EPA-650/4-75-014
April 1974
Environmental Monitoring Series
          EVALUATION  OF THE PROPOSED
               AMBIENT AIR  MONITORING
  EQUIVALENT AND  REFERENCE METHODS
                            U.S. Environmental Protection fl-
                             Office of Research and Development
                                 Washington, D. C. 20460

-------
                                         EPA-650/4-75-014
     EVALUATION  OF  THE  PROPOSED
         AMBIENT AIR  MONITORING
EQUIVALENT AND REFERENCE METHODS
                          by

            J. J. Wesolowski, E. R. DeVera, Y. Tokiwa,
             W. Wehrmeister, K. Smith, andM. Imada

              Air and Industrial Hygiene Laboratory
                 Laboratory Services Program
              State of California, Department of Health
            2151 Berkeley Way, Berkeley, California 94704
                   Contract No. 68-02-0744
                     ROAP No. 26AAF
                  Program Element No. 1H1327
               EPA Project Officer: Franz Burmann

                    Special Studies Staff
              National Environmental Research Center
            Research Triangle Park, North Carolina 27711
                       Prepared for

            U.S. ENVIRONMENTAL PROTECTION AGENCY
             OFFICE OF RESEARCH AND DEVELOPMENT
                  WASHINGTON, D. C. 20460

                        April 1974

-------
                        EPA REVIEW NOTICE

This report has been reviewed by the National Environmental Research
Center - Research Triangle Park, Office of Research and Development,
EPA,  and approved for publication.  Approval does not signify th^t the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
                    RESEARCH REPORTING SERIES

 Research reports of the Office of Research and Development, U.S. Environ-
 mental Protection Agency, have been grouped into series. These broad
 categories were established to facilitate further development and applica-
 tion of environmental technology.  Elimination of traditional grouping was
 consciously planned to foster technology transfer and maximum interface
 in related fields.  These series are:

           1.  ENVIRONMENTAL HEALTH EFFECTS RESEARCH

           2.  ENVIRONMENTAL PROTECTION TECHNOLOGY

           3.  ECOLOGICAL RESEARCH
           4.  ENVIRONMENTAL MONITORING

           5.  SOCIOECONOMIC ENVIRONMENTAL STUDIES

           6.  SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS

           9.  MISCELLANEOUS

 This report has been assigned to the ENVIRONMENTAL MONITORING
 series.  This series describes  research conducted to develop new or
 improved methods and instrumentation for the identification and quanti-
 fication of environmental pollutants at  the lowest conceivably significant
 concentrations. It also includes studies to determine the ambient concen-
 trations of pollutants in the environment and/or the variance of pollutants
 as a Junction of time or meteorological  factors.
 This document is available to the public for sale through the National
 Technical Information Service, Springfield, Virginia 22161.

                 Publication No. EPA-650/4-75-014
                                  11

-------
                               Table of Contents




                                                                            Page




     Acknowledgement                                                         vii




1.0  Introduction                                                             1




2.0  Purpose and Scope of Contract                                            3




3.0  Test Facility and Equipment                                              3




     3.1  Test Chamber                                                        3




     3.2  Gas Generation System                                               3




     3.3  Sampling Probe and Manifold                                         6




     3.4  Automated Methods or Analyzers                                      7




4.0  Evaluation Procedure                                                     8




     4.1  Evaluation of the Regulations                                       8




     4.2  Review of Respondents'  Comments                                     9




5.0  Results and Discussion                                                  10




     5.1  Preliminary  Review                                                 11




     5.2  Evaluation of  Subpart  B                                           12




     5.3  Evaluation of  Subpart  C                                           34




     5.4  Review of Respondents' Comments                                    36




6.0  Manpower  Requirements and Cos ts                                        37




     6.1  Subpart B                                                         37




     6.2  Subpart C                                                         42




     6.3  Overall Times                                                      44




     6.4  Certification                                                     45




7.0  Summary and Conclusions                                                 46




8.0  References                                                              50
                                      iii

-------
                             List of Tables



                                                                               Page



  I.  Analyzers Used  for  Evaluating  EPA's  Proposed Equivalency Regulations      51







 II.  Questions and Suggestions on  Subpart B of the Regulations  Submitted



      to the Project  Officer                                                   52







 III.  Recommended Changes in Subpart C of the Equivalency Regulations           56







  IV.   Changes Authorized or Requested by EPA                                   60







   V.   Evaluation of Subpart B Using S(>2 Analyzers  (Phase I) .                    &~







  VI.   Evaluation of Subpart B for Zero and Span Drifts Using  the Melpar



       Total Sulfur Analyzer - Phase I.                                         64







 VII.   Evaluation of Subpart B for Zero and Span Drifts Using  the Technicon



       S02 Analyzer - Phase I.                                                  65
                                       (a)
VIII.  Evaluation of Improved Subpart Bv ' Using the Bendix  Ozone Monitor



       (Phase II).                                                              66
  IX.  Suggested Zero and Span Drift Test Plan a:  Table B-4                     71







   X.  Evaluation of Subpart C for Oxidant Methods:  Manual KI vs Bendix



       Analyzer - Phase I.                                                       72
                                   iv

-------
                                                                               Page




  XI.   Evaluation of Subpart C for Sulfur Dioxide Methods:




       p-Rosaniline Manual Method vs  Melpar Analyzer  -  Phase  II                  73









 XII.   Evaluation of Subpart C for Sulfur Dioxide Methods:  Manual




       p-Rosaniline Method vs Technicon Analyzer -  Phase II                     74









XIII.   Times to Conduct Test Procedures in Subpart  B for Performance            75









 XIV.  Times to Conduct Test Procedure in Subpart C for Consistent




       Relationship                                                             7 7
  XV.  Overall Times to Determine Equivalency
                                                                                 78
                                    v

-------
                        LIST OF ILLUSTRATIONS









Figure           	Name	               Page









  1              Gas Generation System                               79









  2              Gas Sampling  Manifold                               80









  3              Dynamic  Calibration  Curve of  the  Technicon




                 Air Monitor IV  for S02                              81









  4              Dynamic  Calibration  Curve for the Melpar




                 Total Sulfur  Analyzer,  Model  LL-1100-1B              82









  5              Calibration Curve for the Melpar  Total




                 Sulfur Analyzer,  Model LL-1100-1B                   83









  6              Dynamic  Calibration  Curve for the Bendix




                 Ozone Monitor                                      84








Appendix I     (Copy ^0 CFR  53)                                       85




Appendix II    (Revised Draft of Subpart  B)                           97




Appendix III   (Revised Draft of Subpart  C)                          127




Appendix IV    (Review  of Respondents'  Comments)                     139
                                vi

-------
                           ACKNOWLEDGEMENT









The authors wish to acknowledge the assistance of Mr.  Franz Burmann




who served as the project officer; Dr. Peter K. Mueller who was the




initial principal investigator; Mr. Gordon Ortman who acted as the




liaison officer between EPA and the Air and Industrial Hygiene




Laboratory; Mr. Seymour Hochheiser, chairman of the committee that




drafted the proposed equivalency regulations, for his assistance




in the clarification and interpretation of some aspects of the pro-




posed equivalency regulations; and Ms. Suzanne Twiss who assisted in




the statistical analysis of the experimental data.
                              vii

-------
1.0  INTRODUCTION









     When the federal Environmental Protection Agency (EPA) promulgated the




     national primary and secondary ambient air quality standards on April




     30, 1971 (40 CFR 50) for six pollutants, the promulgation was accomp-




     anied by prescribed "reference methods" for the measurement of the




     pollutants.  To avoid obsolescence of existing analytical instruments




     and to  encourage development of superior methods, EPA also designated




     that  the pollutants may be measured by "equivalent methods".  An




     "equivalent method" was defined as "any method of sampling and analyzing




     for an  air pollutant which can be demonstrated to the Administrator's




     satisfaction to have consistent relationship  to  the  reference method".




     This  definition was  later  amended in  40  CFR 51  (Aug.  14,  1971)  to




      require that  an  equivalent method must "demonstrate  a consistent  re-




      lationship and/or meet performance  specification with the reference




      method".   This change, therefore,  allows three mechanisms by which an




      analytical method can be established to be equivalent to a reference




      method:  (1)   meet performance specifications, (2)  demonstrate con-




      sistent relationship, or (3) meet both (1) and (2).









      To provide a standardized set of test procedures by which any  ambient air




      analytical method can be established as an equivalent method,  EPA developed




      the "Ambient Air Monitoring Equivalent and Reference Methods"  referred to




      in this report as "equivalency regulations" or simply "regulations".  The




      first draft of the regulations was completed by EPA in February 1972

-------
and reviewed internally.  A revised draft was then submitted in May




1972 to members of the Scientific Apparatus Manufacturers Association,




EPA Regional Offices, selected  state  air pollution agencies including




the Air and Industrial Hygiene  Laboratory  (AIHL)  and other scientists




within EPA for review and comment.  A final draft of the regulations




incorporating the relevant comments was completed by EPA with  the




assistance of AIHL staff in April  1973 and proposed in the October




12,  1973  issue of the Federal Register (40 CFR 53)1.








The  proposed regulations consist of  a preamble and  three subparts.




The  preamble describes  the legal basis, purpose and application  of  the




regulations.   Subpart A (General Provisions)  describes the general  re-




quirements that  must be met,  the information and affidavits that must




be supplied and  the  appeal procedures to  be  followed.   Subpart B (Test




Procedures for Measuring Performance Characteristics)  lists the




performance specifications that automated reference and candidate




methods must meet and  describes the test  procedures to be followed.




Subpart C (Test  Procedures for Demonstrating Consistent Relationship




Between Reference Methods and Candidate Methods) contains the  test




procedures,  specifications and sampling plans to be used.  It  is the




intent of EPA to apply  these  regulations  to  both manual and automated




ambient air monitoring  methods  for oxidants  (Ox), sulfur dioxide (802)




and  carbon monoxide  (CO).








.To assess the adequacy  of the proposed regulations, the Air and  Industrial




Hygiene Laboratory of the California Department of  Health under  contract




with EPA  evaluated the  April  1973  version of  the proposed regulations.
                                - 2  -

-------
    This report describes the  (1) purpose of the contract,  (2) procedures




    used to evaluate  the regulations,  (3) results obtained,  (4) deficiences




    uncovered and  the changes  recommended to improve or  correct them,  and




     (5) approximate cost and man-hours required to  conduct  the various




    test procedures and to  establish equivalency of an analytical  method.




    In addition,  this report also contains  a copy of the proposed  regu-




     lations  (Appendix I), AIHL's recommended drafts of Subparts B  and  C




     (Appendix II  and  III, respectively) and AIHL's  review (Appendix IV)




     of the respondents* comments on the proposed  regulations.









2.0  PURPOSE AND SCOPE OF CONTRACT









     The purpose of the contract is to:  1)   evaluate the test procedures,




     specifications and sampling plans  described in EPA's proposed "Ambient




     Air Monitoring Equivalent and Reference Methods" using three automated




     methods, 2)   maintain  records of  the test results of the evaluation,




     3)  delineate any  problem areas in the test procedures and make recom-




     mendations for their correction and/or clarification,  4) determine the




     approximate cost and total man-hours required  to perform each of  the




     test procedures  and to establish  equivalency,  5) provide an estimate




     of man-hours  necessary to review  the test data for  certification




     or rejection  of  a candidate method tested in accordance with  the  regulations









     In addition,  comments  submitted to EPA by interested parties  regarding  the




     proposed regulations were reviewed.









3.0  TEST FACILITY AND EQUIPMENT
                                     - 3 -

-------
3.1  TEST CHAMBER





     The regulations specify that the temperature of the gas generation


system shall be maintained within ±2°C  to assure stable output [153.3(g),


Table B-4].  To satisfy  this requirement, the  gas generation system shown


in Fig. 1 was installed  in a 13 X 7 X 8 feet temperature and huaidity-

                       9
controlled  test chamber  .  The temperature  in  the chamber  can b* main-


tained within ±0.5°C  from 0  to 45°C and the relative humidity to within


±5%  from 5  to 85%.  An air-conditioned  room capable of meetiag*th«  test


facility  requirement  may be  used as well.   All the  tests for measuring


performance characteristics  were conducted  in  the chanter.





3.2   GAS GENERATION SYSTEM





      The system used to  generate the  test atmospheres needed for the


evaluations is  shown schematically in Fig.  1.   The  test atmospheres


were generated according to  the procedures  specified  in Table"B-2 and


fl53.21 (a)  of the regulations.  The system consisted  of the pollutant


 generators  and four test atmosphere ducts or lines.  Test  atmosphere Z


 (zero air), P (pollutant gas)  and I (interferent gas) were generated


 through the first,  second and third lines.   The atmosphere PP  (pollutant


gas  of different concentration from P)  was  generated through  the fourth


line.





To insure stable  test atmosphere concentrations, borosilicate glass


tubing  (3/8 inch  I.D.) was used  to construct the test lines where necessary,

                                                                       o
Copper tubing and brass  fittings were used  in other areas. Flowmeters


were used to measure  the flow  of the  test atmospheres.  In addition,  a


mass flowmeter* coupled  with a strip-chart  recorder, was installed in


                                - h -

-------
the pollutant test line (P) to monitor the pollutant  gas  flow at  all  times .




The recorder trace provided a continuous record of the pollutant  rlowrate




which was particularly useful during periods of unattended tests.  Dow a




stream of the flowmeters, the test lines were fitted  with three-way




valves so that the gas flow in one line could be. combined with the other




as required.  Kjeldahl-type connecting bulbs installed immediately after




the valves were used as mixing chambers for the gas streams.









Cylinder zero air^ was used to dilate the test gases  as specified in the




regulations.  At a nominal flow of 2 liters per minute, the volume of




zero air contained in a size 1A cylinder  (about 220 cu. ft.) was




sufficient to provide about 48 hours of continuous operation per test




line or only 12 hours when all four lines were in  use.  To permit sustained




operation, the cylinders were manifolded  as shown  in Fig. 1  so that when




one cylinder was exhausted another could  be switched on without  interrupting




the gas flow in any  of  the test lines while the empty  cylinder was replaced




with a full one.  To conserve  cylinder  zero air,  the  test lines were also




connected to a source of  filtered ambient air  during  idle test periods.




The gas generation system was  designed  such that  the  changeover  from one




air source to the other could be  accomplished  conveniently  and rapidly




by turning a valve on each test line.









The test concentrations were verified using the reference methods  given




in Table B-2 and in  accordance with the test directions in  Subpart B.
                               - 5 -

-------
3.3  SAMPLING PROBE AND MANIFOLD









     The test procedures for demonstrating consistent relationship




[fl53.30(b)] provide specific requirements for the collection and




transport of the ambient air test samples.  An ambient air sampling




system which met the requirements was designed and installed at the




California  State Department of  Health's  laboratory facility at 2002




Acton Street in Berkeley.  The  sampling  probe was constructed from




borosilicate glass pipes**, 1.5  inches I.D. by 5  feet  long segments,




joined  together by Teflon  sleeves supported by snap-on plastic shells.




The probe  consisted  of a vertical section that extended by about 17




feet above  the roof  of a one-story  building  and  a horizontal section




which was  suspended  parallel  to the ceiling  of the test room.  A single




90° connector  was  used to  join  the  vertical  and  horizontal sections.




The single  bend minimized  loses of  airborne  particles in the ambient




air samples.   A specially  designed  glass sampling manifold (Fig. 2}




with 4  radial  sampling ports  (1/4 inch  I.D.)  to  provide essentially




identical  samples  to the methods under  test  was  connected at the end




of the  horizontal  section  of  the probe.   About 10  feet upstream of the




sampling ports, a  spiking  port  of 1/4 inch I.D.  glass was installed at




the lower  end  of the vertical section of the probe  for introducing zero




air (for diluting) or pollutant for spiking  (enriching with pollutant) the




ambient  air stream.   Ambient  air was drawn continuously  through the




probe and manifold by  a centrifugal fan  at about 1 CFM.  The air  flow
                               - 6 -

-------
was monitored by a mass flowmeter.  The inlet of the sampling probe was




fitted with a sheet metal low conical hat to exclude rain, and with 100




mesh stainless steel screen to prevent debris and insects from getting




drawn into the air sampling system.









The concentrations of  the pollutant of interest in the air samples were




adjusted  to the desired levels by diluting the air stream with zero air




or spiking with pollutant from a generator.  The pollutant generators  used




were  fabricated according to the systems designated  in Table B-2 of the




 regulations.  The  spiking rates were  adjusted such that  at least 60%




 of the  test atmosphere consisted of ambient  air at all times.









 3.4  AUTOMATED METHODS OR ANALYZERS









      The contract required  that  the  regulations be  evaluated using three




 automated methods.  The analyzers used were the Technicon Air Monitor IV'




 for sulfur dioxide, Melpar total sulfur analyzer^ and Bendix ozone monitor^




 and were furnished by the Quality Assurance and Monitoring Laboratory of




 EPA.   The identity, detection principles and some performance characteristics




 of the three analyzers as  given by the manufacturers are given in  Table I.









 The Technicon analyzer represented one of the older types of instruments




 which uses the wet-chemical colorimetric measurement principle,  generally




 characterized by slower response (tj_= 5 to 30 min.)  and  by sensitivity to




 large number  of interferents.   The Melpar and the Bendix units represented




 the later monitors which are technologically more advanced,  generally

-------
    fast-responding (t^  =  1  sec  to 3 min.),  utilize  detection  systems  that



    directly  measure some  physical properties  of  the pollutants of interest




    and  in general insensitive to interferents.









    All  three analyzers  were used units and all  required extensive maintenance




    and  repair before they could be used in the  evaluation.  Some of  the problems




    encountered with the analyzers are discussed in Section  5.3 of this report.




    Additionally, they were not  necessarily the  latest models  in  their respective




     categories.









4.0  EVALUATION PROCEDURE









     4.1  EVALUATION OF THE REGULATIONS









          Subparts A, B and C of  the April 1973 version of the  proposed




     regulations were first inspected for obvious errors such as misprints,




     inconsistent directions and  statements and inadvertent omissions.  These




     deficiencies were first resolved with the project officer  before the




     actual testing was initiated.









    Subpart A did not contain any procedures that required testing.  Subparts




    B and C were evaluated in two phases.  In Phase  I, the test procedures in




    Subpart B were tested  using  the Technicon and Melpar analyzers for S02-




    Deficiencies discovered during the testing were  corrected  and the  corrected




    test procedures were retested in Phase II  using  the Bendix monitor for 03




    to validate  the changes  made and to further  identify any deficiencies not




    discovered in Phase  I.









    Subpart C consists of  three  separate  but essentially identical test




    procedures for demonstrating consistent relationship  for oxidant, sulfur

-------
dioxide and carbon monoxide methods.   The initial evaluation (Phase I)




of Subpart C was conducted with oxidant methods using the Bendix ozone-




monitor against the manual potassium iodide (KI) method^.  Based on




the results of this test, Subpart C was redrafted into a combined test




procedure applicable for all three (oxidant, sulfur dioxide and carbon




monoxide) methods.  The validity of the combined test procedure for




Subpart C was then verified in Phase II using the Technicon and Melpar




analyzers as candidate methods against the manual p-rosaniline as the




reference method.









During the  testing process, a record of  the approximate  costs and man-




hours required to conduct each of  the  test procedures was maintained




to  obtain an estimate of the total cost  and man-hours required to




demonstrate equivalency.









4.2  REVIEW OF RESPONDENTS' COMMENTS









      The regulations  as  proposed by  EPA in the Federal  Register  on




October 12, 1973 (40  CFR 53)  included an invitation for comments  from




interested  parties.   This  resulted in the receipt  of 29 communications




by  EPA; 16  from  governmental  agencies, 10 from industry representatives




and 3  from  private  individuals.









Although not specified  in the  original contract, at the request  of  the




project officer,  these  communications  were reviewed by  AIHL.   The review




was initiated by randomly  assigning  each respondent's communication a
                                -  9  -

-------
    code number  from 1  to 29  for convenience in  referencing.  The communications




    were then  individually  reviewed for  the comments  pertinent to the pro-




    posed  regulations.   In  most cases  the  comments  were  paraphrased, when




    possible,  into one  or two sentences  or into  a simple paragraph, and




    then categorized by subject matter,  tabulated and indentified with the




    part,  paragraph,  section or item in the  regulations.  The categorized




     comments were reviewed  and recommendations were prepared based on




    AIHL's expertise  and familiarity with the regulations.









5.0  RESULTS AND DISCUSSION









     As discussed earlier,  the proposed regulations stipulates that the equiv-




     alency of a candidate method may be determined by one of three mechanisms:




     (1)  meeting the performance specifications  (Subpart B), (2)  demonstrating




     consistent relationship  (Subpart C) or (3)  meeting  both (1) and  (2).




     Subpart B, further, specifies  that the tests for each performance parameter




     is to be conducted seven times.  When all seven test results for  each




     parameter meet the specification given in the regulations,  the automated




     method or instrument is passed for that parameter.  When the instrument




     fails three or more of the seven trials, the instrument is  rejected  for




     that parameter.  When one or two of the test results do not meet  the




     specification, then a second series of eight tests for that parameter




     is to be conducted and the same pass-fail criterion is applied to the




     combined total of 15 test results.  Except for the number of  repeat  deter-




    minations , similar criterion  for the interpretation of Subpart C  data  for




    consistent relationship is followed.
                                    -  10 -

-------
In Subpart B however, while the procedures  for zero  and span  drills,




lag time, rise time, fall time and precision [53.22(f)] stipulate that




the test for each parameter is to be conducted once  a day over seve^;




days, the number of trials to be performed  is not stated in the re-




maining procedures for noise, lower detectable limit (LDL)  and inter-




ference equivalent (IE).  In the evaluation process, therefore, this




omission was interpreted to mean that the general provisions  applied




to those procedures in which the number of  trials were specified and




only one trial was conducted for the other  procedures  (noise, LDL and




IE).  The values reported for these procedures in Tables V and VIII are




based on the results of the single trials.   The deviation might affect




the  instrumental results obtained, but we do not believe it compromised




the  overall objective of evaluating  the procedures  for these three




parameters.









5.1   PRELIMINARY REVIEW









      The initial inspection of Subparts  A.,  B and C  of  the  April  1973




draft of the  regulations  uncovered a large  number of typographical




errors,  omissions,  unclear directions  and  inconsistent statements.




These findings  gave rise to several questions regarding the  intent




and  application of  the  procedures.  Table  II lists  the questions  that




pertain  to  Subpart  B as well as the suggestions  for correcting them.




Table III  delineates the deficiences found in Subpart  C and  the  recommended




changes.   In  response to some of the foregoing questions,  suggestions  and




recommendations,  several changes were  authorized or requested  by the project




officer.   These are also noted and discussed in  Table  IV,






                                - 11 -

-------
5.2  EVALUATION OF SUBPART B









     The analyzer performance data obtained in the course of evaluating




  Subpart B are shown in Tables V and VI along with the summary of con-




  clusions drawn and the changes recommended.  The bases for the recommen-




  dations are discussed below:









5.2.1  General Provisions









       a.  U53.20(a):  The first sentence can be misinterpreted to mean




           that the automated methods are for developing observations on




           Interference	etc. rather than the test procedures.  For




           clarity it should read "Test procedures applicable to automated




           candidate and reference methods for observations on Interfer-




           ence	etc."









       b.  The general provisions also do not state when the method satisfies




           the requirements of Subpart B of the regulations.  We therefore




           recommend that the following statement be added after the first




           sentence, "The methods must pass all the specifications for




           performance to satisfy the requirements of the subpart".  More-




           over, the provisions do not indicate the ambient conditions at




           which the tests are to be conducted.  We recommend that "and




           at normal ambient temperature of 20 to 30°C and normal line




           voltage of 110-125 VAC" be added after the word "section" in




           the second sentence.
                                 - 12 -

-------
c.  1153.20(c):  This paragraph states that an initial series  of  seven




    tests is to be performed on all the performance specifications,




    In the individual procedures, however, this requirement is stated




    only in the procedure for drifts, response times and precision




    [53.22(f)(l)] and further emphasized in Table B-4 [53.22(g)].




    Unless it is the intent of EPA to require all tests  to be




    repeated seven  times, we recommend that the second sentence be




    revised as  follows:  "For each performance specification for




    range, noise, lower detectable limit and interference equiva-




    lent, perform the tests as directed in the test procedures,




    The method  or instrument must pass all the above specifications




    before  other tests are begun.  For each performance specification




    for drifts, response times and precision,  an initial series of




    seven (7)  tests shall be  conducted".









 d.  H53.20(d):   Table  B-l:









    1.   Only  one specification is  listed  for noise while  the  test




         procedure  (1F53.22(b) (2)) requires the test be conducted at




         both  0 and  at  80%  of  the upper range limit. The  missing speci-




         fication should  be provided  and included in the table.









    2.   The total  interference equivalent is calculated as the  sum  of




         the absolute values of the individual interference equivalents




         Therefore,  the minus  signs before the specification for total




         interference equivalents do  not apply and should be deleted.
                           - 13  -

-------
3.  Only one specification is listed for span drift.  The span




    drift tests, however, are conducted both at 20 and 80% of the




    upper range limit  [1T53.22(f) (4) ].  The missing value should




    be supplied and included in the  table.









4.  Superscript "(d)"  after "12-hour" for zero drift should be




    changed to "(c)" to be consistent with the footnote.

-------
5.2.2  Gas Generation









       a.  The design of the gas generation system inferred from H53.2.na),




           Table B-2 and fl53.22(d) in the regulations provides for one




           pollutant line.  This permits only one pollutant concentration




           to be generated at a time.  Most of the test procedures,




           however, constantly require two concentrations  (usually




           at 20 and 80% of full scale).  This requires repeated




           resetting of the diluent air flow to obtain the concen-




           trations needed.  This process is time-consuming and makes




           production  of consistent concentrations impossible.  To




           provide a constant supply of each test  gas concentration, we




           recommend that  a second pollutant test  line be  specified  (See




           3.2  of  this report).   The generation  system will  then  require




           4 instead  of the  present  3  test  lines;  one for  zero  air  (Z),




           one  for the interferents  (I)  and two  for  the  pollutant of




            interest (P and PP).   This  change will  assure constant test  gas




            concentrations  and increase reliability as well as efficiency.









        b. H53.21(a):   This  section specifies  that the  "temperature  of




            the  gas generation system shall be  maintained within ±2°C




            to assure  stable output" and that the "test  gas concen-




            tration shall be consistent from day  to day  [(f)(4)(vi)]".




           The  test procedures for zero and span drifts  (Table B-4),




           however,  are conducted both at 20 and 30°C.   This implies




            that two temperature-controlled test  rooms or chambers,  one




            for  the instruments under test and  one for the gas generation
                                  - 15 -

-------
           system, are required.  This question should be clarified.




           Also, when a change in the gas concentration does occur,




           how much can be tolerated?  Can correction factors be used?









       c.  H53.21(b):  The concentrations specified in Table B-2 of the




           regulations for the stock cylinders of  carbon dioxide, carbon




           monoxide, methane  and xylene  correspond to the interferent test




           concentrations listed in Table B-3.   Some tests, however, re-




           quire mixing  the interferent  with  the pollutant gas which




           would reduce  the concentration.  Therefore, the concentrations




           should be increased by  at least a  factor of 2 to allow for the




           dilution [K53.22(d)(2)(iv) (c)] of  the gases for the interference




           tests.









       d.  The reference procedure stipulated in Table B-2 for verifying




           the concentration  of ammonia  is inadequate as the excess acid




           in the absorbing solution must first  be neutralized before




           reacting the  collected  ammonia with the phenate reagent.




           Instructions  for neutralization with  0.1 N NaOH solution




           should be added or better yet, the procedure should be




           changed to  the Intersociety Committee Method, "Tentative




           Method of Analysis for Ammonia in  the Atmosphere  (Indophenol




           Method), ISC Method 42604-01-72T,  Health Laboratory Science,




           Vol 10, No. 2 (April 1973).
5.2.3  Calibration
       a.  1T53.2l(c) (2) :  The procedure states that the analyzer




           baseline be offset by adjusting the recorder controls.



                                 -  16  -

-------
   However, some analyzers do not indicate below zero output




   response with less than zero input when their controls are




   set at  zero.  In these instances, offsetting the recorder




   by 5% without adjusting the analyzer zero output to an up-




   scale value would be meaningless.  We recommend that the




   procedure  be modified so  that the adjustment to obtain 5%




   offset  zero reading on the chart be made, when possible, with




   the  analyzer  controls rather than with  the  recorder's.




   Therefore, the  third sentence of the paragraph should be




   rewritten as  follows, "adjust the instrument controls,




   when possible,  to  obtain  a 5% offset  zero  reading  on  the




    recorder chart."








b.  The output signals of  all analyzers  are not always in volts




    or millivolts,  but may  also  be  described in some  other  units.




    Therefore, revise the last sentence  of the paragraph  to




    read "....volts, millivolts, milliamps or other appropriate




    units".








c.  The procedure stipulates that the calibration curve is  to




    consist of a graphical plot of pollutant concentrations




    versus analyzer response.  The procedure [53.22(b) - (f)]




    subsequently specifies that the analyzer readings are to be




    converted to concentrations by reference to the curve.   This




    use of the curve, however, is subject to bias and may yield




    inconsistent or erroneous data.  To minimize such inconsistencies,




    we recommend that the pollutant concentrations be determined by






                         - 17 -

-------
dividing analyzer readings by the slope of the calibration




curve and that such slope be obtained by the method of least-




squares .









Figure 3 shows the Technicon analyzer response to SC>2 when




treated by  the least  square, best-fit technique.  A straight




line with a slope of  118.7  and  a y-intercept at 1.68 was




obtained.   The greater than 100 slope indicated that the




analyzer's  response  to SC>2  gas  was  greater than the response




predicted by standard calibrating solutions.  While the




 cause of this discrepancy could not be  resolved,  this  did




not affect  the performance of the analyzer and  the cali-




bration was corrected to a slope of 100 by spanning the




 instrument  to read  the span 862 gas concentration.









 However, a plot  of  the Melpar analyzer readings versus con-




 centrations yielded, on linear coordinate paper,  a  convex




 curve which appeared logarithmic (Figure 5).   The readings




 and concentrations  were then converted to their respective




 logarithms.  The least-square line then yielded a straight




 line with a slope of 0.322 and a y-intercept of 1.97  as




 shown in Figure  4.









These results confirm that the best-fit technique can be




applied to  both  linear and logarithmic response and indicate




that the mathematical conversion of the instrument outputs  to




pollutant concentrations by reference to the slope of the




curve will  yield more consistent results.





                       -  18  -

-------
5.2.4  Noise
           This procedure calls for determinations at 0 and 80% of




           full scale by taking 25 random readings with a digital




           meter  (DM) over 1-hour periods.  The permissible tolerances




           are given in Table B-l [53.20(d>] of the regulations.  The




           evaluation revealed the following changes were needed:








       a.  1F53.22(b)(l) line 8:  Put period after  "units" and add




           "The specifications are" before "given".








       b.  H53.22(b)(2)(ii):  The recorder is not  required in this




           test.  Delete the entire procedural step  and  renumber  (iii)




           to (viii)  to  (ii) to  (vii) .








       c.  1153.22(b) (2) (iii) :  No  chart tracings  can be  submitted since




           a DM is  used  instead  of  a  recorder  for the test as directed




           in step  (ii).   Delete the  second  sentence.








       d.  U53.22(b)(2)(vii):  The  procedure calls for determining the




           mean (£),  deviations  (di)  and standard deviation  (s)  directly




           on the digital  meter  (DM)  readings  obtained.   While  this




           method is  valid for linear response where the deviations




           are proportional to the  DM readings,  it cannot, however,




           be applied where the  deviations are not proportional  to




           DM readings,  as in  the  log-linear response of some instru-




           ments  (e.g. Melpar).




                                  -  19 -

-------
We recommend instead that the individual DM readings be




converted first to concentration units before determining




noise.  To accomplish this, steps  (vi) to  (viii) should be




revised to read [Refer  to H53.22(b)(2)(ii  in 5.2.4(b) above




for  recommendation in numbering of stepsjas follows:








    "(v)   Convert  the individual DM readings




    k]_, k2	k£5 to concentration units  cj_,




    C£	025  by reference to  the previously




    determined  calibration curve of the analyzer,




    and determine  the mean by:
                c =
                            25
    Deviations about the mean are determined by
                        - c
    and the standard deviation or noise is,
                              .(dj)'
    Report  the noise at 0% of the upper range limit in ppm.









    (vi)  Repeat steps  (iv)  through  (v)  using a pollutant




    test gas concentration producing an output of  80 ± 5%




    of the specified upper range  limit  and report  the noise."



                       - 20 -

-------
5.2.5  Lower Detectable Limit



       a.  H53.22(c)(2)(ii):  The test procedure requires that the test gas

           concentration equal to the specified lower detectable limit

           value be generated [Table B-l, K53.20(d)] and verified

            [H53.21(a)].  Our experiences indicate that generation of

            such low concentrations  (0.01 ppm) accurately is difficult

            and, moreover,  that the  sensitivity of most reference methods

            is insufficient at such  low  levels (i.e., 0.02 ppm).  To

            alleviate  this  problem,  we recommend  that the test procedure

            be modified to  read  after procedural  step  (i):



               "(ii)   Generate a test atmosphere equal  to 20*  ±  5% of the

               upper range limit.  Verify the concentration with the

               reference procedure.  Calculate and increase the diluent

               air flow rate accordingly to obtain the concentration

               specified for lower  detectable limit in Table B-l.  Sample

               the gas stream.   Record  the stable reading.



               (iii)  Convert the reading to concentration unit.  Pass

               the instrument for this  specification, if this value is

               equal  to or greater  than twice the noise at zero obtained

               in 1T53.22(b)(2)(vi)."
    *Note:   The 20% value is suggested since  this  concentration is  also
            needed for many of the other tests.

-------
5.2.6  Interference Equivalent









           Interference Equivalent is determined by challenging the




           instrument with substances that are known or suspected to




           cause interference.  The procedure involves mixing indi-




           vidual  interferents with a specified pollutant concen-




           tration and  comparing  the instrument's  output to the




           output  caused by  the pollutant alone.   Where the inter-




           ferent  is known to  react with the pollutant in the gas




           phase,  the response to the interferent  is determined in




           the absence  of the  pollutant.  The responses of the Technicon




           and Melpar analyzers to the  various  interferents specified




           in Table B-3 of the regulations  are  shown in Table V.









        a.  U53.22(d)(2) line 8:   "An interferent"  should be changed to




           "The interference".









        b.  1F53.22(d)(2) (iii) :  The procedure calls for generating three




           test atmospheres; pollutant, interferent and zero gases




           arbitrarily  labeled X,Y and  Z.   Except  for Z, the other




           labels  do not correspond to  the  identity of the test gases.




           Also, in 5.2.2 (a)  of  this report, we recommend that the




           design  of the gas generation system  allow  for a second source




           of pollutant gas  to permit concurrent generation of two




           different pollutant gas  concentrations. This will require a




           new designation for the second gas.  To facilitate identification










                                  - 22  -

-------
    of the various test gases  during  the  testing, we suggest that




    the two pollutant gases  be labeled as P  and  PP,  the  interferent




    gas as I and the Z for zero gas be retained.








c.  K53.22(d)(2)(iv)(b) and (c):  Directions here stipulate that




    the concentrations of the pollutant and the interferent should




    be adjusted such that when each is mixed with either zero  air




    or the interferent atmosphere, the specified test concentrations




    are obtained.  No provision for establishing the concentrations




    are provided.  We recommend that  the concentrations be first




    verified by  the  appropriate reference methods before tests  are




    begun.








 d.   H53.22(d)(2)(ix):   Add  "equivalent"  after the word  "interference".









 e.   1f53,22(d)(2)(xiv> line 2:  Delete "s" in "interferences"  and add




     "equivalents".








 f.   1f53.22(e):  Our experiences indicate that the 25,000 ppm of




     water vapor specified in Table B-3 is impractical because of




     condensation in the test  line.  We recommend that this con-




     centration be reduced to  18,000 ppm  (80% RH) which is readily




     generated at 30°C without problems.








  g.  Footnote  (a) of Table  B-3 specifies  that the interferent concen-




      tration shall be  prepared to ±  20%  of  the  required value and controlled

-------
          to ± 10%.  We believe  that  this requirement is confusing, too




          liberal and inconsistent with  the ±  5%  required  for pollutant




          test concentrations.   Furthermore, manufacturers whose instru-




          ments  are  borderline with  respect to interferences will be tempted




          to select  interferent  levels that will  minimize  the interference




          equivalent.  We recommend  that footnote (a) be revised to read




          "Concentrations of interferents listed  must be prepared and




           controlled to  ± 5% of the  stated value".








       h. Table B-3 does not provide a test for interference of airborne




           particulate matter.  Recent information suggest  that analyzers




          with spectrometric detectors that measure physical properties




           of pollutants  also respond to aerosols.  We recommend that a




           test procedure and specification for particulate interference




           be developed.









5.2.7  Test Procedures for Zero Drift, Span Drift, Precision, Rise Time,




       Fall Time and Lag Time.









       a.  fl53.22(f):  The general direction [(U53.22(f)(l)J specifies  that




           the tests for  zero drift,  span drift, precision, rise  time,  fall




           time and lag time are to be performed seven times over a period




          of at least seven days following the sampling plan given in




          Table B-4 of the regulations and that each test  shall  be run




          only once a day.  (See also 5.2.1(c) of this  report).

-------
   The completion, however, of the required daily tests for these




   response elements depends on the speed of response of the




   method  (analyzer) under test.  Our experiences with the




   Technicon  and Melpar  analyzers show that even analyzers




   with response time  characteristics equal to the specifications




   given in Table  B-l  in the regulations would require at least




   14 continuous hours under ideal  conditions .  Since 12 hours




   of the  24  hours in  a  day is  allotted  for sampling zero air




   for the 12-hour zero  drift  test,  it is  evident that all of




   these tests cannot  be completed  within  the  time period allotted,




   In fact,  our calculations  indicate that the daily tests specified




   can be  completed only with  analyzers  that  can  attain  100% response




           in 12 minutes  or less.
    These results indicate that the requirements in the test plan




    appear incompatible with the performance specifications in the




    regulations and are thus inapplicable to many existing instru-




    ments.  Our recommendations in *53.22(f) (6) (ii) to H53.22(f) (i) (ii)




    in Appendix II will help alleviate these problems.








b.  Our examination also indicate that some of the details regarding




    the test directions and the test plan given in Table B-4 are




    inadequate and give rise to questions such as:









    1)  Are the test procedures for response times (lag, rise




        and fall  time) and precision conducted daily in addition




        to the zero and span drifts tests?





                           - 25 -

-------
           2)  What  specific  line voltage  and test room temperature are




              used  when making the initial zero and span  adjustments?









           3)  Are the subsequent instrument adjustments at the end of




               the 3-day periods made at the test conditions in (2) above?









5.2.8  Zero and Span Drift








           To reduce the possibility of misinterpretation we recommend




           the following changes be made in  this  section:









       a.  U53.22(f)(4)(ii):  The  actual line voltage and temperature to




           be used  at the beginning of the 7-day  test should be definitely




           stated.  Rephrase the second sentence  to  "Instrument is operated




           at normal line voltage  of 125 VAC and  normal temperature of




           20°C".








       b.  153.22(f)(4)(iii) :   In  some instruments,  the span control interacts




           with the zero control.  To insure that no shift  has occurred in




           the zero baseline whenever the  span  control  is  adjusted, add




           the following after  the last sentence, "Rezero  the  instrument.




