COMPARISON  OF AMBIENT  AIR MEASUREMENT  AND
SOURCE MEASUREMENT

Herbert C.  McKee

Southwest Research Institute
Houston, Texas

April  1971
         NATIONAL TECHNICAL INFORMATION SERVICE
                                                Distributed ,., 'to foster, serve
                                                   and promote the nation's
                                                      economic development
                                                         and technological
                                                            advancement.'
                                                U.S. DEPARTMENT OF COMMERCE

-------
 EPA/OAP  SSPCP  LIBRARY
  411 W CHAPEL HILL ST
 DURHAM  N

UNITED STATES
GOVERNMENT
PROPERTY
Thi3      he property of
the United Sl-'tos Government
National1 Air Foliation Control Administration

-------

                                           PB  205  935
        COMPARISON OF AMB1KN I  AIR MKASUKKMF.NT
        AND SOURCE MEASUREMENT
        Contract CPA 70-40
        SwRI Project 21-2811
        Prepared for:

        Office ol Me.isureinenl S(aiidai(li/a(i<»ii
        Division of Cheinisliy and Physics
        Air Pollution Control Olfiee
        Knvironniental Protection Agency
        April, 1971
Sw

           NATIONAL'T'ECHNICAL
          INFORMATION SERVICE

SOUTHWEST  RESEARCH  INSTITUTE
SAN  ANTON IO                         HOU', ION

-------
BIBLIOGRAPHIC DATA
SHEET
                    1. Report No.
                      APTD-0902
3. Recipient's Accession No.
4. Title and Subtitle

     Comparison  of  Ambient Air Measurement and  Source Measurement
                                                                     5' Report Date
                                                                       April 1971
                                                                     6.
7. Author(s)
     Herbert C. McKee
                                                                    &• Performing Organization Kept.
                                                                       No.
9. Performing Organization Name and Address

     Southwest Research Institute
     3600 Yoakum
     Houston, Texas   77006
                                                                     10. Project/Task/Work Unit No.
                                                                       SwRI  Project 21-2811
                                                                     11. Contract /GfeS* No.

                                                                          CPA 70-40
1 2. Sponsoring Organization Name and Address
     Office of Measurement Standardization
     Division of  Chemistry and  Physics
     Air Pollution Control Office
     Environmental Protection Agency
                                                                     13. Type of Report & Period
                                                                       Covered
                                                                     14.
15. Supplementary Notes  DISCLAIMER;   This report was  furnished to  the  Environmental  Protection
     Agency, Air Pollution Control Office, in  fulfillment of  Contract No. CPA  70-40.
16. Abstracts
Many differences  exist between methods used  for ambient air  measurement  and source
measurement,  related to the  differences in concentration,  temperature and humidity of
the sample streams, interferences that affect analytical results, and other factors. The
differences  in the two types of methods have been substantial,  therefore results obtainejd
with these different methods likely are not  equivalent. Emissions cannot be related ac-
curately to  ambient air measurements, which  makes the job  of planning and developing
standards much more difficult. Emission standards have little meaning unless a method
of measurement is specified, and specific methods or general guidelines  have usually
been published with such standards. The need for a standardization program  is outlined.
17. Key Words and Document Analysis. 17o. Descriptors
 Air pollution
 Gas analysis
 Particles
 Measurement
 Sources
 Exhaust emissions
 Atmospheric  composition
 Temperature
 Humidity
 Standards
17b. Identifiers/Open-End ed Terms
17c. COSATI Field/Group
                           14/Q2
18. Availability Statement

    Unlimited
                                                         19.. Security Class (This
                                                            Report)
                                                              UNCLASSIFIED
                                                         20. Security Class (This

                                                             UNCLASSIFIED
           21. No. of Pages
              23
           22. Price
FORM NTIS-38 ('10-70)
                                                                               USCOMM-OC 40329-P71

-------
COMPARISON OF AMBIENT AIR MEASUREMENT
AND SOURCE MEASUREMENT
Contract CPA 70-40
SwRI Project 21  2811
Prepared for:

Office of Measurement Standardization
Division of Chemistry and Physics
Air Pollution Control Office
Environmental Protection Agency
By:

Herbert C. McKee


April, 1971

APPROVED BY:
 Herbert C. McKee
 Assistant Director
 Department of Chemistry
 and Chemical Engineering

-------
                  SUMMARY AND CONCLUSIONS
        This  report was prepared to aid in planning a program to stan-
dardize methods of source measurement used in air pollution control.
Standardization of methods  used for ambient air measurement has been
in progress for some time, and new requirements established by recent
federal legislation make it necessary to provide standard methods for
source measurement as well.

