A GUIDE FOR THE EVALUATION OF ATMOSPHERIC ANALYZERS

                           by

                   P.K. Mueller, Ph.D.
                        ¥. Tokiwa
                       E.R. deVera
                    W.J. Wehrmeister
                        1. Belsky
                        S. Twiss
                        M. Imada

                 EPA Contract 68-02-0214


                    Project Officers

                  J.B. Clements, Ph.D.
                      T.W. Stanley
                      Prepared For
             Environmental Protection Agency
            Research Triangle Park, NC  27711
Air and Industrial Hygiene Laboratory, Laboratory Program
      California State Department of Public Health
               Berkeley, California  94704
                        June 1973

-------
This report has been reviewed by the Environmental




Protection Agency and approved for publication.




Approval does not signify that the contents




necessarily reflect the views and policies of




the Agency nor does mention of tradenames or




commercial products constitute endorsement or




recommendation for use.

-------
                      TABLE OF CONTENTS






CHAPTER                                                     PAGE






            Acknowledgement                                  i




            Abstract                                         ii




            Preface                                          ili-lv




1.0     Introduction                                          1




2.0     Manufacturer's Specifications                         2




  2.1       Introduction                                      2




  2.2       Class I                                           3




  2.3       Class II                                          4




3.0     Operating Instructions                                7




  3.1       Introduction                                      7




  3.2       Identification                                    7




  3.3      ' Operation and Hiintainanca Section                3




  3.4      .Field Repair or Service Section                  15




4.0     Procedures for Determining Performance               19




  4.1       General                                          19




  4.2       Physical Characteristics                         27




  4.3       Instrument Operating Instructions               - 28




  4.4       Calibration                                      30




  4.5       Linearity                                        32




  4.6       Measuring Range                                  36




  4.7   -    Calibration Accuracy                             37




  4.8       Drift at Zero                                    38;




  4.9     .  Drift at Span                                    39




  4.10      Noise                                            41

-------
                      TABLE OF CONTENTS






CHAPTER                                                     PAGE






  4.11      Precision and Response Times                     44




  4.12      Pulse Time            .                           49




  4.13      Lower Detectable Limit                           50




  4.14      Analyzer Deadband                                52




  4.15      Warm-up Time                                     53




  4.16      Interferences                                    55




  4.17      Operating Temperature Range                      58




  4.18      Operating Voltage S.ange                          60




  4.19      Unattended Operation                             62




5.0     Gas Generation                                       80




  5.1       Principle and Scope                              80




  5.2       Apparatus                                        82




  5.3      ' Operation                                        93




  5.4       References                                       94




6.0     Calibration Procedure for Automated Atmospheric




        Oxidant and Ozone Analyzers                         107




  6.1       Principle and Scope                             107




  6.2       Range                                           108




  6.3       Interferences                                   109




  6.4       Precision, Accuracy and Stability               109




  6.5       Apparatus                                       110




  6.6       Reagents and Gases                              114




  6.7       Spectrophotometer Calibration                   120




  6.8       Dynamic Calibration                             121

-------
                       TABLE OF CONTENTS






 CHAPTER                   •                                  PAGE




   8.5       Apparatus                                       175




   8.6       Reagents and Gases                              178




   8.7       Spectrophotometer Calibration                   179




   8.8       Dynamic Calibration                             180




   8.9       .Static Calibration                              187




   8.10      Reconciling Static and Dynamic Calibrations     188




   8.11      References                                      190




 9.0     Calibration Procedure for Automated Atmospheric




         Carbon Monoxide Analyzers                           194




   9.1       Principle and Scope                             194




   9.2       Range                                           195




   9.3       Interferences                                   195




   9.4     •  Pre'cision, Accuracy and Stability               196




   9.5       Apparatus                                       '197




   9.6       Reagents and Gases            •                  199




   9.7       Preparation of Calibrating Gases                201




   9.8       Dynamic Calibration                             203




   9.9       References                                      207




.10.0     Air Analyzer Terminology                            210

-------
                      TABLE OF CONTENTS


CHAPTER                                                     FACE


  6.9       Static Calibration                              126

  6.10      Reconciling Static and Dynamic Calibrations     132

  6.11      Basis for Calculations                          133

  6.12      References                                      135

7.0     Calibration Procedure for Automated Atmospheric

        Nitrogen Dioxide and Nitric Oxide Analyzers         140

  7.1       Principle and Scope                             140

  7.2       Range                                           141

  7.3       Interferences                                   142

  7.4       Precision, Accuracy and Stability               142

  7.5       Apparatus                                       143
       <
  7-6       Reagents and Gases              .                149

  7.7       Spectrophotometer Calibration                   152

  7.8       Dynamic Calibration                             153

  7.9       Static Calibration                              160

  7.10      Reconciling Static and Dynamic Calibrations     165

  7.11      References                                      166

8.0     Calibration Procedure for Automated Atmospheric

        Sulfur Dioxide Analyzers

  8.1       Principle and Scope

  8.2       Range

  8.3       Interferences                                   2.74

  8.4       Precision, Accuracy and Stability       '

-------
                       LIST OF TABLES






                  Title                                     Page






4-1     Manufacturer's Data Sheet                            65




4-II    Analyzer Summary Sheet and Test Results              68




4-III   Instrument Operating Instructions                    71




4-IV    Linearity Summary Sheet                              73




4-V     Worksheet for Linear Equation        .                74




4-VI    Worksheet for Response Times                         75




4-VII  , Worksheet for Precision                              76




4-VIII  Interferent Test Concentrations               *       77




4-IX    Ppm vs. yg/iu  Conversion Factors for Selected Gases  79




5-1     Measuring Ranges and Sampling Rates of Common




        Air Analyzers                                        96




5-II    Methods far Generating and Determining Gas




        Concentrations                                       97-98




5-III   Absorbers for Producing Zero Air Up to 30 1/min.     99-101

-------
                    LIST OF ILLUSTRATIONS






FIGURE                                                      PAGE






 5-1    Diagram of gas mixing apparatus                     102




 5-II   Configurations of gas mixing chambers               103




 5-III  Gas permeation apparatus                            104




 5-IV   Permeation device holder                            105




 5-V    Ozone Generator                                     106




 6-1    Gas generating apparatus for calibrating




        ozone analyzers.                                    137




 6-II   Ozone generator                                     138




 6-III  Sampling train for referee ozone analysis           139




 7-1    Gas generating system for calibrating NO and




        N02 analysers.                                      169




 7-II   Sampling train for referee NO and N02 analysis.      170




 7-III  Calibrating solution dispenser                      171




 8-1    Gas generating system for calibrating S02




        analyzers.                                          192




 8-II   Sampling train for referee S02 analysis.            193




 9-1    Primary dilution system for CO in bags.             208




 9-II   Secondary dilution system for calibrating CO




        analyzers.                                          209




10-1    Diagram showing visual representation and




        interpretation of noise an time delays in




        analyzer response.                                  235




10-11   Diagram of pulse time                               235

-------
                            ACKNOWLEDGEMENT






The authors wish to acknowledge the assistance of many individuals




and organizations whose comments and suggestions made this guide




possible.  The contributions of Dr. John B. Clements and




Mr. Thomas W. Stanley who served as project officers; the technical




staff of the National Environmental Research Center, RTF;




Mr. Jay Sinnett, Office of Monitoring, EPA; the technical specialists




of the Air and Industrial Hygiene Laboratory; Dr. Ernest Yeager,




Case Western Reserve University: and Messrs. John Kinosian and




K. Nishikawa of the California Air Resources Board were especially




helpful.  Drs. -Albert Rocklin and Lionel Farber served as editorial




consultants. 'Our deepest thanks go to the many air analyzer




manufacturers and representatives for their assistance and loan




of their instruments without which the validity of this guide could




not have been established.

-------
                           ABSTRACT






Intelligent selection and application of atmospheric analyzers




require complete, consistent,  cohesive and unequivocal  information




regarding the physical and performance characteristics of these




instruments.  This guide is designed for use by operating personnel




and provides 1)  definitions for selected terms in air monitoring ?




2) a listing of  physical and performance characteristics for which




Information is needed, 3) recommended criteria for operating




instructions, 4) test procedures for evaluating physical and




performance characteristics of air analyzers and 5) procedures




for the calibration of analyzers for carbon monoxide, nitric




oxide, nitrogen  dioxide, ozone and oxidant, and sulfur dioxide.

-------
                             PREFACE









To assess air quality and to determine compliance with legally




established air quality standards, noxious components such as carbon




monoxide, oxidant, nitrogen dioxide, nitric oxide, sulfur dioxide,




hydrocarbons and suspended particulate matter have been monitored




in many areas of the country.  Effective April 30, 1971, the federal




Environmental Protection Agency  (EPA) has promulgated primary and




secondary air quality standards  and regulations for all the above




pollutants except nitric oxide (Federal Register, 36, No. 84, April




1971).  Additional regulations (Federal Register 36, No. 158, August




1971) require the States to follow a demanding time schedule to




achieve compliance with the primary and secondary standards.  The




regulation!? r.learly state that these rules :rs.y not in any way be




considered so as to permit deterioration of existing air quality that




is already well within the confines of the standards.  The federal




regulations further specify that pollutants shall be measured by




prescribed reference methods or methods which have been demonstrated




to be equivalent to the satisfaction of EPA.  This requirement




indicates that equivalency with  the reference procedure must be




demonstrated for current and. future air pollutant sampling and




analysis systems. ,









A substantial increase in air monitoring programs is thus in-




evitable in order to determine short and long term trends, to assess
                              ill.

-------
the effectiveness of control programs, to substantiate compliance


and to insure maintenance of existing clean air environments.  Fara


lei to this increase in surveillance activity will be a growing


demand for air monitoring instrumentation accompanied by foreseeable


advances and changes in air analyzer technology.  In order to deal


competently with this anticipated situation, meaningful information


is necessary to provide guidelines for the selection and use of air


monitoring instrument including'instructions on proper instal-


lation and maintenance.  Only by proper application of well-tested


performance evaluation procedures for air analyzers can one expect
;

to acquire valid and reliable aerometric data.





It is therefore gratifying that EPA chose to provide our laboratory,


which has decades of collective and productive experience with air


pollution measurements, with resources to develop and test system-


atically procedures for the evaluation of atmospheric analyzer per-


formance.  -It is also rewarding to realize that some results of our


efforts have been included in the promulgation of federal register


specifications for carbon monoxide and oxidant analyzers even be-


fore details of our work were finalized.  I trust that results being


produced now will be even more generally useful.  Any comments re-


garding the utility of these procedures and deficiencies and


omissions in this guide are welcome.
                                Peter K. Mueller, Ph.D.
                                Principal Investigator


                                June 1973
                             iv

-------
1.0    INTRODUCTION
       For measuring air pollutants, a variety of commercial monitoring

       instruments are available.  For intelligent selection and appli-

       cation of these instruments, reliable information concerning the

       physical and performance characteristics of these instruments is

       necessary.  The information provided must be complete, consistent,

       cohesive and unequivocal.  The terms used to describe these

       characteristics should be specific and well-defined.  The com-

       parison, selection and application of such instruments are

       enhanced when the specifications cover the same parameters
                                                           »
       and are based on uniform instrument evaluation.




       To meet some of these needs, this guide, which is designed for use

       by operating personnel, 1) provides definitions for selected terms

       commonly used in the description of air monitoring practices and

       instrumentation, 2) lists specific physical and performance character-

       istics for which information is routinely needed, 3) recommends the

       format and information to be included in instrument operating

       instructions, 4) describes test procedures fcr evaluating physical

       and performance characteristics of automated analyzers for atmospheric

       monitoring and 5) provides procedures for the calibration of atmo-

       spheric analyzers for carbon monoxide, nitric oxide, nitrogen dioxide,

       ozone and oxidant, and sulfur dioxide.

-------
2.0    MANUFACTURER'S SPECIFICATIONS









2.1    INTRODUCTION








       The purpose of this chapter is to achieve uniformity by reducing




       the wide variations that now exists in published air analyzer




       specifications by delineating a list of required specifications




       and other pertinent information under two categories, designated




       here as Class I and Class II.  The Class I category is intended




       to cover the minimum number of parameters the manufacturer




       should furnish so that the prospective user can judge the potential




       suitability of the instrument and can decide if further investigation




       is warranted.  It is not expected to provide sufficient information




       to justify a purchase.  Class I specifications are designed for adver-




       tisements in technical journals, product bulletins, new product




       flyers, and news articles.









       The Class II category includes information required for complete




       identification and characterization of the instrument on which a




       decision concerning the purchase can be based.  It is intended




       primarily for presentation in manufacturers' instrument brochures,




       specification sheets and instruction manuals.  In all cases,




       manufacturers should provide realistic and accurate performance




       data and describe their instruments in clearly defined terms, such
                                  —2—

-------
2.3.1  Instrument Description




       a.  Measurement principle - a detailed description of the principle


           on which the measurement is based.  Where appropriate, diagrams and


           photographs should be used (see b below).




       b.  Drawings - block, logic, or schematic diagrams of the basic com-


           ponents and of the important electrical, sample, solution, and

           optical pathways, with an explanation of the function of each.

           Photographs may also be used.


                V

       c.  Optional auxiliary equipment for use with the instrument.




       d.  Sale price of basic unit, optional equipment and accessory


           items.



            i
2.3.2  Installation and Operation




       a.  Physical size and shape of the instrument, with space require-

           ments for its installation, operation, and maintenance including

           allowances for auxiliary items and swinging space for doors and


           access panels.




       b.  Weight of each subunit making up the complete instrument.




       c.  Environmental requirements including the range of temperatures,


           humidity and, especially in the case of portable instruments,
                                   —5~

-------
           vibration conditions over which the instrument will meet  stated^




           performance specifications.








       d.  Air sampling requirements including sample flow rates and




           permissible differences in sample line pressure.









       e.  A list, with consumption rates, of all solutions, reagents,




           and gases required for start up, operation, and maintenance,




           but not including lubricating oils, or other materials for




           preventive maintenance.









       f.  A brief description and recommended frequency of .the calibration




           procedure(s).









2.3.3  Performanc e Charac ter is tics









       a.  Dynamic response including lag time,  rise time, fall time, and




           pulse  time.









       b.  Output signal  characteristics with respect to:   1) type (elect-




           rical, mechanical, optical,  etc.), 2)  magnitude (voltage, current,




           frequency,  resistance,  etc.), 3)  linearity (arithmetic, exponential,




           other  function), 4) impedance, and 5)  noise.








       c.  Audible noise.
                                  -6-

-------
3.0    OPERATING INSTRUCTIONS









3.1    INTRODUCTION









       Most instruments require information and directions concerning their




       proper operation and use.   To insure this, each instrument should be




       delivered with a manual of operating instructions.  The manual should




       contain a section  on:  1) operation and maintenance and 2) field repair




       or service.  These sections may be separately bound.









3.2    IDENTIFICATION









       The manual should  be identified with the model number of the instru-




       ment and should be current for that model.  Separately bound portions




       should be individually identified with the model number.  Loose leaf




       manuals should have the model number appear on every leaf.  Corrections




       or revisions should be clearly marked.  Where the manual covers several




       models, the. model  number of the instrument to which it applies should




       be clearly indicated.









       The manual should  be provided with a convenient reference form listing




       the model number,  serial number or run number, special options and




       modifications, location and telephone numbers of the manufacturer's




       representative and nearest service or repair center.
                                     —7—

-------
 3.3    OPERATION AND MAINTENANCE SECTION








       This section should provide information and instructions regarding




       principles, installation, initial start-up, hazards, calibration,




       sampling, maintenance actions, and limited trouble shooting, and




       should describe or contain any warranties or guarantees.  Addi-




       tionally, a section should be provided dealing with delivery, unpacking,




       and installation of the instrument.  This section should be accessible




       to the purchaser at or before the time of delivery of the instrument




       and should define whether the purchaser may, in the absence of an




       authorized manufacturer's representative:









       s.  Unpack the unit




       b.  Install the unit




       c.  Operate'the unit




       d.  Relocate the unit









       The Operation and Maintenance Section should cover,  but is not limited




       to the following topics:









3.3.1  Delivery and Unpacking Instructions









       The manual should furnish details concerning shipment acceptance.  A




       set of instructions for the orderly unpacking and inspection of the



       instrument should cover the following topics:
                                    -8-

-------
       a.   Unpacking




       b.   Warnings concerning dangers to the instrument and/or personnel




           during' unpacking.




       c.   Inspection




       d.   Procedures to be followed when the instrument arrives in damaged




           condition, when parts are missing, or when wrong parts have been




           sent.









3.3.2  Installation and Assembly









       The manual should describe the utilities and the physical environment




       that are required and should provide instructions for moving,  mounting,




       and installing the instrument, including details for connecting




       together various components and for attachment to utilities.  The




       following topics should be covered:









       a.  Electric requirements - voltage, power consumption (or current




           drain), frequency, number of phases, grounding.









       b.  Other utilities - fuel gas, hydrogen, Oxygen, other gases, water,




           vacuum, drains.'  These should be qualified by specifications for




           the quality or purity and limitations on flow capacity and pres-




           sure .








       c.  Physical environment - limitations on temperature, humidity,




           vibration, light; requirements for shielding or isolation of the
                                     -9-

-------
           instrument; provision of exhaust vents for safety and comfort;




           clearance or space needed for the operation and maintenance of




           the instrument.








       d.  Working surfaces and material of construction of surroundings -




           floor and bench support, floor covering,  bench top,  bench height,




           supporting struts or other suggested bracing.








       &.  Interconnection of instrument sections and attachments to utilities




           electrical, mechanical,  pneumatic.









3e3.3  Hazards and Precautions









       The manual should specify any hazards which could potentially injure




       personnel or damage the instrument.   Warnings should be  placed both




       in the manual and at appropriate locations on or in the  instrument.




       The instrument should be equipped with, or provisions made for the




       attachment of, effective exhaust and effluent disposal devices.  Clear




       instructions and warnings concerning proper disposal of  any potentially




       hazardous discharges should  be provided where appropriate on the




       instrument arid in the manual.  Tests for leaks on all lines, valves,




       and control devices exposed  to hazardous fluids such as  compressed,-




       flammable, toxic or corrosive gases  and liquids must be  specified.




       Emergency shut-down procedures should be clearly described.
                                    -10-

-------
3.3.4  Application, Principle of Operation, and Scope









       The manual should describe what the instrument measures, state the




       useful measuring range(s) and explain the principle of operation.




       Where appropriate, the description should include sketches, block




       diagrams or other pictorial representations of the instrument, the




       units comprising it,  and the function of each.  The effect of




       potential interferents should be described.









3.3.5  Air Sampling Requirements









       The method of  air sampling and any additional equipment required should




       be described.  Where  the instrument is intended for connection to an




       external sampling probe, the recommended probe materials, velocities




       and residence  times,  and dimensions should be specified.  The pre-




       ferred range of air sampling rates should be stated, and the permissible




       variations in  sampling rate and air pressure should be given.









3.3.6  Ou;tpu: t_S i gna1








       Most instruments are  designed to be coupled with signal processing




       and presentation devices such as recorders, integrators, analogue




       and digital displays,  computers, and telemetry systems.  The selection




       of an appropriate signal processing device depends on the characteristics




       of the output  signal  from the instrument.  Therefore, the manufacturer




       should specify the signal as to:
                                    -11-

-------
       a.  Type  - mechanical, electrical,  optical,  acoustical




       B.  Characteristic  - voltage,  current,  frequency,  continuity




          (steady or  chopped)




       c.  Linearity - arithmetic,  logarithmic,  etc.




       d.  "Matching requirement - output impedance, resistance,  back




          pressure.   Instructions  should  be given  for  detecting and




          isolating response difficulties arising  from mismatching of




          the signal  source and the  signal presentation  device.









3.3.7  Specification








       A technical summary of the instrument's physical and performance




       characteristics under various  conditions, as well  as information




       concerning input,  output, power  requirements,  reagents, and




       utilities as  listed in 2.2 and 2.3, should be  provided.








3.3.8  Start-up  and  Operation









       A precise, abbreviated summary of the operational  procedures should be




       provided  in  the manual and printed  on the instrument or on a permanent




       card affixed  to the instrument.  Photographs or  illustrations de-




       picting various views of  the instrument including  the controls suitably



       labeled are  helpful.









       Instructions  for placing  the unit in operation should include:
                                     -12-

-------
       a.   Procedures for procuring,  preparing,  and installing reagents,

           gases,  and other consumable items.



       b.   Identification, purpose of and procedures for operating all

           controls,  and adjustments.



       c.   Prestart-up checks



       d.   Step-wise  start-up and shut-down procedure



       e.   Directions for making response checks and adjustments.
3.3.9  Calibration
       Instructions and recommended frequency for precise  calibration of

       the instrument should be provided.   These should  include  a  description

       of the reference method and a list  of the necessary equipment and

       calibrating standard materials.   Sources, quality,  specifications,
                                     /
       and literature references should be included where  appropriate,
3.3.10 Maintenance
       A routine servicing schedule should be provided.   This  should  in-

       clude a check list of inspection points and maintenance procedures,

       a timetable for carrying out the servicing operations,  and  a form

       for keeping maintenance and repair records.
                                    -13-

-------
      Where the manufacturer wishes to restrict certain maintenance




      operations to specially qualified or authorized personnel,  the




      nature of the qualifications or authorization should be described,




      the responsibilities of both the manufacturer and the purchaser




      should be stated clearly, and the arrangements for maintenance




      should be specified.








      A recommended list of spare parts and subassemblies to be  kept on




      hand  should also be provided, with emphasis on those replacement items




       that  fail frequently or are difficult to procure.









3.3.11 Troubleshooting








       Check procedures  to be  followed in the  event  o±  instrument ±ailure




       should be provided.   Procedures should  include instructions for




       disassembly and  reassembly  along with warnings concerning  possible




       damage or contamination to  components and  steps  for  isolating and




       checking components,  sections  or  areas  to  determine  the  probable




       cause of failure,  and methods  for remedy.   This  may  be accomplished




       with a chart which outlines the symptoms,  the typical cause of




       failure, the unit or  units  which may be involved,  and the  recom-




       mended course of action.








3.3.12 Service Information









       A detailed statement  regarding the. manufacturer's repair policy




       should include:
                                    -14-

-------
       a.   Repairs which may be performed by the user or local service-
           man.
       b.   Repairs which are to be done only by the manufacturer because
           they are too difficult and/or hazardous.
       c.   Locations and telephone number of manufacturer's repair centers,
       d.   Acceptable packing procedures for returning the instrument or
           parts thereof to the manufacturer for repair.


3.3.13 Warranty


       Any guarantee or warranty should be a part of the manual and should
       clearly state which actions by the user will negate such warranty
       either completely or in part.


       Availability of routine maintenance and emergency repair contracts
       should be stated.  The manufacturer should also specify the time
       during which repair parts will continue to be available.
                *

3,4    FIELD BEPAIR OR SERVICE SECTION


       The Field Repair or Service Section should provide detailed infor-
       mation to diagnose instrument failure which will permit qualified
       instrument servicemen to repair and restore the. instrument to oper-
       ation.  It should include, but not be limited to, the following
       topics:
                                  -15-

-------
3.A.I  Physical Description



       A general description of the physical and operational characteristics

       of the instrument should include:



       a.  A physical description of the complete instrument.

       b«  A diagram showing the external view with all items such as meters,

           knobs, switches, lights, jacks and plugs labeled.

       c.  A physical description, of each subassembly.

       d.  Identification, effect of controls, and adjustments for ope-

           rating, calibrating, tuning, aligning, etc., each subassembly.

       e.  An illustration of the interior view with the cover removed and

           with adequate labels for tubes, transistors, resistors, circuit

           boards, capacitors, photometers, cells, motors, etc.  Color

           coding and views showing more than one level may be helpful.


                             /
 3.4.2  Mechanical Description



       A description of the function and operation of significant optical

       or mechanical units should include:



       a.  Diagrams of optical systems, explanations of their function,

           and detailed instruction for adjustment.



       b.  Diagrams of sampling and solution flow systems, descriptions  of

           their operation, and procedures for calibration and adjustment.
                                    -16-

-------
       c.   Descriptions and diagrams of servo-motors,  pumps,  fans,  cir-




           culators, heaters, etc.









       d.   A list of mechanical parts and acceptable commercial substitutes




           (if any) along with parts, numbers and descriptions.









       e.   A list of recommended spare parts to be kept on hand to  replace




           those that fail frequently or that are difficult to procure.









3.4.3  Electrical/Electronic Description









       A general discussion and description of the purpose and operational




       characteristics of each unit and subassembly should include:









       a.   Legible, comprehensive schematic and block or logic wiring




           diagrams for the power and the electronic circuits.









       b.   The physical location of each part in the instrument.









       c.   Identity of the test points to be used in making voltage, re-




           sistance, capacitance, continuity and waveform checks.









       d.   The voltages and waveforms expected or required between  test




           points.








       e.   A list of suggested spare electrical or electronic parts to be
                                   -17-

-------
           kept on hand, especially those that are likely to fail without




           warning or that are difficult to obtain.









       f.  Warnings concerning potential hazards.









3.4.4  MaJor_Calibration and Adjjas^tTnent









       Major overhaul, repair or replacement of components may cause a shift




       in the instrument's response which exceeds the range of normal oper-




       ational adjustments.  Some instruments provide wider range or coarse




       adjustments which can be used to compensate, for such shifts.  These




       adjustments may be locked or factory sealed to prevent unauthorized




       or inadvertent tampering.  NOTE:  The analyzer calibration should be




       checked with calibrating gases when changes in the coarse adjustments




       are made.









       Procedures for major calibration and adjustment should include:









       a.  A list of recommended equipment, special tools, chemicals,




           gases, materials and standards.









       b.  Description and location of adjustment screws, potentio-




           meters, trimmers, etc.









       c.  An explanation of the purpose for each step in the calibration




           procedure and directions for making necessary adjustments and




           response checks during calibration.







                                    -18-

-------
4.0    PROCEDURES FOR DETERMINING PERFORMANCE









4.1    GENERAL









4.1.1  Application









       This chapter contains  test procedures for determining the performance




       of air monitoring instruments.  The procedures are designed to estab-




       lish the analyzers' physical  and  operational  characteristics such as




       size, measuring range, noise, lower detectable limit, drifts, lags in




      • response, etc.  Procedures for  assessing the  adequacy of the instru-




       ments' operating instructions are also provided.









4.1.2  Gas Generation'









       Suggested procedures and  systems  for generating the  test gases required




       for determining analyzer  performance are given in Chapter 5 (Gas Gener-




       ation) .  To insure a stable output, the temperature  of the environment




       surrounding the gas generation  system should  be maintained to within




       ± 1 C  (2°F).  When permeation devices are employed,  the temperature of




       the device and the air passing  over the device should be maintained




       within ± 0.1°C.  The concentration of all test gases generated should be




       verified as directed in Step  4  of 4.1.6.
                                    -19-

-------
4.1.3  Test Facility Requirement^








       1.  Conduct all tests except that for the operating temperature




           range at 21 ± 1°C (70 ± 2CF).








       2.  To assess the effect of temperature changes on analyzer per-




           formance, the test facility should be capable of providing temper-




           atures between 4°C (39°F) and 43°C (110°) and within ± 1.0°C  (2°F).




          - An air conditioned room may be used, provided the range of temper-




           atures required for the tests can be attained.









       3.  The performance of most analyzers does not appear to be affected




           by exposure to the range of humidities normally encountered in




           air conditioned environments.  Control of humidity in the test




           environment, therefore, is not required.









       4.  The test facility should be sufficiently large to permit the




           testing of the analyzer(s) and have ample cooling capacity to




           absorb the heat generated by work personnel and the analyzer(s).









       5.  Conduct all tests except those involving response to powerline




           voltage at 117 ± 2 VAC and 60 ± 1 Hz.









4.1.4  Data Recording,  Key Words and Symbols









       1.  Forms;  Suggested forms for entering and recording results of the




           tests are shown in Tables 4-1, II and III at the end of this  chapter
                                   -20-

-------
2*  Keywords and Symbols;  Except as noted in Test 4.5 and 4.11, the



    following words and symbols are used in all equations and test



    procedures.  For definitions of related terms, see Chapter 10,



    Air Analyzer Terminology.







    a = intercept in the general form Y = a + bX







    b = slope:  the ratio of the analyzer output to



    pollutant concentration
    b  = the slope derived from the manufacturer's calibration
     m          r


    procedure.
                                          •r\

    C - pollutant concentration  (ppm, pg/nr , etc.)








    CV = coefficient of variation








    FS = full scale:  the maximum pollutant concentration that




    can be measured on a given range.  It is usually specified by




    the instrument manufacturer.








    I = interferent








    IE - interference equivalent
                              -21-

-------
LDL = lower detectable, limit









OTR = operating temperature range









OVR = operating voltage range









ppin = the volume (y£) of pollutant per volume (liters) of




air at the standard conditions of 25°C and 760 torr.









R = response:  the instrument's output readings after con-




version to concentration units (pg/m^, ppm,  etc.)









Referee method = one of the procedures in this guide used




to verify the concentration of the test gas.









Span = an instrument reading or test gas concentration equal




to 80 ±5% of full scale









t = time (hr, min, sec)










H» Ss5 V fc-r t-95> tf = See Test 4-n







T = temperature (°C, °F)









UO - unattended operation









X = symbol for pollutant concentration in Test 4.5
                         -22-

-------
           Y « reading or readout:  the instrument's output in the data

           display units such as chart or meter readings before conversion

           to concentration units



           A _
           Y - expected reading:  the analyzer output reading expected,

           based on the response previously obtained with concentrations

           of calibrating gases.




           Units:  the concentrations for all gases are expressed as

           ppm (pl/e,).  Convert ppm to yg/m^ as follows:



                           3 _       mol. wt. x 10-*
                       yg/nr = ppm x
                                         24.47
           or refer to Table 4-IX.




4.1.5  Installation and Start-up;




       1.  Install the analyzer(s) in the test facility described  in 4.1.3

           according to the instrument operating instructions.   Maintain an

           operations log book for the analyzer and the test facility.




           Include in the analyzer log the following information in addition

           to the items listed in Table 4-1,  Manufacturer's Data Sheet:
             (


           a.  Date of delivery

           b.  Description of special options obtained with or  modifications

               cade to the analyzer

           c.  Service information including the name of person to contact



                                    -23-

-------
    d.  Summary of warranty conditions and requirements




    e.  Record of adjustments and calibrations performed including




        settings of flowmeters, graduated dials, step switches, etc.




    £,  Record of all routine and non-routine maintenance and overhaul




        actions




    g.  Descriptions of instrument malfunctions encountered and corrective




        actions taken.  State cause of malfunction when known.









2.  Connect the output of the analyzer to a data acquisition system




    such as strip chart recorder, printer,  digital readout, etc.