           Where  the zero and span controls  are not  electrically independent,




           the adjustments may  require repeated trials".   Also add after




           "percent" in the first  sentence,  "(Span 1)"  to  conform with  the




           test direction in Table B-4.








       c.  «53.22(f)(4)(iv):  Add  "(Span 2)" after "percent" to conform




           with the test direction in Table  B-4.   Also  add  after the last



                                  - 26 -

-------
    sentence,  "Start  Day  0".  This  addition will alert the tester




    when to start time zero.  Furthermore  the  change in designation




    from Day 1 to Day 0 will  be more consistent with the terminology




    of the 7-day or 15-day test.  At the end  of  7-day period  the




    results are examined, not on the "8th day" as  directed in the




    test procedures.








d.  f53.22(f)(4)(vi) to 53.22(f)(4)(xi):  For greater  clarity, we




    recommend that the steps in the test procedure be  rearranged




    somewhat and additional information added as follows:









        "(vi)  In the evening of Day 0, sample zero air for a




        minimum  of  12 hours.  Each  morning, starting on Day 1




        determine the minimum and maximum  readings during this




        period and  record the difference of the readings in




        concentration units as  the  daily  12-hour zero drift.




        Starting in the evening of  Day  1  and  each  evening there-




        after,  adjust the line  voltage  and room  temperature




        according to  Table B-4  before this test  is begun.








        (vii)   Each day,  twenty-four (24)  hours  after  the previous




        day's  readings, and starting on Day 1 sample zero air and




        test atmospheres  containing pollutant concentrations  equal




        to 20 ± 5% and 80 ± 5% of the upper range  limit.  The test




        concentrations shall be consistent from day to day.   Record




        the stable readings at zero and at each of the span  con-




        centrations .





                           -  2? -

-------
              (viii)   Convert the readings to concentration units.  Subtract




              each day's values respectively from that of the previous day




              and record as the 24-hour zero drift and span drifts at 20




              and 80% of the upper rangp limit for that day.








              (ix)  Determine lag time, rise time, fall time and precision




              as outlined in 1f53.22(f) (5) to 1153.22(f) (8) after the zero




              and span drift tests have been completed for the day.








              (x)  Repeat steps  (vi)  to  (ix) daily for an uninterrupted




              period  of seven  (7) or  fifteen  (15) days during which no




              adjustments to the instrument shall be  permitted.  To avoid




              testing during weekends, the tests may  be performed in three-




              day periods where  adjustments are made  on Monday and deter-



              minations are made on Tuesday, Wednesday and Thursday or




              adjustments may be made on Tuesday and  determinations are




              made on Wednesday, Thursday and  Friday. All adjustments




              to  the  instrument  shall be made  at the  line voltage and




              room temperature of the previous day.








              (xi)  At  the end of Day 7,  examine the  test results and determine




              whether to continue or  terminate further testing as directed  in




              H53.20."
5.2.9  Lag Time
          The deficiencies found in  this  procedure were minor.   The




          changes suggested to alleviate  these  are:





                                 - 28 -

-------
       a.   1F53.22(f)(5)(i),  line 5:   Insert "equal to  twice  the noise" after




           "change" in the definition.   This qualification will insure that




           the response observed is  due to 'che input concentration,  not




           instrument noise.









       b.   U53.22(f)(5)(ii)(b) change "(iv)" to "(iii)" in  the sentence.




           The change will require the determination of lag time  by




           sampling a pollutant concentration equal 80 ± 5% instead  of




           the 20 ± 5% of the upper range limit and will eliminate the




           need to generate another concentration for the determination




           of rise time which immediately follows.









       c.   1T53.22(f)(5)(ii)(c):  Insert "equal to twice the noise" after




           "change" for the same reason as in f53.22(f)(5)(i) above.









       d.   1T53.22(f)(5)(ii)(d):  Add, "as the lag time", after "time".
5.2.10 Rise Time
       a.  H53.22(f)(6)(ii) :  Steps (b) through (e) eon-tain several re-




           dundant steps, i.e. determining the analyzer's response at 80




           ± 5% of the upper range limit when this has already been




           established in 153.22(f)(4)(ill), calculating 95% of .this




           reading and rezeroing the instrument.  We recommend the




           test procedure be revised, as indicated by the underscores,



           to read as follows:








                                 - 29 -

-------
               (a)  Sample zero air until stable reading is obtained.








               (b)  Calculate 95% of the stable reading obtained in




                    H53.22(f)(4)(iii).








               (c)  Sample the same concentration  (80 ± 5%) used in




                    H53.22(f)(4)(iii) and use a stopwatch to time the




                    interval starting with the first observable response




                    and ending when the response equals the reading in




                    (b) above.








               (d)  Record the elapsed time, and allow the analyzer to




                    continue sampling until a stable reading is obtained




                    in preparation for the fall time test  [see H53.22(f)




                    (7)(11)1.








               (e)  Repeat steps  (a), (c) and (d)  above after completion




                    of fall time  test.








               (f)  Report the average of the values obtained in steps




                    (c) and (e).








5.2.11  Fall Time




        a.  f53.22(f)(7)(ii):  The test procedure  for fall time may be




            likewise shortened by combining the rise and fall time tests




            into one continuous operation.  The changes suggested, which




            are indicated by the underscores, are  as follows:






                                  - 30 -

-------
              (a)  Calculate 5% of the stable reading In  (f) (6) (ii) (d)




                   of the rise time procedure.









              (b)  While the analyzer is  at  stable reading  from sampling




                   the  concentration in  (f)(6)(11)(d) ,  switch  to sample




                   zero air and use a stopwatch  to time the interval




                   starting with  the first  observable  instrument response




                   due  to the  step  change and ending when the  response




                   equals the  value calculated in (a)  above.









               (c)  Record the  elapsed  time.









               (d)  Repeat steps  (a),  (b)  and (c).









               (e)  Report  the  average  of the values  obtained in steps (c)




                   and (d).
5.2.12  Precision
        a.   H53.22(f)(8)(ii):  The procedure for precision requires 6




            determinations each at 20 and 80% of the upper range limit,




            totaling 12 determinations per day or 84 for seven days.




            The number of daily determinations specified appears to be




            excessive and not easily completed in one day.  We suggest




            that the number of determinations be reduced to 3 each at




            20 and 80% of the upper range limit.  This will give a total




            of 6 precision values per day; 42 for the seven-day test.





                                   - 31 -

-------
This reduction in the number of determinations will allow




more time to complete other tests.









We therefore recommend  that this  procedure be  revised  as




indicated by the underscores  to read as follows:









    (a)   Allow  sufficient time for instrument warm-up and




         stabilization.









    (b)   Sample zero air until a stable reading is  obtained.




         Do  not record.









    (c)   Sample the same pollutant concentration of 20  ± 5%




         of  the upper range as used in 1T53.22(f) (4) (vi) .




         Record the stable reading.









    (d)   Sample a pollutant concentration at least  50%  higher




         (30% of upper range limit) than in (c) above until




         a stable reading is obtained.  Do not record.









    (e)   Quickly switch to the pollutant concentration  used




         in  (c).  Record the reading.









    (f)   Add to the readings obtained in steps (c)  and  e), the




         reading on the same day for the span at 20 ±  5% of the




         upper  range limit in 1T53.22(f) (4) (vi) .








                       - 32 -

-------
(g)   Convert the readings to concentration units and calculate




     the mean and the standard deviation for the three values.









(h)   Repeat step (b).








(i)   Sample the same pollutant concentration of 80 ± 5% of




     the upper range limit  as used in H(53.22(f)(4)(yi).




     Record the stable  reading.








(j)  Sample a pollutant concentration  20% higher  than  in  (i)




     but not  to exceed  97% of full  scale.  Do  not record.









 (k)  Quickly  switch to  the pollutant concentration used in




      (i).   Record the reading.








 (1)   Add to the readings obtained in steps (i) and (k), the




      reading obtained on the same day for the span at 80




      ± 5% of the upper range limit in 1F53.22(f) (4) (vi) .









 (m)  Convert the readings  to concentration units and




      calculate the  mean and  standard deviation for the




      three values.








 (n)  Report  the values obtained in  steps (g)  and (m)  as




      the  precision at  20  and 80% of the upper range  limit,




      respectively.








                     - 33 -

-------
    b.   f53.22(f)(8)(g):   Our experiences indicate that the zero  and




         span drift tests  (Table B-4)  is subject to misinterpretation




         [See also 5.2.1(c) of this report].  Additional directions  to




         augment those in the test procedures are needed.  Refer to




         Table B-4 of Appendix II or Table IX of this report for our




         version which we believe is clearer.









5.3  EVALUATION OF SUBPART C








     Subpart  C provides separate  test procedures designed to demonstrate




consistent  relationship between candidate and reference methods for




oxidants, carbon monoxide  and sulfur dioxide.  The test procedures




require  that  the candidate and reference methods be operated side-by-




side while  sampling  ambient air containing  pollutant concentrations




at three specified levels.  The candidate method is considered to




have a consistent  relationship with the reference method when the




differences between  the respective hourly average results at each




concentration level  do not exceed the specified tolerances.









The test data obtained in  the Phase I and II evaluation of Subpart C




are shown in  Tables  X through XII.  The results show that while none




of the seven  measurements  made in the initial tests with the Bendix




Oo and Technicon S02 analyzers exceeded the tolerance  specifications,




the Melpar  analyzer  exceeded  the  tolerance  specification once  in




the 24-hour measurements and  five times in  the  1-hour  measurements.




All the  failures,  further, occurred at  the  high S0£  (O.to  to 0.50 ppm)




level.   According  to the S02  and  oxidant acceptance plan [H53.34(b)

-------
of the regulations] ,  the Melpar analyzer failed to demonstrate con-




sistent relationship  with the reference method while the Bendix and




Technicon analyzers demonstrated consistent relationship with the




respective reference procedures.  Since the Bendix and Technicon




passed the initial 7 tests and the Kelpar failed, the second test




consisting of 8 additional measurements was not required and there-




fore not conducted.  The Phase 1 evaluation, further, showed that




the three test procedures (CO, S02, and Ox) were essentially the




same except in the test concentration  ranges and in the sampling time




periods required and that much of the  information and test directions




were repetitious.  Based on  these findings, the procedures in this




subpart were redrafted and a single general procedure applicable to




the three pollutants was developed.  The  revised  draft  is shown in




Appendix III.  The evaluation  in Phase II of  the  revised procedure




confirmed that the improved  procedure  was shorter,  less confusing,




informationally  more complete  and more direct.   It  has  the advantage




of being adaptable to other  methods by simply appending the  application




and by adding the  specific information, test  specifications, appropriate




sampling plans and appropriate interpretation criteria  to  the directions.









The  above  results, however,  does  not  mean that all Melpar  units  are




 incapable  of demonstrating consistent relationship.  As stated in 3.4




 of  this report,  all  the test instruments were used units  and all required




 some repair and  maintenance  to make them operational.  In particular,  a




 shorted igniter  and a frosted heat-shield window of the reaction chamber




 were replaced  in the Melpar unit, just prior to initiating evaluation  of




 Subpart C.   Subsequent calibration curves no longer corresponded to  the

-------
normal log-linear response obtained in the earlier evaluation of Subpart




B.  The response, moreover, appeared  to be disproportionately less




sensitive at  the higher  (>0.35  ppm) test  gas  concentrations in comparison




to earlier  data.  We  believe  however, that these  deficiencies did not




interfere with the  objective  of evaluating the test  procedure and not




the  analyzers.









 5.A   REVIEW OF RESPONDENTS'  COMMENTS








     The cotmnents of the respondents are categorized and tabulated in




 Appendix IV by subject matter.  In most cases, it was found that the




 comments were already covered in the regulations and/or in the AIHL's




 progress reports to EPA.  In some instances, new information were cited.




 The merits of which should be considered in the final version of the




 regulations.  The AIHL remarks and recommendations are appended  to  the




 comments.








 In general the comments indicated that the proposed regulations  have  been




 well received by most of the respondents.  Five respondents (#19,  23,  24,




 26 and 27) offered commendations for a piece of work well done while  six




 expressed support for the intent and provisions of the regulations.  Only




 one respondent questioned the basis  and need for such regulations.   Two




 respondents  thought the development  of such regulations should include or




 be a function of professional standard organizations such as ASTM, ANSI




 or NBS while one (#22) voiced a fear that the provisions could jeopardize




 the  outcome of analyzer evaluation studies currently conducted by Southwest




 Research Institute, ASTM and Midwest Research Institute if the test procedures




of the proposed regulations are not  congruent with those of the studies.





                                 -  36  -

-------
6.0  MANPOWER REQUIREMENTS AND COSTS









    The time to establish equivalency obviously depends on 1)  the  mechanism




required;  i.e., whether Subpart 3, C or both are required, 2)  the  analyj-




response time characteristics, 3) the speed with which the test gas  con-




centrations can be verified, 4) the number and kinds of interferent  test




gases required and, 5) the performance of the analyzer; i.e.,  number




of failures encountered.  The man-hours required to conduct the procedures




in Subparts B and C are shown in Tables XIII through XIV and discussed




below.









6.1  SUBPART B









    As explained in 5.0 of this report, the evaluation of the procedures




for noise, lower detectable limit and interference equivalent were con-




ducted only once.  While the values for these parameters in Table XIII




refer to the time required to conduct each test seven times, they are,




in fact, extrapolated values based on the results of single trials.









The data in Table XIII show that  the total man-hours required to conduct




the seven trials in Subpart B ranged from 482 to 1452 man-hours.  The




data further indicate that the slow responding Technicon unit required




the most time  (1452 man-hours) and that the faster responding Melpar




and Bendix units were approximately equal (482 and 502 man-hours,




respectively).  The data also indicate  that approximately 50% of the




Technicon and Melpar times and 72% of the Bendix times could be attributed




L.) the tests for interference equivalent.  The values  reported in Table XIII

-------
 include the  times  for  familiarization with  the  analytical  procedures  for

 verifying  the reference and interferent test gas concentrations,  preparation

 and  standardization of reagents, etc.  In ongoing, routine programs,  some

 of these activities may occur only occasionally.  For such cases,  the

 man-hours  allocated for this purpose may be subtracted and thus reduce,

 to some extent, the time to demonstrate equivalency.



 Details regarding the costs and man-hour requirements for  the  individual

  tests and the gas generation system are discussed below.



6.1.1   Gas Generation System



       To provide the pollutant and interferent test gas concentrations

       specified in Subpart B, a gas generation system is required.  Our

       records show that the design, fabrication and installation of the

       system described in 3.2 of this report required about 176 man-hours

       at a cost of about $4700 for materials.  This cost is broken down

       as follows:



                      $1000   Cost of parts and materials

                       3400   Cost of instrumentation (i.e., mass flowmeter,
                              strip-chart recorder, constant temperature bath)

                        300   Cost of permeation tubes and  cylinders of other
                     	   reference and test gases.
                      $4700   Total



      In addition,  about  $1000 should be added to the above to  cover  the

      cost of approximately  34 cylinders of  zero air needed to  conduct

      the performance test procedures.


                                 - 38  -

-------
6.1.2  Calibration [53.21(c)]









      Table XIII shows that to calibrate each analyzer separately would




  require about 80 man-hours for the Technicon, 64 for the Melpar and 16




  for the Bendix.  These times do not include a variety or "overhead"




  items, such as staff leave time, erroneous starts and supervision.




  The times for the Technicon and Melpar may be allocated as follows:




  1)  24 man-hours to calibrate the spectrophotometer for the pararo-




  saniline reference method   , 2)  24 man-hours to prepare the Technicon




  for operation and calibration, 16 aian-hours  for the Melpar, and 3)   the




  remaining time to generate  the necessary S(>2 gas streams, and to conduct




  the calibration.









  The time for calibrating  the spectrophotometer  includes  the time to




  prepare all glasswares, reagents, and standardization of the reagents.




  The times to prepare  the  analyzers  for operation and calibration in-




  clude  familiarization with the  instruments,  interfacing  the outputs




  with strip chart recorders, photometer alignment  (Technicon), maintenance,




  installation of auxilliary accessories and calibration  of  the  flow-




  metering  devices.   It does not  include the time spent  to troubleshoot  and




  repair malfunctions.   The time  required to obtain  the  actual calibration




  data  includes  the  generation  of seven calibrating  gas  concentrations,




  manual collection  of  samples  for verification,  analysis  of the collected




  samples  and  sampling  the gas  streams with the analyzer.









 6.1.3  Noise [53.22(b)] and Lower Detectable Limit [53.22(c)]:









       The  data in Table XIII indicates that the times to conduct the

-------
 seven trials for noise and lower detectable limit with the Melpar and




 Technicon units were up to twice that of  the Bendix.  Although the




 speed of response of the Bendix and  the Melpar  are  similar, the




 minimum test period for the Melpar unit is  restricted by  the test gas




 verification scheme.  The manual pararosaniline reference method for




 S02 requires up to  30 minutes for  sample  collection and additional




 30 minutes  for analysis whereas the  verification scheme for ozone




 requires  five  to ten minutes  for sample  collection and 15 minutes




  for analysis.









6.1.4  Interference Equivalent [53.22(d)]:









      The determination of interference equivalent depends  not only on  the




  response time characteristics of the analyzer but also on the  measure-




  ment principle used to detect the pollutant since this  fact will




  determine the number and types of interferents to be tested.   Of the




  12 potential interferents listed in Table B-3 of the regulations,




  Table XIII shows that 700 man-hours would be required to  evaluate the




  effect of the seven interferents listed for the colorimetric principle




  (Technicon) for S02 whereas 252 man-hours are required for the three




  interferents listed for the flame photometric (Melpar)  and 364 man-hours




  to test the four interferents for the chemiluminescent ozone  (Bendix).




 As observed in 6.1 above,  these times constituted approximately 50% of  the




 total time  required for determining the performance of the Technicon




 and Melpar  units and  72% of the time for the Bendix.  The times given




 for each interferent  equivalent includes familiarization with  the




 individual  method,  preparation of  spectrophotometer calibration curve,





                                 - kO -

-------
  generating the desired interferent test concentration,  verifying  the




  concentration, and all calculations.









6.1.5  Zero Drift, Span Drift, Precision, Rise Time, Fall Time  and  Lag  !!:=.=




       [53.22(f)]:









      This procedure calls for conducting the tests for these parameters




  once a day over a period of at least seven days, of which at  least  three-




  day periods must be contiguous.  The procedure further provides  separate




  test procedures for each parameter inferring that each test is to be




  conducted as a separate entity.  The procedure further stipulates that  as




  a part of the test for zero drift, the test instrument is to sample zero




  air daily for a period of no less than 12 hours.  But, since this test  is




  conducted during the night-time hours unattended, the man-hours  for




  this test indicated in Table XIII were not included in the total man-hours.









  The evaluation of this procedure revealed, as expected, that with instru-




  ments with response time characteristics on the order of tg5 = 13 sec




  such as the Bendix and Melpar  analyzers, all the  daily tests could be




  completed during a normal workday.  However, with the slow-responding




  Technicon analyzer (tg5 = 41 min) ,  only the drift and response time




  tests could be  completed in one normal workday, leaving insufficient




  time to perform the test for precision.  Furthermore, with the Technicon,




  at least 32 man-hours would be required to obtain the precision values




  specified in  Table B-4  for  a single day and 224 man-hours to repeat each




  test each day for seven days.

-------
On the other hand, the data in Table XIII also show that the tests for




response times  (lag, rise and fall times) required a total of 63




man-hours for the Technicon and 18 man-hours  for  the Melpar and Bendix.




However, these  tests can be conducted in conjunction with the tests




for precision with no increase in man-hours.  By  combining these tests,




the net man-hours could be reduced somewhat  to  1388 vs  1452 for the




Technicon,  464  vs 482  for the Melpar and 484 vs  502 for the Bendix.









6.2   SUBPAKT C









     The data in Table  XIV show  that  in  contrast to  the  evaluation of




 Subpart B,  the  test analyzer  performance were of little consequence




because the test procedure  calls  for making 1-hour  and  24-hour average




measurements.   Thus, the man-hours  required to establish consistent




 relationship is primarily  a function of the test plan requirements as




 indicated by the identical  values for  the  862 analyzers and the somewhat




 lower values for the Ox analyzer.  Finally, Table XIII  shows that the




 additional time to  conduct  the  second  test series is only about 25% even




 though the number of samples  to be  taken is somewhat greater since many




of the preparatory-type operations  have already been completed during  the




first test  series.









Table XIV also  shows the man-hours  required to design,  fabricate  and




install  the sampling probe  excluding any  gas generation system.
                               - 1*2 -

-------
6.2.1  Sampling Probe and Manifold









      The cost of the air sampling system described in 3.3 of this  report




  was relatively inexpensive.  The glass pipes for the ducting are  available




  from most plumbing supply houses and can be purchased in about 2  or 3




  standard lengths of up to 10 ft.  The sampling manifold was the most ex-




  pensive component  ($70.00) of the air sampling system because it  was




  specially designed and fabricated.  As shown in Table XIV, approximately




  80 man-hours were  required to design, construct and install the entire




  sampling system.   The cost was  about $180.00 excluding labor.  The costs




  of the mass  flowraeter, strip-chart recorder and constant temperature bath




  were not considered since  they  are a part  of the  gas generation system




  described in 3.2 of this report.









 6.2.2  Oxidant Methods  [53.31]









      As  can  be seen in Table  XIV,  exclusive of  the time  to  construct  and




  install the sampling  probe,  approximately  104  hours  were required  to obtain




  the initial 14 measurement for consistent  relationship  with the  oxidant




   (Bendix)  analyzer.









  When  the second sample set consisting of 18 measurements is required, it is




   estimated about 72 additional man-hours will be needed.  The time  to conduct




   the second sample set is less because the instruments need not be  calibrated




   again.  This will make the total number of man-hours about 256 or  176 ex-




   cluding the sampling probe.

-------
6.2.3  Carbon Monoxide (CO) Methods:









      This procedure  (1F53.32 of the regulations) is essentially identical




  to that for oxidant methods.  Therefore, not only do the recommendations




  for improving the oxidant procedure  apply to this method but due to the




  similarity in the response time characteristics of these instruments to




  the Bendix oxidant  analyzer, it is estimated that the man-hours required




  to conduct this  test would essentially be the same as that for the oxidant




  method.









 6.2.4   Sulfur Dioxide Methods









      Based on our experiences with  the Melpar and Technicon analyzers, it is




  estimated that excluding the sampling probe, it would require a minimum cf




  154 man-hours to conduct the first sample set and, in the event of one or




  two failures, an additional 72 man-hours to obtain the eight measurements




  required for the second set for a  total of about 226 man-hours.  With the




  time  to design,  construct and install the sampling probe systems included,




  the time needed  to  conduct the first sample set is increased to about 234




  man-hours, and to 306 man-hours in the event the second sample set is




  required.









  6.3  OVERALL TIMES









      Table XV summarizes  the data shown in Tables XIII and XIV and shows




  the total man-hours required by each analyzer to conduct the tests for

-------
performance and consistent relationship.  The table shows that in addition




to the man-hours for the design, fabrication and installation of the




gas generation system and the sampling probe, the Technicon analyzer




would require about 1540 man-hours to conduct the initial test set




for both performance and consistent relationship, the Melpar about




570 and the Bendix about 590.  In the event a second test set is




required, the total man-hours increases to about 3110 man-hours for




the Technicon and 1200 man-hours for the Melpar and Bendix.









These results indicate that  cost of determining equivalency depends on




the operating principles and characteristics as well as  the performance




of the candidate analyzers and that the determination process may be




as short as 104 man-hours when consistent relationship only is  re-




quired to as long as 3110 man-hours in  the event all contingencies




are required.  The data in Table XIII  further indicate that a




substantial proportion  (between 50  to  75%) of the  effort to determine




performance may be attributed to the determination of interference




equivalent.









6.4   CERTIFICATION









     For  equivalency  certification,  the regulations require,  in addition




 to specified  sworn statements,  other  data and  information that must be




submitted  to  EPA.  These  include  a)  all appropriate performance and




 consistent  relationship data obtained  in Subpart B and  C including




 recorder chart tracings,  test data and any other documentation, b)
                                 - 1*5 -

-------
detailed description of the candidate method including the measurement




principle, manufacturer's identifications, components, diagrams and other




relevant items,  c)  a description of  the  quality  control program used to




insure  that  all  the measurement  systems  will have  the same performance




characteristics  as  the candidate method, and d)  a  description of the




calibration  and  operational procedures prescribed  for field use of the




method.









Based on  the experience encountered  in  the evaluation of  the regulations,




it is estimated  that between  80  to 120 man-hours will be  required to check,




verify  and interpret  the submissions for each  method provided all necessarv




information are  included and  that confirmatory laboratory testing is not




required.









7.0  SUMMARY AND CONCLUSIONS









     The federal  Environmental Protection Agency  has proposed the equivalency




regulations  as a standard set of procedures  for  establishing the equivalency




of non-reference analytical methods  to reference methods which are prescribed




by the  Federal Clean Air Act  for the measurement of ambient air quality.  To




assess  the applicability, validity and workability of the regulations and to




determine  the costs to demonstrate equivalency,  the provisions, specifications




and  test procedures of the regulations were evaluated in  two phases using




three automated  methods or analyzers and two manual methods.  The comments




regarding  the proposed regulations submitted to  EPA by respondents were




also reviewed.

-------
The evaluation revealed that the provisions and specifications  given




are by-and-large consistent with prevailing ambient air measurement




technology.  Some of the test procedures and general instructions,




however, were found to be deficient.









The deficiencies found consisted of unclear test directions, inconsistent




test requirements, insufficient description of  the  gas  generation system,




redundant  test information  and procedures, missing  performance specifica-




tions,  poorly organized procedural  steps  and  typographical errors.









Subpart A  of  the regulations was  found to be  adequate  and only requires




a minor improvement  by  the  inclusion of a clear statement stipulating




that  the instruments or analyzers  and not measuring principles are to be




certified  for equivalency.   Substantial changes were made in Subparts B




and C resulting in the redrafting of these two subparts.









The changes made in Subpart B included, a) clarifying or supplying  the




 missing or proper test information, b) reorganizing some of the  proce-




 dural steps for proper sequence, c) developing a more reliable gas




 generation and verification procedure  for lower detectable limit, d)




 revising  the test procedure for noise  to require conversion of the DM




 readings  to concentration  units first  before treatment of the data to




 accommodate instrument outputs that  are  not  proportional to the input




 concentrations, 3)  recommending that  the conversion of the analyzer




  readings  to  concentration  units be accomplished by reference to the




  slope  of  the best-fit  curve rather than  by the graphical reference  to




  the  plot  of  the curve  for  better  accuracy, f)  improving the procedures

-------
for drifts and response times by eliminating unclear test directions




and by making the test information correspond to that in the prescribed




test plan, and g) modifying the test procedure for precision by reducing




the number of trials required to accommodate slower responding instruments.









In subpart C, the redundant test procedures for SC>2, CO and Ox were combined




into a general procedure applicable to all three methods.









The cost  in man-hours  to establish equivalency depends on several factors




such as,  a) the  mechanism  required  (meet performance only, consistent




relationship  only or both), b)  the number and kinds of interferent test




gases  required,  c)  whether additional tests are required, d) speed of




response  of the  method under  test, and e) time needed to generate and




verify the test  gas concentrations.  Thus, two analyzers of different




configuration but using the same measurement principle can differ greatly




in their  equivalency time  requirement.  On a routine basis, it is esti-




mated  that equivalency determination may require from as little as 100




to much as 3110  man-hours  depending on the performance of the candidate




method.









It is  further estimated that  to review the voluminous test data that will




be submitted  for equivalency  certification will  require  between 80 to




120 man-hours per method assuming all necessary  information is available




and no additional tests are required.









Lastly, considering the significance of the proposed rulemaking and  the




number of air analyzer manufacturers and users throughout the  country,




the 29 communications  received by EPA from respondents appeared quite small.





                               - he -

-------
The comments, however, were generally favorable.  Only a small minority




believed such regulations should be a function of organizations such as




ASTM or NBS.  Many of the technical questions raised in the communications




had already been answered in  the AIHL's bimonthly progress reports to EPA.




In some cases, suggestions and information previously not considered were




discussed by the respondents.  EPA should consider  their merits in the




final version of the regulations.

-------
8.0  REFERENCES:









     1.  Federal Register, Volume 38, no. 197, (Oct.  12,  1973)




     2.  Environmental Chamber, Frigitemp Corp., Brooklyn,  NY  11212




     3.  Fisher and Porter Co., Lab Crest Div.,  Warminster, PA




     4.  Hastings - Raydist, Hampton, VA  23361




     5.  AIRCO  Industrial Gases, P.O. Box 785, Oakland,  CA  94604




     6.  Corning Glass Works, 500 Smith Main St., Orange, CA  92668




     7.  Technicon Industrial Systems, Tarrytown, NY




     8.  Meloy  Laboratories, Inc., 6631 Iron Place, Springfield, VA  22151




     9.  Bendix Corporation, 1400 Taylor Avenue, Baltimore, MD  21204




    10.  Federal Register, Vol  35, No. 2273, (November 25,  1971)

-------
                                                        TABLE 1

                         Analyzers Used for Evaluating EPA's Proposed Equivalency Regulations
Manufacturer


Bendix Corporation

Ronceverte, W Va
                       Model
Pollutant
Principle
                                                       Chemiluminescent
Sampling Rate   	
   1/min	   Lag
                                                                                               Response Times
                                                                                                              (a)
                                       1.0
Rise
                                                                                                                  Fall
                                <3secs     "iQsecs     <10secs
Melpar, Inc.

Falls Church, Va
                       LL-1100-1B      Total Sulfur    Flame Photometric      0.200
Technicon Instr. Corp. IV

Tarrytown, NY
                                       SO-
                Colorimetric

                (pararosaniline)
                   0.40      11.5 min     3 min
                                     4.0 min
^'Manufacturer's specifications

-------
                              TABLE II

      Questions and Suggestions on Subpart B of the Regulations
                  Submitted to the Project Officer

                         (April 1973 Draft)
PAGE        SECTION                         ITEM



 15         53.20(d)    In Table B-l, the specifications for zero drift

                        are given for 12 and 24 hours.  No time interval

                        is specified for span drift.  What is the time

                        element required for span drift?



 16         53.21       This section specifies that the "temperature of

                        the gas generation system shall be maintained

                        within 2°C to assure stable output" [1F (a)] and

                        that the "test gas concentration shall be con-

                        sistent from day to day  [53.22(f)(4)(vi)]".  In

                        the test procedures for performance, however,

                        it is required that tests for zero and span

                        drifts (Table B-4) shall also be conducted at

                        20 and 30°C.  Does this mean that two temperature-

                        controlled test rooms or chambers, one for the

                        instruments under test and  one  for the gas

                        generation system, are required?  If a change

                        in the gas concentration does occur, how much

                        can be tolerated?  Can correction factors be

                        used?

-------
                          TABLE II  (con't)
PAGE
SECTION
                    ITEM
 16
53.21
Since, our test facility houses both the gas




generation system and the analyzers, this




situation poses a problem because the change




in chamber temperature may affect the test




gas concentration as well as the analyzers.




We propose either one of the following alter-




nate solutions:  1) adjust the concentration




of the test gas back to the reference concen-




tration after  the test room has equilibrated




to the new test temperature.  This  technique




might require  several determinations and




consequently time-consuming, or 2)  it may be




more expedient to calculate the resulting




concentration  of the test  gas  at the new  test




temperature.   The instrument output can  then




be related back  to  the  reference gas concen-




tration.
  20
 53.22(b)
                         The test procedure calls for use of "Digital




                         Voltmeter  (DVM) accurate to four significant




                         figures with one-second time constant on the




                         input...."  A voltmeter with these specifica-




                         tions and with  lower measuring range of 0 to
                                -  53 -

-------
                          TABLE II (con't)



PAGE        SECTION                         ITEM



 20         53.22(b)
            (2)(iii)    100 mv costs approximately $1700 and is not

                        readily available.  The cost increases sub-

                        stantially when a lower range capability of

                        0 to  1.0 mv is required for some analyzers.

                        The response times of reasonably priced

                         (about $650) DVM's are on the order of

                        milliseconds and microseconds.  How critical

                        is the one-second time constant requirement?



 33         53.22(f)(5) Lag time is defined in the regulations as

                        "the  time interval between a step change in

                        input concentration at the instrument inlet

                        to the first observable change equal to twice

                        the noise in the instrument output."  However,

                        in the test procedure, only the time to the

                        "first observable change" is cited.  For con-

                        sistency, the working in the test procedure

                        should correspond with the definition.  Also,

                        the same criterion should be used in the test

                        procedures for determining rise time [U(6)(ii)

                        and fall time 1F(7)(ii)].

-------
                          TABLE II (con't)
PAGE
SECTION
ITEM
 20
 to
 39
53.22...
53.30       Numbering skipped from 53.22 to 53.30.  Is

            53.23 to 53.29 missing?
 15
53.20(d)
(7)         Delete 95%  after  lag  time.
 25
53.22(d)
(2)(ill)
                         Change  the designation of  the  interferent  gas

                         to "y"  and zero air to "Z".  This  change will

                         make the latter easier to  remember.
 35
53.22(f)
(7)(ii)(d)   Should read "Repeat paragraph (a)  to  (c)".
  36
53.22(f)
(8)(ii)
(1)  &  GO
                         "Switch to sample" should be used in place of
                         "Rapidly reduce".  This is to insure unchanged

                         test gas concentration for the repeat determination,
                                -  55  -

-------
                              TABLE III

   Recommended Changes in Subpart C of the Equivalency  Regulations

                         (April 1973 Draft)



              I.  Editorial and Typographical  Changes:
Page

 40
 40
 40
 40
 41
 41
 42
 43
 43
 43
       Paragraph

 53.30 (c)
 53.30 (c)


 53.30 (e)




 53.31 (a)(l)
 53.31 (a)(3)


 53.31 (b)(l)


 53.31 (d)


 53.31 (c)(Table C-l)


 53.31 (d)(Table C-l)




 53.31 (d)(Table C-l)
Line                Change

  1     Change "party" to "applicant".
        The latter is defined in the
        general provisions of the Regu-
        lations .

  5     Correct the 'spigj3.ing of "identi-
        fied" to "identified".

  1     Correct "relationshiip" to
        "relationship".

  4     Correct "vallues" to "values".

  2     Correct "^accordance" to
        "accordance" .

  3     Correct "appropriate" to
        "appropriate".

  6     Correct "permissable" to
        "permissible".

  3     Correct "0.06-.1.1 ppra" to
        "0.06-0.1 ppm".

  1     Relabel paragraph from "(d)" to
        "(c)".

  1     Relabel Table C-l from "(c)" to
        "(d)".

        Delete the "<"sign in the toler-
        ance column and replace with "+"
        sign above the minus sign to read
        Correct the sample size for the
        first samples in the medium con-
        centration range from "5" to "4"
        and, in the high range from "4" to
        "5".
 46
53.32 (3)
  7      Correct "permissable" to "permissible"

-------
 51

 52


 54
 55

 56
 57


 57

 60



 60


 62
53.33 (a)(3)

53.33 (b)(3)


53.34
53.35 (a)(l)(iii)

53.35 (b)(Table  C-3)
53.35  (b)(l)(i)(a)


53.35  (b)(l)(i)b

53.35  (d)(l)(v)



53.35  (d)-(D(vi)


53.35  (d)(3)
   3     Correct "permissable"  to  permissible".

   2     Correct "section  53.31(c)"  to
        "I 53.31 (d)".

   2     Add "for sulfur dioxide"  after
        "method" for clarity.

   3     Change "party" to "applicant".
        Applicant is defined in the general
        provisions of the Regulations.

   8     Correct "permissable"  to  permissible".

        Correct "allowable number of differ-
        ences which can  fail tolerance  speci-
        fication" to "allowable number  of
        tolerance differences  to  pass con-
        sistent relationship specifications".

   4     Change "instrument" to "method" for
        consistency in  terminology.

   4     Same as above.

   2     Change "means of" to "the".

  11     Correct "3-hour"  to "1-hour".

   1     Correct "section" to "II"  for
        consistency in  terminology:

   4     Correct "four"  to "three".

   8     Correct "3-hour" to "1-hour".

  10     Correct "3-hour" to "1-hour".
              II.  Technical Changes and Improvements:
Page

 40
            Paragraph

     53.31  (a)(2)
            Change

Add after the last sentence, "In the absence
of a manufacturer's calibration procedure,
perform dynamic calibration with at least
7 identifiable points including 0 and 90%
of full scale using the appropriate method
in 40 C.F.R. Part 50."
                                 - 57 -

-------
41
53.31 (a)(3)
41
 42
 45
 53.31  (b)(2)
 53.31  (b)(3)
 53.31  (f)(2)
 46



 46



 47



 50



 51


 52
 53.32  (a)(2)



 53.32  (a)(3)



 53.32  (b)(3)



 53.33  (a)(2)



 53.33  (a)(3)


 53.33  (b)(3)
Add after the last sentence of the paragraph,
"When control samples indicate deviations of
greater than + 5% in the instrument response
from the latest calibration, discontinue
testing.  The deviations indicate possible
calibration shift or instrument malfunction.
Consult the instruction manual or the manu-
facturer and correct the cause of the problem.
Perform calibration and restart the testing
from the beginning of the operational period"
The inclusion of the above will provide a
course of action when control checks indicate
that the instrument is out of performance
specification.

Add after "methods", (i.e., eyeball, piarii-
meter, etc.)".  The addition of this phrase
will alert the applicant of the types "f
methods needed.

Delete the whole paragraph.  It is a 133-,
of what is discussed in detail in the
sampling procedure (11 53.31 (c)).

Revise the first sentence to, "If the
number of measurements from the series
of 14 which fail the stated speclficat.  >n
is one or two, conduct a second set of
18 measurements".  This will make the
sentence conform in intent and format to the
procedures and for the other pollutants.

Add after "manufacturer" the change
recommended for page 40 of the Regulations,
U 53.31  (a)(2).

Add after the  last sentence of the para-
graph  the change recommended for page 41,
f 53.31  (a)(3).

Delete the  paragraph.  It is redundant  since
the content is discussed in detail in
fl 53.32  (c).

Continue at the  end  of the sentence with
the recommendation given for page 40,
H 53.31  (a)(2).

Add at the end of the paragraph the change
recommended for  page 41, H 53.31  (s)(3).

Delete the paragraph.  The content is a repeat
of what is discussed in detail in 1i 53.33  (c).

-------
55           53.34  (a)(l)(ii)           Add at the end of the paragraph the change
                                        recommended for page 40, fl 53.31 (a)(2).

55           53'34  (a)(l)(iii)          Add after "specifications" on line 14 the
                                        change developed for page 41, H 53. Jl (a) (3) .
                                        Delete the last sentence of the paragraph.
                                        The reference method is a manual method
                                        requiring a 24-hour integrated sample.   It
                                        is virtually impossible to introduce control
                                        samples in the collected sample itself.

             53.34  (a)(2)               Revise the last two sentences of the paragraph
                                        to read, "The average response of the auto-
                                        mated method over the 1-hour span corres-
                                        ponding to the 1-hour reference sample shall
                                        be determined by integration of the continuous
                                        signal output or by the average of at least
                                        12 instantaneous signal levels at equally
                                        spaced intervals over the 1-hour sampling
                                        period.  In like manner, the 24-hour average
                                        response of the automated method is obtained
                                        by integrating at least 22 hours of the
                                        continuous output signal during the 24-hour
                                        measurement."  The average over 22-hours
                                        is recommended since, some time is needed to
                                        introduce control samples and when waiting
                                        for the system to come to equilibrium when
                                        the step changes in pollutant concentration
                                        are made.  A full 24-hour average means that
                                        the operational period will be longer than
                                        the 72 hours.

58             53.34  (b)(2)             Delete the first two sentences.  The content
                                        is a repeat of what is discussed in the
                                        following paragraph.

62             53.34  (d)(3)             Delete the first sentence.  It is superfluous.
                                 - 59 -

-------
                              TABLE IV




               Changes Authorized or Requested by EPA









SECTION                       	CHANGES	
11 53.21 (a)      increase the allowable fluctuation of the environment




                around the gas generation system from ± 1°C to ± 2°C.