       Many differences exist between methods used for ambient air
measurement and source measurement,  related to the differences in
concentration, temperature and humidity of the sample streams, inter-
ferences that affect analytical results, and other factors.  Because of
these factors,  methods  used in the past for  ambient air measurement
have been substantially  different from methods used for source measure-
ment,  and therefore results obtained with these different methods likely
are not equivalent.  This in turn means that various emissions as mea-
sured  by the available methods cannot be related accurately to ambient
air measurements, which makes the job of planning and developing
standards much more difficult.

       Emission standards have been developed in the past by a number
of air  pollution control  agencies,  and will be required in the future as
a part of all implementation plans.  The method used most frequently in
the past has been the Ringelmann chart for controlling emission of black
smoke.  Other emission standards have also been used to  control visible
emissions based on equivalent opacity, particulates  measured on a weight
basis,  and sulfur dioxide as measured by a  chemical procedure. A few
agencies have developed unique emission standards to control  specific
problems in their jurisdictions; examples include fluoride, beryllium,
hydrocarbons, and others.

       Most agencies have  recognized the fact that emission standards
have little meaning unless a method of measurement is specified, and
specific methods or general guidelines have usually been published with
the standards.  Various scientific and engineering societies have also
made some effort at standardization of source measurement methods.
None of these efforts have included either a detailed laboratory evaluation
of available methods or  a collaborative test to establish statistical limits
of accuracy and precision.

       A study of existing regulations emphasizes the need for an extensive
program of standardization of source measurement methods.   Available
                                11

-------
methods should be compared to select the variations best suited for
different source measurement applications.   Collaborative testing is
also needed, but will be difficult because of practical limitations on
the number of participants and number of samples that can be accom-
modated by the sampling facilities that can be provided on any reason-
able basis.  Further development of new methods is also  needed, so
that adequate emission standards can be established for many different
types of stationary sources.  These methods  should include both refer-
ence methods and techniques suitable for continuous monitoring and
process control.

-------
                   TABLE OF CONTENTS


                                                           Page

 I.    PURPOSE OF MEASUREMENTS                         1

 II.    DIFFERENCES IN REQUIREMENTS                      2

III.    EMISSION STANDARDS AND MEASUREMENT             4
       METHODS

IV.    PRESENT USE OF STANDARD METHODS                10

 V.    CURRENT EFFORTS AT STANDARDIZATION             11

VI.    FUTURE NEEDS                                       13

       BIBLIOGRAPHY                                       17
                             IV

-------
                I.  PURPOSE OF MEASUREMENTS
       In order to understand the differences between methods used for
ambient air measurement and those used for source measurement, it is
important to examine the reasons for making any measurement related
to air quality.  A most important distinction here is the difference in
legal requirements.  Ambient air measurements are used to evaluate
overall air quality within a  community,  usually over relatively long
periods of time.  These measurements  are useful for planning purposes
to set goals, to document progress in reaching those goals, and to identify
areas of a community in which additional control measures may be needed.
On a shorter time basis,  they may also be useful in indicating when cer-
tain  emergency measures need to be instituted,  although most com-
munities have not yet developed or implemented comprehensive emergency
plans,  and the  difficulties involved in such implementation are quite
formidable.  However, one important point here is  that no immediate
legal action against an individual  source of pollution is likely  to occur,
based on ambient air measurements  alone.

       Source  measurements present a different picture.  Stack sam-
pling,  in particular,  is performed to measure emission rates over a
very short period of time, to determine whether or not that particular
source is exceeding some legally established emission rate.   With con-
tinuous instruments,  stack  monitoring  in some cases is possible on a
minute-by-minute basis.  In other cases,  limitations on available methods
of sampling may make it necessary to take several hours to complete a
single  sample,  but the objective  is still to determine emission rates
over as short a period of time  as possible.  The results also provide
an immediate indication of legal  or illegal operation of the source in
question,  and may lead to demands for changes in the operation of the
source or, in the case of disagreement, to legal action.

       These differences in legal requirements place a greater require-
ment on source sampling methods for short-term or single measurement
accuracy and reliability.  Errors in  a single measurement of ambient
air quality are  of only minor importance when a large number of mea-
surements are  averaged, unless a single measurement is high enough
to initiate some emergency action.  A consistent error in long-term
measurements  means primarily that planning may be done less efficiently
than desired.  While  this is  regrettable, the errors inherent in the pro-
cess of developing ambient  air standards are likely to be greater than
the errors of measurement. However,  with source sampling methods,
any controversy over the error of even  a single measurement may lead
to legal action  under  some circumstances.

-------
              II.  DIFFERENCES IN REQUIREMENTS
       In addition to the different legal requirements for measurements,
many other differences exist between ambient air measurements and
source measurements.  Obviously,  there are exceptions,  but a considera-
tion of the general differences is helpful in understanding  some of the
problems involved in developing methods for source measurement.

       The most obvious difference usually is one of concentration. As
a rule, the concentration of any given pollutant in a stack  or  duct in an
industrial plant can usually be several orders of magnitude higher than
levels that are tolerable in a community atmosphere.  For example,
ambient air measurements of sulfur  dioxide must be made in the range
of 0. 01 to perhaps 0. 5 ppm; measurement in a stack carrying tail gases
from a sulfuric acid plant may frequently show several hundred parts
per million,  with  legally acceptable concentrations in some  areas running
as high as  2000 ppm.