    When a strip chart recorder is used,  the following minimum




    specifications are needed:









    &.  Chart v?idth — 10 inch









    b.  Chart speeds - 1,2 and  5 inch/hr   or




                       2.5, 5 and 12.5 cm/hr








    c.  Zero offset -  ± 10% FS









    d.  Span adjust -  ± 50% FS









    e.  Linearity  -   0.5% FS









    f.  Deadband  -    0.5% FS









    g.  Accuracy   -   i 0.1% FS
                            -24-

-------
           Verify the recorder's accuracy, linearity and deadband using a




           millivolt potentiometer or other appropriate test equipment.









           Any other type data acquisition system should be capable  of




           providing data equal to or superior to those obtained with the




           recorder.









           NOTE:  During the operating temperature range tests,  locate the




           data acquisition system outside the test facility.









       3.  Calibrate or verify all gas and reagent flows which affect analyser




           response with measuring devices such as a wet test  meter,  dry gas




           meter, a calibrated rotaraeter for gas flows and  a graduated buret,




           cylinder or volumetric pipet for liquid reagent  flows in  conjunction




           with a timing device such as a stopwatch,









       4.  Turn the instrument on.  Operate for 24 hours or longer before




           tests are begun.  The 24-hour warm-up time may be shortened




           when so stated in the operating instructions. For  determination




           of warm-up time, see Test 4.15.









4.1.6  General Procedures








       1.  After proper warm-up and stabilization while sampling zero air,




           proceed with the testing.  Begin recording analyzer readings




           when stable response has been attained.  (Look for  a  rapid
                                   -25-

-------
    response an apparent steady state which subsequently shifts




    upward or downward one or two hours later.)   Inability of the




    analyzer to attain stable response within two hours after the




    start of any test except Test 4.15, Warm-Up  Time, indicates




    possible instrument malfunction or faults in analyzer design.




    Discontinue testing and correct any problems before resuming




    the test.  Consult the instrument operating  instructions and/or




    manufacturers to resolve the problem(s).   At the conclusion of




    each test, check the instrument response  with zero air.   For




    some tests, the zero response may be set  at  a level greater




    than zero.  Average the initial and final zero responses to




    obtain the analyzer baseline.








    Unless otherwise specified,  when the difference between  the




    initial and final baseline readings in  a  test is greater than 2%




    of full scale the test results are considered invalid and the




    test should be repeated.









    A second invalidation indicates excessive instrument drift.




    Discontinue the test and correct the unit as above before




  -  resuming.









2.  Set the instrument controls to provide  the desired measuring




    range.  For an instrument with multiple range selection, Tests




    4.4 through 4.19 should be performed on each range.  When




    setting the analyzer measuring range, sampling a pollutant gas




    concentration (span gas) is required.  Refer to Chapters 6, 7,




    8 and 9 for details.





                              -26-

-------
       3.   Unless otherwise specifled„ set the instrument to read 5% of full




           scale while sampling zero air and 85% while sampling a span gas




           equal to 80% of full scale.  This permits observation of




           negative responses.









       4.   Verify concentration of each test gas by a referee method.




           Collect the samples from the delivery system as close as practical




           to the sample intake of the instrument under test.  Suggested




           procedures for verification are given in Chapter 5 (Gas Generation)









       5.   At the conclusion of each test, subtract the average zero reading




           from the analyzer reading Y and convert to concentration units




           R by dividing Y by the calibration slope b obtained in Table  4-IV




           as follows:
                               R-I
       6.  Enter all results obtained in Tests 4.2 to 4.19 under the  appro-




           priate sections in Table 4-II, Column B, and in Table III.









4.2    PHYSICAL CHARACTERISTICS









4.2.1  Definition:  the manufacturer, model, description, application,




       detection principle, and published performance specifications  of an




       analyzer.








4.2.2  Procedure:  during installation and start-up, enter the manufacturer's
                                   -27-

-------
       published  information  under  the  appropriate items in Table 4-1 and




       in Column  A of  Table 4-II.   For  range,  see Test 4.6.









       1. Weight:   determine the weight  of  each unit of the instrument and




          the  total weight of the  complete  instrument.









       2. Size:   measure  the overall dimensions of each unit of the instru-




          ment.   Include  all protruding  items such as knobs, meters, handles,




          flanges, and  feet  which  project beyond the instrument case or




          housing.









       3. Space  requirements:  measure the  dimensions required for the




          operation and maintenance of the  instrument.   These include the




          space  needed  for all auxiliary items and equipment such as pumps,




          reagent-containers, and  swing-out space for cabinet doors and




          panels.









4.3    INSTRUMENT OPERATING INSTRUCTIONS









4'3.1  Definition:  a set  of  directions which provides information regarding




       the measuring principle, details of  design, proper installation,




       application, operation, maintenance,  service and repair of an analyzer.




       (See  Chapter 3 for  specifics which ought to be furnished.)



                                       ;





4.3.2  Procedure;  based on the experiences encountered during installation




       and start-up, assess the operating instructions as follows:
                                    -28-

-------
1.  Study the requirements given in Chapter 3 (Operating Instructions>.









2.  Inspect the operating instructions to determine which of the




    topics listed in Table 4-III are included.









3.  Assess the adequacy and completeness of the information given




    for each topic in the instructions by assigning a numerical value




    between 0 and 10 to each topic according to the following scale.









                        SCORE               BASIS
0
1-2
3-5
6-8
9-10
Missing
Poor
Fair
Good
Excellent
NOTE:  Assign a score of 2 to instructions that are not current.




       Assign a score of 1 to instructions which do not apply to




       the instrument designated.









4.  Enter the results in Table 4-III.  The maximum score is 320.




    In general, a score of 300 or greater is excellent, between




    200 to 300 is adequate, and less than 200 is poor.
                             -29-

-------
4.4    CALIBRATION
4.4.1  Definitions:
       1.  Calibration:  the determination of the analyzer response when a




           series of calibrating gas concentrations are introduced to the




           analyzer inlet or artificial stimuli are presented to the detector.









           a.  Calibration, dynamic:  a performance test of the entire




               analyzer under simulated conditions in which the response to




               a calibrating gas over a known concentration range is deter-




               mined,  When reconciled with a static calibration, dynamic




               calibration also serves to verify 1) the correctness of




               reagent and sample air flow rates, 2) the efficiency of




               sample collection, 3) the integrity of the analyzer's




               plumbing and 4) the quality of any reagents aud/or reactants,








           b.  Calibration, static:  the determination of the analyzer, response




               when artificial stimuli such as standard calibrating solutions,




               resistors, screens, optical filters, electrical signals, are




               applied directly to the analyzer detector.  It is a performance




               test for the detection and signal presentation components of




               the instrument and is primarily applicable to analyzers using




               colorimetric and conductimetric detection schemes.  It is not




               a substitute for the dynamic calibration.
                                    -30-

-------
4.4.2  Procedure:



       1.  After proper warm-up'(See Test 4.1.6, Step 1), calibrate the

           analyzer for each measuring range as directed in the instrument

           operating instructions.



       2.  In the absence of specific calibration instructions, use the

           applicable calibration procedures in Chapters 6, 7, 8 or 9-



       3.  Set any instrument controls so that the instrument output reads

           5% of full scale while sampling zero air and reads 85% of full

           scale while sampling a calibrating gas equal to 80% of full scale.
      /


       4.  The calibration should consist of at least six concentration

           levels including 0, 50, and 80 percent of full scale.



       5.  Plot the net instrument readings Y on the vertical axis and the

           pollutant concentrations C on *the horizontal axis of an appropriate

           graph paper (rectilinear, semilog, log).



4.4.3  Calculations;


       1.  Use the method of least squares to determine the slope b and

           intercept a of the straight line that fits the data best as

           directed in Test 4.5.
                                   -31-

-------
       2. ,  The slope of the dynamic calibration curve should correspond within



           10% of any static calibration curve.  Differences greater than



           10% indicate problems with accuracy which may be related to sampling



           efficiency, leaks, malfunction in the analyzer, or reagent quality.



           Consult the operating instructions and/or the manufacturer to



           solve the problem(s)  before proceeding with the tests.







4.5    LINEARITY






       The following procedure is designed to discriminate between linear



     .  and non-linear relationships and to determine the nature of arithmetical!

                              /

       or exponentially linear response functions.
       1.  Deviation Pattern:  the configuration of the differences (+ or -)



           of the individual calibration points from the best-fit, least



           square line, based on the calibration points.  The pattern may



           be random or non-random.





                                                               A
       2.  Maximum Deviation:  the absolute difference (Y    - Y) of the point
                                                         max


           of maximum departure from the best-fit line expressed as percent



           of full scale.






       3.  Linear Response:   when the analyzer response to a series of inputs



           covering a specified concentration range is expressed as the



           of a straight line and the maximum deviation of the calibration
                                   -32-

-------
           points from this line is less than 4%, the response is considered



           linear.









       4.   Non-linear Response:  when the deviation pattern suggests non-




           random scatter of the calibration points about the best-fit




           calibration line and the maximum deviation observed is 4% or




           more, the response is considered non-linear.
4.5.2  Procedure:
       1.  Prepare a seven-column table as illustrated in Table 4-IV,




           Linearity Summary Sheet.









       2.  From the data collected in Test 4.4, enter the input pollutant




           concentrations X in column a and the corresponding analyzer




           readings, Y, in column b.









       3.  From the data in columns a and b, calculate the sums, sums of




           squares and sums of cross products necessary for fitting  the least




           squares line to the calibration points.   Enter the results in




           columns c, d and e.








       4.  In Table 4-V, Worksheet For Linear Relation, enter the vsums




           from Table 4-IV and perform the operations indicated in Steps  1




           through 11 to calculate the least squares line best fitting  the
                                   -33-

-------
    calibration points.  This worksheet produces a line  of  the form:
                      A
                       Y = a -f bX
5.  Calculate the expected reading Y predicted by the  equation  for each


    input concentration and enter, the results in column f, Table  4-TV.





6.  Calculate the difference between the analyzer reading Y  (column b)

                             A
    and the expected reading Y (column f) and record the positive and


    negative differences in column g,


            <


7.  Inspect the sequence of the +'s and -'s in column,g.  Assess  the


    general pattern of the sequence.  This pattern can be:





    a.  a random series of positive and negative values, e.g.,


        H—hf-—(-, suggesting a linear function.





    b.  negative, positive, then negative, e.g., —HH— suggesting


        a convex function.





    c.  positive, negative, followed by positive, e.g., -H	H-,


        suggesting a concave function.
                                                        A
8.  Select the largest value Y    from column e (Y    - Y) .
                              max             6   max    '

    Calculate the maximum deviation as % of full scale (FS) :
                            -34-

-------
                                             Y) x 100
                           Max dev =
                                            FS


4.5.3  Determination of Linearity



       1.  Inspect the results for the deviation pattern and maximum

           deviation obtained in Steps 7 and 8 above.



       2.  When the response is linear, proceed to Step 5 below.



       3.  When the response is non-linear, logarithmic transformation

           of the concentration and/or the response readings may yield

           a linear function.  Log-linear response patterns may be of the

           forms:



                       log Y = a + b log X                 (logarithmic)


           and


                       log Y = a + b X

                               or                          (semi-logarithmic)

                           Y = a + b log X



           To test these possibilities, convert the values in columns a

           and/or b of'Table 4-IV to their logarithms and proceed with

           Steps 3 through 8 of 4.5.2.



       4.  When the results still suggests the response is non-linear, place

           a check mark in the square designated "non-linear" in Table 4-IV and


                                    -35-

-------
           enter  the word  "non-linear" and a dash for "maximum deviation"




           in column B  of  Table 4-II.








           NOTE:  Due to the wide  variety  of relationships possible, the deri-




           vation of equations  that  fit hyperbolic and transcendental relation-




           ships  are usually not feasible.  Such relationships may be repre-




           sented by graphically fitting the best line through the calibration




           points.  The resulting  curve may then be used as a template to




           transform the net analyzer  reading to concentration.









       5.   'A linear relationship exists when the maximum deviation is less  than




           of 4%  or when there  is  a  random deviation pattern.








       6.   Based  on the conclusions  from Step 5  above, place a check mark in




           the appropriate square  (linear, non-linear or linear with poor




           precision) in Table  4-IV  and enter the corresponding term and




           maximum deviation value in  column B of Table 4-11.








4.6    MEASURING  RANGE
4.6.1  Definition:   the nominal minimum and jnaximum concentrations which the




       instrument is capable  of measuring.  Many  analyzers  provide multiple




       range selection capability  for  greater  accuracy and  ease of interpre-




       tation.   Range is usually specified by  stating the lower and upper




       pollutant concentrations that can be measured, as for example 0 to 1 p




       or 0 to  5 ppm.
                                    -36-

-------
4.6.2  Procedure;  Record the manufacturer's stated measuring range in con-




       centration units in column A, Table 4-II by stating the lower limit




       first, then the upper limit:  e.g., 0 to 1 ppm, 0 to 10 ppm.  For




       instruments with multiple range selection, record all ranges.  No




       test is required.









4.7    CALIBRATION ACCURACY









       This test applies to instruments calibrated according to the manufacturers'




       procedures and for those the manufacturers state no calibration is




       required.









4.7.1  Definition.;   the deviation between the slope b  of the curve obtained by




       the. inanufacturer or with his calibration procedure and the slope b




       obtained with a calibrating gas.  It is expressed in percent.









4.7.2  Procedure:








       1.  From the  manufacturer's calibration, determine the slope bm using




           the procedure described in Test 4.5.









       2.  Compare bn with the slope b obtained in Test 4.5.









4.7.3  Calculation;








       1.  Calculate the accuracy in percent as follows:
                                     —37—

-------
                                      b — bjn
                        Accuracy, % ~	  x 100

                                       bm
       2.  Enter the result in column B of Table 4-II.
4.8    DRIFT AT ZERO
4.8.1  Definition:  the deviation in analyzer output during a stated  time

       period, usually 24 hours, of unadjusted continuous operation when

       sampling zero air.  It is expressed in percent of full scale.
4.8.2  Procedure:
       1.  After proper warm-up (See 4.1.6, Step 1) on zero air, record the

           reading Y  and the time t ,
       2.  Continue sampling zero air for 24 hours.  Make no isanual adjust-

           ments to the electronic and/or gas and reagent flows during this

           test.   Automatic adjustments which are a part of the normal

           instrument operation are permitted.




       3.  Determine and record the reading Ym when the maximum departure

           from the baseline occurred and the reading ¥24 at the end of 24

           hours.




       4.  For instruments with multiple range selection, repeat Steps 1

           through 3 for each range.
                                    -38-

-------
4.8.3  Calculations:
       1.  Convert YQ, Yffl and Y2/,  to  concentration  units  RQ, Rffl and  R24  as


           directed in Step  5 of 4.1.6.




       2.  Calculate the maximum zero drift  in percent  of full  scale and


           the drift at the  end of 24 hours  as follows:
                                           R  - R
                                            m    o
           a.          Max  zero  drift,  % = 	——  x 100
                                             FS
                                             RO / —' R
                                              24    o
           b.           24 hr  zero  drift, % = —	  x 100
                                                FS
           c.  Enter  the  results  in Table 4-II.  Report values less than twice


               the noise  as  zero  (see Test 4.10 for noise at zero).





 4.9    DRIFT AT SPAN  (80% FULL  SCALE)




 4.9.1  Definition;  the deviation in analyzer output during a stated time


       period, usually 24 hours,  of unadjusted continuous operation when


       sampling a span gas equal  to 80% of full scale.  It is expressed as


       percent full scale.





 4.9.2  Procedure:




       1.  After proper warm-up (See 4.1.6,  Step 1) on zero air, record the


           zero baseline  reading  YQ.  Generate a pollutant test gas equal to
          \

           80 ± 5% of full scale.  Verify the pollutant concentration COQ by




                                   -39-

-------
           the  referee method specified in Chapter 5, Table 5-11.  Switch




           the  analyzer inlet to sample the test gas.  Sample until the




           instrument reading indicates a stable reading YgQ at the test  .




           gas  concentration.  At this time, start and record the time tQ.









       2.   Continue to sample for 24 hours.  Make no manual adjustments to




           the  electronics,  gas or reagent flows during this test.  Automatic




           adjustments which are part of the normal instrument operation are




           permitted.









       3.   At the end of the 24 hour test period, verify the pollutant concenti




           tion as in Step 1 above and switch the inlet to sample zero air.
       4.  Determine and record the reading Ym when the maximum departure




           from Y   occurred and the reading Y£  at the end of 24 hours.
       5.  For instruments with multiple range selection,  repeat Steps 1




           through 4 for each range.








4.9.3  Calculations;









       1.  Convert YgQ,  Ym and Y24 to concentration units  Rg0, f^ and £94




           as directed in 4.1.6, Step 5.









       2.  Calculate the maximum drift in percent of full  scale and the drilt




           at the end of 24 hours as  follows:
                                   -40-

-------
                                           Rm " R80
                       Max span drift, % - —	—-  x 100

                                              FS
                                             R   ~ R

           b.           24 Hr span drift, % = —	9.  x 100

                                                FS
           c.   Enter the results in Table 4-II.  Values less than twice the


               noise are reported as zero (see Test 4.10 for noise at span).





4.10   NOISE






4.10.1 Definition:  unwanted, spontaneous, short-term variations in analyzer


       response about the mean output, not caused by variations in pollutant


       concentration.  It is expressed as percent of full scale.





4;10.2 Procedure;






       1.   Noise at zero:





           a.   Examine the analyzer readings during the first 30-ininute interval


               from the data in Test 4.8.  Select the maximum reading Y,  and


               the minimum reading Y£ during this period.





           b.   Convert Y^ and Y? to concentration units R-^ and R£ as directed


               in 4.1.6, Step 5.
                                    -41-

-------
    c.  Calculate the noise at zero in percent of full scale as follows;
                                    i -  o
                Noise at zero, % = 	  x 100
                                     FS
    d.  Enter the result in Table 4-11.

                              i


    e.  For -analyzers with multiple range selection, determine the


        noise at zero for each range.




2.  Noise at 50%




    a.  After proper warm-up (See 4.1.6, Step 1) on zero air, generate a


        pollutant test gas concentration equal to 50 ± 5% of full scale.


        Switch the analyzer inlet to sample the test gas.  Verify the


        pollutant concentration and record the reading YCQ when a stable


        rea'ding is obtained .




    b.  Continue to sample for one hour.  Examine the readings during


        the first 30 minutes from the one-hour period.  Select the


        maximum reading Y, and the minimum reading ^L^ during this


        interval.
    c.  Convert Y^ and Y- to concentration units R, and R«.
    d.  Calculate the noise at 50% of full scale in percent of full seal


        as follows:
                             -42-

-------
                Noise at 50% of FS » -1-1—1  x 10Q

                                        FS
    e.   Enter the result in Table 4-II.





    f.   For analyzers with multiple range selection,  determine  the


        noise at 50% on each range.





3.  Noise at Span:





    a.   Examine the readings during the  first 30 minutes  from the data


        in Step 2 of Test 4.9.2.   Select the maximum  reading  Y^ and


        mini mum reading Y£.





    b.   Convert T^ and Y£ to concentration units Hi and R2.





    c.   Calculate the noise at span as follows:
                                R! - R2
                Noise at span = 	— x 100
                                  FS
    d.  Enter the result in Table 4-II.





    e.  For instruments with multiple range selection,  determine


        the noise at span for each ranges
                            -43-

-------
4.11   PRECISION AM) RESPONSE TIMES









       The tests for determining time delays in analyzer response  (lag,  rise,




       fall, etc.) are conducted in conjunction with the measurement  of




       precision.









4.11.1 Definitions:








       1.  Precision:  the variability in repeated measurements of  the same




           pollutant concentration expressed as the coefficient of  variation,




           i.e., the standard deviation of the individual results expressed




           as a percent of the mean.









       2.  Response Times (See Figure 10-1)









           a.  Lag time (initial response time), t.:  the interval  between the




               time to, when a step changer (increase or decrease) in  pollutant




               concentration is made, to the time t.^ when the instrument  indi-




               cates a response equal to twice the noise.
           b.  Time to 95%, tg5:  the interval between the time  tQ, when  a




               steP increase in pollutant concentration is made,  to the time tjj




               when the instrument indicates a response equal  to 95%  of the step




               chance.
                                    -44-

-------
                      t   f= t-    ~ f-
                       95   C95%    o
    Similarly, tgQ corresponds to the  time  to  indicate  90% and


    t!00 ttie time to indicate 100%  of  the step change.




c.  Rise time, tr:  the interval between the time  to 100% (t100)


    and the lag time (t.).
                    £r "
d.  Time to -95%, t_qc:  the interval between the time t  when


    a step decrease in pollutant  concentration is made to the time


    t new when the instrument indicates -95% of the step change.
    Similarly, t_QQ corresponds  to  the time to indicate -90% and


    t-100 ttie time to indicate -100% of the step change.




e.  Fall time, t,.:  the interval between the time  to -100%  Ct_-,Qg)


    and the lag time (t,,) .  Fall time is not necessarily  equal to


    rise time.
                    tf = t-100 "  t£
                         -45-

-------
4.11.2 Procedure:
       Note:  For this test, a timing device is required with  a  resolution of


       at least 0.1 second, such as a watch with a sweep second  hand or  a sto]


       watch.




       1.  After proper warm-up (See 4.1.6, Step 1), sample  zero air until a


           stable baseline concentration RQ (after conversion  from reading-Y)


           is recorded,




       2.  Generate a test gas equal to 50 ± 5% of full scale.   Switch the


           instrument inlet from the zero gas line to the  test gas line.


           (NOTE:  Some instruments are equipped with controls for switching


           from sampling ambient air to sampling zero air.   These should not


           be used for determining response times as they  may  introduce

                                          v
           additional lag times which are not part of normal operation.)


           Verify the pollutant concentration C.-Q and note the instrument


           response ^Q-  Switch the instrument inlet to sample  zero air.





       3.  When a stable response on zero air is obtained, switch the


          , inlet to sample the test gas and begin the time to(up) by


           starting the timer or noting the time.





       4.  When the analyzer indicates a change in response  equal to twice


           the nqise  (see Step 1 in Test 4.10.2 for noise  at zero), note


           and record the time t.
                                     -46-

-------
      5.  Similarly,  note  and record the times tg^ and t-^Q when the instrument




          indicates 95%  and  100% of R5Q.









      6.  Switch the  inlet to sample zero air and begin the time tQ(down)




          by  starting the  timer or noting the time.









          Similarly record the times for the analyzer to indicate a change




          equal to twice the noise (tj_) and the times to reach -95% (t_^^)




          and -100% (t_100)  of %Q*  Note:  See SteP 2 in Procedure 4.10.2




          for noise at 50% FS.









       7.  Repeat Steps 3 through 6 nine more times.









4.11.3 Calculations:









       1.  Response Times:








           a.   From the data in Steps 3 through 7, determine the time intervals




               required for the instrument to respond to a step increase in con-




               centration for each trial as follows:









               Lag time up, t^ = t^ - to(up)









               Time to 95%, tg5 = tg5% - to(up)
               Time to 100%, t100 - t100% - t0(up)
                                    -47-

-------
        Rise time, tr - t1QO - t&








    b.  Record the values in Table 4-IV.








    c.  Conversely, determine the time intervals required for the




        instrument to respond to a step decrease in pollutant concen-




        trations for each trial as follows:








        Lag time, down, t_^ = t. - tQ(down)








        Time to -95%, t_95 = t_95% - to(down)








        Time to -100%, t_100 = t_100% - tQ(down)
Fall time, tf = t_
                          10Q
    d.  Record the values in Table 4-VI.









    e.  Perform the operations indicated in Table 4-VT and determine




        the mean time and coefficient of variation for each response




        time characteristic.  Enter the results in Table 4-II.









2.  Precision:









    Perform the operations indicated in Table 4-VII and determine the




    coefficient of variation CV at 50% of full scale.  Enter the res«-




    in Table 4-II.
                             -48-

-------
4,12   PULSE TIME


4.12.1 Definition:  the minimum time a pollutant concentration must persist for
       the analyzer to register a peak response equal to the pollutant
       concentration (See Figure 10-11).

4.12.2 Procedure;

       1.  After proper warm-up (See 4.1.6, Step 1) on zero air, record the
           zero baseline response RQ.

       2.  Generate a test gas equal to 50 ± 5% of full scale*.  Switch the
           analyzer inlet to sample the test gas and record the time t .
           Continue timing and record the time t-jnny when the instrument
           indicates 100% of R5Q.
                   /
       3.  Switch the inlet to sample zero air.  From the data in Step 2,
           determine
                               t!00 ~ t!00%   to
       4.  Switch the inlet to sample the test gas and record the time tQ.
           Continue sampling for a. period equal to 90% of t-^Qy and quickly
           switch the inlet to zero air.  The peak instrument reading may be
           equal to or less than RT-
                                    -49-

-------
       5.   When the reading Is less then R5Q,  repeat Step 4 but increase



           the sampling time.   This time can be approximated by interpolating



           between t9Q and t10Q.   Repeat Step 4 until the minimum sampling time



           t .  for the analyzer  to indicate RCA is obtained.
            min             J                  50






       6.   When the reading in Step 4 equals R50, repeat Step 4 but reduce



           the sampling time by increments of 10% of tj_QQ until a response of



           less than R5Q is obtained.  Repeat Step 5 above until t^^ is



           obtained.







4.12.3 Calculations:
       Determine the pulse time t  from the data in Steps 5 and 6 as follows:
       Enter the results in Table 4-II.







4.13   LOWER DETECTABLE LIMIT







4.13.1 Definition;  the smallest pollutant concentration which produces a



       signal equal to twice the noise.   It is expressed in concentration



       units.
                                   -50-

-------
4.13.2 Procedure:
       1.  After proper warm-up (See 4.1.6, Step 1) on zero air, record the


           zero baseline response R .  Obtain the noise at zero from the data


           in Step 1 of 4.10.2.
       2.  Generate a test gas concentration equal to between 5 and 10% o':


           full scale.  Switch the inlet to sample the test gas.  Verify


           the pollutant concentration C and record the response R.





       3.  Based on the response R, calculate the approximate concentration C


           which will produce a response R^ equal to twice the noise.  Adjust


           the pollutant and/or diluent gas flowrates to produce this concen-


           tration.  Continue to sample and note the analyzer response.





       4.  When the response R^ obtained in Step 3 is greater than twice the


           noise, decrease the pollutant concentration Cm stepwise until a


           response equal to twice the noise is obtained.





       5.  Conversely, when the response R_ obtained in Step 3 is less than


           twice the noise, increase the concentration Cm stepwise until a


           response R_ equal to twice the noise is obtained.





4.13.3 Calculations;





       From the data in Steps 4 or 5 above, enter the smallest concentration


       C    that produced a response equal to twice the noise in Table 4-II as
        min
                                    -51-

-------
       the lower detectable limit (LDL).
                                   LDL «* C .
                                          mxn
4.14   ANALYZER DEADBAND








4.14.1 Definition:  the range, in percent of full scale, through which the
       -•—"""•	•• ••	—..—                                            \



       pollutant concentration may be varied without initiating a response




       equal to or greater than twice the noise.  (The term deadband is




       commonly applied to servo-systems such as recorders but can be  applied




       to continuous analyzers as well.)









4.14.2 Procedure;








       1.  After proper warm-up (See 4.1.6, Step 1) on zero air, record the




           zero baseline response Ro.









       2.  Generate a test gas equal to 50 ± 5% of full scale.  Switch the




           analyzer inlet to sample the test gas.  Verify the pollutant concen-




           tration CCQ and record the response £50•









       3.  Increase the test gas concentration to 65 ± 5% of full scale by




           varying the pollutant and/or diluent gas flowrates.  Verify the




           pollutant concentration CW and record the analyzer response Rgc-









       4.  Obtain the noise at 50% FS from the data in 4.10.2, Step  2.

-------
      5.  By interpolating  between R^Q and R,.^,  determine the concentration


          CJL that will  produce  an increase in response (AR)  from RcQ equal


          to twice  the  noise.   Generate concentration CT  as  in Step 3 above.


          Note the  instrument response.  The change from  RrQ should be equal


          to twice  the  noise.   If not, repeat the procedure  but adjust the


          concentration so  that the analyzer produces a change in response


          equal to  twice the noise.






       6-  Restore  the test  gas  to the concentration C^Q as used in Step 2  and


          record the response R5Q.






       7.  Repeat Steps  3, 4, and 5 using a test  gas concentration equal


          to 35 ±  5% of full scale.  Verify the  pollutant concentration C_


          and  similarly determine the concentration Co that  will produce a


          decrease  in response  from R-, equal to twice the noise.






4.14.3 Calculations:  determine  the analyzer deadband in percent of full scale


       as  follows:



                                              (C1 - C2>
                      Analyzer  Deadband, % = 	 2 100

                                                  FS




       Enter the results in  Table 4-II.






4.15   WARM-UP  TIME





4.15.1 Definition:  ^the  elapsed  time necessary after start-up for the analyzer


       to  meet  performance specifications when it has been shut down for at



       least 24 hours.
                                   -53-

-------
4.15.2 Procedure:
       1.  After proper warm-up  (See 4.1.6,  Step 1) on zero air, record the



           zero baseline  response  R .
       2.  Generate a  test  gas  equal  to 50 ± 10% of full scale.  Switch the



           analyzer inlet to  sample the test gas.  Verify the pollutant con-



           centration  Cr0 and record  the instrument response B-cn-






       3.  Switch  the  inlet to  sample zero air.   When a stable respor.se is



           obtained, turn the instrument off for at least 24 hours.  Continue



           generation  of the  test  gas.






       4.  Restart the analyzer, switch the inlet to the test gas and record



           the  time t  .






       5.  Record  the  time  t  when  a stable response at R,-,, is obtained.



           (Precaution:  Watch  for a  rapid response to an apparent steady



           state reading which  subsequently shifts upward or downward within



           an hour or  two later to a  new stable  reading.)






•4.15.3 Calculations:   determine x
-------
4.16   INTERFERENCES



4.16.1 Definition:



       1.  Interference equivalent IE:  that portion of the indicated pollutant

           concentration caused by the presence of an interferent.  It Is

           expressed as the ratio, in percent, of the indicated concentration

           R^ with respect to the concentration of interferent C^.



       2.  Interferent:  a substance or condition other than the one being

           measured which modifies or perturbs the output of an analyzer.


                                                            9
       3.  Prefilters:  filters or scrubbers designed to remove interferents.

           These are used by some air analyzers to increase the specificitv of

           detection.



4.16.2 Procedure;



       1.  General:  Test atmospheres containing interferents  are generated

           by mixing each interferent with the pollutant.  The interferents

           and their concentrations to be used for most of the currently avail-

           able detection methods are listed in Table 4-VIII.   Procedures for

           verifying their concentrations are given in Table 5-II in Chapter 5.

           Tests with interferents known to react in the gas phase with the

           pollutant are conducted in the absence of the pollutant.  The

           addition of the interferent to the gas system should not alter the

           pollutant concentration.
                                    -55-

-------
2.  Prefliters:  Any prefliters should be replaced or serviced as




    directed in the instrument's operating instructions.  In addition,




    the instrument should sample an interferent concentration equal




    to the value specified in Table 4-VIII through the prefilter for




    at least six continuous hours.