53.22(f)(5)(i)  delete the clause  "equal to twice the noise" in the




                technical definition of lag time to read, "the time




                interval between a step change in input concentration




                at the instrument  inlet to the first observable




                corresponding  change."  Although this change will make




                the definition consistent with the  test procedure, in




                our opinion,  this  will  not  give  a  true value for lag




                time because  of the difficulty in  discerning whether




                the first observable change in  response  is  due to the




                introduction  of pollutant  or due to analyzer noise.  We




                would still recommend that  the  directions in  the  test




                procedure and  not  the definition be revised.   See  5.2.9,




                Item (a) for  additional comments.









 fl 53.30(a)      delete the Priority Region  1 site requirement for con-




                ducting  the consistent  relationship tests.   This  will




                eliminate the  necessity for testing at sites  which  may




                be several hundred miles away from the manufacturer's




                location and may prove  to be prohibitively expensive.
                                 - 60  -

-------
II 53.30(a)      increase the maximum allowable volume dilution of the




                ambient air sample from 20 to 40% when spiking to




                obtain the desired pollutant concentration for con-




                sistent relationship tests.









U 53.34(a)(2)   In the consistent relationship test  for S02, reduce the




                3-hour measurements  to 1-hour for the tests at the 0.40




                to 0.50 ppm range.








                The  1-hour requirement appears inconsistent with the




                federal air quality  standard for S02, which calls  for




                3-hour  average.
                                - 61 -

-------
                                                       TABLE V

                                 EVALUATION  OF  SUBPART B USING S02 ANALYZERS  CPHASE I)
PROCEDURE TESTED
ANALYZER USED
                                           SPECIFICATION
TEST RESULT     PASSED OR FAILED
STATUS OF PROCEDURE
Calibration
Range
Noise, 0%
Noise, 80%
i
ro
( Lower Detectable Limit
Lag Time
Rise Time
Fall Time
Precision, 20%
Precision, 80%
(T)
(M)
(T)
(M)
(T)
CM)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)


0.005 ppm
0.005 ppm
0.005 ppm
0.005 ppm
0.02 ppm
0.02 ppm
20 min
20 min
15 min
15 min
15 min
15 min
0.02 ppm
0.02 ppm
0.03 ppm
0.03 ppm
See Fig. 3
See Figs. 4 &
0-1.0 ppm
0-1.0 ppm
<0.01 ppm
<0.005 ppm
0.008
<0.005
0.01 ppm
0.01 ppm
15.57 min
0.05 min
24.87 min
0.16 min
26.68 min
2.89 min
NC
0.003 ppm
NC
0.008 ppm
5

Failed
Passed
Failed
Passed
Passed
Passed
Passed
Passed
Failed
Passed
Failed
Passed
Passed
Passed
                                                                                                      Improvements are
                                                                                                      needed.  See 5.2.3

                                                                                                      No change needed
                                                                                                      Modifications are
                                                                                                      needed.  See 5.2.4

                                                                                                      Specification missing
                                                                                                      In Regulations.

                                                                                                      Improvements are
                                                                                                      needed.  See H 5.2.5

                                                                                                      Change needed
                                                                                                      Improvement needed.
                                                                                                      See I 5.2.7

                                                                                                      Improvement needed.
                                                                                                      See 5.2.7

                                                                                                      Modifications needed.
                                                                                                      See 5.2.7

                                                                                                      Modifications needed.
                                                                                                      See 5.2.7

-------
 Interference Equivalent:
                           Improvements needed.
                           See 5.2.2 and 5.2,6
                           of this report
HC1
NH3
H2S
N02
i
-» C02
1
03
H2C
CO
Total Interferent
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
(T)
(M)
± 0.02 pprc <0.02
NR
± 0.02 ppm <0.02
NR
± 0.02 ppm <0.02
+ 0.045
± 0.02 ppm <0.02
NR
± 0.02 ppm <0.02
0.065
± 0.02 ppm <0.02
NR
± 0.02 ppm NR
<0.02
± 0.02 ppm NR
<0.02
0.06 0.045
0.06 0.114
Passed
Passed
Passed
Failed
Passed
Passed
Failed
Passed
Passed
Passed
Passed
Failed
                                                                                                       Collected samples nee<
                                                                                                       to be neutralized be-
                                                                                                       fore analysis.  See
                                                                                                       Item (d) under dis-
                                                                                                       cussion in 5.2.2
                                                                                                       Higher concentration
                                                                                                       of stock C02 needed.
                                                                                                       See Item (c) under di
                                                                                                       cussion in 5.2.2
                                                                                                       Specified concentrati
                                                                                                       of H20 impractical.
                                                                                                       See Item (g) under di
                                                                                                       cussion in 5.2.2

                                                                                                       Higher concentration
                                                                                                       of stock CO needed.
                                                                                                       See Item (d) under di
                                                                                                       cussion in 5.2.2
NOTE:  (T) - Technicon Air Monitor IV; (M) - Melpar sulfur analyzer; (NCi -
Analyzer response times did not permit completic
in time allotted; (NR) - not required.

-------
                         TABLE VI

EVALUATION OF SUBPART B FOR ZERO AND SPAN DRIFTS USING THE
          MELPAR TOTAL SULFUR ANALYZER - PHASE I

   (Specification for zero and spau drifts = ± 0.02 ppm)

INPUT
S02
PROCEDURE TESTED (ppm)
24-hour at zero 0







12-hour at zero 0







24-hour span 0.20
20% of full scale






24-hour span at 0.77
80% of scale






TEST
PERIOD
DAY
#
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
0
I
2
3
4
5
6
7
CHANGE
IN
RESPONSE
(ppm)
:i [TI 	
0
0
0
0
0
0
0
*..._..
0
0
0
0
0
0
0
_ —
-0.01
+0.01
-0.02
40.03
-0.01
-0.03
+0.04
____
-0.02
+0.0
+0.01
___«
-0.04
-0.04
+0.04
RESULTS
(PASSED
OR
FAILED)
«.«.».«
Passed
Passed
Passed
Passed
Passed
Passed
Passed
_»_*.
Passed
Passed
Passed
Passed
Passed
Passed
Passed
— —
Passed
Passed
Passed
Failed
Passed
Failed
Failed
	 _
Passed
Passed
Passed
«.-»._
Failed
Failed
Failed
                                                       COMMENTS

                                                   Instrument passed for
                                                   this specification.
                                                   Instrument passed for
                                                   this specification.
                                                   Instrument failed this
                                                   specification.
                                                   Instrument failed this
                                                   specification.

-------
                              TABLE VII

     EVALUATION OF SUBPART B FOR ZERO AND SPAN DRIFTS USING THE
                  TECHNICON S02 ANALYZER - PHASE I

        (Specification for zero and span drifts = ± 0.02 ppm)
24-hour zero
   test
12-hour zero
   test
24-hour span at
   20% of scale
 24-hour  span  at
    80% of  scale

INPUT
S02
(ppm)
0.0







0.0







0.21







0.83







TEST
PERIOD
DAY
#
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
CHANGE
IN
RESPONSE
(ppm)
	
0.0
+0.035
-0.01
0.0
+0.01
-0.01
+0.05
— _——
0.05
0.015
0.03
0.07
0.02
0.0
0.04
____
0.0
0.0
+0.04
-0.02
+0.05
-0.01
+0.04
	
+0.005
+0.015
+0.03
-0.01
+0.02
+0.005
0.0
Passed
Failed
Passed

Passed
Passed
Passed
Failed
Failed
Passed
Failed

Failed
Passed
Passed
Failed
Passed
Passed
Failed

Passed
Failed
Passed
Failed
 Passed
 Passed
 Failed

 Passed
 Passed
 Passed
 Passed
    COMMENTS

A second series of !
tests is indicated.
Instrument failed
this specification.
 Instrument  failed
 this specification.
 A second series of  8
 tests  is indicated.
                                 - 65  -

-------
                                                         TABLE  VIII




                        EVALUATION  OF  IMPROVED  SUBPART  fi(a)  USING THE  BENDIX OZONE MONITOR (PHASE II)
ON
Procedure (a) Tested
Calibration
Range
Noise, 0% of Range
, 80% of Range
Lower Detectable Limit
Interference Equivalent
H2S
(X>2
H20
TOTAL
Zero Drift (*>), 12-Hr:
Day 1
2
3
4
5
Specification

0.0-0.5 ppm
0.005 ppm
0.01 ppm
0.01 ppm

±0.02 ppm
±0.02 ppm
±0.02 ppm
0.06 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
Test
Result

0.0-0.5 ppm
0.0 ppm
0.0 ppm
0.01 ppm

0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
Passed/ Status of Procedure^
Failed
Procedure O.K.
Passed
Passed Procedure O.K.
Passed
Passed Procedure O.K.
Procedure O.K.
Passed
Passed
Passed
Passed
Procedure O.K.
Passed
Passed
Passed
Passed
Passed

-------
TABLE VIII (cont'd)
Procedure Tested Specifications
6 ±0.02 ppm
7 ±0.02 ppm
Zero Drift (b) , 24-Hr
Day 1 ±0.02 ppm
2 ±0.02 ppm
, 3 ±0.02 ppm
~^ 4 ±0.02 ppm
i
5 ±0.02 ppm
6 ±0.02 ppm
7 ±0.02 ppm
Span Drift(b), 80% of Range
Day 1 ±0.02 ppm
2
3
4
5
6
7
Test
Result
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.02 ppm
0.00 ppm
0.01 ppm
0.01 ppm
0.01 ppm
Passed/ Status of Procedure
Failed
Passed
Passed
Passed Procedure O.K.
Passed
Passed
Passed
Passed
Passed
Passed
Passed Procedure O.K.
Passed
Passed
Passed
Passed
Passed
Passed

-------
TABLE VIII (cont'd)
Procedure Tested
Lag Time(b) :
Day 1
2
3
4
5
ON ,
CD 6
7
Lag Time(b):
Day 1
2
3
4
5
6
7
Rise Time'"' :
Day 1
2
Specification

1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees

1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees
1200 Sees

900 Sees
900 Sees
Test
Result

3 Sees
4 Sees
3 Sees
3 Sees
3 Sees
3 Sees
3 Sees

3 Sees
4 Sees
3 Sees
3 Sees
3 Sees
3 Sees
3 Sees

1-j Sees
31 Sers
Passed/ Status of Procedure
Failed

Passed Procedure O.K.
Passed
Passed
Passed
Passed
Passed
Passed

Passed Procedure O.K.
Passed
Passed
Passed
Passed
Passed
Passed

Passed "r~>cedure O.K.
Passed

-------
TABLE VIII (cont'd)
Procedure Tested
3
4
5
6
7
1 Fall Time^k) :
vo Day 1
2
3
4
5
6
7
PrecisionOO , 
-------
TABLE VIII (cont'd)
Procedure Tested
4
5
6
7
( Precision (b) , @ 80%
of Range
o
i Day 1
2
3
4
5
6
7
Specification
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm

±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
±0.02 ppm
Test
Result
0.0 ppm
0 . 0 ppm
0.01 ppm
0.01 ppm

0.0 ppm
0 . 0 ppm
0 . 0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
0.0 ppm
Passed/ Status of Procedure
Failed
Passed
Passed
Passed
Passed
Procedure O.K.
Passed
Passed
Passed
Passed
Passed
Passed
Passed

-------
Day
                                                       TABLE IX

                                  Suggested Zero and Span Drift Test  Plana:  Table B-4
Morning
Every Night6 Adjust Test
	Conditions  to:	

  Line      Room
Voltagec   Temp,°Cd  Day
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Adjust zero and Span 1 (80 ± 5% of range). Sample Span 2
(20 ± 5% of range). Record the stable readings. No in-
strument adjustments permitted hereafter until after Day
3 readings .
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2, 24 hours after Day 0 readings.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift, sample 0, Span 1 and
Span 2; adjustments may be made to the instrument, if
necessary, after readings are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2; instrument adjustments permitted after readings
are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2b
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1,
Sj>an 2; adjustments permitted after readings are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span I and
Span 2; adjustments permitted after readings are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2; End of test
125
105
125
105
125
105
125
105
125
105
125
105
125
105
125
.
20
20
30
30
20
20
30
30
20
20
30
30
20
20
30
-
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
aThese  tests  shall  be  run only once each day.  Tests shall also be run either on  consecutive days
 with adjustments permitted every three days or in increments of three consecutive days during
 which  adjustments  to  the instrument are not permitted.  In the afternoon, other  tests not
 requiring  adjustments to the instrument may be performed, i.e., for rise time, fall  time, lag
 time and precision,

bAt this time examine  test results to determine if further tests are required.

Voltage specified  shall be controlled to ± 1 volt.

^Temperature  specified shall be controlled to ± 2°C.

eSample zero  air continuously for a minimum of 12-hours.
                                                  - 71 -

-------
                              TABLE X

                      EVALUATION OF SUBPAKT C
                  FOR OXIDANT METHODS:  MANUAL KI
                   VS BENDIX ANALYZER - PHASE I
TEST INPUT 03
PERIOD LEVEL
DAY 1 LOW
MED
HIGH
DAY 2 LOW
MED
HIGH
DAY 3 LOW
MED
HIGH
DAY 4 LOW
MED
HIGH
DAY 5 LOW
MED
HIGH
OUTPUT-ppm 03 DIFF.
REFERENCED) BENDLX^b) (a)-(b)
PPM 03
0.090
0.22
0.41
0.069
0.21
0.40
0.082
0.23
0.38
0.077
0.21
0.41
0.066
	
0.38
0.088
0.23
0.40
0.071
0.22
0.40
0.087
0.24
0
0.082
0.21
0.42
0.065
	
0.40
-0.002
+0.01
0.0
+0.002
+0.01
0.0
+0.005
+0.01
+0.02
+0.005
0.0
+0.01
-0.001
	
+0.02
TOLERANCE
PPM 03
±0.02
±0.03
±0.04
±0.02
±0.03
±0.04
±0.02
±0.03
±0.04
±0.02
±0.03
±0.04
±0.02
±0.03
±0.04
PASSED/
FAILED
PASSED
PASSED
PASSED
PASSED
PA^S^l'
TOSSED
PASSED
PASSED
PASSED
PASSED
PASSED
PASSED
PASSED
	
	
NOTE:  (a) = Manual KI method, (b) = Bendix 03 Monitor

-------
                             TABLE XI

                      EVALUATION OF SUBPART C
             FOR SULFUR DIOXIDE METHODS:  p-ROSANILINE
            MANUAL METHOD VS MELPAR ANALYZER - PHASE II
TEST
PERIOD
DAY 1
2
3
4
5
6
7
DAY 1
1
1
4
4
7
7
INPUT S02 OUTPUT-ppm S0£ DIFF.
LEVEL REFERENCED MELPAR (a)-(b)
PPM S02
HIGH
LOW
MED
HIGH
MED
LOW
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
0.37
0
0
0
0
0
0
0
0
0
0
0
0
0
.027
.11
.39
.14
.04
.41
.35
.37
.37
.39
.39
.42
.41
0,
0
0
0
0
0
0
0
0
0
0
0
0
0
.40
.03
.14
.43
.15
.04
.64
.41
.44
.43
.43
.43
.66
.68
+0.03
+0.
+0,
+0,
-K).
0
+0
40
+0
+0
+0
+0
+0
+0
,003
.03
.04
.01
.0
.23
.06
.07
.06
.04
.04
.24
.27
TOLERANCE
PPM S02
±0.04
±0.02
±0,
±0,
±0
±0
±0
±0
±0
±0
±0
±0
±0
±0
.03
,04
.03
.02
.04
.04
.04
.04
.04
.04
.04
.04
PASSED/
FAILED
PASSED
PASSED
PASSED
PASSED
PASSED
PASSED
FAILED
FAILED
FAILED
FAILED
PASSED
PASSED
FAILED
FAILED
NOTE:  (a) Manual p-rosaniline, (b) Melpar Analyzer
                                - 73  -

-------
                             TABLE XII

                      EVALUATION OF SUBPART C
                FOR SULFUR DIOXIDE METHODS:  MANUAL
        p-ROSANLINE METHOD VS TECHNICON ANALYZER - PHASE II
TEST
PERIOD
DAY 1
2
3
4
5
6
7
DAY 1
1
1
4
4
7
7
INPUT S02 OUTPUT-ppm S02 DIFF.
LEVEL REFERENCE (a)TECHNICON(b) (a)-(b)
PPM S02
HIGH
LOW
MED
HIGH
MED
LOW
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
HIGH
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.37
.0
.11
.39
.14
.041
.41
.35
.37
.37
.37
.37
.42
.41
0.40
0
0
0
0
0
0
0
0
0
0
0
-
0
.051
.14
.39
.12
.032
.44
.39
.40
.40
.40
.40
.42
.44
+0.03
-0.024
0
0
.03
.0
-0.02
-0
40
+0
+0
.009
.03
.04
.03
+0.03
+0
+0
0
+0
.03
.03
.0
.03
TOLERANCE
PPM S02
±0.04
±0.
±0
±0
±0
±0
±0
±0
±0
±0
±0
±0
±0
±0
.02
.03
.04
.03
.02
.04
.04
.04
.04
.04
.04
.04
.04
PASSED/
FAILED
PASSED
PASSED
PASSED
PASSED
PASS' '•'
PA_,bED
PASSED
PASSE*)
PASSED
PASSED
PASSED
PASSED
PASSED
PASSED
NOTE:  (a) Manual p-rosaniline method, (b) Technicon Air Monitor

-------
                 TABLE XIII
Times to Conduct Test Procedures in Subpart  B
               for Performance
Task or Procedure


1.  Prepare and calibrate
    test analyzers

2.  Noise, at 0%^

3.         at 80%(a)

4.  Lower Detectable Limit(a

5.  Zero Drift

6.  Span Drift, at  20%

7.             , at  80%
        Subtotal


8.  Lag Time

9.  Rise Time

10.  Fall Time
         Subtotal


11.   Precision,  at 20%

12.            ,  at 80%
         Subtotal


13.   Interference Equivalent

             HC1

             NH3

             H2S

             NO 2

             C02

              03
                                    Time in Man-Hours
                (a)
Technicon
80
21
28
56
98(b)
28
28
241
7
28
28
63
224
224
448
140
140
140
84
84
112
Melpar
64
14
28
28
98Cb)
28
28
190
4
7
7
18
11
11
22
-
-
84
-
-
_
Bendix
16
14
14
18
980>>
18
18
98
4
7
7
18
11
11
22
-
-
84
84
84
-
                     - 75 -

-------
                          TABLE XIII (con't)

             Times to Conduct Test Procedures  in  Subpart B
                            for Performance
 Task or Procedure
             H20

             CO
         Subtotal
   Time in Man-Hours
                                           Technicon  Melpar   Bendix
700
 84

 84
252
112
364
14.  Total Man-Hours
      (1st test set)

15.  Net Man-Hours
      (14 - 8, 9, 10)

16.  Conduct 2nd test set
1452
482
1389
464
502
484
1496
     (c)
457
    (c)
  (a)  Time for seven trials extrapolated from results of one trial.

  (b)  Unattended operation; data not included in totals.

  (c)  Net Man-Hours less Item 1.
                                -  76  -

-------
                              TABLE XIV

            Times to Conduct Test Procedure in Subpart C
                     for Consistent Relationship
    Task
1.  Design, construct and
    install sampling probe
    and distribution manifold

2.  Prepare and calibrate
    reference and candidate
    methods or analyzers

3.  Conduct initial test set
    and reduce data.
                Subtotal (2 + 3)

4.  Conduct second test set
    and reduce data.
                    Total
                                                Time in Man-Hours
S02

 80



 80
154
 72
306
 03

 80



 40



 64

104
 72
256
                                                                CPU)
 80
 40
104
 72
256
          extrapolated from evaluation of procedure for oxidant (03)  methods,
                                  -  77  -

-------
                              TABLE XV

               Overall Times to Determine Equivalency
    Task
1.  Design, fabricate and
    install gas generation system

2.  Design, fabricate and
    install sampling probe
    and manifold.

3.  Conduct initial test set

    (a)  Performance
    (b)  Consistent Relationship
                Subtotal

4.  Conduct second test set

    (a)  Performance
    (b)  Consistent Relationship
                Subtotal

5.  Total  (3 + 4)
	Time in Man-Hours	
Technicon   Melpar   Bendix

   176       176      176
    80
  80
  80
   1389
    154
   1543
   1496
   	72
   1568

   3111
 464
 104
 568
 457
  72
 529
1197
 484
 104
 588
 535
  72
 607
1195
                                 - 78  -

-------
                                          Figure  Is
                                                       GAS GENERATION SYSTEM
                            TH
                                            TH
                                                          Test Atmosphere Z
      FM




1

J ,, 	
1
1 MF
O O 1
TH MF
^N
fi ~ n n

^ I i

I
, —
PP



,

j ' •





' v v v v v v
MFM
-O
vo
SCR
                Cylinders of Zero Air
                                                                                             To Vent
                                                                                                      LEGEND













Z


j
e






1

&






p


i
' 	 (,









\



> Rd


1
t
\




i








I


i /
IMT


. A
, B





^



^o
''r* - i •T**
1 Cn ' '

^
i











L
PP


i

v
1






i



SV
¥
H' 1
) TTH (I
Hf
A



- LT




^



L




T L
A








;

_/
<
i
/


^




pl Sampling Ports
MFM - mass flowmeter
FM - flowmeter
MF - mixing flask

\
TH - three-way-valv
MT - MFM transducer
NV - needle valve
SV - shut-off valve
SCR - strip-chart re<
• A
i B
' U
t


^ /

^ (
TH


^
¥ I







"\
^__
,


^
JT
Y
TH j
T
A






L






1



PH - permeation tub«

y






















K
A,

- regulator
B - connections foi
generator
PP - auxilliary pol]






R(
¥ I
NV (
T '

>5c

)
-3~
A A















5 R(



&


•=r
^












)























TH
a\

FM fl
_<£ 	

^"^














r^-CD —
NV

> O
" "^fv
Ls V


PH











I
t
j



t
1

j
^



!i>
/

y




i |
i \

A
•j


\

' B,



\
r
L
^
^
» A<



test attuosphei


To vent

") TH






A«_ _ _ _.

B»-~ -e — .,
„,
PH NV i
(^i
,_9
1
t
                                                                                                                63
                                                                                                                     I
                                                                                                                     I
                                                                                                                     I

                                                                                                                     A
                                                                       Constant Temperature Bath

                                                                      and Permeation Tube Systems
                                                                                                              Generator

-------
                                                Figure 2:   Gas Sampling Manifold
                                                                20"
                                                                t
                                                                2"

                                                                I
                                                                                                 "-4* 1"'
                                                                             Materials:  glass,  Pyrex, heavy well
03
o
                                              12/5
O
                                                                                                        !:RD 02017

-------
FIG* 3:  Dyna ic Calibration Curve of the Technicon Air Monitor IV for SOg
                                 - 81 -

-------
FIG. 4;  Dynaaic Calibration Curve for the Melpar Total Sulfur Analyzer^
                                  -  82  -

-------
Calibration Curvefor the Melpar Total  Sulfur Analyzer. Model  LL-1100-1B
                    ~~~   T— -• —	••*-'—•	I—•—•——r- -i—.—,	-"	1—J"T—| | |   •  •-.-;  —*—. i
                         - 83  -

-------
 FIG.  6;  Dynamic Calibration Curve  for the  Bendix Ozone Monitor
i—i  • i '—:—•	' * r i—' T T  * i T i  i—I—\—i '  • • i—•—T—;	1—i—!—r- —i—i—•—r—'—1~*—i—'—i—i—i—•—i—'—r—•—
                                 - ak -

-------
                  FRIDAY, OCTOBER 12, 1973
                  WASHINGTON. D.C.

                  Volume 38 • Number 197

                  PART II
                   ENVIRONMENTAL
                      PROTECTION
                        AGENCY
                    AMBIENT AIR MONITORING
                   EQUIVALENT AND REFERENCE
                          METHODS
No. 197—Pt. n	1
                  85

-------
28438

  ENVIRONMENTAL  PROTECTION
              AGENCY

         [ 40 CFR Parts 50, 51 ]
AMBIENT AIR MONITORING EQUIVALENT
      AND REFERENCE  METHODS
      Notice of Proposed Rulemaking
  Notice is hereby  piven  that  the  En-
vironmental Protection  Agency Is  con-
sidering amendments to Parts SO and 51
of Title 40, Code of Federal Regulations,
as set forth below.
  Elsewhere in this  issue of the FEDERAL
REGISTER, EPA is proposing the addition
of a new Part 53, entitled "Ambient Air
Monitoring  Equivalent  and  Reference
Methods," to Title 40 of the Code of Fed-
eral Regulations. As more fully described
Jn the Notice of Proposed Rulemakins;
concerning the  proposed  new  Part 53,
that Part would establish definitive re-
quirements and procedures applicable to
d>. terminations   whether   methods of
sai-ipling and analyzing the ambient air
may be designated "equivalent methods"
or "reference methods" for the measure-
ment of specified air pollutants. It is in-
tended that such determinations be gov-
erned In the future by the provisions of
the proposed Part 53 as finally promul-
gated. The amendments proposed in this
Notice  would revise existing provisions
pertinent to such  determinations  and
would,  to effect, provide that for  pur-
poses of Parts 50 and 51 such determina-
tions will be governed In the future by
the provisions of Part 53.
  Interested  persons  may submit  writ-
ten  comments on the proposed amend-
ments in triplicate  to the Quality Assur-
ance  and  .Environmental  Monitoring
Laboratory.  United   States  Environ-
mental  Protection  Agency,  National
Environmental  Research  Center,   Re-
search  Triangle Park, North Carolina
27711. All relevant comments  postmarked
not later than 45 days after  publication
of this notice will be considered. All  com-
ments  will be  available for public in-
spection  during normal business hours
at the address specified above.
  This notice of proposed rulemaklng is
Issued  under the  authority  of sections
109 and 301  of the Clean  Air Act, as
amended (42 C.S.C. 1857C-4, 1857g(a)).

  Dated September 26,1973.
                  JOHN QUARLES,
             Acting Administrator,
     Environmental  Protection Agency.

   Part 50 of Title 40 of the Code of Fed-
eral Regulation*  is  proposed  to be
ariended as follows:
   rci $ 50.1, paragraphs (f) and (g) are
revved to read  as follows:
 § 50.1   Definitions.
     •      •      •       •       •
   (f)  "Reference   method"   means  a
method of sampling  and analyzing the
 ambient  air  for an air pollutant as de-
 scribed in an appendix to this part or a
 method that has beer, designated  as  a
 reference method  in accordance  with
part 53 of this chapter.
   (g) "Equivalent  method"  means any
 method pf sampling  and analyzing the
          PROPOSED RULES

ambient air for an air pollutant that has
been designated as an equivalent method
In accordance with part 53 of this chap-
ter.
  Part 51  of Title 40 of the Code of Fed-
eral   Regulations  is  proposed   to  be
amended  as follows:
  1.  In 5 51.14. the table  in subpara-
graph (1)  of paragraph (e) Is  revised
by revising footnote 1 to read as follows:
§ 51.1-1  Control  Slralpjry: Carbon mon-
     oxide, lijtlrwarixms pliotorlicniical
     oxidunts, and nitrogen ilioxide.
     *      »      •      »      •
  (e) * *  *
  (1) • *  *
  1. Other methods ot measurement will be
considered equivaient methods if they have
been designated equivalent methods in ac-
cordance with part 53 of this chapter.

  2.  In f 51.17, the table  In sub-para-
graph (1) of paragraph (a) is rev::^U by
revoking footnotes d, e.  and f, by revok-
ing the table of  performance specifica-
tions  for  pollutants   sulfur  dioxide,
carbon monoxide and photochemical oxi-
dant (corrected for NO- and SCO  ap-
pearing  after  the colon in the  second
sentence  of footnote 1, and by revising
the second sentence of footnote 1 to read
as follows:
§51.17   Air quality surveillance.
   (a)(l)  • *  '
     *      »      •      »      *
   1  • •  • Other methods of measure-
ment will be considered equivalent meth-
ods tf they have been designated equiv-
alent methods In accordance with part
 53 of this chapter.
     »      «      •      •      •
  [FR Doc.73-21338 Filed 10-11-73:8:45 am]
            [ 40 CFR Part 53 ]
 AMBIENT AIR MONITORING EQUIVALENT
      AND  REFERENCE METHODS
      Notice of Proposed Rulemaking
   Notice  is hereby given that the En-
 vironmental Protection Agency Is con-
 sidering the addition of a new Part 53
 to Title 40 of the Code of Federal Regu-
 lations, as set forth below. The proposed
 new Part 53 would establish requirements
 and procedures applicable to determina-
 tions whether,  for  purposes  described
 more fully  below, methods of sampling
 and analyzing the ambient air may be
 designated "equivalent methods" or "ref-
 erence methods" for the measurement of
 specified air pollutants.
   As required by section 109 of the Clean
 Air Act. as amended December 31. 1970
 (Public Law 91-604). the Administrator
 of the Environmental Protection Agency
 published on January  30, 1971 (36 FR
 1502) and March 30,1971 ((36 FR5S67)),
 proposed national primary and secondary
 ambient air quality standards for six pol-
 lutants arid, after consideration of com-
 ments  received, promulgated  the same
 on April  30, 1971, 36 FR 8186. The  na-
 tional ambient air quality standards are
 now codified at 40 CFR Part 50. In addi-
 tion to prescribing  ambient air quality
standards, the Administrator published
certain te^t procedures to be utilized by
the  Environmental  Protection Agency
and State and local agencies to measure
the levels of ambient air quality for each
of the six pollutants. The measurement
procedures are described in detail  in ap
pendices to Part 50 ami are defined in -1C
CFR 50.1 (f) as "reference  methods."
  The ambient air quality standards, as
prescribed, provide that measurements
are to be made by the appropriate refer-
ence  method  or   by  an  "equivalent
method." "Equivalent method" is defined
in 40 CFR 50.1ig>  as "any method of
sampling and analyzing for an air pol-
lutant which  can be demonstrated to trie-
Administrator's  satisfaction to have a
consistent relationship  to the  reference
method." 40 CFR 51.17, discussed.
below, contains  additional  requirements
applicable  to a determination that a
method Is an "equivalent method." In
addition to the requirement that  a con-
sistent  relationship  to the  reference
method be shown,  a candidate method
is required under  § 51.17(aHl) to meet
certain performance specifications de-
scribed in that section.
   Within 9 months after  promulgation
of the National Ambient  Air  Qur.Iir"
Standards, i.e.. no later  than January 3G
1972, each State was required by Seer' •• .
 110 of the Act to adopt and submii to the
Administrator a plan which provided for
the  Implementation, maintenance, ana
enforcement of  the national ambient air
quality standards within each air quality
control region tor portion thereof) witliin
 the  State. The Act requires the Adminis-
trator to approve  an  imDlemeutatioii
 plan, or any portion thereof, if he deter-
mines that the plan (or portion thereof)
 was adopted after reasonable notice a; -I
heating and that it satisfies detailed cri-
 teria set forth in § 110fa) (2) (A)- ox
the  Act. To  assist the States In the de-
velopment of the implementation p;ans,
and  to implement the  requirements of
 5 110. EPA proposed and promulgated
regulations  entitled "Regulations  for
Preparation, Adoption, and Submittal oi
Implementation Plans." See 40 CFR Part
51 (36 FR 13486).
   The purpose of the proposed new Part
53 is to assist State and  local govern-
ments and  ambient  air monitoring In-
strument manufacturers with  respect to
one  of tlie requirements applicable to ap-
proval oi implementation plans. Section
 llOfa) (2) 
-------
  State to  monitor the  ambient  air  be
  either the appropriate, reference method
  or a method that if. equivalent to the
  reference method. The proposed new Part
  53 would establish  definitive require-
  ments and procedures by which Suites or
  instrument  mar.uiaoturers  may  have
  their monitoring mt-Uiods determined to
  be  equivalent  nietliods  or  reference
  methods.
    As indicated  previously. 40 CFR 50.1
    and  51.17'a>< 1>  presently  contain
  proviiiuns poriincnt, to  deter:::ir.auons
  whether mor.itorin; methods are o.iuiva-
  lent methods, and iO CFR 50.Kit  defines
  specified methods as  reference methods.
  In addition.  rqu:v-/.ent methods are re-
  ferred to in 40 CrH, 51.14(e) Q). It, is in-
  tended   that determinations  whether
  candidate  methods are equivalent meth-
  ods or reference methods be governed in
  the future fay the provisions of Part 53 as
  finally promulgated. Accordingly.  EPA is
  proposing  elsewhere in this issue of the
  FEDrRAL  REGISTER appropriate  amend-
  ments to 40  CFR 50.1 (f), 50.1 (g), 51.14
  (e)(D.and51.17(a)(l).
    It is  recognized that some State  or
  local  governments may have  purchased
  monitoring instruments or systems  that
  purport to b« reference methods or equiv-
  alent methods (or elements thereof)  in
  accordance with  the  existing  provisions
  of 40 CFR 50.1(f), 50.1 (g). and 51.17(a)
  (1) but which -will not conform  to the
  requirements of  Part  53  if it is  promul-
  gated as proposed. Because purchase of
  such equipment may have been relatively
  costly, EPA is considering allowing: its
  use on some basis for purposes of 40 CFR
  51.17(a) (1), provided that the equipment
  was purchased prior to publication  of this
 notice in the FEDERAL REGISTER. For ex-
  ample, use of such equipment might be
  permitted,  on a  showing satisfactory to
  the Administrator that  the equipment
 meets  the pertinent  existing require-
 ments of 40  CFR SO.l(f), 50.1 (g),  and
  51.17(a)(l), for a period that would al-
 low a reasonable time for replacement of
 the equipment with new equipment satis-
 fying the requirements of Part 53 as fi-
 nally promulgated. State and  local gov-
 ernments  and other  interested parties
 are specifically invited  to comment  on
 this proposal, to suggest other  courses of
 action that EPA misrht follow in this re-
 gard, and to submit any information that
 they  consider pertinent to the proposal
 or to  alternative courses of action.  In
 particular.  State  or  local  governments
 that have purchased instruments or sys-
 tems purporting to be reference methods
 or equivalent methods   for  elements
 thereof) in accordance with 40 CFR 50.1
 (f), SO.Kg), and 51.17(a)(l) are invited
 to comment on the number,  type, and
«t»st of the Instruments or systems they
 have purchased, the possibility that such
 equipment might conform to the require-
 ments of the proposed Part 53 If it is pro-
 mulgated as proposed, and the possibility
 that nonconforming equipment might be
 replaced with  conforming equipment in
 time to permit compliance with 40 CFR
 51.17(a) (1) within the time specified to
 that section.
           PROPOSED  RULES

   The proposed new Part 53  is divided
 into  three  subparti,  the  contents  of
 which may be surr-nuri.:ed as follows.
   Subpart A  contains the general re-
 quirements to be satisfied for an equiva-
 lent or reierence method determination
 for both automated and manual methods.
 It specifies  the  descriptive information
 and affidavits, that  must be  submitted
 whenever such  a  determination is re-
 quested. The primary responsibility for
 r.n  equivalent or reference method de-
 twi-ininaiion  r^i-s with tiie  applicant.
 Candidate automated methods that pur-
 port to .measure sulfur dioxide, carbon
 monoxide, br  photochemical  oxidants
 r.:i:st satisfy the performance specifica«;
 lions described in Subpart 3 and demon-
 strate  a consistent  relationship  as re-
 quired by Subpart C. Candidate manual
 methods need only demonstrate a  con-
 sistent relationship as required by Sub-
 part C.
   Subpart A  also  provides  that  any
 method determined to be an equivalent
 or  reference method  will be officially
 designated  as  such  in  the FEDERAL
 REGISTER. Any applicant whose method
 is rejected may  appeal the rejection by
 requesting reconsideration. Designations
 of  methods may  be .cancelled if the
 methods are subsequently found not  to
 be in compliance with the provisions  of
 Part 53. Prior to cancellation, the manu-
 facturer of the method in question will
 receive written notice  of the facts  that
 the  Administrator believes warrant can-
 cellation and an opportunity to demon-
 strate or come into compliance. In ad-
 dition,  a hearing may be granted.  The
 above procedures will assist the 'Agency
 in obtaining reliable data and in making
 fair and sound judgments when grant-
 ing, rejecting, or canceling a method.
   Subpart B of the proposed Part 53
 specifies   performance  characteristics
 that must be met by automated methods
 with respect to  interference, lower de-
 tectable limit, precision,  rise time,  fall
 time, zero drift, span drift,  lag time,
 and noise.
   Subpart C contains  the general pro-
 visions and test procedures necessary for
 demonstrating a consistent relationship
 between  candidate  methods,   whether
 automated  or  manual, and  reference
 methods.
  EPA recognizes that  for carbon mon-
 oxide, and photochemical oxidants, each
 reference method designated in 40 CFR
 Part 50 is a combination of  a refer-
 ence principle and  a calibration proce-
 dure and not a  single specific method.
 Therefore, it is possible for various in-
 strument  manufacturers  to  construct
 different air monitoring equipment based
 on  different adaptations  of the  same
 reference measurement principle.  If an
 applicant's candidate automated method
 utilizes  the appropriate reference  prin-
 ciple and calibration  procedures cor-
rectly and his instrument satisfies an
the performance  characteristics of Sub-
part B, his method may be determined
to be a reference method.
  It  is the Intent of EPA  to encourage
the advancement of the art of monitor-
                                 28139

 ing  pollutants in ambient air.  If  the
 Administrator  becomes convinced  that
 a "candidate  method  has  the necessary
 measurment capability and is superior to
 the  existing  reference method, he will
 consider naming it the ret wc;ue method.
 In this manner, EPA intends to keep the
 best available analytical methodology ns
 the reference method.
   Interested persons may submit written
 comments  on the proposed regulation*
 in triplicate  to the Quality Assurar-.ve
 and  Environmental Monitoring Labora-
 tory, United States Environmental  Pro-
 tection Agency. National Environmental
 Research   Center,  Research  Triar.'-!e
 Park. North Carolina  27711. All relevant
 comments postmarked not later than 45
 days after publication of this notice \vi:i
 be considered. All comments will  be
 available for  public  inspection  during
 normal  business  hours at the address
 specified above. The regulations,  modi-
 fied as the  Administrator  deems appro-
 priate after consideration of comments,
 will be effective on October 12, 1973.
   This notice of proposed rulemaking is
 issued  under  the authority  of section
 301 (a)  of the Clean Air Act, as amended
 (42U.S.C. S1857g(a)).

   Dated September 26, 1973.

                   JOHN QtTARLES.
              Acting Administrator,
    Environmental Protection Agency.

   A new Part 53  Is added to Chapter I.
 Title 40, Code of Federal Regulations, as
 follows:

 PART 53—AMBIENT AIR  MONITORING
   EQUiVAUNl     ANO    REFERENCE
   METHODS
        Subpart A—General Provision!
 See.
 53.1   Definitions.
 S3.2  Applicability.
 53.3  Request for  equivalent  or reference
        method determinations.
 63.4   General requirements for an equiva-
        lent method determination.
 53.5   General requirements for a reference
        method determination.
 53.6   Procedures for equivalent and refer-
        ence method determinations.
 53.7   Right of entry; tests and inspections.
 53.8   Publication of equivalent and  refer-
        ence methods.
 53.9   Appeal.
 53,10  Cancellation of an equivalent or  ref-
        erence method determination.
 53.11  Hearings on cancellation.
 53.13  Modification of an equivalent method.
 63.13  Modification of a reference method.
 53.14  Trade  secrets and confidential  and
       privileged Information.
   Subpart B—Test Procedures for Measuring
        Performance Characteristics
Sec.
63.20  General provisions.
53.21  Test   procedure*  for  automated
        methods.
53.22. Procedures for  measuring  perform-
        ance specifications.

