United States
           Environmental Protection
           Agency
               Office of Mobile Sources
               Emission Control Technology Division
               2565 Plymouth Road
               Ann Arbor, Michigan 48105
EPA 460/3-85-009b
September 1985
           Air
c/EPA
Outdoor Smog Chamber Experiments:
Reactivity of Methanol Exhaust
Part II: Quality Assurance and Data
Processing System Description

-------
Outdoor Smog Chamber Experiments:
     Reactivity  of Methanol  Exhaust
     Part II: Quality  Assurance and
  Data  Processing  System Description
               H. E. Jeffries, K. G. Sexton,
              R. M. Kamens, M. S. Holleman

              Department of Environmental Sciences
                      and Engineering
                   School of Public Health
                 University of North Carolina
                   Chapel Hill, N.C. 27514
                 Prepared under Subcontract with
                   Southwest Research Institute
                    Contract No. 68-03-3162
                     Work Assignment 30

                EPA Project Officer: Craig A. Harvey
              Technical Representative: Penny M. Carey

                       Prepared for

           U. S. ENVIRONMENTAL PROTECTION AGENCY
                    Office of Mobile Sources
                Emission Control Technology Division
                     2565 Plymouth Road
                     Ann Arbor, MI 48105


                       September 1985

-------
                                     DISCLAIMER
    This report has been reviewed by the Emission Control Technology Division,  U.  S. Environ-
mental Protection Agency, and approved for publication.  Approval does not signify that the contents
necessarily reflect the views and policies of the U. S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or recommendation for use.

-------
                                    Abstract
     This report describes the Quality Assurance and Data Processing procedures and systems used
at the UNC Outdoor Smog Chamber Facility. The  primary product of research conducted at this
facility is information in the form of measurements of reactants and products in photochemical
systems  and  measurements of the critical parameters  that influence the chemical transformations
system.
     Generating useful data begins  with understanding the goals  of the project  and the special
needs and concerns of conducting a successful smog chamber operation.  The system components
are designed  to collect, transfer, process, and report  accurate,  high  resolution data without loss
or distortion.  The system components in the Quality  Assurance and  Data Processing system are:
people, hardware, software, checklists, and data bases.
     Quality assurance checks are made at every level of the program. Pressurized gas tank and liquid
mixtures were  used to establish experimental conditions of HC assuring consistency throughout
the program.  Several NBS  traceable standards  and  liquid injections into  the chamber  used for
calibration have been intercompared  and show good agreement. Resulting calibration data indicates
good instrument stability during the program.
     The document describes  the calibration techniques and data processing procedures,  and the
processing of experimental data. A description and examples of the format of the final documentation
and data file  are given. The last chapter summarizes  the Quality Assurance steps and important
accuracy and precision aspects of the data quality.
                                            in

-------
IV

-------
                                  Contents
1   Overall QA Steps                                                          .1

    Introduction	    1

    Clearly Established Goals	    1

    Facilities and Analytical Support Requirements	    2

    Planning of Activities  	    2
       Calibrations	    2
       Characterization Runs	    2
       Special Mixtures	    5
       Special Techniques	    5

    Reviewing and Scheduling Methods	    5
       Priority of Runs	    5
       Use of NOAA Weather Radio	    5
       Feedback Scheduling	    6

    Site Operation	    6
       Lab Air Conditioner	    6
       Smog Chamber	    7
       Chamber Sampling Manifold	    7
       Chamber AC Drying System   .	    7
       Ethylene Converter	    7
       Mass Balance	    7
       Supplies	    8

    Check Lists	    8
       Setup Checkout Lists	    8
       Operator Checkout List   	    8
       Run Checkout List	    8
       Run Packup List   	    8

   Preventative Maintenance	     9

   Characterization Experiments	     9
      Types of Characterization Experiments	     9
      Frequency   	    10

-------
 2   System Components                                                      11

     Location of Work	   11
        Experimental Site	   11
        Data Processing Location	   11

     Personnel   	   12
        Advisory Group, AG	   12
        Administrative Research Assistant, AA	   12
        Project Coordinator, PC	   12
        Site Operator, SO	   12
        Peak Picker, PP	   12
        Computer Technician, CT   	   12

     Computers Used	   12
        DEC PDP-11/40   	   13
        DEC LSI-11/23	   13
        IBM PC   .  . •	   13
        DEC VAX-11/780	   13

     Computer Programs Used	   14
        Description of General Purpose Programs	   14
        Description of Specific Programs	   14

     Description of Data File Formats and Contents	    15
        "U" file	    15
        "G" file   	    15
        "P" files	    16
        "C" files	    16
        "K" files	    17
        "R" files	    17
        "Q" files	    17
        "A" files	•	    17
        "E" files	    17
        ACDTR.COM file	    17

     Description of Databases	    17

     Description of Database Procedures	    18
       List of all DTR Procedures by Name and Function	    18
       Worksheets/Databases	    19


3   Calibration                                                                  21

    Approach	    21

                                         vi

-------
     NBS-Traceable Sources	    21
        NO and O3 Sources	    21
        Hydrocarbon Sources	    22
        Transfer Standards Created from Primary Sources	    23

     Primary Source Comparison	    23
        NO and O3 Sources	    23
        Hydrocarbon Sources	    27

     Transfer Calibration to working  "AUTOCAL" Sources	    27
        NOx and O3 Calibration Sources   	    27
        Manual Gas Phase Titration   	    27
        Hydrocarbon Sources	    31

     Routine NOx and  O3 Calibrations Methods	    40
        Automatic Source Sampling	    40

     Routine HC Calibrations	    40
        Automatic Source Sampling	    41
        Manual Precision  Liquid Injections	    41

     Meteorological Sensor Calibrations	    42
        Solar/UV radiation Calibrations	    42
        Temperature Calibrations	    42
        Dewpoint Calibrations	    42

     Flowrate Measurements -  Calibration  	    43


     Time Measurement - Clock Calibration   	   43


     Calibration Documentation and Processing Databases	   43
        Paper Forms	   43
        Electronic Forms and Other Documentation	   44
        Calibration Data Processing Status  Database (fALEUTRY	   48
        Calibration Databases	   48
        Security of Calibration Databases	   52

     Calibration Data Processing	   53
        NOX and O3	   53
        HC Processing  	   54


4   Experimental Data                                                         71

                                          vii

-------
     Experimental Results Documentation   	    71
        Data	    71
        Forms	    73
        Plots	    77
        Run Data Processing Status Databases	    77

     Description of Data Processing Status Database	    89
        Description	    89
        Use of Data Processing Status Database	    89

     Digital Voltmeter Data Processing  .  .  . '	    91


     Strip Chart Data Processing	    95


     Documentation Processing	    99
        General Documentation Form	  100
        General Description and Purpose of Experiment	  100
        Initial Conditions	  100
        Meteorological Conditions   	  102
        Data Times, Data Exceptions, Special Problems and Concerns   	  102
        Quality  Assessment	  102
        Documentation File Production	  104

     Final Segmented File Production	  105


     Security of Data	  107
       Inventory	  107
       Retention of Data	  107
       Backup of Data	  107


5   Segmented  Data File                                                      108

    General Description	  108


    File Formats	  109


    ANSI Tape File Format   	  112


    Tape Contents	   113


    Final Data Plots made from SegFile	   113
                                         Vlll

-------
 6   Summary of QA                                                        124

     Summary of Calibration and Data QA	  124

     List of QA Steps	  125
        Before Runs	  125
        After Run-Experiment Logging	  125
        During Calibration	  126
        During DVM Data Processing	  127
        During GO Data Processing	  128
        During Documentation Stages	  129
        After SegFile Made   	  130

 APPENDICES

 A   Site  Checklists                                                          132

     Checklist For 0900 EDT Site  Operator	  132
        Outside Check   	  132
        TSR and  UV zero check  	  135
        GC Check	  135

     Checklist for Leaving the Site Set Up for a Run	  137
        Computer  	  137
        Formaldehyde	  137
        NOx and  O3  Instruments	  137
        TSR and  UV Chart	  138
        Charts	  138
        Carle I	  139
        Carle II and Carle III   	  139
        Sigma 10  Integrator	  140
        Dew Point Meter  	  141
        ATC and  Varian	  141
        CO Meter  	  141
        Air Generators  	  141
        General   	  142

    Early Morning Checkout	   144

    Run Pack-Up Checklist	   146


B   HC Tank Calibration Source Certification                           147


C   Official Calibration Sources                                            168

D   CALANA Plots and  Analysis Reports                               178

                                        ix

-------
                                       Tables
  1.  Methanol Target Experiments	     3
  2.  NO Tank Calibration Sources	    22
  3.  HC Tank and Liquid Calibration Sources   	    22
  4.  OCS from Injections in Chamber	    24
  5.  Species ID Numbers and Properties	    26
  6.  Comparison of Primary NO Calibration Sources  	    29
  7.  Intercomparisons Among Sources	    39
  8.  Comparison of Primary HC Calibration Sources  	    39
  9.  Processing System for Instrument Auto-Calibration  Data   	    49
10.  CALSUM Example	    50
11.  NOX and 03 Calibrations for 1984   	    56
12.  Processing System for Instrument Auto-Calibration  Data   	    61
13.  Summary of CALANA statistics	    67
14.  Record Description for RUNENTRY Database System	    81
15.  RUNENTRY Processing Summary for September 14, 1981	    82
16.  Processing System for DVM Data	    93
17.  Processing System for Instrument Data	    97
18.  Documentation Steps	   100
19.  Final Segmented File Production Steps	   106
20.  Example Segmented Data File	   110
21.  VAX/VMS ANSI VOLl Label	   120
22.  First File Header Label (HDRl) Fields  	   121
23.  Second File Header Label (HDR2) Fields   	   122
24.  Third File Header Label (HDR3) Fields	   123

-------
                                      Figures
  1. Comparison of Primary NO Calibration Sources   	    28
  2. NOX and O3 Cal Processing Flowchart.   	    32
  3. LOTUS Spreadsheet for Manual Cal Processing   	    33
  4. LOTUS Analysis Spreadsheet of Manual Cal	    35
  5. LOTUS Analysis Spreadsheet Manual Cal Plots   	    37
  6. Official Calibration Source Form	    45
  7. Cal Pick Instruction Form	    47
  8. AUTOCAL Processing Flowchart	    55
  9. HC Autocal Processing Flowchart	    60
 10. CALANA Example Figures   	    64
 11. Calextr Examples	    69
 12. Runsheet   	    74
 13. Instrument Checklist Form	    75
 14. Run Folder Inventory Checksheet Form	    76
 15. Data Processing Instruction Form	    78
 16. Runentry Welcome and Main Menu Screens   	    83
 17. RUNENTRY Date and Display Selection Screens	    84
 18.  RUNENTRY General, Documentation, and File Status Screens	    85
 19.  RUNENTRY DVM and Picked Instrument Status Screens	    86
 20.  PC Report Example	    87
 21.  DVM Processing Flowchart (Computers and Files)	    92
 22.  Instrument Processing Flowchart (Computers  and Files)	    96
 23.  Documentation Processing Flowchart (Computers and Files)	   101
 24.  Basic Layout of a VAX/VMS ANSI Labeled Volume    	   114
 25.  Single File/Single Volume Configuration	   115
 26.  Multifile/Multivolume Configuration  	   116
 27.  VAX/VMS ANSI VOL1 Label Format	   116
 28.  HDRl Label Format	   117
 29.  HDR2 Label Format  	   118
30.  Blocked Fixed-Length Records	   119
31.  Variable-Length Records	   119
                                           XI

-------
                        Acknowledgements
    This project was assisted by several university staff and many students.  Kenneth  Crossen,
Systems Programmer, wrote the data collection computer code, the data transfer computer code and
many utility programs. Thomas Morris, Computer Programmer, maintained the site data collection
computer code and hardware.  Randy  Goodman, Research Technician, renovated the laboratory
bench and instrument support systems, the chamber/laboratory manifold, the smog  chamber and
the chamber drying system. The site operations were managed by John Suedbeck.

    The following students performed data processing  tasks:  Lynn Clark, Jeffrey Hoffner, Charles
McDowell, Jennifer Jeffreys, Greg Yates, and Cindy Stock. The following students performed ex-
periments at the site:  Jeff Arnold, David Benham, Lisa McQuay, and Joe Simmeonssen.
                                         xn

-------
 Overall   QA   Steps
Introduction

There are a number of general activities that contribute to overall quality in the
experimental data. These include:
  • having clearly established experimental program goals;
  • establishing facilities and analytical support requirements;
  • planning activities for experimental and quality assurance programs required to
   achieve goals;
  • reviewing and scheduling activities;
  • maintaining basic Research Site Facility;
  • using checklists;
  • using preventative maintenance;
  • performing special test experiments routinely.

   Each of these will be discussed  further.

Clearly Established Goals
The most basic part of a QA program is to assure that efforts are properly designed
and directed towards clearly defined goals. This assures that the data obtained will
satisfy the needs of the program.  The approach and design of the experimental
program and the selection of experiments are  as important  as the accuracy and

-------
 Overall QA Steps	Facilities and Analytical Support Requirements

 precision aspects of the final data. A range of experiments conducted with appro-
 priate chemical conditions must be identified. Minimum data requirements must
 be determined beforehand to assure that  the results will be useful to achieve the
 program goals.

    The experimental conditions of interest must be identified and prioritized. The
 number of experiments to be performed are decided based  upon program goals,
 approach, design and resources.

    Once  an experimental program is designed, a table of Prioritized Target Condi-
 tions can  be made to guide the experimental program. An example table is shown
 in Table 1.

    Specific experimental conditions of species (and sources)  and amounts  (cham-
 ber concentrations; milligrams of solids, microliters of liquids, or seconds at specific
 flowrates  of gases) are calculated by the project coordinator and recorded on the
 Run Sheet used to conduct the experiment to assure that the proper experiment  is
 conducted. This form will be discussed further in the chapter  entitled  "Experimen-
 tal Data Processing".

 Facilities and Analytical Support Requirements

 The facilities and analytical support requirements must be identified. Special mod-
 ifications or implementation of new analytical methods consume resources. These
 activities could potentially hinder the attainment of the project goals if problems are
 encountered. Facilities and analytical capabilities are described in two Appendices
 attached to each Final Report.

 Planning of Activities

 Calibrations
 All instruments must be calibrated frequently to assure accuracy and to determine
precision.  Along with scheduling the experimental program, scheduling of calibra-
tions are required to assure that the data can be processed with minimal uncertainty.
Calibration procedures will be discussed in the next chapter.

Characterization  Runs
Several types of smog chamber experiments are generally conducted to  characterize
aspects  which  can effect interpretation of experimental results and aid modelers.