3.  Test Gases:









    In addition to the zero air (Z) for analyzer warm-up, generate




    three test gases:








    a.  P = flow of zero air containing pollutant which when mixed




        with I or A will produce the desired pollutant test




        concentration.









    b.  I = flow of interferent gas required to produce the inter-




        ferent concentration specified in Table 4-VIII when mixed




        with P.









    c.  A ~ flow of zero air equal to flow I.









    To permit measurement of negative responses, adjust the analyzer




    zero baseline, if possible, to read 10% of full scale.
                            -56-

-------
      4.  Test Procedure Without  Pollutant;
          a.  After proper warm-up  (See  4.1.6,  Step  1)  of  the  analyzer  on zero




              air, record the  baseline response R .
          b.   Generate  P (containing no  pollutant)  and  I.  Mix P with I.




               Switch  the analyzer  to sample  the  gas mixture.   Verify  the




               interferent concentration  C^ and record the  response RJ.









          c.   Switch  the analyzer  to sample  zero air Z.  Record the




               response  RQ.









       5.  Test Procedure with Pollutant  (Omit if interferent and pollutant




          cannot be mixed):









          a.   Generate  and mix P and A to  produce a pollutant  concentration




               equal to  50 ±  5% of  full scale.  Switch the  analyzer to  sample




               the test  gas.   Verify the  pollutant concentration Cp and record




               the response R .








          b.   Generate  I and substitute  for  A to produce the interferent




               concentration  C-^ used in Step  4b above.   Record  the response Rp









4.16.3 Calculations:
       Determine  the interference equivalent  IE  at  pollutant  concentration  C
                                                                            P
                                   -57-

-------
       as follows:
                                          R. ~ R
                       IE (when C  = 0) = - - -  x 100
                                             V»
                       IE (when Cn ^ 0) = ~	~ x 100
                                 P            Ci
4.17   OPERATING TEMPERATURE RANGE



4.17.1 Definition:  the range of ambient temperatures through which  the  analyzer

       will meet performance specifications.




4.17.2 Procedure;



       1.  Install and operate the analyzer in the test facility (4.1.3).  Set

           the test' facility controls to provide a temperature  (T^i)  of  21 ± 1°C

           (70 ± 2°F) .  Verify the temperature with an appropriate  temperature

           measuring  instrument such as a .laboratory mercury thermometer.




       2.  After the  analyzer is properly warmed up  (See  4.1.6,  Step 1)  or.

           zero air,  record the zero baseline response R  .




       3.  Generate a test gas concentration equal to 50  ±  5%  of full scale.

           Switch the analyzer inlet to sample the test gas.   Verify the

           pollutant  concentration CCA and  record the response R01 ,
                                     -58-

-------
 4.  Reset the test facility controls to provide the lowest possible



     temperature T^ at or above but not lower than 4°C (39°F).








 5.  Continue to operate the analyzer until a stable response is



     obtained.  Record the analyzer response R/.








 6.  When the difference between R^i and R, is equal to or less than



     twice the noise, record the temperature T, and proceed to Step 9



     below.








 7.  Conversely, when the difference is greater than twice the noise,



     estimate the temperature T,, at which the change in response will



     eaual twice the noise bv interpolation between Tm and T/.
      *                     "                        ^ JL      -r







 8.  Reset the test facility controls to provide the temperature equal



     to T,.  Continue to sample as in Step 5 above and note the analyzer
         As


     response.  Repeat Steps 6, 7 and 8 until the temperature T, that



     causes a change in response equal to twice the noise is obtained.








 9.  Restore the temperature of the test facility to T2,.  Continue to



     sample as in Step 5 until a stable response at R£J is obtained.








10.  Reset the test facility controls to provide a temperature T^



     as  high as possible, but not exceeding 43°C (110°F).  Follow



     the procedures in Steps 5 through 8 and similarly determine



     the upper temperature Th which will produce a change in res-




     ponse equal to twice the noise.





                             -59-

-------
4.17.3 Calculations:
       Report the operating temperature range OTR by stating the  lower




       temperature T£ and the upper temperature Th at which the changes




       in response were equal to twice the noise:
                               OTR = T£ to Th
4.18   OPERATING VOLTAGE RANGE









4.18.1 Definition;  the range of powerline voltages through which the




       analyzer will meet performance specifications.









4.18.2 Procedure:  This test is conducted by first determining the analyzer respm




       to a decrease in voltage (from 117 VAC) and second by determining the




       response to an increase in voltage.









       1.  Install a variable transformer between the source of electrical




           power and the analyzer.  The transformer should have an adjust-




           able output between 0 and 140 VAC and be capable of delivering




           the amperage requirement of the analyzer.









       2.  Set the transformer to provide 117 ± 1 VAC.  Verify this voltage




           V117 and a1"^ subsecluent: t£st voltages with a voltmeter.








       3.  After proper analyzer warm-up on zero air, record baseline  response





           V
                                    -60-

-------
4.  Generate a test gas concentration equal to 50 ± 10% of full  scale



    Switch' the analyzer to sample the test gas.  Verify the pollutant



    concentration C^g and record the analyzer response RIIT-







5.  Set the transformer to provide 90 V  (VQQ) .  Continue to



    sample until a stable response is obtained and record RgQ.








6.  Determine the difference between the response Rj^y and RQ(V
    a.  When the difference is equal to or less than twice the noise s



        record VQQ and proceed to Step 9 below.








    b.  When the difference is greater than twice the noise, estimate



        the voltage V0 at which the change in response will equal twice
          " .          Af


        the .noise by interpolating between V    and V^,-,.








7.  Reset the transformer to provide V^.  Continue to sample as in



    Step 5 until a stable response is obtained.  Repeat Steps 6a and



    6b until the voltage V.  that produces a change in response equal



    to twice the noise is obtained.  Record V^x.








8.  Restore the transformer to V^-jj.  Sample as in Step 5 until a stable



    response at RCQ is obtained.








9.  Adjust the transformer to provide 130 VAC (Vj^g) .  Continue to



    sample the test gas and determine RJ^Q as in Step 5 above.
                             -61-

-------
       10.   Determine the difference between the response R;QJ and
            a.   When the difference is equal to or less than twice the noise,




                record V    and proceed to 4.18.3.
            b.   When the difference is greater than twice the noise, similarly




                determine the highest voltage V,   which will produce a change




                in response equal to twice the noise as in Steps 6 and 7.









       11.  Repeat Step 8.









4.18.3  Calculations ;:








        Report the operating voltage range OYR by stating the lower voltage




        (either VQQ or V,,x) first,  and then the upper voltage V,™ or V,  ,




        e.g.,:









                               OVR = 90 to 130 V








 4.19   UNATTENDED OPERATION









 4.19.1 Definition:  the period of  time during which the analyzer can be




        expected to operate unattended within the specifications.








 4.19.2 Procedure:









        Start  this  test with the maximum amount of consumable items  (reagents,




        gases,  chart paper, etc.) recommended in the instrument operating





                                     -62-

-------
instructions.  No manual adjustments to the electronic, gas  or reagent




flows are permitted during this test.  Automatic adjustments which are




a part of normal instrument operation are permitted at any time.









1.  After proper analyzer warm-up  (See 4.1.6,  Step 1) on zero air,




    record zero baseline response  R  .









2.  Generate a test gas concentration equal to 80 ± 5% of full scale.




    Switch to sample  the test  gas.  When a stable response is obtained,




    verify the pollutant concentration Cgg and record the response Rq^




    and the time t  .









3.  Switch the analyzer to sample  ambient air, but continue  to generate




    the test gas.









4.  At the end of 25  hours, switch the analyzer to sample zero air.  •




    Record the zero baseline and switch to the test gas.  "Verify the




    pollutant concentration COQ and- record the response B-gQ-









5.  Continue Steps  3  and 4 for 100 hrs or until the response at




        either:
    a.  exceeds  the manufacturer's  stated  specifications for noise or




        drift at zero and  span, or








    b.  exceeds  the noise  or drift  values  obtained in Procedures 4.8,




        4.9 or 4.10 or






                             -63-

-------
           c.  the unit is no longer operational (i.e., out of supplies,




               analyzer stoppage, analyzer breakdown, etc.).
       6.  Hote the time t
4.19.3 Calculations:
       Determine the unattended operation  time UO as follows;
                       U0 '
                                   -64-

-------
                                   TABLE 4-1

                           MANUFACTURER'S DATA SHEET
    Date:
    Test  Date(s):
A.  IDENTIFICATION

    1.   Manufacturer
         Phone
        Address
    2.  Vendor
        Address
   ' 3.   Tradename(s) :
    5.  Serial No(s') :
    6.  Measurement Principle(s):

B.  APPLICATION

    1.  Pollutant(s) :
C.  PHYSICAL

    1.   Size,  Weight:

          Description  of Unit
       Overall Size, and Weight
         Phone
         4.  Model(s):
          •Measuring Range
                 ppm
  Dimensions
(W x H x D,  cm)
Weight
                                     -65-

-------
TABLE 4-1 (Continued)


2,  Space required (H x W x D, cm):

3.  Mobility:   Mobile 	    Portable

4.  Signal Output
                              Output
         Pollutant            Range
                                                Stationary
                                                 Units
                                            (volt, ma, 0)
                                        Linearity
D.  REQUIREMENTS

    1.  Sampling Rate, 1/min

    2.  Reagent(s):

                Item
    3.  Utilities

                Item
    4.  Power
          Voltage
        (AC or DC)
                  Consumption
                 Rate (ml/lain)
                                                          Storage
                                                        Captacity
                  Consumption Rate
                       (£/min)
Range
                                Current
                                (amp, ma)
Regulation
Required, %
                                     -66-

-------
   TABLE 4-1 (Continued)




   5.  Calibration




       Dynamic 	      Static  (specify method)




E. SPECIAL FEATURES
                                     -67-

-------
                               TABLE 4-II

                ANALYZER SUMMARY SHEET AND TEST RESULTS
Test Procedure                            (A)
     ,                                Manufacturer's            Test Results
                                     Specifications

4.2  Physical Description

     a.  Weight (kg)                _______ _       _________ _ _

     b.  Size, W x H x D (cm)       ___ _ ______       ____________
     c.  Space Requirements
          W x H x D (cm)
4.3  Instrument Operating Instructions

     a.  Supplied ?	
     b.  Overall score
         (see Table 4-III)

4.5  Linearity

    a.  Response (linear, concave,
                  convex)

    b.  Maximum deviation (%FS)

4.6  Range(s), (ppm)

     a.

     b.

     c.

4.7  Accuracy (%)

4.8  Drift at Zero  (% FS)

     a.  24 hours

     b.  Maximum

4.9  Drift at Span  (% FS)   -

     a.  24 hours

     b.  Maximum
                                  -68-

-------
TABLE 4-II (Continued)

Test Procedure                            (A)                       (B)
                                     Manufacturer's           Test Results
                                     Specifications
4.10  Noise (% FS)

      a.  at 0%		

      b.  at 50%		

      c.  at-span  (80%)
4.11  Precision and Response Times-
      taean of 10 trials  (sec, min)
      and CV, %

      a.  Lag time  (up),  t£

      b.  time to 95%,  t

      c.  time to 100%,  t
                          100
      d.  rise time, t

      e.  lag time  (down). t
                            "**J6

      f.  time to -95%,  t_93

      g.  time to -100%,  t

      h.  fall time, t^

      i.  precision at 50% FS,
          (CV, %)

4.12  Pulse Time, t  (inin)

4.13  Lower Detectable Limit (ppm)

4.14  Analyzer Beadband  (% FS)

4.15  Warm-up time  (min)

4.16  Interferences (IE, %)

           Interferents

      a. T ^	

      b.	
                      V
      c. 	

      d.

-------
TABLE 4- II (Continued)
Test Procedure                                       ,            „ ^. D
                                        Manufacturer's           Test Results
                                        Specifications
4.17  Operating Temperature Range,

        OTR (°C)


4,18  Operating Voltage Range,

        OVR (volts)



4.19  Unattended Operation, UO  (hr)
                                 -70-

-------
                             TABLE 4-III

                  INSTRUMENT OPERATING INSTRUCTIONS


                                                        SCORE

1.   Instructions supplied

2.   Current (up-to-date)	

3t   Applies to instrument designated                    ______

4.   Includes Operation & Maintenance Section
      and Field Repair or Service Section               _____

5.   Identification

    a.  Model                                           ______

    b.  Name, address & phone of manufacturer           	

    c.  Name, address & phone of representative or
          vendor	

    d.  Name, address & phone of nearest service
          center                                        	

    e.  Form for recording serial nos., modifications,
          date of purchase,  etc.                           _

6.   Operation & Maintenance Instructions

    a.  Delivery & unpacking instructions	

    b.  Assembly & installation instructions            	

    c.  Electrical requirements                         _____

    d.  Utilities requirements                          _____

    e.  Air sample flow and pressure requirements       	

    f.  Environmental requirements                      	

    g.  Output signal characteristics                   	

    h.  Start-up & operation instructions	
               i
    i.  Calibration procedure & intervals	

    j.  Maintenance procedures & intervals              	

    k.  List of recommended spare parts                 	
                               —71—

-------
TABLE 4-III (Continued)


                                                        SCORE

    1.  Troubleshooting                                 	

    m.  Service information	_,

    n.  Warranty	

7.  Application

    a.  Purpose & use of instrument                     	

    b.  Measuring range(s)                              	
    c.  Description of operating principle & flow
          diagram

  •  d.  Physical and performance specifications

8.  Repair Instructions

    a.  Physical description

    b.  Mechanical description

    c.,  Electrical/electronic description

    d.  Calibration & adjustment procedures

    e.  Field repair procedures
                                                            TOTAL SCORE
                                -72-

-------
                                              Table 4 - IV

                                     LINEARITY SUMMARY SHEET
(a)
X
(input)









2X=
(b)
Y
(analyzer
reading)




" " 	




2Y=
(c)
X2









2X2=
(d)
Y2









2Y2=
(e)
XY









2XY=
(0
Y
(expected
reading)








1

(8)
Deviation
(b) - (0










Equation of the line:
  ANALYZER
   READING
=	+ 	 X   InPut
 (intercept)    (slope)      Cone
              Deviation pattern
                                   Maximum deviation
 LINEARITY:    '—'   LINEAR - maximum deviation <4% and any deviation pattern
                Q   NON-LINEAR - maximum deviation">4%, and convex or concave delation pattern
                Fl   LINEAR (poor precision) - max deviation ^4% and random deviation pattern
                                              -73-

-------
                              Table 4 -  V
                WORKSHEET FOR LINEAR RELATION'
X denotes (input concentration]

SX=	
                   Y denotes (analyzer reading)
                            Number of Points:  n =
                  Step  (1)  ZXY

                        (2)  (2X)(SY)/n

                        (3)  Sxy = (l)-(2
(4)  X = 2X/n

(5)  Y = 2Y/n

(6)  SX2

(7)  (SX)2/n

(8)  Sxx = (6)- (7)  =

(9)  b=_JLy = [3)   =
       Sxx  (8)
(slope)
                   (10) bX = (9) X (4) =

                       a = Y-bX
                                        (intercept)
                           Equation ot the line:
                          9 = a + bX
    *    adapted from Natrella, MG: Experimental Statistics. National
         Bureau of Standards Handbook No. 91 (1966)

  **    enter value in Table 4 - IV
                                 -74-

-------
         TABLE 4-VI
WORKSHEET FOR RESPONSE TIMES
(a) Et
Trials
n = 10
1
2
3
h
5
6
1
8
9
10
I

(b) Mean = (a) " =
n
(c) Et2
(d) (Zt)2 _ (a)2 =
n n
(e) = (c) - (d)
(+\ , (e)
(f ' ~ n-_i
(g) Std Dev - /(f ) =
(h) cv = 4|4 x 100 =
\°)
Times to Respond (Units )
*i










Ss










*100









... -
t
r










fc~l










'-95









t-100









1
	 , i . . i

























































fcf


— —


. . I












            -75-

-------
                     TABLE 4-VII


               WOEKSHEET FOR PRECISION
(a)  Input Concentration =
ppm
Trials
n = 10
1
2
3
h
5
6
7
8
9
10

TOTAL
Y
net analyzer
reading












R
(c) times Y












H2












    Mean Analyzer Reading =
                             n
(c) Conversion Factor  =  Ob)
                          (aj


(d)   2R

(e) Mean Response  = (d)
                    (n)
(g)

(h) = (f)  - (g)

(i) - J^LL.
u;    n-1

 5) .Std Dev = /(i)
    CV - (J)    X 100

        "tef

-------
          TABLE 4-VIII

INTERFERENT TEST CONCENTRATIONS
                   Interferent Concentration-ppma
                                                 •.n
                                                     Ozone
Pollutant
CO
CO
N02
K02
N02
°3
03
°3
03
03
S02
Measuring
Principle Airanonia
NDIR
Mercury Replacement
Coloriraetric - azo dye
Amperometric
Chemi luminescent - Gas 0.5
Phase
Amperometric
Colorimetric - KI
Electrochemical
Concentration cell
Chemi luminescent - Gas
Phase
Spectrcmetric
Amperometric
Carbon Hydrogen Nitric Nitrogei
Dioxide Chlorine Sulfide Oxide Dioxide
1000
0.1
0*5 0.1 0.5
0.5 ' 0.1 0.5
0.5
0.5 O.lb 0.5b 0.5
' 0.5 0.5b 0.5
0.5 0.5
0.5

0.5 0.1 0.5
                                                      0.5

                                                      0.5
Sulfur   Water
Dioxide  Vaporf;

          80

          80

  0.5

          80

  0.5     80
                                                               0.5

                                                               0.5

                                                               0,5
          80

          80

          80


          80


          80
                                                      1.0

-------
  TABLE  4-VIII  (Continued)
                                                              Interferent Concentration-ppma
Pollutant
S02
S02
S09
Measuring
Principle
Coloriinetric
Conductimetric
Flame Photometric
Ammonia

0.5
0.5
Carbon
Dioxide Chlorine
0.5
0.5
0.5
Hydrogen
Sulfide
0.1
0.1
0.1
• Nitric Nitrogen
Oxide Dioxide
0.5
0,5
0.5
Sulfur Water
Ozone Dioxide Vaporc
1.0
1.0

      a.  Prepare  concentrations to stated value ± 5%,



      b.  Do not mix with pollutant.



      c.  % RH.
CO
I

-------
                                                               TABLE 4-IX



                                             PPM vs yg/m-3 Conversion Factors For Selected Gases
\Desired
GiveiK
yg/m3
mg/m3
PPM at 25 °C and 760 Torr
M3
1.44 X 10- 3
-
C12
3.45 X 10~4
-
H2S
7.18 X 10~4
-
NO
8.15 X 10"4
-
N02
5.32 X 1(T4
- 	
°3
5.10 X ID"*
	
so2
3.82 X 10~4
_
CO
-
0.874

co2
-
0.556
j
vo
I
N. Desired
\
Given\v
I__A,
PPM
!

yg/m3
1
NH3
696

C12
2900.

•V
H2S
1390


NO
1230 v


N02
1880


°3
1960


S02
2620
i.

mg/m^

CO
1.14
i ,

co2
1.80

           Multiply the given value by the factor under the appropriate gas to obtain the desired unit,.

-------
 5.0    GAS GENERATION









 5.1    PRINCIPLES AM) SCOPE









        To conduct the analyzer performance tests described in Chapter 4




        requires the availability of stable test atmospheres which contain




        known concentrations of pollutant and interferent gases for extended




        periods (days).   In general, a test atmosphere is produced by mixing




        a flow of the desired test gas with flows of clean, dry air (zero air).




        The test gas may be obtained by dilution of a. concentrated gas from




        a permeation device or a cylinder, further dilution of a dilute gas,




        or generated from another medium as in the case of oxygen photolysis to




        produce ozone.  The dilutions are prepared in a gas mixing apparatus




        prior to entering a delivery manifold.  This manifold is designed to




        accommodate simultaneously the sample probe of the analyzer and the




        independent referee method (Figure 5-1).









        This chapter provides instructions for generating the various pol-




        lutant and interferent test gases that are needed.  In using these




        instructions, previous experience and training in gas dynamics and




        trace gas analysis are helpful.









5.1.1   Range









        Ths concentration and volume of the test gases required depend on




        the measuring range and the sampling rates of the analyzer(s) to be
                                     -80-

-------
       tested and the range of the referee method.









       Typical analyzer measuring ranges and sampling rates are summarized




       in Table 5-1.  The concentrations range from 0.01 to 100 ppm and




       sampling rates vary from 0.1 to 5.0 1/min.








5.1.2   Interferences








       Pure pollutant and interferent gases can be purchased with a purity




       of 99% or better.  Selective absorbers can be used  to remove further




       unwanted trace substances.  Descriptions of absorbers that have been




       found useful are given in Table 5-III.  See 5.2.2 for zero air speci-




       fications .








5.1.3   Precision








       The precision (coefficient of variation) with which the test gas




       concentrations can be prepared depends on 1) the nature of the test




      • gas and 2) the precision of the method used to establish the concen-




       tration.  With careful work, a precision between 2  to 5% is achievable




       in the preparation of gases such as CO, S02 and N02«  A portion of



       the error is in the determination of air volumes stemming from the




       use of rotameters.  A minimum>of 2% error is produced for each rotameter




       and the errors are usually additive.  Pulsating flows and pressure




       differences in the gas generation systems produce some flow variation




       which also tends to decrease precision.
                                     -81-

-------
       Another source of error is the precision of the referee procedure;


       there is little information in this regard.  When the referee method


       is a colorimetric procedure, the limiting factors are often in the


       measurement of the sample air volume and the absorbance of the


       colored reagent.
5.2    APPARATUS
       This section specifies the equipment components and materials needed


       for generating and verifying the test gas concentrations.



                                                             *

       The gas generating system consists of a) sources of the pollutant


       and interferent gases (permeation system, cylinder of gas. ozone


       generator, etc.), b) source of zero air, c) a gas mixing apparatus,


       d) a delivery manifold and e) provisions for collecting samples for


       independent analysis.  The individual gas flows - zero air, pollutant


       and interferent - are homogenized and delivered to the test analyzer(s).




       The procedures for generating the test gases are summarized in Table


       5-II along with the referee methods for the verification or establish-


       ment of the test gas concentrations.  Additional information and dis-


       cussion concerning the theoretical and practical aspects of gas gener-

                                                                             2
       ation can be obtained in "Controlled Test Atmospheres" by G. 0. Nelson •




       Any gas generation system should meet the following general specificatior
                                     -82-

-------
1.  Produce the required pollutant and tnterferent gases.








2.  Generate gas concentrations to cover the measuring ranges of the




    analyzers to be tested.








3.  Provide sufficient gas flow to accommodate the sampling




    requirements.








4.  Maintain stable output for the required test duration.








5.  Simulate rapid (seconds) step changes in gas concentration.








6.  Generate multicomponent gas mixtures.








7.  Deliver flows of  zero air and the test gas simultaneously




    through separate  ducts.








8.  Maintain the difference in pressure at the point of delivery to




    ± 0.5 inch 1^0 (0.012 Newton/cm^) of atmospheric.








9.  Utilize ducts, valves, rotameter, mixing chambers etc. with




    surfaces that will not alter the test gas composition.








Mount the components  on a chassis for convenience.
                              -83-

-------
5.2.1   Gas Mixing Apparatus





        The gas mixing apparatus consists of an array of flowmeters, metering


        valves, a glass mixing chamber and delivery manifold as shown in Figure


        5-1.





        *•  Flowmeters;  Calibrated flowmeters are needed to measure a) a flow


            of zero air equal to the sum of the flow required by the test


            analyzers, the referee procedure and 10% excess, and b) the range


            of gas flows required to produce the pollutant and interferent


            test concentrations.  Glass rotameters are commonly used.  Cali-


            brate all rotameters in the system with a primary or secondary


            standard such as a soap bubble meter, frictionless piston, wet-


            test meter, spirometer or dry gas meter.





        2.  Mixing Chamber;  This is a vessel with inlet and outlet ports for


            mixing the individual test gases.  The configurations of some


            typical mixing chambers are shown in Figure 5-II.  Design A is


            the simplest and is usually adequate,  Design B permits the


            addition and mixing of several gases simultaneously.  When


            chambers of other configurations are used, make sure they


            accommodate the required gas volumes and flows and provide

                                2
            homogeneous mixtures .
                                      -84-

-------
       3.  Delivery Manifold:  This consists of a borosilicate glass duct

           with several ports for sampling the test gas mixture simultan-

           eously with the test analyzer (s) and the referee procedure.
                                        i
           The inside diameter of the duct and ports should be suffi-
                                                             (.
           ciently large so that the pressure and flow conditions within

           the analyzer will not be altered by more than 2% during sampling



       4.  Metering Valves:  Fine control needle valves should be stain-

           less steel or Teflon for pollutant and' interf erent gases and

           any material for zero air.



5.2.2   Zero Air Production



       The zero air for diluting the test gases should be free of any sub-

       stances that will in any way a) alter the test analyzer response,

       b) react with the pollutant or the interf erent and c) interfere

       in the referee procedure .  (NOTE :  It is not necessary to remove

       substances that are known not to interfere in the test analyzers

       and referee measurement methods.)  The zero air may be supplied

       from cylinders or by filtering of ambient air.
       1.  Cylinder Zero Mry;  Cylinders should be fitted with a two-stage

           pressure regulator and a metering valve.  For uninterupted operation,

           several cylinders may be connected to the same line.  While it
                                     -85-

-------
    is possible to prepare zero air in .cylinders in the laboratory,




    this requires access to a high pressure (1,000 to 2,000 psig)




    air compressor.









2.  Filtered Zero Air;  For diluent flow requirements ^ 7 1/min,




    zero air can be prepared by passing ambient air through a




    commercial breathing mask filter  designed for removing acid




    gases and organic vapors.  This filter, installed in a special




    holder-*, removes NC^j S02, 0^, most other acid and organic




    gases and vapors and partially removes NO.  It does not re-




    move CO or C0£.








    To remove substances not eliminated by this filter and to




    purify larger air volumes (up to 30 1/min.) the ambient air




    is passed through a series of chemical filters designed to




    remove specific constituents.  The filter materials required,




    the purpose of each and the method of preparation are summa-




    rized in Table 5-III.  To trap particulate matter that might




    be released into the air stream, install a glass fiber mat




    particle filter just prior to discharging the zero air.









3.  Mr Pump (for transport of ambient air) ;  Vane, piston or dia-




    phragm air pumps are used to furnish the diluent air flow re-




    quired for the gas mixing apparatus.  Diaphragm-type pumps are




    recommended since carbon ring and carbon vane-type pumps may
                              -86-

-------
           release particulate matter into the air  stream.  Oil lubricated




           pumps should not be used.  A particle  filter  is necessary for




           vane pumps and is optional with diaphragm types.  Diaphragm and




           piston pumps often deliver pulsating flows.   To minimize or




           eliminate such variations, install an  expansion tank in the




           pump discharge line.









5.2.3   Gas Permeation Apparatus  (Figure 5-III)









       Using permeation devices  as sources for gases require a constant.




       temperature bath, a temperature conditioning coil and a permeation




       device holder.
        I.  Constant Temperature B^.th;   A. bath in the rangf? of 20 fo Tn°C and




           controlled to ± 0.1°C is needed.









        2.  Temperature Conditioning Coil:  A coil of metal tubing is used for




           equilibrating the  temperature pf the carrier gas.  A section of copper




           tube  1/4 to 3/8 in. ID  (0.64 to 0.95 cm) by 25 to 50 ft (76 to 152 m)




           long  is usually adequate.









        3.  Thermometer ; Use a laboratory type thermometer suitable for




           measuring between  20 to 30°C with an accuracy of 0.1°C.









        4.  Permeation Device  Holder: The holder is made from a boro-




           silicate glass screw-cap test tube with added inlet and out-




           let ports (Figure  5-IV) .  All connections must be x^ater-tight .
                                     -87-

-------
       5.  Permeation Device;  This is a source for some of the  test  gases


           listed, in Table 5-II.  The devices may be purchased  *   or

                                     o n
           prepared in the laboratory0'^.
5.2.4  jCompressed Cylinder Gases





       High pressure cylinders containing dilute concentrations of test


       gases such as carbon monoxide (CO), carbon dioxide (CO-), nitric


       oxide (NO), nitrogen dioxide (N02) and sulfur dioxide CSO£) can be


       purchased^!10 or prepared in the laboratory.  The concentrations


       should be at least twice and preferably 10 to 100 times the desired


       test concentration.  The maximum useful cylinder concentration is


       determined by the gas volumes and rates that can be reasonably


       handled by the gas mixing system.





       A cylinder -of gas is prepared in the laboratory by transferring the


       'appropriate volume of the pure test gas to an evacuated cylinder.


       The cylinder is then pressurized with the desired diluent gas.  The


       transfer is made with a closed system vacuum-pressure manifold"  .  Care


       must be taken to avoid contamination of the pure gas with air and water


       vapor or accidental release of the toxic gas into the work environment.





       The losses of some test gases are minimised by using stainless steel


       containers or containers coated with a non-reactive lining such as


       chromium molybdenum, alloy, paraffin or Teflon, or containers with
                                    -88-

-------
      "conditioned" surfaces.  Conditioning involves repeating filling




      and emptying of the container with the test gas until referee analysis"




      shows no further changes in concentration occur with time.  In




      any event, a container previously filled with a particular test




      gas should be refilled only with the same test gas and concentration.




      In an emergency, cylinders known to have contained gases such as air




      or nitrogen may be refilled with a different gas.









      Cylinder gases should be analyzed before each use.  CO concen-




      trations should be verified monthly by independent analysis for




      the first six months after purchase or preparation and about every




      6 months thereafter.  Certification or analysis by the supplier




      at the time of preparation is often of little value because changes




      have frequently been observed.  When cylinder pressures deplete




      to 200 to 500 psig; (138 to 345 Newton/cm2) the concentrations of




      some gases increase due to wall desorption.  The gases should be




      analyzed more frequently or preferably the cylinders should be




      refilled.  After equilibration and confirmation or verification




      of the concentrations by referee analysis, the gases are ready for




      use.  In the case of some gases, notably N02> it may be necessary




      before use to bleed the cylinders at 30 to 50 ml/min for up to




      two hours to condition the delivery lines.









5.2.5  Ozone Generation








      Several commercial ozone generators have recently become available.




      An adequate generator can also be assembled in the laboratory with
                                    -89-

-------
readily available components.  See Figure 5-V and  6.5.1 in Chapter 6


for the list of equipment.  NOTE:  The generator in Figure 5-V uses


a larger lamp and chamber than in Figure 6-II to furnish larger gas


volumes for testing several analyzers simultaneously.




Ozone up to 2 ppia is produced by irradiating a stream of zero  air


to ultra-violet light from a mercury vapor lamp.  A portion of the oxyger


in the air is converted to ozone.  The air may be passed through a


chamber in which the lamp is inserted or the lamp may be located


outside the chamber as in Figure 5-V.  In this latter design,  the


chamber should be of quartz to allow the radiation to pass through.