Subpart C—Test Procedure* for Demonstrating •
  Consistent  Ralationship  Between  Candidate
  Methods and Reference Methods
Sec.
53.30  General provisions.
63.31  Test procedure for oridant methods.
                               FEDERAL tECISIEt, VOl. 3«,  NO.  197—FIIOAY, OCTOBER 11, i»73
                                                           87

-------
 28440
                                                  PROPOSED  RULES
 Sec.
 53.32  Test procedure for carbon monoxide
         methods.
 5333  Test  procedure  lor  sulfur dioxide
         methods.
   AUTHORITY : Sec. 301 (a) of the Clean Air
 Act (43 TJ.S C. section 1857g(a)), as amended
 by sec. 15(c)(2) of  Public Law 91-604, 84
 Stat. 1713.

       Subpart A—General Provisions
 § 53.1   Definitions.
   (a) As used in this part, all terms not
 defined herein shall  have the meaning
 given them by the Act.
    of this
 section shall be  supplemented as neces-
sary. Candidate  methods determined to
 be equivalent methods shall be accom-
 panied by such manual when delivered to
 the ultimate purchaser.
   (f) Durability of automated  methods:
   U> Approved automated methods shall
 function  within the  limits of the per-
 formance specifications listed at 5 53.20
 (d> for at least one year after installa-
 tion in the field,  when  properly main-
 tained  and operated.
   (2)  If  the  Administrator determines
 that a representative sample of a manu-
 facturer's  approved automated method
 fails to perform as required by paragraph
 (f) (1) of this section, such manufactur-
 er's equivalency determination shall be
 cancelled in accordance with § 53.10 of
 this part.

 § 53.5  General requirement- for a refer-
     ence method de(ermiuatiuii.

   (a) Utilization of the reference prin-
 ciple and calibration procedures: All can-
 didate methods submitted for a reference
 method determination shall utilize the
 appropriate reference  principle and cal-
 ibration procedures as described  in the
 appendices to part 50 of this chapter.
   (b) Automated methods: All candidate
 automated  methods  shall comply  with
 the requirements  of  subpart  B of tins
 part.
   ic) Descriptive information and sworn
 statements  to be submitted:
   (1)  A detailed description of the can-
 didate  method  including  measurement
 principle,  manufacturer,  name,  model
 number and other forms of identification,
 including but not  limited to a listing of
 the components and schematic diagrams,
 ^iiUil U'C bU~H-.Lil ICU.
   (2) A  detailed  description  of  the
 quality control  progrr.m utilized  to as-
 sure that all measurement systems that
 are or  will  be made  available have the
 same performance characteristics as the
 candidate method  tested, shall be  sub-
 mitted.
   (3) A comprehensive description  de-
 fining  the calibration and  testing  pro-
 cedures which have been followed for the
 candidate method, shall  be submitted.
   i4) A statement that  the candidate
 method described  in  accordance  with
 paragraph (CM!) of this section is rep-
 resentative of the measurement system
 tested, shall be submitted.
   (5) A statement that the candidate
 method has been  tested in accordance
 with the procedures described in subpart
 B of these regulations and Part 50 of this
 chapter, shail be submitted.
   Operation   manual.  A  manual
 v;hich describes ho'.v a candidate method
 and the instruments utilized r.s part of
 that method should be operated shall be
 submitted. This  manual may be utilized
 to satisfy the requirements of paragraphs,
 (c> (1)  and  (3)  of  this  section.  Any
 manual which does not provide the in-
 formation and data required  by para-
 graphs (c) (1; and <3) shall be supple-
 mented as necc-ssary. Candidate methods
 determined to be reference methods shall
 be  accompanied by such  manual when
 delivered to the ultimate purchaser.
  te>  Durability  of  automated  refer-
 ence methods:
  (1)  Approved   automated   methods
shall function within  the limits of the
                              FEDERAL REGISTER, VOL. 38, NO. 197—FRIDAY, OCTOBER 12, 1973
                                                           88

-------
                                                  PROPOSED RULES
                                                                                                                 2&441
performance   specifications  listed   at
5 53.20' d>  for  at  least  one year  after
installation in  the  iv.d. when  properly
maintained and orcr.;:.. cl.
  (2i  If  r.l:C A.inyir.: '•.': iio'r oetennir.es
that a. reprcr enii-.Uve s:im-.-l> of  a manu-
facturer's  approved •.'.uio-r-.-.-cd meiuod
fails to pcr.jrtn AS required  by \.-^:\\-
graph :
  (a'  Notify  the  ap;/:.c>aut  that  his
method has  b?en c:j:ern:ir.cd to be  an
equivalent or reference method:
    Nc::fy the appi-ant tha-- liis re-
quest   for  an  equivalent   cr  reference
method deic-i-muuuion has oeen rejected.
including j statement  of reu.-ov.o for re-
jection;
  (c>  Notify  the applicant  that his sub-
mission is  incomplete  ar.ci  request  the
additional  informaiion  that is needed
(incomplete submissions will  not start
the 60-day period that will be utilized by
the Ac:::iii-:i.'trator to  determine if  the
candidate H.etiiod is  an  equivalent  or
reference method • ;
  (d>  Notify  the applicant that  addi-
tional tests v.-iii be conducted by the Ad-
ministrator,  specifying  the reasons for
the luiiii::..i.i-il tr.sis ithe 60-dav period
will coirnrier.ee one calendar  day after
the ackiiii^al  tests  have been  com-
pleted) .

| 33.7  l?i^iil of entry;  lr*U and in  the  applicant  may  request the Ad-
inin:.-;.-:Uor to reoon-uior the data and
inti-rrr.-.uion alre.iCly submiited. vc'  Che
uplj'.icant may request that the Artmin-
i.itrat.'r  repeat  any test which he con-
ducted that was a  material factor in his
decision to reject the method. A repre-
sentative of the mcuuiiacturer  may  be
present  during  the performance  of any
such retest.
§ 33.10   (".aiicpilation nf an equivalent or
     reference im-iJuul determination.
  If the Administrator makes a prelimi-
nary finding on the basis of evidence
and data available  to him that a repre-
sentative sample of a method designated
as  an equivalent  oi- reference method
does not fully satisfy the requirements of
this part, he shall  cancel his determina-
tion for  that mclhc-d in accordance with
the fo'lowing procedures.
   i a)  Notification :.;:d opportunity to
'/ifl.Ve  necesfary adjustments. <1> The
manufacturer sh.i"; be  noticed in writ-
ir.ic of the evidence and data which the
Administrator  has  found  to  warrant
cancellation.
   12)  Within 60 days following receipt
of  notification, the manufacturer shall
demonstrate to the Administrator's sat-
isfaction that  the method it. yuestion
satisfies the requirements of this part or
commence making  the adjustments nec-
essary to brinjj the method into compli-
ance.  If adjustments are  required,  all
instruments 'methods^ needing the ad-
justments  shall  be  adjusted within  a
rensoiitibio ume, is determined by  the
Administrator.
   (b>  Request  JOT  a hearing. Within 60
days following receipt of the notifica-
tion, the  manufacturer may request  a
hcarir.g on the Administrator's prelimi-
nary finding. The request shall be  in
writing, signed by  an authorized repre-
sent live of the manufacturer, and shall
include  a statement specifying the nianu-
fact-urer'.s oWeciior.s  to the Administra-
tor's findinsr. and data In support of such
objr-^vonM. If. after a review  of  the  re-
fiUCH  and supporting data, the Admin-
istrator :"-.;;,s that tV.? request raises  a
suosfj'r.rial factual  i.;?ue.  lie shall pro-
vide the manufacturer a hearing in  ac-
cordance  with  § 53.11 with  respsct  to
 such, issue.
    (c)  Publication <~>' c.Tnce'.'Mtion. It. at
 the end of the C0-d;.y neri.od the Adniln-
 \stiator determines li'.^t the ..v«prov:'! of
 ;.!:;• r.-.etncd should be canceled, h? s-hall
pu'siish a notice of  c-.ince:i:vtion in  the
 F^nrrAi RrriSTSR.  If a hw.ring h-v; been
requested and prranted in accordance wish
 § ?.3 10,  a cancelUtion notice will not be
 published until the hoarinsr proceedincs
 have been completed including any  ad-
 ministrative   review   and  unless   the
 A^ncy's  determination  to cancel  has
 been sustained by the Administrator.
 § 53.11  Hearings on cancellation.
    (a) (1)  After granting  a request for
 hearing under 5 53.10. the Administrator
     designate  a presiding olficer for the
  r2) The General Counsel or his repre-
sentative will represent the Environmen-
tal  Protection Asency in  any hearing
under this section.
  13) If a time and place for  the hear-
ing have  not  been fixed by the Admin-
istrator pursuant to § 53.10. the hearing
shall be held as soon as practicable at a
time and  place fixed by the Administra-
tor or by  the  presiding officer.
  (bHl>  Upon  his appointment pursu-
ant  to paragraph (ai of tills section, the
Presiding Officer will establish a hear-
ing  file. The file shall consist of  the no-
tice issued by the Administrator under
§ 53.10.  together with any accompanying
material,  the request  for a hearing  and
the  supporting data submitted therewith
and correspondence and other material
data relevant to the hearing.
   f2> The hearing file will be available
for  inspection by the manufatcurer or
its  representative  at the  office of  the
Presiding  Officer.
   (c) The manufacturer of the method
in question may be represented by coun-
sel  or  by  any  other duly  authorized
representative.
  "(d) (H  The Presiding Officer upon the
request of  any party, or  in his discre-
tion, may arrange for a prehearing con-
ference at a time and place specified by
him to consider the following:
   i'i) Simplification  of  the issues:   .
   Ui)  Stipulations, admissions  of  fact,
and the introduction of documents:
   liii) Limitation of  the number of ex-
pert witnesses:
   (iv^ Possibility of agreement disposing
of  all or  any of the issues in  dispute:
     The results of the conference shall
be reduced to writing by  the Presiding
OfPcer  and made pan of the record.
   I'e)' 1)  Hearings shall be conducted by
the Presiding Officer in an Informal but
orderly  and  expeditious  manner.  The
parties may  offer oral cr written  evi-
dence, subject to exc'.usion  by the Presid-
ing Oficer of irrelevant, immaterial, or
repstrii.ins evidence.
   (2'>  "Witnesses  shall be  placed  under
oath.
   i.3)  Any witness may be examined or
cross-exomined by the Presiding Officer.
the parties or their repre?|?nwtives. The
Pre.-id-.nT Officer may. in his  discretion.
 limit c:-o?s-exarr;i:iat:on to reievarst and
material issues.
   '4i  Hearings shall be  reported  ver-
 batim.  Copies of  transcripts  of   pro-
 ceedings  may  be  purchased  by the
manufacturer from the reporter.
   f5>  All  written  statements,  charts,
 tabulations, and data offered in evidence
 at  the hearing shall,  upon  a  showing
satisfactory  to the Presiding: Officer of
 their authenticity, relevancy,  and  mate-
 riality, be received in evidence and shall
 constitute part of the record.
   (6) Oral argument shall be permitted.
 The Presiding  Officer may  limit  oral
 presentations to relevant  and  material
                               FEDERAL REGISTER. VOL 38,  NO. 197—FRIDAY. OCTOBER 12,  I»73


                                                          89

-------
2RI12
PROPOSED  RULES
 issues and designate the amount of time
 allowed for oral argument.
    The  Presiding  Officer   shall
 make an initial dec:.-.on which shall in-
 clude written  find.i ;s and conclusions
 and  the  reasons  there-tor  on  ail  the
 material issues of fact, l.iw, or discretion
 presented on the record.  The findings.
 conclusions,  and \\nuen  decision  shall
 be provided to the parties and made part
 of the record.  The initial  decision  shall
 become the decision of ihe Administra-
 tor  without  further proceedings unless
 there is an appeal  TO  the  Ad:r.:::iit:v.:or
 or motion for review by the Administra-
 tor within 20 calent-ar cays c: the date
 the initial decision w.ts filed.
   <2> On appeal from or  review of the
 initial decision the Administrator shall
 have a'l the po~.*ers v.iiith  he would have
 in making the initial  decision inducing
 the  discretion to require or allow briefs.
 oral  argument, the taking of additional
 evidence or the ren-.anuuig to the Presid-
 ing  Officer  for additional proceedings.
 The decision by the Adr-iinistmcr shall
 include v.Titten ii::ri::r -  and ccnclv.'iur.s
 and the reasons or "or  ; therefor  en all
 the  material  ;>s::c-s c: fact. IKW. or  dis-
 cretion presented on the appeal or con-
 sidned in the review.
 §53.12  Modiflrulioll  of  an  equivalent
      n.i llio<; nr-plicant. within
 30 cp.JeHd.TT days,  from ci'.tc of the report
 of his intention  to c:",J:-. el the ces ora-
 tion  in accordance v/it.i 5 53.10.  If ihe
 Adaiir.i«tritor  deterr-.uies  that   addi-
 tional tests are Kcec.'£d to  enable him to
 determine v\lK..ier 10  continue the cit?-
 ifnation  in eiiect. h^ skull so notify the
 applicant within 15 calendar days of the
 report.

 §5:'.. 13  Mmliliralion  of   n   rrfrrcm-c
     incllicii.
   After a cnn'.rdrto method he.?  been
 dei-i.-.ated a  rt:c-rcv?c method inuto-
 rna'.-;  o.- r;:">.m;r'i>. :  Test procedures applicable to can-
                                         didate methods and reference meriiods
                                         which are au^nat-ed methods for devel-
                                         oping   obiei'vations   on  Interference.
                                         Lover Detectable Limit. Prec:sion. E;se
                                         T;rr.e,  Fail lime. Zero DriCt, Sixui Drift.
                                         Lag Tir^e,  ar.d Noise ore described in
                                         this subpart. Measurement systtms must
                                         be  operated  in  the ranfre  si.eciried in
                                         Table  B-l  of paragraph icU of th.iK sec-
                                         tion during  conduct of these tests.
                                 (b) All  recorder  chart tracings,  test
                               data and other documentation obtained
                               for the purpose oi  those tests sh:;!! tc-
                               identified,  dated and signed by t!ie ana-
                               lyst performing tiie test, and mi'!;i.'.>? If the instrument pas.-es r.ll seven
                               (71  tests for a piven performar.i-e ^;:^.•ci-
                               fication. pass the  in.--irurr.em  ;or  tii.,1
                               specif.cation.
                                 (2) If the instrument fails th:tit-rm:iJcJ i>> j;. ..tfu-'iin; z*ro ;ur co^iliiMjo*:^)
§ 53.21 T«'-t procc
-------
              Oil TEST ATIUMFHI-.K7.-: TABI.F n ? ^RKI-
                                                              PROPOSED  RULES
                                                       . -EM I.HTKD AT ENO or TABLI)
                                          Generation
                                                                           Verification
 1. Ammonia.	 (,T:n..;iii,,n nil* -y^-oni. Sinn ir pro-  Coll.vt «mpies In bubbler con'.aining
                                >•. -luretoin 11 -1. -,i:U-i m r-\..",ui'.«    o.!N'  lit. 1. An-iiya- ximp,.-s by tu»
 .,,,,.                   *:   .                              1'!. ''iv...  mi •Ji.'.-J, p. -.3-J. r-.' relic* 1.
 ICarlk.n ..:.n.[,•		 CV.N>I. . ,.r;,-ro:.ir..-:i' -uuri :lh(.rin-  Cortiii,-.|  aiUy.-ns by  ni-li-ufai-tutM.
                                ',.':,!raiuMl 0{ '  '>: si" ^ll-"'t m 1-T'it.'    vthil'T oniiu.'peni:l->nt Ui'.-oratory.

 3. Cvl.ou :;:oi.oi. ! •	 C v.:i ,!.•;•:,  i,f ?.,•:.. air i-om:iii.i:!ir  l!>.>  CerliiiM analysis  by  mai\'ihi:turi n( (U ,|,,/i:-..,l  m    v,Mi.i,,r .,r iii'!-'|k-ti.|i-iil ItilwJlory.
                                1-lMeB 3.                          Kfcr-rli'.ica'.iuM   rvquirt-d   evi-ry  ti

 i F'.hLias	 ry-!;;u1ir « .'f 7., .••)  ilr <-,.nt.iii iu^  r)  f ^--nii.-i" n" I'yvi* 1-y  iTia-infu'liiri-r
                                p.p.r;:. ..•.],.,•.*  li.i'j:.-  nilli i. P. :i.r    vvh IM;- .,;  ..; I,.;),.:,".!,-!!!  Ijuoialoty
                                B 3.  "'  '---''-I'l--   i" «•>"••-    i.-....- ^... .iM.uriun

 I. Etiu:,-!',...	 C'\. :,.!-r'  -f "i^.-p-in: ,-.|" n;..-,,..•.,  tVri'il .-!  :i-. ,iO.:/i,v  i,!:,iiiil-..-umT,
                                vx'i:a:ni!..- -' p.piu. .•!;:;.••-'•-.  l>.-   v.::.lor 'ir i'.; i;.'jkil'!i'lit l.iU>r:Uury
                                nir • 'AI-'I 7 TI> ii- ' • ti'i..•••.:;..!:,••!!    u..ii-.; i.i:--.. 'i-iiiz.iiiaii  di-t.-.-'-ir i-iii-
                                sj .ri!,,-Ji'i Taliii- H 3.                '.T.ii-i «•;': ::..-'!.:i!!->. it-. ••:l:iii-.tlioi!

 C Hydro,;-1, cj, :..>.:•]«	 C~: '-.ril.T*  . r "M-.-pur.!!.'.!" nilnj-.-.-Ti  Co!---  >'i:ii:-';fi'i"n l'Ui.'i''l(:r'c.i:!liii:iir.i;
                                i •'.:• i .   .. U".'|-i.;i.n. ni - .. •,.'.!< lir;    ,t: -•,:: ..| w; ,-..1. \r j.yz,. .ju-.p-i.; i,y

                                «i-' !•<• li'i T:, l>li- i. 3.      '          A'-INi li'V'.'ji. p'.l".': r.:. r'.-:-.o'-.'. l> '
 i. Hycl'O^a ^ii.'nin	 P,n;.  ;.t:Pi: tul»: .-> '• in. p. Urt. r,-!.-r-  Toiii-iiiv.- :;;. ••!:«!  of an-ilyyis for HjS
                              C\lmd»r» of /.t-ro ;vii coi.l vi'.ii;.^ [it!-  Co
                                d.-Ti'-'-i^riitiua of :M''llut:;t' .--^.-(.liicti
                                in T.l'-l>- b 3.
                                                                         .-i. ..ysis  by niL\;i:;':u-tnrer,
                                                                         or  iii-it p i v. i;:i at"thun->. i'.rc»-r!ili«.fition
                                                                  i.- r • i1: .-••u r-\ ••:>• 6 nirmt!:-..
I. Nilricri! -In ..................... Cy!i':d- r • of "[.ivpurili.-!" ni'rfi-ji'n  Tcr.liilnv :n,-!i.'i.l of ,iriiiy,-ii (,.r nil-ic
                                c-'T.'.iiiiiiiir I-.KJ p. p. in. Nf'-  !n;u'>>    tn.U-.  r.Ki'. :IL of the ali;i!*|.lieie, p.
                                wi:t! /..-ro :iir ;o . (.tu-i'in r '• lfvi -[*._• i-    3-''.. t-  f--' net 3.
                                lioj ir. T.ib:.- ii-3.
10 . N:t:ni!iT. il.iai k- ............... 1. t; i-  [il. .,-..-  tU'-inou--'!.-- i ipta.n  1.  R-f-ivni--  ).  ApiienUix  II;  Os is
                                a\ Liil.ibU1 from  N K liC H't i'.         i,!Lr.ir -1 ,i.;.i.a*c No which is rq'.iiva-
                                                                  liir'C I') N 1 1; ir.fine 4 ami !n.)di:i,'U L.y N K lit -
U.
   Su'fii'
                              C .;.t> r it- ! r./< n;  '/•"l-frntor. ItrvK-.''.lurc  Kl  :ii-'l!hHl.  Prnc'-durP d<-:-iil«-J in
                                --I'lr-1  lj-ro>u!i:. •• ::.- timd.  Prrt-.^lur*  de-
                               d'-ci:i>f.l  in n f-n-i.i'e nieli1.'' 1  f..r   s^Ti^sl i;i r'-UT^nc*1 ni^ttu.«l for SOj,
                               SO-. ri.-.Vr-Mic* 4.                    r^ftren-t' l
                              r*-7. r.. J;r!t,r(u..;li.I!.stii!i-.la-.-i|.T .-,  M.- i,tir.  r.-; .tiv.;  hun.iiiity hy mcsiii
                               <, llvo, I t, ,..„„>. .. iu..,>t,:... f.-.Tiv....i.   r.l i .1 .«• ;.,,int iiudi-intr  clIinnitMl
                                                                          "
14. Xyi»n»
                            .
  t'OOOiri-.'c; s i!;ir j^-a. J >il;.L- wr t; /-r-i
  a^r  ro  o >.'•' ••-. .L'.ii.i-iii  >:«-'-.:-- i  ia
  T.ii-lf li-3.
C'yhn.l-r" of "pr- :-iir;lif:|" T.:U'-;..!I
                                                                 eter  i
                                                                                 .
                                                                          .-i.Jry bulb


                                                                          :i.::vv;s  by
                                   .
                               with z*-ro iir to f ..n. onl:;ii .on ,-it.' .-I-
                               r. il in T..''i.' H-3.           '
                              Air I. \V-.KZ ;•. w .*!••; . 'I'.t^va '^ : s< f- tn
                                                                                             cuii-
                                .      . ...  .  .                .      ...           .           .
                               ^s  in'-O: >nc 01" ;-• ^ t:i .i:  -y >• ;u-\, I/i   u.-.;:.^  rli.. f i^iu/.-itj,..:! -it-t.--vt'>ip  i.-.i:l-
                               0.1  p.p !•;.. '.-lrb.jp  -il-.xi'l'' ci.nt-nt   b-utcl »:U ;:iivti. li'i- all.) i;n':.;r-il
                               n(3."M-3iiii p.m. »n.l nx v  . •iir.uiu-i:t   'l-i.'-.or ciii'-'iWJ wiih c.irbon rli-
                               of JI.O— ii.'t p-'ri'Htl.                 OTi.l-   i': I  liy  i,>;!:^  -.i;.i.ro[-riaU
                                                                 aiLllvtii-Lil   t" 'i:::l'Ki.'S  I.;!'  01V-X-.1
                                                                 and v. i:«-r i.; .L'-nt.
                                   i!--i>-  H'd to. !':l» l-j'.ll'l ^i-t n:-:l-1l.-"i|.
                                   :i-.i ,,f W:it,-r ;ir"i V. ^ .  W H.T. I it i K.I.. A::i-:i..m PubMi Health Asso-

                                  1 .!:i'.!-  S'.irl .'3. Ali-i -'in f-x-i-'ly (or IVslifv; ami Materials, V-18 Hace

 ii ;. :i- v        r .vT S -.iiir!.;^' an.'. Vr.iiy.-i-. Ii'.t^tso-.'i-.'iy Coniiniit'''-. iv'72. Ain^i icAn i'ublio Heaitri A-socia-
tlrj. l.il.;. I-;.'- -I •- . '* ".. •• ,'-:;::\ '. . i. !) i '.
 ::•:.!- NJ.;.  . i'.. .: y .1: -I .-  .i;..i,-y An, i.i. :.-. A:r Q'li'ity Slauil irJj.  40 CFi: 'A

  ic'i  S:-j!;-r,p. rjlibratioa.  nr.c  -afcty   and deadband  of le?.5  than 0.25 percent
requirerif  ' •:                                   of  scpie.  Allow  arfr-.-iaaie  warrn-up  or
  '!> .Vci-ii'; .-r.d st.irt-un of rho mst.ru--   stobiii/ution  iir>i?   as  indicated  in  the
mc-nt il.a!1. I-  \.\ strict acco.'ihir.L-o  vith   instruction  mar.tra.
tli"-  ;n^;''    c  r's   ia-iu(';..!v    ;'-mua!       t2>  C.'ili'srat;.;:: 01 the instrument shall
                                        28443

 operators  or environment or  discharge
 any  hazardous  or  detrimental  effiuent
 gas. Instruments shall include saiety de-
 vices to  prevent  hazards  or d:;rnage  to
 the internal components or performance
 of the instrument.
 § 53.22   J'rocedim's for  measuriu£ per-
      formance spccilu*:iiions.
    (a)  Range — n>  Technical definition.
 Nominal  minimum  and maximum  con-
 centrations  v.hich  the system shall   be
 capable oi measuring. The nominal range
 which shall be  used during conduct of all
 tests is specified in  units  of par's-per-
 million by volume in Table B-l of 5 53.20
 idi . Po]lut.Lint concentration at tixe lower
 and upper scale  limits are specified.  a.s
 for example: 0 to 1 ppm.
    (2 1  7'e?f procedure.  No  test procedure
 is  required.
    (b)   Noise — (1)  Technical  definition.
 Spontaneous,  short  duration deviations
 in the instrument output about the mean
 output,  which  are not caused  by input
 concentration  changes.  Noise  ii  deter-
 mined  as  the standard deviation about
 the  mean and expressed  in concentra-
 tion  units  given   in   Table   B-l   of
 §53.20(d..
    (2) Test procedure. U> Allow sufficient
 time  for instrument warm-up  and  sta-
 bilization.  Noise is determined u.v;ng both
 zero air and a pollutant test gai concen-
 tration a.s  indicated  below.
    (ii>  Recheck strip  chart recorder   to
 vertify  and  demonstrate  that  recorder
 deadband is less than 0.25 percent of full
 scale.
    ;iu>  .for this  teot,  i:  is ncco.viary   to
 disconnect the  strip  chart recorder from
 the instrument output, and list instead
 an  integrating type  Digital  Volt  Meter
 fDVM) accurate to three sieni:'.c?.nt fig-
 ures on the instrument output. The strip
 chart tracings  shall   be submitted along
 with the DV..I  readings.
   (iv> Sample zero air for 60 minutes.
   (v) Dunns this 63-miaute interval, re-
 cord  twenty-five  r25>  random  DVM
 readings.
   (vi)  Calculate the noise from the  25
DVM readings using the following equa-
 tions.
   (vii)  If  k ,,  k ..  •   •  * fc ,. are  the  in-
dividual  DVM  readings,  the  mean   is
determined by
                   _  Zfc,

                        25

   Deviations about the mean  are,
Connor*
                        to a  i.-jita
                                                    '. ^ '  0:1 n.-ra'..,;.: ui L::L :u:>i.i uiuci;v Mian
                                                 be  ::..s ir.dicared by  the  instrument's cp-   and the standard deviation (noise) is.
chart  ;:> :,;c!,--r.  .NOTE.—This  require-   eration ni.i:i-J.\l.  A dyn.iruic
men is  i-.o:  u-.i-.-nd'-d to  limit d..ta  ac-   cuno which  ha.-;  at  !e^i 7  S.'!cntin:ible
tbi.LiLU  V.-.v   iiitLii;'..'. wt;cvt;it.i*Ai»i^>ii\^A   •
data sulv.n-Via  to £PA for the  purpose   The calibration curve snail include ft plot
Of  (Jc-m."'-!-a:::i-;   i:- u    n  c,\nd:;'tu   of poU'K.i^t co:.i;e::u:i!io;i asainsi chart
method  is an  equivalent  or  reierer.  t-   divi-ion or porc;-;;t o-. icule and a::unst
method.)  The recorder shall he of the   inr,truuu-:it  output  in terms of  Vulw or
servo, :.i;ll-;>:'.i"'ire  tyne  )uiv::i~  a  ch-'.rt   millivolts.
Wi;iia of : I1  i  ,-.\."s  or  ^it-a;.-r.  r_.  o^nse      <3'  i-Toi'inal ii-v.r;i:ne:it operations or
tinie of  1  .>e';ond or  !.:.•.>,  at lea:-;. 5 i,or-   failures -shall u  it   civ.-rate or  prL'jent
ceut below >vro  or zero oii'set capability,   any hazards or h.i7.irdous conditions to
                                                                                                                          24
                                                                                                  Convert cl, and S to concentration in ppm
                                                                                                  by reference to the previously determined
                                                                                                  calibration curve for  the  instrument.
                                                                                                     (vlii)  Repeat step.:  (iv)  through (viU
                                                                                                  u.--n;;; a  pollutant tost, .'Ms  rot;cc-:uration
                                                                                                  prociucins  an output of Ci)±5 percent of
                                                                                                  the spi-cified upper range limit.
                                                                                                     (c) T^iscr iefcctabli limit — (1)  Tech-
                                                                                                  nical definition. The niinimua pollutant
                                                                                                  concentration  which,  produces  a signal
                                                                                                  of twice  the noise level.
                                     FEDERAL REGISTER,  VOL  38, NO.  197—FRIDAY, OCTO&ER  n,  1973

                                                                   91

-------
2S-H4
                                                      PROPOSED RULES
   (2) Test jtTorf'lnrs. (1> Allow sufficient
time for warns-u;> nr.d stabilization. Sam-
ple zero conconrraiion and allo'a- for sta-
ble reading t.> v: 'abU. h ;••. :o ba^-liiie.
   (li! Gentr.ite .ir.d sarr.::le a w:u atmos-
phere equal  10 the vr.Uu; for ii;e  lower
detectable  limit given in the perform-
ance specifics.ions [Table  B-l,  5 53.20
 I.
   UU">  Record instrument  rerulnicr.  An
instrument re;: <*••;«•? of at lei'.ot twice the
noise level previously  rietermir,*.-! ?.t  0
ppm is nee.- :.ry to satisfy  the r.-.-T.v're-
mc.'it  for  lower detectable limit sen-
sitivity.
  (d) Wcrfcrttrc csuir^lciit—(1) Tech-
nical (ic/ir-iiio?:. JK-r>:ve nr r,-r.vive one-
put cau-od r  • a s-.i: ar.c? oth-. T ili;>n the
one boiiir; mea^.:: -1
   (2) Test .w:h r b3 c.'livr nositivc or nega-
tive,  dt-ijcndir.s  on rvhet'acr the iv.r-'.ra-
ment's outr-'U is ir.crtv.s.-d  or d«.cre;'..*ed
by the presence of the  ir.t--r.crent. In-
terference tests nre performed by mixing
each int-M-fereiit vi:h t'-.c polhit:«u ccn-
cc-ntrstlfrus s.vrci'u-d  i.i  Tr.:/:c B-3 of
parr»'rrR!>h  cf this sr-ction urd com-
pnrin!;  the in:-;t-umrnfs  ouin-.n to the
output caused by  the  pollutant  alone.
Knov.n gas :-h.-.so rcjctions  that occur
bctr;ccn  the  intcrferent and  the pollut-
ant are  tle?:?natt-d in Table  B-3. In
thf-^e CASC', tbe ir1  ".•for*07 respon.^p shall
IMS dvUTiiiiii.Hl in Hie a^.-r-nce oi ;i.c  yoi-
lutant.
    Allow siiificipnt time  for  \varm-np
and stabilization.
   (ii) For !r_'.i:i:iiKnts usir.ir a  prtS'ter
or scrubber br.sad ot; a cl:«.::iical reaction
to derive p-rt of its specificity, :-nd Uiat
       requires periodic service or maintenance,
       the  instrument r-lv.ill  be "conditioned"
       prior to each interference trst as follows:
          (c <  Son ire or perform  the  indicated
       rsn'.iiit' luince  on the scrubber or  pte-
       lil'.vr a.-- riirected in t;-o manu::!.
          * b> Before testiiif1. for e;icii ir.U'i'i'erent,
       allo\v the instr.ur.^»;t to ?air.ple through
       the E«:n:i;bi-r a test atmcspliere contaia-
       ir.T  the inicrr^renc ac  a concentration
       eoual ; j the vrluo specified in Table B-3
       of p-inu'raph  «ci  of t!v ; section. Sara-
       piinj: :-..all be at the nor.:.:il flo-.v iv.ie and
       shall' be performed  tor C  continuous
       hours prior  to tesnns.
          'UP  Three  tost  nuno-syiieres shall  be
       CC"er..tod ?s fLltov..-:
          • a? Tv.st :v.ri.c=r'-M.o-•": Specified pol-
       lutant conceni-rrticn.
          (7,1  IV-T a;.i;o :••: re Y: Specified in-
       te: i'c!'••-:-.I conctii'-i'ai ion.
          (c> Tt">'  atmosphere Z: Zero  air iu-
       terfertai ccncentratior..
          (ivi  Adj»:--s the  individual Cow rates
       ar.;l i.'.e poll'uunt or mterforent genera-
       tors for  the three  test atmospheres  as
       foilov.;:
          (r>  The flcvv rates of test atmospheres
       Y a:Vi 2 shall fc-  i.xi'h either
       test asmo-phere Y or  Z.  the resulting
       eor.r-'.nirati^Ti  of  po'.i';' -nt  shall be  ns
       si'-.r^i-'c! ir.  Table B-3 o: panvyr-.yh >e)
       of th-X   The  co.-.centratlcn of intcrfcrent
       in te:; r-inio  ohtre 2  rhii'.l  be adjusted
       such that when mixed (di'.uisd.1 r. ;tii lost
       auno.-vjnero 1C.  the resuitirig cor.ceutn,-
                in^rf'-rent  shrill be equal tc the
                                   a stable reading, and record the read-
                                   in;; in concent-ration units.
                                      Sr:r,;;ie  i'.nd n'.t-asurc n  m;x: .re
                                   of test atirosfjiere  X and Z.  Allow  .. r a
                                   stable renc'ir.s, r.nci  recorii tlie ro:ic^i.'; in
                                   concentration ururs.
                                     iix> Calculate the1 intirfricr'^1 ;..--  the
                                   Et.-;:.>nd  rf  •!:".•,' minus tlie fi'vt  r-v.'diiv^.
                                     (X)  Ropr.it  r.ti"is iii!)  thr;>ii"'.i  'ix>,
                                   in turn, fur ra-.-ii u'.teri'ereni.
                                     'xn For  lii'.v-e  U'.ierlei'i-nt.-;  \\hich
                                   c:'.nnot  bo  li'.ixcd with the rolliit.-:;:. ;-s
                                   indicated in Tul-'.c E -:5 of p:\i-a; r ,; ::  ••.•)
                                   of this  -ict-'i, :! arijM;it the- cor.'  • ..  '^.r.
                                   of tho i:iter;(.-rc:ii hi tlie t^ ,-r  „' • • ~; ;n.-re
                                   to  the specified value  v.uhout bt.a-:5
                                   r,:i:-:otl  or   c:::u:td  L>y   any  cUicr  test
                                   atmos;jheros.
                                     <:cii)  Sample and nit.1 sure the inter-
                                   ferent test atmosphere and  allo-.v ior a
                                   stable rer-flii'i--:.
                                     fxiii) Record UM^ rc^dir?. in P';l;ive
                                   or n.vrtive  con-cntr.iiicn unit.--, as  the
                                   interiV-reiico t-guivalrnt.
                                     (xivi T'r.2 sum of the  ;'.b o'v.li> v.ihir*
                                   of the i'.v.livirl:i:>l interferences f.li-.-.U  not
                                   e\cr;u  the \aiL;n t,iven  in :\^rforrr. mce
                                   specification^ fTable B-l ef  $53.:o>d'l
                                   for thr loL.t! irti'rfercnce. -|jp"i-
                                   fi-.:;ii:ons  rosulis frcm  ciiailC'.!- iivj  ti:->
                                   mc:isv.ren:»j!\!. .\-\-'^m v.ith niaxin'..-! con-
                                   ce'irrnti--n cf  potential  intcrierv. •..'>..•>. If
                                   ainbiL-at air cont?.i:.fd :!-.e c;.;:nb:r- .i re:^-
         of this section.
          (v) Mix tost at:«o,pp..-res X ;'.nd Y by
        pr-vsinj the total  f.o.r of both tJyou^h a
        mi\ii"r. fiark.
          (vi>  g.'mp]-? and mcr.siue the mixture
        of te:.t a;uc?pheres X and Y. Allow for
                                   for the purpo.-e of thcce tests, i1, i.  r,:o:-l
                                   likaly that the concentration of t1.-: \--~A-
                                   lutant  of interest  as  uportr.-f:  by  Trc
                                   measurement  rystein  won''.', far c.\c':cfl
                                   the air quality standard.)
                         (e) lKTf.Kff.':F.M TBST COSCFSTH.'TIOV H.E APTOMATKU MSTHOHS IN
                                                                                    r«:p. Mauo»«: TABLE Ii 3
fcll.Tt- ( Hy
6Ot
SO,
SO,
J-

phy-iH i".
PI- i-iK-i houaii"!-
e*i«i tua K- <>
suj title

0. 1
aa »ai OLI
SlU-:r N:iro- Nitric i:-ir'«n Kl'-v!- in- W-'i-r Cr.ri-nn .V.-rh-
liii'ii ic »:<-n o*i'!«- d:(;t:ti-- euc Oxone Xyionr -.^in-r i^on- tn- r',!;iiT

§0^14 • 7jO 	


, ^^ fl>) -^
>"\KO 5-- 	 - 	
O.S

FO,
90,
bUl

O,
TIM •*.--:. -









-<-.

— ..... 	 .......










••** 	 ...