-------
Table 1. Methanol Fuel Reactivity Experiments
           (AM experiments have 0.35 ppm NOX)
Num
1
2
3
4
5
6
TYPE
Reduction, 33%
mid-ratio
Substitution
mid-ratio
normal HCHO
Substitution
mid-ratio
low HCHO
Substitution
mid-ratio
high HCHO
Reduction, 33%
low-ratio
Substitution
low-ratio
normal HCHO
REACTANTS
First Side
3.00 ppmC SynAuto
3.00 ppmC SynAuto
3.00 ppmC SynAuto
3.00 ppmC SynAuto
1.000 ppmC SynAuto
1.000 ppmC SynAuto
REACTANTS
Second Side
2.00 ppmC SynAulo
2.00 ppmC SynAuto
0.89 ppmC MeOH
0.01 ppmC MeNO2
O.lOppmC HCHO
2.00 ppmC SynAuto
0.99 ppmC MeOH
0.01 ppmC MeN02
2.00 ppmC SynAuto
0.79 ppmC MeOH
0.01 ppmC MeNO2
0.20 ppmC HCHO
0.666 ppmC SynAuto
0.666 ppmC SynAuto
0.300 ppmC MeOH
0.003 ppmC MeN02
0.030 ppmC HCHO
PURPOSE
To determine the effect of 33% reduction in HC in an auto-
ez/iaust-like environment at atypical HC-to-NOx ratio. Expect
30% reduction in ozone maximum.
To determine the reactivity of the most likely methanol fuel
exhaust in an auJo-ez/if!t<«>-like environment at a typical HC-
to-NOx ratio. Expect 20% reduction in ozone maximum.
To determine the reactivity of the lowest reactivity methanol
fuel exhaust in an «u(o-ez/iaust-like environment at a typical
HC-to-NOx ratio. Expect 30% reduction in ozone maximum.
To determine the reactivity of a highly reactive methanol fuel
exhaust in an outo-ex/ioust-like environment at a typical HC-
to-NOx ratio. Expect less than 10% reduction in ozone maxi-
mum.
To determine the effect of 33% reduction of HC in an auto-
ei/iaust-like environment at a low HC-to-NOx ratio. Expect a
large reduction in ozone maximum.
To determine the reactivity of the most likely methanol fuel
exhaust in an auto-ezJioust-like environment at a low HC-to-
NOx ratio. Expect a large reduction in ozone maximum.

-------
Table  1, cont'd.  Mcthanol Fuel Reactivity Experiments
               (All experiments have 0.35 ppm NOX)
Num
7
8
9
10
It
12
TYPE
Reduction, 33%
in id-ratio
Substitution
mid-ratio
normal HCHO
Substitution
mid-ratio
high HCHO
Substitution
low-ratio
normal HCHO
Substitution
high-ratio
normal HCHO
Chemistry
REACTANTS
First Side
3.00 ppmC SynUrban
3.00 ppmC SynUrban
3.00 ppmC SynUrban
1.000 ppmC SynUrban
6.00 ppmC SynAuto
l.OOppmC HCHO
REACTANTS
Second Side
2.00 ppmC SynUrban
2.00 ppmC SynUrban
0.89 ppmC McOH
0.01 ppmC MeN02
O.lOppmC HCHO
2.00 ppmC SynUrban
0.79 ppniG McOH
0.01 ppmC MeNO2
i_ 0.20j>pjnC HCHO
0.6GC ppmG SynUrban
0.300 ppmG McOH
0.003 ppmG MeNO2
0.030 ppmG HCHO
4.00 ppmC SynAuto
1.78 ppmC MeOH
0.02 ppmC MeN02
0.20 ppmC HCHO
l.OOppmC HCHO
1.00 ppmG MeOH
PURPOSE
To determine the effect of 33% reduction in HC in an urban-
like environment at a typical HC-lo-NOx ratio. Expect 30%
reduction in ozone maximum.
To determine the reactivity of the most likely methanol fuel
exhaust in an ur6an-likc environment at a typical HC-lo-NOx
ratio. Expect 20% reduction in ozone maximum.
To determine the reactivity of a highly reactive methanol fuel
exhaust in an ur&an-like environment at a typical HC-lo-NOx
ratio. Expect 10% reduction in ozone maximum.
To determine the reactivity of a typically reactive melhanol
fuel exhaust in an ur/ian-like environment at, a low HC-lo-NOx
ralio. Expect large reduction in ozone maximum.
To determine the reactivity of the most likely methanol fuel
exhaust in an aufo-ei/iousf-like environment at a high HC-to-
NOx ratio. Expect 20% reduction in ozone maximum.
To illustrate the chemistry of methanol in a highly reactive
environment.

-------
 Priority of Runs	                                   Overall QA Steps

 These are performed throughout the programs. A discussion of chamber background
 reactivity was given in Jeffries and Sexton12.

 Special Mixtures

 The experiments often utilize synthetic mixtures.  These mixtures require extensive
 experimental support. For example, previous experiments may compare the reactiv-
 ity of these mixtures with real auto exhaust or with other synthetic mixtures which
 are further compared with simpler photochemical systems.  This previous experi-
 mental  experience is an additional QA step which assures reference to well-studied
 photochemical systems.

    These mixtures are either ordered from scientific gas suppliers or are mixed
 and prepared on site. In either case, the resulting mixtures are  compared  with
 NBS-traceable calibration sources.

 Special Techniques

 Some project goals require the development of special techniques such as the pro-
 duction of a particular dilution rate profile or the use of highly  labile compounds
 such as methyl nitrite.  To use methyl nitrite  requires  that the synthesis, storage
 and use (injection into chamber) techniques  be developed and implemented.


 Reviewing and Scheduling  Methods

 To maximize the quality of program results, reviewing and careful scheduling are
 needed throughout the program. Important considerations are the priority of the
 experiments, status of the smog chamber site  facilities, weather, and previous ex-
 perimental results. Inherent in this process is a continuous review of experimental
 results to assure that target conditions are being attained and experimental condi-
 tions were satisfactory.

Priority of Runs

The results of experiments are affected by the initial concentration conditions, even
for the same composition of ozone precursors. For most studies, several experiments
are desired. Higher priorities are assigned to those  experiments which will yield the
most  information for the  more  average conditions.  Later experiments are used
to establish the range of conditions or to establish more basic results to assist in
interpretation.

-------
 Overall QA Steps	Use of NOAA Weather Radio

 Use of NOAA  Weather Radio
 Outdoor  Smog Chambers are subjected to real weather situations.  This leads to
 difficulties in scheduling runs. To provide up-to-the-minute information needed to
 schedule runs, we use a weather radio to receive the airport NOAA weather radio
 station. Weather reports are updated every hour and forecasts are made every 3
 minutes.  The most desirable conditions are clear, sunny, dry and warm.

    When clear weather is forecast,  the site operator is contacted by telephone. He
 gives a status report of the site facilities and instrumentation.  The next highest
 priority experiment which can be conducted considering the current status of the
 facilities is then scheduled.

    When cloudy weather is forecast ("partly cloudy") with a low probability of
 rain, a characterization experiment is scheduled unless a set of calibrations are
 needed and the long range forecast  is for good weather.

    When rainy or "mostly cloudy" weather is forecast, a "CAL DAY" is scheduled.

 Feedback Scheduling
 Experimental  results  are  reviewed  frequently.  Experiments are repeated or new
 experiments are designed and scheduled depending on the conditions or  outcome of
 experiments already conducted.

 Site Operation

 A great deal of QA occurs in  the  routine  operation  of the smog  chamber site.
 Procedures more basic than calibrations are required to assure useful and complete
 data.  The laboratory itself must be maintained.  Adequate supplies must be in
 place, all documentation must be complete, instruments must be in good operating
 order, support equipment  must be checked-out, and the chamber must  be cleaned
 and leak-checked.

 Lab Air Conditioner
 The  laboratory air conditioner maintains a consistent environment needed for the
operation of instruments and computers in the typical 95°F, high humidity, summer
conditions in North Carolina. Lab temperature  is monitored continuously by the
computer  data acquisition system and is reported in the final distributed exper-
 imental results segmented data file.  The lab temperature is maintained at 75°F.
The air conditioner is checked at the  beginning of every run season by a professional

-------
 Ethylene Convertor	Overall QA Steps

 service. The filter is changed every three weeks during most of the season and every
 two weeks during the warmest part of the season.


 Smog Chamber

 The smog chamber itself requires routine maintenance.  It  is frequently checked
 for holes  and  leaks, which are  repaired.  Occasionally  insects find their way  into
 the chambers and must be removed. The floor is swept and sponged with distilled
 water when needed.  The automated  venting doors and air drying ductwork is
 checked daily.  Leaks are often discovered by analysis (least-squares exponential fit:
 see section on computer programs below) of the inert tracer  data  collected during
 every run on both sides of the dual chamber.


 Chamber Sampling Manifold

 The sampling  manifold  is routinely checked for breaks.  It is  warmed with electric
 heat tapes the entire  length from the chamber to the  fans inside the laboratory.
 Minor leaks can be  discovered by analysis of the inert tracer  data collected during
 every run on both  sides of the  dual chamber. The manifold is checked daily for
 breaks,  inoperating heating lines by touch, and  by  appearance of condensation
 (warm chamber air  into a cooler laboratory).


 Chamber AC Drying System

 An automated chamber air drying system in place under the chamber is used before
 experiments to prevent  condensation which might form during  the  cool morning
 hours after venting the chambers.  A Freon check and general maintenance checkout
 is performed at the beginning of every run season. Testing is performed early  in the
 season to measure the dewpoint depression capability. Routine monitoring of daily
 performance (for each experiment) is the most common QA procedure.  Dewpoint
 is continuously monitored. Condensation is checked by the morning operator.


Ethylene Convertor

The  ozone monitor  utilizes ethylene. The excess is routed through a heated con-
vertor which oxidizes it. Untreated excess ethylene can  be a severe contamination
problem. To check for  leaks in the ethylene treatment system,  two GC-FID systems
 (Carles 1 and 2) monitor for this possible contamination in background chamber
air analyses before every experiment.

-------
 Overall QA Steps	Mass Balance

 Mass Balance

 A high precision Mettler balance is used for weighing chemicals used for  reagents
 and paraformaldehyde used for injection of formaldehyde into the chamber.

 Supplies
 An inventory is performed at the beginning of every run season to determine supplies
 which must be ordered long in advance of the run season. Special mixtures of HC
 and other precursors must also be ordered. More common supplies are tracked and
 stocked routinely. During the run season, the use of expendable supplies is carefully
 monitored and reordered when necessary.


 Check  Lists

 Many details must be checked to assure that  a  smog chamber experiment will be
 successful. Several checklists are  used before, during, and after experiments.

 Setup Checkout Lists

 To make sure that things are ready for an experiment the next day, a setup checkout
 list is used. All supplies (gases, etc.), computer and instrument switch settings, and
 injection gases are checked; syringes and liquid chemicals are set out; the run sheet
 is rilled out; venting is checked; the manifold is checked for leaks, etc. This checklist
 is shown in Appendix A.

 Operator Checkout  List
 For the early morning operator responsible for initiating the experiment,  another
 checklist is used. This is somewhat redundant to the setup list to check extremely
 important details that might have been overlooked by the setup operator. In  addi-
 tion, check items include details about proper  operation of instruments, manifold,
 and chamber, injections of chemical precursors, and recording documentation about
 details of experiment. This list is also shown in Appendix A.

Run  Checkout List
Another checkout list is used by site personnel that arrive at mid-morning. Check
items include checking on status of instruments, stripchart recorders, computer,
chamber,  etc. to assure successful continuation of the experiment and data acquisi-
tion. This list is also given in Appendix A.

-------
 Types of Characterization Experiments	Overall QA Steps

 Run Packup List
 This list is primarily an inventory checklist to insure that all data is properly packed
 up, documented, and returned to the School of Public Health data processing office.
 This list is included in Appendix A.

 Preventative  Maintenance

 To assure proper operation of all instruments, the smog chamber, support facilities,
 instrument stability (reflected  in "cal factor variation"), and maximum "up-time,"
 preventative maintenance is routinely performed.

 Characterization Experiments

 As mentioned above, characterization experiments are performed routinely. These
 experiments are  used by modelers to determine the importance of chamber artifact
 processes.  Background,  reactivity,  surface processes (such as off-gasing  of NOX,
 HCHO), decay of 03  and NOX are of primary interest. Previous experiments conducted
 over several years  have been studied  to characterize these and other aspects of
 chamber behavior.  Benchmark experiments are also routinely  performed. These
 include propylene and NOX experiments.

 Types of Characterization  Experiments

 Background
 This experiment shows how much O3 is  formed when no chemical precursors are
 injected into the  chamber except for the rural air used in venting and any material
 which may remain  on the chamber walls  after venting.

 O3 Decay
 O3 is injected into otherwise clean air and allowed to decay. The rate is determined.
 This experiment  is usually performed at  night. Some modelers  include this decay
 process although the rate is very low.

 JVO Oxidation, Daytime

 NO is injected into clean  chamber air (nothing else injected)  to detect sources of
radicals.  The rate  of oxidation is observed.  The magnitude of and types chamber
radical sources required to achieve the observed rate can be estimated by modeling.
To modify  the test, CO can be added  to one side to help  determine the type of

-------
 Overall QA Steps	Frequency

 radical source. The CO changes the balance of HO and HO2 radicals without adding
 any radicals to the system itself.

 NO Decay, Nighttime
 This experiment is conducted like the one described above except that no sunlight
 is involved.

 Acetaldehyde and CO
 Matched injections of acetaldehyde (usually 1 ppmC)  are made into each chamber.
 One side has CO injected (usually 50 ppm). This experiment forms PAN which can
 aid in  detecting any wall NOX  off-gasses.  The CO changes the balance of HO  and
 H02 radicals without adding any radicals to the system itself.

 Formaldehyde and CO

 Matched injections  of formaldehyde  are made into each chamber. One side has
 CO injected (usually 50 ppm).  Formaldehyde,photolyzes to produce HO, and CO.
 Without injected NOX, only the  background NOX in the rural air or from the chamber
 walls can participate in this system. Any HO produced can react with formaldehyde
 or with the additional CO added to one side.  The CO changes the balance of HO
 and HO2 radicals without adding any  radicals to the system itself.

 Propylene and NOx
 Matched conditions of propylene and NOX have been used for years as a bench-
 marking experiment.  Experience  has shown that  experiments conducted on the
 same day a year apart can be  duplicated  if the sunlight  quality is good.  Demon-
 strating matched  results from the two chambers is  another reason for conducting
 matched condition experiments.

 Frequency
 These experiments should be conducted routinely. However, they are not conducted
as frequently when  the program falls behind schedule because of poor weather.
These experiments have been studied and the range of magnitude of these chamber
processes is well known.
                                    10

-------
 System   Components
This section lists and describes the system components. Major types of tasks and
responsibilities have been organized and assigned to different people. Computers,
instruments and other hardware are used to generate and process data. Many com-
puter programs (software) are used. Policies and procedures have been  developed
to assure proper and consistent treatment of the data. Databases have been created
to organize and to aid in retrieving and processing the large amounts of calibration
data.
Location of Work

The research work occurs in two different locations. The experiments are conducted
at the Smog Chamber Site located in a rural area with relatively clean background
air, and the data is processed on the University campus.


Experimental Site

The site is located in Chatham County, North Carolina, approximately 32 kilo-
meters from the University of North Carolina at Chapel Hill, and approximately
10 kilometers from the small town  of Pittsboro. Chatham County is one of the
most rural, least  industralized counties  in North Carolina and is heavily wooded.
The background concentrations of NOX and NMHC are usually less than 5 ppbC and
less than 80 ppbC. More importantly, the air exhibits very low reactivity in the
chamber.

                                 11

-------
 System Components	Data Processing Location

 Data Processing Location

 The data are processed in the School of Public Health on the campus of UNC, where
 all computers and data processing facilities are easily accessible.