Depending on the generator design, changes in concentration are obtained


by a) the adjustment of a movable shade over the lamp which controls the


amount of radiation, b) varying the airflow through the  chamber,


c) adjusting the lamp voltage or d) dilution of the ozonized air with


zero air.  In some designs, the distance between the lamp  and  chamber


can be varied.  The concentrations can be further increased by using

oxygen in place of air.




Although ozone can be obtained dissolved in liquid chlorotri-


fluoromethane (Freon 13)3, it is not recommended because of a)  cost,


b) limited durability (days), c) toxicity and d) possibility of
                                     •I n
explosion at temperatures above -20°C  .
                              -90-

-------
5.2.6  Humidity Generator









      The generator consists of a glass vessel (humidifier) containing




      distilled water and a moisture sensor.  Humidified air is produced




      by passing zero air over the distilled water in the humidifier main-




      tained at a fixed temperature between 20 to 30°C so that a




      relative humidity between 70 to 80% is produced.









       1.  Humidifier;  This can be a 250 or 500 ml Erlenmeyer or




           equivalent flask with an inlet tube extending to within




           1 to 2 cm above the water level.









       2.  Moisture Sensor:  A commercially available moisture sensor such




           as a LiCl salt type sensor, dew-point detector or continuous




           wet and dry bulb recorder may be used.









5.2.7  Sampling and Analysis








       The referee procedure generally is a manual method and requires




       a sampling train consisting of sampling probes, absorbers, metering




      valves, flowmeters, vacuum pumps and traps as shown in Figures 6-III,




       7-II and 8-II.








      1.  Sampling Probes:  All sampling lines for the referee procedure




           should be of glass or Teflon, and should be as short as possible




           to prevent transit losses.  Short sections of polyvinylchlori.de




           tubing (like Tygon) may be used for-butt-joints only.
                                    -91-

-------
2.  Absorbers: All-glass bubblers or midget impingers are described




    in the respective referee procedures.  The absorbers may be




    purchased from major glassware suppliers.  For some tests, two




    absorbers in series may be needed to insure complete collection




    of the air sample.  Standard ball-joints are commonly used to




    facilitate on-off connections.









3.  Air-Vacuum Pump;  This is needed to draw up to 2.0 iiters/min




    through the absorbers.  The pump should be equipped with a




    trap and a needle valve at the inlet side to regulate flow.









4.  Thermometer;  An accuracy of ± 2°C is needed to correct the sampled




    gas volumes to standard conditions (25°C, 760 torr).









5.  Metering Valve: A fine control needle valve is used to regulate




    sample flow.









6*  Barometer:  This is used to correct the sample gas volumes to




    standard conditions and should be accurate to the nearest torr.









7.  Flowmeter;  A calibrated rotameter to measure gas flows between




    0.5 to 3 liters/min within ± 2%.









$•  Trap:  A vessel which contains glass wool to protect the needle




    valve and the vacuum pump from moisture.
                              -92-

-------
      9«  Sgectroghotometer;  This is used to measure the absorbance of




          solutions between 350 to 600 im.  The references listed in Table




          5-II cite bandpass requirements.









     10*  Other;  Chemicals, reagents and other apparatus needed to perform




          the referee analysis are specified in the references listed in




          Table 5-II.









5.3   OPERATION OF THE SYSTEM









      The test gases and  the concentrations needed and the instructions for




      generating them are described under the respective tests in Chapter 4.




      Information regarding the criteria of components for the gas generating




      systems and additional details concerning other methods and equipment




      for generating test atmospheres are found in Reference 2.
                                    -93-

-------
5.4    REFERENCES
       1.  Thomas MA., and Amtower RE:  Gas dilution apparatus for




           reproducible dynamic gas mixtures in any desired concentration




           and complexity.  J Air Poll Coiit Assn, 16-618, 1966









       2.  Nelson GO:  Controlled Test Atmospheres.  Ann Arbor Science




           Publisher, Inc. (1971)








       3.  Matheson Gas Products, 61 Grove Street, Gloucester, MA  01930









       4.  Filter Cartridge, (CMC fill), Cat. No. DZ-78006,-Mine Safety Applian




           201 North Braddock Avenue, Pittsburgh, PA  15028









       5.  Filter Holder, End of Line, Cat. No. DZ-78006, Ibid








       6.  Analytical Instrument Development, Inc., 250 South Franklin Street,




           West Chester, PA  19380









       7.  Metronics Associates, Inc., 3201 Porter Drive, Stanford Industrial P




           Palo Alto, CA  94304









       8.  O'Keeffe AE , Ortman GC:  Primary standards for trace gas




           analysis.  Anal Chem 38, 760 (1966)
                                     -94-

-------
 9.   Scaringelli FP,  O'Keeffe AE,  Rosenberg E,  Bell JP:




     Preparation, of known, concentrations of gases and vapors with




     permeation devices calibrated gravimetrically.  Anal Chem 42,




     871 (1970)









10.   Scott Research Laboratories,  Inc., Plumsteadville,  PA









11.   Belsky T:  Preparation of low-concentration mixtures of gases,




     Air and Industrial Hygiene Lab, Cal State  Dept of Publ Hlth,




     Berkeley, CA  94704, AIHL Report No. 117 (Jan. 1972)









12.   Braker W, Mossiaan AL:  Effects of exposure to toxic gases.




     Matheson Gas Products, Div Will Ross Co, NJ (July 20, 1970)
                              -95-

-------
                                                     TABLE 5-31

                             MEASURING RANGES AND SAMPLING RATES OF COMMON AIR ANALYZERS
  POLLUTANT
     CO
     0-.
v
     N0/N02
     so2
MEASUREMENT
 PRINCIPLE
NDIR
Hg replacement

Atnperometrlc
Chemilumines cent
Colorimetric
Electrochemical Cell
IN Absorption

Amperometrie
Chemiluminescent
Colorimetric

Amperometric
Colorimetric
Conductimetric
Flame Photometric
Electrochemical Transducer
  MEASURING
    RANGE
(Upper Limit)
  25 to 100
  5 to 50

  0.1 to 2
  0.02 to 2
  0.5 to 5
  0.025 to 1
  0.5 to 2

  0.1 to 2
  0.1 to 2
  0.1 to 2

  0.1 to 2
  0.25 to 5
  0.5 to 2
  0.01 to 1
  1 to 5
SAMPLING
  RATE

 1/min

0.5 to 2.0
8 to 41a

0.15 to 4.0
1.0 to 2.0
0.25 to 4.0
0.10
3.0

0.15 to 3.5

0.2 to 3.5

0.15 to 3.5
0.4 to 1.0
0.1 to 5
0.15
0.2 to 0.5
  a Rate of interferent prefilter section.  Detector section samples at 1 to 2 liter/mln.

-------
                           METHODS FOR GENERATING AND DETERMINING GAS CONCENTRATIONS
   TEST GAS
   1.   Ammonia
   2.  Carbon Dioxide
   3.  Carbon Monoxide
!  4.  Chlorine
VO
•-j
    5.  lydrogen Sulfide
    6.   Nitric Oxide
    7.   Nitrogen Dioxide
GENERATION

Permeation tube and procedure
described for S02 in footnote 3.
A cylinder of zero air containing
10,000 ppm C02.  Dilute with zero
air to concentration specified in
Table 4-VIII.

A cylinder of zero air containing
CO concentrations 10 to 100
times those required in the test
procedure.  Dilute with zero air
to concentration needed.

Permeation tube and procedure
described for S02 in footnote 3.
 Permeation  tube, and procedure
 similar  to  that described for
 S02  in footnote 3.

 Cylinder of nitrogen containing
 20 ppm NO and  less than 4 ppm
 02.   Dilute with  zero  air to
 desired  concentration.

 Cylinder of nitrogen containing
 20 ppm N02. Dilute with zero
 air  to desired concentration.
 (NOTE: To condition regulator
 and  delivery lines, discharge
 the  N02  at  a low  (~ 30 ml/min)
 rate until  the concentration is
 stable  (usually within one  hour).
REFEREE METHOD

Tentative Method of Analysis for Ammonia
in the Atmosphere (Indophenol Method).
42604-01-72T, footnote 1.

Certified analysis by manufacturer, vendor or
independent laboratory.
Certified analysis by manufacturer, vendor  or
independent laboratory.
Tentative Method of Analysis for Free Chlorine
Content of the Atmosphere (Methyl Orange Method).
42215-01-70T, pg 282, footnote 2.

Tentative Method of Analysis for Hydrdgen Sulfide
Content of the Atmosphere, 42402-01-70T, pg  426,
footnote 2.

Tentative Method of Analysis for Nitric Oxide
Content of the Atmosphere (Greiss-Saltzman
Reaction), 42601-01-7IT, pg 325, footnote 2.
Tentative Method of Analysis for Nitrogen Dioxide
Content of the Atmosphere (Greiss-Saltzman Reaction)!
42602-01-68T, pg 329, footnote 2.

-------
                                                TABLE 5-II  (Cont'd)
 TEST GAS

 7.  Nitrogen Dioxide
         (Cont'd)
 8.  Ozone
GENERATION
Permeation tube and procedure
described for S0£ in footnote 3.
Air in contact with permeation
tube must have a RH <5%.  Tube
should be weighed weekly.

Ozone generator - procedure des-
cribed in footnote 3.
                                                                        REFEREE METHOD
                                                                        Ibid (preferred) or emission rate certified by
                                                                        manufacturer, or by independent laboratory.
                                                                        Tentative Method of Analysis for Oxidizing Sub-
                                                                        stances in the Atmosphere, 44101-01-70T, pg 351,
                                                                        footnote 2.
 9.   Sulfur Dioxide
    10.  Water
CO
I
Permeation tube and procedure
described in reference method for
     footnote 3.
                             Pass zero air over water at a fixed
                             temperature between 20 to 30°C to
                             obtain 80 to 90% RH.  Dilute with
                             zero air to desired RH.
                                                                        Pararosaniline procedure described in
                                                                        footnote 3.
                                        Dewpoint detector, calibrated hygrometer or a
                                        wet/dry bulb thermometer.
11.   Zero Air
Air containing 21.0 ± 0.5% oxygen and
no, constituents that will 1) react
with the pollutant or interferent gas
2) alter the test analyzer response
and 3) Interfere in the referee pro-
cedure.
                                                                        Cylinder zero air: certified analysis by
                                                                        manufacturer or vendor.  Filtered  zero air:
                                                                        compare response of referee method and test
                                                                        analyzer(s) with response  obtained with
                                                                        certified cylinder zero air.
     1   Health Lab Sciences, 10:2, 115-118, Apr 1973.

     2   Manual of Methods of Ambient Air Sampling and Analysis.  Intersociety Committee, 1972.  Amer.
         Public Health Assn., 1015 - 18th Street N.W., Washington, DC.

     3   Part 50 - National Primary and Secondary Ambient Air Quality Standards, Federal Register, Nov. 25,
         1972 (26 FR 22384)

-------
                                                       TABU.  ,'5-IXI

                                     ABSORBERS FOR PRODUCING  ZERO AIR UP TO  30  1/mi.n
IATERIAL
PURPOSE
   Activated  charcoal
   (6 x 16 mesh)
Removes many gases such as ozone,
N02, higher molecular weight organic
vapors.  Does not remove CO, CC.
   NO Oxidizer .

    (a)  Chromium  trioxide
Converts NO to N0£ for subsequent
removal with TEA or soda lime^.
     (b) Humidifier
 Furnishes water vapor for proper
 operation of oxidizer.
PREPARATION

Commercially activated coconut hull charcoal
6 ic 16 meshX  Place in 2 inch I.D.  x 18  inc
long section of plastic pipe.  Use  glass
wool plugs to retain charcoal in  place.
Plastic pipe caps on each end of  cylinder
are drilled and threaded to accept  standard
1/2 inch O.D. tube fittings.  Containers
of other materials and similar configuration
and volume may be used.
Soak firebrick or alumina 15-40 mesh in a
solution containing 16 g Cr03 in 100 ml of
water.  Drain, dry in an oven at 1U5-115°C
for 30 to 60 minutes and cool.  Spread a thin
layer of the dry pellets in a dish and
place in a desiccator containing a saturated
salt solution which maintains the relative
humidity between 50 to 70%.   The reddish
color changes to a golden orange when equi-
librated.  Place in a 1/2 inch I.D. x 15 inch
glass tube.  Use glass wool  plugs to hold
pellets in place.  (Caution:  Protect eyes and
skin when handling this material.  Do not
breathe oxidizer dust.)  Before using,  pass
air containing 30 to 70% RH  through the
oxidizer for about 1 hour at 0.5 £/min to
condition.  Discard when more than
3/4 of the oxidizer bed depth turns brown.

Pass dilute NO air stream over water at a
fixed temperature such that  the humidity of
the air stream is maintained within 50 ±
20% R.H.

-------
                                                  TABLE 5-HI (Cont'd)
MATERIAL

3.  Triethanolamine  (TEA)
PURPOSE
Removes NO,
4.  Desiccant
Remove water
    Carbon Monoxide
    Oxidizer  (Hopcallte)
Catalytically oxidize CO to C02 for
subsequent removal with Ascarite or
soda lime.
6.  C02 Absorber
Remove C02, H20
PREPARATION

Soak firebrick 10-20 mesh in a 20% aqueous
solution of TEA.  Drain, spread on a dish.
and dry for 30 to 60 min at
95°C.  Pellets should be free-floxd.ng.
Place in a 1/2 inch O.D. (7/16 inch X.D.)
•x 15 inch long polyethylene or stainless
steel tube with standard 1/2 inch O.D,
tube fittings on each end.  Use glass
wool plugs to hold the pellets in place.

Commercial 6 x 16 mesh silica gel with
color indicator".  Place in 3 inch I.D.
x 24 inch long clear plastic cylinder
capped at both ends.  Caps are drilled
and threaded to accommodate standard 1/2
inch O.B. tube fittings.  Use glass wool
to hold granules in place.  When the. change
in color exceeds 3/4 of the desiccant bed
depth, regenerate by exposing the silica  gel
to 120°C atmosphere overnight.

Commercial mixture  of  copper and manganese
oxides-^.  Place  granules  in a  1/2 inch  O.D.
 (7/16  inch I.D.) x  15  inch long section of
copper pipe  with standard  1/2  inch  O.D. tube
fittings  on  each end.   Use glass wool to
hold  granules  in place.
 a.
                                                                            b.
Soda Lime^:  commercial mixture of
calcium and sodium hydroxides; 4  to
8 mesh.

Ascarite^ :  a commercial preparation
of sodium hydroxide in an asbestos
matrix, 8-30 mesh.

-------
                                                 TABLE 5-IIZ (Cont'd)


ATERIAL                        PURPOSE                                     PREPARATION

                                                                           Place soda lime or ascarlte in a 2 Inch
                                                                           I.D. x 15 inch long section of plastic pipe.
                                                                           Plastic caps at both ends are drilled and
                                                                           threaded to accept standard 1/2 inch O.D.
                                                                           tube fittings.


   1    Activated coconut hull charcoal, Type PCB 6 x 16 mesh, Pittsburgh Activated Charcoal, Merk & Co., Inc.
        Pittsburgh, PA 15230

   2    Silica gel, indicating, 6 x 16 mesh, Grace Davison Chemical, Baltimore, MD 21226

   3    Carbon monoxide purifier, (Hopcalite) Model RAF-BCHDI, Robbins Aviation, Inc., Vernon, CA 90058

   4    Soda  lime, 4-8 mesh, J.T. Baker Chemical Co., Phillipsburg, NJ 08865

   5    Ascarite, 8-30 mesh, ibid.

-------
I
M
O

I
                      Zero Gas Line
      Pollutant/Interferent Gas Line
   O
                    Vent«~O
O
Flow
Meter
Zero Air   Pollutant   Interferent
                                       &
               Mixing
                Cnamber
                                      Vent
                                               To Test

                                               Analyzers
                                               To Referee
                                               Method
                                Vent
        Figure  5—1.  Diagram of gas mixing apparatus.

-------
                                                           Ill
o
U)
I


                                                                 B
                             Figure 5-II.  Configuration of gas mixing chambers.

-------
o
-p-
I
                                             Thermometer


                                                      fl
                                                          Stirrer
                         Flowmeter
                            Control

                            valve
                                                      Water bai:b
                                                                                          Permeation device
                                                                                          Device holder

                                                                                          (glass)
                                  Zero air
                                                                    Temperature conditioning coil (copper)
                                              Figure 5-III.  Gas permeation apparatus.

-------
I
J_a
o
25mm
          Outlet
           Port
                               200mm















f
^
Inlet
Port
1

Screw Cap Test Tube
                                                   or  Culture  Tube*
                                                                 Plastic Screw Cap
                                                                 (w/Teflon Link)
                                                           Threaded to Fit Cap
    * Kimble Glass Co., Div. Owens-Illinois, Toledo, Ohio.
       Cat. No. 45066.
                  Figure 5-IV.  Permeation device holder.

-------
                                                              Aluminum box enclosure
!
M
O
I
              Hi-voltage
             power supply
        Clean air       ~~
        or oxygen
                                               Adjustable  sleeve
                                          9-in. UV lamp        ^
                                 £-tezz2
                                                                   -1
                                                                    I
                                                                   .J
                                            Quartz  tube  35mm O.D.
                                                                                           Collar
A
                                                 Figure  5-V.   Ozone  Generator

-------
6.0    CALIBRATION PROCEDURE FOR AUTOMATED ATMOSPHERIC
        OX3DANT AND OZONE ANALYZERS
6.1    PRINCIPLE AND SCOPE



6.1.1  This procedure is for the calibration of automated atmospheric oxidants

       and ozone analyzers.  The calibration -may be of two types, dynamic and

       static.  The dynamic calibration must always be performed by determining

       the analyzer response to a series of ozone (03) concentrations.  The

       procedure is applicable to all oxidant and ozone analyzers and is a

       performance test of the entire analyzer under simulated service

       conditions.  The static calibration is performed by determining the

       analyzer response to artificial stimuli such as standard calibrating

       solutions, optical filters, screens, electrical signals, resistors,

       etc.  This calibration is a test of the detection and signal presen-

       tation components only and is not a substitute for the dynamic cali-

     •  bration.  It is primarily applicable to analyzers using wet-chemistry

       and colorimetry.



6.1.2  Atmospheric oxidants are those airborne compounds that will oxidize

       iodide ions in solution to iodine.  The predominant oxidant is ozone.

       Other oxidants are:  nitrogen dioxide, halogens, peroxy compounds,

       hydroperoxides, organic nitrites, peroxy nitrates, and hydrogen

       peroxide.  Sulfur dioxide reduces the iodine in solution.  Oxidant

       data should be corrected for the effect of nitrogen dioxide and sulfur

       dioxide.
                                    -107-

-------
6.1.3  The ozone concentrations for the dynamic calibration are  generated


       by irradiating a stream of purified air (or oxygen) with  ultraviolet


       light.  Each ozone stream is sampled simultaneously with  the instru-


       ment and with the manual referee method (neutral potassium iodide1)


       to establish the concentration.





6.1.4  Static calibration is accomplished by flowing reagent solutions


       containing amounts of iodine (12) equivalent to concentrations of


       ozone through the analyzer's detector at the reagent flox-7 conditions


       encountered during normal operation.  The analyzer responses are


       plotted versus equivalent ozone concentrations to obtain  the static


       calibration curve.  The instrument variables, e.g., air and liquid
                                           I

       flow rates, may be adjusted after static calibration to make the


       output response conform to the pollutant concentration or a simple


       multiple or fraction of the concentration in parts per million (ppra)


       or micrograms per cubic meter (pg/m^) (spanning).  When a static


       calibration is not performed, the spanning may be done during the


       dynamic calibration.





6.2    RANGE





       The range of the calibration procedure is determined by that of the


       manual referee method.  For a 100 liter sample collected  in 10 ml


       of absorbing reagent and measured in a 0.5 inch cuvette,  the range is


       between 0.02 to 1.0 ppm 03 (40 to 2,000 yg 03/m3).  The lower limit
                                      -108-

-------
      can be extended down  to about 0.01 ppm (20 yg/m3)  by using a




      one inch path  cuvette and a 30 minute sampling period.   At con-




      centrations below 0.05 ppm (100 yg/m3) the manual  method may be




      less accurate  because of the losses of iodine on the glass surface .




      The upper  limit can be extended up to 10 ppm 03 (20,000 yg 03/m3)  by




      using 20 ml of absorbing reagent and decreasing the  sampling rate




      and collection period.









6.3   INTERFERENCES









6.3.1 The zero air used for ozone generation must be free  of  reducing and




      oxidizing  substances.  Commercial zero air in cylinders is usually




      free of such interferents.  Chemical absorbers are available that




      will remove most  interfering compounds when ambient  air is used




      for zero air.  Details for preparing these absorbers are given  in




      Table 5-III.









6.3,2 Glassware  should  be cleaned with chromic acid solution  since dust




      or foreign matter may interfere.  Rinse thoroughly with distilled




      water.  Vessels to contain a dilute iodine solution  should be rinsed




      just before use with  a portion of the solution to  avoid losses  of




      Iodine  .








6.4   PRECISION, ACCURACY AND STABILITY









6.4.1 The coefficient of variation of the manual referee method  within  the




      range of 0 to  2 ppm is about ± 5%.  The -major error  is  from the loss
                                    -109-

-------
       of iodine.  This is reduced by using a second impinger in tandem.





6.4.2  When spanning is possible, any discrepancy between the input and


       output may be resolved by adjusting the output of the instrument


       to correspond with the value of the calibrating gas or solution.


       Where analyzers have no spanning controls, a correction factor


       nay be calcxilated to convert the analyzer readings to ozone con-


       centrations .





       The calibration is based on the 1:1 stoichiometry of the reaction of

                                      Q / t
       ozone in the absorbing solution^' > .





               03 + 2KI + H20 •> 12 + 2KOH 4- 02             (1)





6.4.3  The absorbance of the reacted KI solutions in the referee procedure


       should be read between 15 to 30 minutes after sampling to insure


       complete I2 formation and yet avoid fading of iodine color, especially


       at low absorbance levels.





6.5    APPARATUS





6.5.1  Ozone Generator (Figure 6-1 and 6-II)  The generator should be capable


       of producing ozone concentrations in the range of 0.01 to 1.0 ppm


       (20 to 200 yg/m3) at a flow rate of at least 5 1/mln.





       Commercial ozone generators have recently become available.  The ozone


       concentration produced varies inversely with the airflow rate, and
                                     -110-

-------
directly with the exposed area of the ultraviolet lamp.









An adequate generator consists of:









1.  Ultraviolet Source - Low-pressure mercury lamp such as Ultraviolet




    Products, Inc., San Gabriel, California, model PCQ9G-1 or equivalent




    (available in lengths from 6.3 to 20 cm or 2-1/2 to 8 inches), with




    matching power supply (ballast transformer).









2.  Lamp Control - The lamp ballast should be connected to a source




    of constant AC voltage to provide a constant lamp output.









3.  Lamp Housing - Figure 6-II shows a recommended system for exposing




    the air stream to the ultraviolet lamp.









4.  Pump (for transport of ambient air) - capable of furnishing the




    total airflow requirements of the calibration system plus 10%.




    Diaphragm-type pumps are recommended since pumps with carbon




    vanes and rings may release particulate matter into the air




    stream.  A particle filter (glass fiber mat) should be installed




    downstream of carbon vane pumps and is optional with diaphragm




    pumps.








5.  Activated Charcoal Trap - For filtering cylinder air.




    See Table 5-III, Item 1.
                              -Ill-

-------
                     - F°r measuring airflow past the lamp; meter should be
           capable of measuring flows from 1 to 10 liters/min within ± 5%.




           Rotameters are most commonly used.  They should be calibrated




           frequently (at least monthly) with an appropriate flow measuring




           device such as a wet test meter, dry gas meter, soap bubble




           meter or calibrated rotameter.









       7.  Sampling Manifold - The manifold is fabricated from borosilicate




           glass tube with sampling ports to distribute the calibrating




           gas simultaneously to the analyzer (s)  and the manual sampling




           train.  See Figure 6-1.









6.5.2  pH meter - Capable of measuring with a precision of ± 0.02 pH units.









6.5.3  Sampling Probes - The gas lines for conducting the ozone calibrating




       mixtures to the analyzer should be of Teflon or glass, and they should




       be as short as possible to minimize wall losses.  Short sections of




       polyvinylchloride tubing may be used to butt join sections of inert




       tubing .









6.5.4  Manual Sampling Train:









       Assemble the ozone sampling train as shown in Figure 6-III.  Use




       ground-glass joints upstream from the impingers.  Other connections




       may be butt-joined with polyvinylchloride tubing.
                                     -112-

-------
      1*  Absorber - All-glass midget  impingers  as  shown  in Figure  6-III




          are recommended for collecting  the  samples  for  referee analysis.




          See Reference 1 for specifications.  The  impingers may be




          purchased from glassware  suppliers.  Two  absorbers in series




          are needed to insure complete collection  of the sample.  Ball-




          joint connections are recommended for  convenience.









      2.  Air-vacuum Pump - Capable of drawing 1 liter/min through the




          absorbers.  The pump should  be  equipped with a needle valve and




          trap at the inlet side to regulate  flow.









      3.  Flowmeter - A calibrated  metering device  to measure gas flows of




          0.5 to 2 liters/min within ± 2%.  See  Item  6 in 6.5.1 for




          calibration requirements.









      4.  Trap - A vessel containing glass wool  to  protect needle valve




          from moisture.









6.5.5  Miscellaneous









      1.  Spectrophotometer - Equipped with a bandpass filter of 20 run




          or less for measuring the absorbance at 352 nm.  Matched 1-cm,




          0.5 or 1.0 inch cuvettes  should be used.









      2'  Timer - 0 to 30.0 minutes  -  for measuring the sampling period




          of the referee sample.
                                    -113-

-------
       3*  Thermometer - With an accuracy of ± 2°C to correct sample volumes



           to standard conditions (25°C, 760 torr).






       4.  Barometer - Accurate to the nearest torr to correct sample volumes



           to standard conditions.






6.5.6  Static Calibration






       I-  Volumetric Flasks - 25, 100, 250, 500, 1000 ml - low actinic



           glass.
                                        x





       2.  Buret - 50 ml.






       3.  Pipets - 0.5, 1, 2, 3, 4, 10, 25, and 50 ml volumetric.






       4.  Separatory Funnel - 125 ml, low actinic glass.






       5.  Iodine Flasks - 300 ml., (3).






       6.  Oven - maintained at 105°C.






  6.6  REAGENTS AND GASES






       Purity of chemicals - Unless otherwise indicated, all reagent specifi-



       cations shall conform to the Committee on Analytical Reagents of the
                                     -114-

-------
      American Chemical  Society6.   When such reagents are not available,

      ascertain that  they do  not lessen the accuracy of the determination.
6.6.1  Potassium  Iodide  (KI)  - Reagent grade KI often contains a reducing

      agent  to prevent  the oxidation of iodide to iodine on standing.   KI

      meeting "USP  specifications usually contains little or no reducing

      agent.  While this agent keeps the KI from turning yellow,  it

      interferes significantly with oxidant and ozone measurement and

      calibration.   Every lot of KI purchased should be tested before

      use.   To test, prepare a 10% solution as directed in 6.9.1, Step 2,

      in a clear 100 ml volumetric flask.  Make sure the pH is between 6.5

      to 7.0.  Add 0.10 ml of iodine working solution (Item 6.6.11) and mix

      well.   Observe the iodine color visually or transfer 10 ml to a

      cuvette and read  at 352 ma in the spectrophotometer.  When reducing

      agent  is present, the absorbance will fade within minutes.   Add  ^ 10%

      hydrogen peroxide (H202) solution dropwise (with a graduated pipet

      or buret)  to the  10% KI until the loss of absorbance stops.  To  check

      the end-point, again add 0.10 ml of the I2 solution (Item 6.6.11).

      The color  should  not fade.  Based on the volume V]_ of the H202 required,

      determine  the volume V2 of the H202 needed to neutralize the reductant

      in 1 liter of 50% KI (Item 6.6.9).
                            .    (50%) (1000 ml)           (2)
                          = CV1' (10%)  (100 ml)
       Add  V2 ml  H202 to each liter of the 50% KI 6.6.9 prepared.  Verify

       that the  proper amount of H202 has been added to the batch.  A yellow
                                     -115-

-------
       color Indicates too much H202 was added and a. new batch should be




       prepared.  When the solution is colorless, add 0.1 ml of the I2




       working solution (Item 6.6.11) forming a slight yellow color that




       should not fade.  When it does, determine the additional amount of




       H202 required.  Other KI solutions (1%) should be tested and treated




       in a similar manner.








6.6.2  Distilled Water - Double-distilled water should be used for all




       reagents.  It can be prepared in an all-glass still.  Before the




       second distillation, add potassium permanganate to the water to




       produce a faint pink color and one crystal of barium hydroxide per




       liter to make the water alkaline,








6.6.3  Absorbing Solution  (1 % KI in 0.1 M phosphate buffer) - Dissolve 13.6 g




       of potassium dihydrogen phosphate (KH2PO^), 14.2 g of disodium hydrogen




       phosphate (l^HPO^) or 35.8 g of the dodecahydrate salt (Na2HP04'12 B^O)




       and 10.0 g of KI in sequence in approximately 750 ml distilled water,




       and dilute to 1 liter.  Keep at room temperature for at least one day




       before use.  Measure the pH and adjust to 6.8 ± 0.2 with NaOH pellets




       or 10% H-^PO^.  This solution can be stored for several months in a




       glass-stoppered brown bottle at room temperature without deterioration.




       It should not be exposed to direct sunlight.









6.6.4  Standard Arsenious  Oxide As20-^ Solution  (0.05 N) _ T)ry reaeent  grade




       As203 at 105° to 115°C for one hour immediately before using.   Cool




       to room temperature in a desiccator.  Accurately weigh 2.473 g  in  a

-------
      small glass-stoppered weighing  bottle.   Dissolve in 25  ml 1 N NaOH

      in a flask or beaker on  a  steam bath.   Add 25  ml 1 N ^SO, .   Cool,

      transfer quantitatively  to a 1000 ml volumetric  flask and dilute  to

      volume.  The solution should be neutral to litmus.
                                    (wt.  AsoOo)
              Kor-ality  °
6.6.5  Starch Indicator  Soulution (0.2%)  - Triturate  0.4  g  soluble starch

      and approximately 2  mg mercuric  iodide (preservative) with a few ml of

      water.  Add  the paste slowly to  200 ml boiling water.   Continue boiling

      until the  solution is clear, allow to cool,  and  transfer  to a glass-

      stoppered  bottle,  Alternatively,  a commercial indicator  (Thyodene3)

      can be used .
6.6.6  Stock Iodine Solution (0.05 N
       1.   Preparation - Successively dissolve 16  g KI  and  3.173 g resublimed

           I2 in 25 ml distilled water.   When the  I2 dissolves, transfer  the

           solution to a 500 ml glass-stoppered volumetric  flask.  Dilute to

           volume with distilled water and mix thoroughly.   Keep at room

           temperature at least one day before use)  Store  the  solution in

           a dark brown, glass-stoppered bottle away from light.  Re-

           standardize before use.
       ^Indicator,  Thyodene,  Cat.  No.  T-138,  Fisher Scientific Company
                                     -117-

-------
       2.  Standardization - Accurately pipet 20.0 ml of standard As203



           solution (Item 6.6.4) into a 300 ml iodine flask and acidify with



           one drop of 10% l^SO^ solution.  Add about 2 g solid sodium



           bicarbonate (NaHCX^) , and 5 ml starch solution (or 0.1 to 0.5



           gm thyodene)  as indicator.  Titrate with the stock I2 solution



           adding 1 ml of 10% I^SO^ near the end-point to saturate the



           solution with carbon dioxide.  Continue the titration until



           a pale blue color is obtained that persists for 30 sec.  Cal-



           culate the normality of the stock I2 by equation 4.
                                  ml AsoOo X normality AsoOo   , ,.
                   Normality 1.2 = - ^-3 - - - ^-^   (4)

                                          ml I2
           The stock I2 solution may also be standardized with 0.025 N



           sodium thiosulfate (Na2S2Oo).   The ^32820^ solution is in turn


                                                                           7
           standardized with primary standard potassium biiodate (KH(IOo)?)



           or potassium dichromate (K^Cr^y) .