— —*""•










U i

& 3


OiS






a2 0.5 .
750 	
	 a*
7T-0 «d08 ...



fc **; '"* ^

0,2 	 	
	 	 *-r..ooo 	 	



O,

NOj
NO,
       S-J1.--I!.'
»ai
«.s     as    '0.5	   «o.o"	
as    «ai     as	'zs.ooo.
as    «ai     as	    as	
NO,
                                                    as
                                                    as
                      • 0.1
                      • ai
               o.s ...
               as	
                                                                                                               50
                                            0.2 .
                                 rCOEIAL BEGISTEt, VOl.  38,  KO. 177—fRIDAY, OCTOBER 1J,  1973
                                                                 92

-------
                                                    PROPOSED RULES
                                                                                                                   28115
 I'ollut-                Hydrochloric Amrao-  Hy.lf*.  Sul(nr   Nitro-   Nttrt«   Carhoa  Ethyl-           m-   Watsr  Carbon  Mcih-
 ant  Mi':isurifij; principle     acij       nia     cr'i    dloiida   »»n.    oiide       O«>n»   Xylena  vapor   mono-    am<
                                        suuUla         d)ou>i«                                             oiiija
co
CO
CO
CO
CO
CO
IR 'al'wt tlvv.v

F!'> irxl>i>:>:K '!--- 	 - 	 - 	 	
Cai.iK-,.:..-,ii-.:!>ns- 	 0.1 	 	 	
IR f!i o^---fr -i-
Msrioiry r.'K •.•*- 	 	
750 - — —

	 0-S 	 £2.. 	
	 750 tt 2 	
7.X)
	 a 2 	
25.000
— . 25.0UO
	 25,ono
.. . 25.000

• 10 	 ,..
'10 	
• (0 	
• 10 	
• 10 	 .• 	

0.5

0.5
0.5
THCY  C'ataiviii- onm!>tis-
 CtU    £ioa-til>.
                                                                              tt2	  25.000
                                                                                                               '2.0
                                                                                                                        0.5
• Concentrat'.iNis <•< ir.i.-r; ;•• :u
fc Do not :-.'.\ %;•..; • ,i".nn."'.
• Conreiitratioti in' ;•
                         iite-i r-.n-t K- rvr- ;,.ived
                         ' i for t..-7. Tli'v^ r^Hui

  (f) Test procedures for zero drift,, spaa
drift, precision, ri.-e  time, izll  rime,  las
time:
  (1) Determination of these r°rnonse
elements  shall be  accomplished  over  a
period of  at least "even o.ays. Du.-;n?  the
test, manual  adjustments to the  slcc-
tronics, gas  or reagent flo--.s Flvli bo
permitted only orce ev--ry  t'uoc'  day.--.
Automatic adj\istments which ave a part
of the normal instrument operations  are
permitted at p.nv 'ime. A s-ampUng p'.au
Is given in Table B-4 of paragraph  
of this section.
  (2) Technical definition of zero drift:
The chan.ce an in.-rtrument o">i'JUt over s.
stated time period of unac;;us;ed con-
tinuous operation,  v;hen th-3 input con-
centration of pol.'urant is zero.
  (3) Technical   dcfini'.ion  of  span
drift: The ch'n^e  in instrument output
                                          to i- -'•) r-'r-'T.t of s', .:>•! 7u!ue -inJ

                                          ar.t c.:.;,-, .:i'rn!^.[:j muf- I--1 [nvparo
•oniro'iieO. to ± 10 ]»rcent.

i t" i 10 r^rivnt of stui' d value ar.d cor,troli''d to :
                                                                                                        5 percent.
continuo:is  operation,  ivhrn  the input
pollutant concentration is a stated up-
scale value.
  (4> Test procedure for  zero snd span
drift:
  (1) Allow sufficient time  for instrument
to warm up arid  stabilize.
   Sample t«r..i atmosphere of pollu-
 tant concentrations of  C. I'O. r>n;! ?0 per-
 cent of  U-e upper ra::o:e I:.ail each day.
 24 hours after the previous day's read-
 ings. Tlie.se test concentrations  shall be
 consistent  from day  to day. Allow suf-
                                       ficient time for stabilization of zero and
                                       span output and record the readings in
                                       con- eatrattin. ur-.i'.n. Subtract each day's
                                       reu.iiin?s from iViC- previous day's respec-
                                       tive readings  to o'/iain the zero and span
                                       drift i'or th:.t  cay.
                                          (vii>  £a*-.'.;'e  anri  measure zero  air
                                       continuwsiy  for at  lenst 12 heAjrs each
                                       day. Detenr.lne the minimum and maxi-
                                       mum reau:nss  during this  period of 12
                                       consecutive hour? and record the "peak-
                                       to-peak" d:"erence.  regardless  of  si.yn
                                       in  concentration  units, as  the 12-hour
                                       drift.
                                          (viii)  At the  beginning of the next 12-
                                       hour continuous  zero  air  test,  change
                                       the line vc!c'\~.e and room temperature to
                                       the vp.lue givea in t-3f.t plan.
                                          fix)  On  tr.e  -Jtli.  Teh.  10th, and  13th
                                       days r-s i indicated in the test plan, the
                                       instrument may  bs serviced end  ad-

                                          (x) On the  8th day,"the test  results
                                       shall be examined to determine  if fur-
                                       ther testing ir, required.
                                            Tliio t-2st sh?.ll  be conducted ei-
                                       ther in an uninterrupted !>eriod of fe'-'en
                                       or  fifteen rtsys. ov periods of three con-
                                       secutive davs C'Mrit'-S;  which  no adjust-
                                       ments ;:l:;-.li be  pv'V!i:ii'«d. For exarr.ple.
                                       if  the  t.?;t is i:-;?rru;j*ed by weekends,
                                       it rvvr>.y be- •.•.•••/formed in threo-vU;.-. ana Thuvsday.
                                          <5 i l..ajr uii'.e:
                                          »-;•: V.-^hnical d-:rini'-ion:  The time in-
                                       tcrv;i.l  b3t-.vey;i  a step change  in input
                                       coiiec-nti-3tica  as the  iriStrurnent  inlet
                                       to   the  fii-.-il  ob.ser.-s.U!e corresponding
                                       chC'.r.ce in in.;tru:t:?:;r ou::Jtit.
                                          (ii> Test :>:'oceciuve:
                                          (c.i  Sample   zero  air until  a stable
                                       rcadiD-j is oblninea.
                                          i?), Switch  to Uie  test  atmoiphere
                                       used in par^rapis.  (,f)(4)(iv) while si-
                                       rnulu-.^eoi-ijiy itartin?  a stopwatch.
                                          '.c> Observe   ti-.e  in.-cr-ment  oi::p-.it
                                       and  .slop  the  .•/lO.v.vatch  iiv.rnediui.eiy
                                       when the fir.t  observable change occurs
                                       due  to lUe step concentration change in
                                        the output.
                                          (d> Kecord  the elapsed  time.
                                           ,:c> r:i--e time:
                                       .    ci)  rec-;-inicaI  r:..-.1n:tiors:  Tlie  time
                                        Interval between  initial response and 55
                                        percent of final response afr.er & step in-
                                        crease in input concentration.
                  (ii) Test procedure:
                  fa> Sample zero air until stable read-
                ing is  obtained.
                  ib> Sample a concentration of 80^5
                percent of tl->e upper ranse  limit,  and
                obtain a stable reading.
                   Calculate 95  psrcent of the read-
                ing obtained  in paragraph (f; (6,> (iii  (6) (ii) (a)
                of this section.
                  (e> Sample the same concentration as
                in paragraph (b) antl use a stopwatch to
                time the interval starting with the rirst
                observable instrument output response.
                and ending when the response equals the
                reading calculated in paragraph (f > <6>
                (ii) (c) of this section.
                  (/) Eecorci Uie elapsed tinje.
                  (7) Fall time:
                  (i> Technical definition:  The tiine

                percent of final rer-ponse after a step d;-
                crease in input corccnlration.
                  (ii)  Test  procedure:
                  (a)  Sample a concentration of 80i:5
                percent of the upper range limit, and ob-
                tain a stable readina:.
                    (Ki^e
                Time), step  (ii> (a>.
                  (c> Record the elapsed time.
                  (ct> Repeat steps  >rz>, ti>),  and (c>.
                  (3)  Precision:
                  (i)  Technical  definition:  Variation
                about  the mean  of repeated rr.ca^^rv-
                ments of the. same concentration.  This
                var'acion is  expressed as one  standard
                deviation about a mean.
                  (ii)  Test procedure:
                  (a)  Alloy suiT.eient time  !;>r instru-
                ment v/arm-up and st.ibilizntioJi.
                  (b) Sample and measure zero air until
                a stable reading  is  obtained.  Do  not
                record.
                  (c)  Sample a pollutant concentration
                of 20±5 percent of the upper rmsa limit
                and obUun a stable rcjci;;;;j. Convsn to
                concentration units and  record value.
                    Sample a pollutant concentration
                at least 50  percent  higher i.30 percent
       No. 197—Pt.
                                 FEDERAL REGISTE*, VOL.  38, «O.  197—FHSDAY, OCT05«  15, t973



                                                             93

-------
                                                     PROPOSED  RULES
of full •.••.-.lie;  liv ;i -.;i  noi record
   te>  Quickly  swircri to thr poMutaus
concentration used  in (c). Record the
reading.
   (/>  Repeat steps   Calculate  and  report the mean
and standard  deviation  for  these six
value:-.
   (h)  Repeat step fb).
   !e rcadhv-. C.V-nu-:t- to
                                           Cv>no":.irv.-;o-; ur.iV-  rru  r:1 •'•• d value.
                                             (j  Sart:;M3  a  pollutant  ccr.centraticn
                                           20 percent h:,rhcr than In paragraph (f >
                                           (8) <.i> of this section. But not to exceed
                                           97 percent of full scale.
                                             (fc)  Quickly switch to  the pollutant
                                           concentration used in (?').  Record the
                                           reading.
                                             (1) Repeat steps (i) throuch  (fc)  two
                                           additional times.
                                              T, C'ljllStlE.-II'Ji TCiy l«
niadr to the ins? r siurut, U ntfe*-
sary, afii-r rra*lii-i*> arc Infrru.
E .i.,).lc and trad 0. spin i aiid
spar. 2.
Fampl? and read C, rr-an 1 ami
span 2.
Fai'?I-l<- »v. »:i-
S.ir.ir'1* und trad 0. -|V.n 1 and
s.p:in 2".
£ai!i;>k s'id tuJ 0. .-(.an 1 and
S.niij.:v and road 0. s,.s.i> 1. span
V; h-JJ'iyniW'Kt^ jr-r:aitlfd aftT
r«Ti'l:i:f* are ink' i.
F;:i.|.le and rr.i-1 d. span 1 and
Fr'.:|Jf. »n>l r,a-I t. =t:n 1 and
f^j"""irlft -Mr: '•,'!»?. *

F.in.|.!^«J pa! !>. f(wt 1 »».:.|i ii:rroiTi<-i.
                                         li al>o be run eiiiier Oil f'U!«onitlve days » ith a£Jjit*lri.'ent
                                         «cm]ve s   Selection of test  silef. Tlie exact
location of the sair.p'.uig sites is left to
the discretion of  the manufacturer: the
sampling sites shall i ?  loeattd hi nrtrs
of expectant maxinnmi pollutant  con-
ccntraiions. It Is esix-etec1 tbrxt refeionce
and equivalent mc-!''c«ds pre applicable
to a'i iiinbient air environment' vs,lnch
are likely to be  cn^ouKt^rcd. Ti;o Ad-
i.ii:-jst;s.tor  rctr-.l!.; ihc dlicretio-.i to ?c-
lect  tr.mpJJng sites  for tes.t.5 be clcf-er-
miuc ? r.e«-ssari' to conduct.
   (b)  Test  ctr.iospf'.src.  Unf'itvjred CJA-
biont air slisll be used for the test How-
                                           ever, when  necessary  the ambient air
                                           may be spiked usinj a pollutant genera-
                                           tor.  The  ambient air saBipie shall not
                                           be altered in a:iy way except by spikins
                                           to brins l>:s  pc.Hutiii;t level i'!> to spec-
                                           ified test concentrations and/or by ad-
                                           dition   of  potential  Inter ft'iiis   sub-
                                           stances novm.'tlJy found ir. an;b;cnt air.
                                           The intake  and distribution :uarsifo!d
                                           shsll bs de;-!?r.cd so as not to ren»ove
                                           particles or trace jwsfa rnci i:L$urc that
                                           idcntli.-al p..!'v4^c:i rf.icr. trie reference and
                                           cr.ndid.uc  Piii;5;i!ins  :r.stn.ui:»-ri.s. Wh^n
                                           spikiv.s is i:.-cd. t-.t leas; 60 pcreer.t of the
                                           si>rj;..!c p..-•:.••:iUed TO the hJ.v;:-<:mf>ni s!;=il
                                           be iimV.'.-pl air  ana the rc;:!"-.i'u:cr  CMI
                                           oriclnj-tp from the j->oliutn;':L £c;^c:-ator.
                                           Schematic dro.v.inns :in«l complete  rro-
                                           codiu-E,! details of the test s;. stein shall
                                           be provided.
   <( >  in mill t'- ;;'*. ••; s f::ni]}Ii''-':. Tfii.- ;.;i:ly
.-!i;-.: coi:duct .- ii.i'ii'ancou.s s'tiy.oiii:,? oi
the;  test atmo.-'i>iirre by  both the refer-
ence and candidate methods. Ail rerord^r
chart  tracings and other documentation.
resulting from conduct of tests shall be
identified and submitted along \vith test,
data.
   id)  Operation  o/ measurement  ii/s-
tcms. Measurement systems shall be op-
erated in the range specified at  | 53.20
(d)  (Table B-l).
   (e)  Demonstration o/ consistent rela-
tionship.  A consistent  relationship is
demonstrated when the difference in re-
sults between candidate method and ref-
erence method meet the tolerance values
specified in this subpart.
§ 53.31  Tost  procedure   for   u\ii!ant
   (a)  Calibration:
   <1)  Calibrate  the  reference  instru-
mental method prior to conducting the
tests  in  accordance  with  procedures
specified in the appropriate appendix to
40 CFR Part 50.
   (2)  Calibration  of the candidate in-
strument shall be  done  in  accordance
with a method specified  by the mar.u-
facturer.
   (3)  Calibration data, procedures, snd
calculations  for both the reference and
candidate methods shall be submitted.
  NOTK.-- Calibrations  sha:'.  he  per TH: r«>ti
only once every 72 bci'.ru (p|:f ratioiial period)
for bolh methods. Manual :iUJusnn.:i.t 10 the
Instruments shall riot b? permli-sibie cl'irlng
the  f^atccl operational poric*!. Contro'  sam-
ples shall bo ir.trodivTC'.i to bol!i the reference
and.  cn!)dic.."ite  icsti-.in-.cats  r.r  cjj'.-ratioii.il
checks shall bo p!_rio.ri>.«: dAirir.g s;i o;:ora-
tior.:;: period.

   (4)  Both the ref.rreijce and canciiuiite
Instruments shall be operated  coutiiiu-
ously during an operational period.
   (b)  Ambient air tests:
   (1)  Tests over 1-hour avcrapin.™ ti-r.es
shall be conducted  at three ambient, air
or s;:-:I?ed ambient air oxidanl concentra-
tion ranges: O.OC-1.1 ppm, 0.15-0.25 ppm,
and 0.35-0.-15 ppm. Hourly avorngc.- r-si-
dant concentration es meaauied by the
reference method  shall  fall vithr.i the
range specified. 1'hfc hourly avevaiie shall
be based upon integration of the output
signal over the en lire hourly span.
   (2)  Description  of the methods  u-ed
to  obtain  hourly   averages  shall  be
si'brjUled.
   (3)  The following tests shall be  per-
f«-.:-!n-.-d. Tab'e C-3 of 5 53.31 paragraph
-f-  for the fir*:- sample sot i.~ 14 nr.d the-
a!io.-.::blf number of times that sp?cifl-
cal•<';•;a can be exceeded Is G. fj;t!i:!cr!y,
the s:/c of the second sample set, if re-
quired, is ir. and the nilowrvble in^r.l-er
of times thfit  specifications can be  tx-
Cixicd  for  the  combined total o: bxh
samples (32) is 2.
                                         fEGSSUK, VOL. SC,
                                                             .  W— !iV.'.>.\Y, OCIC21K It,
                                                            94

-------
                                                   PROPOSED RULES
                                                                                                                  28 U7
(c) TF.«T
         t iFK'.tTiONS von OXU»AMI
          VONOXIDE: TABIK C-l
rolluteut raage
(p.p.m.)

Orldnnt:
1. Low 0.00-0.10—
2. Med.O.li-0.25..
3. High 0.035-0.45.
Totals 	
Carbon mouoiide:
2. M«d 20-30
3 High 35-15

Totals

Tolenwicc,
i.e. ikllow»iic»
betwetu
C'auJ. »'id
Kol. niertois
(p.p.m.) i
<0.02
<0.03
<0.04

<1 5
<2.0
<3.0



Remp:
1st
S
5
4
14
5
6
4

14

to site
Vi
B&mpla
6
6
6
18
6
6
6

18

  Allowable nr.tiYier c! diSwerrra wbicti can ciweJ
tolerance speeiCcalioui for tacU pollutant.

  (d)  Sampling procedure: Datr.  shall
be  coUected according to  a stratified
sampling  plan. Sampling  shall be con-
ducted in sets of three ranees. Within
each set sampling without replacement
shall  occur. For  example,  if the first
range selected  ii low,  then the next one
Is cither medium or high.  Once the sec-
ond range is selected, the third one  is
automatically determined. In  the  above
example the first one is low. if the sec-
ond js high, then tht- third  must  be
medium. This completes the first  set of
three. In starting the second set oi three,
the fourth range cannot be tho same as
the third. So if the third ranee v as me-
dium,  then  the fourth shall  be  cither
lov,' or high. This begins the secund. set
of three ranges. If the fourth run&e was
high, then the fifth can be  either medium
or low. Thus,  oncp ih» -n't1: :v:i£c hna
be?" "lr.~.trd.  His  ;.l.\Lh range i.3  cicUi-
mined. This procedure is  followed  to
obtain H sample size  of 14 for the first
saniple and  a sample size of IS for the
second sample, if  needed. As an addi-
tional  constraint on the sampling plr,n,
at least 3 and no  more than  6 samples
per day shall be  collected ir* each of
three successive days. A  sampling p!nn
5s specified  in § 53.31 > size Time rcnuirc.'iinit falls t*tw?nn 5 oii
     of!«                10 days
 P!>i No 1    TIT..M L'JH. MKL. MLlI, Ml.
 Plbnko. ? ".. LMH. 1.1'M, I.IIM. H'.rL, HL.
 1-ian No. ?   . ifJIL, Milt, IIMI., Milt. 1!J,.
 l'orsi:-ipl«5ire Time n-ijuirencnl falls 1>C .>/..VM 7 suJ
     of IS               13 day*
 PianNo. <	 LM1I. MU1-,  HMT-, MI.U. l.MH,

 IVanNo. 5	u'l.M."'u.VL,  MHI., Mil)., MTU.,

 J'bnNo. 6	 l.MH. 'l.-UM,  ITI.M. I-'.UI, MI.II,
              J.HM.
 tnnwntralinn rs-.ir*. J)--
 SpiKifiCd m T»W-> C-l o' }.
  (f) Acceptance sampling plan, neccs-
sa.ry to perform tho statistical tests:
  (1) The first sample set shall consist
of an initial set of 14  measurements.
The results of these measurements shall
be interpreted as follows:
  (i> it the  number of measurements
from the series of 14 which fail the stated
specification does not exceed zero, pass
the  instrument for consistent  relation-
snip.
  (ii) If  the number of measurements
from the series of 14 which fail the stated
specification is three or more, reject the
instrument for consistent relationship.
  (2) If 12 or 13 of measurements from
the  series of  14  pass the stated specifi-
cation conduct a second set of 18  test
measurements. If the number  of meas-
urements failing the consistent relation-
shi!) tost  in  the combined total  of 32
test measurements is:
   (i) One or two, pass the instrument for
consistent relationship.
   (ii) Three  or more, reject the instru-
ment for consistent relationship.

§ 53.32   Test procedure for carbon inon-
      o\itle method*.
   (a) Calibration:
   (1) Calibrate the reference instrumen-
tal method prior to conducting the tests
in accordance v/ith procedures specified
in the appropriate appendix to 49 CPR
part 50.
   (2) The candidate instrument shall be
calibrated  in   accordance  with  the
method .specified by the manufacturer.
   (3) The calibration data,  procedures,
and  calculations  for  both  the  refer-
ence and candidate meiriods shall be
submitted.
  »;r.-,x.—C'-iibi*;-ioftS sV-.r.iV  t>£  permiVled
only once  every 72 hours (ojiCTfttional pe-
riod) for both, methods. V.Ana&l adjustments
to the instruments sliall not be permissible
during the stated operational  period.  Zero
and span  tests shall b-  conducted oo  both
the  reference ana candidat-? i'lstruments  to
document  that  Instruraenfs e.re operating
within j^.-formanre specifications.

   (4) Both the reference and  candidate
instruments  shall  be operated continu-
ously during an operational period.
   (b) Ambient air tests:
   (1) Tests over 1-hour  averaging times
shall br condiKted at threo ambieuf- air
or spiked ambient air carixva  monoxide
concentration rance.s: 7-11 ppm. 20-30
pp::i, and 3S-45 ppm. Hourly average CO
concent ration as measure; by  the refer-
ence method sh?ll fall v.itlu'n  the range
specified.  The  hourly avtiasc shrill be
basc-d unon continuous integration of the
ou.put signal over the entire hourly span,
or  shall  be the avt-racfe of  si le.ist 12
inslunt-tneous sLav-il levels which are at
equally spaced intervals over the entire
hourly span.
   (2) Description of the method \i>t-d to
obtain  liourly averases shall be submit-
 ted.
   (3) The following tests shall be per-
formed. Soi.Sl'.c)  delineates tlie munber
of ine:'.sui-.ts needed for each oC the
 concentre lion rAiiP.es f.nd th^ir respec-
tive tolerances. It ii.ciieatM Ui;.t tlie total
 sample si-^e for the first sample is ]-•( anj
the allowable number of times specifica-
tions that cf-n be exceeded is 0. Similarly,
the second sample, if required, is 18 and
the allowable number of tirne.5 specifica-
tions can be  exceeded lor the combined
total of both samples (32) i.s 2.
  (c) Sampling procedure:
  (1) Data shall be collected according
to a stratified sampling plan. Sampling
shall be conducted in sets of three ranges.
Within each set sampling  \\1thout re-
placement  shall occur. For example,  if
the first  range selected is low, then the
next one is either medium or high. Once
tlie second range is selected the third one
is automatically determined. In the above
example  the first one is low, if the second
is high, then the third must be medium,
This completes the first set of three. In
starting  the second set  of  three, the
fourth ran?e cannot be the same as the
third. So if the third range was medium,
then the fourth shall be either lov,- or
high. This  besins the second set of three
ranees. If  the fourth range was  high,
then the fifth can be either medium or
low. Thus, once the fifth ran-;o has been
selected  the  sixth range  is determined.
This procedure is followed to obtain a
sample stee of H for the first i-ample and
a sample size of IS for the second sample,
if needed. As an additional constraint on
the sampling plan, at least 3 and r.o more
than 6 samples per day shall be collected
in each of  three successive d:\ys. A sam-
pling  plan  is  specified  In Ji53.31(e)
(Table C-2). Table C-2 shall be used  to
select  either plan 1, 2, or 3 for a sample
sije of 14  and if arv  additional sample
of 18 is required, then either plan 4, 5, or
Gsliallbeuocd.
   (2)  Ti.: i-CCcpttaict;  sainn!ir,f<  ;>'.?.»>
necessary U> peviovm tne statistical tests
is specified as follows:
   (i) The first sample set shall consist
of an initial set of 14 measurements. The
results o£  thesa  measurements tUaU be
Interpreted as follows:
   (a)  If the mimber of measuieniems
from the 14  v.'hich fail the stated speci-
fication  does not exceed zero,  p.xss the
instrument for consistent rela'jouship.
   (b)  If the number of mciuutements
from the 14  which fail the stated speci-
fication  is tlu-ee or  more, reject the in-
strument  for consistent  relativ«*hip.
   (ii)  If the number of inea-'Wt'-ments
from the 14  which fail the stated .speci-
fication  is either one or two conuu.ct i%
second set of  JS te*t measurement.  If
the number of measurements f aUins the
consistent relationship test in the  com-
bined  total of 32 test measurements is:
   (a>  One or  two,  pass the  instrument
for consistent relationship.
   (b)  Three or more, reject the instru-
ment  for  consistent relationship.
§ 53.33  Te&l procedure for sulfur tliox-
     irlr  methods.
   A manual reference meitiod has been
published  in 40 CPR Part 50. Tne  party
shall demonstrate ? consistent  relation-
ship between the candidate method and
the reference method in accordance with
the .following  procedures.
   (a>  Automated  methods.—(1)  Cali-
brc'.ion.    u)   Calibrate  the  reference
                                        r.EGISTEf!, VOL. **, NO. 177—FKiOAY, OC.OGCR \>, 15X3
                                                              95

-------
284-18
           PROPOSED RULES
method In  accordance  with the proce-
dure specified in the appropriate spjxm-
dix to 40 CFB Part 50.
  (fi>  The candidate automated method
shall  be calibrated in  accordance viih
the procedure specific f.  by the manufac-
turer.
  (iii) The calibn.ijc:: data for both the
reference and canOjcInte methods and,
the calibration procedure and calcula-
tions for the candidate method shall  be
submitted.  NOTE.—Instrument calibra-
tion shall be permitted only once  every
72  hours (operational  period) for the
automated method. Mammal adjustment
to the instrument is not permissible dur-
ing the stated opera.t:c'nal period. Con-
trol samples shall be introduced to the
candidate   instrument   or  operational
checks shall be perfoj .;;ed during an op-
erational period  to  document  that the
Instrument is operator within perform-
ance   specifications.   Control  samples
shall   also   be used in  the  reference
method to verify that the method is  in
calibration and that vu-:d data are being
obtained.
  (2) Operation  of ll.c  instruments. The
candidate  instrument shall be opera-led
continuously during U;e operational p2-
riod.   Simultaneous  measurements  of
pollutant  cor«ccntrs».Mo3;s shall bs ob-
tained l-y  means of the. reference *nd
candidate methods lo;  the coacenli-aUoii
range^. nnd averaging  t.;aies specified Jn
J 53.340;)   (Table C-&;.  On any  fci-.cn
day urOy one 24-hour test result shul) be
reported  for a  specified concentration
ranr.c-. In the lange 0.4'—0.50 jipni, 1-hour
Mid 24-hour measurements shall be ob-
tained concurrently. O:i?-ho'.ir &w?.f"C
rmiriu;  r.h».;«inro   in-   ui«   ailtonuUfti
method shall be based on continuous ia-
tcgralion of the outpot Mgr.al over the
1-fcour span or &hall be the average of «t
least   12  Instantaneous  signal  level*
vhich aic  ftt equally  spaced  intervals
over the  1-hour  spun.  In like manner,
results of 24-hour measurements shall be
obta-ined.
  (b)  Test specifcati -:.? /or tul/ur di-
oxide. (1) The acci-ptn:  ce sampling p!sn
necessary to perform i';.e  ftatitUcal tests
Is srecilied as fallows:
  <1)  The- first- sample set shall consist
of en initial tr.l of 7 measurement*. The
results of  these  mew.i-cHiraits shnlJ he
Interpreted as foJlo\vs:
  (a)  If  the nuittbfr of  iiy-asurenwjj**
Irom i>!« i-trics tf 1 v.hich fall the st^tM
specifica'.ion does not exceed zero, pat-s
the iii.;-uumr:it for consistent relation-
ship.
  (b)  )T  the number of  nu3>v;:rj:.?r.ts
from the series of V V..-K-U ia« the st.-irvJ
specification is three or niv.o. reject the
Instrument for cor.Msu-ai  reJptlrwhi;;.
  (lit  If  the number  of  inca.iiirciiifiit-v
fro;n the rc-ries of 7 vr.UJx iaJi the M-siMl
specification is cither one or U'.o con»!'irt
a second set of 8 te*', meoiViMwienlr-.  Jf
the nujr»--er of mc.isi:i t rvi^its fa;ii):g the
consiK'tnl rtlution»hi:i-ttj;t in the  com-
bined  tot.il of  15 le.a :;j«>.suj-e!i;n!s i>:
  ti)  One  or  two,  pu.<-« tlie i;:-triiuic-nt
for cons'strjrt  re5-".tior.>tiip.
   (1>>  Three or more reject the instru-
ment for consistent relationship.
   (2)  (i) Data for 24-hour comparisons
shall be collected according to the sam-
pling  plan given  in S 53.33 (c)  (Table
C-4).  Table C-4 shall be used to select
either plan 1,  2, or 3 for a sample size
of 7 and if an additional sample of 8 is
required, then  either plan 4, 5, or 6 shall
be used.
   (ii)  Sampling shall be conducted  in
sets  of three  ranges  as specified  in
§53.33ntiRlion RMir.r, ns a)>ccified li 'I'abte  C-3
 o.' «£?.33(t).
Table  C-3  iDdic«.tcs  that  the   tot
sample size for the fust sample k 7 rr.
the allowable number of times sp-j-jinj-,:
tioiis can btj exceedeo j; 0. S!mil?.rl\- ;•-.
second rninplc. if rcfjuircci, is 8 rov.; v
allov.r^blc number of  tin-its specific   :.
can be exceeded  for  ti:» combined lotn
of both .« ainples (15; l~- 2.
   (2) The  wcceptarice  s^ntplins  p:;i
ni-ccs<;:uy  to  pc;-l<;'ni  the stati.sticr.
te^.ts ij specified a°. ,'o;j>.r.vi-;
   (O The fii-cl saiopl*1 set shall co.'vis
of ar> V-iiilr1 Fet of 7  ? . :ftr.Fur&nient^. 'i .i
   ta; Aiantia
   (1) Calibration:
   (i)  Calibra't ti:c rcfert-ncc method in
 nccoi clance \vith the procedure sviccilicd
 for  Uie reference  uicthixi in 40 CFR
 Part 1.0.
   (ii) The  candidate  manual uicUiod
 shall  be  calibrated In  accordance with
 the    pioccuv-ra    specified    ty    the
'manufacturer.
   (Hi) The i-r.l&ration data for both the
 reference  ar.a  candidate  incthoSs  end.
 the  c:'ilibictioT\ piocediu'e rnd '•a-.ciua-
 tions for the candidate meil-.s-r: sb;-.li be
          .
   (iv) A Uc.tiiii-.d proc..c'iir.--5 thseriptloa
of the candi'ir.vc metl-.od shall be sub-
mitted,
   (v)  SiniuMwieous  lucasuieuients   of
jx»n;5tant  c<.:u-entration  by  jaor-ns  of
      -i.re aivi candidate ;aetJ:o;ls for the
    i-irr. coj-u-e'.waUoii vwr.fo* ar-.d tver-
           .
 (T«b'o C-3 1  :hr.U be  ob:aii5cd.  Not.--:
 Conlrc.1  s.'J«pit.s nir..\ bo  lu-;:'. *.o dofi-
 ini-o' liir. v-ii'.dii.v of the ie::oUs. On sr.y
 fher.  clay only e/ie 24-houv  U^ rebuilt
 shall be rcpovt'-c! for a spc-cin«l coaccn-
 trrli-n  iT.rc-e.  In  tlie  ran?.f. O.ii'-C.r-O
 pp:n. test rfi.!:- foi b-.iiU 3-l;o.i; uvei-
 asir." ti?nes wuJ 2-!-ho';!' r.vcr. ^i:ig tir,-.e'
 shuU be oVV:*:nrd in a s-vcn cjy.
   fv:< f-c;ior.  f.3.o;?-M   (Tai'le  C-3)
 dt'inrr.lc» I"";" i.ur-.-orr of  mcas^rcscejiJs
 r.f-'-i'-ci' for <:.?h i-i the SO.  i;.n?es r.nci
 thci.- rorx-c'iv-e t^icrnujes for t^ih 24-
 hcur i);cr.nirc:.-t;.'s u:id  l-lun:r  mo«-
 urcrr.or.ts.
interpreted as follow:. :
   (H)  If  th» numfcrr. of measure mc^t-
fj-om the 7 which fi'.ii the Ktateci sj.c< ;r>
cf.tion docs not e>:cc» -.1 /ere, pass ti:e i "-
strumcnt for consisU-!i!,
   (b)  If  the number of
from the 7 which la;! the ststed spocia-
ca;Jon is three or ino;?, reject the  in-
stiianc-iit for co->.=i-tpnt relation ship.
   (ii)  Jf  the iiumljtr of mcasurejrioius
frora the 7 which foil fhe slated s.oc '.•:,]-
ration is cither one or two conduct a sec-
ond set of  8 test moasurenitrits. IT tiie
number  of  ineasnTc-nicnts  lailin?  the
consistent rt^ntior.^hJp test  in the com-
bined tot;\l of  15 tt-jit measurements i-.- :
   (n) One  or two, p-.ss the instruTiu---.il
for consistent relationship.
    (T..!.r
C— i> . ,S '::ipl'iv.; s}!:>.;i be conductt-d in ici.s
of four is.ii.v-es as sjw,?if,-n  in § f'3.32 •>:•'•
(Tab'c  C-3>  ar.d,  vithin each   stv.
sanr,.l:nsr  vitiiout.   rc:>'nccmcnt  f^i'.'M
ocruv. J''or example-, !:  the  firit i.T:Le
seU-jiCit  ts  lo\v,  the  next one  is cither
niccuiir.i. high or \c-ry ];i?h. \Viien o'otair. -
in^ 3-hour mefiSUrcnK-uis,  the- dstr. p.: e
roliecic-d  eonseculivcly  over  the  hi;-!:
ranre 0.-10-0.50 vith no more than four
3 • ii.ii"- measurements in a ^ivrn 24-hou:
   R Doc.TJ-2;3-lI> Piled lC--ll-7i.8:45
                                         iGisrrx,  VOL sr, \o. 197—WDAY,  OCTOEFS 12, J9/-3
                                                              96

-------
                APPENDIX II






  Air and Industrial Hygiene Laboratory's









             Recommended Draft






                    of






 SUBPART B:  Test Procedures for Measuring




        Performance Characteristics






                  of the






PROPOSED AMBIENT AIR MONITORING EQUIVALENT




           AND REFERENCE METHODS
                       97

-------
        Subpart B - Test Procedures for Measuring Performance



        Characteristics



153.20      General Provisions




            (a)  Test procedures applicable to automated  candidate and



            reference methods for observations on Interference, Lower



            Detectable Limit, Precision, Rise Time, Fall  Time, Zero




            Drift, Span Drift, Lag Time and Noise are described in this



            Subpart.  The methods must pass all the specifications for



            performance to  satisfy the requirements of  this subpart.  Measure-



            ment systems must be operated  in  the range  specified  in Table  B-l




            of paragraph  (d) of this  section  and at normal ambient temperature




            of 20 to  30°C and normal  line  voltage of  105  to 125 VAC during the




            conduct of these tests.



            (b)  All  recorder chart tracings, test  data and other documentation



            obtained  from these tests shall be Identified, dated  and  signed by




            the analyst performing the test,  and submitted.



            (c)  Performance specifications are delineated in  f(d) of this




            section.  For each performance specification  for range, noise,



            lower detectable limit and interference equivalent, perform the



            tests as  directed in the  test  procedures.   The method or  instrument



            must pass all these specifications before other tests are begun.



            For each  performance specification for  drifts, response times  and



            precision, an initial of  seven (7)  tests  shall be  conducted.  The



            results of these tests shall be interpreted as follows:
                                           98

-------
(i)  If the instrument passes all seven (7) tests for a




given performance specification, pass the instrument for



that specification.




(ii)  If the instrument fails three (3) or more of the




seven tests for a given performance specification, reject




the instrument for  that specification.




(iii)  If the instrument  fails one or two of the initial




series of seven tests  for a  given performance specification,




conduct a second series of eight tests  for that performance




specifications.  If the number of tests results failing that




performance specification in the combined  total of 15 test




results is:




     (a)  One or two, pass the instrument for that




    specification.




     (b)  Three or more, reject  the  instrument for




     that specification.
                               99

-------
                                     (d)   Performance Specifications for Automated Methods:  Table B-l
o
o
 1.  Rangeb

 2.  Noise, 0%
          • 20%

 3.  Lower Detectable Limit

 4.  Interference Equivalent

         Each Interferent

         Total £f Interference

 5.  Zero Drift

         12 Hourc

         24 Hour

 6.  Span Drift, 20%
               , 80%

 7.  Lag Time

 8.  Rise Time, 95%

 9.  Fall Time, 95%

10.  Precision

         20% of Upper Range Limit

         80% of Upper Range Limit
Jnits*
IMMMMHInil*
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
minutes
minutes
minutes
ppm
ppm
SQ2
0-0.5
0.005
0.005
0.01
± 0.02
± 0.02
± 0.02
± 0.02
± 0.02
20
15
15
0.02
0.03
£3
0-0.5
0.005
0.005
0.01
± 0.02
± 0.02
± 0.02
± 0.02
± 0.02
20
15
15
0.02
0.02
00
0-50
0.50
0.50
1.0
± 1.0
± 1.0
± 1.0
± 1.0
± 1.0
10
5
5
1.0
1.0
 aTo convert from ppm to
  weight of the gas .
                                               at. 25°C and 760 mm, multiply by M/. 02447 where M is  the roolecular
                 No  performance  test  required.   All other performance spec: f : .-ations are tested on  instrument
                 operating  in the range  specified.
                  TJ&
                             t>y nve aa ti -tc in
                                               adLir c=OT\t
                                                                fox- 3. 2

-------
1F53.21      Test Procedures for Automated Methods




            (a)  Test Facility




                A test room capable of providing controlled  temperature to




            within ± 2°C at ambient conditions will be required for the gas




            generation system  (fl53.21(b)).  A second  test  room or chamber capable




            of providing temperatures at 20 and  30°C  and within ± 2°C will be




            required for the instrument  under test  (See f53.22(g), Table B-4).




            (b)  Generation of  Test Atmospheres




                Methods for generation of test  atmospheres and for verification




            of the test concentrations are given in Table  B-2 of  this paragraph.




            The test gas delivery system shall  be  designed and constructed so as




            not to alter the test gas  composition.   The  delivery  system shall be




            fabricated  from borosiUcate glass  or FEP Teflon.  All  diluent air




            shall be free  from contamination.   The temperature surrounding the




            gas generation system shall  be controlled within ±  2°C.   The  output




            of  the gas  generation system shall  be stable in order to  obtain




            valid results  from the tests.  If a permeation device is  employed




            during generation  of the test atmosphere, the device  as well  as  the




            air passing over jit shall be controlled to ± 0.1°C.   Samples  for




            verifying  test concentrations shall be collected utilizing the same




            delivery  system as close as  possible to the sample intake port of




            the automated method under  test.  The concentration of each test




            atmosphere shall be verified by analysis of duplicate samples before




            and after each series of tests.  Schematic drawings and complete




            procedural details of the test gas generation and delivery system




             shall be provided.  A suggested system is shown in Figure 1 and




             as described in 1153.22(d) (iii) .
                                            101

-------
                          (c)   Test Atmospheres:   Table B-2 (References listed at end of table)
         Test Gas
                   Generation
1.   Ammonia
2.  Carbon Dioxide
3.  Carbon Monoxide
4.   Ethane
5.  Ethylene
6.  Hydrogen Chloride
7.  Hydrogen Sulfide
Permeation tube system.  Similar procedure
to that described in reference 4.
Cylinder of zero air containing 2X the con-
centration of C02 specified in Table B-3.
Cylinders of zero air containing 2X the
concentrations of CO specified in Table B-3.
Cylinder(a' of zero air containing 10 ppm
ethane.  Dilute with zero air to concen-
tration specified in Table B-3.


Cylinder(a) of "prepurified" nitrogen con-
taining 20 ppm ethylene.  Dilute with
zero air to concentration specified in
Table B-3.


Cylinder^3' of "prepurified" nitrogen con-
taining 100 ppm of gaseous HC1.  Dilute with
zero air to concentration specified in
Table B-3.

Permeation tube system, p. 426, reference 3.
                Verification
Collect samples in bubbler containing 0.1 N
HCl.  Analyze samples by the Phenate method,
reference 1, after neutralizing excess acid
with 0.1 N NaOH.

Certified analysis by manufacturer or vendor
and recertified to be within ± 5% by an
independent laboratory just before use.

Certified analysis by manufacturer or vendor
and recertified to be within ± 5% by an
independent laboratory.  Recertification
required every 6 months.

Certified analysis by manufacturer or vendor
and recertified to be within ± 5% by an
independent laboratory using flame ionization
detector calibrated with methane.

Certified analysis by manufacturer or vendor
and recertified to be within ± 5% by an
independent laboratory using flame ionization
detector calibrated with methane.  Recerti-
fication is required every 6 months.

Collect samples in bubbler containing
distilled water.  Analyze samples by the
mercuric thiocyanate method, ASTM (D5.2),
p. 29,  reference 2.

Tentative method of analysis for ^S content
of the atmosphere, p. 426, reference 3.

-------
                                            Table B-2 Test Atmospheres (Con't)
          Test  Gas
                   Generation
 8.  Methane
 9.   Nitric Oxide
10.   Nitrogen Dioxide
11.  Ozone
12.  Sulfur Dioxide
13.  Water
14.  Xylene
Cylinder^ of zero air containing 2X the
concentration of methane specified in Table
B-3.
Cylinderof "prepurified" nitrogen
containing 100 ppm NO.  Dilute with zero
air to concentration specified in Table B-3.

1.  Gas phase titration - description available
from NERC-RTP.
2.  Permeation tube system - procedure similar
to that described in reference 4 and modified
by NERC-RTP.

Calibrated ozone generator.  Procedure
described in reference method for photo-
chemical oxidant, reference 4.
                                                                    Verification
Permeation tube system.
in reference method for
                         Procedure described
                         0» reference 4.
Pass zero air through distilled water at a
fixed known temperature between 20° to 30°C
such that the air stream becomes saturated.
Dilute with zero air to concentration
specified in Table B-3.