       Temperature and humidity variations also may be  greater within
a stack or  process stream than in the ambient atmosphere.   Excluding
only infrequent periods in cold climates,  ambient air temperatures
normally will be in the range  of 20 to 100° F.   Stack temperatures,  on
the other hand,  may range all the way from below  ambient to 1500° F  or
above.  Humidity  can vary correspondingly  since the absolute humidity
required for saturation is temperature-dependent.

       Interferences in  analysis may also change drastically in going
from a source sampling situation to an ambient air problem.  Again, to
use sulfur  dioxide as an example, interference with a chemical method
of measurement in a stack may occur due to the presence of several
hundred parts per million of nitric oxide.  In the atmosphere, however,
much of the NO will have been converted to  NOg, while a portion of the
SOs will disappear by  oxidation to SQa or reaction with alkaline particu-
lates.  Thus,  the  absolute quantities of the NO and other interfering
substances change drastically as well as the ratios of these  various
substances to the  sulfur dioxide being measured.  The formation of
secondary pollutants in an ambient atmosphere by  photochemical reaction
and in other ways  also introduces other intereferences not originally
present in the process gas measured at the  source.

       For these  various  reasons, measurements at the source and
measurements in  the ambient atmosphere may not be related to each
other. In other words,  a measurement which indicates a pound of SOg

-------
in a stack may not correspond to an ambient air measurement indicating
a pound of SQs distributed in a certain volume of the atmosphere.  Like-
wise, a source measurement indicating a pound of SOa  in the gases from
a sulfuric acid plant may be subject to entirely different errors than a
measurement indicating a pound of SOa  in the flue gas from a coal-fired
boiler.

       One  suggestion for reducing the impact of these variations  in
interfering substances is to  use  the same method for stack sampling
that is used for ambient air  measurements,  and dilute the stack gas
sufficiently to reach the same concentration range.  This would allow
the use of continuous monitoring instruments and other techniques of
measurement commonly used in ambient air sampling for  stack measure-
ments.  While this suggestion has some merit, differences in interfering
substances and the difficulties involved in maintaining an accurate dilu-
tion ratio make this something less than a cure-all for all of the problems
which exist.

       The  differences in working conditions for sampling  also increase
the cost of making measurements and exert a psychological effect  on
persons involved in such work.  Ambient air measurements are usually
made at or near ground level.  By bringing  air into  a building through
a suitable air inlet, an investigator can work in air-conditioned comfort.
In most stack sampling, however, the  actual work of sample collection
must be conducted by  manual methods  immediately adjacent to a duct
or stack carrying the  gases  to be emitted.   The only suitable sampling
points may be located on a stack 100 feet or more above ground level.
Construction of access ports, sampling platforms, and other mechanical
facilities may require major construction and alteration of facilities,
and may only be possible when the entire unit is shut down  for routine
maintenance.   During sampling, much lost time and personal incon-
venience occurs due to the necessity for climbing up and down ladders,
hauling equipment up and down with a block  and tackle, running extension
cords for several hundred feet to obtain electrical power,  and other
similar tasks.   This means  that workers must be more highly motivated
in order to obtain high quality data, and the cost per sample will usually
be much higher than the cost of ambient air sampling and analysis.  The
only partial solution for these problems in the near future is the develop-
ment of automatic instruments for stack monitoring, which can reduce
the amount of work which must be done by manual methods but cannot
eliminate such work.

-------
   III.  EMISSION STANDARDS AND MEASUREMENT METHODS
       In order to understand the problems in evaluating source sampling
methods, it is helpful to examine the regulations which are now in use
and the methods of measurement which are specified.   These two must be
considered together, since no emission standard has much meaning unless
the method of measurement is specified.  To do this, a number of air
pollution control ordinances and regulations were examined to identify
the emission standards and methods of measurement that are now in use.
"While not complete, this brief survey gives a general summary of the
methods now being used throughout the country.  These are divided  into
two categories: (1) those that are in general use in many jurisdictions
and (2) those in use in only a few jurisdictions because  of special problems
or as a result of pioneering efforts by the agencies involved.

       The most general emission standard is a limit on black smoke
based on evaluation by the well known Ringelmann chart.   This standard
appears  in practically all air pollution control ordinances  and  regulations
that deal with emission standards of any type.  For many years, emissions
were  limited to No. 2 Ringelmann in most jurisdictions, but recently
there has been a tendency to impose a No.  1 limitation.

       Perhaps the second  most used method is the  regulation of visible
emissions other than black smoke by the concept of equivalent opacity.
There are valid scientific objections to  this concept, and margins for
error in measurement are greater than desired; nevertheless, the con-
cept is a valid one and can be used  satisfactorily if inspectors are trained
properly.  It is also legally acceptable, having been upheld in  courts of
law up to and including the U.  S. Supreme Court.

