should be related  to overall influence.  This method is expected
to be much more informative at the five-digit zip code (neighborhood) level than the
three-digit (metropolitan area) level, since the use of five-digit area data will pro-
vide much more accuracy in the location of concentrations of population and indus -
trial odorant sources . However, the availability of additional data (especially
population data on a five-digit zip code basis) is required before potential odor
problems can be identified in terms  of neighborhoods .

-------
Projection of Results to the Nation as a Whole

       In developing the set of potential odor problem factors, principal attention
was focused on metropolitan areas.  Under the criteria outlined in Chapter VII,
seven metropolitan areas were  selected for investigation from among the most
likely odor problem areas in the nation.  Although the seven metropolitan areas
were representative of many geographic and atmospheric regions, kinds of indus-
trial odorant sources,  and types of people in the United States, it is unknown
whether the results of the investigations conducted in those areas are representative
of all metropolitan areas.  However, it is expected that the results are generally
representative of most metropolitan areas, assuming that most metropolitan areas
contain industries known to be potential sources of odorant emission.

       In accordance with the objectives  of the technical field program, the
Engineering-Science, Inc., team of engineers confirmed the locations, areal extent,
and odorant sources of actual odor problem areas identified by the local air pollu-
tion control agencies in the principal cities of each of the metropolitan areas.  The
results of the public opinion surveys conducted in these areas indicated that a large
proportion of residents and businessmen were aware of noticeable odors.  Aware-
ness of odors was found not only in the survey test areas but also in the relatively
odor -free control areas.

       Difficulty arises in attempting to extrapolate the awareness level of percep-
tion to the problem level. There appears to be a very nebulous transition between
an expression of awareness and an expression of problem. In this study, the
differentiation between these levels was attempted through careful development of
the public opinion survey questionnaires.  It was planned that the responses to the
questionnaires would provide the necessary data for an assessment of the extent of
odor problems in the survey areas.

       Although a further difficulty arises in attempting to project the extent of odor
problems beyond the survey areas, assumptions were formulated to assist in over-
coming this problem.  Based on these assumptions, the survey findings were used
to assess the extent of odor problems in the  metropolitan areas of the United States.

       The survey findings of all cities combined (Appendix G) show  that 65 percent
of test area residents stated, without prompting, that air pollution was a problem.
Furthermore, 82 percent of these residents believed odors to be a major element of
air pollution.  From this, it may be assumed that about one-half of the test area
residents would state that air pollution is a problem and odors are a  major element.
Similarly,  about one-quarter of the control area residents would state that air pollu-
tion is a problem and odors are a major element.

       There are an estimated  137 million persons in the  155 metropolitan areas of
the nation. Assuming that two-thirds of these persons are male and  female adults
in households (residents), that only 10 percent live in actual odoT problem areas
(similar to the public opinion survey test areas), and that  90 percent live in relatively

-------
odor-free areas (similar to the survey control areas), then a projected 25 million
residents in the United States could be expected to state that air pollution is a pr$b-<
lem and that odors are a major element.

       Based on these assumptions, further projections of the extent of odor prob-
lems can be made:

       (1)     About one and one-half million residents would voluntarily
              state that odors are a disadvantage to living in their areas
              of their cities.

       (2)     Almost five million residents would admit that odors bothered
              them very much.

       (3)     More than five and one-half million residents would state that
              odors are a continuously serious problem.

       (4)     About three million residents would feel that odor pollution has
              reduced the value of their home property.

       (5)     Almost five million residents would have seriously considered
              moving away from their areas because of odor pollution.

       (6)     Only one-half million residents would have requested some
              authority or agency to take action concerning air pollution.
              The number of. residents that would have requested action
              concerning odor problems would be a small fraction of this.

       Similar projections of the response  of businessmen were  not possible be-
cause of a lack of information concerning the number of small businesses in the 155
metropolitan areas.

       On the basis of the projections given above, a large number of residents
perceive odors as a problem.  Yet only a small percentage of these residents are
motivated to seek recourse. The reasons for this apparent apathy cannot be
inferred from the public opinion survey findings.  Instead, additional study of
attitudes is required to develop meaningful  conclusions.   The remaining section
presents a few observations concerning the legal recourse that is currently avail-
able ,  should such recourse be sought.
LEGAL RECOURSE TO ODOR PROBLEMS

The Foundation of Legal Recourse

       Following the 1955 federal act which authorized the Department of Health,
Education and Welfare  to develop an air pollution control program, numerous state

-------
and local governments responded in kind by establishing the mechanisms by which
research and control programs could be implemented to curb air quality degradation.
In most states the ultimate agency-level authority is vested in the local counterpart
to the Department of Health,  Education and Welfare.  On the urban level there is
often a split in the responsibilities and authority consignments between the local
public health and air pollution control boards.

       The Clean Air Act of  1963 provided the first substantial federal grants to
local air pollution agencies with funds awarded on a three-to-one basis.  More
important, however, from the standpoint of legislation, the Act has been instrumen-
tal in the unprecedented expansion of control programs  on both a  state and local
level.  Now with the increased structural impetus provided by the Air  Quality Act
of 1967, state and local legislative enactments will inevitably give recognition to
the inter-dependency of cities, states, and entire geographic regions in air  quality
management.

       It is worthy of note, however, that although states like California and New
York forged ahead of federal timetables in setting air quality standards consistent
with or exceeding established criteria for particulate matter and  sulfur oxides,
none have succeeded in establishing effective legislation to abate  the emission of
environmental odors.

Enforcement Practice

       The enforcement of established laws, regulations, and policies designed to
control the level of air pollutants generally rests with the designated air pollution
control agency in each urban area.  In instances involving odorant conditions the
most frequently used legislation are local or state ordinances regarding "nuisance"
situations. An excellent example of such a law — one which is more carefully
worded than many,  yet still retains the weaknesses found in all — is the California
Health and Safety Code CH. 2.5, Article 10, Section 24360:

       "A person shall not discharge from any source whatsoever such
       quantities of air contaminants or other material which cause  in-
       jury,  detriment, nuisance, or annoyance to any  considerable
       number of persons or to the public or which endangers the comfort,
       repose, health, or safety of any such persons or the public or which
       cause or have natural tendency to cause injury or damage to busi-
       ness or property."

The requirement of a "considerable number" of persons to constitute a nuisance or
annoyance renders the approach of limited utility. This stipulation,  though  not
present in all similar state and local laws, typifies, nonetheless, the difficulty in
establishing an enforceable case when dealing with the diffuse problem of odor
emissions .  As indicated in Chapter VI,  formal complaints about many problems
are few and infrequent.

-------
       Unlike situations wherein property damage and/or personal health is in
acknowledged jeopardy, strong public reaction to odorant conditions is generally
non-existent. Notable exceptions to the general attitude were found in the cities of
Portland and Tampa.  In one city, major activist groups evolved out of an initial
concern about an odor emission, while in the other,  efforts are being exerted at
this time around a major theme of clean air.

       Enforcement efforts to reduce odor emissions generally await specific com -
plaints.  The attitude of enforcement officers regarding such complaints varied in
each metropolitan  area investigated.  These different attitudes were also reflected
in the speed and manner in which the complaints were handled.  Those agencies
which viewed the complaints as an inconvenience or harassment generally responded
more slowly and with less apparent effort to effect a resolution to the problem.  The
more aggressive agencies tended toward greater promptness in response, odor
source identification, and problem resolution via odor abatement. However, none
of the agencies interviewed by Engineer ing-Science, Inc., could claim significant
control of odorant  conditions  in their areas  even though the sources were fairly
well pinpointed and identifiable. All those interviewed agreed that their regulatory
frameworks were too poorly defined to effect proper enforcement among already
existing industries which were responsible for the emissions . As confirmed by the
survey of local air pollution control agencies, present-day abatement activities are
principally aimed at resolution through the voluntary cooperation of the sources in-
volved .

       Beyond more definitive enforcement  criteria, agency officials have expressed
hope in the efficacy of plan review power, which some have been accorded through
recent legislation.  This authority permits air quality agencies to review plans for
new facilities and to stop issuance of construction permits if reasonable steps are
not taken to insure adequately clean operations.

Judicial Precedence

       Adjudication of cases  involving air pollution emissions is still in the embryonic
stage.  In most instances,  intolerable conditions are resolved through compromises
and concessions at the hearing level (either formal or informal) of the air pollution
control boards.

       To the extent this study was able to determine, none of the metropolitan
areas investigated had records of any substantial fine or other judicial penalty
involving environmental odor levied against an air pollution violator.