6.6.7  Zero Air - This may be obtained from cylinders of high-pressure,



       compressed synthetic air, or by filtering ambient air.  The air



       for generating and diluting the test gas should be free of ozone



       and any substances that will in any v?ay a) change the test gas



       concentration, b) interfere in the analyzer response or c) interfere



       in the referee procedure.  An activated carbon filter will remove



       most interferents, but when nitric oxide or sulfur dioxide is present
                                    -118-

-------
       in the ambient air, a Cr03  oxidizer followed by a soda lime or tri-


       ethanolamine filter is necessary  to remove them.   See Chapter 5,  Gas


       Generation, for details.





 6.6.8  0.5 M Buffer Solution (for  static calibration)  -  Dissolve 71 g (0.5


       mole) anhydrous Na2HP04  or  179  g  Na2HP04'12 H20,  and 68 g KH2P04,  in


       approximately 750 ml distilled  water.   Make up  to almost one liter.


       Measure the pH.  When different from pH 6.8 to  7.0,  adjust with


       NaOH or KOH pellets, granular KH2P04 or 10% H3P04>   Dilute to one


       liter.  Store in an amber bottle  and out of direct sunlight.   Discard


       after two years.  When exposure to sunlight is  unavoidable,  add


       0.025 gm of a non-reducing  preserving  agent, e.g., sodium o-phenyl-


       phenate, sodium pentachlorophenate, Trichlorophenate (Dowicide B).





 6.6.9  50% Alkaline KI Solution (for static calibration)_ -  Dissolve 500 g


       of KI (See 6.6.1) in. approximately 750 ml of water.   Add a sufficient


       amount of concentrated NaOH solution or NaOH pellets to obtain a pH


       9 to 11 and make up to one  liter.  Mix well and store in an amber


       bottle out of direct sunlight.  Discard after two years.





6.6.10  Iodine Dilute Solution (0.004N  I2)  - Transfer 40.00  ml of the 0.05  N


       I2 stock solution (Item  6.6.6)  to a 500 ml low-actinic volumetric


       flask with a volumetric  pipette.   Dilute to volume with distilled water,


       Mix well and store out of direct  sunlight.   Discard  after two days.


       Calculate the normality  of  the  I2 by equation 5:
                                    (40.00 ml)  (N  of  stock  12)
                  N of dilute I2  =	—	—      (5)
                                           (500 nil)
                                      -119-

-------
6.6.11  Iodine Working Solution (0.0008N I2:  equal  to 10  yl  03/ml  of solution)




        Calculate the volume V^ of dilute I2  (Item 6.6.10) necessary to make  •




        250 ml of iodine working solution by  equation 6:







                            0.204                           (6)
                     a   N of Dilute I2









        where 0.204 is a combined constant (See Basis for calculations, 6.11).









        Transfer the volume of 0.002 N I2 solution calculated by equation 6




        to a 250 ml low-actinic volumetric flask and dilute to volume with




        distilled water.  Mix well and store out of direct sunlight.




        Prepare daily.
        1.  Prepare 'a series of calibrating solutions between 0.1 to 0.7




            Hi 03/rnl by diluting aliquots (i.e., 1, 2, 3 ml) of the I2




            working solution (Item 6.6.11) to 100 ml with the 1% KI absorbing




            solution (Item 6.6.3).  Obtain the spectrophotometer readings at




            352 ran.  Plot the absorbances versus equivalent yl Oo per ml




            on rectilinear graph paper as a check for linearity.  Calculate




            the slope b of the best-fit curve for the data using the method




            of least squares .









        2.  When a static calibration is to be performed on a colorimetric




            Instrument using iodide absorbing solution, prepare a calibration
             Table 4-V in Chapter 4.




                                      -120-

-------
          curve  for  the  spectrophotometer as in step 1 above bat using




          the same reagent  formulation as used in the analyzer.   This




          curve  will be  used  when performing a secondary check on the




          accuracy of the calibrating solutions during static cali-




          bration  (6.9) .









6.8   DYNAMIC CALIBRATION









6.8.1 General









      1.  When a static  calibration (6.9),  is to be performed on the




          analyzer,  it should be done before the dynamic calibration




          (6.8)  to assure proper operation of its detection  and  signal




          presentation components.









      2.  When the analyzer to be calibrated has been operating  as  a




          continuous monitor, it is useful to determine the  response




          near the span  level first without changing the span setting




          (auditing). When the analyzer  response is within  10%  of




          that obtained  in  the previous calibration, the calibration




          is still valid and  a new calibration is not needed.  When




          the response is greater than 10% from the previous calibration




          proceed with the  complete calibration.  The audit  data provide




          a record of the calibration drift.  Instruments with non-linear




          response require  the full calibration.
                                    -121-

-------
       3.  The static and dynamic responses may be compared  (reconciled)




           to verify the proper operation of the analyzer.   See  6.10




           for details.
6.8.2  Procedure
       1.  The analyzer should be in good operating condition and in-




           stalled in accordance with manufacturer's instructions.




           Operate the analyzer for at least 24 hours to warm-up.  This




           24 hour warm-up time may be shortened when so stated in the




           operating instructions.  Adjust air and reagent flows to their




           recommended rates or the rates determined from the static




           calibration data and verify the flowrates as described in




           6.5, Item 6.  Record data only after stable response has




           been attained.  Refer to 4.1.6 in Chapter 4 for determining




           stable responses.









       2.  Place the ozone generation app'aratus (Item 6.5.1) as close as




           practical to the analyzer to prevent losses and to minimize




           pressure changes in the analyzer sampling duct.  Determine the




           sample airflow requirements of the analyzer(s) and add the




           airflow (1.0 £/min) needed for the referee analysis.  Add




           about 10% of the total to insure an excess.   (NOTE:  The




           excess ozone stream should be vented or absorbed by a charcoal




           filter to avoid exposure to personnel.)  In a proper assembly,




           connection or disconnection of the analyzers  sampling ducts




           should not alter the airflow settings of the  instruments.
                                     -122-

-------
3.  The ozonizer lamp and power  supply should  be  sufficiently




    warmed up to Insure  a steady ozone output.  Some commercial




    ozone generators may require -up  to two  hours  for warm-up.




    With the ozonizer power  supply on  and the  lamp  completely




    shielded, flush the  system with  zero air for  several minutes




    to remove residual ozone.









4.  When a steady  analyzer reading is  obtained, pipet 10 -ml of




    the 1% KI absorbing  reagent  (Item  6.6.3) into each of the 2




    impingers.  Assemble the impingers and  connect  to the




    sampling train as shown  in Figure  6-III.   Draw  a sample




    from the sampling manifold through the  train  au 1.0 liter/




    min for 30 minutes.  Allow the solutions to stand for 15




    minutes.  Transfer the exposed solution from  each impinger,




    in turn, to a  clean  spectrophotometer cuvette and determine




    the absorbance of the solution at  352 nm against unexposed




    absorbing reagent as the reference blank.








    When the absorbance  of the solution in  the first impinger




    is greater than the  blank, continue flushing  the apparatus




    until the absorbance obtained is the same  as  the blank.




    NOTE:  Inability to  obtain a reading equal to the blank




    indicates possible presence  of oxidants in the  zero air or




    low pH (<6.5)  of the absorbing solution.   Correct problem(s)




    before continuing.
                               -123-

-------
    Zero the analyzer by adjusting the analyzer zero controls


    so that the output reads zero or reads the desired baseline



    value.






5«  Generate an ozone gas concentration equal to 80 ± 5% (span


    gas) of the analyzer's full scale measuring range.  When


    the analyzer response is steady, collect two samples of the


    gas stream and analyze as described in Step 4 above.  Adjust


    the sampling period so that the absorbance of the sampling


    solution is about midscale (between 0.2 to 0.5) on the


    spectrophotometer.  Five to ten minutes are usually sufficient.


    Do not exceed 30 minutes to avoid iodine losses.






6.  Add the two spectrophotometer readings to obtain total absorbance.


    From the volume of air sampled and the slope of the spectrophoto-


    meter calibration curve (See Step 1 in 6.7), calculate the


    ppm 0-j (or yg Og/m ) by equation 7:
            n    (A) (10 ml)
        ppia 03 =	                        (7)
    where:   A = absorbance of the solution


             b = slope of the spectrophotometer calibration curve


                 obtained in 6.7


            Va = volume in liters of sample collected (liters/ruin


                 X min)
                              -124-

-------
    To convert ppm (yl/1) to pg 03/m3, use equation 8.









        Vg 03/ia3 = (1960) (ppm 03)                  (8)







    Usually, the change in the gas volumes of the samples due to




    deviation from the standard conditions of 25°C and 760 torr




    are small and may be neglected.  When the deviations are large




    (sufficient to cause a change in gas volume greater than about




    5%) correction should be made.









7.  The concentrations of the two samples obtained should be within




    ± 5%.  Differences greater than 5% indicates possible unstable




    gas concentrations or error in the sample collection and analysis.




    Correct these problems before proceeding with the calibration.









8«  Determine the net analyzer reading by subtracting the analyzer




    baseline reading from the span reading.









9.  Adjust, when possible, the analyzer upper limit (span) controls




    to give a reading equivalent to the span concentration (spanning).




    When the instrument has no span controls, proceed to Step 10




    below.  Generate zero air and note the analyzer reading.  When




    the reading is different from the original reading by > 2%, reset




    the analyzer to the original baseline and repeat Steps 5 through




    9 above.  NOTE:  When the span and zero controls are not electri-




    cally independent, it may be necessary to rezero and respan




    iteratively until the proper zero and span settings are obtained.
                              -125-

-------
      10.  Generate, in turn, four additional ozone concentrations (10, 20,




           40 and 60% of full scale) and determine the ozone concentrations




           in duplicate as described in Steps 4 through 8 above.  Record




           the net analyzer readings by subtracting the baseline reading




           from the individual readings.









6.8.3  Treatment ofDynamic Calibration Data









       1.  Plot the net analyzer readings on the vertical axis versus the




           ozone concentrations on the horizontal axis of an appropriate




           graph paper (rectilinear, log, semi-log, etc.).  Calculate the




           slope b
-------
       also  6.1.4.   The procedures can also be used on other color imetric




       KI  instruments by using the appropriate reagent formulations.
6.9.1  Procedure
       1.   Turn the analyzer air and liquid pumps off.  Clamp off and




           disconnect the reagent line at the air-liquid separator or




           other convenient place.  Install a 125 ml low-actinic separatory




           funnel supported by a ring stand and connect the funnel to the




           inlet of the analyzer sample cell with 8 mm ID glass tubing.




           Butt-join the tubing with polyvinylchloride tubing.  Dis-




           connect the sample cell exit line at the inlet to the reagent




           reservoir and lead the line to a waste bucket.








           When the analyzer uses a solution reference, disconnect the




           reagent line at the exit port of the reference cell and recycle




           the solution back to the reagent reservoir so that the reagent




           flows through the reference cell only.  Turn the analyzer




           reagent pump on.








       2.   Prepare 10% KI absorbing solution blank by placing 20 ml of




           the 0.5 M buffer reagent (Item 6.6.8) and 20 ml of the 50% alkaline




           KI reagent (Item 6.6.9) in a 100 ml low-actinic volumetric flask.




           Dilute to volume with distilled water and mix well.
                                     -127-

-------
    NOTE:  This, and all subsequent 10% KI solutions should be




    prepared as needed and used within minutes to prevent inter-




    ference from oxidation of iodide to iodine by exposure to air




    or light.









3.  Rinse the separatory funnel, connecting lines and sample cell




    with a small portion (= 10 ml) of the 10% KI blank solution




    twice.  Fill the separatory funnel with the remainder of the




    reagent and adjust the flow through the sample cell to




    approximately the flow rate at which the instrument is to




    operate (1 ml/min - 20 drops/min).   The separatory funnel




    stopcock may be used to control the flow.  When the analyzer




    reading is stable, set the recorder baseline slightly above




    the pen indicator stop (or zero) to permit indication of




    negative deviations from zero.  Record the analyzer baseline




    reading and the position of the zero control.









4.  Prepare a calibrating solution equal to about 80 ±5% of full




    scale with the 10% KI solution as follows:  Make up 100 ml




    of 10% KI solution and mix well.  Transfer a portion to the




    cuvette to zero the speetrophotometer at 352 nm.  For a.




    full scale of 1 ppm, 80% of full scale is 0.80 ppni Oo.
            v» =
                             -128-

-------
    where:  Vw - volume (ml) of iodine working  solution  (6.6.11)


            Vc = volume of calibrating solution (100 ml)


            C  « concentration desired, in yl 03/101


            C^ = concentration of working solution (10 pi 03/ml)





    For example:  When the desired calibrating  solution is 0.20


    ul 03/ml, the volume Vw of the iodine working solution re-


    quired is:
                          (0.20 yl Do/ml)
            Vw - (100 ml) —	i—- = 2.0 ml      (10)
             w            (10 ul 03/ml)
    Pipette the required volume (8.0 ml) of the iodine working


    solution (6.6.11) to the remainder of the 10% KI in the


    Volumetric flcisk.  ZelTO tu5 SpeCtropaOtOUietfcr witu tile


    reagent removed earlier then dilute the flask to volume with


    the same' blank reagent.  Mix well.  This solution contains


    0.80 yl Oo/ml.  Discard the excess blank reagent in the


    cuvette.





5.  Rinse the cuvette with some of the above solution.  Read a


    portion of the solution in the spectrophotometer to verify


    that the solution contains 0.80 yl 03/ml.  When the proper


    reading cannot be obtained, check for error in the T£


    concentration, error in the dilution process or poor quality


    reagent.  Correct problem(s) before proceeding with the


    calibration.
                              -129-

-------
6.  Rinse, the separately funnel with a small portion  of  the remaining




    0.80 yl 03/ml solution and fill the separatory  funnel.   Introduce




    the solution through the sample cell in the same  manner as  Step 3




    above.  Note the analyzer reading and adjust, when possible,  the




    analyzer to give a reading equivalent to the spaa concentration




    (spanning).  For instruments with non-linear response,  proceed




    to Step 8 below.









7.  Prepare and introduce blank 10% KI solution and note  the analyzer




    reading.  When the response is different from the original base-




    line by > 2% of full scale, reset the analyzer  to the original




    baseline and repeat Step 3 above.  When the zero  and  span controls




    are not electrically independent, it may be necessary to rezero




    and respan iteratively until the proper zero and  span settings




    are obtained.









8.  Prepare, in turn, solutions equal to 10, 20, 40,  and  60% of full




    scale (e.g., 0.10, 0.20 yl O^/ml, etc.) and determine the analyzer




    response to each concentration as described in  Steps  4  and  5  above.









9.  Determine the net responses by subtracting the baseline reading




    froia the individual readings.  Replumb the analyzer  to  original




    operating condition.
                              -130-

-------
6.9.2  Treatment  of Static Calibration Data









      1.   Plot  the net analyzer readings on the vertical axis versus ul




           O-j/ml  of the calibrating solutions on the horizontal axis of




           an appropriate graph paper (rectilinear, semi-log, log, etc.).




           Calculate the slope bg of the static response curve by the




           method of least squares0.  To determine the linearity of the




           response, see Test 4.5 in Chapter 4.  A non-linear response




           from an instrument normally linear indicates malfunction in the




           analyzer or error in the preparation of the calibrating solutions.




           Correct the problem(s) before recalibrating.









       2..  When a non-linear response is normal, prepare a template as




           directed in Step 2 of 6.8.3.  An alternate -method is to attempt




           to linearize the instrument output by adjusting the electronics




           of the photometer by some combination of photocell voltage and




           span upper-limit setting until a linear output is found over




           the concentration range of interest.









6.9.3  Determination of Sample. Air Flqwrate









      For  instruments not equipped with adjustable -upper limit or span




      controls or when the range of span adjust is insufficient, the slope




      of the static calibration can be used to determine the sample airflow




      rate that  will make the analyzer output correspond to the pollutant







   cSee Table 4-V  in Chapter  4.
                                     -131-

-------
       concentration or a simple fraction or multiple of the concentration
       as follows:
           where:  f = fraction or multiple of the analyzer range desired.

                       (e.g., 0.5, 1, 2, 3)
                  Qa = airflow rate, ml /rain
                  Q  = reagent flowrate, ml/min
                  bs - slope of the static calibration curve obtained in 6.9.2


6.10   RECONCILING STATIC AND DYNAMIC CALIBRATIONS


       1.  The static calibration slope bg (6.9.2, Step 1) and dynamic cali-
           bration slope bj (6.8.3, Step 1) are compared in equation 10
           below.
                   R » -a-      x 100
                          bs
       2.  Large values of R (> 10%) are indicative of a) error in the
           analyzer's air or liquid flowrate, b) leaks or malfunction in
           the analyzer, c) poor quality reagents and/or reactants, d)
           error in the static or dynamic calibration process/or e) change
           in efficiency of sample collection.  Consult the analyzer
           operating instructions and/or the manufacturer and correct
           problem(s) before recalibrating.
                                    -132-

-------
6.11   BASIS  FOR CALCULATIONS





      The relationship between the standard iodine and ozone and the


      constant used in equation 6 (6.6.11) was obtained from the stoichio-


      metric equation 1.  One mole of ozone produces one mole of iodine


      according to the reaction:





               03 + 2KI + H20 -*• I2 + 2KOH + 02             (1)





       One mole of ozone, however, contains two equivalents (there is a


       two electron exchange in the half reaction) :
                    H20 + 2e~ -»• 20H~ + 02                  (11)
       Therefores one-half mole of ozone is equal to one equivalent ^eq.j


       of iodine.  Since one mole of ozone occupies, at 25°C and 760 torr,


       24.46 liters, then





               I eq. 03 = 12.23 liter 03 - 1 eq. I2


       or


                             12.23 E! 03 = 1 milliequivalent (meq) I2


       therefore


                                10 yl 03 -  	 veq I2       (12)

                                            12.23



                                         _  10 x 10"3 meq I7    (13)
                                         ~   12.23      H  2





       or in terms of solution:
                                     -133-

-------
                      10 yl 03/ml = 0.8177 x  10~3 meq I2/ml   (14)









The volume V,, in ml, of the dilute iodine solution  Nj needed




to make 250 ml of a working solution Nw equal to 10  yl O^/ml is




then:
                            V -, N. = V  N                       (15)
                             d  d    w  w
V  = V   "
y,   v  -


 d
                                     (250 ml)  (0.8177)  (IP"3)
                                    - . -


                                              (Nd)
                               rr    0>204

then                           Vd=  Nd
  ich is the basis for the constant in equation  6.
                              -134-

-------
6.12   REFERENCES
       1.   Intersociety Committee:   Tentative method for the manual


           analysis  of oxidizing substances in the atmosphere,  no.  44101-


           02-7OT.   Methods of Air  Sampling and Analysis.   Washington,


           D.  C., American Public Health Association, 1972




       2.   Saltzman  BE, Gilbert N:   lodometric micr ode termination of


           organic oxidants and ozone.  Anal Chem 31:1914-1920,  1959




       3.   Hodgeson  JA, Baumgardner RE,  Martin BE, Rehme KA:  Stoichioiuetry


           in the neutral iodornetric procedure for ozone by gas-phase titra-


           tion  with nitric oxide.   Anal Chem 43:1123-1126,  July 1971




       4.   Kopczynski SL, Bufalini  JJ:  Some observations  on stoichiometry


           of iodometric analyses of ozone at pH 7.0. Anal Chem 43:1126-


           1127, July 1971




       5.   Dietz RN, Pruzansky J, Smith JD:  Effect of pH  on the stoichiometry


           of the iodometric determination of ozone.  Anal Chem  45:402-404,


           Feb.  1973




       6.   ACS Reagent Chemicals, American Chemical Society Specifications.

                     »
           American  Chemical Society, Washington, B.C.  For suggestions  on
                                    -135-

-------
    the testing of reagents not listed by the American Chemical




    Society, see:  Rosin J:  Reagent Chemicals and Standards.




    New York, D. Van Nostrand Co., Inc.,  and The United States




    Pharmacopoeia.









7.  G Fredrick Smith Chemical Co., Columbus, OH  43223









8.  JT Baker Chemical Co.,  Phillipsburg,  N.J.   08865
                             -136-

-------
1
(-•
to
I
                                     Sampling manifold
             Ozone source
                                                                        Vent
                                      Analyzers
   Rotameter
      Flow
      controller

      Particle
      filter
     or
                                                                  Syn.
                                                                  zero
                                                                  air
                                                        Ambient air
           Air pump
                                                 Activated
                                                  carbon
                                                  filter
Figure 6-1.  Gas generating apparatus for calibrating ozone analyzers,

-------
U)
00
I
          Hi-voltage
         power supply
   Clean air •

    or oxygen
                                                           Aluminum box enclosure
                                          Adjustable sleeve

                                     9-in.  UV lamp
                                                                                      Collar


-o

X




1
-«-«_- — «.-«.: 	 1


rfffiiajinm»nuj-.._r.T...?"ar

^ f
                                        Quartz tube  15mm O.D.
                                             Figure 6-II.   Ozone generator.

-------
                                              flexible tubing
Flowmeter
                                            Trap
Needle
valve
          Figure 6-III.  Sampling train for referee ozone analysis.

-------
7.0    CALIBRATION PROCEDURE FOR AUTOMATED ATMOSPHERIC

        NITROGEN DIOXIDE AND NITRIC OXIDE ANALYZERS
7.1    PRINCIPLE AND SCOPE







7.1.1  This procedure is for the calibration of continuous analyzers used



       for monitoring atmospheric concentrations of nitrogen dioxide (N02)



       and nitric oxide (NO).  Analyzers for monitoring both pollutants



       concurrently are common.  The calibration may be of two types, dy-



       namic and static.  The dynamic calibration must always be done and



       is performed by determining the analyzer response to a series of NO or



       K02 concentrations.  The dynamic calibration is applicable to all NO,



       K02 or NOX, analyzers and is a performance test of the entire instrument



       under simulated service conditions.  The static calibration is



       performed by determining the analyzer response to artificial stimuli



       such as standard calibrating solutions, optical filters, screens,



       electrical- signals, resistors, etc.  This calibration is primarily



       applicable to analyzers using wet-chemistry and colorimetry and is



       a test of the detection and signal presentation components only.



       It is not a substitute for the dynamic calibration.







7.1.2  Each calibrating gas for the dynamic calibration is generated by



       mixing a dilute NO or N0£ gas stream or diluting with clean air a



       concentrated source such as N02 from a permeation tube or from a



       cylinder.  Each gas is sampled simultaneously with the analyzer and


                           1 2
       with referee methods: '   to establish the gas concentrations.
                                     -140-

-------
7.1.3   The  static- calibration is accomplished by adding known concentrations


       of  sodium nitrite to measured volumes of the analyzer absorbing


       solution.  The solutions produce colors with absorbances equivalent


       to  concentrations of N02-  These solutions are flowed through the


       analyzer's detector at the reagent flow conditions encountered


       during normal operation.  The analyzer readings are plotted versus


       N0£ concentrations to obtain the static calibration curve.  The


       instrument variables, e.g., air and liquid flow rates, may be


       adjusted after static calibration to make the output response


       conform to the pollutant concentration or a simple multiple or


       fraction of the concentration in parts per million (ppm) or micro-

                                  *3
       grams per cubic meter (yg/m ) (spanning) .  When a static is not


       performed, the spanning may be done during the dynamic calibration.






 7.2    RANGE





       The range of the calibration procedure is determined by that of the


       referee method.  For a 4-liter sample collected in 10 ml of absorbing


       solution and measured in a 1.0 inch cuvette, the range is between

                                         o                        33
       0.01 to 2ppm (10 to 3,760 yg N02/iir' and 12 to 2,450 yg NO/m ) .  The


       sampling rate may be reduced or the sampling period increased or


       decreased to extend or reduce the range.






       Continuous colorimetric analyzers for N02 and NO usually measure


       these gases in the same range.  The linear range for chemiluminescent


       NO~N02'analyzers is about 0.005 to 15 ppm (9 to 28,000 yg N02/m3 and
                                     •141-

-------
       6 to 18,400 yg NO/m3)^.  Amperometric analyzers usually measure  con-




       centrations from 0.005 to 1 ppm (9 to 1,880 yg N02/m3; 6  to  1,220




       yg NO/m3).   Sampling rates of these instruments range from 0.15  to




       4 1/min.








7.3    INTERFERENCES








7.3.1  The main interference arises from traces of NO in N02 and vice




       versa.  Zero air and pure streams of NO or N0£ can be obtained by




       the use of various combinations of chemical absorbers and con-


              ") S 6
       verters  ' * .  Details for preparing these absorbers are given in




       Table 5-III, Chapter 5.
                  «**







7.3.2  Relative humidities in excess of 90% for extended periods (> 3 hr)




       inactivate the chromium trioxide oxidizer used for converting NO




       to N0£ in the manual referee analysis and in some analyzers.




       Operating the oxidizer 5°C above ambient temperatures (e.g., by




       heating with a 7 W lamp) usually eliminates this problem.








7.4    PRECISION, ACCURACY, AND STABILITY








7.4^.1  A coefficient of variation of 1% can be achieved in the manual




       referee analysis with careful work7; the limiting factors are in




       the difficulty in measuring the volumes of the air samples and in




       reading the absorbance of the colored reagent.  Instrumental precision
                                     -142-

-------
      will usually decrease  between 2 to 5% because of airflow variations.




      These variations  are produced by pressure changes and unstable or




      pulsating  flows.









7.4.2  When spanning  is  possible,  any discrepancy between the input  and




      output may be  resolved by adjusting the instrument output to




      correspond to  the value of  the calibrating gas or solution.   Where




      analyzers  have no span controls, a correction factor  may be calcu-




      lated  to convert  the analyzer readings to pollutant concentrations.








7.4.3  The absorbing  solution used in the manual referee procedure or




      the static calibration should be protected from direct sunlight.




      Allow  15 minutes  for maximum color development and measure the




      absorbance within one  hour  after sampling to minimize any color loss.









7.4.4  A detailed discussion  of the various sources of error in calibrating




      gas preparation is given in Reference 1, Part I:  General Precautions




      and Techniques.  The minimization of the sources of error is  important




      to assure  high levels  of accuracy and precision.









7.5   APPARATUS








      A gas  generating  system (Figure 7-1) capable of providing calibrating




      gases  in the range of  0.01  to 2 ppm (19 to 3,760 yg N02/m3 or 12  to




      2450 yg NO/m^) is needed in the dynamic calibration.   The system




      consists of a  gas dilution  apparatus, sources of zero air and NO
                                    -143-

-------
       and N02 gases and a sampling train (Figure 7-II) for referee,  analysis.




       The components and connecting lines used should be sized and  assembled




       so that the differences in the pressure between the various flow com-




       ponents do not exceed 2% overall to prevent errors in flowrate measure-




       ments .









7.5.1  Gas Dilution Apparatus









       1-  Flowmeters - Rotameters with sapphire or stainless steel  balls




           are used to measure the flow of zero air (0 to 5 liter/min)




           and pollutant gas (0 to 1 liter/min).  They should be




           calibrated frequently (at least once a month)} with an




           appropriate flow measuring device such as a wet test meter,




           dry gas meter, soap bubble meter or calibrated rotataeter.









       2.  Connecting gas flow lines - Use 8 mm ID (minimum) borosilicate




           glass tubing or 10 ran ID polytetrafluoroethylene (TFE) tubing




           for all connections.  Flexible" polyvinyl chloride (PVC) tubing




           may be used to butt-join sections of inert tubings only.  Ball




           and socket joints (e.g., 12/5 mm) with 8 mm stems are preferred




           for connections that are frequently made and broken such  as  the




           calibrating gas lines to the dilution apparatus and the sampling




           ducts to the analyzer(s).









       3.  Mixing chamber - A cylindrical Kjeldahl type connecting bulb of




           200 to 300 ml volume works well.  This can also be fabricated
                                     -144-

-------
          from bprosilicate glass as shown in Figure 7-1, with standard



          12/5 ran balj. connections.








      4.  NO Oxidizer  (CrOo) .   p-m _ 1nA .   -, ,-n ,   . c    TTV  n       ,
          ____	    J/ •   till a 100 to 150 by 15 urn ID glass tube



          with ball connecting ends (see Figure 7-1) to a length of 20




          to 80 mm  with Cr03 granules (Item 7.6.6). Use glass wool plugs




          to hold granules in place.  (Caution!  Protect eyes and skin




          when handling this material.  Do not breathe the oxidizer dust.)




          Before use,  pass air containing 30 to 70% RH at 0.5 1/min




          about 1 hr.  to condition.  Discard when more than 3/4  of the



          packing bed  depth turns brown.








      5.  J£uaf~r_, 0  to  30.0 minutes - for measuring  the sampling  period in



          the referee  method.








7.5.2  Zero Air Source








      The air for diluting the calibrating gases should be free  of SO and




      N02 and substances that  will in any way 1) change the calibrating



      gas concentration, 2)  interfere in the analyzer  response or 3)  inter-




      fere in the referee procedure.  The zero air  -may be furnished from




      a gas cylinder or by filtering ambient air as indicated in Figure 7-1.








      1.  Cylinder  Zero Air  -  Cylinders  should be fitted with a  two-stage




          pressure  regulator and  with fine control  needle valves.  A




          Wv filter (Item 3a)  is used when  NOX contaminants are present.
                                     -145-

-------
    The humidifier (Item 3b) should be used with a pressure  relief

                                                       ty
    valve set between 2 to 5 psig (1.4 to 3.4 newton/cmz)  to prevent*


    expolsion of glass vessel in case of flow stoppage.




2.  -Air Pump (for transport of ambient air) - An oil-less  (carbon


    vane or diaphragm) air pump capable of delivering between 5


    to 10 liters/iain is used.  The pump is fitted with a vibration


    pad, a fine control needle valve, and a bleed-off valve.  A


    particle filter should be installed downstream of carbon vane


    pumps and is optional with diaghragm types.