       Another type of emission standard which appears frequently  is
the imposition of emission limits based on the weight of material emitted.
A major  limitation here  is that the adverse  effects of an emission may
not be directly related to weight because of differences  in particle size,
corrosivity, or other characteristics; nevertheless, weight is a  conve-
nient  unit of measure that has been widely used.  Regulations of  this
type appear to  have been developed initially to limit  emissions of fly
ash and other  particulate matter from coal-burning equipment, and  more
recently weight limitations  have been applied to  incinerators and other
combustion sources as well as process  units not involving combustion.
Weight limits  applicable to  combustion sources are expressed in many
different ways  such as grains per cubic foot, pounds per hour, pounds
per thousand pounds of air, etc.  Units based on volume of gas discharged

-------
are usually corrected to some standard condition such as  12% COg to
compensate for changes in the amount of dilution air which may be present;
this  correction also avoids circumvention through the deliberate addition
of dilution air.  The ASME Power Test Codes provide the basic standard
methods that have been widely used in measuring the particulate content
of stack gases. Various modifications have been made, especially to
collect condensable vapors as in the PHS Method for incinerator testing.
Obviously, these different methods will give substantially different results
depending on the temperature of filtration to collect samples,  and on
other variations such as whether or not condensable vapors are also
collected and measured.

        With some variations in mechanical equipment and procedures,
all of these methods are similar in nature.  First, a velocity traverse
is made with a pitot tube to determine velocity distribution across the
duct or stack.   Isokinetic sampling is  then performed,  by controlling
the sampling rate so that the linear velocity of the gas  stream entering
the sampling probe is equal to the linear velocity of the stack gases  at
that point; this avoids centrifugal effects which would result in samples
not representative of the flue gas stream.   Particles are collected with
an alundum thimble,  cyclone collector, filter,  or some combination
of these.  Water is condensed and measured,  and other condensable
vapors may also be collected for measurement.  Simultaneous  velocity
measurement and sample collection is also possible, using a probe
which  is equipped with both a pitot tube and a sampling inlet.  After
collection, samples are dried and weighed, and further laboratory work
may be conducted to determine particle size distribution,  chemical
composition, or other necessary data.

        When attempts were made to apply  some of these emission stan-
dards  to process emissions,  difficulties arose because of great variations
in the  volume of gas emitted along with the  particulate matter.  In some
jurisdictions, units of measurement such as weight per unit volume were
retained despite these difficulties, but in others the so-called process
weight principle has been used.  This  concept was originally developed
in Los Angeles  County to apply to the foundry industry.  It allows for
consideration of the fact that a small foundry cannot do as good a job of
controlling emissions, on a percentage basis, as  a large foundry.  Since
its original adoption, the same table with the same numbers has been
copied in many other jurisdictions and applied indiscriminately to any
and all types of processes.  While a detailed study of this matter appar-
ently has never been made,  it would appear likely that the use of the
same table for different types of industrial operations would impose
much more severe restrictions on some industries than on others.

-------
       Sulfur compounds have been regulated in a number of jurisdictions
and the present regulations seem to have been based in part on a per-
centage reduction believed necessary to achieve certain ambient air
quality goals, and in part on limitations of the available technology for
process applications.  An example of the first type is the limitation of
sulfur in fuel, which serves to reduce emissions of sulfur dioxide though
these limits do not constitute "emission standards" in the usual way.
The measurement of sulfur in coal and fuel oil has been well known with
adequate standard methods  for a number of years, and  no significant
problems of measurement have arisen.  For process applications, how-
ever, most of the emission standards now  in use depend on either a
maximum concentration in the stack as measured by stack sampling,
or on a maximum concentration in the ambient atmosphere surrounding
the plant as measured by conventional ambient air measurement methods.
For many years, stack gas concentrations from process sources have
been limited to 2000 ppm, with a  recent trend in some jurisdictions to
reduce this to 500 ppm.   This concept was originally developed  for appli-
cation to sulfuric acid manufacturing  and has been applied to other process
operations involving sulfur  dioxide.  Most  regulations specify one or
more of the stack sampling methods contained in PHS Publication 999-
AP-13, or their equivalent.   These are essentially acid-base titration
methods.

       Turning now to the more unique emission standards,  special
problems in certain localities have led to the development of new emis-
sion standards by the responsible agencies.  Since cities  in California
led in the development of photochemical smog,  it was only natural that
the control districts in Los Angeles County and the San Francisco Bay
area have led in the development  of source sampling relative to hydro-
carbons.  Los Angeles  County, with some  exceptions, has operated
primarily on the basis of total hydrocarbons, limiting emissions from
storage tanks,  loading docks,  and other facilities handling large quantities
of various hydrocarbon mixtures.  The San Francisco Bay area, on the
other hand,  has followed a slightly different approach which places more
emphasis on photochemical reactivity, and therefore measurement of
individual hydrocarbons is required.  This is not a clear-cut distinction,
as shown by the fact that Los Angeles County's  Rule 66  relating to solvent
vapor is also based on reactivity.  However, this partial difference in
philosophy has led to differences  in methods of measurement as outlined
in the applicable publications of these two agencies.