        (a)
Cylinder v ' of "prepurified" nitrogen con-
taining 100 ppm xylene.  Dilute with zero
air to concentration specified in Table B-3.
                                                    Certified analysis by manufacturer or vendor
                                                    and recertified to be within  ±  5% by an
                                                    independent laboratory using  flame ionization
                                                    detector calibrated with methane.  Recerti-
                                                    fication is required every 6 months.

                                                    Tentative method of analysis for nitric
                                                    oxide content of the atmosphere, p. 325,
                                                    reference 3.

                                                    1.  Reference 4, Appendix D; 03 is
                                                    titrated against NO which is equivalent
                                                    to N02 formed.

                                                    2.  Reference 3, page 329.
KI method.  Procedure described in
reference method for photochemical
oxidant, reference 4.

P-rosaniline method.  Procedure described
in reference method for S02» reference 4.

Measure relative humidity by means of a
dew point indicator, calibrated electro-
lytic or piezoelectric hygrometer or wet/
dry bulb thermometer.
                                                    Certified analysis by manufacturer or
                                                    vendor and recertified to be within ± 5%
                                                    by an independent laboratory using flame
                                                    ionization detector calibrated with
                                                    methane.

-------
                                            Table B-2 Test Atmospheres (Con't).


  	Test Gas	   Generation	  	Verification	

15.  Zero Air                Air having a water content of less than 1000 ppm,   Certified analysis by manufacturer or
                             hydrocarbon content as methane of less than or      vendor,  and recertified to be within ± 5%
                             equal to 0.1 ppm,  carbon dioxide content of 350     by an independent laboratory using flame
                             ± 30 ppm and oxygen content of 21.0 ± 0.5%.         ionization detector calibrated with methane
                                                                                 and infrared detector calibrated with carbon
                                                                                 dioxide  and by using appropriate analytical
                                                                                 techniques for oxygen and water content.



 'a'Use stainless steel pressure regulator.

 Ref.  1 - Tentative Method of Analysis for  Ammonia in the Atmosphere (Indophenol Method), Intersociety Committee Method
             42604-01-72T,  Health Laboratory Science, Volume 10, No. 2 (April 1973)
 Ref. 2 - 1970 Annual Book of ASTM Standards. Part 23, American Society for Testing and Materials, 1916 Kace Street,
             Philadelphia,  Pennsylvania.
 Ref. 3 - Methods of Air Sampling and Analysis. Intersociety Committee, 1972, American Public Health Association, 1915
             18th Street, N.W., Washington,  D.C.
 Ref. 4 - Part 50 - National Primary and Secondary Ambient Air Quality Standards, Federal Register. November 25, 1971,
             [36 FR 22384].
  o
  .e-

-------
(d)   Start-up, Calibration, and Safety Requirements




    (1)  Set-up and start-up of the instrument shall be




    in strict accordance vli_n the instrument's instruction




    manual.  Connect the output to a suitable strip chart




    recorder.  (Note:  This requirement is not intended to




    limit data acquisition components.  Other components




    may be used along with the recorder during conduct of




    these tests.  It is intended only to facilitate evaluation




    of data submitted to EPA for the purpose of demonstrating




    that a candidate method is an equivalent or reference




    method.)  The recorder shall be of the servo, null-balance




    type having a chart width of 10" or greater, response time




    of 1 second or  less, at least 5% below zero or  zero offset




    capability, and deadband of less than 0.25% of  scale.  Allow




    adequate warm-up or stabilization time as indicated in




    the instruction manual.




    (2)  Calibration of the instrument shall be as  indicated




    by the instrument's operation manual.  Adjust  the  instru-




    ment controls,  when possible,  to obtain a 5% offset zero




    reading  on the  recorder chart.  A dynamic calibration curve




    which  has at  least  7  identifiable points  including 0 and




    90% of full  scale shall be  submitted.  The  slope of the




    curve  shall be  determined by  the best-fit least square




    irethod and the  values  plotted  in  terms of pollutant con-




    centration against  chart  division or  percent of scale and
                                105

-------
                against  instrument  output  in  terms  of volts,  millivolts,




                milliamps  or  other  appropriate units.




                (3)   Normal  instrument operations or failures shall not




                generate or  present any hazards or hazardous conditions




                to operators or environment or discharge any hazardous




                or detrimental effluent gas.   Instruments shall be




                equipped with safety devices to prevent hazards and/or




                damage to  the internal components of such instruments.




K53.22      Procedures for Measuring Performance Specifications




            (a)  Range




                (1)  Technical Definition:  Nominal minimum and maximum




                concentrations which  the  system  shall be capable of




                measuring.   The nominal range which  shall be used during




                conduct of all tests  is specified  in units of  parts-per-




                million by volume  in  Table B-l of  1153.20(d).   Pollutant




                concentrations at  the lower and  upper scale  limits are




                specified, as for  example:  0 to 1 ppm.




                (2)  Test Procedure:   No  test procedure  is required.




            (b)  Noise




                (1)  Technical Definition:  Spontaneous, short duration




                deviations in the  instrument  output  about  the  mean output,




                which are not caused  by input concentration  changes.   Noise




                is  determined as the  standard deviation  about  the mean




                and expressed in concentration units.  The specifications are




                given in  Table B-l of U53.20(d).
                                           106

-------
(2)   Test Procedure:




    (i)  Allow sufficient time for instrument warm-up




    and stabilization.  Noise is determined using both




    zero air and a pollutant test gas concentration as




    indicated below.




    (ii)  For this  test, it is necessary  to disconnect




    the strip chart  recorder from the instrument output,




    and use instead  an  integrating  type Digital Meter




    (DM) accurate  to three  significant figures.




    (iii)  Sample  zero  air  for 60 minutes.




    (iv)  During  this 60-minute  interval, record




    twenty-five  (25) random DM readings.




    (v)  Convert  the individual  DM readings k]_, k2	k25





    to concentration units  c-^ c2	c25 by  reference to





    the previously determined calibration curve of  the




    analyzer, and determine the  mean by:
             c =
                         25
     Deviations about the mean are determined by
                           107

-------
        and the standard deviation or noise is,
                               24
        Report the noise at 0% of the upper range limit




        in ppm.




        (vi)   Repeat steps (iv)  through (v) using a




        pollutant test gas concentration producing an




        output of 80 ± 5% of the specified upper range




        limit and report the noise.




(c)   Lower Detectable Limit




    (1)  Technical Definition:  The minimum pollutant




    concentration which produces a signal of twice the.




    noise level.




    (2)  Test Procedure:




        (i)  Allow sufficient time for warm-up and




        stabilization.  Sample zero air and allow for




        stable reading to establish zero baseline.




        (ii)   Generate a test atmosphere equal to




        20 ±  5% of the upper range limit.  Verify the




        concentration with the reference procedure.




        Calculate and increase accordingly the diluent




        air flow  rate to obtain the concentration




        specified for lower detectable limit in Table
                            108

-------
        B-l.  Sample the gas stream.  Record the




        stable reading.




        (iii)  Convert the reading to concentration unit.




        Pass the instrument for this specification, if




        this value is equal to or greater than twice the




        noise at zero obtained in H53.22(b)(2)(v).




(d)   Interference Equivalent




    (1)  Technical Definition:  Positive or negative output




    caused by a substance other than the one being measured.




    (2)  Test Procedure:  The candidate instrument shall be




    tested for all interferents specified in Table B-3 of




    11 (e) of this section for the pollutant of interest.  The




    basis for the test is that each candidate method is




    challenged with those interfering agents known or suspected




    to cause interference.  The interference; may be either




    positive or negative, depending on whether the instrument's




    output is increased or decreased by the presence of the




    interferent.  Interference tests are performed by mixing




    each interferent with each of the pollutant concentrations




    specified in Table B-3 of fl(e) of this section, sampling




    the mixtures and comparing the instrument's output to the




    output caused by the pollutant alone.  In cases where the




    pollutant and the interferent are known to react in the




    gas phase as designated in Table B-3, the interference




    equivalent shall be determined in the absence of the pollutant
                              109

-------
(i)  Allow sufficient time for warm-up and




stabilization.




(ii)  For instruments using a prefilter or




scrubber based upon a chemical reaction to




derive part of its specificity, and that requires




periodic service or maintenance, the instrument




shall be "conditioned" prior to each interference




test as follows:




    (a)  Service or perform the indicated




    maintenance on the scrubber or prefilter as




    directed in the manual.




    (b)  Before testing for each interferent,




    allow  the instrument  to sample through  the




    scrubber a  test atmosphere containing the




    interferent at a  concentration equal to the




    value  specified in Table B-3 of fl(e) of this




    section.  Sampling shall be at the normal




    flow rate and shall be performed  for 6




    continuous hours  prior to  testing.




(iii)  Three test atmospheres  shall be generated  as




follows:




    (a)  Test atmosphere  P_:  Pollutant concentra-




    tion.




    (b)  Test atmosphere  li  Interferent concentra-




    tion.




    (c)  Test atmosphere  Zi  Zero air.
                      110

-------
   (iv)  Adjust  the  flow rates of P, I and Z




   as  follows:




   (a)   The  flow rates  of  test atmospheres




   I and Z shall be  equal.




   (b)   The  concentration  of  pollutant in  test




   atmosphere P  shall be  adjusted  such that




   when mixed (diluted) with  either test




   atmosphere I  or Z, the  resulting concentra-




   tion of pollutant shall be as specified in




   Table B-3 of  11 (e) .




    (c)  The  concentration of  interferent in test




   atmosphere I_ shall be  adjusted such that when




   mixed (diluted) with test  atmosphere  £, the




    resulting concentration of interferent shall




   be  equal  to the value specified in Table B-3




   of  paragraph (e)  of this section.




(v)   Mix test atmospheres P and Z by passing the




total flow of both atmospheres through a mixing




flask.




(vi)   Sample the mixture of test atmospheres P




and Z.   Record the stable reading.




(vii)  Mix test  atmospheres P and I by passing




the total flow of both  atmospheres  through a




mixing  flask.



(viii)    Sample the mixture  of test  atmospheres




P and  I.  Record the stable reading.
                         Ill

-------
(ix)  Convert the first and second readings to




concentration units, C^ and €2, respectively t




Calculate the interference equivalent. IE =
C2 "
(x)  Follow steps (iii) through (ix) to determine




the IE  for each  interferent.




(xi)  For those  interferents which  cannot be mixed




with  the pollutant,  as indicated  in Table B-3  of




fl(e)  adjust  the  concentration  of  the interferent




in the  test  atmosphere to the  specified value




without being mixed or diluted by any other test




atmospheres .




 (xii)  After stable zero, sample the interferent




 test atmosphere until a stable reading is obtained.




 (xiii)   Convert the net reading to  concentration




 unit and report as the interference equivalent.




 This value can  either be positive or negative.




 (xiv)  The sum  of the absolute values of the




 individual interference equivalents for the




 specified interferents shall  not exceed the value




 given  in performance  specifications [Table B-l of




 153.20(d)] for  the total interference.   (Note:




 Specifications  for  total interference  are  not




 intended to  indicate  allowable measurement in-




 accuracy at  the levels of  air quality  standards.




 The  allowable interferent  response stated  in  the




 performance specifications results from challenging






                            112

-------
(e)      Interfercnt Test Concentration for Automated Methods in Parts Per Million3:  Table B-3
Pollu-
tant
SO2
SO2
SO2
SO2
: s°2
S02
°3
°3
03
NO,
Measuring Hydrochloric
Principle Acid Ammonia
Flame Photometric
(FPD)
Gas Chromatography-
(FPD)
Spectrophotometric-
Wet Chemical (Pararo-
sanih'ne Reaction) 0.2 O.lb
Electrochemical 0.2 0.1 b
Conductivity 0.2 0.1 b
Spectrophotometric-
Gas Phase
Chemilumine scent
(other than ref.
method)
Electrochemical O.lb
Spectrophotometric-
Wet Chemical (po-
tassium Iodide
Reaction) 0.1 b
Spectrophotometric-
Gas Phase
Chemiluminescent O.lb
Hydrogen Sulfur Nitrogen Nitric Carbon Ethy- Water Carbon
Sulfide Dioxide Dioxide Oxide Dioxide lene Ozone m-Xylene Vapor Monoxide Methane Ethane
0.1 0.14C 750 18,000b 50
0.1 0.14C 750 18,000b 50
0.1 0.14C 0.5 750 0.5
0.1 0.14C 0.5 0.5 0.2 0.5 18,000b
0.1 4C 0.5 750
0.14° 0.5 0.5 0.5 0.2
O.lb 750 0.08C 18,000b
0.5 0.5 0.08C 18,000b
0.5 0.5 0.5b 0.08°
0.5 0.5 0.5b 0.08°
0.5 O.lc 0.5 18,000b

-------
                                              (c)
                                             Interfcrcnt Test Concentration for Automated Methods in Parts Per Million":  Table B-3 (Continued)
Pollu
tant
NO?


NO2
Measuring
Principle
Spcctrophotometric
Wet Chemical (Azo-
Dye Reaction)
Electrochemical
Hydrochloric Hydrogen Sulfur
Acid Ammonia Sulfidc Dioxide


0.5
0.5
Nitrogen
Dioxide


O.lc
O.lc
Nitric Carbon
Oxide Dioxide


0.5
0.5
Ethy
lene Ozone


0.5
0.5
Water
m-Xylene Vapor



18,000b
Carbon
Monoxide



50

Methane Ethane




NO,
Spectro photometric-
Gas Phase
                                                                             0.5
                                                                                         O.lc
                                                                                          o.s
                                                                                                                                    0.2
CO

CO
CO
CO
CO
CO
THC/
CH4
IR (other than
reference method) 759
Gas Chromatography
FID
Electrochemical 0 5 0.2
Catalytic Combustion-
Thermal Detection 0.1 750 0.2
I!" fluorescence 750
Mercury Replaccrnent-
UV Photometric 0.2
Catalytic Combustion-
FPD 0.2

18,000
18,000
18,000
18,000
18,000

18,000

10C
10C
10C
10C
10C
10C
50


0.5

5.0 0.5
0.5
0.5
2.0C 0.5
Notes:   Concentrations of interferent listed must be prepared and controlled to i 5% of the stated value.
          Do not mix with pollutant.
         cConccnlration of pollutant used for test.  These pollutant concentrations must be prepared and conti oiled to i 5%of the stated value.

-------
    the measurement system with maximal concentration




    of potential interferents.  If ambient air contained




    the combined concentration of challenging agents




    specified for the purpose of these tests, it is most




    likely that the concentration of the concentration




    of the pollutant of interest as reported by the




    measurement system would far exceed the air quality




    standard.)









(f)  Test Procedures for Zero Drift, Span Drift,




Precision, Rise Time, Fall Time, Lag Time




    (1)  Determination of these response elements shall




    be accomplished over a period of at least seven (7)




    days.  During  the test, manual  adjustments  to the




    electronics, gas or reagent flows  shall  be  permitted




    only once every three days.  Automatic adjustments




    which are a part of the normal  instrument operations




    are permitted  at any time.  A sampling plan is  given




    in Table B-4 of fi(g) of  this section.




    (2)  Technical Definition of  Zero  Drift: The  change




    in instrument  output over a stated time  period  of




    unadjusted  continuous  operation, when the input




    concentration  of pollutant is  zero.




    (3)  Technical Definition of  Span Drift:  The  change




    in instrument  output  over a stated time  period of




    unadjusted  continuous  operation,  when the input




    pollutant concentration is a stated upscale value.
                              115

-------
(4)   Test Procedure for Zero and Span Drift:




    (i)   Allow sufficient time for instrument




    to warm up and stabilize at normal line voltage




    of 125 VAC and normal room temperature of 20°C.




    (ii)  Sample zero air until a stable reading is




    obtained, and adjust baseline to 5% of chart




    [see H53.21(c)(2)].




    (iii)  Sample test atmosphere of pollutant




    concentration equal to 80 ± 5% (Span 1) of the




    upper range limit.  Adjust span control as




    necessary.  Rezero the instrument.  Where the




    zero and span controls are not electrically




    independent, the adjustments may require repeated




    trials.




    (iv)  Sample test  atmosphere of pollutant con-




    centration equal  to  20  ±  5%  (Span  2) of the  upper




    range limit and record  the output.  Start Day 0.




    (v)  Hereafter, no manual adjustments  shall  be




    made to  the instrument, except as  indicated  in  the




    test plan, and then only  after the zero and  span




    readings have been recorded  for that day.  Allow




    the  instrument to  sample  ambient air or perform




    other tests during idle periods.




    (vi)  In the evening of Day  0. sample  zero air  for




    a minimum of 12 hours.  Each morning,  starting  on
                         116

-------
Day 1 determine the minimum and maximum readings during




this period and record the difference of the readings




in concentration units as the daily  12-hour zero drift.




Starting in the evening of Day 1 and each evening




thereafter, adjust the line voltage  and room temperature




according to Table B-4 before this test is begun.




(vii)  Each day, twenty-four  (24) hours after  the




previous day's readings and starting on Day 1,  sample




air and test atmospheres  containing  pollutant  concen-




trations equal to  20  ± 5% (Span  2) and 80 ± 5% (Span 1)




of the upper range limit. The test  concentrations  shall




be consistent  from day to day.   Record the stable readings




at zero and at each  span  concentration.




(viii)  Convert  the  readings  to  concentration  units.




Subtract each day's  values respectively  from that




of the previous  day  and  record  as  the  24-hour  zero




drift and  span drifts at  20  and  80%  of  the upper




range limit for  that day.




(ix)  Determine  lag  time, rise  time, fall time and




precision  as outlined in 1T53.22(f) (5)  to H53.22(f)(8)




after  the  zero and span drift tests  have been  completed




for  the  day.




(x)  Repeat steps  (vi)  to (ix)  daily for an uninterrupted




period  of  seven (7)  or fifteen (15)  days during which no




adjustments  to the instrument shall be permitted.   To avoid




testing  during weekends,  the tests may be performed in
                      117

-------
   three day periods where adjustments are made on Monday




   and determinations are made on Tuesday, Wednesday




   and Thursday or adjustments may be made on Tuesday




   and determinations are made on Wednesday. Thursday




   and Friday.  All adjustments to the instrument shall




   be made  at the normal line voltage and room  temperature




   of the previous day.




    (xi)  At the end of Day  7, examine the test  results and




    determine whether  to  continue or  terminate  further  testing




    as  directed  in f53.20.




(5)   Lag Time



    (i)   Technical Definition:   The time interval between




    a step change in input concentration at the instrument




    inlet to the first observable corresponding change




    equal to twice the noise in instrument output.




    (ii)  Test Procedure:




        (a)  Sample zero air until a stable reading is




        obtained.




        (b)  Switch to the test atmosphere used in




        f(f)(4)(iii) while simultaneously starting a




        stopwatch.




        (c)  Observe the instrument output and  stop




        the stopwatch immediately when the first




        observable change equal to twice the noise




        obtained in ^53.22  (a)(2)(viii) occurs  due to




        the step concentration change in the output.




        (d)  Record the  elapsed time as  the  lag time.
                          118

-------
(6)  Rise Time




        (i)  Technical Definition:  The  time interval between




        initial response and 95% of final  response after a




        step increase in input  concentration.




        (ii)  Test Procedure:




            (a)   Sample zero air until stable  reading is




            obtained.




            (b)   Calculate 95%  of  the stable reading in




            H53.22(f)(4)(iii).




            (c)   Sample the same  concentration (80  i 5%)




            used  in  1!53.22(f) (4) (iii) and use  a stopwatch




            to  time  the interval  starting with the  first




            observable  response and ending when the response




            equals  the  reading in (b) above.




             (d)   Record the elapsed time, and allow the analyzer




            to  continue sampling until a stable reading is




            obtained in preparation for the fall time test




             [see H53.22(f)(7)(ii) below].




             (e)   Repeat steps  (a), (c) and  (d) above after




             completion of fall time test.




             (f)   Report the average of  the values obtained in




             steps (c) and  (e).




 (7)  Fall Time



         (i)  Technical Definition:   The time  interval between




         initial response  and 95% of  final response  after a




         step decrease in  input concentration.
                           119

-------
        (ii)   Test Procedure:




            (a)   Calculate 5% of the stable reading in (f)




            (6)(ii)(d)  of the rise time procedure.




            (b)   While  the analyzer is at stable reading




            from sampling the concentration in (f)(6)(ii)(d)




            above, switch to sample zero air and use a stop-




           watch to time the interval starting with the first




            observable  instrument response due to the step




            change and  ending when the response equals the




            value calculated in (a) above.




            (c)   Record the elapsed time.




            (d)   Repeat steps (a), (b) and (c).




            (e)   Report the average of the values obtained




            in steps (c) and (d).




(8)  Precision




        (i)   Technical  Definition:  Variation about the mean of




        repeated measurements of the same concentration.  This




        variation is expressed as one standard deviation about




        a mean.




        (ii)   Test Procedure:




            (a)   Allow  sufficient time for instrument




           warm-up and stabilization.




            (b)   Sample zero air until a stable reading is




           obtained.   Do not record.




            (c)   Sample the same pollutant concentration of




            20 ± 5% of  the upper range limit as used in




           1I53.22(f)(4)(vii) .   Record the stable reading.
                         120

-------
(d)  Sample a pollutant concentration at




least 50% higher (30% of full scale) than




in U(c) above until a stable reading is




obtained.  Do not record.




(e)  Quickly switch to the pollutant con-




centration used in (c) .  Record the reading.




(f)  Add to the readings obtained in steps




(c) and (e) , the reading on the same day for




the span at 20 ± 5% of the upper range limit




in 1T53.22(f)(4)(vii).




(g)  Convert the readings to concentration




units and calculate the mean and the standard




deviation for the three values.




(h)  Repeat step (b).




(i)  Sample the same pollutant concentration




of 80 ± 5% of the upper range limit as used




in TI53.22(f)(4)(vii) .  Record the stable reading.




(j)  Sample a pollutant concentration 20% higher




than in 11 (i) but not to exceed 97% of full scale.




Do not record.




(k)  Quickly switch to the pollutant concentration




used in (i).  Record the reading.




(1)  Add to the readings obtained in steps  (i)




and (k). the reading obtained on the same day




for the span at 80 ± 5% of the upper range limit




in 153.22(f)(4)(vii).
              121

-------
(m)  Convert the readings to concentration units




and calculate the mean and standard deviation for




the three values.




(n)  Report the, values obtained in steps  (g)




and (m) as the precision at 20 and 80% of




the upper range limit, respectively.
               122

-------
Zero and Span Drift Test Plan3:  Table B-4
                                               Every Night6 Adjust Test
                                                     Conditions to:
Day
0
1
2
3
4
5
6
7
8
9
10
11
12
Morning
Adjust zero and Span 1 (80 ± 5% of range) . Sample Span 2
(20 ± 5% of range). Record the stable readings. No in-
strument adjustments permitted hereafter until after Day
3 readings.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2, 24 hours after Day 0 readings.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift, sample 0, Span 1 and
Span 2; adjustments may be made to the instrument, if
necessary, after readings are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2; instrument adjustments permitted after readings
are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2b
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1,
Span 2; adjustments permitted after readings are taken.
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Span 2
Determine the 12-hour zero drift; sample 0, Span 1 and
Voltage0
125
105
125
105
125
105
125
105
125
105
125
105

Temp, Ca
20
20
30
30
20
20
30
30
20
20
30
30

Day
0
1
2
3
4
5
6
7
8
9
10
11


-------
     Span 2; adjustments permitted after readings are taken.                        125	20	1.2
13   Determine the 12-hour zero drift; sample 0, Span 1 and
     Span 2	                     105        20        13
14   Determine the 12-hour zero drift; sample 0, Span 1 and
     Span 2	                     125        30        14
15   Determine the 12-hour zero drift; sample 0, Span 1 and
     Span 2; End of test		                     -	-	15.
 aThese tests shall be run only once each day.   Tests shall also be run either on consecutive days
  with adjustments permitted every three days or in increments of three consecutive days during
  which adjustments to the instrument are not permitted.  In the afternoon, other tests not
  requiring adjustments to the instrument may be performed, i.e., for rise time, fall time, lag
  time and precision.

 t>At this time examine test results to determine if further tests are required.

 cVoltage specified shall be controlled to ± 1 volt.

 ^Temperature specified shall be controlled to ± 2°C.

 eSample zero air continuously for a minimum of 12-hours.

-------
                                                                        A Ltno ^ p =•  i e
K)
Ln



1
FM
Z
MFM /

	

	 	 ., i
- -. I 1 '

SCP


t
R Aco
X
V s
KM ()
r
!
i

_j



» 4
P
i t
*- „., i
i



<

f\ /
f
i

t 	
i ^

1. H
£




1 J
1 '
I
1 I
1
[J MT
4
t A i

. B <

^ /


t

t

k
I
1


L
1 i!
L
pp


v /
\

i

i
^


5V
¥
I

^


) ' ^ o p !
1 f^U I1 1 I I L t
A O . . . k V >» k k k -s.
\±J \_J - ~ — 3 — y- — J y j r
TH MF PP
1 MIJMflfM
» vvyyvyvYYyyy
J Sampling Ports
FI
MI
Tl
* Ml


S\
sc
PF
' D
R
1 B A ' A
r\ ,
	 -"'I
I PF
4<
< TH
	 *^ - 	 	 	 ..... . <; _ (Ty 	 3k 	 p. .
t - C i

¥ i m ¥ " V
,ni, T .... , . ifc / ^ ....A.— ^ \/
TH NV ( ^ ^\d, A r T »
r R- LT LT t A
M M M t t ! A'n
PH 6 o
! 1 § gl

! [ ! | Cons tan L Temperature
•J LJ LJ i— J d Permeation Tube Sv

To Vent

LEGEND
rM - mass flowmeter
4 - flowmeter
- mixing flask
I - three-way-valve
r - MFM transducer

? - needle valve
' - shut-off valve
;R - strip-chart recorder
'. - permeation tube holder
- regulator
B - connections for 03
generator
- auxilliary pollutant
test atmosphere

To vent
'H



.
B« 	 0 	 	 - -, !
y i
1 s
1 i _
j
{,.., 	 .,., „,,, _ • 1 , -
ftatVi !> 'i Generator
s terns
                       Cylinders  of Zero Air

-------
                APPENDIX III




   Air and Industrial Hygiene Laboratory's




              Recommended Draft






                     of






SUBPART C:  Test Procedures for Demonstrating




           Consistent Relationship






                   of the






  PROPOSED AMBIENT AIR MONITORING EQUIVALENT




            AND REFERENCE METHODS
                      127

-------
Subpart C - Test Procedures for Demonstrating A Consistent Relation-




        ship Between Candidate and Reference Methods








H53.30  General Provisions








        (a) Selection of Test Sites:  The exact location of the




        sampling sites is left to the discretion of the applicant




        but shall be located in the areas of expectant maximum




        pollutant concentrations.  The Administrator retains the




        discretion to select sampling sites for tests he determines




        necessary to conduct.









        (b) Test Atmposhere:  When possible unaltered ambient air




        shall be used for the tests.  The tests shall be conducted




        at three pollutant  levels (low, medium and high) as specified




        in Tables C-l and C-3.  To obtain the specified test con-




        centrations, the pollutant levels of the ambient air may be




        increased by using  a pollutant generator or decreased by




        using zero air.  The volume dilution resulting from the




        addition of pollutant or zero air shall be adjusted such




        that the test atmosphere shall consist of at least 60%




        ambient air at all  times.








            Oxidant levels may be adjusted with an ozone generator,




        sulfur dioxide levels with a permeation tube system and that




        for carbon monoxide a cylinder containing 100% gas may be used




        (see Table B-2).
                                 128

-------
(c) Test Conditions:  The reference and candidate methods shall




be operated in the measuring ranges specified in Table B-l of




H53.20(d).  They shall be properly calibrated before the tests




are initiated.  All measurements  shall be  conducted at normal




temperature of 20 to  30°C and  normal  line  voltage of 105 to




125 VAC.









The sampling  system shall consist of  an  intake  and distribution




manifold which have been designed and constructed so as not  to




remove airborne particles and  trace gases  and insure that  identi-




cal samples reach the reference and candidate methods.  Schematic




drawings and  complete details  of the  system shall be submitted




by the applicant,









(d) Sampling  Procedure;  The test atmospheres  shall be simul-




taneously  sampled by  both  the  reference and candidate  methods.




The tests  shall be  conducted at three specified pollutant levels




and in accordance with the  sampling plan provided for  each test




procedure.  The  tests shall be performed in one or two sample




sets   The first  set consists of the minimum number of measure-




ments necessary  tu  pass or fail a method for consistent relation-




ship. The second set consists of additional measurements required




when  the method failed in the first set in one or two  specifications




only.  The results  are averaged over the sampling periods specified




 for  each tefct procedure.
                           129

-------
            All recorder chart tracings and other documentation resulting




        from conduct of the tests shall be identified and submitted along




        with the test data.








        (e) Demonstration of Consistent Relationship:  A consistent




        relationship is demonstrated when the differences in the average




        results obtained by the reference and candidate methods meet the




        tolerance specifications given in this subpart.








f53.31  Test Procedure for Oxidant, Carbon Monoxide and Sulfur Dioxide




        Methods.








        (a) Calibration








            (1) When the reference method is an  instrumental method,




            calibrate the  instrument prior  to conducting the tests in




            accordance with the procedure specified in the appropriate




            appendix to 40 C.F.R. Part 50 or succeeding codes.  When the




            reference method is a manual method, prepare a standard cali-




            bration curve with at least 7 points in the measuring range




            specified in Table B-l of f53.20(d).








            (2) When the candidate method is an  instrumental method.




            calibrate the  instrument before conducting the tests in




            accordance with the procedure specified by the manufacturer.




            In the absence of such a procedure,  perform a dynamic cali-




            bration with at least  7 identifiable points including 0 and




            90% of the upper range limit and in  accordance with  the
                                  130

-------
   appropriate  procedure  specified  in the  appendix  to 40 C.F.R.




   Part 50 or succeeding codes.   For a manual  candidate method




   prepare a standard calibration curve with at  least  7 points




   in  the measuring range  specified  in Table B-l of  H53.20(d),









   (3)  Submit  calibration  procedures, calculations and d^ta.









   (4)  Subsequent calibrations shall be permitted only during




   the time  interval between operational periods for both methods.




   An  operational period is 72 hours of continuous operating .luring




   which  no  manual adjustments to the electronic, gas or reagent




   flows  of  the instrument are permissible.  Automatic adjustments




   which  are a part ot the normal operation of the instruaent-s are




   permitted at any time.









(b)  Sampling Procedure:








    (1) After proper warm-up and  atablization. connect the sampling




    probes of the  reference and candidate methods  to the  distrri-




                L.ld of the ambient air sampling  system,
    (2) ttP.fer  to  Table C-l or C-3 for the test concentrations




    required.   Select one of the test sampling plans in




    Table  C-2  or  C-3.
                           131

-------
(3) Sample each test atmosphere simultaneously with the




reference and candidate methods in accordance with the




selected sampling plan.








No substitution in the plan shall be made until 14 samples




have been collected for the first test set or 18 samples for




the second set for oxidant and carbon monoxide methods.  For




sulfur dioxide methods, no substitution in the selected plan




shall be made until 7 samples have been collected for the




first set or 8 samples for the second set.  When necessary




adjust the pollutant concentrations in the sample air to the




desired levels with a pollutant generator or  zero air  [see




fl53.30(b)] such that the hourly average concentrations as




measured by  the reference method  fall within  the specified




ranges.








(4) At each  concentration range,  collect one-hour samples




for oxidant  and carbon monoxide methods.  For sulfur dioxide




methods collect 24-hour samples and in addition collect 1-hour




samples at the high concentration range.








(5) Conduct  the tests by operational periods  [see 1F53.3l(a) (4) ].




Introduce control samples or perform operational checks during




these periods to verify that the  methods are  performing within




specifications.  Control samples  or checks that show deviations




of more than ± 5% from the latest calibration indicate possible




calibration  shift or instrumental malfunction.  Discontinue
                       132

-------
testing.  Consult the operating Instructions or the manu-




facturer and correct the cause of the problem.  Recalibrate




the method and restart the testing from the beginning of




the operational period.









(6) During an operational period, collect at  least  three (3)




but no more than  six (6) samples per day for  oxidants and




carbon monoxide methods  and  one  (1)  24-hour sample  per day




for sulfur dioxide  methods.   Additionally, concurrent with




the 24-hour samples, collect two  (2) or  three (3)  1-hour




samples  at the high pollutant concentration  range  for sulfur




dioxide  methods.









(7) The  hourly  results shall be averaged based upon integration




of the  continuous output signals or by averaging at least  12




instantaneous  signal  lev?.Is at equally spaced interval-6; ovei




the 1-hour sampling periods.  For the 24-huur measurements, jl




least 22 hours ot continuous output signals shall be integrated




and reported as the 24-hour average tesults.  Description of




the methods  (i.e. planimetry, etc,..) used to obtain the  average




results shall be submitted.
 (8) Examine the results after completing the first test set.




 Analyze the results in accordance with  the statistical acceptance




 sampling plan in H53.31(e) or 153.3100.  Determine Whether




 the method has passed or  failed  to  demonstrate consistent
                        133

-------
relationship with the reference method or whether further




testing with the second set is required.








(9) From the combined results of the first and second test



sets, determine whether the candidate method has demonstrated



consistent relationship with the reference method according




to the acceptance sampling plans in H53.31(e) or H53.31(h).








(10) Submit all recorder tracings and other supporting




documents together with the test data.
                     134

-------
        (c) Test Specifications for Oxidant and Carbon Monoxide:   Table C-l
Pollutant Range
    (ppm)	

Oxidant
                  Tolerance,  i.e.
              Allowance Differences
            between Cand.  & Ref. Methods
            	(ppm)	
            Sample Size
      1st  Sample     2nd Sample
1.  Low 0.06-0.10
2.  Med. 0.15-0.25
3.  High 0.35-0.45
                       ±0.02
                       ±0.03
                       ±0.04
         5
         4
         5
Totals  14
 6
 6
 6
18
Carbon Monoxide

1.  Low 7-11
2.  Med. 20-30
3.  High 35-45
                       ±1.5
                       ±2.0
                       ±3.0
         5
         4
         5
Totals  14
 6
 6
_6^
13
         (d)  Sampling Plans  for Oxidant and Carbon Monoxide  Test
         Specifications:   Table C-2
         For  sample size
              of  14

         Plan #1
         Plan #2
         Plan #3

         For  Sample size
              of  18

         Plan #4
         Plan #5
         Plan #6
                            Time requirement falls between
                            	5 and 10 days	

                                HLM, LMH, MHL, MLH, LH
                                LMH, LHM, LHM, HML, HL
                                MHL, MHL, HML, MHL, HL

                            Time requirement falls between
                            	7 and 13 days	

                            LMH, MHL, HML, MLH, LMH, MHL
                            LHM, HML, MHL, MHL, MHL, MHL
                            LMH, LHM, HLM, LMH, MLH, LHM
      Note:
L = Low Concentration Range, M = Medium Concentration Range,
R = High Concentration Range as specified in Table C-l of
H53.31(c) .
                                    135

-------
(e)  Acceptance Sampling for Oxidant and Carbon Monoxide Methods:








    (1) The results of the first test consisting of 14 measure-




    ments shall be interpreted as follows:








        (i) If the number of measurements from the series of




        14 which fail the stated tolerances does not exceed




        zero, pass the method for consistent relationship.








        (ii) If the number of measurements from the series of




        14 which fail the stated tolerances is three or more,




        reject the method for consistent  relationship.








     (2) If  the number of measurements  from the series  of  14




    which fail the stated specification is one or  two, conduct




    a  second set of 18 measurements.   If  the number of measure-




    ments failing the stated tolerances of the combined tests of




    32 measurements is:








        (i) One or two, pass the method for consistent relation-




        ship.








        (ii) Three or more, reject  the method  for  consistent




        relationship.
                           136

-------
    (f)  Test Specifications for Sulfur Dioxide:   Table C-3
                                    Number of Measurements Required
                                     24 Hr.  Meas.
                                              1 Hr. Meas.




1.
2.
3.



S02 Range
(ppm)
Low 0.02-0.05
Med. 0.10-0.15
High 0.40-0.50

Tolerance, i.e.
Allowable Diff.
Between Cand.
& Ref. Methods
±0.02
±0.03
±0.04
TOTAL

1st
Sample
Set
2
2
3
7

2nd
Sample
Set
3
2
3
8

1st. 2nd
Sample Sample
Set Set
_»_»^ 	 u-r
	 	
7 8
7 3
    (g) Sampling Plans for S02 Test Specifications:  Table C-4
For Sample Size of 7

    Plan #1
    Plan #2
    Plan #3

For Sample Size of 8

    Plan #4
    Plan #5
    Plan #6
                                     H L M H M L H
                                     L M H M H L H
                                     M H L H M L H
                                     MHLHLMLH
                                     HLMLHMHL
                                     LMHMHLHL
Note:
L = Low Concentration Range, M = Medium Concentration Range,
H = High Concentration Range, as specified in Table C-3 of
1152.31(f).
                              137

-------
(h)  Acceptance Sampling Plan for Sulfur Dioxide Methods:








    (1) The results of the first test consisting of 7 24-hour




    and 1-hour measurements each shall be interpreted as follows:








        (i) If the number of measurements from each series of




        7 which fail the stated tolerances does not exceed



        zero, pass the methods for consistent relationship.








        (ii)  If the number of measurements from each series




        of  7 which fail the stated tolerances is three  or



        more, reject the method for  consistent relationship.








     (2) If  the number of measurements  from each series  of  7




    which fail the stated tolerances is either one or two,




     conduct a second test of 8 measurements.  If the number  of




    measurements  failing the stated  tolerances  from the combined




     total of  15 measurements for each  series  is:








         (a) One or two, pass the method  for consistent




         relationship.








         (b) Three or more,  reject  the  method  for  consistent




         relationship.
                            138

-------
                                 APPENDIX IV
     Air and Industrial Hygiene Laboratory's Review of the Comments and
         Recommendations Submitted by Respondents on EPA's Proposed
           Ambient Air Monitoring Equivalent and Reference Methods
            As Published in the Federal Register October 12, 1973
                                 Prepared by

J.J. Wesolowski, E.R. de Vera, Y. Tokiwa, W. Wehrmeister, K. Smith and M. Imada
                       AIHL, Laboratory Services Program
                          State Department of Health
                2151 Berkeley Way, Berkeley, California
                                 Prepared for

                        Environmental Protection Agency
                 Research Triangle Park,  North Carolina  27711

                                  March 197U


                       Federal Contract No.  68-02-07^
                                     139

-------
                         LIST OF RESPONDENTS






Agency/Location






Government Agencies:






1.  California State Air Resources Board




    Sacramento, California
Author
Kinosian
2.  Commissioner of Air Pollution Control




    St. Louis, Missouri
Copley
3.  Kansas State Department of Health




    Topeka, Kansas
Sides
4.  Los Angeles County Air Pollution Control




    District, Los Angeles, California
Lunche
5.  Maryland State Environmental Health




    Administration, Baltimore, Maryland
Ferrer!
6.  Montgomery County Department of Environmental   Mendelsohn




    Protection, Rockville, Maryland








7.  New York State Department of Environmental      Diamond




    Conservation, Albany, New York
8.  Oklahoma State Department of Health




    Oklahoma City, Oklahoma






                                140
Blanche

-------
 9.   Oregon State Department of Environmental




     Quality,  Portland, Oregon
Myles
10.  San Bernardino County Air Pollution             O'Malley




     Control District, San Bernardino, California
11.  Tennesse Valley Authority




     Chattanooga, Tennesse
                                                     Gartrell
12.  Wayne County Department of Health




     Detroit, Michigan
 Warner
 13.  U.S.  Department  of  Interior




     Washington,  DC
                                                     Seidl
   .   U.S.  EPA,  Office of Air Quality Planning        Schueneman




      and Standardization, Research Triangle Park,




      North Carolina
 L5.  U.S. Office of Environmental Affairs




      Washington, DC









 Lf>.  U.S. EPA Region III




      Philadelphia, Pennsylvania
                                                      Cromwell
  Belanger
                                   141

-------
Private Agencies:








 17.  American Electric Power and Service Corp.        Reeves



      New York, New York
 18.  American Smelting and Refining Co.