 Personnel

 A number of people with different tasks contribute to the process of conducting
 smog chamber experiments  and processing the results.  The names and roles of
 these people" are listed below.

 Advisory Group,  AG
 An advisory group,  consisting of the principal investigator, the co-principal inves-
 tigator,  the  project  coordinator, and  the administrative research assistant meet
 frequently to review the progress of each project.

 Administrative Research  Assistant, A A

 The administrative research assistant is responsible for logging the experiments and
 calibrations, ordering supplies, and managing the budget.

 Project Coordinator, PC
 The project coordinator oversees the experimental program and the processing of
 the data.

 Site Operator, SO

 Site operators are responsible for maintenance,  calibration, and operation of the
 research site. They work closely with the PC to determine which experiments should
 be conducted next or if time should be scheduled for calibration or maintenance.

 Peak Picker, PP
 The "peak pickers" process most of the data under the guidance of the PC.

 Computer  Technician, CT

The computer technicians are responsible for unpacking the data from  the smog
chamber data acquisition system and for making tapes of segmented data files and
a backup of all data processing files.

                                     12

-------
 DEC VAX-11/780	System Components

 Computers Used

 Several computers are used in the program. They help conduct experiments, process
 data, plot data, analyze results, and aid in writing.

 DEC PDP-11/40
 The DEC PDP-il/40 is used to control the experiments and perform the data acqui-
 sition for much of the data. This computer writes data to 8 inch flexible diskettes
 for transfer to the data processing office.

 DEC LSI-11/23
 Two  DEC LSI-11/23 computers are  used to process most of the experimental and
 calibration data.  Most programs are written  in PASCAL. A major component is the
 DIGITIZER which permits rapid cind accurate conversion of stripchart data to com-
 puter  data  files.  Each computer  has  at least  one 8  inch floppy drive  and a 10
 Mb hard disk.   These computers can  communicate to the IBM PC and to  the DEC
 VAX-11/780.

 IBM PC
 The IBM PC is used with LOTUS  123 to implement many spreadsheets and databases
 discussed throughout this report. It is  also used for word processing.  Some PASCAL
 programs are also used. The IBM PC uses 51/4 inch floppies and can "upload" and
 "download" data from the DEC VAX-n/780.

 DEC VAX-11/780
 A DEC VAX-11/780 mainframe computer  is used to maintain several large databases
 discussed in this report.  The VAX system also makes plots and prints reports  using a
 large, high resolution laser printer.  Modeling input decks are set up and submitted
from the VAX to be run on the university's IBM 4381 mainframe computer.  "Printouts"
and "punched"  output are returned to the VAX as standard ASCII files.  The VAX can
also read and write to 8 inch floppies and industry standard magnetic tape.
                                     13

-------
 System Components	Computer Programs Used

 Computer Programs Used

 Description of General Purpose Programs
 LOTUS 123
 LOTUS 123 is an electronic spreadsheet, database, and graphic program that runs on
 the IBM PC. The spreadsheet in 123 has 256 columns and 2096  rows; each column
 can be a different width and each cell can hold a text label, a numerical value, or a
 formula that includes references to other cells. When new data is entered into cells,
 the program automatically  updates the values of all the formulas. Ranges of cells
 can be selected to be graphed. Graphs include line, bar, and  pie; up to 6 lines can
 be plotted on a graph. In  addition, the spreadsheet contents can be treated like
 a database table in  which the rows are records and the columns are fields in the
 record.  Records can be selected, deleted, copied, and sorted to  an output location
 on the spreadsheet. The whole spreadsheet or selected parts of the spreadsheet can
 be printed to a printer or  file. Data can also be 'imported' from other programs or
 files.

 DATATRIEVE
 DATATRIEVE is a data base query, update, and  report writing system running on
 the VAX-ll.  DATATRIEVE is used to maintain calibration and  data processing sta-
 tus databases and serves  a  tool to coordinate effort among the PC, PP, and CT.  In
 DATATRIEVE, the PC can ask the system to print a list of all runs that  had  all in-
 strument  data "picked" but did not have calibration factors determined for those
 instruments. From this list, the PC determines which calibration factors must be
 produced  to complete the processing of the runs.  Typical run tracking  data bases
 grow to be  500,000 characters during a run  processing season.  The total size  of
 data managed by DATATRIEVE is more than 20 million characters.

 Description of Specific  Programs
 UNPACK converts packed binary DVM data from the chamber site's DA system to
 ASCII format ("U  files").

    FASTDV produces temporary ("P" file) files of raw DVM data for making plots
 for preliminary QA procedures,  and strips NOX and O3  calibration data from DVM
 data file.

    DIGPIK allows the digitizer to be used to convert experimental peak height data
for all species for a given instrument and writes  data to ("P"  file) files on LSlii/23.

                                     14

-------
 "U" file                                                         System Components

 It also is used to digitize calibration data and produce calibration response ("R")
 files.

    PLOP1C plots raw ("P" file) experimental data from both DVM or stripcharts.

    ACALAV automatically produces average NOX  and  Os instrument calibration
 data for zeros and spans stripped from the DVM data with FASTDV.

    CALFAC calculates calibration factors from digitized calibration data and the
 official calibration source file.

    CALLOK lists and  plots calibration data for a given instrument  and species
 combination.

    CALANA performs statistical analysis on calibration data for any combination
 of instrument and species.

    DVMFIX applies calibration factors to DVM ("U"  file) data and outputs concen-
 tration and physical measurement  ("G") files.

    CALCON applies calibration factors  to digitized ("P" file) data from stripcharts
 and outputs concentration ("C" files).

    PLOCON plots concentration ("C" file) experimental data from stripcharts.

    HCANAL reads time/concentration data files of HC data and computes compo-
 sition analysis.

    CSTAR reads tracer data files  and computes chamber dilution rate by fitting
 with exponential least-squares method.

 Description of Data File Formats and  Contents

 «U" file
The primary data file originating from the Data Acquisition system is  "IT'npacked
after transport from the chamber site to the data processing office. This  file con-
tains "raw" data from the NOX and Os monitors, chamber, ambient and laboratory
temperature,  total solar and ultraviolet light,  and dewpoint monitors. The data
is taken every minute alternating between the two  sides.  Calibration data from
automatic calibrations of the NOX and Os monitors is also contained in this file.

                                     15

-------
 System Components                                                           "G" file

 "G" file

 The final data file processed from the primary "IT  file is referred to as the  "G"as
 Chamber file. This file contains corrected (adjusted with calibration factors) data.
 The final data is for four minute intervals alternating between the two  sides.

    The file is comprised of two parts: documentation and data. The documentation
 identifies the experiment. Next information pertaining to the processing of the data
 is given:
  • source of calibration factors;
  • data last processed;
  • explanation of data identifiers (labels);
  • explanation of units;
  • explanation of data alteration.

    Finally the actual calibration factors are listed. Calibration factors are specified
 for the beginning  and ending of periods of valid data.

    The data are preceded by a "header" card identifying the columns of data. Data
 are listed in the same order as the header card. An example of this type of file is
 shown in Chapter 6.

 "P" files
 The data files of  "P"icked data from other individual instruments  which are not
 connected to the Data Acquisition System are produced and processed separately
 with more manual effort (e.(/.measuring peak heights from gas chromatographs).
 This file is comprised of two parts: documentation and data. The documentation
 indicates:
  •  the source of the data;
  •  the data processor ("peakpicker");
  •  attenuation;
  •  units;
  •  species identification.

    The data follows with one data point per line: the format allows for easy editing
and correction during processing.

                                      16

-------
 ACDTR.COM file	System Components

 "C" files
 These are the files of final corrected "C"oncentration data processed from the  "P"
 files. The format of these files are identical to the data portion of the UG" files. An
 example of this type of file is shown in Chapter 6.

 "K" files
 For every "C" file there is a matching "K" file containing documentation pertaining
 to the source and processing of the data. In addition to the documentation originally
 in the "P" file  is the maximum amplitude and concentration of each species in the
 file and the calibration factor (and source) used in processing the data. An example
 of this type of  file is shown in Chapter 6.

 «R" files
 Calibration "R"esponse data to official calibration  sources other than from  the
 NOX and 03 monitors are stored in  these files, which are produced from the same
 program which produces the  "P" files (DIGPIK). They have very similar format and
 documentation but also include the DCS identification number.

 "Q" files
 The resulting calibration factors from comparing the "R" files  to the Official Cali-
 bration Sources are stored in the "Q" files along with all associated documentation
 from the "R" file. This file also contains DATATRIEVE commands  which will store  the
 data in the calibration DATATRIEVE database HCCAL.

 "A"  files
 Calibration response data from  the "A"utomatic calibrations of  the  NOX and 0$
 monitors are  produced with FASTDV from the data stored in the "U" file.

 "E" files
 The "A" files often contain data which represent a transition reading between the
chamber air and the calibration source. This unwanted data is "E"dited from the
file.

ACDTR.COM file
The "E"  files are processed (averaged) to a single value with a program called
ACALAV which is  output to a file Ceilled ACDTR.COM which also contains DATATRIEVE
commands to store the data in the DATATRIEVE database AUTOCAL.

                                      17

-------
 System Components	Description of Databases

 Description of Databases

 Several databases are used in processing the data and calibration data.  The fol-
 lowing are created with DATATRIEVE on the VAX-11/780. These are described in more
 detail  in Chapter 3.

    RUNENTRY is the primary database identifying the experiment and tracking
 the processing of the data from each instrument used.

    CAL.ENTRY is the database for tracking the calibration data processing.

    CAL.SOURCES is the database of the  "O"fficial "Calibration "S"ources ocs)
 including all supporting information and corresponding identification number.

    HC.CAL is the database of the final processed calibration factors resulting from
 calibrations performed on the analytical instruments producing data which is not
 collected from the Data Acquisition System.

    AUTO.CAL is the database of the calibration data from "auto"matically per-
 formed calibrations of the NOX and Os monitors.

    RUN.CAL.USED is the database of the calibration  factors actually used in
 processing the manually processed ("P" file) data.

    SPECIES is the database of official species names and corresponding identica-
 tion numbers used in data processing to assure consistency and proper spelling and
 identification.

 Description of Database  Procedures

 List of all DTR Procedures  by Name and  Function

 Various procedures are used with the DATATRIEVE databases  to assist in accessing and
 using the  data and producing reports or files to be used by other data processing
 computer  programs.

   PPREPORTS is a VAX command  procedure which creates a task report for Peak
 Pickers (data processors running DIGPIK using DATATRIEVE and the RUNENTRY data base.
The report lists each data processing step presently needing PP action, such as
instrument stripcharts needing to be digitized and plotted.

                                     18

-------
 Worksheets/Databases                                                System Components


    PCREPORTS is a  VAX command procedure which  creates a task report  for
 Project Coordinators using DATATRIEVE and the RUliEHTRY data base. The report lists
 each data processing step presently needing PC action, such as various QA steps.

    CTREPORTS is a  VAX command procedure which  creates a task report  for
 Computer Technicians using DATATRIEVE and the RUtiEl.'TRY data base. The report lists
 each data processing step presently needing CT action, such as moving data to the
 VAX or vice versa.

    PRINT-SPECIES prints a list of the official species  names and corresponding
 identification number.

    PRINT.CAL-SOURCES prints a list of the Official Calibration Sources includ-
 ing name, serial number, source date, validation  date and personnel, manufacturers
 and the validated concentrations.

    PRINT.OCS-FILE-FOR-LSI produces a file  of the official calibration sources to
 be used by another program (CALFAC) used to calculate calibration factors from
 the calibration response data.

    PRINT.CALS-AND-FACTORS calculates and produces a report of the  cali-
 bration factors calculated from the automatic calibration data from the NOX and O3
 monitors from the AUTO.CAL database.

    PRINT-HC-CAL.FACTORS produces a report of the  calibration data for  all
 the other instruments other than NOX and O3monitors from the HC.CAL database.

    PRINT-RUN-CAL.USED- produces a report of the calibration factors used  by
 instrument, species, and date from the RUll.CAL.USED database.

    PRINT.CALENTRY.SUM.ON.FILE produces a summary report of all cali-
bration data processing steps completed by date, by instrument from the CAL.EIITRY
database.

Worksheets/Databases
Several worksheets implemented with the LOTUS 123 software on the IBM PC are
used.  These  are backed-up on 5.25 inch floppy diskettes.  Each of these will be
discussed in detail in Chapter 3.

    OCSICyy - Injections into the smog chamber are used as  calibration sources.
This worksheet/database calculates the chamber concentration from the  amount

                                     19

-------
System Components                                                 Worksheets/Databases

injected, the chamber temperature, the ideal gas law, and compound physical data
from a lookup table in the worksheet.  A different worksheet is maintained for each
year (yy).

    CAL - All  hydrocarbon calibration sources are compared and  "validated" in
this worksheet/database.  It is listed in Appendix B.

    MANmmddyy - "Man"ually performed gas phase titration calibrations for the
NOX and ©3 monitors are  processed in this worksheet. One worksheet is produced
for each calibration identified by date (month, day, year).

    MANCALyy - The calibration factors resulting from the MANmmddyy work-
sheets is analyzed in this worksheet.
                                     20

-------
 Calibration
Approach
All instruments are calibrated at a minimum of two reference levels: a "zero" and a
"span" point. The latter is generally near the expected maximum concentration in
the experiments. "Zero"  air samples are obtained by sampling from a zero air gen-
erator or by sampling from a filter specific for the species being monitored. "Span"
values are generated from several sources that will be described below.  During the
programs, different "span" levels are used. These different levels assure that linear-
ity can be checked in final QA procedures, to be discussed in later sections. Several
different calibration standard sources, including NBS traceable sources and sources
from outside groups, are  utilized and compared. Calibrations for most instruments
are performed immediately before an experiment. NOX and  03 monitors are  cali-
brated twice daily. These calibration data are used to adjust the data collected in
final data processing steps after the calibration data itself has been processed and
reviewed  in the context of the entire calibration database for the season.

NBS-Traceable Sources

NO and O3 Sources
Two techniques are used to calibrate the chemiluminescent NOX  and 03 monitors.
The primary technique is  the EPA gas-phase titration method3 which utilizes a high
concentration NO source (NBS-traceable) as the primary standard. Comparison of
this primary source with  other sources is performed regularly. In one recent study,
three of these high concentration sources were compared and found to agree within
a few  percent of each other. Low concentration sources of NO and O3 are also used

                                    21

-------
 Calibration
                                                                Hydrocarbon Sources
 to perform span checks more frequently and automatically (AUTOCAL source).  Low
 concentration  (<1 ppm) pressurized gas cylinders are used for NO sources, and a
 generator built into the Os monitor is used for the  O3 AUTOCAL source. These are
 calibrated and used as transfer standards. Table 2 lists the calibration sources  used
 for NOX.
                   Table 2. NO Tank Calibration  Sources
           Name
Hanuf.
Manf.  ID  UtIC ID
DCS ID
           NO 46.0 ppm  Matheson  SX-13759  UNCAB2113    94
           MO 52.4 ppm  Airco     1122590   UIJCBC9045   169
           HO 52.6 ppm  Scott     BAL177    UNCAB0348   170
           HO 0.823 ppm          CC-15860               4
           HO 0.288 ppm                               171
           MO 0.665 ppm Airco
         CC-15780  U11CAB2478   172
    The OCS ID heading will be explained below.
Hydrocarbon Sources

Commercially prepared NBS traceable calibration sources are also used for hydro-
carbons. Five different tanks purchased by UNC were used. In addition, two other
tanks were borrowed from the Research Triangle Institute (RTI) who is involved
in a multiyear intercomparison of commercial hydrocarbon calibration sources. A
low concentration tank  prepared from diluting a small  portion of one of the com-
mercially prepared high concentration tanks was calibrated as a transfer standard
for automatic calibration. These calibration sources are listed in Table 3. Another
source of hydrocarbon standards was prepared in the chambers by injecting liquid
hydrocarbons.  These are listed in Table 4.  The identities and  properties of the
species calibrated this way are given in  Table 5.  All sources were compared and
results are described below.