3.  Air Filter




    (a) NOX filter - This is used to remove any NO and NO? from


        the air used for zero air.  Fill a 30 mm OD by 300 mm long


        glass tube to one-half its volume with CrOg oxidizer (Item


        7.6.6) and the rest with soda lime (Item 7.6.5).  Place the


        soda lime on the downstream side of the CrOo.  Use glass wool


        plugs in both ends to retain the granules.  The orange Cr03


        granules turns brown with use.  Discard when more than 3/4 of


        the packing length is brown.




    (b) Humidifier - This is used to humidify the zero air for the props


        operation of the Cr03 oxidizer (7.5.1, Item 4 and Item 3a) and


        consists of a 250 ml flask with the inlet tube extending to 2 a

        above water level as shown in Figure 7-1 to give a relative
                              -146-

-------
              humidity of 70 to 80%.  Install upstream  of  the  NOX filter


              (Item 3a) and the air pump  (Item  2) and downstream of the


              charcoal filter (Item 3c).






           (c) Charcoal Filter - See Item  2 in 5.2.2 and Table  5-III for


              requirements and preparation.






7.5.3   N02  Permeation Apparatus






       Proper  operation of permeation devices or tubes (Item 7.6.2)  requires


       a constant temperature bath maintained within 0.1°C  at 20  to  30°C


       and  equipped with a temperature conditioning coil and permeation

                                                                  Q
       tube holder.  The gas permeation  systems  described for S02° are


       »>1 ocv QTIT>I-i ^oKI o fr\r WO,-, cro-no-ra t--i rvn-'   C.otrme>f oT a 1  evc:t-ATn<2 J5TP  avfl T.1'-iM P
       Vb-l-u —f «.£-£.'_«-	 ^«.l— -I- 	 — ~ -_ ~. V ^ c>>^*«*^*_.	,	..-  .  —	—	~ ~~	— j ^ ~- -      -   . —  -  —






7.5.4   Ozone Generator (when calibrating chemiluminescent analyzers)





       An ozone  generator is used in the zero air line to furnish ozone  for


       converting NO to NC^-  This provides a mixture of NO and N02  which


       is used to test the efficiency of the N02-to-NO converter  in  chemi-


       luminescent analyzers10.  See 7.8.5 for details.  Ozone  generators


       for  conversion of NO to N02 are commercially available.  One  can


       also be fabricated by using a low pressure mercury lamp  (Ultra-


       violet  Products, Inc., San Gabriel, Cal., or equivalent) and  a


       lamp power supply.  See 6.5.1 in  Chapter  6 for details.
                                     -147-

-------
7.5.5  NO2 Absorber (TEA) -  Fill a 20 mm  ID  by 100 to 150 mm long polyethylene
       .inm,,.—,*-V  -----—   -i ,,  m-^, i_ _,  INI                                                    ^


       tube with tubing  fittings at both ends with TEA granules (Item 7.6.4),



       Hold the granules in place with glass  wool  plugs.   Install the absorber



       in the line between the NO source tank and  the  dilution apparatus.








7.5.6  Spectrophotoaeter or colorimeter - With  2.54 cm (1 inch)  cuvettes



       for measuring the absorbance at 550 nm.







7.5.7  Manual Samp1ing Train:







       Assemble the sampling train as shown in  Figure  7-II.   "Use ground-



       glass joints upstream from the bubbler.   Other  connections may be



       butt-joined with polyvinylchloride tubing.







       1.  Absorber - All-glass bubbler with  a  maximum frit pore diameter of



           60y as shewn in Figure 7-II or in  References 1 and 2  are needed



           to collect the samples for referee analysis.   Bubblers may be



           purchased from suppliers of air pollution equipment and many



           glassware suppliers.  Ball-joint connections are recommended  for



           convenience.








       2.  Air-vacuum Pump - Capable of drawing 0.40 liter/uiin through the



           bubbler.  The pump should be equipped with  a needle valve  and



           trap at the inlet side to regulate flow.








       3.  Flowmeter - Calibrated to measure  gas flows  of  0.5 to 2 liters/



           min within ± 2%.   See Item 1 in 7.5.1 for calibration requirements.

-------
      ^*  ;*•£§£ ~ A vessel containing glass wool to protect needle valve from




          moisture.









7.5.8 Miscellaneous;   Other equipment needed to perform the manual reference




      sampling operation described in Reference 1 and 2,









7.5.9 CalibratingSolution Dispenser(for static calibration);  A solution




      dispenser, consisting of one or two lengths (about 50 cm)  of small




      diameter (about 1.5 mm I.D.) clear polyvinylchloride (F?C) tubing,




      a plastic anti-syphon device and a 125 ml low-actinic separatory




      funnel as shown in Figure 7-III, is used to deliver the calibrating




       solutions under controlled flow conditions by means of furnished




       screw clamp(s)  on the tubing(s).  This (hypodermoclysis) set,




      which is intended for administering fluids under the skin, is




       inexpensive and can be purchased from medical supply houses.




      A similar device can be fabricated from small diameter plastic




       tubing, Y connector and screw clamps.









7.6    REAGENTS AND GASES








      Purity of Chemicals - Unless otherwise indicated, all reagents




       specifications shall conform to the Committee on Analytical Reagents




       of the American Chemical Society11.  When such reagents are not




       available, ascertain that they do not lessen the accuracy of the




       determination.
                                     -149-

-------
7.6.1  Cylinder NO;  A cylinder of compressed N2 gas containing about 100


       ppm NO (122 mg NO/m3), <10 ppm H20 and <3 ppm 02:  This mixture is


       quite stable when care is taken to exclude air during preparation.


       It may be purchased or can be prepared in the laboratory using a

                                              1 O
       closed-systems vacuum-pressure manifold  .  Further dilution of the


       gas brings the concentration within the normal range of NO^ analyzers.






       Permeation tubes containing NO are not available because of its


       low boiling point (-152°C).






7.6.2  NO2.Permeation Devices;   Permeation devices containing N0£ may be


       purchased13'  .  Their use usually requires the use of a temperature


       controlled bath.  While S02 permeation devices can be used as a primary


       standard because of their reliable emission rates, HO^ permeation


       devices have not yet achieved the same level of reliability.  The


       exact concentrations produced, therefore, must be established during


       calibration by a referee method .  Exposure of the tube to water


       vapor and wide variations in temperature during operation or storage


       may produce marked changes in permeation rates.  N02 devices rmay


       also be used for calibrating NO analyzers when a N02~to-N0 converter


       is used.
7'6'3  °ther Sources of NO and N02;  Lower concentrations of NO and N02 in


       steel cylinders (about l/10th of those listed in 7.6.1) may be used


       but since these substances are unstable at these low levels, the
                                     -150-

-------
       concentrations resulting from such -mixtures must be determined by




       referee analysis just prior to and/or during calibration.









7.6.4   Triethanolamine (TEA.) ;  Soak 1.5 mm (1/16") diameter or 10 to 14 mesh




       granules of porous, inert material such as firebrick, alumina, or




       xeolites, in 20% aqueous TEA6.  Drain, spread on a dish, and dry




       for 30 to 60 min at 90 to 95°C.  The granules should be free flowing.









7.6.5   Soda Lime - Granules, 4-8 mesh.









7.6.6   Chromium Trioxide (CrOa) CbdLdizer2'6 - Soak 15 to 40 mesh porous
       firebrick or alumina in a 16% w/w aqueous CrO^ solution.  Drain,




       spread on a dish and dry in an oven for 30 to 60 min at 105 to 115°C,




       then place the dish in a desiccator containing a saturated salt




       solution which maintains the relative humidity between 50 to 70%.




       The reddish color changes to a golden orange when properly equili-




       brated.  CautionI  Protect eyes and skin when handling this material.




       Do not breathe the oxidizer dust.









7.6.7  NaN02 Stock Solution.  Dissolve 2.03 g NaNO£ (a correction is applied




       if purity is less than 100%) in distilled water and dilute to 1000 ml




       in a volumetric flask.  This solution contains the equivalent of




       1000 yl N02/ml.  The weight of the nitrite is calculated based upon




       the empirical observation that 0.72 mole of sodium nitrite produces




       the same color as 1 mole of nitrogen dioxide >J--).
                                     -151-

-------
7.6.8  NaN02 Working Solution (10 yl N02/ml) .  Dilute 5.00  ml  of the stock




       NaN02 solution (Item 7.6.7) to 500 ml with distilled water (nitrite




       free) in a low-actinic volumetric flask.  This working  standard  is




       equal to 10 yl of N02/ml,
7.6.9  A^Si^i^H^l—lSESHLL^,5  Reagents needed to perform the manual  referee




       analysis are described in Reference 1.









7.7    SPECTROPHOTOMETER CALIBRATION









       1.  Prepare a series of calibrating solutions equivalent to  between




           0.04 to 0.20 yl N02/ml by diluting 1,2,3 ml etc of NaN02 working




           standard (Item 7.6.8) in 250 ml of the azo-dye absorbing reagent as




           directed in Reference 1 and in 7.9, Static Calibration.  Allow




           15 minutes for full color development and then read the  solutions




           in the spectrophototueter at 550 nm.  Plot the absorbances on the




           vertical axis versus yl N02/ml on the horizontal axis of a recti-




           linear graph paper as a check for linearity.  Calculate  the slope




           b of best-fit curve for the data using the method of least squares3.









       2.  When a static calibration is to be performed on the analyzer and




           the formulation of the analyzer absorbing solution is different




           from that of the referee procedure, obtain a spectrophotometer




           calibration factor by diluting aliquots of the NaN02 working




           solution with the analyzer reagent as described in Step  1 above.
       aSee Chapter 4, Table 4-V.
                                     -152-

-------
7.8    DYNAMIC  CALIBRATION









7.8.1  General










      1.   When a static calibration (7.9) is to be performed on the analyzer,




           it should be done before the dynamic (7.8)  to assure proper  operation




           of its detection and signal presentation components.  The static  and




           dynamic responses may be compared (reconciled)  to verify the proper




           operation of the analyzer.  See 7.10 for details.









      2.   When the analyzer has been operating as continuous monitor,  it




           is useful to determine the response near the span level first




           without changing the span settings (auditing).   When the response




           is within 10% of the. previous calibration;  the  calibration is




           still valid and a new calibration is not necessary.  When the




           response is greater than 10%, proceed with  the  complete cali-




           bration.  The audit data provide a record of the calibration




           drift.  Instruments with non-linear response require the full




           calibration.









      3.   Chemiluminescent NOV analyzers with single  detectors that measure
                              JV



           N02  by difference (N02 = NOX - NO) are usually equipped with




           converters for reducing the N02 to NO.  Such instruments must




           first be calibrated with NO because the analyzers response to




           N02  depend on the efficiency of the converters.  See 7.8.5




           for  details.
                                     -153-

-------
4.  The analyzer to be calibrated should be in good operating condition




    and installed in accordance with the manufacturer's  instructions.




    Operate the analyzer for at least 24 hours to warm-up.   This  24-hr




    warm-up period may be shortened when so stated in  the operating




    instructions.  Adjust the air and reagent flowrates  to  their




    recommended rates or to the rates determined from  the static




    calibration (7.9) data and verify the rates as described  in 7.5.1,




    Item 1.









5.  Record all data only after stable analyzer response  has been




    attained.  Refer to 4.1.6 in Chapter 4 for determining stable




    response.









6.  Assemble the gas generating system for generating NO or-N02




    as needed.  Place the system as close as practical to the




    analyzer to prevent losses and to minimize pressure  changes




    in the analyzer sampling ducts.  Calculate the airflow of




    the analyzer(s) and add the airflow (400 ml/min) needed for




    the referee analysis.  Add about 10% of the total to insure




    an excess.  (NOTE:  The excess calibrating gas should be vented




    or absorbed by a non-restricting filter to prevent exposure to




    personnel.)  In a proper assembly connection or disconnection of




    the analyzer sampling line should not alter the airflow settings.
                              -154-

-------
7.8.2  Analyzer Zero









       1.   Generate a flow of zero air equal to the rate determined in




           7.8.1, Step 6 above and sample with the analyzer.  When a




           stable analyzer reading is obtained, pipet 10 ml of the




           azo-dye absorbing reagent in the bubbler (7.5.7, Item I)1.




           Assemble the bubbler and connect to the sampling train as shown




           in Figure 7-II.









           Sample the gas stream according to the manual referee method




           for 10 minutes at 400 ml/min.  Wait for 15 minutes for color




           development and then transfer the solution to a clean 1.0 inch




           (2.54 cm) cuvette.  Measure the absorbance of the solution at




           550 nm on the spectrophotonieter against unexposed absorbing




           reagent as blank.









       2.   When the absorbance of the solution is greater than the blank,




           continue flushing the generation system until the absorbance




           obtained is the same as the blank.  Inability to obtain a




           reading equal to the blank indicates interferents in the zero




           air.  Correct the problem before proceeding.









       3.   Zero the analyzer by adjusting the analyzer controls so that the




           output corresponds to zero or the desired baseline value.
                                    -155-

-------
7.8.3  Generation of N02



       1.  Connect the source of the NC>2 (NC>2 permeation apparatus  7.5,3  or

           the NO cylinder 7.6.1 and ozone generator 7.5.4) as  shown  in the

           gas generation system (Figure 7-1).  It -may be necessary before

           hookup to bleed the cylinder outside the generation  system at  a

           low flow (30 to 50 ml/min) for one to two hours to condition the

           delivery tubing.  (NOTE:  Bleed the N0£ into a filter to avoid

           exposure to personnel).



       2.  Generate a N02 concentration equal to 80 ± 5% (span  gas) of full

           scale.  When a steady analyzer reading is obtained,  collect two

           samples of the gas stream and analyze as described in Step 1 of

           7.8.2.  Adjust the sampling period to keep the absorbance  of the

           samples at about midscale (0.3 to 0.5 A) on the spectrophotometer.

           Five to ten minutes are usually sufficient.



       3.  From the volume of air sample and the slope of the spectrophotometer

           calibration curve, calculate the ppm N0£ as follows:
                                 (A) (10)                  (1)
                           ppm = (b) (vj
           Where:   ppm = concentration (u/D N02 or NO in air

                     A - absorbance of the solution

                     b = slope of the spectrophotometer calibration  curve

                         obtained in 7.7.
                                     -156-

-------
             Va = volume, In liters, of  the gas  sample collected  (liters/

              min X rain) .
                          <5
    To convert ppm to pg/m , use equation 2 for NO and equation 3 for



                Ug NO /in3 = ppm NO X 1230            (2)


               pg N02/m3 = ppm N02 X 1880           (3)
    Usually, the change in the gas volumes of the samples due to deviation

    from the standard conditions of 25°C and 760 torr is small and may be

    neglected. ,- When the deviations are large (sufficient to cause a

    change > 5% in gas volume) , correction should be made.



4.  The concentration of duplicate samples should be within 5%.

    Differences greater than 5% may be due to unstable gas concentration

    or error in sample collection and analysis.  Correct these problems

    before proceeding with the calibration.



5.  Determine the net analyzer reading by subtracting the baseline

    reading from the span reading.



6.  When possible, adjust the analyzer to give a reading equivalent

    to the span N02 concentration obtained in Step 3 above (spanning) .

    When the instrument has no span controls, proceed to Step 7 below =

    Generate zero air and note the analyzer reading.  When the reading

    is different from the original baseline reading by > 2%, reset the
                              -157-

-------
           analyzer to the original baseline and repeat Steps 2 through 6




           above.  NOTE:  When the span and zero controls are not electrically




           independent, it may be necessary to rezero and respan iteratively




           until the proper zero and span settings are obtained.









       7.   Generate^ in turn,  four additional concentrations (10, 20, 30,




           40 and 60% of full scale are suggested).   Determine the N02




           concentrations in duplicate as described  in Steps 2 through 4




           above.  Record the net analyzer responses by subtracting the




           baseline reading from the individual readings.









7.8.4  Generation of NO









       1,   Install the TEA NO^ absorber (Item 7,5.5) at the outlet of the




           NO (Item 7.6.1) cylinder to remove any traces of N02«  Install




           the NO (Cr03) oxidizer (Item 4 in 7.5.1)  in the manual sampling




           train as shown in Figure 7-II. 'This oxidizes NO to N0£ for




           referee analysis.  The humidity of the gas stream is controlled




           by the humidifier in the zero air line as shown in Figure 7-1.




           Connect the cylinder of NO as shown in Fig. 7-1.  The rest of




          .the calibration follows the same procedure as described for N02




           in 7.8.3 above.









       2.   Determine the NO concentration by equation 1 in Step 3 of 7.8.3.




           This procedure is valid because the conversion of NO to N02 is




           100% when the CrO-j is used as specified .
                                     -158-

-------
7.8.5   Determination of Converte^j^ffj-ciency (for chemiluminescent analyzers)







       1.   To determine the converter efficiency of cherailuminescent NCL



           and NO  analyzers, generate an NO concentration equal to 80 ± 5%
                 ji


           of full scale.  Verify the concentration and record the analyzer



           readings YNQ and Y    in both the NO and NO  modes.
                               2C





       2.   Oxidize a portion of the NO to NO,, by turning the ozone generator



           (Fig 7-1) on and adjust the ozone output so that the NO reading
Y   is reduced to about one-half.  Record the new NO and NO


            'NQ
            readings Y'NQ and Y'NO  .
                                                  1
           The decrease in the NO reading  (YNQ - Y NQ) is equal to the


           input N02 concentration.  The analyzer output as N02 is equal


           to the difference between the new NO^ and NO readings or
            '    - Y*
             NO,,   '  NO'
       3.  Calculate the conversion efficiency fc by equation 5.
               _
             f  -
                            Y*     - Y'
                              N0x
                                . - .


                            YNO ~  Y. NO
           'A ratio f  of the  output N02  to  the  input N02  of  less  than  1.0


           indicates the conversion is less than  100%  and the  converter


           should either be corrected or  preferably replaced.
                                     -159-

-------
7.8.6  Treatment of DynamicCalibrationData









       1.  Plot the net analyzer readings for each channel on  the vertical




           axis versus the ppm NC>2 (7.8.3) or ppm NO  (7.8.4) obtained by




           manual sampling on the horizontal axis of an appropriate  graph




           paper (rectilinear, semilog, log, etc.).  Calculate the slope




           b,j of best-fit curve for the data using the method of least




           squares^.  To determine the linearity of the response, see Test




           4.5 in Chapter 4.  A non-linear response from an instrument




           normally linear indicates malfunction in the analyzer or  error




           in the calibration process.  Correct problem(s) before




           recalibrating.








       2.  When a non-linear response is normal, draw a smooth line  through




           the calibration points that fits the data best.  From this,




           prepare a template to convert the net analyzer readings to




           cone en trat ion uni t s.









7.9    STATIC CALIBRATION









       The following section pertains to colorimetric analyzers using azo-dye




       forming reagents for monitoring NO, N02 or NO^,  See also 7.1.3.









7.9.1  Procedure









       1.  The reagent used in the following procedure should be formulated




           in accordance with the manufacturer's recommendations.






       bSee Chapter 4, Table 4-V.




                                     -160-

-------
2.  Turn the analyzer air and liquid pumps  off.  Disconnect  the




    reagent line at the exit port of the photometer reference




    cell and recycle the solution to the reagent reservoir.  Dis-




    connect the spent reagent line at  the used-reagent reservoir,




    or at the inlet to the reagent recovery carbon column, when so




    equipped, and lead it to a waste bucket.  Disconnect the




    reagent line at the sample cell inlet and connect the delivery




    end of a calibrating solution dispenser (Item 7.5.9).  Support




    the separatory funnel near the instrument on a ring stand.




    Turn the analyzer reagent pump on.  The reagent now flows only




    through the, reference cell.  Siphon absorbing solution from




    the analyzer reagent reservoir into six 250 ml low-actinic




    volumetric flasks and fill almost  to the mark.









3.  Mark one of the flasks as blank.   Pour  about 15 ml of the




    reagent from the blank into  the separatory funnel.  Squeeze




    out any air bubbles trapped  between the funnel and the screw




    clamps.  Open the screw clamp(s) and flush the connecting




    line and cell.  Repeat with  a second 15 ml portion.  Fill




    the separatory funnel with the blank reagent and adjust  the




    discharge rate into the cell at about the same flow rate at which




    the analyzer is to operate (1 ml/min equals about 20 drops/




    minute).  For dual channel analyzer that monitor both pollutants




    (NO and NO ) concurrently, Steps 3 through 8 can be performed




    simultaneously using the  same calibrating solution for both




    channels.
                              -161-

-------
4.  While waiting for the instrument reading to stabilize, determine,




    by equation 6 s the volumes of working NaN02 solution  (Item 7.6.8)




    required (in 250 ml) to produce a series of calibrating standards




    equal to 10, 20, 40, 60 and 80% of full scale.







                                  (C)               C6)
    where:   C = concentration (in yl N02/ml) desired




            Cy = concentration of working solution (10 ul N02/ml)




            Vc = volume of calibrating solution (250 ml)









    Pipet the required volumes, e.g., 1.0, 2.0, into the five




    remaining 250 ml volumetric flasks containing the absorbing




    reagent-  Dilute .to the mark with the reagent and mix well.




    Let stand for 15 min for full color development .









    Read the color of a portion of each standard solution with




    the previously calibrated spectrophotometer (Item 7.7) to verify




    that each standard contains the equivalent yl N02/ml calcu-




    lated in Step 4.  When the proper readings cannot be obtained,




    check for error (s) in the NaN02 concentration, error in the




    dilution process or poor quality reagent(s) .  Correct problem(s)




    before proceeding with the calibration.









5.  After the analyzer response has stabilized on the blank reagent,




    adjust the analyzer controls to obtain the desired baseline




    reading.
                              -162-

-------
      6.  Introduce the standard  equal  tc  80%  of  full scale  as  directed  in

          Step 3 above.  When a steady  reading is obtained,  adjust, when

          possible, the analyzer  reading to  correspond to  the concentration,
          of standard.  For  instruments with no span  controls,  proceed to
          Step 8 below.


      7.  Introduce blank  reagent and note the analyzer  reading.  When the

          reading is  different from  the original  baseline  by >  2%, reset
          the analyzer  to  the original  baseline and repeat Steps 3 and 5.
          When the span and  zero  controls  are  not electrically  independent,
          it may be necessary to  rezero and  respan iteratively  until the
          proper zero and  span settings are  obtained.


      8.  Introduce the remaining calibrating  standards  in turn and record
          the corresponding  analyzer readings  as  described in Step 5.
          Determine the net  readings by subtracting the  baseline reading
          from, the individual readings.


7.9.2  Treatment of Static  Calibration Data


      1.  Plot the net  analyzer readings on  the vertical axis versus ul
          N02/ml of the calibrating  solutions  on  the  horizontal axis of
          an appropriate graph paper (rectilinear, semi-log, log, etc).
          Determine the slope bs  of  the static response  curve by the
          method of least  squaresc.  To determine the linearity of the
          response, refer  to Test 4.5 in Chapter  4.   A non-linear response
          from an instrument normally linear indicates malfunction in
          the analyzer  or  some error in the  preparation  of the  cali-

      cSee Chapter 4, Table  4-V.
                                    -163-

-------
           brating solutions.  Correct the problem(s) before proceeding




           with the calibration.









       2.  When a non-linear response Is normal, prepare a template as




           directed in 7,8.6, Step 2.  An alternate method is to attempt




           to linearize the instrument output by adjusting the electronics




           of the photometer by some combination of photocell voltage and




           span upperlimit setting until a linear output is found over the




           concentration range of interest.









7.9.3  Determination of Airflow Rate









       1.  For instruments not equipped with adjustable upper limit or




           span controls or when  the range of span adjust is insufficient,




           the slope of the static calibration curve can be used to establish




           the sample airflow rate that will make the analyzer output




           correspond to the pollutant concentration or a simple fraction




           or multiple fg of the  concentration range as follows:








                                 (fs>


                       Qa " (Qr>  TTT                      (7)
                                 (OQ)







           where:  Q  = airflow rate, ml/rain
                    ct



                   Qr = reagent flow rate, ml/min




                   f  = fraction  or multiple of the analyzer range desired
                    s



                   b  - slope of  the static calibration curve obtained in 7.9.2
                                    -164-

-------
         The  NO channel is handled in like manner.  Equation 7 is also


         useful if the analyzer read-out is in yg/nr^ since the static



         calibration curve is then units of absorbance versus yg N02/ml.






7.10  RECONCILING THE STATIC AND DYNAMIC CALIBRATIONS






      The static calibration slope bg (7.9.2, Step 1) and the dynamic


      calibration slope b^ (7.8.6, Step 1) are compared by equation 8:




                              lbg - brr|
                          R »    b	   X 100             (8)
                                  S





      Large values of R (£ 10%) are indicative of a) error in the analyzer's


      air or reagent flowrate, b) leaks or malfunction in the analyzer,


     . e) poor quality reagents, d) error in the static or dynamic calibration


      process or e) change in efficiency of collection.  Consult the


      analyzer operating instructions and/or manufacturer and correct


      the problem(s) before recalibrating.
                                    -165-

-------
7.11   REFERENCES
       1,  Intersociety Committee:  Tentative method of analysis for




           nitrogen dioxide content of the atmosphere (Griess-Saltzman




           reaction).  Hlth Lab Sci 6:106, 1969.  Reprinted in Methods




           of Air Sampling and Analysis.  Intersociety Conimitteej p 329-336.




           American Public Health Association 1015 Eighteenth Streets N.W.,




           Washington, B.C., 1972.









       2.  Intersociety Committee:  Tentative method of analysis for nitric




           oxide content of the atmosphere.  Hlth Lab Sci 9:71, 1972.




           Reprinted in Methods of Air Sampling and Analysis.  Intersociety




           Committee, p. 325-328.  American Public Health Association, 1015




           Eighteenth Street, N.W., Washington, B.C., 1972.









       3.  ASTM:  Continuous measurement of nitric oxide, nitrogen dioxide,




           and ozone in the atmosphere.  Method D 2012-69.  Annual Standards.




           Part 23, 620-27, November,,1970.









       4.  Hodgeson JA, Rehme KA, Martin BE, Stevens RK:  Measurements for




           atmospheric oxide of nitrogen and ammonia by chemiluminescence.




           Presented at Air Pollution Control Association Meeting, Miami, FL,




           June, 1972.









       5.  Intersociety Committee:  Tentative method for calibration of




           continuous colorimetrie analyzers for atmospheric nitrogen dioxiue




           and nitric oxide.  Hlth Lab Sci 9:4, October, 1972.
                                     -166-

-------
6.  Levaggi DA,  SIu W,  Feldstein M5  Kothny EL:   Quantitative




    separation of  nitric oxide from  nitrogen dioxide  at  atmospheric




    concentration  ranges.  Env Sci Tech 6:250,  1972.








7.  Thomas, MA.,  Amtower RE:  Gas dilution apparatus for  preparing




    reproducible dynamic gas mixtures in any desired  concentration




    and complexity.  J  Air Pollut Contr Assn 16-618,  1966.








8.  Intersociety Committee:  Tentative method of analysis for sulfur




    dioxide  content of  the atmosphere.  Hlth Lab Sci  7:4, 1970.




    Reprinted in Methods of Air Sampling and Analysis, Intersociety




    Committee,  p 447-455.  American  Public Health Assoc., 1015




    Eighteenth St., N.W., Washington, B.C., 1972.








9.  Saltzman BE, Burg WR, Ranaswamy  G:  Performance of permeation



    tubes as. standard gas sources.  Env Sci Tech 5:1121, 1971.








10.  Wright B, Jeung E:   Simplified calibration  procedure for




    chemiluminescent NO-NC)  analyzers.  Presented at  the 13th
                           X



    Conference on Methods in Air Pollution and  Industrial Hygiene




    Studies,  California State Department of Health, Berkeley,




    California,  October 30-31, 1972.








11.  ACS Reagent Chemicals, American Chemical Society  Specifications.




    American Chemical Society, Washington, D.C.  For  suggestions on
                               -167-

-------
     the testing of reagents not listed by the American Chemical




     Society, see:  Rosin J:  Reagent Chemicals and Standards.




     New York, D. Van Nostrand Co., Inc., and The United States




     Pharmacopoeia.









12.  Belsky T:  Preparation of low concentration mixtures of gases,




     Air and Industrial Hygiene Laboratory, California State Depart-




     ment of Health, Report No. 117, Berkeley, California, January




     1972.









13.  Metronics Associates, Inc., Palo Alto, CA  94304.









14.  Analytical Instrument Development,  Inc., 250 South Franklin Street,




     Westchester, PA  19380.









15.  Scaringelli FP, Rosenberg E, Rehine  KA:  Comparison of permeation




     devices and nitrite ion as standards for the colorimetric determinat




     of nitrogen dioxide.  Env Sci Tech  4-924, 1970.
                              -168-

-------
              Gas mixing
                chambe r
                                   (for NOo stream)
                                                   JJ-
                               CrO~ (for NO  stream)
                                    Sampling manifold
                                       M   M   M
      Analyzers
Zero air
                                                         eferee
                                                        sampling
                                                       !train
                               Air pump

                               	Dilute NO
                                                  Humidifier


              Dilute N02  (from N02 permeation apparatus)
                                                                      Ambient
                                                                        air
Figure 7-1.  Gas generating system i'or calibrating  NO  and  N02  analyzers.

-------
                                 Cr03 (for NO  stream)
o
i
                                                    Absorber
                                                    (fritted)
                                                                                     Flexible tubing
                                                                                                                    To
                                                                                                                   .air
                                                                                                                    pump
                                                                                                  Needle  valve
        From gas manifold

-------
              Screw Clamp
Connect to Analyzer
Solution Cell Inlet'
                                        Separatory Funnel
                                          125 ml  Cap
                                           Hypodermoclysis  Set
Connect to Analyzer
Solution Cell Inlet
                               Flexible
                                Tubing
            Figure 7-III.  Calibrating solution dispenser.

-------
8.0    CALIBRATION PROCEDURE FOR AUTOMATED ATMOSPHERIC SULFUR

        DIOXIDE ANALYZERS
8.1    PRINCIPLE AND SCOPE







8.1.1  This procedure is for the calibration of continuous atmospheric



       sulfur dioxide analyzers.  The calibration may be of two types,



       dynamic and static.  The dynamic calibration must always be done



       and is performed by determining the analyzer response to a series



       of sulfur dioxide (SO,.,) concentrations.  The dynamic calibration is



       a performance test of the entire analyzer under simulated service



       conditions and is applicable to all S02 analyzers.  The static



       calibration is performed by determining the analyzer response



       to artificial stimuli such as standard calibrating solutions,



       optical filters, screens, electrical signals, resistors, etc.  This



       calibration is a test of the detection and signal presentation com-



       ponents only and is primarily applicable to S02 analyzers using wet-



       chemistry such as colorimetry and conductimetry.  It is not a substitute



       for the dynamic calibration.







8.1.2  The calibrating gas for the dynamic calibration may be generated in



       two ways.  The preferred method is by mixing a stream of S02 from


                                  1 *) *\
       an S02 permeation apparatus ''J (Figure 8-1) with clean air.