The  Enforcement Agency

       Public reaction to odor problems usually takes the form of individual com-
plaints.  Although  conservationist groups are ever-present (in varied strength,
number, and sophistication), few such groups have concentrated on the problem of

-------
environmental odor emissions. Most enforcement agencies are responsive, never-
theless, to the individual complaints reported by local citizens.  They are usually
fully cognizant of the nature and location of the static odors in their community, but
are somewhat frustrated by their inability to effect odor abatement in the majority
of cases.  The enforcement agencies' perception of problem magnitude becomes
accentuated when they occasionally bear the brunt of newspaper criticism.  Some
have attempted to offset unfavorable publicity by educating special interest groups
and the general public via public information programs and speaker bureaus . How -
ever, this has sometimes backfired by creating demands which they have been
unable to fulfill.

       Many of the officials interviewed expressed the wish that industry would
cooperate more fully by doing a better job of countering public pressure for odor
abatement.  It was believed that public information and education by private industry
emphasizing what they are doing to reduce air pollutant by-products would be
immensely helpful.

       To date, the tools  and devices available to enforcement agencies to deal
with odor problems have been virtually nil. Legislative enactments have been
inadequate and measuring devices almost non-existent.  Both the federal govern-
ment and local air pollution agencies, however, recognize the need to acquire
answers .  This study is an indication of such concern.

-------
                                CHAPTER XI
                   CONCLUSIONS AND RECOMMENDATIONS
       The conclusions of this study are presented below.  They are drawn from
the analyses of the study data and the experiences of the study team.  Consistent
with these conclusions  and the study tasks,  recommendations for further research
directed toward an assessment of the social and economic impact of odors are also
presented.
CONCLUSIONS

       Based on the assessment of the national odor problem, it is concluded that
odor problem potential exists in most metropolitan areas in the United States.  But,
primarily because of meteorological factors, the probability of encountering odor
problems is greatest in the Appalachian and California Coastal metropolitan areas.
It is further concluded that a large number of residents — perhaps as many as 25
million — perceive odors as problems  and desire some form of abatement.

       The results of the mail survey and personal interviews of local air pollution
control agencies indicated that when concentrations of air pollutants can be related
to amounts of property damage and/or ill health, abatement criteria (e .g., minimum
levels of emissions) can be established and enforced.  However, in cases of air
pollutants causing only odors,  such relationships are generally non-existent. Con-
sequently, odor abatement criteria, if any, are typically based on number of
complaints and inadequately backed by nuisance laws.

       In order to support more effective odor abatement,  some indication of the
social and economic impact of odors on individuals must be developed. Perhaps
relationships between the presence of odors and a few social and economic  indicators
(e .g., differences in property values) could be estimated.  However,  only a study of
the  attitudes of individuals who may be  affected by odors could provide an assessment
of the overall social and economic impact.

       The basis  for this assertion lies in the conception that attitudes show evalu-
ative personal reactions — basically, tolerance or intolerance — to a situation such
as an odorified environment.  Furthermore, it is the attitudes of individuals toward
the  presence of odors as being intolerable that  is the key to defining the existence of
odor problems.

       It is recognized that there are difficulties associated with obtaining  a social
and economic assessment through an attitude survey.  For example, great  care

-------
must be taken in the development of the survey questionnaire, the selection of the
representative sample of individuals, and the interpretation of results in order that
the assessment could be admissible in judicial proceedings.  However, a single
procedure could be professionally prepared to reduce survey complexities to an
easily applied set of instructions, which could be used by relatively inexperienced
authorities in any community in the nation.  Although other difficulties exist with
regard to assessing the social and economic impact of odors,  it is felt that they
provide no permanent obstacle to success.
RECOMMENDATIONS

       Taking into consideration the present state of knowledge, it is recommended
in cases where air pollutants are found to cause adverse effects on property and/or
health,  as well as intolerable odors, that abatement should be directed to relieve the
adverse effects on property and/or health.  Hopefully, such abatement would result
in eliminating the problem of odors as well.  In cases where odors alone must be
relieved, it is recommended that abatement should be supported with an assessment
of the social and economic impact on the individuals that may be affected.

       The following discussion provides an approach that could be used  in develop-
ing a method of assessment. Because most odor problems are local in terms of the
area affected, the approach is focused on a specific odor problem. The discussion
is meant to suggest the content of a continuation (Phase II) of the present  study.

Objective and Methodology

       The principal objective of further study would be to develop an efficient and
inexpensive method that would provide an adequate assessment of the social and
economic impact of a specific odor problem on a typical community in the United
States.  In order to develop this methodology, a minimum of two odor-affected test
areas and one control area would be established within residential sectors of a
metropolitan area.   The metropolitan area would be representative of many types
of people and kinds of industrial odorant sources in the nation. It would be chosen
within a region of average atmospheric vulnerability to serious odor problems .

       Because of the availability of information obtained during this study, initial
consideration would  be given to the seven metropolitan areas investigated in Phase
I. However, the choice would not be restricted to those areas.

       One of the test areas would be located near an odorant  emitting industrial
plant that has been in operation for a year or more. To minimize the effect of odor
adaptation by residents, the second test area would be located near an identical
industrial plant that  has recently begun operations. The two test areas would be
as near to each other and as similar as possible in all characteristics except for
term of operation of the plants.  The control area,  also with similar characteristics,

-------
would be established nearby but would not be affected by odor ants from either of the
plants.  This arrangement would allow a valid comparison of attitudinal variables
between the residents of each area.

       Recorded panel discussions with residents of each of the test and control
areas would be conducted to identify attitudinal variables that may be related to the
problem  level of odor perception. This would be a preliminary step toward the
development of an effective public attitude survey questionnaire.

       To determine how odors influence the attitudes of residents, a number of
public attitude surveys would be conducted by telephone periodically during the
investigation. (It is recommended that the period of investigation extend up to six
months to account for seasonal influences.) Different sample sizes would be used
to determine the smallest size  permissible to obtain meaningful survey results.
This would provide a guideline  for minimizing survey time and expenses .

       Technical field studies using field sensory techniques would be used to mea-
sure the  intensity, duration,  frequency,  and temporal variation of odors throughout
the test areas .  The results of  scentometer measurements would be compared to the
results obtained by trained odor panel evaluations . Different sized odor panels (of,
say, 8,  12 and 16 panelists) could be used at different times to determine the effect
on the correlation of the two techniques as evidenced in Phase I.  These studies
would be conducted concurrently with the public attitude surveys so that the data
obtained  from both could be used to develop close approximations to existing rela-
tionships .  Meteorological data relating to atmospheric stability and wind charac-
teristics would also be gathered for  the period of investigation.

Development of an Odor Problem Index

       The results of the public surveys and technical field studies between each of
the areas (test area A vs. control area;  test area B vs. control area; test area A
vs . test area B) would be examined for significant differences . Should significance
be found, multivariate techniques (including factor analysis) would  be used to relate
the differences in attitudinal variables to the differences in odor presence (intensity,
duration, etc.), atmospheric (wind speed and direction, etc.), and demographic
(age, income, etc.) variables.

       From the above,  an odor problem index would be developed to represent
an overall attitude differential between any given  test and control area. In addition,
a set of generalized charts would be developed to relate significant odor presence,
atmospheric and demographic variables to the odor problem index.  Finally, a pro-
cedure would be developed to provide step-by-step instructions for the establishment
of an odor problem index for a  specific odor problem. The instructions would include:

-------
       (1)     Specification of the information required.
       (2)     Techniques for collecting and analyzing this information.
       (3)     Approximate costs of various phases of the assessment.

       This method would provide as brief and inexpensive an investigation as
possible to obtain an adequate assessment of the social and economic impact of a
specific odor problem on a typical community in the United States.

-------
                         ANNOTATED BIBLIOGRAPHY
A.     IDENTIFICATION OF MAJOR ODORANTS AND ODORANT SOURCES
American Society for Testing and Materials .  Symposium on Odor.  Special Techni-
       cal Publication No. 164.  Philadelphia: American Society for Testing and
       Materials,  1954.

       A collection of six papers on odor sources, measurement and control
       together with a discussion.
City of Philadelphia,  Pennsylvania, Department of Public Health.  Air Pollution from
       Diesel Powered Vehicles in Philadelphia, 1965.

       A discussion of the major pollutants of diesel engines .  Recommends cita-
       tions for diesel powered vehicles emitting excessive odors.  11 pages.
           .  Air Pollution from Fuel Combustion Processes in Philadelphia, 1966.
       An analysis of major categories of combustion processes in terms of air
       pollution problems. Recommendations are made regarding control measures
      	.  Air Pollution Problems from Refuse Disposal Operations in Philadelphia
       and the Delaware Valley, 1969.
       A discussion of the causes of air pollution from refuse combustion and the
       measures which may be taken to reduce emission, including odor control
       at incinerators . Ten pages .
           .  Sources of Air pollution Emissions in Philadelphia, no date .
       A graphical summary of sources of major pollutants; sulfur dioxide,  nitro-
       gen dioxide, hydrocarbons, and particulates.