                                      22

-------
 NO and O3 Sources	Calibration


            Table 3.  HC Tank and Liquid Calibration Sources
IIEV.' AIRCO Airco
LOV: HW (HC) Scott
LOV.' MW (LC) U11C
HIGH MW Scott
RTI- ethylene
RTI- propylene
82HCCALMIX
NEV/SCOTTCALTANK
I.IQ STD 8/14
X-3362
AAL11557
1IA
AAL11551
2H
3A
XA-1336

NA
U1ICAB2467
U11CAB2136
HA
UNCAB2139





112
109
111
108
167
168
1
110
123
Transfer Standards Created from Primary Sources
For NOX, O;j, and hydrocarbons, several sources traceable to NBS are used.  These
are high concentration tanks for stability and accuracy which require dilution or
instrument attenuation adjustment when used.  Accurate liquid injections into the
smog chamber are also considered as primary standards. Several  low concentra-
tion  NO tanks  were  purchased as a  source for automatic calibrations  ("AUTOCAL"
sources).  A low concentration HC tank was prepared by  diluting a portion  of the
LOW MOLECULAR WEIGHT (high concentration) calibration tank into an  empty
tank with nitrogen to be used as an AUTOCAL source. AUTOCAL sources do not require a
morning operator to  be present to change the attenuation settings and are not sub-
ject to possible errors from dilution. These calibration sources were intercompared.
The  RTI  tanks were considered absolute because of their extensive comparative
analysis history. In the case of the hydrocarbon sources, some judgements resulted
in small adjustments to the "validated" concentrations to make all the UNC sources
more consistent. The intercomparison procedures result in calibration of the AUTOCAL
sources.

Primary  Source  Comparison

NO  and  O3 Sources
This test is conducted  to compare the manufacturers stated analyzed gas concen-
trations of the NO tanks. Nitrogen oxide tanks are available for use in the EPA Gas
Phase Titration method (see description  in "Routine NOX and Os Calibrations").
Three tanks with NO concentrations of 52.6, 52.4, and 46 ppm  (see 2) were diluted
with clean air (zero air generator) to concentrations of approximately 0.2, 0.6 and

                                     23

-------
                                  Table 4.
OFFICIAL CALIBRATION SOURCES FROM INJECTIONS  INTO SHOS CHAHBER




LAST UPDATE:   13-JUN-B5




CHAHBER VOLUME   150000 (L)  STP VOL:     22.414 (L)
DCS OCS SOURCE
ID t NANE DATE
95
94
97
98
99
100
101
102
103
104
105
106
107
115
116
117
118
118
118
118
118
118
119
120
123
123
123
123
123
123
123
124
125
126
127
128
129
130
131
i/25
4/23
6/27
6/26
6/26
6/19
6/19
7/10
7/10
7/11
7/11
7/27
7/27
7/31
6/28
6/28
7/16
7/16
7/16
7/16
7/16
7/16
7/27
7/27
3/14
8/14
8/14
8/14
8/14
8/14
8/14
6/27
7/13
7/13





SPEC
ID t
68
55
55
68
71
55
68
139
139
139
139
139
139
139
55
68
40
55
68
71
67
89
82
40
19
28
42
82
55
68
71
68
139
139
205
205
205
205
205
C
t
8
7
7
8
8
7
8
1
1
I
1
1
1
I
7
8
6
7
8
a
8
a
8
6
5
6
7
a
7
8
a
a
i
i
i
i
i
i
i
DENSITY UK

0.864
0.867
0.867
0.864
0.897
0.867
0.864
0.815
0.815
0.815
0.815
0.815
0.815
0.815
0.867
0.864
0.878
0.367
0.864
0.897
0.861
0.906
0.692
0.878
0.626
0.653
0.695
0.692
0.867
0.864
0.897
0.364
0.315
0.815
0.791
0.791
0.791
0.791
0.791

106.1
92.14
92.14
106.1
106.1
92.14
106.1
30.05
30.05
30.05
30.05
30.05
30.05
30.05
92.14
106.1
78.11
92.14
106.1
106.1
106.1
104.1
114.2
78.11
72.15
86.18
100.2
114.2
92.14
106.1
106.1
106.1
30.05
30.05
32.04
32.04
32.04
32.04
32.04
UL
or 6
283.000
556.000
463.000
283.000
243.000
370.500
282.900
.0.110
0.210
0.192
0.192
0.192
0.038
0.192
556.000
283.000








416.000
420.000
372.000
373.000
274.000
276.000
238.000
188.600
0.212
0.212
224.000
224.000
80.100
74.700
196.700
T

F

70.0
70
70
71
71
70
70
70
.0
.0
.8
.8
.0
.0
.0
70.0
70.0
70.0
70.0
70.0
70.0
77.3
77.3








85.
85.
85.
85.
85.
85.
85.
70.
68.
68.
70.
74.
74.
75.








0
0
0
0
0
0
0
0
0
0
4
0
0
5
73.0
PURITY

100Z
1001
1001
100!
1001
100Z
100Z
921
92Z
92Z
92Z
92Z
92Z
92Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
92Z
92Z
100Z
100Z
100Z
100Z
100Z
CONC
UL
2.9679
5.8955
4.9093
2.9780
2.6530
3.9285
2.9668
NA
NA
NA
NA
NA
NA
NA
5.9767
3.0083
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
o.oooo
2.9875
3.1609
2.9897
2.9925
2.9876
2.9765
2.6630
1.9779
NA
NA
0.8914
0.8974
0.3209
0.3001
0.7866
CONC
S
NA
NA
NA
NA
NA
NA
NA
0.5421
1.0350
0.9463
0.9463
0.9463
0.1893
0.9463
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.0384
1.0384
NA
NA
NA
NA
NA
                                       24

-------
Table 4, cont.
OCS DCS SOURCE
ID 1 NAME DATE
132
133
134 '
135
134
137
138
140
143
144
147
148
159
162
162
162
162
163
163
163
163
164
165
16S
165
165
165
165
166
166
166
166
166

10/7
10/7
10/9
10/9
10/16
10/16
8/25
9/1
9/3
9/17
9/25
9/5
5/14/85
5/14/85
5/14/85
5/14/85
5/17/85
5/17/85
5/17/85
5/17/85
5/20/85
5/22/85
5/22/85
. 5/22/85
5/22/85
5/22/85
5/22/85
5/24/85
5/24/85
5/24/85
5/24/85
5/24/85
SPEC
ID 1
205
139
139
139
139
139
139
205
205
205
205
139
139
55
40
68
82
55
40
68
82
139
82
40
55
68
71
78
82
40
55
68
71
C
i
t
1
I
I
'1
1
1
1
I
1
I
1
1
7
6
8
8
7
6
8
8
1
a
6
7
8
3
9
8
6
7
8
8
DENSITY IW

0.791
0.815
0.815
0.815
0.815
0.815
0.815
0.7914
0.7914
0.7914
0.7914
0.315
0.815
0.867
0.378
0.864
0.692
0.367
0.373
0.864
0.692
0.815
0.692
0.378
0.367
0.864
0.897
0.876
0.692
0.878
0.367
0.364
0.897

UL
or 6
32.04 65.500
30.05 0.192
30.05
30.05
30.05
30.05
30.05
32.04
32.04
32.04
32.04
30.05
30.05
92.14
78.11
106.1
114.2
92.14
78.11
106.1
114.2
30.05
114.2
78.11
92.14
106.1
106.1
120.2
114.2
78.11
92.14
106.1
106.1
0.097
0.192
0.192
0.048
0.096
74.000
249.000
66.900
151.000
0.192
0.192
92.000
45.000
185.000
240.000
100.000
50.000
200.000
300.000
0.201
100.000
150.000
200.000
400.000
500.000
500.000
100.000
150.000
200.000
400.000
500.000
T F

74.5
54.0
54.0
55.0
55.0
61.0
61.0
66.2
59.0
66.7
45.0.
62.0.
73.0
36.0
86.0
86.0
36.0
72.0
72.0
72.0
72.0
82.0
30.0
80.0
80.0
80.0
80.0
80.0
65.0
65.0
65.0
65.0
65.0
PURITY

100!
921
921
921
921
921
921
100Z
100Z
1001
1001
921
921
100Z
100Z
1001
100Z
100Z
100Z
100Z
100Z
95Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
100Z
CONC
UL
0.2627
NA
NA
NA
NA
NA
NA
0.2921
0.9695.
0.2644
0.5721
NA
NA
1.0050
0.5033
1.9988
1.9290
1.0643
0.5449
2.1054
2.3494
NA
0.7949
1.6593
2.1607
4.2741
5.5431
5.3791
0.7728
1.6132
2.1006
4.1553
5.3889
CQNC
6
NA
0.9177
0.4636
0.9195
0.9195
0.2325
0.4651
NA
NA
NA
NA
0.9320
0.9606
NA
NA
NA
NA
NA
NA
NA
NA
1.0461
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
   25

-------
Calibration	NO and O3 Sources






    Table 5. Species ID numbers and Properties
                              ID    NAHE    I  DENS    NU  PURITY




                               19 N-PENTANE 5  0.624 72.15     1001
28
40
42
55
60
67
68
71
82
89
139
2-HETHYIP
BENZENE
2,3-DI«ET
TOLUENE
N-OXTANE
P-XYLENE
H-XYLENE
Q-XYLENE
2,2,4-TRI
STYRENE
HCHO
6
6
7
7
8
8
8
8
8
8
1
0.
0.
0.
0.
653
878
695
867
0.703
0.
0.
0.
0.
0.
0.
861
864
897
692
906
815
86.18
78.11
100.2
92.14
114.2
106.1
106.1
106.1
114.2
104.1
30.05
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
IOOZ
92Z
                             205 HEOH      1 0.7914 32.04     IOOZ
                                           26

-------
 Manual Gas Phase Titration	Calibration

 0.8 ppm. The exact  calculated concentrations were determined by flow measure-
 ments with soap-bubble meters for the air and NO.  The measured concentrations
 were plotted against  the calculated concentration.  All data were  fit to a straight
 line with the least-squares method. The data is plotted in Figure 1 and is shown in
 Table 6. Agreement between the sources is excellent. The values assigned to these
 tanks determine the absolute accuracy of our reported NOX and O3 data.

 Hydrocarbon Sources
 Because there is much less of a distinction between primary and working sources for
 hydrocarbons than for IMOX and 03 sources, the comparison of primary hydrocarbon
 sources will be discussed below in the next section on transfer standards.

 Transfer Calibration to working "AUTOCAL"  Sources

 NOx .and O3 Calibration Sources
 Experience and analysis of past calibration data indicates that calibration factors
 estimated from the results of the "manually" performed gas-phase titration method
 vary  more than the calibration factors obtained from the automatic span checks
 using low concentration sources.  But when the manual values are averaged, they
 yield the same  results.  The low concentration sources are convenient, but require
 careful cheiriicterization, since their concentration  decreases slightly during the  run
 season.  Our experience and analysis also shows that the instruments are more stable
 if the zero and  span settings are not adjusted. Therefore, the following strategy is
 followed to maximize  the quality of the data from these instruments.

    The  NOX and  O$ instruments are not adjusted after the initial  start-up at  the
 beginning of the experimental  season.  Manual gas-phase  titration calibrations  are
 performed to estimate calibration factors for the instruments throughout the exper-
 imental season. The low concentration sources are used twice a day.  The responses
 are plotted for the entire run season to establish a decay rate for the low concentra-
 tion sources. The manual calibration  factors establish the concentrations of these
 low concentration sources at the beginning of the season. The calibrated responses
from the low concentration sources are then used to calibrate the instruments. The
flowchart illustrating  the procedures  associated with the transfer  of NOX and O3
calibrations is shown in  Figure 2.

Manual Gas Phase Titration
The manual Gas Phase Titration calibrations are performed occasionally through-
out the  season as  discussed above. The technique is the EPA reference method as

                                     27

-------
    Calibration
                                  Manual Gas Phase Titration
     COMPARISON  OF  PRIMARY NO CAL  SOURCES
 I
 8
 1
      .  0


   Q  1984
               0.4
0.6
0.8
   CALCULATED CONCENTRATION (TANK LABEL)
+  1983      o  Aeroaol        	 Linear Fit
Figure 1.  Comparison of Primary NO Calibration Sources


                                 28

-------
Manual Gas Phase Titration	Calibration

        Table 6. Comparison of Primary NO Calibration  Sources

                               CALC. ESTIMATE MEASURED    DELTA  PERCEi.'T
            TANK               cone.    cone.    couc.     cone.

              1
            1984
          52.4 ppm

              2
            1983
           46 ppm

              3
          Aerosol Project
          52.6 ppm

                                                              ave  2.3%
0.197
0.392
0.803
0.202
0.422
0.886
0.192
0.377
0.786
0.213
0.418
0.849
0.218
0.449
0.936
0.208
0.402
0.831
0.218
0.423
0.855
0.229
0.419
0.930
0.215
0.403
0.840
0.005
0.005
0.006
0.011
-0.030
-0 . 006
0.007
0.001
0.009
2.3%
1.2*
0.7%
4.8%
6.7%
0.7%
3.4%
0.2%
1.0%
                                      29

-------
 Calibration	Manual Gas Phase Titration

 described in the CFR3.

    This calibration technique is based upon the rapid gas phase reaction between
 NO and O3 to produce stoichiometric quantities of NO2. The quantitative nature of
 this reaction is such that when the NO concentration is known (primary calibration
 standard), the concentration of N02 and O3 can  be determined. O3 is added to excess
 NO in a dynamic calibration system, and the  NO/NOX/NO2 analyzer is used as an
 indicator of changes in  NO concentration.  Upon the  addition of O3,  the decrease
 in NO concentration observed on the calibrated NO channel is  equivalent to the
 concentration of N02 produced and 03 consumed.

 Calibration Factors from Primary Sources
 Each "manual"  calibration is processed with the aid  of an electronic form imple-
 mented with LOTUS  123 on an IBM PC called MAHCALSH.MS in  Figure 2 (see figure Fig-
 ure 3. Information  is transfered from the paper form rilled in during the gas phase
 titration calibration procedure. The results of each manual  calibration are  inspected
 for notes of problems that were encountered, such as odd or unstable readings. The
 electronic form  assures  that the calculations are  performed correctly.  The com-
 pleted form is printed out and the file is saved on diskette and backed up to another
 diskette, thus resulting in at least four sources  of the calibration  data.

    Another LOTUS 123 electronic form and database is used to collect the final cal-
 ibration factors  for  all calibrations (called MANCAL.WKS in Figure 2). Output
 from  this worksheet is shown in Figure 4. The data are fitted with a  straight line
 using a least-squares method.  The data  and the straight  line fit are  plotted  as a
 function of date; these are shown in Figure 5. The plots are used to visually inspect
 the data for trends and large  variations.  If necessary, the individual calculation
 electronic forms can be inspected for errors in transfering data. The original paper
 form can also be examined.  If there are problems, the data in the database may  be
 invalidated.