                                                                           *3

       Alternatively, streams of dilute S02 (50 to 100 ppm; 131 to 262 yg/m )



       from a cylinder may be mixed with clean air.
                                    -172-

-------
      The calibrating  gas  is  sampled simultaneously with the analyzer and  -




      with the  referee method^ to establish the concentration of the gas.




      A permeation tube with  a known emission rate can be used as a primary




      'standard  (see 8.6.1)  source of SC>2 gas.









8.1.3  The static  calibration  is performed by adding known concentrations




      of a standard reagent to measured volumes of the analyzer absorbing




      solution  which provide  an effect equivalent to concentrations of S02-




      These  solutions are flowed through the analyzer detector at the actual




      reagent flow conditions encountered during normal operation.  The




      analyzer  readings are plotted versus equivalent S02 concentrations




      to obtain a static calibration curve.  The instrument variables, e.g.,




      air and liquid flow rates, may be adjusted to make the output response




      conform to the pollutant concentration or to a simple multiple or




      fraction  of the concentration in parts per million (ppm) or micrograms




      per  cubic meter (ug S02/m^) (spanning).  When a static calibration is




      not  performed, the spanning may be done during dynamic calibration.









8.2   RANGE








      The  range of the calibration procedure is determined by that of the




      referee method^.  For a 10-liter sample collected in 10 ml of absorbing




      solution  and measured in a 1.0 inch cuvette, the range is between




      0.01  to 10 ppm S02 (0.026 to 26 yg S02/m3).
                                    -373-

-------
       The measurement range  and  the  sampling  rates  of  continuous SC^




       analyzers vary greatly depending  on  the detection methods.  The




       upper limit of the measuring range can  vary from 0.2  to 10 ppm




       (0.52 to 26 yg/m3) .  Sampling  rates  may vary  between  0.015 to 5 1/min,




       A descriptive compilation  of most of the currently available S02




       analyzers is given in  Reference 5.









8.3    INTERFERENCES









8.3.1  Interferences are a function of the  detection principle as shown in




       Table 5-IV, Chapter 5.  S02 with  purity greater  than  99.9% is readily




       available.  Higher purities are obtainable when  required.   Zero air




       for dilution must be free  of SC>2  and other substances that can




       potentially interfere  in the analyzer detection  principle.  Selective




       absorbers (drying agents,  Ascarite for  C02> etc.)  can be used when-




       ever a particular measurement  principle requires it.   Details for




       preparing such absorbers are given in Table 5-III,  Chapter 5.









-8.4    PRECISION, ACCURACY, AND STABILITY









8.4.1  With careful work the  coefficient of variation at the 95%  confidence




       level of the pararosaniline referee  method is 4.6%^.   Careful attention




       to the details of the  method is critical.









8.4.2  When spanning is possible, any discrepancy between the input and output




       may be resolved by adjusting the  instrument output to correspond to the
                                      -174-

-------
      calibrating  gas  concentration.   Where analyzers have no spanning




      controls,  a  correction factor may be calculated to convert the




      analyzer readings to S0£ concentrations.









8.4.3  A detailed discussion of the various sources of error in the prepar-




      ation  of calibrating gases is given in Reference 7,  Part I:   General




      Precautions  and  Techniques.  The appreciation and minimization of




      the  sources  of errors is important to assure high levels of  accuracy




      and  precision.









8.5   APPARATUS









      A gas  generation system consisting of sources of S02 and zero air,




      .flcvseters,  gas  mixing chamber,  sampling manifold and a sampling train




      for  referee  analysis is needed in the dynamic calibration (Figure




      8-1).   The system should be capable of providing calibrating gases




      between 0.002 to 1.0 ppm (0.005  to 2.6 yg S02/m3).  The S02  may be




      furnished  from an S02 permeation apparatus or from a cylinder of




      dilute S0'2 gas.








      The  components and connecting lines making up the system should be




      sized  and  assembled so that the  differences in the gas pressure




      between the  various components do not exceed 2% overall to prevent




      errors in  flowrate measurements.  Ball and socket joints are con-




      venient for.  connections that are frequently made and broken.
                                     -175-

-------
8.5.1  S02 Permeation Apparatus;  See Figure  8-1.   This can also be purchased,









       I-  Flowmeters ;  To measure the  flows  of  zero air (0 to 1 liter/inln)




           over the permeation tube.  They  should  be calibrated frequently




           (monthly) with wet or dry test meters,  soap bubble meter or




           calibrated rotameter.









       2.  Temperature-contro lied Bath ;  Maintained at 20 to 30 ± 0.1°C.




           It is needed for proper operation  of  permeation tube 8.6.1.




           The baths described for S(>2  '  are acceptable.









       3.  Need! e valves :  for controlling  the rate of gas flows.  Stainless




           steel type is recommended for S02«
       4.  Thermometer :  A laboratory  type  or  other  temperature-^raeasuring




           device is needed to measure with a  precision of 0.1°C or better




           the temperature of the constant-temperature bath and the zero




           air (carrier gas) flowing over the  permeation tube.









8.5.2  Gas Dilution Apparatus (when cylinder S(>2  is  used)









       •"••  Flowmeters ;  To measure the rate of S02 and zero air flow.




           Calibrate frequently  (at least monthly) as  indicated in 8.5.1,




           Item 1.









       2.  Needle Valves :  to control  gas flow rates.   Stainless steel types




           are recommended for S(>2.
                                     -176-

-------
       3.   Mixing Chamber:  A cylindrical Kjeldahl type connecting bulb of




           200 to 300 ml volume works well.  This can also be fabricated




           from borosilicate glass as shown in Figure 8-1.









       4-   Sampling Manifold;  Fabricate from borosilicate glass (see Figure




           8-1)N.  It should contain three or four ports to permit simultaneous




           sampling of the calibrating gas stream with the analyzer(s) and




           the referee method.









8.5.3  ZeroAir Source









       The zero air for diluting the calibrating gas should be free of




       S02 and substances that will in any way 1) change the calibrating




       gas concentrations, 2) interfere in the analyzer response and 3)




       interfere in the referee method.  The zero air may be furnished from




       a cylinder or by filtering ambient air as indicated in Figure 8-1.









       1.   Air pump (for transport, of ambient air) - A diaphragm or




           carbon vane pump capable of delivering flow rate reqxiireinents




           of the total generation system (5 to 10 liters/min) is needed.




           A particle filter should be installed on the downstream side of




           carbon vane pumps and is optional for diaphragm pumps.









       2.   Filters for Ambient Air;   See 8.6.4









8.5.4  Manual Sampling Train and Apparatus for Referee Analysis:  See Reference 4




       and Figure 8-II.
                                    -177-

-------
8.5.5  Analytical balance;  A laboratory type with a sensitivity of  10 yg


       or better is needed for weighing the permeation tube.





8.5.6  Calibrating Solution Dispenser (for static calibration):  See 7.5.9


       and Figure 7-III in Chapter 7.





8.5.7  Apparatus for static calibration;  refer to the manufacturer's in-


       structions for a list of the required apparatus.





8.5.8  Absorber (Impingers) ;  See Item 1 in 6.5.4 and Figure 6-III in Chapter 6





8.6    REAGENTS AND GASES





       Purity of chemicals - Unless otherwise specified,  all reagent


       specifications shall conform to the committee on Analytical Reagents

                                       o
       of the American Chemical Society .  When such reagents are not


       available, ascertain that they do not lessen the accuracy of the


       determination.





8.6.1  S02 Permeation Tubes•  Permeation tubes containing S02 are commercially


       available in a variety of sizes and permeation (emission) rates^>   .


       The rates may be either nominal or certified.  Certification can be


       done by the supplier at additional cost or by the user in his own


       laboratory.  To assure maximum reliability, these tubes should be


       weighed regularly (e.g., at least once a month) and just before use.


       Certified tubes can also be obtained from the National Bureau of


       Standards.
                                      -178-

-------
8,6.2  Cylinder S02 (50-100 ppm; 131-262 yg/m3):  This gas mixture is




       further diluted to produce the desired concentrations for calibration.




       The concentration produced should never be considered as primary




       standard and must always be standardized by the referee method or




       compared to the analyzer response obtained with a certified S02




       permeation device.









8.6.3  Zero Air:  A high-pressure cylinder of synthetic zero air, or




       filtered ambient air from an air pump may be used.









8.6.4  Zero Air Filter;  Activated charcoal and soda lime used together will




       remove residual SO? arid most interferents from the zero gas stream.




       Excessive amounts of nitric oxide in ambient air can be removed by




       placing a Cr03 oxidizer (Item 4 in 7.5.1) before the soda lime.




       Refer to Chapter 7 (calibration method for NO2 and NO) and Table 5-III




       for details.









8.6.5  Reagents for Referee Method for 502=  see References 4 and 7.









8.6.6  Reagents for Static Calibration;  refer to the manufacturer's operating




       instructions for a list of required reagents.









8.7    SPECTROPHOTOMETER CALIBRATION









       1.   Prepare a series of calibrating solutions containing the equivalent




           of  0.05 to 0.4 yl S02/ml (i.e.,  1,2,3 ml) as directed in the manual
                                    -179-

-------
           referee method^.  Treat the solutions (Ref 4) and wait 30




           minutes for full color development.  Determine the absorbance




           of the solutions on the spectrophotometer against reagent blank.









       2.   Plot the net absorbances on the vertical axis versus yl S02/ml




           on the horizontal axis of a rectilinear graph paper as a check




           for linearity.  Calculate the slope b of best-fit curve for the




           data using the method of least squares3.









8,8    DYNAMIC CALIBRATION








8.8.1  General









       1.   When a static calibration (8.9) is to be performed on the




           analyzer, it should be done before the dynamic to assure proper




           operation of its detection.and signal presentation components.




           The static and dynamic responses may be compared (reconciled)




           to verify the proper operation of the analyzer and the validity




           of the calibration.  See 8.10 for details.









       2.   When the analyzer has been operating as a continuous monitor,  it




           is useful to determine its response near the span level first




           without changing the span settings (auditing).  When the response




           is within ± 10% of the previous calibration, the calibration is




           still valid and a new calibration is not necessary.  When the




           response is greater than ± 10% proceed with the complete calibration









       aSee Chapter 4, Table 4-V




                                     -180-

-------
    The audit data provide a record of the calibration drift.




    Instruments with non-linear response require the full




    calibration.









3.  The analyzer to be calibrated should be in good operating




    condition and installed in accordance with manufacturer's




    instructions.  Operate the analyzer for at least 24 hr to




    warm-up.  This 24-hr warm-up period may be shortened if so




    stated in the operating instructions.  Adjust the air and




    reagent flowrates to their recommended rates or to the rates




    determined from the static calibration (8.9) data and verify




    the rates as described in 8.5.1, Item 1.









4.  Record all data only after stable analyzer response has been




    attained.  Refer to 4.1.6 in Chapter 4 for determining




    stable response.









5.  Newly prepared SC>2 permeation "tubes may be used after one day




    (24 hr) of equilibration at a constant temperature provided the




    concentrations produced are established by the referee method.




    A minimum of 30 days is required to establish that the emission




    rate is stable and before the emitted concentration can be used




    as a primary standard .  The tube should be equilibrated for




    at least 12 and preferably 24 hours whenever the temperature is




    changed by more than ± 5°C.  See Reference 7 for general pre-




    cautions pertaining to permeation tubes.  It is generally desirable
                              -181-

-------
          1 to precondition all 862 gas lines for % to 1 hour by  flowing  a




           stream of dilute S02 through them.
8.8.2  Procedure
       1.  Place the gas generation system as close as practical  to the




           analyzer to prevent losses and to minimize pressure changes in




           the analyzer sampling duct.  Calculate the airflow of  the




           analyzer and add the airflow needed for the referee analysis




           (1.0 to 2.0 liter/min).  Add about 10% of the total tp insure




           an excess.  (NOTE:  The excess calibrating gas should  be vented




           to a hood or absorbed by a soda lime trap to avoid exposure




           to personnel.)  In a proper assembly, connection or dis-




           connection of the analyzer sampling line should not alter the




           airflow settings.









       2.  Generate a flow of zero air equal to the rate determined in




           Step 1 above.  Pipet the required volume of absorbing  reagent




           in the impinger (Item 8.5.8) according to the manual referee




           method .  Connect the impinger to the sampling train as shown




           in Figure 8-III.  When a stable analyzer response is obtained,




           sample the gas stream from the manifold for 30 minutes at the




           rate directed in Reference 4.  Transfer the exposed solution to




           a 25 ml volumetric flask.  Develop the color of the solution in




           accordance with the referee procedure  and read the absorbance




           at 560 nm.  When the absorbance of the solution is greater than




           the blank, continue flushing the g^3 generation system until




           the absorbance is obtained the same as the blank.  Inability  to

-------
    obtain a reading equal to the blank indicates interferents in


    the zero air.  Correct the source of the problem(s) before


    proceeding.





    Zero the analyzer by adjusting the analyzer controls so that


    the output corresponds to zero or the desired reading.





3.  Generate an SCU concentration equal to 80 ± 5% (span gas) of the


    full scale reading (0.80 ppm for full scale of 1.0 ppm; 2.10


    yg SC^/nr* for full scale of 2.62 yg/nP) .  When the analyzer


    response is steady, record the analyzer reading.





4.  When a permeation tube system is used and has been shown by


    frequent referee analysis to provide reliable S02 concentrations;


    the collection of referee samples may be omitted.  The S02 con-


    centrations are calculated from the diluent gas flow and permeation


    rates by equation 1 or 2.
a)  To determine the concentration of SC^ in ppm (yl/1) fr


    the permeation tube emission rate, use equation 1.
                          p                          (1)
                ppm = 	                      x
                VV    2.62 (Qd)
    where  P = permeation rate in  yg/min.


          Qd = rate of zero  (diluent)  air in  liters/min.
                                                              om
                              -183-

-------
    b)  When the concentration units are desired  in  yg SC>2/!iP instead-


        of ppo> then:
               c. 3   (R X 103)
               S02/m  = -———	             C2)

                              (Qd>
5.  To establish the S02 concentration by referee analysis:





    a)  Collect duplicate samples of the gas stream and analyze


        as described in the referee method  .  Adjust  the  sampling


        period to keep the absorbances of the samples at  about


        midscale (0,3 to 0.5A) on the spectrophotometer.  Five to


        30 minutes are usually sufficient.  For maximum precision,


        place the flasks containing the reacted solutions in a bath


        maintained within 2°C of the temperature used during the


        development of the spectrophotometer calibrating  solutions.





    b)  From the volume of air sampled and  the slope  of the spactro-


        photometer calibration curve, calculate the ppm SOo as


        directed in equation 3.
                          (A) (10)
                    ppm =
        where:  ppm = concentration (yil/1) S02


                  A - net absorbance of the solution


                  b = slope of the spectrophotometer  calibration curve


                      obtained in 8,7
                             -184-

-------
                 Va = volumes in liters, of the gas sample collected




                      (liters/min X rain)



        To convert ppm to yg/m^, use equation 4
                yg S02/m3 = ppm S02 X 2620          (4)
        Usually, changes in the gas volume of the samples due to




        deviation from the standard conditions of 25°C and 760 torr




        are small and may be neglected.  When the deviations are




        large, (sufficient to cause a change in gas volume greater




        than about 5%) correction should be made.









    c)  The concentrations of the duplicate samples should be within




        5%.  Differences greater than 5% may be due to unstable gas




        concentrations, error or inconsistencies in sample collection




        and analysis.  Correct these problems before proceeding with




        the calibration.









6.  When possible, adjust the analyzer to give reading equivalent




    to the span S02 concentration (spanning).  When the instrument




    has no span controls, proceed to Step 7 below.  Generate zero air




    and note the analyzer reading.  When the reading is different




    from the original baseline reading by > 2%, reset the analyzer




    to read the original baseline and repeat Steps 3 through 6




    above.  Analyzers with zero and span controls not electrically




    independent may have to be respanned and rezeroed iteratively




    until the proper zero and span settings are obtained.






                             -185-

-------
       7.  Generate, in turn, four additional SC^ concentrations between




           the blank and span range (e.g., 10, 20, 40, and 60% of full




           scale)  and determine the S02 concentrations in duplicate and




           analyze as directed in Steps 3 through 5 above.  Determine




           the net analyzer readings by subtracting the baseline reading




           from the individual readings.









8.8.3  TreatmentofDynamic Calibration Data









       1.  Plot the net analyzer readings on the vertical axis versus the




           corresponding SO?, concentrations on the horizontal axis of an




           appropriate graph paper (rectilinear, semi-log, log, etc.).




           Calculate the slope b^ of the best-fit curve for the data by




           the method of least squares .  To determine the linearity of




           the response, see Test 4.5 in Chapter 4.  A non-linear response




           from an instrument normally linear indicates analyzer malfunction




           or, possibly, errors,in the preparation of the calibrating




           gases.   Correct the cause of the problem(s) before recalibrating.









       2.  When a non-linear response is normal, draw a smooth line through




           the calibration points that fits the points best.  From this,




           prepare a template to convert the net analyzer readings to S02




           concentrations.
            Chapter 4,  Table 4-V.
                                     -186-

-------
8.9    STATIC CALIBRATION
       The following section pertains to S02 analyzers using wet-chemical




       methods (colorimetry and conductimetry) .  See also 8.1.3.









8.9.1  Procedure









       Prepare calibrating solutions (equivalent to 10, 20, 40, 60, and 80%




       of full scale and perform the static calibration according to manu-




       facturer's instructions.  In the absence of such instructions the




       static calibration procedures described in Section 6.9, Chapter 6,




       for automated oxidants and ozone analyzers may be used as a guide.









8.9.2  Treatmen^ of_S ta/tic _Calib ration .Data









       1.  Plot the net analyzer readings on an appropriate graph paper




           (rectilinear, seial-log, etc.) against the. equivalent yl S02/ml




           of the calibrating solutions.  Calculate the slope bs of curve




           that best-fits the data by the method of least squares0.  To




           determine the linearity of the response, see Test 4.5 in Chapter




           4.  A non-linear response from an instrument normally linear




           indicates malfunction in the analyzer or error in the static




           calibration process.  Correct problem(s) before recalibrating.









       2.  When a non-linear response is normal, prepare a template as




           directed in Step 2 of 8.8.3.  An alternate method is to
       cSee Chapter 4, Table  4-V.



                                     -187-

-------
           attempt to linearize the instrument output by adjusting the




           electronics (i.e. photometer, etc.) until a  linear  output  is




           found over the range of pollutant concentration of  interest.









8.9.3  Determination of Airflow Rate









       For instruments not equipped with adjustable upper limit or  span




       controls or when the range of span adjust is insufficient,  the




       slope of the static calibration can be used to establish the




       sample airflow rate that will make the analyzer  output  correspond




       to the pollutant concentration or a simple fraction or multiple




       fs of the concentration range as follows:
                          «= (Qr)
       where:  bs'=. slope of the static calibration curve obtained in




                    8.9.2




               Qa = analyzer airflow rate, ml/min.




               Qr = analyzer reagent flow rate, ml/min.




               fs ~ range factor (e.g., %, 1.0, 2.0)









8.10   RECONCILING THE STATIC AND DYNAMIC CALIBRATIONS









       The static calibration slope ba (8.9.2) and the dynamic calibration
                                     s



       slope b, (8.8.3) are compared by equation 6:
                                    rl88-

-------
             |ts - bdj
        R «	X 100                          (6)
              b.
               s
Large values of R  (> 10%)  are  indicative  of  a)  error  in  the  analyzer's


air or reagent flowrate, b)  leaks  or malfunction  in the  analyzer


c)  poor quality reagents,  d)  error in the static or  dynamic calibration


process or e) change in  the sample collection efficiency of  the


analyzer.  Consult the analyzer  operating instructions and/or


manufacturer and correct the problem(s) before  recalibrating.
                              -189-

-------
8.11   REFERENCES.
       1.  O'Keeffe AE, Ortman GO:  Primary standards lor  trace  gas
                                                           >



           analysis.  Anal Chem 38:760, 1966.









       2.  Scaringelli FP, Frey SA, Saltzman BE:  Evaluation of  teflon




           permeation tubes for use with sulfur dioxide.   Amer Ind




           Hygiene Assoc J _28_:260, 1967.









       3.  Scaringelli FP, O'Keeffe AE, Rosenberg E, Bell  JP:  Preparation




           of known concentrations of gases and vapors with permeation




           devices calibrated gravimetrically.  Anal Chem _42_t871, 1970.









       4.  Environmental Protection Agency, National primary and secondary




           ambient air quality standards, Appendix A:  Reference method for




           the determination of SO^ in the atmosphere.  Fed Reg 36;  No. 84,




           Friday, April 30, 1971.









       5.  Environmental Instrumentation Group, Lawrence Berkeley Laboratory,




           Univ. of Calif., Berkeley, Calif.:  Instrumentation for environ-




           mental monitoring, Air, LBL-1, vol.1 (SO-,) Dec. 1, 1971.









       6.  Pate JB, Ammons BE, Swanson GA, Lodge JP, Jr:   Nitrite inter-




           ference in spectrophotometric determination of  atmospheric




           sulfur dioxide.  Anal Chem _3J7_:942, 1965.
                                    -190-

-------
7.   Intersociety Committee:   Methods of air sampling and  analysis.




     Amer  Publ Health Assoc,  1015 18th St.,  NW,  Wash.,  DC,  1972.









8.   ACS Reagent Chemicals,  American Chemical Society Specifications.




     American Chemical Society, Washington,  B.C.  For suggestions  on




     the testing of reagents not listed by the American Chemical  .




     Society,, see:  Rosin J:  Reagent Chemicals  and Standards.




     New York, D. Van Nostrand Co., Inc., and The United States




     Pharmacopoeia.









 9.   Metronics Associates, Inc., Palo Alto,  CA  94304.









10.   Analytical Instrument Development, Inc., 250 South Franklin Street,




     Westchester, PA  19380.








11.   National Bureau of Standards, Office of Standard Reference




     Materials, Washington, DC  20234.
                               -191-

-------
                Gas nixing
                   chamber
Sampling manifold
                                                                 Vent
                                                     r                i
                                                     { Referee sampling i
                                                          train       i
            Particle   Air pump    Filter
             filter
                                                           Ambient air
              Gas Dilution Apparatus
                                                     SOo Source Apparatus
           Permeation device
                            7
or
       Rotameter
 Dil,
 so2
 in
 N2

                              Thermometer-
                        Constant
                        temperature bath
                        (air or water)
                                                         Rotameter
   Figure 8-1.  Gas generating system for calibrating  SC^  analyzers,

-------
                                                Flexible tubing
Flowmeter
    From sampling
      manifold
                                                                            *• To air pump
                                                            Needle
                                                            valve
                   Figure 8-II. Sampling train for referee  S02 analysis,

-------
9.0    CALIBRATION PROCEDURE FOR AUTOMATED ATMOSPHERIC

        CARBON MONOXIDE ANALYZERS
9.1    PRINCIPLE AND SCOPE









9.1.1  This procedure is for the calibration of continuous atmospheric




       carbon monoxide (CO) analyzers that use physical measurement




       methods such as nondispersive infrared absorption, etc.  The




       calibration, known as a dynamic calibration, is performed by




       determining the analyzer response to a series of calibrating




       gas concentrations containing CO.  It is a performance test of

                              <~ •


       the entire analyzer under simulated service conditions.









9.1.2  The calibrating gas concentrations may be generated in several



           1 2
       ways ' .  The most coamon method is to prepare a series of CO




       concentrations in plastic bags (primary or batch dilution) or to




       prepare or purchase cylinders of compressed air containing CO




       concentrations between 10 to 80% of full scale (primary standard).




       A wider variety of concentrations may be generated by diluting




       50 to 100 ppm (57 to 115 mg/m^) CO from a cylinder with zero air




       (secondary dilution).









9.1.3  The instrument operating variables, such as airflow rates, output,




       gain, span, can be adjusted during calibration to make the output




       reading conform directly to concentration in ppm (yl/1) or mg CO/m




       or a simple fraction or multiple of the concentration.
                                    -194-

-------
9.2    RANGE



       The range of the calibration procedure depends on the measuring

       range(s) and sampling rate(s) of the analyzer(s).  The analyzer's

       measuring range and sampling rates vary greatly depending on its

       design and measurement principle (See Reference 2 and its lists

       of references for further details).  The commonly used nondispersive

       infrared (NDIR) analyzers typically measure in the range of about

       0 to 25, 50, or 100 ppm CO (0 to 28.6, 57.3, or 115 mg CO/m3).

       Sampling rates are typically between 1 to 2 1/min, but can go

       as high as 42 I/minute (See Table 5-1).  The semi-continuous gas-

       chromatographic CO and methane analyzers measure CO concentrations

       between 0 to 1000 ppm (0 to 1145 ing/iir) .  Electrochemical analyzers

       measure in the same range although they are not much used for

       continuous ambient monitoring.



9.3    INTERFERENCES



       Carbon monoxide is available with purities of 99.7% or greater.

       Higher purities are obtainable for special work.  Zero air for

       dilution must be free of CO, CC2 and water.  Selective, absorbers

       (drying agents, Ascarite for C02, etc.) can be used whenever a
                                                           i
       particular measurement principle requires it.  Details for pre-

       paring these absorbers are given in Table 5-III.
                                     -195-

-------
9.4    PRECISION, ACCURACY, AND STABILITY









9.4.1  A precision (coefficient of variation) from 2 to 3% is achievable




       when preparing repeated primary dilutions of CO in non-rigid, CO-




       inert bags in the range of 10 to 100 ppm (12 to 115 rog/m3).  The




       chief source of error is often in measuring the air volumes.  A




       minimum of 2% error is produced for each rotameter used and the




       errors are approximately additive.  When greater precision is




       needed, positive displacement, metering pumps can be used.









9.4.2  When spanning is possible, any discrepancy between the input and




       analyzer output may be resolved by adjusting the instrument output




       to correspond to the concentration of the calibrating gas.  Where




       analyzers have no spanning controls, a correction factor may be




       calculated to convert the analyzer readings to CO concentrations.









9.4.3  Primary dilutions made in non-rigid containers usually cannot be




       kept for more than a few (< 5) days due to concentration changes




       with time.  The concentration of dilute CO (> 50 ppm)  stored at




       high pressure in steel cylinders lined with chromium molybdenum




       alloy should be checked periodically (^ monthly) for one to three




       months after preparation and semi-annually thereafter.  See 9.6.2




       for additional comments.









       A detailed discussion of the various sources of error in the




       preparation of calibrating gases is given in Reference. 1 in
                                    -196-

-------
      Part I:   General Precautions and Techniques.  The minimization

      of  the sources of error is important to assure high accuracy

      and precision in calibration.



9.5   APPARATUS



      The apparatus needed depends on the method chosen.  The procedures

      listed under 9.7 are the easiest as well as among the most reliable.

      .For delivering large (> 40/min) flows, procedure 9.7.3 is preferable,

      The fourth, indicated in 9.5.4 and discussed in Reference 1, is less

       commonly used, although equally reliable.  The zero air may be

       furnished from a cylinder or by filtering ambient air as shown

       in Figure 9-1.



9.5.1  Primary Standards
       Eefer to 9.6.2 for equipment and gases.
9.5.2  Primary Dilution in Bags



       1.  Bagsi	(Scotchpaka)i_:  Fabricate or purchase six bags of sufficient

           volume (10 to 20 1) to satisfy the volume rate requirement of

           the analyzer.  Use one bag for each gas concentration.



       2.  Dilution system (Figure 9-1):  consists of a flowmeter for

           measuring gas flows between 1 to 5 liters/rain, a needle valve,


       Registered trademark of Minnesota Mining & Manufacturing Co.  C3M)
        for their metallized laminate consisting'of polyester, aluminum,
        and polyethylene, Cat. No. 20A20.


                                    -197-

-------
           a three-way stopcock for diverting flows, a tee connection with




           a rubber septum cap for injection of known volumes of pure CO




           and the necessary glassware and ball-and-socket joints.  Poly-




           vinylchloride tubing may be used to butt-join connections.









           The rotameter should be calibrated frequently (monthly) with.




           a wet or dry test meter, soap bubble meter or calibrated




           rotameter.








       3.  Gas Sampling syringes";  A 1.00 ml syringe is needed for pre-




           paring the  CO concentrations in bags.








       4.  Stopwatch or accuratetiming device.









9.5.3  Secondary Dilution of CylinderCO









       1.  Dilution system (Figure 9-II);   This consists of a mixing




           chamber, two flownieters capable of measuring the maximum




           flow required by the analyzer(s), two needle valves for flow




           control and a delivery manifold containing sampling ports




           and a vent  for discharging excess calibrating gas and the




           necessary glassware and ball-and-socket joints.  Rotameters




           should be calibrated frequently (monthly) with a wet or dry




           test meter, soap bubble meter or calibrated rctameter.











        Precision Sampling Corp., Baton Rouge, LA, Series A, or equivalent.
                                     •198-

-------
      2.  Air pump  (for transport of jugbieni^air) :  An oil-less (carbon




          vane  or diaphragm)  air pump  capable of delivering the flow




          rate  requirements of the total calibration system plus 10%.




          A particle filter should be  installed downstream of carbon




          vane  pumps and is optional for diaphragm types.









9.5.4  C0_ln. Cylinders (Low Pressure);   See Reference 1 for apparatus




      and method.
9.6
REAGENTS AND GASES
9.6.1  100%  Carbon Monoxide,  Chem.  Pure (> 99.7%):   For preparing CO in




       bags  by Method 9.7.2,  Primary Dilution.









9.6.2  Dilute CarbonMonoxide (Primary Standards):   High pressure steel




       cylinders containing CO concentrations equal to 10,  20,  40$  60 and




       80 ±  5% of full scale in air.  Cylinders should be lined with a




       chromium molybdenum alloy.   Alternatively,  the 80% of full scale




       cylinder may be used for spanning and for preparing  intermediate




       Concentrations by Method 9.7.3 (Secondary Dilution).  These




       mixtures may be purchased or may be prepared in the  laboratory




       with  suitable gases and high-pressure manifold .  The CO concen-




       trations are established by  comparing the analyzer readings




       obtained with this gas versus those obtained with 1) a primary
                                    -199-

-------
       dilutions 2) certified CO ga.s from another source, or 3)  analyzed


       -gravimetrically^'^.  Certification or analysis by the supplier


       at the time of preparation is of little value because of  unpre-


       dictable changes that frequently occur with time.  Verify the


       concentration monthly for one to three months after purchase or


       preparation, and occasionally (setni-annually) thereafter,  NOTE:


       Since changes in the CO concentration may occur after the  cylinder

                                                      /                  
-------
9.7    PREPARATION OF CALIBRATING GASES









9.7.1  Primary Standards









       Equip the zero air cylinder (Item 9.6.3) and each CO cylinder




       (Item 9.6.2) with an appropriate pressure reducing regulator, metering




       valve and flowmeter.  These gases inay be fed directly to the analyzer.