-------
Grouse, W. R. and Flynn, N. E .  "Source Inventory IBM System for Particulate and
       Gaseous Pollutants," Journal of the Air Pollution Control Association, XVII,
       No. 8 (August,  1967), 508-511.

       Describes the rationale and methodology for compiling an emissions inven-
       tory,  and the use of a computer for data processing and information retrieval.
Fieldnap, A. C., et al. Warning Agents for Fuel Gases. U.S. Bureau of Mines,
       Monograph No. 4,  1931.

       A reference of perceptible concentrations and characteristics of various
       substances in air.
Gould, Robert F. (ed.)  Flavor Chemistry. Washington, D.C.: American
       Chemical Society Publications, 1966.

       Presents a collection of 15 papers dealing with practical problems associated
       with the presentation and enhancement of tastes and odors in foods .
Kerka, William F. and Kaiser, Elmer R.  "An Evaluation of Environmental Odors, "
       Journal of the Air Pollution Control Association, VII,  No. 4 (August, 1957).

       A survey of most common sources  of environmental odors, together with
       a brief outline of physiological response variables, methods of measure-
       ment, and odor control methods. Five pages.
Magill, Paul I. (ed.) Air Pollution Handbook .  New York:  McGraw-Hill Book
       Company, Inc.,  1956.

       A single volume  introduction to air pollution; its causes, effects, and
       methods of control.
Niagara County, New York.  Department of Health.  Bureau of Air Pollution Control.
       Industrial  odors and Emissions - Niagara Falls Area, no date.

       List of the most serious sources of odor-emissions together with descrip-
       tions of characteristic odor quality. One page.

-------
Odors, the Problem, its Definition and Control.  Proceedings of the Mid-Atlantic
       States Section Semi-Annual Technical Conference, Wilmington,  Delaware,
       November 12, 1965. Pittsburgh, Pennsylvania: Air Pollution Control
       Association,  1965, 80 pages.

       A compilation of papers dealing with odor sources, evaluation, and tech-
       nical and legislative control.
Randolph, T. G.  Human Ecology and Susceptibility to the Chemical Environment
       Springfield, Illinois: Charles C. Thomas, 1962.

       Includes a list of chemical contaminants that produce odors and their
       subsequent effects on health (case study).  Also contained are some
       examples of questionnaires with emphasis on reactions of individuals
       to chemical odors.
Schumann, Charles E. Air Pollutant Emissions in Cincinnati. Cincinnati, Ohio:
       City of Cincinnati, Division of Air Pollution Control and Heating Inspection,
       1968.

       Compilation of emission data based largely on fuel consumption data and
       standard emission factors. No data on odors .  15 pages .
Stern, A. C. (ed.)  Air Pollution.  3 vols .  2nded.  New York:  Academic Press,
       1968.

       The most comprehensive treatment of all phases of air pollution available.
       Sections of most interest relative to odor problems are Part III, Air Pollution
       Meteorology, and Part V, Air Quality and Meteorological Measurements .
Summer, W.  Methods of Air Deodorization.  Amsterdam:  Elsevier,  1963.

       A comprehensive introduction to odors, including theories of olfaction,
       properties of odorants, meteorology, and detailed discussions  of each of
       the common methods of deodorization.

-------
Thring, Meredith W. Air Pollution.  London:  Butterworth Scientific Publications,
       1957.

       Treats air pollution in general, including various sources of air pollution,
       measurement, effects on health, and geographical factors.  Discusses
       elimination of pollution originating from sources such as domestic heating,
       motor vehicle exhausts, and cement manufacture.  Noxious odors are dis-
       cussed in connection with cooking processes such as food preparation,
       tanning,  and fish meal preparation.
U.S. Public Health Service. Air Quality Criteria for Particulate Matter, 1969.

       A compilation of information on particulates in terms of sources and proper-
       ties , and economic and health effects.  One chapter is devoted to odors
       associated with atmospheric particulate matter.  105 pages .
             Air Quality Criteria for Sulfur Oxides, 1967 .  PHS Publ. 19619.
       A compilation of information on sulfur oxides in terms of sources, proper-
       ties, means of measurement, and effects .  Includes discussion of human
       olfactory response to various concentrations.  178 pages .
      	.  Air Quality Data, From National Air Sampling Networks and Contri-
       buting State and Local Networks 1964-1965. Cincinnati, Ohio: U.S. Public
       Health Service, Robert A. Taft Sanitary Engineering Center, 1966.
       A summary of air quality data for 1966, arranged by pollutants.  Includes
       summary plots and detailed tabulations .  No odor data per se. 157 pages.

-------
B.     MEASUREMENT OF INDUSTRIAL ACTIVITY
Adelman, Morris .  "Measurement of Industrial Concentration," Review of Economics
       and Statistics, XXXIII (November, 1951),  269-296.

       A major study dealing with the problem of industrial concentration, empirical
       and theoretical.  Various measures and economic variables are discussed in
       terms of concentration and size.
Bain, Joe S. Barriers to New Competition .  Cambridge, Massachusetts: Harvard
       University Press, 1965.

       A cross-sectional study of twenty manufacturing industries .  Study focused
       on the relation between "the condition of entry" of new firms into an industry
       and the workability of competition in that industry.  Recommendations
       offered for the design of a more effective antitrust policy.
George, Edwin.  "Four Comments on 'The Measurement of Industrial Concentration':
       III, " Review of Economics and Statistics,  XXXIV (May,  1952), 168-172.

       Comments on Adelman's article on concentration.
Kays en, Carl and Turner, Donald F. Antitrust Policy:  An Economic and Legal
       Analysis .  Cambridge, Massachusetts:  Harvard University Press, 1959.

       An evaluation of the limits and defects of traditional antitrust policy and a
       carefully detailed projection of possible reforms .  Surveys the relative
       importance of different kinds of market structure in the economy.
Loquerics, P. and Stanley,  W. J.  Air Pollution Manual of Coding. U.S. Public
       Health Service, 1968.

       This manual provides listings of certain industrial activities by Standard
       Industrial Classification codes and the equipment associated with the air
       pollution generated by those activities.

-------
National Bureau of Economic Research. Business Concentration and Price Policy.
       Princeton, New Jersey:  Princeton University Press, 1955.

       Separate discussions by various authors on subjects pertaining to business
       size, monopoly power, vertical integration, conglomerates, price discrimi-
       nation, effects of taxes on concentration, etc.  Mainly empirical investigations

-------
C .     ATMOSPHERIC FACTORS RELATED TO ODOR PROBLEMS
Bay Area Pollution Control District.  Technical Services Division. An Agricultural
       Burning Control Program Derived From Inversion Climatology in the San
       Francisco Bay Area.  Information Bulletin No. 11-68.  San Francisco:  Bay
       Area Pollution Control District,  1968.

       The development of meteorological criteria for permitting the open burning
       of deciduous fruit and nut trees is described.  Basic criterion is a minimum
       of 2,500 feet vertical mixing.  15 pages .
Bay Area Pollution Control District.  Air Pollution and the San Francisco Bay Area,
       1969.

       A general discussion of air pollution. Provides information on different
       kinds of smogs,  various effects of air pollution,  and statistics on counties
       in and around San Francisco area.  A meteorological and geographical
       explanation of smog in the Bay Area is presented.  Bay Area Air Pollution
       Control District and its functions are described.
Boettger, Carl M.  "Air Pollution Potential East of the Rocky Mountains:  Fall,  1959, "
       American Meteorological Society Bulletin, XLII (September, 1961), 615-620.

       The weather over the eastern two-thirds of the United States was monitored.
       Air pollution potential forecasts were made and disseminated to affected
       cities . The results tended to confirm the validity of criteria developed for
       wind and subsidence conditions .
Danis, A. L.  "Effect of Local Weather on Air Pollution Problems, " Journal of the
       Sanitary Engineering Division, LXXXIII, No. SA-6, Part 1 (December, 1957),
       1463-1 to  1463-10.

       A short introduction to air pollution meteorology with special emphasis on
       diffusion climatology.  The effects of inversion and wind conditions and
       stack height on diffusion of airborne pollutants is discussed.
Fuller, W. J. Specifications for Data Acquisition System Air Monitoring.  Portland,
       Oregon:  Columbia-Willamette Air Pollution Authority, 1968.

       Technical specifications are given for equipment used to monitor air quality
       and local meteorology.  12 pages .

-------
Harding, C.I. and Kelly, T. R.  "Horizontal and Vertical Distribution of Corrosion
       Rates in an Industrialized Seacoast City," Journal of the Air Pollution Control
       Association, XVII, No. 8 (August, 1968).