   If no trend is observed, the calibration factors are simply averaged for the season
 to estimate the true calibration factor for the instrument.  Analysis of manual
 calibrations performed over many years by many people show variation by human
error  but usually no trend.  If a  trend were observed, the calibration factor for a
 particular time of interest would  be estimated by a least-squares  estimate through
the calibration data points.

 Calibration  of AUTOCAL Sources - Transfer Standards
The response  and zero data from the AUTOCAL  procedures (NO, NOX, and 03) are

                                      30

-------
 Hydrocarbon Sources                                                       Calibration

 plotted and listed using a DATATRIEVE database. The first QA step is to investigate
 the large variations or odd values.  Sometimes  a calibration was not conducted and
 regular experimental data was processed as a calibration. A regression is calculated
 to estimate the decay rate and to  estimate the intercept (correct response) at the
 beginning of the season  or start of a new span  check  tank.  The average (or if
 a trend was observed,  the appropriate time specific)  manual calibration factor is
 used to calculate the true concentrations of the AUTOCAL calibration sources from the
 responses at these times (see  bottom of Figure 4and discussion of routine NOX and
 03 calibrations for final report of AUTOCAL calibration factors).

 Hydrocarbon Sources
 A procedure similar to that performed for the different NO and 0$ calibration sources
 was performed for  the  hydrocarbon calibration sources.  The comparison was far
 more complex because of the large  number of species and different sources utilized.
 Five commercially prepared tanks purchased by UNC were compared and analyzed.
 Two additional well-characterized  (several years history) tanks which are part of
 an on-going study of commercial HC  calibration sources  were borrowed from RTI.
 A low concentration tank was prepared from  one of the purchased tanks.  These
 were also compared. Liquid injections into the smog chamber used as calibration
 sources were also compared with these tank sources.  The low concentration tank
 was calibrated as a  working AUTOCAL source. All of the tanks, however, were used
 occassionally for calibration.  A list of the sources that were compared  is given in
 Table 7.

    The borrowed RTI  tanks  were considered to be an  absolute standard. The
 values of the other calibration sources were "analyzed" based on the RTI  tanks as
 the primary standard. In most cases the final analyzed  values for all compounds for
 all sources were within a few percent of the manufacturers reported concentrations
 and the liquid injected  sources from  the UNC chamber.  Only one compound (2-
 methyl-l,3-butadiene ) was outside of the manufacturers stated uncertainty. This
 compound is not important to the present research program.

    The intercomparisons of the HC  sources was performed with a LOTUS 123 spread-
sheet on an  IBM PC. The spreadsheet is listed in Appendix B.

    Table  8 shows the agreement with the manufacturers analysis after "validating"
the tanks  with the RTI NBS-traceable tanks (see Table 7). For each tank, the aver-
age and standard deviation of the ratio of the  "validated" to manufacturers stated
concentration for each species in the tank are listed. The range of agreement is 0.94
to 0.99 with a mean of 0.963 and a relative standard deviation of 1.8 percent; the

                                      31

-------
    Calibration
                                          Hydrocarbon Sources
                  LOTUS   PRINT
                  IBM* PC
                                        •>Ct;
      AUTQCAL
      PROCESSING
CALANA
RESPONSE
  DATA
            INrriAL_AUTOCAL_VALUE
              RATE_OF_CHANGE
                                 (AUTOCALyy)
Figure 2. NOX and O3 Cal Processing Flowchart

                          32

-------
      Hydrocarbon Sources
                                                                                Calibration
       NO-03 Calibration Data Sheet
                                           DATE 8/6/SS
                                                           NAME Suedbeck
       Air Flow:
        volume   0.483    time   0.221    0.223
               0.222
                       0.223   0.221

                             air  flaw
avs 0.22200

    2.18468
       NO Cal Tank Flowttank and 03 turned on
        volume    0.01    time   0.361   0.361
            _30_ min  be-fore  flow rate  measurements)
              0.359    0.360    0.361      av:  0.3604O
                                                                NO  flow
       Predicted NO:   tnk cone 52.453
                     NO Flow
                    0.02775
    0.02775


    0.65783
                                                       2.21243
                                                     NO + Air Flow
                 ZERO READING !ZERO POT SETTING
               	!	
               be-fore/ a-fter! !be-fore/ after!
        NO (1)   -0.015        !     214
                    0.01254

                      SPAN READING !SPAN POT SETTING
       N02 (!)   -0.009

       NOX (1)   -0.013

        03 (2)    0.001
 33

1S1

  0
               !  enter  NA if  no adjustment made
               (1)  zero air with H20 (pura-fil)

       03 SLIDE OUT AND TITRATION BEGINS!

       NO during  titration       0.2133
(2)
(2)
(2)
be-fora/ a-fter!
0.5S6
-0.002
0.642

! be-fora/
! 604
!
i
i
572
714
230
after!



             to pot settings
                    (2)  with NO added
              delta  0.3724
      N02  during  titration
                                0.4112
                                                 delta  0.4130
      Ozone  during titration   0.0665   (measure at  outlet)
       Ozone  with  NO  TURNED  OFF  0.4153
       (wait  for stable value) 	
       (measure at  inlet)
                                0.4162
       (measure at  outlet)     	
                                                       Air  Flow
                                                       2.1S46S
                                            0.41009
                    2.21243
                 NO «•  Air Flow
                    0.98746
       If difference between inlet and outlet is more than  a  few ppb, wait  longer.

       Ozone  (delta corrected for NO dilution):  delta 0.34363

       03 rotoball setting           3 Ethylene pres.       15 02 pres.     20
Figure 3.   LOTUS Spreadsheet for Manual Cal Processing

                                             33

-------
  Calibration
                                                                     Hydrocarbon Sources
  SPAN  CALIBRATION  FACTORS

     NO    1.0944    -  predicted  NO /  (NO span - NO zero)
    N02    0.9869
    NOX
     03
            1.0079
           (delta NO x NO span cal factor)  /  delta  NQ2
           (predicted NO + ((N02 span - N02 zero) * N02  cal  factor))  /
               (NOx span - NOx zero)
1.1861   - (delta NO x NO span cal factor) / delta 03
  IF instrument adjustments ara to be made  -  these  are desired span values
  ADJUSTMENT ORDER: 1) ZEROS, 2) NO SPAN, 3)  NOX  SPAN, 4)  N02 SPAN, 3)  03 SPAN
  NO span value

  NQx span value


  N02 span value


  03 span value
         • predicted NO + NO zero

         - (predicted NO + ((N02 span - N02 zero) * N02 correction))
                 NOx zero

         » (delta NO x NO span correction) +• N02 zero
                       (should be 1.0 now)     (should be 0.0)

         = (((03 value during  titration x 03 span correction) +•
               (delta NO :<  NO  span correction)) x DCF) +• 03 zero
                      DCF
                           (should be 1.0 now)

                             NO -i- Air Flow
                               2.21243        !

                               2.1S468        !
                               Air Flow       !
                                                                   (should  be 0.0)
                                                                   1.01270
  03 rotoball  setting  	

  Time that 03  instrument adjustments made or NA

  Time that NOx instrument adjustments made or NA
  »*»»»»»****»»»»***»•»*»**#*****»**»**»»**»»*»»»***<
          AUTO ZERO               AUTO SPAN DATA
                                  TANK    AB2473
                                  SPAN    CORRECTED
                                                 »•»•»»»•»»**»•»•»•»»
                                               AUTO SPAN & CQRR
                                               TANK
                                               SPAN    CORRECTED
   NO (1)  -0.0202

  NQ2 (1)  -0.0068

  NOX (1)  -0.0217
   03 (1)  0.00088
                   (2) 0.60751 0.68706

                   (2) 0.00023 0.00693
                      '	 0.69400
                   (2) 0.66927 0.69631

                   (3) 0.10S82 0.12446
(2)

(2)

(2)

(3)
          (1)  regular  auto zero   (2)  be sura to check for excess flaw rate
          (3)  regular  auto span gonerator
Figure 3, cont.  LOTUS Spreadsheet for Manual Cal Processing.
                                        34

-------
Hydrocarbon Sources
Calibration
MANUAL CALIBRATIONS
2l-Hay-35 - last update
52.4523 ppi NO Tank

Cal
Date
02-Jul-84
10-Jul-84
li-Jul-84
lS-Jul-84
lB-Jul-84
18-Jul-84
29-Aug-84

day zero
02-Jul-84
0
8
14
16
16
16
58

real

0.7482
0.7693
0.78355



0.7835
NO
delta
response
0.7710
0.8360
0.8540



0.8430
NOx
real/
response
0.9704
0.9208
0.9175
0.9515
0.9615
0.9443
0.9294
real
delta
real/
real
response response
0.7521
0.7707
0.7836



0.7835
0
0
0



.8170
.8810
.9000



0.8840
0.9205
0.8747
0.8706



0.8863
03
delta
response
0.2640 0.2344
0.2385 0.2403
0.2340 0.2520



0.2172



0.2387

real/
response
1.1260
0.9926
0.9284



0.9100
LEAST SQUARES FOR SPAN NO
N
1
1
1
1
1
1
1
7



I
0
a
14
16
16
16
58




Y
0.97042801
0.92080143
0.91750585
0.9515
0.9615
0.9443
0.92941874




m
0
8
14
16
16
16
58
128
AVE CAL:
XBAR:

N*Y
0.97042801
0.92080143
0.91750585
0.9515
0.9615
0.9443
0.92941874
6.59545404
0.94220772
18.2857142

YDEV2
0.00079638
0.00045822
0.00061018
0.00008634
0.00037219
0.00000437
0.00016355
0.00249127



est y
0.9491
0.9461
0.9438
0.9431
0.9431
0.9431
0.9273





SO:


12
o-o
64 0
196 0
256 0
256 0
256 0
3364 0
4392 6
0,


Y2
.94173053
.84787528
.84181699
.90535225
.92448225
.89170249
.86381919
.21677899
.01886519


KY
0
7.36641148
12.3450819
15.224
15.384
15.1088
53.9062870
119.334580
SLOPE:
INTER:
R2:
rel std dev
1.13
-1.34
-1.39
0.45
0.98
0.07
0.11

-3.744E-04
0.94905347
0.11541280
LEAST SQUARES FOR SPAN NOx
N
1
1
1
1
4



X
0
a
14
58

*


Y
0.92050183
0.87474460
0.87061111
0.88631221




NtX
0
8
14
58
80
AVE CAL:
IBAR:

N*Y
0.92050183
0.87474460
0.87061111
0.88631221
3.55216977
0.88804244
20

YDEV2
0.00047114
0.00455127
0.00512607
0.00312430
0.01327279

est
0.
0.
0.
0.


Y
8925
8907
8894
8796

SD:
12
0 0.
64 0.
196 0.
3364 0.
3624 3.
0.
Y2
84732363
76517812
75796370
78554934
15601481
05760381
XY
0
6.99795636
12.1885555
51.4061085
70.5926210
rel std dev
0.49
-0.28
-0.33
0.12

SLOPE: -2.227E-04
INTER: 0.89249673





R2: *
0.06530589
Figure 4.  LOTUS Analysis Spreadsheet of Manual Cal
                                       35

-------
  Calibration
                                                                                     Hydrocarbon Sources
 LEAST SQUARES FOR SPAN 03
         N         I          1
N*»
NtY
YDEV2    est y
Y2
JY rel std dev
1
1
1
1
4

AVE CAL:
NO
0.94221
NOx
0.88804

03
0.98926
0 1.12601851
-8 0.99243298
14 0.92837585
58 0.91000502

AVE CAL:
XBAR;
Auto Tank 1 : 6/21 to 9/1
responsel CAL VALUE
0.30515 0.28751
0.!
responsel CAL VALUE
0.32207 0.29401
Auto Cal Gen : 5/30 to 11/4
response CAL VALUE
0.12218 0.12087
0 1.12401851 0.03378640
8 0.99263298 0.00254270
14 0.92837585 0.00019132
58 0.91000502 0.00103701
80 3.95703238 0.03755744
0.98925809
20.0000 .

decay*
-4.485E-05
531 agreeient
decay*
-5.304E-05

decay*
-7.179E-05
1.0428 0 1.26791769 0 0.86
1.0214 64 0.98532023 7.94106384 -0.30
1.0053 196 0.86188173 12.9972620 -0.79
0.8875 3364 0.82810915 52.7802916 0.23
3624 3.94322881 73.7186174
SO: 0.09689872 SLOPE: -2.479E-03
INTER: 1.04283546
R2: 0.50605026
Auto Tank 2 : 9/1 to 10/16
response2 CAL VALUE decay*
0.70602 0.66522 -5.687E-04
-0.423: agreement
response2 CAL VALUE decay*
0.75224 0.66802 -3.688E-04


responsel -response value for  21-JUN-84 determined froi least squares regression of response data

response2 -response value for  l-SEP-84 detertined froi least squares regression of response data

decay*   -froi least squares  regression of response data

CAL VALUE -VALUE ENTERED INTO  AUTOCAL DATABASE TO GENERATE DAILY CAL FACTORS



Figure 4, cont.    LOTUS Analysis  Spreadsheet Manual Cal Plots.
                                                  36

-------
MANUAL CALIBRATION -  SPAN  - NO
MANUAL CALIBRATION -  SPAN  - NOx



PS

u
k.
X
8
jj
o
u
I





1.4 -
1.3 -


1.2 -

1.1 -

1 -
(
0.0 -
O.8 -
0.7 -
o.e -
o.s -\









^ < 1
^-a 	 a- •






PS
o
B
f
2
s
B
o
u
1


o 20 40 eo
DAYS
D data + lln »f
l.S -
1.4 -
1.3 -


1J8-

1.1 -

1 -
o.o -!
0.8 -
0.7 -
o.e -
A K










>- 	 ^^



0 2O 40 61
DAYS
a data •* lln r«f
                       MANUAL CALIBRATION - SPAN -  03
                   o.s
             Figure i.  LOTUS Analysis Spreadsheet, of Manual Cal Plots

-------
u
I
        MANUAL CALIBRATION - SPAN - NO
    1.0
1.4 -
1.3-
IJt-
1.1 -
 1 -
0.0 -
0.8 -
0.7-
0.0
0.0
                            leas
                —I—
                 20
                     data
                             40
                            DAYS
                                   —I—
                                    00
                                                  80
                                                          O
                                                          s
o
u
I
09
                                                              MANUAL CALIBRATION - SPAN - NO2
                                                                                  ieoo
                                            1.4 -
                                            1.9 -
                                            1.2-
                                            1.1 -
                                              1 -
                                            o.o -
                                            o.a -
                                            '0.7 -
                                            0.6-
                                            0.9
                                                                            20
                                Un r«(
                                                                               d»U
                                                                                   40
                                                                                  DAYS
                                                                                                  00
                                                                                                             BO
                                                                                          Un i
 kt
 X
 O
 u
 1
       MANUAL CALIBRATION - SPAN -  NOx
    1.0
 1.4 -
 1.3-
 1.2-
 1.1 -
  1 -
 0.0 -
 0.0-
 0.7 -
 o.e
 0.0
                            leas
20
 a  data
                             40
                            DAYS
                                        eo
                                                   80
                                                          a
                                                          H
                                                          U
                                        O
                                        u
                                        0.
                                        01
                                                              MANUAL CALIBRATION -  SPAN -  03
                                                              1.9
    1.4 -
    1.3 -
    1.2 -
    1.1 -
     1 -

    0.0-
    0.8 -
    0.7 -

    o.e -
    0.9
                                                                                      less
                                                                           20
                                Un
                                                                               data
                                                                                   40
                                                                                  DAYS
                                                                                   t  Un r»f
                                                                                                  OO
                                                                                                             eo
                            Figrire 5, cont.  LOTUS Analysis Spreadsheet of Manual Cal.