9.7.2  Primary or Batch Dilution in Bags









       1.  Assemble the dilution system described in Item 2 of 9.5.2




           and shown in Figure 9—1.  Generate a convenient flow of




           zero air (i.e.s 1000 ml/min) .  From the volume of the bags




           calculate the volume of pure CO (Item 9.6.1) needed to prepare




           the desired CO concentration by equation 1.  The volume of




           diluent zero air depends on the size of the bag and should




           not exceed about 80% of the bag volume.
                   v0 = i x io-6 (vx) (cx)                 CD
           where:  V0 = volume of pure CO in ml




                   Cj = ppm CO desired




                   Vj_ - volume of zero air and CO added to the bag in ml.
                                     -201-

-------
                                 *5
    To convert ppm CO to mg  CO/irr:





            mg C0/m3 -  (1.15)  (ppm  CO)               (2)





    Usually, the change in gas volumes  due to deviation from the


    standard conditions of 25°C  and 760 torr is small and may be


    neglected.  When the deviations are large (sufficient to


    cause a change > 5% in gas volume), volume corrections should


    be made.





2.  Connect the bag (with its valve open)  to the three-way stopcock.


    Carefully fill the  gas-tight syringe to the calculated volume V0


    tar? f-Ti r>nri=> C.Q.  KtjT.t"rV» fh.p stonrnrk  to fill thp. h^f? with zero
    ..«..-.._ ^-.__- —.  -  _ .  —  - .. « „   __Jt__-    _   .          ^


    air and begin timing.  Insert the  syringe needle through the


    rubber septum cap on the glass  tee  and expel the pure CO so that


    it merges with the  zero  air  stream.  Continue filling the bag


    until the predetermined  volume  Vj  is reached, then switch the


    three-way stopcock  to vent.  Seal  the valve on the bag and knead


    the bag to mix tha  gases thoroughly.  Empty the contents to a


    hood and repeat Step 2 twice.   Prepare the four other intermediate


    CO concentrations in the same manner.   These primary dilutions nay


    be prepared in duplicate or  larger  multiples and the analyzer


    responses averaged  to insure against errors or to acquire


    precision data.
                               -202-

-------
9.7.3
       1.  Assemble  the  dilution system (Item 1 of 9.5.3) as shox^n in

           Figure 9-II.   Connect the cylinder of dilute CO (Item 9.6.2)

           and the zero  air  source  (Item 9.6.3) to the needle valves of

           their respective  rotameters.



9.8    DYNAMIC CALIBRATION



9.8.1  General



       1.  When the  analyzer to be  calibrated has been operating as a.

           continuous monitor it is useful to determine its response

           near the  span level first without changing the span setting

           (a'.'dl fi/OP1) -   Wh#?t» th«? ?Pfl 1 'v?.(?-Y TP-ftponse is within ± 10% of

           the previous  calibration, the calibration is still valid and

           a new calibration is not needed.   When the response is

           greater than  ± 10% proceed with the complete calibration.

           The audit data provide a record of the calibration drift.

           Instruments with  non-linear response requires the full

           calibration,



       2.  The instrument to be calibrated should be in good operating

           condition and installed  in accordance with manufacturer's in-

           structions.   Operate the analyzer for at least 24 hours to
               •i
           warm-up.  This 24-hr warm-up period may be shortened if so

           stated in the operating  instructions.  Adjust the analyzer
                                      -203-

-------
    airflow to the recommended sampling rate and verify the rate




    as described in 9.5.2s, Item 2.









3.  Record all data only after stable, analyzer response has been




    attained.  Refer to 4.1.6 in Chapter 4 for determining  stable




    response.









4.  Assemble the appropriate generating system (.Items 9.5.1, 9.5.2 or




    9o5.3)for generating CO as needed.  Place the system as close as




    practical to the analyzer to prevent losses and to minimize




    pressure changes in the analyzer sampling duct.  Calculate




    the airflow of the analyzer and, when required, add  the airflow




    seeded for the referee analysis.  Add 10% of the total  to insure




    an excess.  (NOTE:  Any excess CO gas streams should be vented




    or passed through a CO pxidizer/C02 absorber filter  to  avoid




    exposure to personnel.  See Table 5-III for details.)   In a




    proper assembly, connection or disconnection of the  analyzer




    sampling line should not alter its airflow reading.









5.  Generate a flow of zero air equal to the rate determined in




   ' Step 4 above.  When a stable analyzer reading is obtained,




    adjust the analyzer zero control so that the analyzer output




    reads zero or the desired baseline reading.









6.  Generate a flow of calibrating CO gas equal to SO ± 5%  of. full r




    scale (span gas).  When a stable response is obtained,  adjust
                              -204-

-------
          the analyzer controls to give a reading equivalent to  the




          span. CO concentration.  Generate a zero gas.  When the zero




          reading is different from the original baseline reading by




          > 2%, reset the analyzer to read the original baseline and




          repeat Steps 5 and 6.  When the span and zero controls are




          not electrically independent, it may be necessary to rezero




          and respan iteratively until the proper zero and span control




          settings are obtained.









      7.  In turn, generate and introduce the four Intermediate CO




          concentrations (e.g,, 10, 20, 40 and 60% of full scale)




          from the bags or from the dilution system and record the




          corresponding instrument readings.  Determine the net




          analyzer readings by subtracting the baseline reeding




          from the individual readings.









9.8.2  Treatment of Dynamic Calibration Data









      1.  Plot the net analyzer readings on the vertical axis versus




          the CO concentrations on the horizontal axis of an appropriate




          graph paper.  Calculate the slope b of the best-fit curve




          for the data by the method of least squares0.  To determine




          the linearity of the response, see Test 4.5 in Chapter 4.  A




          non-linear response from an instrument normally linear indicates
      C8ee Chapter 4, Table 4-V.
                                    -205-

-------
    malfunction or, possibly, errors in the concentrations of the




    calibrating gas.  Correct the problein(s) and recalibrate.









2»  When a non-linear response is normal, draw a smooth line through




    the calibration points that fits the data best.  From this,




    prepare a template to transform the net analyzer readings




    to CO concentrations.
                            -206-

-------
9.9    REFERENCES
       1.  Intersociety Committee:  Methods of Air Sampling and Analysis.




           Tentative Method of Preparation of Carbon Monoxide Standard




           Mixtures, No. 117, Amer Publ Hlth Assn, Wash., DC (1972).









       2.  U.S. Dept. of Health, Ed. & Welfare, Publ. Hlth Serv.,  Envir.




           Hlth Serv.:  Air Quality Criteria for Carbon Konoxide,  Ch.  5,




           NAPCA Publ. No. AP-62 (March 1970).









       3.  Belsky T:  Preparation of low-concentration mixtures of gases.




           Air & Industrial Hygiene Lab., Cal. State Dept. of Publ. Hlth,




           AIHL Report No. 117  (Jan, 1972).
                                    -207-

-------
                       IT 100% co
                            Source
                   Zero
                   air
                                                   Flexible (Scotchpak)
                                                     bag
Figure 9-1.  Primary dilution  system for CO in bags.

-------
10.0   AIR ANALYZER TERMINOLOGY









       This chapter contains selected terms commonly used in air monitoring




       practices and instrumentation and was developed to promote more  effectiv




       communication between the various disciplines in air pollution techno log




       It is not intended as an all-inclusive glossary of air pollution terms,









       The emphasis of terms related to instrument performance applies  to the




       overall analyzer system and not to the individual components.









       For visual representation and interpretation of noise a.nd of the various




       time delays in instrument response, see Figure 10-1.
                                      -210-

-------
Absorbent;  a.medium, which  absorbs  the  pollutant being measured.









Absorber  (contactor,  contact'column,  scrubber):   a device  for bringing a




gas stream containing vapors,  fumes  or particulate matter  into intimate




contact with a liquid absorbent  usually  aqueous.   To minimize absorption




of interferents.  an absorber design  with minimum  turbulence and minimum




surface area consistent with absorption  of  the pollutant is desirable,









    Bubbler:  an  absorber  in which the sample gas is introduced below




    the surface of a  liquid  absorbent.   For increased efficiency, the




    gas stream may be broken up  into small  bubbles by being forced




    through restricted openings  as in a  perforated sparger or a fritted




    glass tip.  Bubblers may be  tall or  short depending on the efficiency




    of the absorbing  process.









    Helical ^coluran;   an absorber consisting of a  coil of tubing with one




    or more turns, usually mounted with  its axis  vertical.  Gas-liquid




    flow  is co-current and may be up or  down the  column.  High wetted—




    surface area  is achieved by  selecting a tube  diameter  sufficiently




    small to insure that the surface tension of the absorbent will cause




    the entire inner  surface of  the  column  to become wetted.  The coil




    is usually glass, but  other  materials may be  used.









    Helix insert  column:   a  high-surface-area absorber composed of a




    narrow vertical tube with  a  solid helix, usually of glass or wire,




    placed axially and in  contact with the  interior column wall.  Liquid
                                 -211-

-------
absorber passes down the column in a. thin  film over  the helix.  Gas-




liquid flow may be co-current or countercurrent.   The helix may also*




be wrapped around a center support with the  gas-liquid mixture passing




between it and the column wall.








Impjlnger;  an absorber in which sample air is  blown  or aspirated onto




or below the surface of the liquid absorbent.   Absorption  is increased




by both high surface area and by turbulence.









Mechanical abs orber:  an absorber designed to  bring  a gas  or gases and




liquid absorbent into intimate contact by mechanical agitation.  Absorp-




tion is increased by both high surface area  and by turbulence.









Packedcolumn;  a high-surface-area type absorber, usually a vertical




tubular column loosely filled with an inert  medium such as Raschig




rings, helices, crushed or sintered glass, beads or  other  materials




of various shapes.  Gas and liquid flow may  be co-current  or counter-




current with absorption taking place on wetted surfaces.   A single




or a series of sintered glass plates through which the gas-liquid




mixture is passed may be used in place of inert packing.









Spray jet:  an absorber in which a liquid absorbent  is forced through




an orifice at high velocity and impacted on  the inner surface of the




absorbing chamber.  Gas is absorbed through  mixing and turbulence




created by the impaction.  Kinetic energy is imparted to the liquid




by hydraulic pressure, sample air pressure,  or by  aspiration caused




by the airflow into the chamber.
                             -212-

-------
Absorption;   a process  in which one or more pollutants are transferred




from a  gas phase  to  a liquid  or solid phase, either by physical  dissolution




in, or  preferably by chemical reaction with the absorbent.   The  driving




force is a function  of  the pollutant partial pressure  at  the gas-absorbent




interface.  The process is favored  by high interfacial areas,  turbulence,




and high diffusion coefficients and, where the  absorption process is




chemical, fast reaction rates and irreversible  reactions.









Absorption spectroscopy:  see detection methods.









Accu_racy_:  (not to be confused with precision)  the  deviation between the




analyser output in concentration units and true pollutant concentration




which has been established by an accepted  reference procedure.  Lt is
expressed  as  a percent  of  the  true 'concentration.
Accuracy,  calibration;   deviation  between  the slope b^ of the curve




obtained by  the  manufacturer's  calibration procedure and slope obtained




with calibrating gases  expressed as- a percent of the manufacturer's




slope.








Accuracy,  monitoring_:   deviation between the analyzer response obtained




under actual monitoring conditions to that of a reference procedure when




samples of the same  atmosphere  are taken at approximately the same time,




It is expressed  as a percent of the reference procedure.









Adsorption;  a process  in which one or more pollutants are transferred from




a fluid to a solid by physical  binding to  the surface of tlie solid.  Adsorpti.o:
                                 -213-

-------
Is usually reversible, the pollutant being released  (desorbed)  by methods




such as solvent elution} evacuation, or heating.









Amperogie.try:  see detection methods.









Analy_zer (also monitor) :  a device or an instrument  for  sampling  and




assaying pollutants.









Analyzer deadband:  see deadband, analyzer.








Anemometer:  an instrument for measuring and indicating  wind  speed.









Aud_ib1e noi s e (also acoustical noise) :  1) any unwanted  sound,  2)  an




erratic, intermittent or statistically random acoustical oscillation;




commonly expressed in units of power (decibels) sound pressure  (newton




or dyne/square area) or loudness (phones, soness noys).









Auxiliary equipment:  accessory items designed but not normally required




for use with an operational analyzer.









Bubbler_:  see absorber.









Calibra_ting gas_:  zero air or zero gas containing a  known  concentration




of a pollutant and used for the calibration of an analyzer.









Jga_Mbratingsolution;  a solution used for static calibration, of  an analyze.




and containing a known concentration of a substance  which  gives a response




equal to that of a known pollutant concentration,




                                -214-

-------
Calibration:  the  determination of the analyzer response when a series




of calibrating gas concentrations  are introduced to the analyzer inlet




or artificial stimuli  are  presented to the detector.









    Cali'braetion_8_jdynaiai_c;   a, performance test of the  entire analyzer




    under simulated service conditions in which the response to a




    calibrating  gas over a known concentration range  is determined.




    When reconciled with a static  calibration, dynamic  calibration




    also serves  to verify  1)  the correctness  of reagent and sample




    air flow rates, 2)  the efficiency of sample collection,  3)  the




    integrity of the analyzer's plumbing and  4)  the quality of  any




    reagents and/or reactaiits.









    Calibration, static:   the determination of the  analyzer response when




    artificial .stimuli such as  standard calibrating solutionss  resistors,




    screens, optical filters, electrical signals, are applied directly




    to the  analyzer detector.  It  is a performance  test of  the  detection




    and signal presentation components of the instrument and is primarily




    applicable to  analyzers using  colorimetric and  conductinetric detectors,









_Chemilumines cence-  (see  chemiluminescent under detection methods):  an




emission of light  during chemical  reaction.








Collection efficiency:  the amount of pollutant measured divided by the




amount sampled, usually expressed  in percent.









Colorimetry:  see  detection methods.

-------
Conduct im e t r ;  see detection methods.
Contac t column :   see absorber .
Contactor :  see absorber.
Correlation spectrometry :  see detection methods
          :  see detection methods.
         Orifice :  see orifice
Deadb and , Ana 1 y z er ;  the range , in percent of full scale, through which




the input pollutant concentration may be varied without initiating an




analyzer response greater than twice the noise.  (Deadband is commonly




applied to servo-systems such as recorders, but can be applied to




continuous analyzers as well.)








Detection methods:  the techniques used to detect and assay pollutants




either directly through physical measurement or indirectly through measure-




ment of a reaction product.  Inherent in most detection devices is a




comparison system in which the output from the detector is, at some




convenient stage, compared to or balanced against a reference quantity




such as a known voltage or current, or a color intensity of a reference




solution.
                                -216-

-------
Absorption spectroaetry (ultraviolet, Visible, infrared, and microwave) :

the  detection and assay of pollutants based on the selective absorption

of electromagnetic radiation by the pollutants.


Amperoiaetry (not to be confused with coulometry) :   The detection and

assay of pollutants based on the measurement of an electrical current

which is the result of a pollutant reacting in an  electrochemical cell.

The  potential and mass transport conditions in the cell are usually

adjusted such that the magnitude of the current between electrodes

immersed in the electrolyte is proportional to the pollutant concen-.
tration.

                     \
' Chemiluminescent Method:  the detection and assay  of pollutants  based

on the emission of light by pollutants during chemical reaction.   The

intensity of the light produced is proportional to pollutant concen-

tration .


Colorimetry:  the detection and assay of pollutants based  on the reaction

with a solution to produce a colored product having an intensity (or

absorbance) proportional to the pollutant concentration.


Conductimetryr  the detection and analysis of pollutants based on its

reaction with a solution to produce a change in conductance that is

proportional to the pollutant concentration.
                             -217-

-------
           »,         :  ':'ne detection and analysis of pollutants by




direct comparison and correlation of their spectra with that of known




substances „









Coulometry (not to be confused with amperometry) :  the detection and




assay of pollutants by an electrolytic method based on Faraday's Law




wherein the quantity of charge (integral of current x time) required




for the reaction is proportional to the pollutant concentration.









Emission spec trome try :  the detection and measurement of pollutants




based on the characteristic spectra emitted when pollutants are excited




by heatj radiation, electrical discharge or other stimuli.
                   the detection and measurement of pollutants based
on the characteristic spectra emitted when pollutants are excited in




a flame.  Narrow band-pass optical filters are frequently used to




isolate a characteristic emission line when -measuring a single




pollutant.









Fluor ime try;  the detection and assay of pollutants based on the




characteristic electromagnetic radiation (usually visible light)




emitted during exposure of pollutants to radiation having a wave




length shorter than that emitted (usually visible or ultraviolet




light) .









lonization detector (flame and radiogenic) :  a device which detects




ions produced by passage of a substance through a flame or a region
                             -218-

-------
    of radioactive emissions.  The flame ionization detector is specific




    for combustible organic compounds.  The radiogenic detector say be




    a highly specific type such as electron capture that responds to




    electrophilic compounds (e.g., pesticides) or nonspecific such as




    helium ionization that responds to all gases.









    Nondispersive infra_red_absor£tion (NDIR) :  the detection and measure-




    ment of pollutants based on their absorption of infrared radiation




    generated in mixed wavelengths.  Spectral specificity is obtained by




    either a selective detector, optical filters, gas filled interference




    cells or by pretreating the sample.









    Potentiometry;  The detection and assay of pollutants based on the




    measurement of the potential of an electrochemical cell under




    conditions where the cell potential is determined by the. concentration




    of the pollutant absorbed.  With unpolarized electrodes, the electro-




    motive force is usually a linear function of the log of the pollutant




    concentration in the electrolyte surrounding one of the electrodes in




    the cell.








Detector (also:  sensor, transducer):  a device which responds to an input




parameter that is a property of, or an effect caused by, a substance




being measured.  The amount of the substance present is indicated as the




deviation of an output parameter from a reference value.  The input para-




meter is often light or other electromagnetic radiation, electrical current
                                 -219-

-------
or voltage.  The output parameter is frequently voltage, current  or resist-

ance.  Typical detectors are photocells, infrared bolometers  and  flame-

ionization detectors.


Drif_t-i- ca lib ration:  the change in analyzer calibration which occurs

during a stated time period, usually one week, expressed as percent  of

full scale.


Drift at span:  the deviation in analyzer output during a stated  time

period, usually 24 hours, of unadjusted continuous operation  when sampling

a span gas concentration equal to some upscale value, usually 80% of

full scale.  It is 'expressed as percent of full scale.


Drift^at zero:  the deviation in analyzer output during stated time  period,

usually 24 hours, of unadjusted continuous operation when sampling zero

air.  It is expressed in percent of full scale.


Dry-gas iseter (bellows or diaphragm meters) :  a device which  measures the

total volume of a gas passed through it without using liquids.


Emission spectroscopy;  see detection methods.

                               \
Fall time., tf'.  see response times.


Flame photometry:  see detection methods.
                                -220-

-------
Flow _ggntr_ol__d_ev 1 c e_s^ :   apparatus for regulating fluid flow, usually by




maintaining constant pressure differentials across in-line flow con-




strictors .
           ing ....... and metering_devic_e_s_:  apparatus used to measure and meter




fluid flow.  Most devices determine flow either by -volume displacement or




by methods based on  the relationship between the pressure differential




across an orifice or constriction and the rate of fluid flow through that




orifice.  Some devices determine total mass transfer by an electrodynamic




method.  Examples of devices using these principles are:









1)  Volume measuring devices - wet test meters, dry test, meters,  bellows




    meters,  diaphragm meters, etc.








2)  Rate measuring devices - rotameters, orifice meters,  venturi tubes,




    pilot tubes j  catharometers , anemometers, volume measuring devices




    used in  conjunction with time measurement.









Flowmeter (see also  rotameter) :  any device which measures a rate of




fluid flow.  In  common practice, the term generally refers to a rotameter.









Fluor ime try:  see detection methods.








Full__scale_  (see  also range) :  the maximum pollutant concentration measure-




able for given range, expressed in units such as yg/m , ppra, }il/l, %.
                                 -221-

-------
Helicalcolumn:  see absorber.




Eelix _insert^c_oltmn;  see absorber.




Humidity operating range:  see operating  humidity range.




_Imgiriger_:  see absorber.




jngtrjgnent inlet:  the opening through which air sample enters the analyzer,
          
-------
    time (initial response  time)  t^j   see response times.
Linearity :  a concept which  expresses  the analyzer  response  as  a mathematical




function of concentration over  a specified range.   For  a linear relationship




this function is  the equation of a  straight line.   A non-linear response




pertains when the deviation  of  calibration points about a straight  line of




best fit suggests non-random scatter and the maximum deviation  is greater




than a specified  percent  of  full scale.









Lower detectable  limit  (not  to  be confused with  sensitivity) :   the  smallest




pollutant concentration XvThich produces a signal  equal to twice  the  noise.




It is usually expressed in concentration units.









Manufacturer :   the company or organization which assembled,  fabricated or




otherwise produced the  instrument.









Mechanical absorb_e_r_:  see absorber.








MobjLlity:  a qualitative  term used  for describing the relative  ease with




which an analyzer may be  moved  about.  It encompasses the  categories:




portable, mobile, and stationary.   All analyzers can be used as stationary




units.  Mobile  and portable  analyzers should have qualities  for 1)  with-




standing the rigors of  vibration and acceleration such  as  occurs during




transport, 2) tolerating  extremes in environmental  temperature variations,




and 3) warming-up in less than  one  hour.   Portable  units are expected
                                 -223-

-------
in addition to be lightweight, self-contained, and independent of  external




utilities such as power, gas, vacuum, and reagents.   Instruments intended




for mobile operation should be able to meet all performance  specifications




while in motion.









Model;  an identifying name or alphanumeric code other than  the trade  name




assigned by the manufacturer to identify the instrument.









Monitor:  sae analyzer.









Noise:  unwanted, spontaneous, short term variations  in analyzer response




about the mean output not caused by variations in pollutant  concentration.




It is usually expressed as percent of full scale (see Figure 10-L).









Koise;  audible:  see audible noise.









Nondispersive infraredabsorption (NPIR):  see detection methods.









Operating humidity range:  the range of ambient relative humidity  (RH) over




which the instrument will meet performance specifications.









OjH5rating__teirperature range:  the range of ambient temperature over -which




the,instrument will meet stated performance specifications.









Operating vibration range:   the range, in type and intensity,  of mechanical




vibration in units of distance/time^ (cm/sec^) ever which  the  instrument
                                  -224-

-------
will  meet performance specifications.  The emphasis is on analyzers designed




for portability in field use in motor vehicles and aircraft.









aperating_voltage_ranget  the range of power line voltage over which the




"instrument will meet performance specifications.









Operational^period^  an indication of the time during which a properly main-




tained instrument can be expected to function without failure.  This -nay be




expressed as a ratio between a specified operating time and the corre-




sponding off-stream, including maintenance, time during that period; as a




percent "of on-time during a specified operating period; or  as an average




time  between failures (mean failure rate) determined by dividing a  specified




total operating period by the number of failures which can  be expected to




occur during that period.








Orifice:  an opening through which a fluid can pass and which is shaped




to provide pressure-flow characteristics that can be translated to  flow




rate.  At rates approaching sonic velocities, the characteristics are




useful for controlling flow (critical orifice).









Output;  the final signal, usually electrical (volts, ohcs, amperes), pro-




duced by the instrument after detection of the pollutant being measured.




The signal is a function of the concentration of the pollutant.  The




function may be linear; if non-linear, it is usually exponential but,




occasionally it is undefined.  (see Lineaiity)    The output is -usually-
                                 -225-

-------
fed to a signal presentation device such as a meter, chart  recorder,  or




data processing device (see signal presentation devices).









Packed column:  see absorber.









pollutant:  the substance in the atmosphere which the analyzer is




designed to detect.









Portable:  see nobility.









PotentiOTietry_:  see detection methods.









Precision (not to be confused with accuracy) :  the variability in repeated




measurements of the same quantity (e.g., pollutant concentration) expressed




as the coefficient of variation, i.e., the standard deviation of the




individual results expressed as a percent of the mean.









primary standard:  a substance with a known property which  can be defined,




calculated or measured, and which is readily reproducible.  The standard




may be traceable to the National Bureau of Standards or  other accepted




standards organizations.









_Pulse_tirae_:  see response times.









Range (see also full scale.) :  the nominal minimum and maximum concentrations




which the instrument is capable of measuring.  Many analyzers provide







                                -226-

-------
multiple range  selection capability for greater accuracy and ease of




Interpretation.   It is  specified by stating the lower and upper pollutant




concentrations  that can be measured as for example 0 to 1 ppm, 0 to 5 ppm.









Readout :   the presentation of the instrument output in an observable




display, as  for example, with meters, printed numerals or words s




light,  stripcharts, etc.









Resolution:   the ability to separate two closely spaced events in




space  or time at a signal-to-ncise ratio of two expressed as percent




of  full scale.









Resp_onse___t_iinea_:  (see Figure 10-1) .
       ime,  t*:   the interval between the time to -100% Ct_-j_Q0}  and
 the lag time (t-j_) «   Fall time is not necessarily equal to rise time.
                     'f " ^100
 Lag time_ (initial response tine), t^:  the interval between the time




 t0, when a step change (increase or decrease) in pollutant concen-




 tration is made, to the time t^ when the instrument indicates a




 response equal to twice the noise.
                             - t0
                                 -227-

-------
_Pulse___tlme_:   the  minimum time a pollutant concentration must persist
for the analyzer  to  register a peak response equal to the pollutant
concentration (see Figure 10-11).


Rise; time,  tr:  the  interval between the time to 100% (tjj}()) an
-------
Rise timg_, tr:  see  response  times.









Rotameter (variable  area  f lowmeter) :   a flow-measuring  device  that




operates at constant pressure and  consists  of  a weight  or  float




in a vertical tapered  tube.   As  the flow rate  increases, the float




moves upward to a new  equilibrium  position  in  a region  of  larger




area, the position being  related to flow rate.









Sample ;  a representative specimen of  air collected for the purpose




of determining its pollutant  content.









Sample_,_ grab ;  a single sample rapidly collected when expedient.




Gaseous grab samples are  usually collected  in  a container  or




absorbed on a solid  or in a solution within one jninute.









Sample, integrated:  a composite of a  series of samples or a




continuous flow of sample collected over a  finite time  period and




representing an average sample for that period.  Although  losses




may occur, an integrated  sample  is often stored for some time




before analysis.








S amp ling, au t oma ted ;   programmed automatic  collection of samples




for later detection  and analysis.
                        1   collection  of  a  composite  sample durin
a predetermined  total  time  or until  a predetermined  total volume
                                 -229-

-------
has been reached, comprising uniformly timed samples  separated




by regularly timed intervals.  The ratio of on- time versus  off-




time is stated.
                 :  collection of grab or of integrated  sampled  at
random intervals.









Samp ling , s e q'u ent ia 1 :  collection of uniformly timed samples  that follow




each other at regularly timed intervals.









_Scrubb£r_:  see absorber.









Secondary standard:  a substance having a property which is calibrated




against  a primary standard, to a known accuracy.









Selectively permeable membrane:  a film which permits certain pollutants




in an air stream to diffuse faster than others.  It is used for




separating or concentrationg those pollutants for subsequent  analysis.









_Sensj.tivity (not to be confused with lower detectable limit) :   the




instrument signal output response per unit of pollutant concentration.









Signal presentation devices;  devices that are used to convert  the




analyzer output (the signal from the detection device) to a usable




form, such as meters, recorders, digital readout units, computers,




and other data acquisition systems (see output).
                               -230-

-------
      -1o-noise  ratio:   the  ratio of  the  magnitude  of  the  input




signal to that of  the analyzer  noise  expressed  as percent  of full




scale.  A ratio  of two  is  usually the smallest  value that  can be used




with confidence.









J?£gCJL requir emen ts_;  the area required for  safe installation, operation




and maintenance  of an instrument  including  allowance for all auxiliary




items and equipment  such as  pumps,  reagent  containers, etc.s and




swingout space for cabinet access doors and panels.









Span gas;  a calibrating gas containing a pollutant concentration




equal to seine up scale  value, usually 80% of  full scale.









Span solution:   a  liquid reagent  used in  the  static calibration of




an analyzer containing  a concentration of a substance which gives a




response equal to  some  known pollutant concentration, usually 80%




of full scale.
Spanning:  adjustment  of  the  instrument  output  to a selected readin




when sampling a span gas  or when  the  detector, is subjected to an




artificial stimulus such  as a span  solution,  resistor,  screen,




optical filter, or electrical signal.









Spray jet:  see ab s or b e r.
                               -231-

-------
          :  see
S_t o ck s o lut 1 on :  a solution which contains a known concentration of




a standard.  This solution may be diluted to provide convenient con-




centrations for use in analysis.









Telemetry_:  the conversion of the instrument output into a standardized




signal s and transmission thereof to a distant location for interpretation,
                          e:  see operating temperature range.
Time to_95%, tnci  see response times
Timejuo	-95_%, t_q~:  see response times.
Trad_e_jname;  the name used by the manufacturer or \rendor to identify




a particular instrument or series of instruments.









Unattended operation:  the period of time during which the instrument




can be expected to operate unattended within specifications.









Vendor:  the independent dealer or distributor of the instrument,




instrument supplies, parts and service.









Ventur1_ tube;  a tube with a constricted throat formed by a gradually




contracting cone followed by a gradually expanding  cone used  for
                               -232-

-------
determining fluid velocity by measurement of differential pressure




at the. throat  as the  fluid traverses the tube.









Vibration:  an oscillatory velocity or acceleration expressed in units




of distance/time  ;  e.g.  cm/sec*".









Vibration  operatir..|^rang_e_;  see operating vibration range.









Voltage  operating range:  see operating voltage range.









Warm-up  titne:   the  elapsed time necessary after start-up for the




analyzer to meet performance specifications when it has  been shut




clown  for at least 24  hours,  The. shortest possible warm-up time




(minutes)  is  desirable for instruments designed for portable field
vise,
Wet-test  meter :   a volumetric flow measuring dex'ice that measures




the  total gas volume by entrapping the gas under a liquid in inverted




cups  or vanes attached to a rotor.  The buoyancy of the gas causes




rotation  of  the  rotor.  The'rotation is proportional to the gas




volume and is indicated directly on a calibrated dial.









Zero  air  (see zero gas):  air containing 21.0 ± 0.5% oxygen and no




substances that  will 1) react with the pollutant or interferent gas,
                               -233-

-------
2) alter the test analyzer  response  or  3}  interfere in the referee




procedure.  .








_Zer£_gas_ (see zero air):  an artificial atmosphere  containing no




pollutants.








Zeroing_ (also zero adjustment):  deliberate  translation of the analyzer




output or recorded data to  a position selected as zero  reference while




sampling zero air.
                                -234-

-------
                             Figure 10-1

     DIAGRAM SHOWING VISUAL REPRESENTATION AND INTERPRETATION
                OF  TIME DELAYS IN ANALYZER RESPONSE
              Pollutant
                  i in
                  f     Input
        loo -T-     r-	—	-?
2X noise
Pollutant
    off
                             Figure 10-11

                        DIAGRAM OF PULSE TIMI
        d
        o
        rt
        O
        C
        O
                                   Output
                           Time

-------