      Salt Lake City, Utah
Nelson
 19.  Beckman Ins truments, Inc.




      Fullerton, California
Chapman
 20.  Coomonwealth Edison




      Chicago, Illinois
Fancher
 21.  Ecology Board, Inc.




      Chatsworth, California
Chand
 22.   Edison Electric Institute




       New York, New York
Crawford
 23.  Environmental Research and Technology Inc.




      Lexington, Massachusetts
Muldoon
  24.  Exxon Corporation




      Washington, DC
Keller
                                   142

-------
25.  Meloy Labs




     Springfield, Illinois
Baer
26.  Southern Services, Inc.




     Birmingham, Alabama
Craig
Private Individuals:
27.  Philadelphia, Pennsylvania
Cropper
28.  San Jose, California
Mage
29.  Birmingham, Alabama
MeRanie
                                  143

-------
£                           REVIEW OF RESPONDENTS' COMMENTS AND RECOMMENDATIONS ON  SUBPART A
Ln

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-l    40 CFR 53/Overall
 12
Test program appears thorough.
Generally we agree.  However, we

do not believe it is complete,

e.g. test procedure for the effetf

of particles should be included.
A-2
                Procedures and specifications

                are so complicated and legalistic

                and leave so little allowance for

                judgment and the cost of certifi-

                cation will be so excessive, es-

                pecially for small manufactures,

                that once certified, there will

                be little incentive for change.
                                        We disagree with respect to the

                                        comment on the intricacies of

                                        the test procedures.   We believe

                                        the specifications are practical.

                                        We agree that cost of certificate

                                        will not be cheap. However, we

                                        disagree that certification will

                                        stifle development of superior
                                                                                          methods.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECOMMENDATIONS
A-3
                The regulations, as proposed, will

                force every agency to establish and

                maintain an expensive calibration

                and test facility to certify existing

                and proposed candidate methods.
                                    40 CFR 53 should state that the

                                    responsibility for certification

                                    of new analyzers rests with the

                                    manufacturer and not with the

                                    user.  For existing equipment,

                                    40 CFR 53 requests information

                                    regarding the respondent's status

                                    and possible replacement of

                                    potentially non-conforming in-

                                    struments and alternate proposal*
 A-4
                To reduce duplication of effort,

                EPA should publish names and models

                of instruments that EPA has already

                found to meet thess specifications.
                                    EPA as yet does not have such In-

                                    formation.

-------
                                                 RESPONSES TO SUBPART A
  ITEM   SECTION/LINE-TOPIC

  A-5
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
 10
San Bernardino County, California

APCD would not be able to conduct

and maintain such an extensive

program without additional funding.
See Item A-3 above.
  A-6
00
                EPA must furnish technical assistance   See Item A-3 above.

                and funding to carryout the tasks in-

                dicated in 40 CFR 53.
  A-7
 24
The elaborate test procedures pro-

posed may cause undue hardship on

smaller manufacturers.  EPA should

consider conducting the testing

either in-house or by contract

to another laboratory.
No comment

-------
                                                RESPONSKS TO  SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
     AIHL REMARKS OR
     RECOMMENDATIONS
A-8
 27
The paper work and test  facility

requirements assume a highly

centralized operation.   Most state

and local agencies will  not be able

to conduct subparts A and B.
 See Item A-3 above.
                                                   The administrator should have  the

                                                   option to declare instruments  in

                                                   place on October 12, 1973 as equiva-

                                                   lent upon presentation of data that

                                                   confirms partial compliance with

                                                   subparts B and C.
 A-9
 22
EPA should delay implementation of

40 CFR 53 until the collaborative

tests of the methods currently in

progress by ASTM (Project Threshold),

Midwest Research Institute and South-

west Research Institute are  com-
The organizations mentioned are

testing the accuracy and precis

of primarily manual methods,  no

equivalency determinations.
                                                   pieted.

-------
                                                    'ONSErj TO SUBFAR'J A
ITEM _  SEUTION/LINL-TOPIC

A-1C-
REST-
ONDEN1
 14
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
40 CFR 53 should not be implemented

until at least two instruments are

known to be certifiable.
We believe the test procedm ?.?

of the regulations as improve;!

by AIHL are practical and the

specifications are attainable.

We also know of 2 automated

methods that meet the requiremt

of the regulations.
A-ll
                EPA should leave the development of

                such testing procedures to pro-

                fessional standards organization

                such as NBS and ASTM.  This would

                avoid duplication of efforts and

                provide a division between metho-

                dology and effort and thus avoid

                any changes of conflict of interest.
                                                                                           No comment

-------
                                                ?JZ~,PON3E3  1C  SUBPART  A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
     AIHL REMARKS OR
     RECOMMENDATIONS
A-12
 22
40 CFR 53 is too harsh on existing

instrumentation since replacement

may be required as soon as 6  to 12

months whereas the analyzer life

may be 20 years.
 No  comment
A-13
 22
EPA should entrust development of

such test procedures to NBS, ASTM

or ANSI and avoid conflict of interest

between the methods and enforcement.
No comment
 A-14
 22
We are concerned 40 CFR 53 may not

accomplish their purpose in a fair

and equitable manner, avoid adding

unnecessary costs, avoid arresting

analyzer technology and may place

undue hardship on smaller manu-

facturers .
EPA has made every effort to be

fair, minimize cost, and encoura^

new innovations in analyzer tech-

nology consistent with the EPA's

obligations.

-------
                                                 RESPONSES TO SUBPART A
  ITEM	SECTION/LINE-TOPIC

  A-15
RESP-
ONDENT
 20
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
40 CFR 53 does not equate candidate     Our experiences indicate that at

methods with reference methods since    least one analyzer which uses th

we believe the reference methods can-   reference principle meets the

not meet these standards.  Hence, 40    performance specifications in

CFR 53 should be called a "performance  Subparts B and C.

standard".
  A-16
 20
Oi
The experience and expertise of

standards authorities such as NBS,

ASTM and ANSI should be included

in the development of such test

procedures.
The regulations was proposed in

the Federal Register in order

that everyone can participate in

the development of the final

version.  The proposed version

has included the input of air

pollution experts in the country
  A-17
 25
The pararosaniline (West-Gaeke) pro-    Total sulfur includes reduced

cedure responds to other oxides other   sulfur compounds (e.g. I^S, mer-

than S02-  The national standard        captans) as well as sulfur oxide

-------
   ITEM   SEC1ION/LINE-TGPIC
RESP -
ONDElif
                                                   RESPONSES TO SUBPART A
    RESPONDENT'S COMMENTS
     AIHL REMARKS OR
     RECOMMENDATIONS
   A-17  (Con't.)
 25
refers to sulfur oxides  measured

as S02 by the reference  method.

Therefore, the reference and

equivalency methods should be

changed to monitor total sulfur.
 The West-Gaeke procedure as pub-

 lished in the Federal Register

 is  relatively specific for S02-
   A-18   Pg.  28438, Col 2 and

          3, II 2 of 40 CFR 53
 19
Ln
This paragraph refers to 40 CFR

50 and mentions the fact that

EPA designated analysis methods

for 6 pollutants as reference

but does not indicate that 40

CFR 53 is to be limited to 3

pollutants.
The paragraph summarizes  the

authority and pertinent basis

for promulgation of 40 CFR 53.

On pg. 28439, Col. 2, 1T53.4(a),

it is stated that 40 CFR 53 applifc*

only to automated methods that

measure S02, CO and photochemical

oxidants.
                                                                                              To  avoid  confusion,  add  "regula-

                                                                                              tions  for the  other  three pollu-

                                                                                              tants , N02,  total hydrocarbon an

                                                                                              particulate  matter are to be es-

                                                                                              tablished later".

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC	

A-19   pg. 29438, Col. 3, H4/

       Reference to 40 CFR 51

       and 53.5(e)(2).
 RESP-
ONDENT
 12
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
We interpret this paragraph to

mean that when a method being used

in an accepted implementation plan

is declared by EPA within 2 yrs.  as

a reference method, then the user

need not demonstrate equivalency

for that method.  If the method,

however, is not accepted within

2 yrs., then the user must act to

validate his instrument.  But,

53.5(e)(2) requires the instru-

ment performance must be main-

tained for at least one year of

operation.  This means the tests

must be conducted over a one year

period.  This requirement appears

unrealistic in view of the fact

that many State's implementation

programs are already underway.
5.3.5(e)(2) requires that approy

methods (analyzers) shall functi

within specifications for at lea

one year.  If the Administrator

finds that a representative meth

fails to perform as required, th

approval shall be cancelled.  No

procedure for testing is given.

-------
                                                  RESPONSES TO BVBffLRT A
  ITEM   SECTION/LINE-TOPIC
  A-20
RESP-
ONDENT
 16
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
The test procedures apply to in-


dependent analyzer units and not


to units that may be incorporated


into a system.  An additional


paragraph (53.15) below is suggested.
We tend to agree.
Ui
Ln
                "53.15  Inclusion of Methods


                         in Automated Systems





                When a designated equivalent method or


                reference method is included in an auto-


                mated system package designed by a con-


                tractor other than the manufacturer of


                the methods, details of the system shall


                be reported by the contractor to the Ad-


                ministrator at least 30 days prior to the


                beginning of hardware construction.  The


                submission shall include specifications,


                drawings, data-handling procedures, and

-------
 ITEM   SECTION/LINE-TOPIC

 A-20 (Con't.)
QNDEi'lT
 16
Ul
                                                 RESPONSES TO SUBPART A
    RESPONDENT'S COMMENTS
operating, maintenance and calibration

procedures.  Where an alteration in a

method has been made, the submission

shall also include a statement that the

modification has been made, state the

reason for the modification, and show in

detail the departures from the method as

approved or designated.  In systems

placing designated/equivalent methods or

designated reference methods under the

control of computers, the submission shall

include a description of the software in-

cluding a statement of functions and flow-

charts.  Because these programs usually

are modified in the de-bugging phase, a

final set of software documentation shall
                                                            AIHL REMARKS OR
                                                            RECOMMENDATION
                                                    be  submitted upon  completion of system

                                                    installation, showing any deviations

-------
                                                  RESPONSES  TO SUBPART A




                                                                        ..                         A1HL REMARKS OR




                                                         P^rONDENT'S  COMMENTS                   RECOMMENDATION
 A-20  «'Con't  )                        16              from  Che  first  submission.  Where




                                                      internal  functions  of approved or




                                                      designated methods  are replaced by




                                                      computer  control,  the submission shall




                                                      compare the  computer control with the




                                                      original  functions.









                                                      If  the Administrator makes a preliminary




i-                                                     finding that  the methods,  as applied to




                                                      the system,  do  not  satisfy the requirements




                                                      of  this part, he shall notify the applicant




                                                      within 30 calendar  days of the report of




                                                      his intention not to accept the methods




                                                      as  applied as designated equivalent  or




                                                      designated reference methods.









                                                      If  the Adtoinistrator finds that the  methods,




                                                      as  applied to the system satisfy the require-




                                                      ments of this part,  he  shall forward to the




                                                      .-, poll'i^iit %  ^.ev-.r,c.sr  ;••..- tK? ;• ec£.--c  "

-------
                                                 RESPONSES TO SUBPART A
  ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECOMMENDATION
  A-21    pg.  28439,  Column  1,


         IF  3  Operation  and


         replacement of


         existing  analyzers.
Cn
00
                The regulations is not clear whether


                existing analyzers are to be accepted


                as equivalent.  Recommend that all


                analyzers in use prior to Oct. 12,


                1973 be accepted as equivalent.




                California has over 200 analyzers

                in operation.  The state has pur-


                chased nearly 40 analyzers that


                are reference or will meet equiva-


                lency.  Many local agencies have


                also replaced older units with


                newer generation instruments.  Up


                to 5 yrs. may be required to re-


                place all non-conforming instruments.
                                    We recommend that all analyzers


                                    in place on Oct. 12, 1973 that


                                    do not conform to 40 CFR 53


                                    should be accepted as equivalent


                                    for a period of 2 yrs. after


                                    promulgation of 40 CFR 53 and fo


                                    a period of up to 5 additional


                                    years upon demonstration of


                                    consistent relationship at the


                                    sampling site every 2 yrs.




                                    Also see respondent's comments


                                    in A-30 below.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECOMMENDATIONS
A-22   pg 28439 Col. 1, f3

       Operation and replace-

       ment of existing analyzers.
                Over 80% of Oregon's S02 analyzers

                are conductimetric.  These do not

                meet the requirements of 40 CFR 53

                because of their response to ions

                such as HC1 and NIK.  Since these

                interferents are not present in

                significant amounts in Oregon,

                these analyzers should be exempted.
                                    See Item A-21 above
A-23
                St. Louis has 5 analyzers for S02

                (conductimetric Davis), 5 for CO

                (NDIR), 4 for total oxidant (neutral

                KI with filter for S02 and data

                corrected for N0£), and one for 03

                (chemiluminescent).  Total oxidant

                units met the requirements in 40

                CFR 51.  S02 units will not meet

                40 CFR 53 due to response to inter-
                                    See Item A-21 above.

-------
                                                 RESPONSES TO SUBPART A
 ITEM   SiiCTION/LINE-lOP 1C
RESP-

ONDENT!
    RESPONDENT'S COMMENTS
                                            AIHL REMARKS  OR

                                            RECOMMENDATIONS
 A-23 (Con't.)
cr.
o
                ferents but since the instruments



                correct  for C0£ and since the other


                potential interferents  (NH3, HC1,


                are absent, we believe  these units


                provide valid data and  should be


                allowed to operate for  at least 5


                yrs. upon demonstration of con-


                sistent relationship at the site.
 A-24
 17
All analyzers in use on Oct. 12, 1973


should be considered equivalent for


the duration of their useful life  (up


to 20 yrs . ) .
                                        See  Item A-21  above.
 A-25
All analyzers in the state of Kansas


conform to the requirements in 40 CFR


51 except for N0£ units.  Assuming


availability of funds, it would take



nn1". il some time in 1.9 7 i.o convert to
                                                        See Item A-21  above,
                                                    40 CFR L>n, equipment .

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-26
  8
All existing analyzers for S02

and 03 in the state of Oklahoma

meet the requirements in 40 CFR 51

but not 40 CFR 53.  These were

selected with EPA Region VI's and

RTF's advice and approval.  Such

equipment should be allowed to

operate until normal attrition

replaces all with conforming

instruments.
See Item A-21 above.
A-27
                All methods used to obtain the data

                on which the air quality standards

                are based should be accepted as

                equivalent.
                                        See Item A-21 above.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECOMMENDATIONS
A-28
                The state of New York has 65

                analyzers; 26 for S02 (16 colori-

                metric, 6 FPD, 5 coulometric, 1

                GC), 16 for CO (NDIR), 22 for

                ozone (chendluminescent) and

                1 for CO-total HC in addition

                to analyzers for NO, N02> total

                oxidant by KI, COH and total

                HC by FID.
                                    See Item A-21 above.
                                                   We estimate 2 yrs. to convert

                                                   to all conforming equipment

                                                   depending on the availability

                                                   of funds, time ard personnel.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-28 (Con't.)
                We believe all existing equip-

                ment that meets most of the

                critical performance specifi-

                cations (substantial equiva-

                lency) should be accepted

                until the end of normal life.
                                        See Item A-21 above,
A-29
 14
EPA should allow use of existing

equipment for up to 3 yrs. for

those based on reference principles

and one year for other principles.
See Item A-21 above.
A-30
 11
Add "Upon promulgation of Part 53,

(1) a period of X months will be

allowed for testing of all existing

equipment for conformance under

Part 53; (2) existing monitoring
See Item A-21 above.

-------
                                                RESPONSES TO 3UBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    A1HL REMARKS OR
    RECOMMENDATIONS
A-30 (Con't.)
 11
data will be allowed during the

test period; (3) a period of T

months will be allowed to replace

nonconforming equipment; and (4)

existing monitoring data will

be allowed during this period".
A-31
 16
American Smelting and Refining

Co. has conductimetrie (Thomas

Autometer) for monitoring S0£

at 50 smelters.  Long term data

indicate potential interferents

such as HC1 and NHj are absent.

The proposed regulations should

be amended to make clear that

they are not intended to pre-

clude the use of Thomas autometers
See Item A-21 above.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-31   (Con't)
 18
or similar instruments in approved

ICS systems where these instruments

are shown to be satisfactory under

local conditions in which they will

operate.
A-32
 26
Consideration by EPA to permit

continued operation of instru-

ments purchased before Oct. 12,

1973 is proper and a procedure

for allowing the use of such

equipment should be delineated.
See Item A-21 above.
A-33   pg. 28439, Col 2, H2/

       Responsibility for

       determination
                It is not clear who-the manu-

                facturer or the control agency

                is to conduct these tests.  It

                is inconceivable that most states
                                        The responsibility for equiva-

                                        lency or reference determination

                                        rests with the applicant

                                        (pg.  28439, Col 2, HI).

-------
                                                  RESPONSES TO SUBPART A
   ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
                                                         RESPONDENT'S COMMENTS
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
   A-33  (Con.t)
                are capable and only the largest

                manufacturers would have such

                facilities.  The burden should

                be on the manufacturer.  For ex-

                isting equipment, some other

                arrangement should be made.
                                        The paragraph should  indicate

                                        clearly  that for instruments in

                                        place on or before Oct.  12, 197'

                                        the applicant in most cases is I

                                        user and that for instruments si

                                        after Oct. 12, 1973,  the applici

                                        is manufacturer.
  A-34
 12
cr>
It is not clear who-the manu-

facturer or the user-is to assume

the responsibility for demonstra-

ting the validity of their instru-

ments.  Our interpretation is that

the burden is on the user either by

instituting his own program or by

appeal to the manufacturer.  We

believe the burden should be on

the manufacturer only or in

collaboration with :"kf:.
See also Item A-33 above.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-35   pg. 28440, 53.4-53.14.

       53.4-53.14.
 12
The capability of equivalence

laboratories should be tested

with EPA test atmosphere.
EPA should establish a quality

assurance program to check the

capability of equivalence labor-

atories.
A-36   pg. 28440, 53.4(c-f)

       and 53.5  (c-1)/General

       Requirements
A-37   pg. 28440, 53.1(e) and       29

       53.5/Reference Methods
                A paragraph should be added re-

                quiring manufacturers to guarantee

                availability of replacement parts

                for some number of years and

                furnish proof of such action to

                EPA with perhaps a performance

                bond.



                When a reference method is replaced

                by another method, the previous

                reference method should be con-

                sidered equivalent.
                                        While this is desirable, EPA

                                        should investigate whether this

                                        would be legally enforceable.
                                        EPA should consider some type of

                                        variance program.

-------
                                                 RESPONSES TO SUBPART A
 ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
 A-38   pg. 28440, 53.1(e);

        53.4(d),(e),(f);

        5.310(2); 53.11(c).
oo
 26
The word "method" is used inter-

changeably with the word "technique".

"Method" should be used to mean

the measurement principle.  "Tech-

nique" should refer to the instru-

ment family.  It is not clear from

the text whether the measurement

principle, instrument family or

individual instruments are to be

certified.
Revise text to make clear that

instrument is to be certified in

cases of automated methods.
 A-39
 27
Use "instrument system or device"

in place of "method".
See Item A-38 above,
 A-40   pg. 28440, 53.1 (c-

        i) and other places,
 28
Stipulate that individual instruments   See Item A-38 above,

are to be certified.  The use of the

word "method" interchangeably with

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-40 (Con't.)
 28
"technique" infers that only

certification of the measurement

principle is required.  Other

paragraphs, 53.4(d),(e),(f), and

53.11(c) infers that individual

instruments are to be certified.
A-41   Pg. 28440, Col. 1,

       U53.3/Request for

       Equivalent or

       Reference Method

       Determination.
 25
A time limit should be specified

so that all manufacturers will

have the same opportunity to be

on the first selection list.
It may be possible for EPA to

announce in the Federal Register

a schedule of publication dates

(deadlines) for submissions to

insure inclusion of the applicants

method on the next announcement

of certified instruments.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOP1C
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-42   pg. 28440, H53.4(a)
 26
No information regarding the status

of N02 is given even though it is

known that EPA is re-evaluating

analytical techniques for N0£.

Some statement regarding this

pollutant should be made.
40 CFR 53 clearly states that

these procedures apply only to

methods for CO, S02 and photo-

chemical oxidant.  See also

Item A-18.
A-43   pg. 28440, U53.4
 25
Some of the information (components

schematic diagrams) is proprietary.

A description of the measurement

principle, manufacturer's name,

general schematics of the sample

flow and electronics should be

sufficient.
Schematic diagrams are usually r

proprietary but the description*

of some components (i.e., in-

gredients of converters and pre-

filters) may be.  Proprietary ii

formation appears to be adequati

covered by K53.14.
                                                   Requirements for model numbers are

                                                   unrealistic since minor changes  to

                                                   interface with buyer's system may

                                                   require different m- di'.'i designation.
                                                        We  do not  agree.

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
A-44   pg. 28440, 53.4(d)

       and (e).
                53.4(d) concerns the descrip-

                tive information and sworn

                statements to be submitted and

                53.4(e) concerns the operation

                manual.  It is suggested that f53.4

                be recorded so that H(d) concerns

                descriptive information to be sub-

                mitted, and (e) concerns sworn

                statements to be submitted.  Then,

                the operation manual, which contains

                descriptive material, would be trans-

                fered to sub 1f(d), and sub t[s(d)(4)

                and (d)(5) to 1F(e).
                                        Separate descriptive information

                                        from sworn statements unless it

                                        is the intent of this section to

                                        require that the descriptive in-

                                        formation in 53.4(d)(l),(2) and

                                        (3) are to be submitted in the

                                        form of sworn statements.
A-45   pg. 28440, 53.4(d)(2)
 26
Since sworn statements can be mis-

interpreted, the determination of

equivalence should be based on the

test data 53.4(c) submitted.
It is the intent to use sworn

statements to supplement the

test data submitted in accordance

with 53.4(c).

-------
  ITEM	S F.CT ION /L IN* -T'"1? ~ C

  A—i5 I'v'o,:" :.)
RESP-
ONDED
                                                  RESPONSES TO SUBPART A
    RE^PONDENT'5 COMMENT5
                        pfeguards ph^u^d be de-

                veloped to insure the quality

                control programs developed in

                53.4(d)(2) are adhered to.
    AIHL-REMARKS OR
    RECOMMENDATIONS	

Appears to be covered by  ')'•.'••
                                      25
NJ
                The quality assurance  and quality

                control programs of some manufac-

                turers may include proprietary  items,

                Release of such information  in  de-

                tail is not proper.
                                        See  Item A-43.
  A-47   pg. 28440, 53.4(d)(4)
 25
The meaning of the statement

"representative of measurement

systems tested" is not. clear.

Sugges*- "representative in con-

struct ton, design, and perfor-

mance jpeciflesticns of the

measurement systems tested."
Insert "in  construction,  desigr.

and performance  specifications'

between  "representative"  and "•_

in 53.4(d)(4)  and  53.5(4).

-------
                                                RESPONSES TO SUSPART A
ITEM   SECTION/LINE-TOPIC

A-48 ,  pg. 28440, 53.4(c)/

       Operation manual
RESP-
ONDENT
 26
    RESPONDENT'S COMMENTS
The operation manual provided to

the purchaser should include all

the data submitted to EPA for

certification purposes.
    AIHL  REMARKS  OR
    RECOMMENDATIONS
The buyer need  only know  that  the

instrument meets  the  requirements

of the regulations.
A-49
 29

 26
The operation manual stipulated

in 53.4(c) and 53.5(d) should in-

clude all the data that was sub-

mitted to EPA to achieve certi-

fication.
See Item A-48 above.
A-50   pg. 28440, 53.4(f)(i)

       and 53.5(e)(i)
 25
The requirement that the "methods

shall function properly	for at

least one year in the field" is un-

necessarily harsh on the manufacturer.

It should be phrased to state "for

at least one year after delivery

to the customer", which is consis-

tent with standard practice.
We agree.

-------
                                                RESPONSES  TO SUBPART A
ITEM   SECTION/LINE-TOPIC

A-51   pg. 28440, 53.4(f)

       and 53.5(e).
RESP-
ONDENT
RESPONDENT'S COMMENTS
                                                                                              AIHL REMARKS OR
                                                                                              RECOMMENDATIONS
                How is durability to be determined?

                Who is to conduct the tests after

                the one year period - the manufac-

                turer or the purchaser?
                                    Under paragraphs 53.4(f)(2) and

                                    53.5(e)C2), the Administrator m*

                                    determine compliance based on

                                    manufacturer's statements and

                                    test data or the Administrator n

                                    conduct his own tests.
A-52   pg. 28440, 53.4(f)(2)

       and 53.5(e)(2)
                                    26
                How does EPA propose to inform buyers

                when certification is cancelled.

                Announcement in the Federal Register

                may not suffice.
                                    According to 53.10(c), cancel-

                                    lation shall be published in the

                                    Federal Register.  Since the in-

                                    strument buyer must rely on the

                                    Federal Register for informatior

                                    concerning changes in Federal

                                    regulations, announcement in the

                                    Federal Register is the proper

                                    means for cancelling certificati

-------
                                                jnjece§e*oMSJDsat  TO
ITEM   SECTION/LINE-1'OPIC
RESP-
ONDENT
A-53
     RESPONDENT'S  COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
                What recourse does the buyer have

                in the event certification is

                cancelled?  Civil suits are ex-

                pensive and time-consuming.

                Possibly a performance bond should

                be required of the manufacturers.
                                         Under paragraphs 53.4(f), 53.5(e)

                                         and 53.10, the purchaser has no

                                         recourse in the- event certification

                                         is cancelled.   Perhaps EPA should

                                         require some type of performance

                                         bond of the manufacturer as a

                                         condition of certification or per-

                                         haps the purchaser can require a

                                         bond as .condition of sale.
A-54   pg. 28440, 53.5(d)/

       Operation manual
 25
A comprehensive description of

the calibration and testing pro-

cedures followed is not needed by

the ultimate buyer because test

procedures are published by the

government and available to the

public and also since the govern-

ment intends to publish those methods

Which have passed the tests and are

therefore acceptable
Substitute the word "operational"

for "testing" in 53.5(c)(3), line

This change will make 53.5(c)(3)

correspond to 53.4(d)(3).

-------
                                                RESPONSES TO SUBPARX A
ITEM   SECriON/LIHE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COmBBTS
    AIHL REMARKS OR
    RECOMMENDATIONS
A-55   pg. 28441, 53.6 through
       53.10.
 26
Notic* of all pending actions re-
garding equivalence and cancellation
should be published in the Federal
Register to warn prospective pur-
chasers .
We do not agree.  The «anufac-
turer is "innocent until proven
guilty".
A-56   pg. 28441, 53.6
 25
Change "60 calendar days" to "15
calendar days" since 60 days is
an unreasonable delay for the manu-
facturer.  Government has a respon-
sibility to process the results of
such restrictive actions promptly.
Do not change; the tine to respon
to the applicant will depend on
the number of submissions being
reviewed and on the completeness
of the applicant's submission.
The 15 days proposed is insuffi-
cient without a large staff of
reviewers.
A-57   pg. 28441, 53.6(d)
 25
Add "the additional tests will
be thoroughly defined for the
manufacturer and a reasonable
The paragraph states that the tea
are to be conducted by the Adminii
tractor, not the manufacturer.  Tl

-------
                                                RESPONSES TO SUBPART A
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
A-57 (Con't.)
A-58   pg. 28441, 53.71

       Right of entry.
 25
 25
specified time limit be given

to conduct the tests".  The 60

day limitation for approval should

also be changed to 15 days.



Right of entry provision should

be deleted as this may involve

entering areas where proprietary

activities and development are

being conducted.
additional  tests  to be  conducted

would be  those published in 40 CFR

53.  See  also Item A-56.
The purpose of right of entry is

to assure proper conduct of the

equivalency test procedures.  This

requirement is analogous to govern-

ment inspectors entering food and

drug processing and similar indus-

tries to insure compliance with

government regulations.   See

also Item A-43.
A-59
 25
Also, it should be stipulated

that the government will reim-

burse any costs incurred by the
The text states that the tests to

be conducted are for the purpose

of assisting the applicant in the

-------
                                                  RESPONSES TO  SUBPART A
  ITEM   SECTION/LINE-TOPIC
                 RESP-
                 ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
 A-59  (Con't.)
                  25
company during conduct of any

tests conducted at the applicant's

facility by EPA.
equivalency or reference deter-

mination as well as to determine

compliance with the regulations.

We do not believe that the govern-

ment should reimburse the company

for such expense.
  A-60   pg.  28491,  53.8
00
                  29
The word "method" should not

be used to mean "instrument".
See Item A--38 above.
  A-61   pg.  28441,  53.10
                                 When an approved method has been

                                 revised by the manufacturer to

                                 avoid cancellation, such revisions

                                 should be published.
                                        We agree.
  A-62
53.10(c)
Reference in 53.10(c), line 8 to

53.10 should be 53.10(2)(b).
Add "(b)" to 53.10  in  line  8  of

53.10(c)

-------
                                                  RESPONSES TO SUHPART A
  ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
  A-63   53.10,  53.13
Nl
O
 25
53.10 and 53.13 permit no right

of appeal until after the lengthy

and therefore costly cancellation

procedures in 53.10 have been ex-

hausted.  Some provision for right

of prior appeal should be provided.
According  to  53.10(a)(2),  the

applicant  (manufacturer) has the

option of  demonstrating to  the

Administrator that  the method in

question does  meet  the require-

ments of this  part  or he may make

any adjustments needed to bring

the method into compliance.  We

believe this provision constitutes

prior appeal.
  A-64   53.11
 26
Adequate provisions should be made      The restrictions in 53.11(d)(l)

for input from the purchasers during    (iii) do not preclude purchasers

these hearings.                         as expert witnesses.
  A-65   pg.  28442, 53.12

         Modification
 25

 17

 29
The definition of modification is

too broad and requires the manu-

facturer to report any trivial

and insignificant changes.
Add "that will affect the perfor-

mance of the instrument" between

"method" and "including" in line 3

of 53.12 and line 4 of 53.13.

-------
                                                 RESPONSES TO SUBPART A
 ITEM   SECTION/LIME-TOPIC
RESP-

ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR

    RECOMMENDATIONS
 A-65  (Con't.)
                1153.12 should be changed to read



                "when significant changes...which



                affect the performance of the



                method".
  ,\-66   pg. 28442, 53.13
oo
o
 27



 26



  4
The statement "any modification of



the construction or design of the



method shall be cause for cancella-



tion" would freeze the design of



the method.  Change the word "shall"



to "may".  This change will make



53.13 conform in intent and meaning



to 53.12.
Delete remainder of paragraph



starting with "Any" in line 9



of 53.13 and substitute remainder



of paragraph 53.12 starting with



"If the administrator	etc."

-------
                            REVIEW OF RESPONDENTS'  COMMENTS AND RECOMMENDATIONS ON SUBPART B
oo

-------
                                                 RESPONSES TO SUBPART B
 ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
 B-l    pg. 28442

            53.20(a)
 26
Terms used in 53.20(a) should

be defined in 53.10.
Respondent probably meant 53.1.

Relevant terms are sufficiently

defined in the test procedures.
 B-2    pg. 28442
oo
to
                Include a test for linearity

                since most telemetering systems

                accommodate only linear outputs.
                                        EPA should consider a test for

                                        linearity.
 B-3    pg. 28442

            53.20(c)
 25

 27
This section indicates that seven

tests shall be performed for each

performance specification, in-

cluding all of those in Table B-4.

This certainly is unnecessary for

noise, lower detectable limit,

and interference equivalent.
Proposed EPA draft reviewed

Feb. 4, 1974 by AIHL stipulates

a minimum of 7 tests for each

parameter except for range.

-------
                                                   RESPONSES  TO SUBPART  B
  ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
  B-4    pg. 28442

             53.20(c)
 29
All of a sudden,  "instrument"  is

used, rather  than  "method".
 In AIHL  Sept.-Oct.  1973 Progress

 Report,  Appendix  I, instrument

 is used  as the preferred word

 for automated method.  EPA

 should clearly define "Method"

 and "Instrument" or "Analyzer"

 in Subpart A.
»  B-5     pg.  28442
o

              53.20(d)

          Table  B-l
  1

  2

  4

 19

 25

 28
Footnotes to Table B-l do not

apply correctly as they appear

out of sequence.  Footnote b,

not c, applies to range, while

footnote c, not d, applies to

zero drift.
This was corrected in the AIHL

suggested draft of Subpart B,

Appendix I, Sept.-Oct. 1973

Progress Report.
   B-6     pg,  28442

              53.20(d)

          Table  B-l
 28

  5

  6
Ranges specified are too low, since     We agree.  The possible relaxatioi

they are below the federal emergency    of standards due to energy crisis

warning level.  They should be at       situations may require more

-------
                                                RESPONSES TO SUBPART B
 ITEM   SECTION/LINE-TOPIC
                                   RESP-
                                   ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECQMMENPATIONS_
 B-6 (Con't.)
                                                   least  at  the emergency  level  and

                                                   preferably 10% higher.   Also, 0.5

                                                   ppm 03 is regularly exceeded  in

                                                   California, and 50 ppm  CO can be

                                                   obtained  in almost any  city.
                                    flexibility in the measurement

                                    ranges of S02 and 03 methods.
00
B-7    pg. 28442

           53.20(d)

       Table B-l
 B-8    pg.  28442

            53.20(d)

        Table B-l



 B-9    pg.  28442

            53.20(d)

        Table B-l
                                                    Increase 0,  range  from 0-0.5

                                                    ppm to  0-1.0 ppm.
                                                   Increase 03 range to 1.5
                     ppm
                                                   Increase 03 range to 0.7 ppm

                                                   (our emergency level is 0-0.6 ppm.)
                                    There seems to be a. plurality

                                    sentiment in favor of increased

                                    range - See Item B-6.



                                    See Item B-6.
                                    See Item B-6

-------
                                                    OiM&ISS "TO SUBPAJCT
ITEM   SECTION/LDIE-TOPIC

B-10   pg. 28442

           53.20(d)

       Table B-l
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
                Increase CO range, from 0 to 50 ppm

                to 0-100 ppm.
                                         See  Item B-6.
B-ll   pg. 28442

           53.20(d)

       Table B-l
                Increase CO range to between 100

                and 150 ppm, since major downtown

                intersections often exceed 50 ppm.
                                        See Item B-6,
B-12   pg. 28442

           53.20(d)

       Table B-l
                Increase CO range to 150 ppm.
                                        See Item B-6,
B-13   pg. 28442

           53.20(d)

       Table B-l
 25

 26
Total sulfur should be used as

the standard, rather than S02,

as the former includes toxic

gases such as 803, H2S + mercaptans,
We disagree.  Later it may be

important to develop separate

standards for each of these gases,
B-14   pg. 28442

           53.20(d)

       Table B-l
 29
Some instruments now produced have

0 to 0.5 ppm S02 range and cannot

be modified to another range.
Required S02 range in Table B-l

is 0-0.5 ppm.

-------
                                                   RESPONSES  TO  SUSPART  B
   ITEM   SECTION/LINE-TOPIC
                                   RESP-
                                   ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
   B-15   pg.  284A2

              53.20(d)

          Table B-l
                                    17

                                    23
Some instruments have log-linear

SC>2 range, and thus can measure 0

to 1.00 ppm S0£ while still having

high precision at low S0£ values.

The S02 range should be changed to

reflect this to allow 0-1.00 ppm.
Multi-range capabilities are

covered in the EPA Feb. 4, 1974

draft.
oo
B-16   pg. 28442

           53.20(d)

       Table B-l
                                       26

                                       29
A more reasonable range for

S02 field sampling is 0 to

1.00 ppm.
See Item B-15 above.
   B-17   pg.  28442

              53.20(d)

          Table  B-l
                                                   The range for SOj should be

                                                   0-5.00 ppm.
                                        Seems excessive - See Items B-6,

                                        B-14 and B-15 above.
   B-18    pg.  28442

              53.20(d)

          Table  B-l
                                                   The 0-0.5 ppm S02 range for S02

                                                   is less than the recommended

                                                   warning level and should be revised.
                                        See Items B-6 and B-15.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOP1C

B-19   pg. 28442

           53.20(d)

       Table B-l
RESP-
ONDENT
 15
    RESPONDENT COMMENTS
                                                                                               AIHL REMARKS OR
                                                                                               RECOMMENDATIONS
The range values given In Table

B-l Imply that 0 ppm Is a measure-

able level which is impossible.
                                                                                           53.22(a)  defines "Range"

                                                                                           as  "the nominal minimum and

                                                                                           maximum concentrations...."
                                                   Some finite level below the lower

                                                   detectable limit should be specified,

                                                   or the definition of range changed.
B-20   pg. 28442

           53.20(d)

       Table B-l
                                    25
                The noise values should have ±

                values.
                                        Corrected in the AIHL suggested

                                        draft of Subpart B in the Sept.-

                                        Oct. 1973 Progress Report,

                                        Appendix I.
B-21   pg. 28442

           53.20(d)

       Table B-l
 26
Noise spec, is acceptable.
                                                                                          We agree.
B-22   pg. 28442

           53.20(d)

       Table B-l
 29
Noise spec, is reasonably tight.
                                                                                          We agree.

-------
    ITEM   SECTION/LINE-TOPIC
                                                    RESPONSES TO  SUBPART B
RESP-
ONDENT
    RESPONDENT COMMENTS
                                            A1HL REMARKS OR
                                            RECOMMENDATIONS
    B-23   pg.  28442

               53.20(d)

           Table B-l.
                Reduce S02 noise spec, from i

                .005 to ± 0.002 ppm.
                                        At the current state of the art,

                                        ± 0.005 ppm is optimal.
    B-24   pg.  28442

               53.20
-------
                                                   RESPONSES  TO  SUBPART  B
   ITEM   SECTION/LINE-TOPIC
                                   RESP-
                                   ONDENT
    RESPONDENT COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
   B-26   pg.  28442

              53.20(d)

         Table  B-l
                                    25

                                    26

                                    12

                                    29

                                     2
Value signs (±) are incorrect

for interference equivalent.

The total interferent should be

given as the absolute value.
Corrected in AIHL Sept.-Oct. 1973

Progress Report, Appendix I, page

3.
00
\0
B-27   pg. 28442

           53.20(d)

       Table B-l
                                       29
The interferent values appear

high, especially the total

interferent.
We disagree.  We think they are

too tight for the present state

of the art.
   B-28   pg.  28442

              53.20(d)

          Table B-l
                                     7

                                     2
Reduce zero drift to the range

of 0.005 ppm to 0.01 ppm for

S02 and 03.  A ± 0.02 ppm tolerance

is 66% of the 0.03 ppm average

annual standard for S02 and 25%

of the 0.08 primary standard for

0^.  In any case, drift should not

exceed 15% of the standard.
Present specifications reflect thi

state of the art.  Normally, it 1$

the practice to subtract the driff

values from ambient air readings

and thus such drifts should not

affect the annual readings.