-------
Hydrocarbon Sources                                                           Calibration

largest individual relative deviation from the average is 2.5 percent. The agreement
between the standard tanks and the RT1 tanks is very good.  The absolute accuracy
of the hydrocarbon calibrations depends directly upon  the concentrations  assigned
to these tanks.
                 Table 7. Intercomparisons Among Sources
          Source                Source                  Date

          RTI ethylene          AIRCO                   8/1-2
          RT1 propylene         AIRCO                   8/1-2

          AIRCO                 LOV/MV (Hi Cone)          7/18
          AIRCO                 HUM (Hi Cone)           7/18

          AIRCO                 LIQ STD 8/14            8/14
          AIRCO                 HIMV; (Hi Cone)           8/14
          LIQ STD 8/14 (C/area) HIMVJ (Hi Cone)           8/14
          LIQ STD 8/14 (C/inch) HIMV! (Hi Cone)           8/14

          LOVJMV: (Hi Cone)       LOV.'MVJ (Lo Cone)          7/13

          LOV/HV; (Hi Cone)       HIMV.' (Hi Cone)           5/30
          LOV/MV! (Hi Cone)       82HCCALMIX              5/30

          LOV/MW (Hi Cone)       SCOTTTAIIK               10/5
          HIM!;; (Hi Cone)         SCOTTTAIIK               10/5
          LOV/1W (Lo Cone)       SCOTTTAIIK               10/5

          LOVJMV (Lo Cone)       SCOTTTAIIK               8/17
          AIRCO                 SCOTTTAIJK               8/17
                                       39

-------
 Calibration	Routine NOx and O3 Calibrations Methods


        Table 8. Comparison of Primary HC Calibration Sources
            DCS  ID  SOURCE       ANALYZED/        DEV  REL DEV ABS REL DEV
                              MANUFACTURER
                                avg      std


                                                         0.8%     0.87.
                                                         2.3%     2.3%
                                                        -0.7%     0.7%
                                                        -2.5%     2.5%


                   ave        0.963    0.077                       1.6%
                   std        0.017    0.038                       0.8%
                   rel std      1.8%
1
108
109
112
82HC
HMWHC
L1WHC
AIRCO
0.971
0.985
0.956
0.939
0.126
0.047
0.034
0.101
0.008
0.022
-0.007
-0.024
Routine  NOx and O3 Calibrations Methods

This section describes the types of calibration methods routinely performed.  The
section "Calibration Data  Processing"  describes how  the  data are processed, in-
cluding the  QA steps. The calibration reports are included or refer to appendices
of calibration reports.

Automatic Source  Sampling
The transfer technique to characterize the AUTOCAL sources  for NOX and Os was de-
scribed above.  Before and after every experiment, these sources plus the AUTOZEROs
are used to calibrate the instruments by scheduling 30  minutes to sample from the
span sources and from the zero sources by computer control (0200-0230 EDT, 1800-
1830 EDT for AUTOZERO; 0230-0300 EDT, 1830-1900 EDT for AUTOSPAli). The two low
concentration NO tanks used for AUTOCAL sources are listed in Table 2. The generator
built into the Os monitor is used for the 03 AUTOCAL source.

    DATATRIEVE is used to maintain an AUTOCAL  database. In this database, the de-
cay rates and the initial concentrations  at the beginning of the season are used to
estimate the concentration of the span sources for each AUTOCAL performed. A data-
base procedure PRUIT.CALSJVIID-FACTORS computes calibration factors from the AUTOCAL
response and the computed tank concentration.

                                      40

-------
 Calibration	Routine HC Calibration?

 Routine HC Calibrations

 Automatic Source Sampling
 Routine HC calibrations were performed automatically before every experiment.
 The GCs are  activated several hours before injections and are allowed to  produce
 several chromatograms. A timer switches on the calibration source tank connected
 to a calibration manifold at approximately 0300 EDT. The flowrate is adjusted to
 provide an excess which is vented.  The timer also switches  a three-way  valve to
 change the GCs from the chamber sample manifold to the calibration source man-
 ifold. The "low concentration" tank listed in Table 3  and  discussed above was the
 source most commonly used for an AUTOCAL, however other tanks  listed in the table
 were sometimes used. This would normally be when the presence of a morning site
 operator was required many hours before sunrise.  Sometimes calibrations were per-
 formed after an experiment when a wider range of calibration species were desired.
 The site operator indicated on the instrument checklist that a. calibration was per-
 formed. When the run is logged in later in the data processing office, these notes
 of calibrations performed will be logged into  the calibration database  (discussed
 below).

 Manual Precision Liquid Injections
 More extensive calibrations were performed on days that were poor for performing
 an experiment (poor sun or rain forecast).  On these  "CAL" days, several tanks
 would be used for calibrating. More importantly, liquid standards could be  injected
 into the unused chamber for calibrating compounds with vapor pressures too great
 to be stored in pressurized cylinders (e.g.l, 2,4-trimethylbenzene).  These calibration
 data are stored in a "CAL" folder. The presence of the  "CAL" folder and a calendar
 entry indicating a "CAL DAY" will trigger the processing of a calibration that was
 not performed during an experiment.

    Liquid injections of reasonably large amounts (100's of /*!) used to produce initial
 injections for a smog chamber experiment were  usually processed as calibration
sources  also. The project coordinator decides if the injections  are appropriate as a
calibration source and then logs in the calibration.

Injection Calibration Procedures
The chamber  is usually vented more than 10 hours  before injections are made.
Hamilton  "Standard Microliter "  gas chromatography syringes are used to inject
known volumes (in  10's to 100's  //I)  range) into the smog chamber. Stated accu-
racy is 1% of the syringe volume. The temperature is recorded for concentration

                                     41

-------
 Calibration                                              Meteorological Sensor Calibrations

 calculations. All data are recorded on the stripchart or by a special note on the RUN
 SHEET or IliSTRUUEl.'T CHECKLIST.

    Injection calibration data are transfered to the Official Calibration Source form
 (OCS), then to the electronic OCSIC84  form which performs ideal gas law  calcu-
 lations, using the  amount injected and the chamber volume.  The electronic form
 automatically looks up density, carbon number, MW etc.A.1] calculations are thus
 recorded. Errors (from improper data transfer or calculation) are therefore mini-
 mized.

    The determination of chamber volume was by a "strapping" technique. The di-
 mensions of the chamber were carefully measured and used to calculate the volume.
 Several people have independently made these measurements with less than 1.5%
 variation in volume.

 Meteorological Sensor Calibrations

 Solar/UV radiation Calibrations
 Two sets of both the Eppley Black and White  Pyranometer (Total Solar Radiation)
 and Eppley Ultra-Violet  Radiometer (Ultraviolet)  sensors are used.   One  set  is
 used continuously  and the other  set is reserved  as  the References.  Both sets are
 calibrated by  Eppley.  The  continuous sensors  are sent back to Eppley annually
 for recalibration. At Epply, the total solar radiation sensor is compared with the
 Eppley  group of reference standards at approximately one half a solar constant.
 The UV sensor is  calibrated by comparison with an Eppley standard  of spectral
 irradiance lamp.

 Temperature Calibrations
 The temperature thermistors are checked against an ice bath and a water bath near
 normal high afternoon temperature, measured with a general laboratory mercury
 thermometer.

 Dewpoint Calibrations
 There is no direct calibration technique for the dewpoint monitor. The  sensor is a
high precision point thermometer sensor.  The electronics of the instrument  must
be frequently adjusted to balance the  sensor bridge.  The  detector must also be
cleaned  routinely. The dewpoint meter is checked by monitoring the saturated air
pulled through a water bubbler of known temperature.  Response to dry air (air

                                     42

-------
 Paper Forms	Calibration

 pulled through Drierite) tests for proper cooling circuit operation needed to reach
 low dewpoints.

 Flowrate Measurements -  Calibration

 Accurate flowrates are required for the manual gas phase titration method for cali-
 brating NOX and 03 and for making injections into the smog chamber. The primary
 standards and routine flowrate measuring devices are a series of 10. 100, 1000, and
 2000 milliliter soap bubble flowmeters.  The volume  of the soap bubble tubes was
 determined by measuring water volume in NBS calibrated graduated cylinders. A
 four place digital stopwatch is used for timing.

 Time Measurement  - Clock Calibration

 All times reported are Eastern Daylight Time, even during winter. The main system
 clock is housed within the computer. Time is set at the beginning of the run season
 against standard time reported at the local weather bureau at the Regional Airport.
 This time is verified several times during  the year. Stripcharts are marked  at the
 beginning and throughout the experiment  with time read from the computer clock.

 Calibration  Documentation and Processing Databases

 A formal system has been developed to document and track the processing of the
 large calibration  database.  This is important to a  QA program to assure that
 all data  is processed and utilized. Various paper and electronic forms, calendars,
 notebooks, and computer databcises are utilized.  Forms help assure consistency
 and completeness. Electronic forms allow for ease of organization, manipulation,
 retrieval, and analysis. Some of the form  fields are actually computed internal to
 the form.

 Paper Forms
 Run and Calibration Calendars
 Calendars at both the site and the office in the School of Public Health are used to
record days on which experiments were conducted. These entries also indicate days
on which only calibrations were performed.

Run Folder Inventory Check Lists
Instrument Checklists are filled  out for each experiment. These lists identify the
calibrations that were performed  (see Appendix A).

                                    43

-------
 Calibration	Electronic Forms and Other Documentation

 Stripcharts

 Stripcharts from an experiment identify the calibrations that were performed, the
 calibration source, the chart attenuation, the chamber temperature, and the number
 of microliters that were injected.


 Manual NOX/03 Calibration Form

 Paper forms are used to record manually performed calibrations.  These are filled out
 at the site and transported to the data processing office where they are transfered
 to electronic form (see below) and stored.


 Official Calibration Source Form

 All information, including the calibration  source concentration  used  for a calibra-
 tion, is recorded on Official Calibration Source (OCS) forms (see Figure 6).  Infor-
 mation concerned with several stages of processing and several different programs,
 worksheets, and databases is recorded. The calibration information comes from sev-
 eral sources including Stripcharts and attached notes. This form has an id number.
 identification information, amounts and temperature, and official validated concen-
 tration. This information is further stored  in a computer database described  below
 which generates a file for processing calibration factors.


 Calibration Pick Instructions

 A paper form is used to give and record processing instructions to Data Proces-
 sors for the raw calibration response data. The location of the calibration data is
 recorded and the corresponding calibration sources used are identified. Notes can
 be made about processing status and special  problems and processing instructions.
 Processing status is recorded on  this form and concurrently in CALEllTRY. a computer
 database for tracking processing status (see below). Figure 7 shows this form.


Electronic Forms and Other Documentation

 COMMAND FILE

 Command files are created at the smog chamber site on the computer used to
control the experiment.  Commands issued to perform auto cals can be used later
to document the exact time the instruments are calibrated. Other instruments not
directly controlled  yet auto calibrated are identified in remarks in the command file.

                                      44

-------
     Electronic Forms and Other Documentation
                                                                Calibration
           OfRcal  Calibration  Source  Form
   Who is filling out this form:
   Date form filled out:
   SOURCE INFORMATION.
              INSTRUCTIONS:
              ID NUMBER:The id number is the next sequential calibration source
              number.
              DESCRIPTION: examples are: 82SCOTHC, or STDMIX2INJ. Please
              print.
              SOURCE DATE: the date tank was acquired or the date of in-
              jection.
              SERIAL NUMBER: the tank  serial number or the initials of the
              injector. Please print.                               /
              VALIDATION OF SOURCE:
              VALIDATION DATE The date that the source was approved.
              INITIALS The person doing the approval.
  ID NUMBER
Three digits.
  DESCRIPTION    Ten chars.
  SOURCE DATE   dd-mmm-yy

  SERIAL  NUMBERTen chars.
Figure 6.  Official Calibration Source Form
                                    45

-------
   Calibration	     Electronic Forms and Other Documentation
SOURCE VALUE :: one page per species.Page	of





SPECIES NAME      Full name  	




SPECIES NUMBER   Three digits          	




STATED CONG.      9.9999               	




ACTUAL CONG.     9.9999               	




CONG. UNITS        e.g. ppmC             	:	
VALIDATION DATE dd-mmm-yj




INITIALS
COMMENTS:
 Figure 6, cont.  Official Calibration Source Form.
                                 46

-------
       Electronic Forms and Other Documentation
                                                                                   Calibration
                      CAL PROCESSING INSTRUCTION FORM


CAL DATE 				          INSTRUMENT  		


CAL INFO IN:    RUN FOLDFR  	  OR CAL FOLDER	


CAL SOURCE  -NAME  		 -10 NUMBER	


DIGITIZE (OR -PICK) CAL  STARTING  (TIME) 	.	   ENDS


M E R 0 E   UITH					


COMPOUND NAME  ID  *   SIDE       ATTN
                              (AND UNITS)       NOTES




























          DATE     INIT

DIGPIKR  	   	

CALFAC   	   	

RTTOWAX  	   	

TO OTR   	   	
                                                                        	of	
Figure 7.  Cal Pick Instruction Form
                                               47

-------
 Calibration                          '    Calibration Data Processing Status Database (CALEHTRY)

 REMARKS FILE from Computer DA system
 Another file created during the experiment by the computer is the remarks file
 which often has comments about calibration events, especially  if deviations from
 the normal procedures are made.

 Manual NOx/Os Calibration Form
 Information from a paper form filled  out during  a manual NOx/Os calibration is
 manually transferred to electronic form (in LOTUS 123 spreadsheet, see description of
 general  purpose programs below) where the calculations are performed automati-
 cally.

 Official  Calibration Source
 A computer database of the Official Calibration Sources also recorded on paper form
 is maintained on a DEC  VAX.  This database can be summarized by a database
 procedure and output in a form used by another program CALFAC to process
 calibration factors when compared to the response data from a calibration.

 Injected Conditions as Official Calibration Sources
 Information from paper forms (primarily from the  OCS  paper  form) of injected
 liquids and  solids for calibration is input,  and processed in electronic form (in LOTUS
 123 spreadsheet) where calculations are performed  automatically  (OCSIC). The
 spreadsheet looks up the molecular weight, density, and carbon number using the
 species id number (from the Official Species List), temperature at time of injection,
 the chamber volume, and  the ideal gas  law. It was shown earlier  as  Table 4.

 Calibration Data Processing Status Database (CALEHTRY)
 A calibration data processing status  database called CALENTRY is maintained on the
VAX-il/780 using the DATATRIEVE database system.  The AA or PC logs in the existence
of calibration data for a given instrument and day. The PC and PP update CALE1JTRY
as the various tasks are performed. DATATRIEVE report procedures  are used to assist
in updating the database, and generating status reports.  The PC  and PP use the
status reports to guide the processing  of the calibration data. Table 9 shows the
CALEHTRY  tasks flowchart and the events explained.  Table 10 shows  an example of
the CALSUM report.