       Corrosion rates of mild carbon steel plates over the metropolitan Jacksonville
       area in a 44-station network were found to provide a good description of 50
       dispersion  patterns from major sources. Three pages.
Holzworth, George C. "Estimates of Mean Mixing Depths in the Contiguous United
       States, " Monthly Weather Review, XCII (May, 1964), 235-242.

       Mean radiosonde observations and normal maximum surface temperatures
       are used with the assumption of a dry adiabatic lapse rate to estimate
       monthly mean maximum mixing depths for 45 stations in the contiguous
       United States.
      	.  "Mixing Depths, Wind Speeds, and Air Pollution Potential for Selected
       Locations in the United States, " American Meteorological Society Bulletin,
       XLVIII, No. 4 (October,  1967).

       Daily estimates of mixing depths and average wind speeds through the mixing
       layers were calculated and summarized for seven locations in the United
       States.  This information was used in an urban diffusion model to calculate
       theoretical values of relative pollutant concentrations for  four major cities .
Hosier, Charles R.  "Low-Level Inversion Frequency in the Contiguous United
       States, " Monthly Weather Review, LXXXIX (November,  1961), 319-339.

       An analysis of data from weather bureau radiosonde stations in the contiguous
       United States allowed development of estimates of the percentage of the time
       in which inversions occur below 500 feet for each of the four seasons .
Miller, Marvin E .  "Semi-Objective Forecasting of Atmospheric Stagnation in the
       Western United States," Monthly Weather Review, XCE, No. 1 (January,  1964),

       Forecasts of air pollution potential by semi-objective means are demonstrated
       and found to be useful in delineating stagnation areas associated with quasi-
       stationary anticyclone and ridge aloft over the western United States .

-------
Miller, Marvin E . and Holzworth, George C.  "An Atmospheric Diffusions Model
       for Metropolitan Areas, " Journal of the Air Pollution Control Association,
       XVII, No. 1 (January, 1967).

       An urban diffusion model is presented, based on assumed uniform distribu-
       tion of pollutant sources, and vertical diffusion from line sources up to the
       top of the mixing layer. A computer is not required.  Five pages.
Niemeyer, L. O.  "Forecasting Air Pollution Potential, " Monthly Weather Review,
       XXXVIH (March,  1960).

       A procedure for forecasting weather conditions conducive to high air pollu-
       tion levels over a large area as a primary alerting system for potentially
       hazardous conditions is presented.  The results of experiments to test the
       procedure are described.
Panofsky, H. A. and Prasad, B.  "The Effects of Meteorological Factors on Air
       Pollution in a Narrow Valley, " Journal of Applied Meteorology, VI (June,
       1967), 493-499.

       A simple mathematical model for prediction of variations in air pollution
       from a large number of low-level sources in a narrow valley is presented
       together with results of tests of the model using data for Johnstown,
       Pennsylvania.
Shaw, A. J. The Climatology of Air Pollution Potential - Hillsborough County 1968.
       Tampa, Florida:  Hillsborough County Department of Health,  1968.

       A thorough and detailed study of the climatology of Hillsborough County,
       especially as it relates to air pollution.  85 pages.
Smith, David B.  "Tracer Study in an Urban VaUey, " Journal of the Air Pollution
       Control Association, XVIII,  No. 9 (September,  1968), 600-604.

       A description of the results of a series of nighttime tracer experiments con-
       ducted during the autumn of 1966 in  the industrialized valley of Johnstown,
       Pennsylvania.  An urban heat island effect is observed which modifies clas-
       sical drainage flow patterns.

-------
Stern, A. C. (ed.) Air Pollution.  3 vols .  2nded. New York: Academic Press,
       1968.

       The most comprehensive treatment of all phases of air pollution available.
       Sections of most interest relative to odor problems are Part III, Air Pollu-
       tion Meteorology,  and Part V, Air Quality and Meteorological Measurements
Summer, W. Methods of Air Deodorization. Amsterdam:  Elsevier,  1963.

       A comprehensive introduction to odors,  including theories of olfaction,
       properties of odorants, meteorology,  and detailed discussions  of each of
       the common methods of deodorization.
Turner, D. Bruce.  Workbook of Atmospheric Dispersion Estimates.  U.S. Public
       Health Service, PHS Publ. No. 999-AP-26,  1967.

       Discusses methods of practical application  of the binormal continuous plume
       dispersion model to estimate concentrations of air pollutants.  Numerous
       special topics are covered,  and 26 example problems are presented. 84
       pages.


U.S. Atomic  Energy Commission.  Division of Technical Information.  Meteorology
       and Atomic Energy. Edited by David H. Slade. AEG Research and Develop-
       ment Reports TID 24190, 1968.

       An introduction to meteorology, followed by more detailed sections on
       diffusion and transport theories and experimental procedures.  445 pages.

-------
D.     DEMOGRAPHIC FACTORS RELATED TO ODOR PROBLEMS
Baier, M.  "Zip Code - New Tool for Marketers, " Harvard Business Review, XLV
       (January-February, 1967),  136-140.

       An introduction to various ways in which postal zip codes can be utilized
       in business activity,  and some examples of the way they are now being
       utilized. Significance as well as some deficiencies of the system are dis-
       cussed.
Breslow, L.  Air Pollution:  Effects Reported by California Residents .  Berkeley,
       California: State of California, Department of Health,  1956.

       Investigation of the social and psychological aspects of daily living. Char-
       acteristics of samples such as sex, age, education, occupation, family
       income, etc.  Major findings:  45 percent of California's adult population
       are bothered by air pollution.  Eye and nasal irritation most frequent.
       People with chronic respiratory conditions are more sensitive.  Most
       serious and extensive complaints  from Los Angeles County.
Commercial Atlas & Marketing Guide .  100th ed. New York: Rand McNally &
       Company, 1969.

       Three-digit postal zip code, SMSA, and individual state geographic and
       business data.
Crowe, M. Jay.  "Toward a 'Definitional Model' of Public Perception of Air Pollution, "
       Journal of the Air Pollution Control Association,  XVIII (March, 1968), 154-158.

       From a public attitudinal survey conducted in a city in southwest Pennsylvania,
       different levels or aspects of public perception of air pollution are suggested.
       This model is  analyzed in terms of selected social characteristics,  such as
       education, socio-economic status, and residence.
Hillsborough County,  Florida.  Department of Health.  Division of Environmental
       Engineering.  Urbanization and Air Pollution, 1969.

       Discusses the  relationships which exist between increases in population
       and economic growth and environmental pollution.  Vehicle emission con-
       trols are explained.  Eight pages .

-------
Jones, Richard W.  "New Developments in Zip Code Marketing, " The Marketing
       Magazine (June,  1969), A1-A6.

       An introduction to the origin of postal zip code usage.  Various uses of
       zip code in business activities are introduced. Some tips as to how to
       organize a marketing  system by using zip codes .
McMullen, Thomas B., et al.  "Air Quality and Characteristic Community Para-
       meters ," Journal  of the Air Pollution Control Association, XVIII, No. 8
       (August,  1968), 545-549.

       Statistical correlations between all pairs of 16 selected air quality measure-
       ments and 13 selected community parameters were analyzed and highest
       correlations discussed. Odors are not included per se.
U.S. Bureau of the Census.  Estimates of the Population of Standard Metropolitan
       Statistical Areas: July 1,  1965.  Current Population Reports, Series P-25.
       No. 371, 1967.

       Contains data on population and population changes in Standard Metropolitan
       Statistical Areas.
U.S. Public Health Service. National Air Pollution Control Administration. Report
       for Consultation on the San Francisco Bay Area Air Quality Control Program,
       1968.

       One of the whole series enhancing Baltimore,  Pittsburgh,  Cleveland, Cincinnati,
       Buffalo, Minneapolis-St. Paul, etc.  The literature introduces the Air Quality
       Act, and discusses the size of the regions chosen,  various technical, socio-
       logical, geographical,  and meteorological factors, and proposes the region
       under consideration as an Air Quality Control region.
Zip Code Area Data, Individual Income Tax Returns.  Washington,  D.C.: U.S.
       Internal Revenue Service,  1966.

       Lists the total number of exemptions and the total tax in each three-digit
       zip code area in the United States .  Data broken down into postal regions
       (first digits).  Three-digit maps for each region are included.

-------
E .     ODOR MEASUREMENT AND EVALUATION TECHNIQUES
Adams, Donald F., Young, Francis A. and Luhr, R. A. "Evaluation of an Odor
       Perception Threshold Test Facility, " Journal of the Technical Association
       of Pulp and Paper Industries,  LI, No .  3 (March,  1968) .

       The design of a facility for  the study of odor perception and objectionability
       thresholds within a large human population is described.  The results of
       tests on nearly 6,000 untrained subjects are reported and analyzed.
American Society for Testing and Materials .  Manual on Sensory Testing Methods .
       Special Technical Publication No. 434.  Philadelphia: American Society
       for Testing and Materials,  1968.