       The first method used for  source measurements of hydrocarbons
was developed by the Los Angeles County Air Pollution Control  District.
Samples are  taken in evacuated glass bulbs and transferred to the gas

-------
cell of an infrared instrument for measurement of total hydrocarbons by
infrared absorption at a wave length of 3. 4 microns.  Results are cal-
culated as ppm hexane,  since a major concern there is evaporation of
gasoline in storage and handling,  and hexane was picked as being repre-
sentative of the molecular weight range and infrared absorptivity of
gasoline vapor.

       With the adoption of Rule 66 relating to organic solvents and
vapors,  a different method was developed for  use with the greater variety
of organic vapors which had to be measured under that regulation.  Sam-
ples are still taken in evacuated glass bulbs to measure process emissions,
vapors in storage areas, or other source  sampling involving organic
solvents and vapors.   These samples are  then passed through a com-
bustion tube in which all organic matter is converted to COa which is
then measured with a nondispersive infrared analyzer.  While based on
a different principle, this  method also  gives a measure of total organic
materials without regard to  individual organic compounds present.  Other
methods have been used for  special purposes to evaluate classes of com-
pounds or individual compounds by the  Los Angeles County Air Pollution
Control District.  An example of the first type is the use  of the fluorescent
indicator analysis (FIA) to measure olefins, aromatics,  and saturates
in hydrocarbon samples.  Gas chromatography has been used to measure
individual chemical compounds  in various organic mixtures but this has
been done primarily for research purposes or to meet unique problems,
and methods based on gas  chromatography have not been widely adopted
for generalized application in connection with  emission standards.

       As mentioned previously,  the Bay  Area Air Pollution Control
District  has emphasized photochemical reactivity  in Regulation 3 relating
to the control of organic emissions. Therefore,  the emission standards
are stated in terms of reactive  components with special emphasis on
olefins and substituted aromatics.  The determination  of these constitu-
ents in gaseous effluents is made by absorbing the olefins and substituted
aromatics on silica gel, after which they are removed with  an  appro-
priate organic solvent and determined by gas chromatography.   A similar
chromatographic procedure is used for other samples  to determine com-
pliance with the various portions of Regulation 3.  For solvents and other
organic liquids, olefins and substituted aromatics are  determined by a
sulfonation procedure which is a modification  of ASTM Method D-1019-62.
For source sampling,  however,  gas chromatography is the method most
frequently used.  Another  interesting variation is a provision for using
a combustible gas indicator for purposes of conducting preliminary tests.
Organic  gases of less than the allowable limit  of 50 ppm (as hexane)
when measured in this way are assumed to be  in compliance with the

-------
regulation.  Emissions showing more than this amount are suspected of
being in violation and additional tests are conducted by gas chromato-
graphy or other suitable methods,  depending on the type of source in
question.

        A limited number of other process emissions have been covered
by emission regulations where specific problems are known to occur.
For example, Florida sets a limit for fluoride emissions based on the
weight of phosphate (PsOg) processed, in order to limit the emission of
fluorides by fertilizer manufacturing operations.   Texas sets a fluoride
limit which is calculated based on meteorological dispersion equations
starting with maximum downwind levels off the plant site thought to be
acceptable.  Very strict limits on beryllium have been adopted in a
number of  states,  based primarily on a comprehensive  study originally
initiated because of specific problems near beryllium processing opera-
tions in Pennsylvania.  Some  limits have been established on lead and
other metals, based at least in large part on previous work done by
industrial hygienists to protect industrial workers from toxic metals.

       "Fugitive dust" may constitute the most important category of
pollutant that is covered in only a very few jurisdictions.  Most of the
various emission standards are based on measurements which must be
made in a stack or duct prior to  emission to the atmosphere. The first
exception to this principle was the use in California of atmospheric
monitoring instruments around industrial plants to measure  sulfur dioxide,
which was originally introduced as a further check on the stack emission
limit of 2000 ppm which was determined by  stack sampling.  This principle
has been further extended by Pennsylvania and Texas in order to permit
quantitative measurements of fugitive dust,  which is defined as dust which
is not emitted through a stack or duct and therefore cannot be measured
by conventional stack sampling methods.  Examples here might be dust
from metallurgical operations which drifts out the side  of a building, dust
raised by vehicle traffic either on a plant site or on streets and freeways
throughout the  community, dust blown from an open conveyor or  storage
pile, dust arising from construction activities, and many others.  Regula-
tions in these two  states provide for measuring fugitive dust by making
ambient air measurements downwind from the suspected source, using
a conventional  high-volume sampler.  Another sampler located upwind  or
in another  suitable location to measure community background levels is
also used,  with the difference between the two results being  the contribu-
tion of the  source in question to the ambient air levels.   Because of
variations  in background between upwind and downwind locations,  margins
of error in such measurements may be significant and therefore  this
                                 8