-------
                                                  RESPONSES TO SUBPART  B
   ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
   B-29    pg.  28442

              53.20(d)

          Table  B-l
                24 hrs. is given for zero drift

                in the table, and 12 hrs. in the

                test.
                                        See Appendix I, page 24, AIHL

                                        Sept.-Oct. 1973 Progress Report

                                        wherein the specifications were

                                        made to correspond to the tests.
   B-30   pg.  28442

              53.20(d)

          Table  B-l
                The 24 hr. drift value is more

                important than the 12 hr. value.
VO
o
                                        The 12 hr. test shows the con-

                                        tinuous trace over 12 hr. whereat

                                        the 24 hr. data only indicates

                                        the Instantaneous values at 24

                                        hr. intervals.
   B-31    pg.  28442

              53.20(d)

          Table  B-l
                The specification should also

                list a value for 3-day zero

                drift.
                                        We believe the 24-hour zero drift

                                        specification is adequate.  In

                                        most monitoring operations, the

                                        instruments are rezeroed every

                                        24 hours.
   B-32    pg.  28442

              53.20(d)
 29
Zero drift specification is

reasonable and somewhat ti^
We agree.

-------
                                                    RESPONSES TO SUBPART B
    ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
    B-33   pg. 28442

               53.20(d)

           Table B-l
 26
Zero drift value is OK.
We agree.
    B-34   pg. 28442

               53.20(d)

           Table B-l
 27
As in Item 28, span drift should

be also reduced from 0.005 to

0.001 for S02 and 03.
See Item B-28.
5   B-35   pg. 28442

               53.20(d)

           Table B-l
 26

 25

  6

 29
No time limit is given for the

span drift.
Corrected in February 4, 1974

EPA draft.
    B-36   pg. 28442

               53.20(d)

           Table B-l
 29
Span drift specification is

reasonable.
We agree.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
                                                                                                   AIHL REMARKS OR
                                                                                                   RECOMMENDATIONS
B-37   pg. 28442

           53.20(d)

       Table B-l
                The specifications for lag, rise,

                and fall times are broad enough

                to cover wet chemical instruments

                and should not be applied to

                direct physical analyzers.
                                        If we start dividing instruments

                                        into artificial classifications,

                                        we defeat the purpose of the

                                        regulations which is to establish

                                        reasonable minimum specificatioc

                                        for all analyzers.
     B-38    pg.  28442

£               53.20(d)

            Table  B-l
                                    29

                                    26
                The lag, rise, and fall times

                should be reduced to 3 minutes,

                and preferably 1 minute.
                                        The current specifications are

                                        adequate for reporting one-hour

                                        averages for ambient air quality

                                        data.
B-39   pg. 28442

           53.20(d)
 12
Response times could be much

shorter.
                                                                                                See  Item B-38.
B-40   pg. 28442

           53.20(d)

       Table B-l
 28
Reduce lag time and rise time

to 10 minutes each and the sum

of both not to exceed 15
                                                                                                See  Item B-38

-------
                                                  RESPONSES TO  SUBPART B
  ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
  B-40 (Con't.)
 28
minutes.  The present 15

minute tolerance for both

can delay emergency alerts

for up to 35 minutes.
   B-41   pg.  28442

              53.20(d)

         Table  B-l
 25
vo
LO
The specifications for rise time,

fall time, and lag time are too

long and would result in excessive

waits during alerts, the lag time

should be 3 minutes, the fall

time 5 minutes.
See Item B-38
   B-42   pg.  28442

              53.20(d)

          Table B-l
                Specifications for precision appear

                too lax for the 20% range and too

                stringent for the 80% range for 03

                and S02
                                        Revised in EPA draft, February

                                        4, 1974.
   B-43   pg.  28442

              53.20(d)

          Table B-l
 29

 26
Specification for precision is

reasonable and fairly tight.
We agree.

-------
                                                  RESPONSES TO SUBPART B
   ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
   B-44   pg.  28442

              53.20(d)

         Table B-l
 25
The precision specifications

require a ± sign.
Corrected in Appendix I, page 3,

AIHL Sept.-Oct. 1973 Progress

Report.
   B-45   pg. 28442

             53.20(d)

         Table B-l
 28
vo
Unless the temperature of the

ppm reading is within ± 2°C of

25°C, footnote "a" should read

"to convert from ppm at pressure

P torr and temperature T°C to

Ug/M3 at 25°C, and 760 torr,

multiply by (104.2)(M)(T+273.15)/p

where M is the molecular weight of

the gas".
In this conversion, ppm is

considered to be at standard

conditions of 25°C and 760 Torr.
   B-46   pg.  28442

              53.20(d)

          Table  B-l
 29

 26
Specifications for NO, N02, and

THC are missing.  If there is a

reason it should be so stated in

the document.
See Item A-18.

-------
                                                    RESPONSES TO SUBPART B
    ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
     AIHL REMARKS OR
     RECOMMENDATIONS
    B-47    pg.  28442


               53.20(d)


           Table B-l
 25
Some of  the  tests can be  run  con-


currently e.g.  the noise  test with


zero drift.
 See EPA draft February 4, 1974


 and Appendix I,  AIHL Sept.-Oct.


 1973 Progress Report.
    B-48   pg.  28442


               53.20(d)


           Table B-l
 20
VO
Ui
    B-49   pg.  28442


               53.20(d)


           Table B-l
 12
These standards are invalid as


they cannot be met even by the


reference methods.  (Respondent


gives no proof or reasons)




Specifications are for the most


part within the capabilities of


existing instrumentation.
One  reference method  that  we


have  tested has passed all the


specifications.
We agree.
    B-50A  pg.  28442


               53.20(d)


           Table B-l
 19
Performance specifications in EPA


publications of August 14, 1971


and November 25, 1971 do not agree


with Table B-l.  The former should


be upgraded.
40 CFR 51, pg. 28438, H51.14 and


1151.17 clearly state that the


table of performance specificatici


has been revised.

-------
ITEM
B-50B  53.21

       Test procedures

       for automated

       methods.
                                                RESPONSES TO SUSPART B
                            RESP-
                            ONDENT
                                                       RESPONDENT'S COMMENTS
                                            Add a test procedure for response

                                            to vibration.
                                                           AIHL REMARKS OR
                                                           RECOMMENDATIONS
                                                       This is not required under current

                                                       regulations for air quality.  Coul

                                                       come under requirements in 53.4(f)

                                                       and 53.5(e).
B-51
vo
pg. 28442

    53.21(a)

Gas Generation
19

 4
This paragraph requires "all

diluent air shall be free from

contamination".  Since this is

impossible, the maximum possible

allowable concentration must be

stated for each gas, within

reasonable limitation.
                                                                                           Corrected in EPA draft, February

                                                                                           4, 1974.
B-52   pg. 28442

           53.21(a)

       Gas Generation
                             19
               "Output of the #as generation

               system shall be stable"   How

               stable?  How demonstrated?
                                        See AIHL remark in 53.22(c) of

                                        EPA February 4, 1974 draft.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
                                                                                               AIHL REMARKS OR
                                                                                               RECOMMENDATIONS
B-53   pg. 28442

           53.21(a)

       Gas Generation
                                    26

                                    29

                                     2

                                     4
                The cylinder gases are supposedly

                certified by the manufacturer, but

                the details of the test analysis

                (e.g., precision and accuracy)

                are missing.  Also, cylinders are

                sometimes 40% in error from supposed

                values.  All gases should be verified.
                                        We agree.  The duplicate samples

                                        provision should also apply  to

                                        all gases.  Also see AIHL remarks

                                        in Table B-2 of EPA Feb. 4,  1974

                                        draft.
1-54   pg. 28442

           53.21(a)

       Gas Generation
 19
"Concentration of test atmospheres

shall be verified by duplicate

samples...".  Does this hold for

gases which have been certified?
                                                                                           See 53.22(a), Gas Generation, in

                                                                                           EPA February 4,  1974 draft.
B-55   pg. 28442

           53.21(a)

       Gas Generation
 27
The test atmosphere is assumed

to be perfectly invariant.  Since

this cannot be so, the tolerance

specifications must be corrected.

to allow for this.
                                                                                           See Item B-53 above..

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
                   RESPONDENT'S COMMENTS
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
B-56   pg. 28442

           53.21(a)

       Gas Generation
                             19
               "Temp... around gas system shall be

               controlled to within ± 2°C".   Does

               this mean ambient air in the  test

               lab?
                                        This requirement deleted in EPA

                                        February 4, 1974 draft.
B-57
 CO
pg. 28442

    53.21(a)

Gas Generation
25
B-58   pg. 28442

           53.21(a)

       Gas Generation
B-59   pg.  28442

           53.21(a)

       Gas  Generation

       Line 12
                             21
                             11
The temp specification of ± 2°C is      See Item B-56 above.

too restrictive and should be

changed to 5°C.



Has commercially available permeation   The invitation should be accepted

devices with a low temperature depen-   by EPA.  The currently used per-

dency of 2 to 3% per °C.  Invites EPA   meation devices are critically

to investigate these products.          temperature dependent.
               Change "the temperature surrounding

               the gas generation system shall be

               controlled to within t 2°C" to "the

               temperature of the gas surrounding
                                        See Appendix I, page 4, AIHL Sept,-

                                        Oct. 1973 Progress Report as being

                                        a clearer explanation of the

                                        requirements.  See also Item B-56.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
                                                                                               AIHL REMARKS OR
                                                                                               RECOMMENDATIONS
B-59 (Con't.)
 11
the gas generation system shall be


within the operating range of the


gas generation system and main-


tained within a variation of ±


2°C during the test period".
B-60   pg. 28442


           53.21(a)

i—>
£      Gas Generation
                                    11
                Change to "if a permeation device


                is employed during the generation


                of the test atmosphere, the per-


                meation device as well as the air


                passing over the device shall be


                maintained at a temperature which


                is within the normal operating


                temperature range of the device


                with a variation of less than ±


                0.1°C of the selected temperature


                during the test period".
                                        So stated in Appendix 1, page 4,


                                        AIHL Sept.-Oct. 1973 Progress


                                        Report.  Also see EPA Feb. 4, 1974


                                        draft.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
B-61   pg. 28442


           53.21(a)


       Gas Generation


       Line 13
                            RESP-

                            ONDENT
                             11
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR

    RECOMMENDATIONS
Add "during", between "before" and


"and" to provide for verification


of the test atmosphere during the


test.
Add "and/or during" between


"before" and "and".
B-62
t-o
O
o
pg. 28442


    53.21(a)


Gas Generation
                                    25
Use of stainless steel or TFE Teflon


should be allowed for some of the


gases.
We suggest, that any inert material


which has been demonstrated not


to alter the test gas concentra-


tions be allowed.
B-63   pg, 28442


           53,21(a)


       Gas Generation.


       Line 18
                                            The reference to the design and


                                            construction of the gai generation


                                            system is vague and should be


                                            deleted.
                                        See Appendix I, page 4 and page 29


                                        AIHL Sept.-Oct. Progress Report


                                        wherein a more elaborate des-


                                        cription is giveii
B-64   pg,  28443


           53.21(b)
                                            Table B-2 requires  certified gas  in


                                            zero air cor CO,  C02,  and
                                        In most cafes the interference


                                        equivalent is determined in the

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-64 (Con't.)

       Table B-2
ho
O
                while the test procedure requires

                mixing the certified gas with

                zero air to achieve the test

                concentration.  This is incon-

                sistent.
                                        presence of interferent and

                                        pollutant gases.  The secondary

                                        mixing with zero air is a pre-

                                        liminary step in the test pro-

                                        cedure for interference equiva-

                                        lent.  The zero air is substituted

                                        with equal flow of pollutant test

                                        atmosphere later in the test.
B-65   pg. 28443

           53.21(b)

       Table B-2
 28
The requirements for the certifi-

cation of the cylinder gases are

inadequate.  A standard could be

obtained by the purchaser and

comparison of 2 cylinders of the

same gas from separate manufacturers,

and/or an additional certification by

an outside laboratory.
See EPA draft Feb. 4, 1974 and

AIHL remarks thereon.

-------
                                                RESPONSES  TO  SUBPART  B
ITEM   SECTION/LINE-TOP1C

B-66   pg.  28443

           53.21(b)

       Table B-2
B-67   pg. 28443

           53.2Kb)

 M     Table B-2
                                   RESP-
                                   ONDENT
                                    12
                                    19
    RESPONDENT *S COMMENTS
                                                                                               AIHL REMARKS OR
                                                                                               RECOMMENDATIONS
Cylinders should contain less than      See Table B-2 in EPA Feb. 4, 1974

700 ppm CO, since concentrations of     draft

around 2000 ppm are known to convert

to iron pentacarbonyl.



(For N02) "gas phase titration des-     Reference added in EPA draft of

cription available from NERC-RTP"       Feb. 4, 1974.

should be included in document.
B-68   pg. 28443

           53.21(b)

       Table B-2
                                    J9
 (For N02) "Permeation tube modified     See Item B-67

by NERC-RTP" should be included in

document.
B--69   pg. 28443

           53.21(b)

       Table B-2
                                    19
Footnote:  Ref. 2 - 1970 Annual book    Included in EPA Feb. 4, 1974 draft.

of ASTM standards should use current

1973 book.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-70
NJ
O
U)
pg. 28443

    53.21(b)

Table B-2
Water vapor test:  Provide de-


tails of bubbler; some designs


do not give saturated air.


Also, the humidity measurement


depends in part upon the tech-


nique used and this should be


detailed.
Design of humidifier is the respon-

sibility of the applicant.
B-71   pg. 28443


           53.21(c)(2)


       Calibration of


       the instrument
                                            Identify all 7 points with a

                                            tolerance for each point to

                                            allow more precise comparison

                                            of calibration curve.
                                        Specified as "approximately equal


                                        spaced intervals" in EPA Feb. 4,


                                        1974 draft.   Comparison of curves


                                        may be done with slopes.
B-72   pg. 28443


           53.21(c)(2)


       Calibration of


       the instrument
                             19
Must the calibration consist of


7 points when the output of the


analyzer is known to be linear?


This calls for more cylinders of
                                                                                    7 points are needed for greater


                                                                                    statistical reliability.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/UNE-TOPIC
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-72 (Con't.)
                             19
gases or a more sophisticated


dilution system.  Would 5 be


enough ?
B--73
 N)
 O
pg. 28443


    53.21(c)(2)

Calibration of


the instrument
"A dynamic calibration curve...


including 0 and 90% of scale...".


What is the meaning of dynamic?
Add definition o  dynamic calibra-

tion in 53.20.
B-74   pg. 28443


           53.21(c)(3)


       Safety
                                            Spell out safety factors or make

                                            reference to applicable OSHA


                                            requirements.
                                        These requirements should be

                                        considered and included in the

                                        sections for operational manual

                                        in 53.4 and 53.5.
B-75   pg. 28443


           53.21(c)(3)

       Safety
                             19
"shall not...discharge ar "


hazardous...effluent gas'.  Is


the discharge of ethylene from
Both gases are hazardous.  Dis-


charges of analyzers should be


properly vented as described by

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-75 (Con't.)
                             19
the chend.luminescent analyzers

hazardous?  Is the discharge

from a CO analyzer hazardous?
 the manufacturers.
B-76
ro
o
Ln
pg. 28443

    53.22(a)

Range

Line 9
Change example for range from

0 to 1.00 ppm to 0-0.5 ppm to

correspond with one in Table

B-l.
So changed in EPA Feb. 4, 1974

draft.
B-77   pg. 28443

           53.22(a)

       Range

       Line 9
                                            Procedures and specifications

                                            apply to operating on a single

                                            range only.  Procedures should

                                            apply to other ranges also.
                                        EPA Feb. 4, 1974 draft also

                                        covers multirange methods.
B-78   pg. 28443

           53.22(a)

       Noise
                             25

                             19

                              5

                              1
Use of DVM precludes submission

of recorder tracings.
So corrected in Appendix I pages

10 and 11 of AIHL Sept.-Oct.

1973 Progress Report and in EPA

Feb. 4, 1974 draft.

-------
                                                RESPONSES TO SUHPART B
ITEM
B-79
SECTION/LINE-TOPIC
pg. 28443
53.22(a)
Noise
RESP-
ONDENT
25
RESPONDENT'S COMMENTS
The noise test cannot be carried
out with a span gas as there are
definite deviations in concen-
AIHL REMARKS OR
RECOMMENDATIONS
We disagree. The noise at span
is often substantially greater
than at zero. See also Item B-52
                                                   tration, e.g., ± 0.004 ppm with


                                                   S02 gas from a permeation system.


                                                   We recommend deletion of the noise


                                                   test at span.
B-80a  pg. 28443


           53.22(b)


       Noise
NJ
o
    We find noise is better defined     a)  Specifications are based on the


    in terms of absolute range of           ranges given in Table B-l and


    chart rather than ppm.                  apply to all ranges.
B-80b  (2)(vii)


       Noise
                                                  Certain noise frequency patterns    b)  We agree.


                                                  may be missed by the random DVM


                                                  readings.
B-81   pg. 28443


           53.22(b)(2)(vii)


       Noise
                                    4


                                   25
    Errors in the noise equation:


1)  The radical sign is missing from


    the standard deviation formula.
So corrected in Appendix I, pages


10 and 11, AIHL Sept.-Oct.  1973


Progress Report.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOP1C
RESP-
ONDENT
    RESPONDENT'S COMMENTS
                                                                                               AIHL REMARKS OR
                                                                                               RECOMMENDATIONS
B-81 (Con't.)
            2)  Small k becomes large K, even

                though both are identical
                                    11
            3)  (dj)2 should be (di)2
B-82   pg. 28443

           53.22(b)(2)(vii)

       Noise
                                    25
                "25 random readings taken during

                60 minutes" does not agree with

                the definition of noise as short

                term, and may include zero drift

                in the result.  Change to 25

                readings taken within 2 minutes

                after a stable baseline is

                achieved.
                                        The zero drift requirements make

                                        any significant drift during the

                                        noise test unlikely.  Suggestion

                                        incorporated in EPA Feb. 4, 1974

                                        draft.
B-83   pg. 28443

           53.22(b)(2)(vii)

       Noise
 19
Noise could easily result from ex-

terior factors such as from outside

electric mains, air sparks, etc.

This should be ruled out.
                                                                                           Noise should apply  to  the analyser

                                                                                           performance  and not reflect  ex-

                                                                                           ternal factors.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-84   pg. 28443-44

           53.22(c)

       Lower Detectable

       Limit
                             11
Add "as determined with zero

air" after "noise level".  The

noise can be influenced by the

magnitude of the detector signal,
Also corrected in EPA Feb. 4, 1974

draft.
B-85
 O
 oo
pg. 28444

    53.22(d)

Interference

Equivalent
Proposal indicates that some form

of pre-scrubber will be required

on our FPD analyzer for S02 to

remove I^S.  Such contaminants do

exist in our atmosphere and such

scrubbers may degrade the pollutant

of interest.  The proposal should

allow more flexibility and judgment.
Paragraph 53.22(d)(2)(ii) does

not require the use of any pre-

scrubers.
B-86   pg. 28444

           53.22(d)
                             26

                             29
No method of flow measurerrent and

control is stated.  The only re-
The procedure and specifications

in the EPA Feb. 4, 1974 draft

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-86 (Con't.)

       Interference

       Equivalent
                             19
striction is that the interferent

concentration be within ± 20% of

the stated value.  This permits

cancelling the effect of any

interferent by flow adjustment.

Also, no flow reading is required.
requires the test gas concen-

trations be prepared and con-

trolled to ± 10%.  We recommend

preparation and control of test

gas concentrations to ± 5% of

stated values.
B-87
N3
o
VO
pg. 28444

    53.22(d)

Interference

Equivalent
Test provides no provision for

synergistic effects.  For example,

in electrochemical S02 analyzers,

NO- together with water gives a

larger response than the response

separately.
This is beyond the capabilities of

the test.
 B-88   pg. 28444

           53.22(d)

       Interference

       Equivalent
                             25
How can "x" be diluted and still        Corrected in Appendix I, p.  14,

have same concentration as specified?   AIHL Sept.-Oct.  1973 Progress

                                        Report.

-------
                                                          iu
ITEM   SECTION/LINE-TOPIC
RESP-

ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR

    RECOMMENDATIONS
B-89   pg. 28444


           53.22(d)


       Interference


       Equivalent.
 25


  1
The gas concentrations must


have a clear cut value reference


to Table B-3.  The words


"specified concentrations1' are


used for x, y, and z, and then


for x 4- y, x + z.  They are con-


tradictory.
Corrected in Appendix I, p. 1,


AIHL Sept.-Oct.  1973 Progress


Report.
B-90  -2(b)(iii)-(iv)
N3
M
o
                Composition of test gas is in-


                determinate .   Procedure is in-


                consistent with the procedural


                description of gases x + y.
                                        See Item B-89 above.
B-91  -2fb)(iii)-(iv)
 25
It is unreasonable to use the sum


of the absolute values of the in-


dividual interferences as some


gases compensate for eac>; other.


A formula should be prepared to


take this into account.
This is beyond the scope of the


test required.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
     AIHL  REMARKS  OR
     RECOMMENDATIONS
B-92  -2(iv)(c)-line 2
 11
Change "z" to "y
 Corrected  in  Appendix 1,  AIHL

 Sept.-Oct.  1973  Progress Report.
B-93  -2(xiii)
 11
Change to:  "make a reading;

subtract from this the reading

obtained from zero gas.  Record

the difference as the "inter-

ference equivalent absolute

concentration limit" to account

for zero effect.
Same as  Item B-92 above.
B-94   pg. 28444

           53.22(d)

       Interference

       Equivalent
 19
Procedure appears over complicated

and tedious.
The information needed requires

a definitive procedure that

cannot be easily made simple.
B-95   pg. 28444

           53.22(d)(2)(e)
 25

 2b
Specifications for NO 2 and THC

are included in Table B-3 even
Editorial error-corrected in

EPA draft of Feb. 4, 1974.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
B-95 (Con't.)
 29

 19
though not discussed elsewhere,

A reason should be so stated.
B-96   pg. 28444

           53.22(d)(2)(e)
                Are interferent levels listed

                realistic?  For example 750 ppm

                C02 is about twice the level

                in ambient air.  How about possible

                quenching effect of G02 in chemi-

                luminescent analyzers for 03?
                                        The concentrations listed are

                                        compromises reflecting experiences

                                        under a wide variety of moni-

                                        toring situations and practical

                                        test limitations.
B-97   pg. 28444

           53.22(d)(2)(3)
                Cone,  of l-^S is unrealistic by

                100X of amount in ambient air.
                                        See Item 96 above.
B-98   pg.  28444

           53.22(d)(2)(e)
 25
The Table does not list a"..l

possible methods, e.g.,  oloumetric

and polarographic for S02.
See Item 99 below.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC

B-99   pg. 28444

           53.22(d)(2,(e)
RESP-
ONDENT
 15
    RESPONDENT'S COMMENTS
Table B-3 should have only the

status of a guide, unless EPA

believes there are no other

possible methods or interferences.
    AIHL  REMARKS  OR
    RECOMMENDATIONS
Table B-3 in  EPA  draft  of  Feb. 4,

1974 allows for future  consider-

ation of other measurement

principles.
B-100  pg. 28445

           53.22(f)(4)

       Test Procedure

       for zero and

       span drift
 11
Instructions are confusing.

Allow too much interpretation;

may be incomplete, and require

back and forth referencing from

the text to Table B-4.  Expand

Table B-4 to indicate timing of

all measurements; including zero,

span 1, span 2, lag time, rise

time and fall time,
A refined plan is given in

Appendix 1, p. 19 AIHL Sept.-Oct.

1973 Progress Report.  Test in-

structions are clarified in same

report.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
                   RESPONDENT'S COMMENTS
                                            AIJHL REMARKS OR
                                            RECOMMENDATIONS
ii-101  pg.  28445

           53.22(f)(4)

       Test Procedure

       for zero and

       span drift
                         26
               Since calibration is permitted

               every 72 hours in Subpart C con-

               sistency tests, then it should

               also be permitted in zero and

               span drift tests in Subpart B.
                                        In the EPA draft of Feb. 4, 1974

                                        the provision .In Subpart C per-

                                        mitting calibration every 72 hrs,

                                        has been deleted.
B-102
53.22(f)(4)(vi)
19
"Test concentration shall be con-

sistent from day to day".  How

accomplished?  Drift in day to

day test gas concentration will

obscure analyzer drift.
See Item B-52 above.
3-103      53.22(f)(5)(i)

       Lag Time
                                        Definition is not adequate for

                                        instruments using GC for separ-

                                        ation of individual components.
                                                       We agree that the performance

                                                       specifications in this pro-

                                                       cedure may not always be

                                                       applicable to GC type analyzers

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
                   RESPONDENT'S COMMENTS
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
B-104  pg. 28445

           53.22(f)(5)

       Lag Time
                         19
               What is the instrument inlet?

               What is first observable change

               in output?
                                    1)  The sampling port so designated

                                        by the manufacturer.

                                    2)  The change equal to 2X noise as

                                        recommended in AIHL Sept.-Oct.

                                        1973 Progress Report.
B-105
 (Jl
53.22(f)(6)
19
It is impossible to know when the

response equals 5% of upper range

limit, since this value is not

known beforehand.
It's possible.
                                                   Propose another method for re-

                                                   sponse times similar to the

                                                   international standard method

                                                   for electronic instruments.
                                                                                No comment
B-106      53.22(f)(6)

           53.22(f)(8)(g)

           53.22(f)(8)(m)
                         27
               An example of each calculation needed   Equations and data sheets  added

               in this section would be helpful.        In EPA draft of Feb.  4,  1974.

-------
                                                RESPONSES TO SUBPART B
ITEM   SECTION/LINE-TOPIC
.,-107  pg. 28445

           53.22(f)(7)(ii)(b)
RESP-
ONDENT
 II
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
Change to "sample zero air and use

a stopwatch to time the interval

starting with the first observable

instrument output response due to

the concentration step change and

ending when the response reaches

95% of the way to the zero reading

obtained in 53.22(f)(6), rise time,

and corrected for zero base line

drift used on said zero reading".
We believe the procedure is

clear enough.
B-108  pg. 28445

           33 22(f)(7)(ii)(b)

       Fall Time
 11
Change "Repeat steps (a),(b),

and (c), to repeat rise time

and fall time tests".
See Appendix I, pp 22 to 24, AIHL

Sept.-Oct.  1973 Progress Report

wherein work definitive test pro-

cedure are recommended.

-------
                                                RESPONSES TO SUBPART R
ITEM   SECTION/LINE-TOPIC

B--109  pg. 28445

           53.22(f)(8)

       Precision
RESP-
ONDENT
 25
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
These sections do not stipulate

at what time you record the

value after "switching quickly

to pollutant concentration used".

We recommend 1 hour.
Corrected  in Appendix  I, p.  24,

AIHL Sept.-Oct., 1973  Progress

Report.
                                                   No time period is given after

                                                   sampling the higher concentration.

                                                   It should read "after stable out-

                                                   put is obtained.
                                                        See Item B-109 above.
B-110  f(8)(ii)(g) & (8)(ii)(i)
 19

 25
"Calculate and report the mean and

standard deviation for these 6

values" is incorrect.  Should be

calculated for each group of 3

data.
See Item B-109 above, p. 25 and

26.

-------
 ITEM   SECTION/LINE-TOPIC

 B-lll  f(8)(ii)(k)
RESP-
ONDENT
  1
                                                 RESPONSES TO SUBPART B
RESPONDENT'S COMMENTS
Procedure should read "switch to

(i)" not to (J).
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
                                                         lee Item B-109 above, p. 25.
 B-.U2  pg. 28446

            53.22(g)

        Table B-4
Ni
M
oo
 B-113  pg. 28446

            53.22(g)

        Table B-4
                Point out in text that the two

                values for line voltage and for

                temp, constitute specifications

                for voltage and temperature.



                Temperature and voltage settings

                for night tests appear to be

                missing.
                                        This seems to be a valid

                                        criticism.
                                        Clarified in Appendix I, p. 27,

                                        AIHL Sept.-Oct.  1973 Progress

                                        Report.
 B-114  pg. 2 446

        Table B-4
 29
Good test plan
                                    We agree.  However, it needs tot

                                    improved.  See suggested version

                                    in the AIHL Sept.-Oct. 1973

                                    Progress Report.

-------
                                                 RESPONSES TO SUBPART B
 ITEM   SECTION/LINE-TOPIC
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
 B-115  Table B-l, B-2

        and B-3
 29
Specifications given for Q.J, S02,

and CO only.  N02 and THC should

be included.
The proposed regulations currently

do not include N02 and THC; there-

fore no specifications for these

methods should be included for

now.
NJ
M
VD

-------
REVIEW OF RESPONDENTS' COMMENTS AND RECOMMENDATIONS ON SUBPART C

-------
ITEM   SECTION/LINE-TOPIC
C-l    pg.  28446, 53.307


       overall proposal
                        RESP-

                        ONDENT
                         19
                                                RESPONSES TO SU3PART C
                   RESPONDENT'S COMMENTS
               The procedure appears over-


               complicated and tedious am


               should be simplified.
                                           "AIHL REMARKS OR
                                            RECOMMENDATIONS
                                        The procedures have been simpli-


                                        fiec! - see recommended draft of


                                        Subpart C in Appendix II, AIHL


                                        Sept.-Oct., 1973 Progress Report.
           53.30-53.33
                         28
  K)
  M
  KJ
               Differences in analyzer hourly


               average readings in excess of


               tolerance due to differences


               in the lag and response times


               of the reference and candidate


               methods should not be cause for


               rejection.
                                        Test procedure should allow compen-


                                        sation for differences in response


                                        time characteristics between


                                        reference ant1 candidate methods.


                                        Has been corrected in EPA draft of


                                        Feb. 4, 1974.
C-3
53.30(a)
27
There is no justification ior re-


quiring that the test sanding site


be located in areas of maximum ex-


pectant pollutant concentrations
Other substances are generally


associated with high pollutant


concentrations in smoggy air.


This requirement is an attempt

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOP1C
RESP-
ONDENT
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
C-3 (Con't.)
 27
since the concentrations are to

be altered by spiking anyway.
to obtain test atmosphere com-

positions that are as close to

realistic as practical.
 to
                Also, does this mean the tests

                should be conducted during periods

                of maximum pollutant concentration

                and that this period might be

                different for each pollutant?
                                        No provision is given whether the

                                        testing shall be conducted during

                                        seasons when maximum pollutant

                                        levels are expected.  More clarifi-

                                        cation is needed here.
C-4    pg. 28446, 53.30(a)
                Requirements for test site are in-

                sufficient.  For instance, CO

                concentrations adjacent to a

                traffic intersection may be

                much higher than elsewhere.

                Then distance from curb and

                sampling site must be known.
                                        This specification applies to

                                        sampling community air.  See also

                                        Item C-5.

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
                                    25
                   RESPONDENT'S COMMENTS
                                        Object to "the Administrator

                                        retains the discretion to select

                                        sampling sites for the tescs he

                                        determines necessary to conduct"

                                        as being costly to some manufacturers

                                        and favoring some companies due to

                                        geographic location.
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
                                                       Re\ision of test site requirements

                                                       should be considered by EPA.
C-6
53.30(b)
12
Performance efficiency of the ozone

generator is not mentioned.
Not relevant since reference and

candidate analyzers are already

calibrated.
                                                   Reliance on manufacturer for standard    See  remarks  in B-65.

                                                   gases  is unreliable and methods  for

                                                   verification should be i  iuded.
C-7
53.30(b)
               The stipulation that "the in-

               take. . .manifold shall be designed
                                        Corrected in EPA draft of Feb.  4,

                                        1974.

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
                   RESPONDENT'S COMMENTS
                                            AIHL REMARKS  OR
                                            RECOMMENDATIONS
C-7 (Con't.)
                         25
               so as not to remove particles or


               trace gases" is impossible.  Change


               to "reasonable precautions shall


               be taken to minimize the removal


               of particles and trace gases".
C-8
53.30(c)
25
 to
 ho
 Ul
Requirements are expensive and time-


consuming.  EPA should conduct these


tests.
No comment
C-9
53.30(e)
25
It is illogical to require any


method bear a  "constant relation-


ship" to a reference method since


the reference may have more drift,


interferences and less accuracy.
Tolerances in the test procedures


should be sufficient to accommmodati


such differences.
                                                   The candidate method may have a much


                                                   smaller absolute error and still be

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOPIC

O9 (Con't.)
RESP-
ONDENT
 25
NJ
N3
ON
    RESPONDENT'S COMMENTS
    AIHL REMARKS OR
    RECOMMENDATIONS
rejected unless the method drifts In

the same direction and suffers the

same interferences as the reference.

This section requires that the manu-

facturer provide methods "as bad as"

the reference instead of the best

available.



The procedures should require meeting

Subpart B and submission of a corre-

lation plot relating the reference

and candidate method readings waich

are based on the same conditions

specified in Subpart C.
C-10   pg.  28446; 53.31

       (a)(3); 53.32(a)
 29
Calibrations are permitted oncf

every 72 hrs.  If required for
See Item B-101.

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
    RESPONDENT'S COMMENTS
    AIHL  REMARKS  OR
    RECOMMENDATIONS
       (b); 53.
                                        Subpart C it should also be in

                                        tests for zero and span drifts in

                                        Subpart B.
C-ll
 to
 NJ
53.31(b)
C-12   pg. 28446; 53.31(b)(l)
Paragraph states "Description of

the methods used to obtain the

hourly averages shall be sub-

mitted".  Will manual integra-

tion of the chart traces des-

cribed in 53.32(b)(l) be accepted?



Change 1.1 ppm to 0.10 ppm to

correspond to Table C-l.
                                                                                           Yes
                                                                                Corrected in Appendix III,  AIHL

                                                                                Sept.-Oct.  1973 Progress  Report.
C-13
53.31(b)(l)
Value for oxidant is based on

1-hour averages only while the

procedure for CO and S02 permits
Same remark as in Item C-12.

-------
  ITEM   SECTION/LINE-TOPIC
                        RESP-
                        ONDENT
 C-13  (Con't.)
                                                 RESPONSES TO  SUBPART C
    RESPONDENT'S COMMENTS
                                        averaging 12  equally spaced


                                        readings  over a 1  hr.  period.


                                        This  option should also .pply


                                        to the oxidant procedure,
    AIHL REMARKS OR
    RECOMMENDATIONS
  C-14        53.31(c)/Tables  C-l,      29

             C-3
                                        Tolerances  are tight and should

                                        assure accurate instruments.
                                        No comments.
  C-15       53.31(c)/Tables  C-2
M
NJ
00
  C-16       53.32(b)
                         29
Good sampling plans.




Range of 35 to 45 ppm CO is


insufficient.  Change 55


to 75 ppm.
We believe them to be acceptable-




Administrator has the discretion


to permit changes in range.
  C-l 7
53.32(b)
The 5-channel Jieckman €  >Q


cannot meet 12 samples/hour.
                                                                                             See Item B-103.

-------
                                                RESPONSES TO SUBPART  C
ITEM   SECTION/LINE-TOPIC
                            RESP-
                            ONDENT
                   RESPONDENT'S COMMENTS
                                            AIHL REMARKS OR
                                            RECOMMENDATIONS
C-18   pg. 28447

           53.33

       Test Procedure, S02
                             25
               Concentrations at low, medium,

               and high, cannot be maintained

               due to variability of ambient

               air.
                                        Concentration levels are maintaine

                                        artificially by adding pollutant

                                        and/or zero air.
 C-19
NJ
NJ
VO
pg. 28447

    53.33

Consistency Test

Procedure, S02
25
Use of West-Gaeke method for S02

is not recommended, since the

precision is only 4.5% at 95%

confidence level.  Use of NBS

permeation tubes allows greater

accuracy.
Corrected in EPA draft of Feb. 4,

1974.
 C-20   pg.  28447

            53.33(a)

        S02  Manual Methods
                                            Control samples,  cannot be in-

                                            troduced or operational checks

                                            be made during an operational

                                            period, and at the same time

                                            obtain the hourly sampling

                                            averages required by the procedure,
                                                       Clarified in EPA draft of Feb.  4,

                                                       1974.

-------
                                                RESPONSES TO SUBPART C
ITEM   SECTION/LINE-TOPIC


C-21   pg. 28448


           53.33(b)
                            RESP-
                            ONDENT
    RESPONDENT'S COMMENT S
                                            Section has two "b" paragraphs:


                                            (b)  Test specifications for S02;


                                            (b)  Table C-3.


                                            (c)  Correct the section designation


                                            of Table C-3 to (c).
    AIHL REMARKS OR

    RECOMMENDATIONS
                                        Corrected in Appendix II, AIHL


                                        Sept.-Oct.  1973 Progress Report
C-22
NS
U)
o
pg. 28448


    53.33(b)


Table C-3
Reduce the S02 tolerance to the


range of 0.005 to .01 ppm, be-


cause a ± .02 ppm tolerance


constitutes 66% of the .03 ppm


annual average for S02-
No comment.
C-23       33.33(d)(3)


       Line 10,
                                            This sentence refers to a "very


                                            high" concentration instead of


                                            the "hig i  concentratrati a" re-


                                            ferred to  in Table C-3 or C-4.
                                        Corrected in EPA draft of Feb. 4,


                                        1974, and Appendix II, AIHL Sept


                                        Oct. 1973 Progress Report.
C-24   pg. 28448


           5.33(d)(3)
                             2,4
This paragraph refers to 24 and


3 hr. sampler anci r\f s tables
Corrected in EPA draft of Feb. 4


1974 -Bid in Appendix II, AIHL

-------
                                                RESPONSES TO SUBPART C
SECTION/LINE-TOPIC
RESP-
ONDENT
RESPONDENT'S COMMENTS
AIHL REMARKS OR
RECOMMENDATIONS
C-24 (Contf.)
                C-3 and C-4.  Line 14 makes

                reference to 3 hr. values, but

                Table C-3 contains only 24 and

                1 hr. values.  Inconsistent;

                needs correction.
                                    Sept.-Oct. 1973 Progress Report.
C-25       53.33(d)(3)

       Line 5
                Sentence refers to 4 sampling

                ranges while Tables C-3 and

                C-4 ci te only 3 ranges.
                                    Same as remark in C-24 above.

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1. REPORT NO.
  EPA-650/4-75-014
             3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
  Evaluation of the proposed Ambient Air Monitoring
  Equivalent and Reference Methods
                                                           5. REPORT DATE
                                                             April 197U
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
   J.J.  Wesolowski, E.R.  DeVera, Y. Tokiwa,
     W.  Wehrmeister, K.  Smith and M. Imada
             8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Air and Industrial Hygiene Laboratory
  Laboratory Services  Program
  State of California, Department of Health
  2151 Berkeley Way ,   Berkeley,  CA  91*70*1
             10. PROGRAM ELEMENT NO.
               1H1327, ROAP 26 AAF
             11. CONTRACT/GRANT NO.
               68-02-07^1*
 12. SPONSORING AGENCY NAME AND ADDRESS

  Environmental Protection Agency
  Research Triangle  Park, N.C.  27711
             13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE
 15 SUPPLEMENTARY NOTES
 16 ABSTRACT

   The report assesses  the adequacy of the Environmental Protection Agency's  proposed
   Ambient Air Monitoring Equivalent and Reference Methods, published  in the  Federal
   Register of October  12, 1973.   To assess  the applicability, validity and
   workability  of the  regulations and determine the costs to demonstrate  equivalency,
   the provisions, specifications and test procedures of the regulations were
   evaluated in two phases using three automated methods or analyzers  and  two manual
   methods.  The comments regarding the proposed regulations submitted to  EPA by
   respondents were also reviewed.
   The cost and man-hours required to perform each of the test procedures  was also
   estimated.
17.
                                KEY WORDS AND DOCUMENT ANALYSES
                  DESCRIPTORS
  Air Pollution
  Tests
  Standards
  Monitors
                                              b-IDENTIFIERS/OPEN ENDED TERMS
                           c.  COS AT I Field/Group
 3 DISTRIBUTION STATEMENT

      Unlimited
19. SECURITY CLASS (This Report)
  Unclassified	
21. NO. OF PAGES
    251
                                              2O. SECURITY CLASS (This page)
                                                 Unclassified
                                                                         22. PRICE
EPA Form 222O-1 (9-73)
                                             232

-------