       A guide for the  technical man who is not an expert in the field, but who is
       confronted with the need to conduct sensory tests.  Includes sections on
       general requirements,  forms of tests, applications, and statistical analyses.
           .  Symposium on Odor. Special Technical Publication No. 164.
       Philadelphia: American Society for Testing and Materials,  1954.

       A collection of six papers on odor sources, measurement, and control
       together with a discussion.
Byrd, J. F.  "Demonstration  - Syringe Odor Measurement Technique, " Journal of
       the Air Pollution Control Association, VII, No.  1 (May, 1957).

       A review of the rationale and methodology of the syringe odor measurement
       technique, which utilizes medical type syringes for aliquot withdrawal from
       stacks and dilution to threshold.  Two pages .
City of Portland, Oregon. Bureau of Health.  Soiling Index.  Information Bulletin
       No. 9, 1966.

       Summary and explanation of soiling index data for Portland are presented.
       Two pages.

-------
Core, John E. and Arnett, D. E.  "The Analysis of C2 Through C- Atmospheric
       Carbons by Gas Chromatography, " Scientific Programmer (November, 1968),
       pp. 21-22.

       An outline of equipment and methodology used to individually monitor at
       least 14 separate hydrocarbons.  18 pages .
Core, John E., Berg, Neil and Roach, M. D. The Relationship of Degree Days and
       Other Parameters on Portland's Soiling Index.  Portland, Oregon:  City of
       Portland, Bureau of Health, no date.

       Soiling index data for Portland is correlated with meteorological data and
       data on other measures of air quality.  Space heating is found to be the
       single most important source of soiling.  18  pages.
Deininger, Nicholas and McKinley, Russell W.  "The Design, Construction, and
       Use of an Odor Test Room, " Symposium on Odor.  Philadelphia:  American
       Society for Testing and Materials,  1954.

       Provides a detailed consideration of the requirements for an odor test room,
       Dimensions, materials, systems of ventilation, and odor injection are dis-
       cussed . Ten pages.
Faulkner, Dean L., Schumann, Charles E. and Gruber, Charles W.  Final Report -
       Particulate Sampling by Adhesive -Coated Materials, October, 1,  1969.
       Cincinnati, Ohio: City of Cincinnati, Division of Air Pollution Control and
       Heating Inspection,  1969.

       Summarizes five-year federally supported project.  Concludes that adhesive
       materials offer an attractive means of measuring particulate concentrations.
Fox, E . A. and Gex, V. E.  "Procedures for Measuring Odor Concentration in Air
       and Gases, " Journal of the Air Pollution Control Association, VII, No. 1
       (May, 1957).

       A detailed procedure for the quantification of odor intensities in terms of
       dilutions to threshold is outlined. All aspects of the proposed method'are
       covered:  sampling and dilution techniques, apparatus, interferences,  and
       calculations. Two pages.

-------
"Gases and Vapors:  Methods Giving Quantitative Results in the Field, " Industrial
       Hygiene and Toxicology, Vol. I.  Edited by Frank A.  Patty.  2nded.  London:
       Interscience Publishers, Ltd., 1958.

       A discussion of practical problems associated with quantitative assessment
       of odors in the field, and of the physiological response patterns of human
       subjects to odors .  Five pages .
Hemeon, Wesley, C. L.  "Technique and Apparatus for Quantitative  Measurement
       of Odor Emissions," Journal of the Air Pollution Control Association, XVIII,
       No . 3 (March, 1968).

       A description of instruments and techniques developed by the author for
       the quantitative measurement of odor emissions.  Description includes
       an odor sampling device and a device for controlled dilutions prior to
       sensory evaluation.
Hillsborough County, Florida.  Department of Health.  Division of Environmental
       Engineering. A Comparative Study of Air Pollution - Davis Island,
       Hillsborough County, Florida, 1968.

       A study of a residential area located near major sources of air pollution.
       Reports include analyses of ambient air data and identifies firms probably
       responsible for complaints .  48 pages.
Horstman, S.W., Wromble, R. F. and Heller, A. N.  "Identification of Community
       Odor Problems by Use of an Observer Corps, " Journal of the Air Pollution
       Control Association, XV,  No. 6 (June, 1965), 261.

       To evaluate the major odors affecting a community, approximately 120
       high school students recorded odor observations .  Simultaneous odor
       observations were made three times daily for 14 consecutive days each in
       a fall and spring survey.  The observers were tested for olfactory sensi-
       tivity.  Population characteristics were recorded for analysis of the human
       variables .  These data were analyzed to show the areas most severely
       affected,  the major objectionable odor types, the time when odors were the
       most frequent, and the associated weather parameters .  A critique of the
       procedures employed and the results of the  survey are presented (JAPCA).

-------
Huey, Norman A., et al.  Odor Measurement Techniques in.  Third Year Final
       Report.  Cincinnati,  Ohio:  City of Cincinnati, Division of Air Pollution
       Control and Heating Inspection, 1960.

       Describes the development and the field testing of scentometer.  15 pages.
Huey, Norman A., Broering, Louis C. and Gruber, Charles W.  Odor Measurement
       Techniques II.  Second Year Final Report.  Cincinnati, Ohio: City of
       Cincinnati, Division of Air Pollution Control and Heating Inspection, 1959.

       Report describes development of scentometer and contains recommended
       procedures for field odor investigations.  Correlations are investigated
       between meteorological variables with complaint frequency.  26 pages.
"It Pays to be Nosy, " Marathon World , VI, No. 1 (Winter,  1969).

       A short description of a mobil facility developed by Marathon Oil Company
       for monitoring meteorological and air quality conditions around Marathon
       installations.  One page.
Jenkins, H. N.  and Harris, T. O.  Interstate Air Pollution Study: Phase II Project
       Report:  IV Odors - Results of Surveys .  Cincinnati,  Ohio:  U.S. Public
       Health Service, Robert A. Taft Sanitary Engineering Center,  1966.

       A survey undertaken to investigate the odor problems in the St.  Louis
       metropolitan area is described. Observations were made at 79  different
       places,  most of which were at fire stations. Odors were classified into
       ten different types and the type  of odor, frequency of detection,  data and
       time of day were reported.  Meteorological conditions were also considered.
Nickol, G. B. "Rank Order Method for Evaluating Stack Gases, " Journal of the Air
       Pollution Control Association,  VII, No. 1 (May, 1957).

       A threshold dilution-based method is presented for the assessment of odor
       intensities in stack gases.  One page.

-------
Odors, the Problem, its Definition and Control.  Proceedings of the Mid-Atlantic
       States Section Semi-Annual Technical Conference,  Wilmington,  Delaware,
       November 12, 1965. Pittsburgh, Pennsylvania: Air,Pollution Control
       Association,  1965.

       A compilation of papers dealing with odor sources, evaluation, and
       technical and legislative control.  80 pages.
Stern, A. C. (ed.)  Air Pollution.  3 vols .  2nded.  New York:  Academic Press,
       1968.

       The most comprehensive treatment of all phases of air pollution available .
       Sections of most interest relative to odor problems are Part III, Air Pollu-
       tion Meteorology, and Part V, Air Quality and Meteorological Measurements ,
Stoldt, Stephen H. and Turk, Amos. Bromination Reactions on Absorbent Surfaces.
       Philadelphia, Pennsylvania:  State of New Jersey, Department of Health, Air
       Pollution Control Program, 1968.

       An investigation of the mechanisms by which brominated activated carbon
       removes olefins from the air.  Six pages.
Sullivan, D. C., Adams, D. F. and Young, F. A. Design of an "Odor Perception
       and Objectionability Threshold" Test Facility , Vol. II.  New York:
       Pergamon Press,  1968.

       Design requirements for facilities to study the thresholds of odor perception
       and objectionability in large populations are discussed.  A mobile facility
       incorporating these features is described together with the procedures used
       for administrating tests to subjects .
Sullivan, Ralph].  Air Pollution Aspects of Odors .  Bethesda,  Maryland: Litton
       Systems, Inc., 1969.  Draft copy.

       The nature and the sources of odors are introduced.  Odor problems and
       their abatement are discussed.  Methods of quantifying the odor problem
       by means of sensory as well as  instrumental investigations are described.

-------
Turk, Amos.  "Concentrations of Odorous Vapors in Test Chambers," Basic
       Principles of Sensory Evaluation. Special Technical Publication No. 433.
       Philadelphia: American Society for Testing and Materials, 1968.