-------
method cannot be used to measure emissions which are relatively small.
However, gross sources of emissions sometimes give  short-term readings
of many hundred or several thousand micrograms per cubic meter,  in
which case the  contribution of the source is much greater than the limits
of accuracy of measurement,  and this method  can be used.  This type of
measurement has been very useful  in making easier the control of severe
dust nuisances  resulting from fugitive dust emissions.

-------
          IV.  PRESENT USE OF STANDARD METHODS
       Most of the emission standards outlined previously contain some
specification of a method of measurement or provide for such specification
by an air pollution control agency or other governmental entity.  In many
cases legal regulations will not specify a method but will state that
measurements are to be made by methods acceptable to the commissioner
of health,  the air pollution control officer, or other official.  Frequently,
printed methods are provided either as a part of the regulation or as a
supplement prepared by the agency itself.  Some of these give specific
directions  for conducting source testing, while others  contain background
information and/or literature references, and leave the specific choice
of methods and techniques to the investigator.

       One of the  most complete and widely used of these publications
is the Source Testing Manual published by the Los Angeles County Air
Pollution Control District.  In the San  Francisco Bay area, appendices
are provided for each regulation which outline details of methods of
source measurement.  These publications might not be considered
"standard methods" in the usual sense of the word in that they are more
guidelines  than standards and leave some  discretion to the judgment of
the  person making such measurements.  This may be more desirable in
the  case of source measurements than ambient air measurements, and
such judgment may always be required to  allow the investigator to com-
pensate for the great differences in temperature, flow conditions, humidity,
and other  variables which affect the results.

       Historically,  the methods for source  sampling which have been
quoted most often  in emission standards are  the Power Test Codes
published  by the American Society of Mechanical Engineers.  Some
effort at standardization has also been made  by ASTM, APCA, and other
professional groups.

       Within the  past few months,  the necessity to develop implemen-
tation plans under the provisions of the federal Air Quality Act of 1967
has forced many air pollution control agencies to take  a closer look at
emission standards and at methods  of source measurement.  Much more
activity along this line can be expected on the part of most, if not all,
air  pollution control agencies.  The requirements  of the  1970 amendments
to the federal Clean Air Act have also  focused more attention on the
development of emission standards  and source measurement methods.
                                 10

-------
         V.   CURRENT EFFORTS AT STANDARDIZATION
       In addition to the efforts of various governmental agencies to
develop emission standards and the  concurrent necessity for specifying
methods of source measurement,  attempts at standardization have been
and are being pursued by various  scientific groups.  The longest "track
record"  is that of the American Society of Mechanical Engineers which
published the first version of the well known Power Test Codes in 1927.

       At present,  the most concerted activity is being conducted by the
American Society for Testing and Materials.   Committee D-22 of that
organization is working in the field of standard methods for measure-
ment of the atmosphere,  and  Subcommittee 06 of that committee  is
devoted to methods for source sampling.   The following methods have
been adopted or are under consideration by Committee D-22:

       Particulate  Matter.  One method, D-2928-71 "Sampling Stacks
for Particulate Matter",  has  been published as a tentative standard.
It is similar to the  ASME Power Test Code method except for some
technical details regarding  sample collection.

       Black Smoke.   The subcommittee has developed a method to
evaluate black smoke by the conventional Ringelmann method, and
approval by the  full committee is  expected within a few months.

       Equivalent Opacity.  Work is underway by the subcommittee  to
develop standard methods for equivalent opacity,  following the  regula-
tions in use in many jurisdictions.  Preliminary indications are that
subcommittee members believe that severe limitations on the use of
this concept will be necessary because of possible sources of error.

       Instrumental Methods  for  Particulates.  Subcommittee 06 is
evaluating the need for a  separate method covering instrumental methods
such as the smokescope,  transmissometer,  etc.  Such methods might
cover black smoke, emissions other than black as indicated by equi-
valent opacity, or both.

       Sulfur Compounds.  Methods  are being developed by the sub-
committee for sulfur  dioxide  and  sulfur trioxide from fossil fuel com-
bustion  sources. These likely will be based on previously published
methods such as the so-called Shell method and Monsanto method (PHS
Publication 999-AP-13).
                                11

-------
       General Considerations.  Work is being planned by Subcommittee
06 to develop standard methods applicable to most types of sources, cover-
ing such items as the measurement of velocity and volume by methods
other than pitot tube, possible changes in the pitot tube method specified
in D-2928-71,  and other general considerations  in sample collection.
Additional methods for source measurement of gaseous contaminants
are also under consideration.