Calibration Databases
Several calibration databases are maintained. Some double as spreadsheets to per-

                                     48

-------
Calibration Databases
                                                                         Calibration
                                    Table 9
        Processing System for Instrument Auto-Calibration Data
          tot inst    cal analysis
\
 !CPI!
     \
                      \
                       -
                                  \
                                  IOCS!
                                      \
                      \
                     ---(Q2V)--(DTR)--(C2L)----!CQA!
          Step
                     !sss!  ==  stages determined by Project coordinator (PC)
                       ==  stages determined by Peak Pickers (PP)
                     (sss)  ==  stages determined by Computer Techs (CT)
  CALENTRY stages
Meaning
PC
PP
PC
PP
CT
CT
CT
PP
PC
PC
PC
CPI
DPR
DCS
CFR
Q2V
DTR
C2L
CLR
CQA
NTB
BAD
CAL pick Inst
Digpik run
DCS ok
CALFAC run
Qfile to vax
DTR file updated
Cals to LSI
CALA11A run
Cal QA
Hot to be proc
Stop processing
Charts marked, cal status sheet in folder
Cal data digitized and R-file exists
Official Cal Source exists in DTR
Cal Factor computed, Q-file created
Q-file moved to VAX
Cal Data added to DTR data base
DTR CAL file loaded on LSI-11
Cal Factors analyzed
Cal Factor Quality Assurance
Not to be processed
Something is wrong with this run
                                      49

-------
Calibration	Calibration Database?



    Table 10. CALSUM Example



   Count of calibrations that have reached  stated  stages  of  processing
.  . '           for METH

   Total calibrations

          265


   ******** HC Instrument data ************

   CPI DPR OCS CFR Q2V DTK C2L CLR CQA C2S NTB BAD

   181 181 181 181 179 179   0   0 144   0  53   3

        Sum of CPI and NTB Is  234
                     CAL  PROCESSING  STATUS               6-May-1985
                                                        Page 5
        CAL                                   Calibration
        DATE   Instrument                         Steps

        Carle  III GC                    CAL  pick Inst
                                        0 Igp 1 k  run
                                        CALFAC  run
    13-Jul-1984
        Carle  I   GC                    CAL  pick Inst
                                        D Igp 1 k  run
                                        CALFAC  run
        Carle  II  GC                    CAL  pick Inst
                                        D Igp 1 k  run
                                        CALFAC  run
        Carle  III GC                    CAL  pick Inst
                                        D Igp Ik  run
                                        CALFAC  run
        CEA Formaldehyde                CAL  pick Inst
                                        D Igp Ik  run
                                        CALFAC  run
    14-Jul-1984
        Carle  I   GC                    CAL  pick Inst
                                        D Igp 1 k  run
                                        CALFAC  run
        Carle  II  GC                    CAL  pick Inst
                                        D Igp 1 k  run
                                        CALFAC  run
        Carle  III GC                    CAL  pick Inst
                                        DIgp I k  run
                                        CALFAC  run
    16-Jul-1984
        Carle  I   GC                    CAL  pick Inst
                                        D Igp 1 k  run
                                        CALFAC  run
        Carle  II  GC                    CAL  pick  Inst
                                        D igp 1 k  run
                                        CALFAC  run
        Carle  III GC                    Not  to  be proc
                                    50

-------
 Calibration Databases	Calibration

 form calculations. The databases serve several functions. Database commands and
 procedures can sort by  instrument, and species, make plots of cal-factors vs time,
 perform regression and  correlation calculations, convert, to a common attenuation
 or range, output  reports, output files for other programs, and in general allow us
 to organize, retrieve and update information easily.


 HC Tank Calibration Source Certification Database ("CAL")

 Many different calibration sources were used for hydrocarbons. Some are "cal" tanks
 and others are analyzed by the manufacturer with a stated uncertainty. Liquids
 injected into the  chamber are another calibration source.  All "cal" sources were
 intercompared to  check for consistency. Data was collected in both peak height and
 integrator area, and organized in a LOTUS  123 spreadsheet database. Calculations of
 the intercomparison were then performed (see Appendix B).


 Official  Calibration Sources (OCS) Database

 The Official Calibration Sources (ocs) Database were mentioned earlier in the elec-
 tronic forms section. The database utilizes DATATRIEVE on the VAX-11/780. The data-
 base appears as a collection of forms on  terminal screens which  makes  updating
 easier. Procedures have been written to generate reports (see  Appendix  C)  and a
 file of the calibration source data required by the program CALFAC used to calculate
 the calibration factors.


 Gas Phase Titration Database ("MANCAL" LOTUS  123 )

 The resulting calculations (calibration correction factors) processed from the raw
 calibration zero and response data are summarized and analyzed in another data-
 base/spreadsheet using LOTUS 123. Plots of calibration data vs time with regressions
 and correlation are produced easily.  Trends and variation are easily recognized.
 Reports  are quickly produced.


 NOX and Oz Auto-calibration Database: AUTOCAL

 The NOX and O3 AUTOCAL data  are  maintained  in a separate DATATRIEVE database.
 Instrument  calibration zeros and responses  to  the AUTOCAL sources are recorded.
 Report procedures compare the corrected responses to the known calibration source
concentration and  generate and list the calibration factors, sorted by date and time
of day  (autocals for NOX  and Os are usually performed twice a day).

                                      51

-------
 Calibration	Security of Calibration Databases

 Hydrocarbon Auto-Calibration Database

 Another DATATRIEVE database similar to the NOX and Os AUTOCAL database is main-
 tained for hydrocarbon calibration data. The database is more complex because of
 the many more instruments which are calibrated. The database is also much larger.
 A calibration factor is recorded for each species. Attenuation is also recorded.  In-
 teractive report procedures can  retrieve and sort calibration  data  by  instrument,
 species, and date, and convert calibration data to a common attenuation.  This
 report is the primary source of final calibration data used to process data. Another
 procedure outputs a file which can be read by another custom program  (CALANA)
 which statistically analyzes calibration data for a given species instrument combi-
 nation and plots the calibration history for the run season.
 Calibration Factors Used Database
                                                   r
 Another DATATRIEVE database is used to record the calibration factors actually used
 to process the final experimental time-concentration data.  Some interpretation and
 choice must be made when the statistics and the various sources of calibration fac-
 tors do not agree. The PC makes a judgement based on all calibration information
 and chooses a calibration factor.  The chosen calibration factor is kept in a separate
 data base. The resulting calibration factor database is maintained as a final source
 of information for making a choice in  a calibration factor. We believe that except
 for rare occurences, calibration factors  for most instruments do not vary much. This
 database is consulted to see if the calibration factor chosen varies much from those
 used on data from days (experiments) before and after the current experiment being
 processed.
Security of Calibration Databases

There are three basic approaches to data security: inventory, retention of original
data in restricted areas, and maintaining several backup copies of all data in different
locations. All data are under lock at night and on weekends.
Inventory
At the site during an experiment, the calibrations that were performed are indicated
on a Instrument Checklist Form. Afterwards, before anyone is allowed to work with
the run folder contents, a run folder inventory  is  performed which documents all
data and forms. This assures that no data or documentation are lost.

                                      52

-------
 NOX and 03                                                              Calibration

 Retention of Data
 All original data in both paper and electronic files are retained.  Paper format data
 (all stripcharts, forms) are not allowed out of the School of Public Health office.

 Backup of Data
 Multiple  copies of data are maintained  in several  formats,  in several locations.
 Backup copies of raw and final data on floppies on IBM PC are kept in several loca-
 tions.  LOTUS  123 worksheet files are backed up in two locations. A hardcopy (print-
 out) of worksheet formulas  using a spreadsheet AUDITOR program is  kept.  Backup
 copies of raw and final data on LSI-11 computers are maintained in several locations,
 including: raw data and processed data sets backed up on two diskettes, processing
 programs (sources) backed  up on two diskettes, and whole image  backup of LSI
 hard disk. Backup copies of final data are maintained on the VAX and  are further
 backed-up on tapes of VAX files.  There are also frequent printouts of final calibration
 data in notebooks. The school operates a daily, weekly, and monthly backup of VAX
 files with  off-site vault storage for ten years  for all VAX data.

 Calibration Data Processing

 This section  describes and illustrates the actual data processing procedures. Pro-
 cessing techniques are discussed separately for NOx/Os and HC data,  because of
 the different  data and processing requirements and the different type of calibration
 sources. The databases are also maintained separately. Another  important distinc-
 tion is the way the different types of calibration data are transported from the smog
 chamber to the data processing office in the School of Public Health.

 NOX and O3
 Manual Calibration Processing for Characterization of AUTOCAL Sources
 The calibration sources used for the NOX and Os monitors were discussed above. Also
 discussed was the relationship between the manual and auto calibrations performed.
 The manually performed gas phase titration calibrations are used  to calibrate the
 auto calibration sources.

AUTOCAL Data Processing
AUTOCAL data  for zeros and spans is obtained twice daily. The PDF/11-40 computer
at the site initiates the calibration from a COMMAND file written by a site operator
 (30 minutes zero; 30 minutes span). The calibration data is recorded along with the
experimental results by the comptiter DA system (one value every minute). Therefore

                                     53

-------
 Calibration	HC Processing

 the calibration data is transported along with the electronic data. Figure 8 shows a
 flowchart of the autocalibration data processing procedures. These tasks are mostly
 performed by the CT.

     A computer program was written (FASTDV) to strip the calibration data from the
 DVM data file (see Figure 8). One file is made for every AUTOCAL performed.  There are
 usually  two autocals per day.  The files are named MMDDYA.DVn where n is 1st
 and 2nd calibration performed. These are called "A-files".  The files are backed-up
 on floppies. Each file is edited if necessary to eliminate readings taken in  transition
 to the calibration value.  These edited files are named MMDDYE.DVn ("E files").
 These are also backed-up on  floppies.  They  are then appended together into a
 file called AUTCAL.NEW. This file is appended to AUTCAL.DVM which is the
 "OFFICIAL" file of appended "E files", which is also backed-up on floppies.

     The AUTCAL. HEV.' file is further processed. It is averaged to a single value for both
 zero and span with the program called ACALAV on the  LSl-ll/23. A  listing of the
 processed averaged data is made and stored in  the NOX/O3 FASTPLOT notebook.  This
 program generates a file called ACDTR.COH which contains the processed averaged
 calibration data  with additional DATATRIEVE commands. The file is transferred to
 the VAX and activated as a command procedure which causes  the data to be auto-
 matically entered into the DATATRIEVE AUTOCAL database.  The database also contains
 the characterized concentration values of the AUTOCAL sources (see discussion above).
 A report procedure in DATATRIEVE called PRIHT.CALSJUJD-FACTORS reports the original
 calibration data  by date and time and reports the calculated  correction factors.
 Table 11 shows the final AUTOCAL calibration factors for 1984.

 HC Processing
 Processing of hydrocarbon calibration  data is more complicated than processing
 NOX and Os calibration data.  Each gas chromatograph monitors dozens, potentially
 hundreds, of different compounds.  Each species used in an experiment requires  a
 calibration factor for each instrument.  Not every species  is  calibrated explicitly
 before every experiment.  Several different calibration sources are used.  Resulting
 calibration data must be reviewed before assigning a value for final data processing.
 An explicit  calibration  source for a compound monitored in an experiment  may
not exist and other techniques must be used to establish an appropriate calibration
factor.  The different types of calibration sources and the documentation, forms,
and data bases associated with this processing  were  described above.  This section
discusses in more detail the procedures of processing the calibration data.

                                      54

-------
   HC Processing
                       Calibration
 .: a '•
 ;S;n8>:f3,
                         XU' file
                   DISK  j mmddyA.DVn

                       EDIT
                         mmddyE .DVn
                       APPEND
                   SLU^i AUTCAL.NEV
                   21^-J AUTCAL.DVM
                    v
                 ACALAV
                    I
        CAL SOURCE \
VALUES
dANCAL
t
stat
rept
/
\
71

c
C


DISK
+
ESTTI:
DISK
J 	 ^
>
J
^
ALANA

^^*

^Crn^ACDTR.COM
 L/IJ3IV
  JL
  AX^»,   , DTR (AUTOCALyy)
                                      REPORT
                                      LISTING
Figure 8.  AUTOCAL Processing Flowchart.