       A presentation of mathematical derivations of expressions which permit
       prediction of odorant concentrations in terms of chamber volume,  time,
       odor injection rate, and ventilation rate.  Five pages.
Turk, Amos and Mehlman, Stanley.  "Correlations Between Instrumental and Sen-
       sory Characterization of Atmospheric Odors, " Correlation of Subjective-
       Objective Methods in the Study of Odors and Taste.  Special Technical
       Publication No. 440. Philadelphia: American Society for Testing and
       Materials,  1968.

       The sampling and characterization of atmospheric odor sources are dis -
       cussed. Sensory evaluation of atmospheric odors and correlation with
       instrumental characterization are included.  Nine pages.
Turk, Amos and Wittes, Janet. "The Selection of Judges for Odor Discrimination
       Panels," Correlation of Subjective-Objective Methods in the Study of Odors
       and Taste.  Special Technical Publication No. 440.  Philadelphia: American
       Society for Testing and Materials,  1968.

       Criteria for the screening of candidates for a sensory odor discrimination
       panel and a screening procedure is outlined which employs three successive
       tests are discussed.  Statistical significance of difference between scores
       is analyzed.  22 pages.
U.S. Public Health Service, Division of Air Pollution, Kansas City, Kansas, and
       Kansas City, Missouri. An Odor Survey of the Two Kansas Citys, 1965.

       An air resource management program is recommended to deal with
       regional odor problems .  Ambient air odor intensity data are included
       based on observations by Fire Department personnel at prescribed
       times . 46 pages.

-------
Wohlers, Henry C.  "Recommended Procedure for Measuring Odorous Contaminants
       in the Field, " Journal of the Air Pollution Control Association, XVII
       (September,  1967), 609-613.

       Describes subjective and objective methods for measuring odorous con-
       taminants in the field. Some weaknesses in using human odor perception
       as a means of measurement are discussed.

-------
F.     COMMUNITY PERCEPTION OF ODOR PROBLEMS
Bates, D. V.  "Health Costs of Diseases Related to Air Pollution, " (Pollution and
       Our Environment) Conference Background Papers,  I.  Montreal, Canada,
       1967.

       It seems very likely that chronic bronchitis is aggravated by urban air
       pollution and its morbidity increased as a result of this factor. About
       20 percent of health costs of chronic bronchitis and emphysema in the
       population might be partly attributed to the effects  of air  pollution.
Breslow, L.  Air Pollution:  Effects Reported by California Residents. Berkeley,
       California: State of California, Department of Public Health, 1956.

       An investigation of the social and psychological aspects of daily living.
       Characteristics of samples such as sex, age, education,  occupation,
       family income, etc.  Major findings:  45 percent of California's adult
       population are bothered by air pollution.  Eye and nasal irritation most
       frequent.  People with chronic respiratory conditions are more sensitive.
       Most serious and extensive complaints from Los Angeles County.
Crowe, M. Jay.  "Toward a 'Definitional Model' of Public Perception of Air Pollution, "
       Journal of the Air Pollution Control Association, XVIII (March, 1968), 154-158.

       From a public attitudinal survey conducted in a city in southwest Pennsylvania,
       different levels or aspects of public perception of air pollution are suggested.
       This model is analyzed in terms of selected social characteristics, such as
       education, socio-economic status, and residence.
DeGroot, Ido.  "Trends in Public Attitudes Toward Air Pollution," Journal of the
       Air Pollution Control Association,  Xn (October, 1962), 679-682.

       An attempt to bring together findings related to people's attitudes toward
       air pollution from a series of studies conducted by or for the U .S. Public
       Health Service.  Trends are noted. Findings are presented in the light of
       planning for the implementation of air pollution abatement measures and
       probable public reaction to them.  No special reference to the odor prob-
       lems.

-------
 'Gases and Vapors:  Methods Giving Quantitative Results in the Field, " Industrial
       Hygiene and Toxicology, Vol. I. Edited by Frank A. Patty.  2nded.  London:
       Interscience Publishers, Ltd.,  1958.

       A discussion of practical problems associated with quantitative  assessment
       of odors in the field, and of the physiological patterns of human subjects to
       odors . Five pages .
Heller,  A. N. Methods of Evaluating Socioeconomic Effects in Air Pollution .  (In
       the Proceedings of the International Conference on Atmospheric Emissions
       from Sulfate Pulping, Sanibel Island. April 28, 1966.) Ed. by E. R.
       Hendrickson, Deland, Florida: E. O. Painter Printing Company, no date.

       An opinion survey used to study public attitudes toward air pollution is
       presented. The questionnaire, sample selection and composition, inter-
       view situation structure, and survey results are explained.
Hillsborough County,  Florida.  Department of Health .  Division of Environmental
       Engineering.  A Comparative Study of Air Pollution - Davis Island,
       Hillsborough County, Florida, 1968.

       A study of a residential area located near major sources of air pollution.
       Reports include analyses of ambient air data and identifies firms probably
       responsible for complaints .  48 pages.
Jennings, Burgess H. Interaction of Man and His Environment.  New York:
       Plenum Press, 1966.

       Mainly concerned with environmental health, how man perceives various
       environmental effects such as air pollution, and how his physiology is
       affected by them.  A cursory survey of man's relation to odor is made.
       Control methods of malodors are suggested.
Medalia, Nahum Z.  Community Perception of Air Quality:  An Opinion Survey in
       Clarkston, Washington. Cincinnati, Ohio: U.S. Public Health Service, 1965,

       In a community with a population of 7,000 and located approximately four
       miles downwind from a pulp mill, a public opinion survey was taken to
       analyze the environmental stress of air pollution on a sample of household
       heads and spouses .  Two principal attitudinal dimensions — awareness
       and concern — were investigated.

-------
Randolph, T. G.  Human Ecology and Susceptibility to the Chemical Environment.
       Springfield, Illinois:  Charles C. Thomas, 1962.

       Includes a list of chemical contaminants that produce odors and their
       subsequent effects on health (case study). Also contained are examples
       of questionnaires with emphasis on reactions of individuals to chemical
       odors.
Stalker, W. W.  "A Method for Using Air Pollution Measurement and Public Opinion
       to Establish Ambient Air Quality Standards, " Journal of the Air Pollution
       Control Association, XVH (March,  1967), 142-147.

       Air quality was comprehensively evaluated by means of sampling at 21
       locations over metropolitan Birmingham during a period of one year.
       Thousands of integrated samples of three common atmospheric gaseous
       pollutants and two common participate pollutants were collected and
       analyzed. The relationships between ambient air quality and neighbor-
       hood opinion effects on health and property are evaluated statistically.
U.S. Public Health Service.  People and Air Pollution:  A Study of Attitudes in
       Buffalo, New York (An Interdepartmental Report),  1962.

       Examines questions relative to the levels of awareness and concern
       about air pollution. Hypotheses relating public attitudes to air pollu-
       tion and certain social factors are tested.
Williams, J. D. Interstate Air Pollution Study:  Phase II Project Report: VI Effects
       of Air Pollution.  Cincinnati, Ohio: U.S. Public  Health Service, Robert A.
       Taft Sanitary Engineering Center, 1966.

       Report of a survey in the St. Louis metropolitan area. Sections are devoted
       to the following: (1) methods available for setting air quality goals, (2)
       deleterious effects of sulphur oxides, HjS, CO,  oxidants and particulate
       matter, (3) odors, (4) effects on asthma sufferers,  (5) aeroallergens, (6)
       deterioration of steel, nylon, and cotton fabrics, (7) vegetation damage,
       (8) effects of air pollution on property values, (9) public opinion survey,
       (10) summary of effects on human health, and (11) summary of economic
       effects.

-------
Zotterman, Y.  (ed.)  Olfaction and Taste.  Vol. I. New York: Pergamon Press,
       1963.

       A collection of research papers dealing with physiological response mech-
       anisms of man and animals to tastes and odors.  Orientation is highly
       theoretical.

-------
G.     ODOR ABATEMENT AND CONTROL
American Society for Testing and Materials.  Symposium on Odor. Special Tech-
       nical Publication No.  164.  Philadelphia: American Society for Testing and
       Materials, 1954.

       A collection of six papers on odor sources, measurement, and control
       together with a discussion.
Bay Area Air Pollution Control Districts .  Air Pollution and the San Francisco Bay
       Area,  1969.

       A general discussion of air pollution. Provides  information on different
       kinds of smogs, various effects of air pollution, and statistics on counties
       in and around San Francisco area.  Meteorological and geographical
       explanation of smog in the Bay Area is presented.  Bay Area Air Pollution
       Control District and its functions are described.
City of Cincinnati, Ohio.  Division of Air Pollution Control and Heating Inspection.
       Ordinance Regulations Pertaining to Air Pollution Control and Heating,
       Mechanical Ventilation and Refrigeration of the City of Cincinnati, 1965.