       A major activity in the standardization of atmospheric measure-
ment methods is being conducted by the Intersociety Committee,* but
this  organization has not yet begun work on  source sampling methods.
There has been some discussion in the Intersociety Committee of the
need for source sampling methods, and some effort  in that direction
may be made in the near future. However,  it is the opinion of various
persons in the  Intersociety Committee that some of the subcommittees
would have to be restructured to bring in the proper knowledge and
capability to attack the problems of standardization of source measure-
ment methods.  No specific plans of this nature  have yet been developed.
   The Intersociety Committee is a cooperative group consisting of
   representatives of eight scientific and engineering societies:

       Air Pollution Control Association
       American Chemical Society
       American Conference of Governmental Industrial Hygienists
       American Industrial Hygiene Association
       American Public Health Association
       American Society for Testing and Materials
       American Society of Mechanical Engineers
       Association of Official  Analytical Chemists
                                12

-------
                       VI.   FUTURE NEEDS
        The previous discussion has served to point up the obvious:  There
is a great need for a much more concerted effort to develop and standard-
ize methods for source measurement.   To many who have been involved
in the air pollution control field for many years, this realization is long
overdue.  Nevertheless,  this need is now widely recognized,  and a  con-
siderable emphasis in this area can be expected in the immediate future.
Much of this emphasis will come from air pollution control agencies who
are just beginning to realize the implications of the federal legislation
which requires emission standards as portions of overall implementation
plans.   Many industrial pollution control people are now reaching the
same realization as they have interpreted for them the implications of
the same federal legislation by air pollution control agencies and by the
companies' own legal counsel.

        The major difference in the overall status of source sampling
methods as compared to  ambient air methods is the greater need for
development work in the  source measurement area, which follows  from
the greater variety of conditions likely to be  encountered.   These varia-
tions, as mentioned previously, may always  require that a standard
method leave a greater amount of leeway for the individual judgment of
the investigator, and the  degree of standardization probably can never
be as complete as in the  case  of ambient air measurement.  In other
words,  setting up and running a high-volume sampler in Cincinnati
involves essentially the same problems that are encountered in setting
up and operating a high-volume sampler in Dallas.  In source  sampling,
however,  every industry is different,  and even  individual plants within
the same industry will  show a considerable degree of variation in operr-
ating conditions which in  turn influence  the problems encountered in
source measurement.  This may mean that a more rigorous control
over the skill and experience of the investigator may be  required in
source measurements than will be the case in ambient air measurements.
If standard methods can be developed to the point where  practically all
details are  specified, then much of the work can be performed by tech-
nicians  and others with a relatively low level of professional skill.
While this has not yet been accomplished in ambient air  measurements
with any assurance of quality,  the  likelihood of accomplishing this appears
to be even less in considering source measurements.

       The need for collaborative  testing as a part of the process of
standardization  is just as great in  the development of methods for source
measurement. In fact, in view of  the more restrictive legal require-
ments,  it may be even greater.  However, many practical difficulties
                                13

-------
can be anticipated in attempting to conduct multilaboratory collaborative
tests of source sampling methods.  For any method involving particulate
contaminants, the generation of a test atmosphere for purposes of labora-
tory evaluation will be extremely difficult if not impossible.   This means
that collaborative testing in some cases will probably have to be per-
formed by bringing a number of collaborators to the same point so that
all can sample the same source.  For the usual type of stack sampling
operation, however,  space on sampling platforms will be limited so
that only one or two groups can work  at any one time and therefore changes
in operating conditions between different sampling periods may introduce
serious sources of error.   Even if these problems can be resolved to
any reasonable degree,  obtaining test results with a large number of
collaborators and/or a large  number  of samples will be  much more
difficult than is usually the case in the collaborative testing of ambient
air methods.  Obtaining a  sufficient number of samples to insure statis-
tical validity may be very  difficult or impossible.

       In developing source sampling methods, there is a great need to
shorten the time  required  to collect and analyze samples. Since source
sampling  is done to meet immediate needs to a greater extent than
ambient air measurement, the several days'  time lag inherent in the
conventional  particulate sampling methods is a severe limitation on the
use of such methods.  This also raises the question of the availability
and suitability of continuous instruments for stack monitoring purposes.
Several companies are now developing and introducing new stack moni-
toring instruments to measure a variety of contaminants,  and continued
activity in this field can be anticipated.

       As with ambient  air measurements, manual reference methods
can be used as a  means  of checking on the  calibration of continuous
analytical instruments.  Here again,  however,  the comparability of the
reference method and of the measurement obtained by a  continuous
instrument must  be established by detailed development  work before
routine use of a stack monitoring instrument is feasible.   Thus,  in the
development  and  standardization of source sampling methods, a division
might be made similar to the division already being observed in ambient
air sampling methods.   That  is, reference methods should be developed
to provide a standard basis for measurement,  and these in turn should
then be used  as a means of checking the suitability of various stack
monitoring instruments  for continuous process use.  Once this com-
parability has been demonstrated,  stack monitors can then be used for
routine operation and checks  by a reference method can  be limited to a
rather infrequent schedule.
                                14

-------
        The use of stack monitoring instruments raises other questions
in connection with the legal aspects of developing and applying emission
standards.  Nearly all of the standards now in effect require the use of
manual sampling and analytical methods which involve a considerable
time lag in obtaining results.  There is a need for new concepts in the
legal field so that regulations and emission standards  can be written
directly in terms of measurements made with continuous instruments.
One example of this concept has already been developed in the state of
Texas,  where an instrumental method has  been written into  a legal
emission standard as  a substitute for the visual estimation of equivalent
opacity.  This  eliminates problems inherent in visual observation, and
legal compliance with this regulation can be demonstrated only on the
basis of an instrumental measurement without the necessity to demon-
strate  comparability between the instrumental method and a reference
method.  In this particular instance,  the instrumental method is more
accurate and reliable  than the conventional visual method, and therefore
the question of comparability is  not applicable.  However, the further
development of legal concepts which would permit the  use of emission
standards tied  directly to instrumental readings would be of considerable
benefit in eliminating  time lag, uncertainty, and confusion in the opera-
tion of various stationary sources  subject to emission standards.

        One other question which seems to  have been partially overlooked
in the past is the comparability of ambient air measurements and source
measurements. No specific  recommendations can be  made  to overcome
this limitation  except  the obvious one that further work done to develop
methods for both ambient air measurements and source measurements
should include  some consideration of this problem.  In the work which
is now in progress to  evaluate methods for ambient air measurements
prior to collaborative testing and adoption  of  standards, some considera-
tion should be given where possible to the nature of the source sampling
methods which are available for the  same contaminant.  One example
here is the effect of particle  size in measuring total particulates.  The
conventional high-volume method for particulates includes very few
particles larger than 50 microns, whereas the conventional  method for
source  sampling is supposed to obtain a  representative sample of the
material in the  stack irrespective of particle  size.   If a large portion of
the emission consists of large particles, it is obvious that many of these
particles will fall to the ground within a  short distance of the stack, partly
on the plant property.  In this special situation, a stack measurement
could not relate directly to an ambient air measurement in the community
using a high-volume sampler.
                                 15

-------
       Another problem may occur with organic matter which would be
measured as a constituent of particulate matter by some stack sampling
methods or would be  recovered in a condensate trap in sampling.  In
actual fact, at least a portion of such organic matter probably would be
diluted in the  atmosphere following  emission and would remain in the
vapor phase.  Thus,  a stack measurement might reflect the presence
of organic matter but give an erroneous indication of the physical state
of such constituents after emission  to the  atmosphere.  Problems of this
nature affect a. legal definition of particulate matter  as well as the pro-
cedures  used  in sampling with respect to temperature, flow rate, con-
densate traps, etc.  Similar examples  of this lack of comparability
might be shown for sulfur dioxide,  hydrogen chloride,  hydrocarbons,
and many other contaminants subject to control through the use of
emission standards and source sampling methods.
                                 16

-------
                         BIBLIOGRAPHY

Laws, Ordinances, Standards, and Regulations of the following states:
       New Jersey                  Pennsylvania
       Florida                      Oklahoma
       California                   North Dakota
       New York                   Iowa
       Texas                       Delaware
Laws, Ordinances, Standards, and Regulations of the following local
governmental entities:
       Los Angeles County Air Pollution Control District
       Bay Area Air Pollution Control District (San Francisco Bay area)
       Allegheny County,  Pennsylvania
       County of San Bernardino, California
       Metropolitan Dade  County, Florida
       County of Sarasota, Florida
       City of Chicago,  Illinois
       City of Cleveland,  Ohio
       City of Poughkeepsie, N. Y.
       City of New York, N.  Y.
       City of Detroit,  Michigan
       City of St. Louis, Missouri
       City of Cincinnati,  Ohio
       City of Cedar Rapids, Iowa
       St.  Louis County, Missouri
Air Pollution  Control District, Los Angeles County, California.  "Air
       Pollution Source  Testing Manual. " November, 1965.
Air Pollution  Control District, Los Angeles County, California.  "Recom-
       mended Test Methods for Organic Solvents and Vapors (Rule 66)."
       April,  1968.
                                17

-------
"Atmospheric Emissions from Sulfuric Acid Manufacturing Processes."
       U. S. Public Health Service Publication 999-AP-13 (1965).
"Determining Dust Concentration in a Gas Stream."  Power Test Codes
       PTG-27, American Society of Mechanical Engineers.  1957.
1970 Book of ASTM Standards.  Part 23,  Water; Atmospheric Analysis.
       American Society for Testing and Materials,  Philadelphia,  Penna.
                                18

-------