       A basic air pollution regulation.  Odors are in violation if found to cause
       injury, annoyance, or detriment to the public or to any person.  25 pages.
City of Kansas City, Kansas.  Kansas City, Kansas Air Pollution Ordinance: 1966,
       Ordinance No. 45834.

       Odors are covered only by implication under nuisance statute.
City of Kansas City, Missouri.  Committee Substitute for Ordinance No. 36539
       An Ordinance, 1969.

       Proposed air pollution ordinance to bring City statute into conformity
       with State regulations.  Odors handled under public nuisance section.

-------
City of Philadelphia, Pennsylvania.  Department of Public Health.  Air Management
       Services . Air Pollution Code.

       Contains the sections of the General Ordinances of the City of Philadelphia
       dealing with air pollution. Delineates powers of City agencies relative to
       air pollution.  Includes general nuisance statute relative to odors .  12 pages.
           .  Air Pollution from Diesel Powered Vehicles in Philadelphia,  1965.
       A discussion of the major pollutants of diesel engines . Recommends
       citations for diesel powered vehicles emitting excessive odors.  11 pages.
           .  Air Pollution from Fuel Combustion Processes in Philadelphia, 1966.
       An analysis of major categories of combustion processes in terms of air
       pollution problems .  Recommendations are made regarding control
       measures.
      	.  Air Pollution Problems from Refuse Disposal Operations in Philadelphia
       and the Delaware Valley, 1969.
       A discussion of the causes of air pollution from refuse combustion and the
       measures which may be taken to reduce emission, including odor control
       at incinerators . Ten pages .
      	.  Amendments of Regulations of the Air Pollution Control Board, City of
       Philadelphia, Relating to Control of Incinerators, 1969 .
       The emissions standards for incinerators .  Includes a provision that gases
       from the primary combustion chamber shall be maintained for a sufficient
       length of time at and as a temperature not to be less than 1400°F in order
       to prevent odorant emission.
           .  Regulations of the Air Pollution Control Board, Revised 1966.
       Air pollution regulations for Philadelphia.  Odors are treated under the
       nuisance statute .   16 pages .

-------
City of Portland, Oregon. Bureau of Health.  Incinerators .  Information Bulletin
       No. 11.  2nd Revision, 1967.

       A summary of City's Air Quality Control Code provisions relative to
       incinerators . One page.
Columbia-Willamette Air Pollution Authority.  Open Outdoor Fires .  Information
       Bulletin No. 3, 1968.

       A basic regulation on controlling open outdoor fires.  Includes ban on any
       open fires which generally cause obnoxious odor.  Four pages .
           .  Prohibited Emissions .  Information Bulletin No. 2, 1968.
       Lists emission standards for visible emissions, particulates and odors,
       Odor standards specify maximum intensities in terms of dilutions to
       threshold.  Two pages.
           .  Rules II, 1969. Draft copy.
       A basic regulatory document.  Includes odor provisions based on both
       nuisance and scentometer criteria.  Scentometer standards vary with
       land use.
      	.  Statutory Authority and Powers of the Columbia-Willamette Air
       Pollution Authority.  Appendix A - Information Bulletin No. 1,  1968.
       Outlines the scope, powers, organization, and financing of the agency.
       Seven pages.
Erie County, New York. Department of Health.  Erie County Sanitary Code,
       Article XIV, Air Pollution Control, 1969.

       A basic air pollution law that treats odors only under nuisance provisions,
       49 pages.

-------
Faith, W. L.  "Odor Control in Cattle Feed Yards, " Journal of the Air Pollution
       Control Association,  XIV (November, 1964), 459-461.

       A discussion of types  of odor problems from cattle feed yards,  some odor
       control methods,  and  the state of cattle-feeding industry in general.
Fink, R . H.  "A Management Approach to Odor Control, " Journal of the Air Pollu-
       tion Control Association , VII, No. 1 (May, 1957).

       A management methodology for dealing with odors is outlined.  The
       approach entails source identification followed by an aggressive and
       persistent control program .  Two pages.
Gruber, Charles A.  Odor Pollution From the Official's Viewpoint.  Cincinnati,
       Ohio: City of Cincinnati, Division of Air Pollution Control and Heating
       Inspection, no date.

       Discusses the unique problems associated with odor abatement, together
       with case histories of specific problems.  An approach to odor quality
       measurement is proposed.  Ten pages .
Gruber, Charles W. andjutze, G. A.  Teamwork in Air Pollution Control.
       Cincinnati, Ohio: City of Cincinnati, Division of Air Pollution Control and
       Heating Inspection,  1960.

       Describes the program by which City of Cincinnati provides air pollution
       control services to neighboring communities on a contract basis.  Two
       pages.
Hatchard, R. E .  Intercommunity Cooperation in Air Resources Management.
       Portland, Oregon:  City of Portland, Bureau of Health, 1966.

       Discusses the factors promoting development of regionalism in air
       quality management, together with an outline of the growth of a regional
       air pollution agency in the Portland area. Seven pages .

-------
Hillsborough County,  Florida.  Department of Health.  Division of Environmental
       Engineering.  Hillsborough County Pollution Control Commission - Chapter
       67-1504, House Bill No. 2127.

       Presents the State of Florida statute creating the Hillsborough County
       Pollution Control  Commission.
           .  Hillsborough County Pollution Control Commission - Rules and Regu-
       lations .

       Contains technical and administrative regulations covering air and water
       pollution. No mention of odors.  34 pages.
           .  Hillsborough County's Program for Clean Air, no date .
       Outlines agency activities including emissions inventories, ambient air
       quality measures, and meteorological studies . Eight pages .
           .  Report of Health Department Services, 1969.
       Discusses major pollution problems and actions taken to achieve abate-
       ment.  26 pages.
Kollatz, Charles W. Annual Report of the Erie County Health Department, Division
       of Air Pollution Control,  Erie,  New York: Erie County Department of
       Health,  1968.

       Public information document notes 48 odor complaints in the year 1968
       out of a total of 282 complaints.  11 pages .
Lemkuhl, Richard B. Intercommunity Air Pollution Control Program Activities for
       1968.  Cincinnati, Ohio:  City of Cincinnati, Division of Air Pollution Control
       and Heating Inspection,  1968.

       An annual summary of activities of Cincinnati air pollution agency program
       to provide service to neighboring communities on a contract basis .  Brief
       description of specific odor abatement cases .  Eight pages.

-------
MacKenzie, V. G.  "Future Outlooks for the Development and Application of Air
       Quality Standards, " Journal of the Air Pollution Control Association, XIV
       (January,  1964), 19-22.

       A discussion of the types of information needed for air quality standard
       guides: effects of pollutants on mortality, acute illness, production of
       chronic diseases,  impairment of important physiological functions, the
       production of sensory irritations (including odors), damage to vegetation
       and property,  and  community dissatisfaction.
Magill, Paul I. (ed.)  Air Pollution Handbook . New York:  McGraw-Hill Book
       Company, Inc., 1956.

       A single volume introduction to air pollution, its causes, effects, and
       methods of control.
Manatee County,  Florida.  Department of Health.  Manatee County Air and Water
       Pollution Control Code.

       A basic code covering responsibilities of Health Department in air
       pollution.
Odors, the Problem, its Definition and Control.  Proceedings of the Mid-Atlantic
       States Section Semi-Annual Technical Conference, Wilmington, Delaware,
       November 12, 1965. Pittsburgh,  Pennsylvania: Air Pollution Control
       Association,  1965.

       A compilation of papers dealing with odor sources,  evaluation, and
       technical and legislative control.  80 pages.
Oregon State Sanitary Authority.  Hot Mix Asphalt Plants.  Information Bulletin
       No. 3, 1968.

       Contains emission regulations for hot mix asphalt plants .  Criteria are
       based on production and location.  No emission standards are set for
       odors .  Two pages .

-------
Roach, Michael D. and Arnett, David E.  Computer Application in Air Resources
       Management.  Portland, Oregon:  City of Portland, Bureau of Health, no date,

       Use of computers is described for data storage and statistical analyses of
       various types of air pollution data compiled within the agency.  19 pages .
Sableski, J. J.  "The Federal Air Pollution Control Program as it Relates to the
       Kraft Pulping Industry," Journal of the Technical Association of Pulp and
       Paper Industries, L (August,  1967), 35A-39A.

       The federal air pollution program as it relates to the kraft pulping industry,
       with particular emphasis on the odorous character of the  industry's atmos-
       pheric emissions,  is described in terms of four individual federal activities:
       (1) research, (2) grants-in-aid, (3) technical assistance,  and (4) abatement
       actions .  (Author's Abs .)
Stanley, William J. and Cranshaw, David D.  "The Use of a Computer-Based Total
       Management Information System to Support an Air Resource Management
     .  Program, " Journal of the Air Pollution Control Association, XVIII, No. 3
       (March, 1968), 158-159.

       The role of computer technology in air resource management is discussed,
       including development of an early warning system for the prediction of
       hazardous levels of pollution.
State of New Jersey, Department of Health.  Air Pollution Control Program.  New
       Jersey Air Pollution Control Code.

       Contains detailed regulations, with separate chapters covering major
       sources of pollution.  Odors are handled under general provision banning
       air pollution.
           .  New Jersey Air Pollution Control Laws - General Provisions of 1954
       Act as Amended.

       Presents the state law establishing the powers of the State Department of
       Health relative to air pollution.

-------
Stern,  A. C. (ed.)  Air Pollution.  3 vols .  2nded.  New York: Academic Press,
       1968.

       The most comprehensive treatment of all phases of air pollution available.
       Sections of most interest relative to odor problems are Part HI, Air Pollu-
       tion Meteorology, and Part V, Air Quality and Meteorological Measurements ,
              'Emission Standards in Air Pollution, " American Journal of Public
       Health,  LV (July, 1965),  1075-1081.

       A discussion of emission standards to be employed in dealing with air
       pollution is presented. It is urged that standards be developed for real
       situations which relate both single and multiple sources to community's
       air quality standards (gas and fine particles).
      	.  "Implications of the Air Quality Act of 1967, " Transcript of the New
       York Academy of Science, XXX (April, 1968), 759-765.
       The Air Quality Act of 1967 is assessed. It incorporates the body of
       federal air pollution legislation that had developed from 1955 through
       1966, setting a firm basis for the outgoing programs of research,
       training, technical, and financial assistances, and interstate and motor
       vehicle pollution abatement.  It establishes a  Presidential Air Quality
       Advisory Board.  The implications of the requirements and limitations
       of the Act are reviewed.  (Chem. Abs.)
      	.  "Summary of the Conference on Odor Control, " Journal of the Air
       Pollution Control Association,  VII, No. 1 (May, 1957).
       A discussion of the nature of odor problems in general, and of conditions
       complicating the problem in Cincinnati.  A four -point program for odor
       abatement in the area is proposed which emphasizes political and finan-
       cial aspects of the abatement process.  Two pages .
Strauss, W.  "The Development of a Condenser for Odor Control from Dry Rendering
       Plants, " Journal of the Air Pollution Control Association, XIV (October,  1964),
       424-427.

       Contains a brief comment on odor from the rendering industry and a dis-
       cussion of the local control requirements around the  Los Angeles  area.

-------
Summer, W.  Methods of Air Deodorization.  Amsterdam:  Elsevier,  1963.

       A comprehensive introduction to odors, including theories of olfaction,
       properties of odorants, meteorology, and detailed discussions  of each
       of the common methods of deodorization.
Teller, A. J.  "Odor Abatement in the Rendering and Allied Industries, " Journal of
       the Air Pollution Control Association, XIII (April, 1963),  148-150.

       Discusses problems created by odor and profitability of odor abatement.
       Describes some of the sources of air pollution in rendering and allied
       industries .  Provides  suggestions as to the odor control methods.
Thring, Meredith W.  Air Pollution.  London:  Butterworth Scientific Publication,
       1957.

       Treats air pollution in general, including various sources of air pollution,
       measurement, effects on health, and geographical factors.  Discusses
       elimination of pollution originating from sources such as domestic heating,
       motor vehicle exhausts, and cement manufacture.  Noxious odors are dis-
       cussed in connection with processes such as food preparation, tanning,
       and fish meal preparation.
Turk, Amos.  "Odor Control, " Kirk-Othmer Encyclopedia of Chemical Technology,
       2nded.,  XIV (1967), 170-178.

       A review of community air pollution problems and methods of odor control
       includes:  ventilation or dispersal, adsorption, air oxidations, conversions,
       washing and scrubbing,  irradiation, use of quaternary ammonium  compounds,
       and counteracting and masking agents.
U.S. Congress. Air Quality Act of 1967.  Public Law 90-148, S780, 1967.

       Amendment to Clean Air Act authorizing, among other things, establish-
       ment of air quality standards.  23 pages.

-------
U.S. Public Health Service.  Compilation of Selected Air Pollution Emission Control
       Regulations and Ordinances.  Rev.  1968.  PHS Publ. 999-AP-43.

       A presentation of air pollution ordinances including odor emission control
       provisions from Chicago, Cleveland, and St. Louis air pollution codes.
       146 pages.
           .  Digest of State Air Pollution Laws, 1966. PHS Publ. 711.
       A compilation of all state air pollution laws is presented verbatim.
       556 pages.
      	.  Kansas City, Kansas - Kansas City, Missouri Air Pollution Abate-
       ment Activity.  Phase II - Pre-Conference Investigations,  1968.
       A summary of local air pollution abatement activities, in terms of ambient
       air quality, climatological studies, and pollutant emissions.  No special
       attention given to odors .  180 pages .
      	.  Progress in the Prevention and Control of Air Pollution. (First Report
       of the Secretary of Health, Education and Welfare to the U.S . Congress,
       pursuant to Public Law 90-148, The Air Quality Act of 1967), 1968.
       A summary of progress in air pollution control in terms of administration,
       abatement, and research.  85 pages .
      	.  Status Report for the Kansas City Metropolitan Interstate Air Quality
       Control Region,  1969.
       An outline of the organization, staffing, budget allocations, and accom-
       plishments of air pollution agencies in the Kansas City region.  Includes
       data on ambient air quality,  emissions, and meteorology.  55 pages .

-------
U.S. Public Health Service.  National Air Pollution Control Administration.  1968.
       Report for Consultation on the San Francisco Bay Area Air Quality Control
       Program, 1968.

       One of the whole series enhancing Baltimore, Pittsburgh, Cleveland,
       Cincinnati, Buffalo, Minneapolis-St. Paul, etc.  The literature intro-
       duces the Air Quality Act, and discusses the  size of the regions chosen,
       various technical, sociological, geographical, and meteorological fac-
       tors, and proposes the region under  consideration as an Air Quality
       Control region.
             State of Pennsylvania and Philadelphia Odor Complaints, 1969.
       A tabulation of odor complaints in the Philadelphia area for 1967-68.
       Complaints are described in terms of postal zone, odor type, and
       name of source.

-------
H.     GENERAL INFORMATION ON THE ODOR PROBLEM
City of Philadelphia, Pennsylvania. Department of Public Health.  The Air You
       Breathe - Air Pollution and its Control in Philadelphia, no date.

       An introduction to causes and the nature of air pollution, and a description
       of Philadelphia's program to control air pollution.
Faith, W. L.  "Air Pollution Research - Reflection and Projection, " Journal of the
       Air Pollution Control Association, XIV (September, 1964), 367-370.

       Contains comments on the odor problem and the various angles from
       which it can be attacked.
Gruber, Charles A. Air Pollution Control in Cincinnati .  Cincinnati, Ohio:  City
       of Cincinnati, Division of Air Pollution Control and Heating Inspection,  1968,

       Public information booklet describes sources of local air pollution and
       activities of agency in controlling air pollution.  12 pages .
Hillsborough County, Florida.  Department of Health. Division of Environmental
       Engineering.  Air Pollution:  The Hidden Danger, no date.

       Public information booklet introduces concepts of oxygen and carbon dioxide
       balances and the potential effects of energy generation, pesticides, etc .
       Six pages.
           .  The Facts About Air Pollution,  1969.
       Public information booklet outlining the causes of air pollution, and the
       effects of major pollutants .  Eight pages.
Lewis, Howard R.  With Every Breath You Take.  New York:  Crown Publishers,
       Inc., 1965.

       A general non-technical discussion of air pollution.  Fairly extensive treat-
       ment of the odor problem.  Different types of odors and odor sources, effects
       of odor on local population, and types of odor-controlling methods are dis-
       cussed .  Includes a brief bibliography on the odor problem.

-------
Middleton, John T. The Air Quality Act of 1967 - A Challenge to the Scientist.
       U.S. Public Health Service, 1968. (Presented to The Chemical Society of
       Washington, Washington, D.C., January 11, 1968).

       The government laboratories working jointly with the academic community
       and private industry under the Air Quality Act of 1967 will make possible
       a greatly accelerated program of research in all our areas of ignorance.
U.S. Public Health Service.  National Air Pollution Control Administration. Air
       Pollution Publications, Selected Bibliography with Abstracts 1966-1968.
       PHSPubl. 979.

       Contains abstracts of approximately 900 federally sponsored investigations
       of air pollution-related topics . 144 pages.
U.S. Public Health Service.  Sources of Air Pollution and Their Control, 1966.
       PHSPubl. 1548.

       Public information booklet discusses major sources on common pollutants,
       and briefly lists means of abatement.  16 pages .

-------