                       55

-------
                               Table  11.    NOX AND 03 CALIBRATIONS FOR 1964
16-M»y-198S
       DATE  NOX ZEPO   NO ZERO  NOZ ZERO   O3 ZERO  NOX SPAN   NO SPAM   03 SPAN  NOX PESP   NO RESP   03 *ESP   NO  CAL    NOX  CAL   O3  CAL
o>
Jun23 -0.0026
JunZS -0.00Z6
JunZS -0.08Z6
Jun?6 0.001Z
JunZ6 -0.0014
Jun27 -0.0013
JunZ7 -0.00Z5
Jun28 -0.0020
JunZB -O.t'019
Jul'JC -0.0002
Jul07 -0.0014
Jul07 -fl.0012
Jul09 -0.0009
Ju)09 -B.0025
Jul 10 -0.0013
JullO -0.0021
Oulll -0.17016
Julll -0.0026
JulZ0 -0.0011
JulZI -0.0003
JulZl -0.0006
JuIZZ -0.e0.37
Jul22 -0.O009
Jul24 -0.0004
JulZ4 -0. H033
JulZS -0.PB08
Jul25 -0.0819
Jul2G -B.O010
Jul26 -0.1016
JulZB 0.0B10
Jul28 0.0023
Jul29 B.H00B
Jul29 -0.P003
Jul3l O.I'OOI
Auo,02 -0.W030
AutjeZ -0.P02I
Au.j03 -0.O018
AugC3 -0.0030
Aug04 -0.C042
AuqLM -0.0832
Auo«5 -0.C039
A.jolTS -0.0034
Aug06 -B.0042
Aug06 -B.1'040
Aug07 -O.f'057
Au'i07 -0.1/034
Augl'B -0.0062
Aug!>8 -0.C038
Au-jlG -0.0008
0.0001
-0.P004
-0.0006
0.0025
-0.0008
0.0005
-0.0009
0.0000
-0.0006
0.0008
3.B01B
0.BO00
0.0013
0.0002
0 . 00 1 0
fl. 0004
0.0008
-0.00.03
0.0007
0.0010
0.0028
0.0013
0.0004
0.0013
-0.0009
0.0039
B . 0 0 'J 1
c.oeod
0.*0u2
e . 03 1 2
0.0012
0.0017
0.0013
0.J08;.'8
-0.0C'05
-0.0010
0.0092
-0.C809
-0.08Z0
-0.0069
-0.0023
-0.0012
-0.0021
-H.CUZ0
-0.00.il
-0.0018
-0.0036
-3.0021
0. «0<71
-0.0015
-0.0010
-0.0004
-0.0004
0.0004
-0.0006
-0.O004
-0..90 10
0.0002
0.0003
-0.0012
-0.0001
-0.0005
-0..0011
-0.0007
-0.3007
-0.0008
-0.B005
0.0002
0.0005
0.0009
0.0005
0.0306
0.0008
0.B003
B.iraee
0 . 0005
0.0005
o.aoes
0..'I018
0.0034
0.0005
0.9068
0.0016
O . Mf 008
0..00B8
0.0004
0./JB04
0.UC02
0.0003
0.0008
0.0005
O.O007
B.0B04
0.0806
0.0004
0.01)04
0.0003
0.0015
-0.0148
0.0008
0 . 00 1 I
0.0008
0.0022
0.0007
0.0B12
0.0007
0 . 00 1 9
0.0002
0.0007
0.0010
0.0007
0.0009
0.0006
0.001 1
0.0006
0.0010
0.0907
0.0096
0.0010
0.0008
0.0012
0.0007
0.0007
0.0006
0.0009
0.0007
0 . 00 1 6
-0.0021
-0.01)08
-0.001 1
-0.0014
0.0006
0.0006
0.0013
0.C007
0.0009
0. 0110 7
0.00U8
0.0008
0.0089
0.0007
0.0088
0.0007
0.0015
0.0007
0.0008
0.0006
0.3259
-9.9999
0.3200
0.3267
0.3190
0.3256
0.3207
0.3248
-9.9999
0.3140
0.3138
0.3105
0.3193
0.3130
0.3163
0.31 19
0.3145
0.3060
0.3205
0.3224
0.3200
0.3229
-9.9999
0.3220
0 . 3 1 3«
0.3200
-9.9999
-9.9999
0.3100
0.3219
0.3237
0.3239
0.3244
0.3193
0.3424
0.3342
0.3249
0.31G3
0.3150
-9.9999
-9.9999
0.3153
0.3141
0.3122
0.3101
0.3111
0.3006
0.3111
0.3230
0.3101
-9 . 9999
0.3061
0.31 16
0.3962
0.3121
B.3t''9
0.3090
-9.9999
0.2988
0.2985
0.2954
B. 31/39
0.2'JBE
0.3H16
0.2274
0.2994
0.2?28
0.3061
0.3'j'77
0.3061
0.3S/1
-9.9999
0.3(171
0.2937
0.3053
- n . 9 11 2 9
-3.9999
0.3030
a. 31! 6 6
0.3079
0.3K94
0.3H93
0.3f.MB
0.3164
0.3155
O. 2/782
.0.3" 14
O.P'ign
-9.9'.!99
-9.9399
0 . 30 1 2
0.2991
a.Z'lUZ
0.2'jr,o
)1.2'.I75
U.2950
0.2973
0.3071
0.1139
-9.9999.
B. 12!i3
0. 1234
0. 1273
0. 1291'
0. 1261
II. 1275
e. 1284
0. 1220
0. 1263
0. 1225
0. 1ZI.-1
0. 1255
0. 1264
a . 1 1 4 r.
0 . 1 1 '. (I
0.1116
0. 11 1 1
0.1142
U.I 146
0. 1175
0.1174
0. 1204
0. 1 H,S
11. 1 IUG
0. 1 1 tip
0 . 1 1 8H
B. 1 173
0. 1 168
H. 1163
il. 1189
0. 1 103
0 . 1 Z 1 fi
0.1212
0 . 1 2«i'
U. I21B
0. 1 IU7
e . i r. i
0. 172
0. 12G
0. 144.
0. 091.
0. 16i:
0 . 148
n . i?;:
0. IG4
H . 1 8 I
0. 12.03
0.3235
-3.9973
0.3226
e.3?ss
0.3204
0.3269
0.3232
0.32GB
-9.9980
0 . 3 1 50
0.3152
0.3117
0.3202
0.3155
0.3176
0.3140
0.3161
0.3086
0.3216
0.3227
0.3214
0.3236
-9 . 9990
0.3Z24
0.3163
0.3?')8
-9 . 9(.1;I0
-9.9'.'B9
0.3196
0.3219
0.3214
0.3239
0.3247
O.3192
0.3454
0.33G3
0.3?fi7
0.3193
0.3?00
-9.9967
-9.9960
0.3107
0.3103
0.3IG2
B.3158
fl.3145
0.3140
0.3119
0.3230
0.3100
-9 . 9995
0.3067
0.3091
0.3070
0.3116
0.3078
0.309B
-9.9993
0.Z9B0
0.2975
0.2954
0.3026
0.2986
0.3006
0.2970
0.2986
0.2931
0.3054
0.3067
0.3053
0.3058

0.3058
0.2996
0.3044
*******

0.302B
0.3054
0.3067
0.3077
0.3000
0.3032
0.3169
0.3165
0.30B0
0.3023
0.3019
-9.9990
-9.9976
0.3024
0.3012
0.3002
0.2991
0.2993
0.29B6
0.2994
0.3070
0.1287

0. 1242
0.1276
0.1251
0.1283
0.1249
0.1268
0.1265
0.1226
0.1256
0.1215
0.1257
0.1246
0.1258
0.1135
0.1142
0.1106
0. 1 104
0. 1 136
0. 1 136
0.1167
0.1162
0.1197
0.1148
0.1180
0.1151
0.1181
0.1157
0. 1189
0.1171
0. IZ00
0.1197
0.1212
0.1206
0.1187
0.1211
0.1178
0.1154
0.1164
0.11 18
0.1135
0.1089
0.1154
0.1141
0.1157
0.1157
0.1173
0.1197
0.9268
-0 . 0Z87
0.9364
0.9290
0.9354
0.9214
0.9328
0.9266
-0.0Z87
0.9620
0.9634
0.9703
0.9469
0.9595
0.9530
0.9645
0.9592
0.9772
0.9364
0.9323
0.9366
0.9349
-0.0286
0.9345
0.9539
0.9387
-0.0286
-0.0286
0.9435
0.9351
0.9312
0.9280
B.9271
0.9414
0.9004
0.9016
0.92G3
0.9437
0.9448
-0.0285
-0.0285
0.9431
0.9467
0.9499
0.9532
0.9525
0.9546
0.9521
0.9272
0.8746
-0.0Z87
0.8903
0.88Z2
0.8962
0.8783
0.8883
0.8784
-0.0287
0.9100
0.9093
0.0195
0.8948
0.9081
0.9026
0.9123
0.9061
0.9ZB1
0.8892
0.8861
0.8896
0.8834
-0.0286
0.8864
0.9035
0.8907
-0.0286
-0.0Z86
0.8939
0.8900
0.8886
0.8816
0.8794
0. 8942
0.0261
0.8415
0.8733
"0.8935
0.8914
-0.0285
-0.0ZB5
0.8949
0.8958
0.9018
0.9028
0.9065
0.9055
0.9052
0.8791
0.92.15
-0.0119
0.9SSB
0.9298
0.9484
0.9Z4Z
0.9493
0.9345
0.9367
0.9619
0.9383
0.9700
0.9364
0.9447
0.9351
.0364
.0295
.0630
.05U0
.0286
.02G6
.0006
.0049
0.9744
I .0163
• 0.9078
1 .0127
0.9863
1 .0068
fl-9705
0.9935
0.96t'9
0.9714
0.9SOZ
0.9617
0.9771
0.9572
0. 9049
1 .1/030
P.995Z
1 .0355
1 .02IVO
1 .06i.'4
1 .00/6
1 .0134
0.9?94
0.99P7
0.91)51
0. 96H6

-------
                          Table 11, COnt.  NOX  AND  03 CALIBRATIONS FOR 1984
ie-May-1985
DATE I NOX ZERO   NO  ZERO   N02 ZERO   O3 ZERO  NOX  SPAN    NO  SPAN   O3 SPAN  NOX RESP   NO RESP    03 RESP  NO CAL   NOX CAL  O3 CAL
AuglO -0.0040
AuglO -0.004S
Au3l9 -0.0047
Augl9 -0.0035
Aug20 -0.0048
Aug20 -0.0047
Aug22 -0.0045
Aug22 -0.0012
Aug23 -0.0036
Aug2S 0.0050
Aug25 0.0020
Aug-'7 0.0016
Aug28 -0.11010
Aug29 -0.0022
Aug29 -0.0007
Sep01 -0.P0I4
Sep01 -0.IJ014
Sep02 -0.0028
Sc-p02 -0.0037
Sep02 -0.P027
Sep03 -0.J.042
Sep06 0.0055
Sep06 0.0061
SepOO 0.0047
Sep08 -0.0016
Sc-p09 -0.0016
Sep£l9 -0.0027
Sepl6 -0.0033
Sepl6 -0.0047
Si-'pl7 -0.0038
Sepl7 -0.0042
Sop 18 -0.0037
Sop I 8 -0. 110 4 4
Sup 19 -0.1)040
Supl9 -0.0030
Sop21 -0.0052
Stp21 -0.0027
S,ip25 -0.0052
Sep2S -0.0001
Oct.03 -0.8028
Oct05 -0.0037
OctOS -0.0036
OctOS -0.0050
Oct06 -9.9999
Oct07 -0.0054
Oct07 -0.0054
Oct09 -0.0040
Oct.10 -0.0045
Oclll -0.0051
-0.0027
-0.0625
-0.0027
-0.0020
-0.0027
-0.0031
-0.0034
-0.0023
-0.0026
0.0000
-0.0010
-0.0014
-0.0618
-0.0026
-0.0018
-0.0022
-0.0022
-0.0028
-0.0031
-0.01)25
-0.0021
0.0003
0.0017
a . 00.45
-0.0013
-0.0056
-B.0014
-0.0023
-0.0029
-0.0023
-0.0030
-0.0026
-0.0032
-0.00.11
-0.0631
-0.0034
-0.0025
-0.0036
-0.0038
-0 . 0028
-0.0033
-0.0041
-0.0016
-9.9999
-0.0045
-0.0065
-0.0043
-0.0055
-0.0048
0 . 00 1 1
0.0004
0.0002
0.0009
0.0006
0.0010
0.0014
0.0030
0.0016
0.0060
0.0050
0.0044
0.0031
0.0026
0.0027
0.0031
0.0031
0.0028
0.0018
0 . 00 1 9
-0.0001
0.0073
-9.9999
0 .O059
0.0015
0.0009
0.0006
0.001 1
0.O004
0.0600
0.0009
0.0005
0.0010
0.0005
0.0020
0.1)006
0.0022
0.0002
0.0014
0 . 0020
0.0020
0.0026
0 . 00 1 9
-9 . 9999
0.001 1
0 . 00 1 9
0.0018
0.0025
0.0019
0.0007
0.0007
0.0008
0.0015
0.0008
0.0010
0.0008
0.0608
0.0008
0 . 00 1 0
0.0010
0.0023
0.0008
0.00OB
0.0008
0.0017
0.11017
0.0008
0.0007
0.0007
0.0009
0.0038
0.0022
-9.9999
0.0021
0.0007
0.0010
0.0031
0.0013
0.0007
0.0011
0.00>I8
0.6008
0.0007
0.0014
0.0008
0.0019
0.0007
0.0009
0 . 000B
0.003B
0.0007
0.0007
-9.9999
0.00O7
0.0009
0.0007
0.0020
0.0407
0.3164
0.3172
0.3140
0.3139
0.314Q
0.3147
0.3183
0.3209
0.3197
0.3140
0.3155
0.3165
0.3144
0.3143
0.3129
0.3153
0.7439
0.7377
0.7377
0.72G9
0. 7390
0.7447
0.7447
0 . 7 5 'JO
0. 76ul
11.7626
0. 753G
M.750D
0.7520
0.75U1
0.7505
0.7541
0.7479
0.7552
0.7453
0. 731'J
0.7273
0.7271
-9.9999
-9.9999
0.7290
0.7257
0.730C
0 . 739C
0.7420
0.7414
0.7448
0.7400
e. 7364
0.3019
0.2099
0.2999
0.2906
0.3002
0.3fi04
0.3H37
0.3049
0.3044
0.2945
0.2965
0.2992
3.2933
0.2900
0.2972
0.2997
0. 7u6u
0.6920
0.6920
0.6U13
0.6049
0.b'.)69
0.6069
0.7140
0.7149
O.7iri6
0.7C.-32
11 . 7U I 0
0.71)25
0. 7JS83
B.G9'JS
0.7050
S.6U7JO
0.7JI4O
0.69 JO
0.G771
0.6706
0.6718
0.G691
-9.9999
0.6742
0.6723
0.6760
0 . 6U 7 2
0.6904
0.6301
0.6910
0.6bS0
0.6034
0. 1 167
0. 1 162
0. 1 17'.,
0. 1 183
0. 1 192
0. 1 ICG
0.1 20f;
0. 1201
6. 1265
0.121V.
0. 12E1/
0. 1255
0. 124G
0. 12SG
0 . 1 1 fj n
0. 1 17;.'
B. I7Z
e. 172
0. 172
0. 139
0. 173
0. 193
tl . 19',:
0 . 1 'J 2
0. 130
a . 194
a. ion
0. 1270
o. i2'ji;
O . 1 2 7 a
0 . 1 2 C li
0.127V.
• 0.1 2SC
II. 12 ('.'.'
0.1213
0 . 1 1 a 7
tl. 1 1U3
K. 1 165
0.1181
-9.999 9
0. 1122
0.1 1 OS
0. 1 1 13
0 . 10G3
0.1076
0. 1073
0. 1074
0. 107S
I). IDbl
0.3204
0.3217
0.3187
0.3174
0.3196
0.3194
0.3228
0.3221
0.3233
0.3090
0.3135
0.3149
0.3154
0.3165
0.3136
0.3167
S.7453
0.7405
0.7414
0.7296
0.7432
0.7392
0.7386
- 0.7543
0.7617
0.7642
0.7E63
0.7533
0.7576
0.7G19
0.7547
0.7578
0.7523
0.7592
0.7433
0.73G7
0.7300
0.7323
-9.9S98
-9.9971
0.7327
0.7293
0.7356
0.7474
0.7468
0.7496
0.7445
0.741S
0.3046
0.3024
0.3026
0.3006
0.3029
0.3035
0.3071
0.3072
0.3070
0.2945
0.2975
0.3006
0.3001
0.3006
0.2990
0.3019
£.7082
0.6948
0.5951
0.6838
0.6970
0.6966
0.6952
0.7135
0.7162
0.7202
0.710G
0.7033
0.7054
0.7106
0.7025
0.7076
0.7002
0.7071
0.G961
0.6805
0.6731
0.6754 -
0.6729
-9.9971
0.6775
0.6764
0.6806
0.6949
0.6946
0.6953
0.6905
0.6882
0. 1 160
0. 1155
0.1167
0.1168
0.1184
0.1175
0.1200
0.1193
0.1257
0.1205
0. 1240
0. 1232
0. 1240
0.1238
0. 1 177
0. 1 155
S.IIS5
0. 1164
0. 1165
0. 1 182
0. 1 164
0. 1 185
0.1171

0.1118
0. 1 187
0.1145
0. 1239
0. 1245
0.1271
0.1255
0. 1267
0. 1250
0. 12C2
0. 1 199
0. 1 179
0. 1 164
0. 158
0. 172
0. 114
0. 102
0. 106
0. 1069
0. 1064
0. 1067
0. 1055
0. 1054
0.9342
0.9410
0.9462
0.9465
0.9391
0.9373
0.9260
0.9257
0.9261
0.9GS1
0.9553
0.9452
0.9466
0.9440
0.9499
0.9403
S.9383
0.9563
0.9559
0.9717
0.9526
0.951 1
0.9530
0.9273
0.9238
0.S18U
0.S304
0.9354
0.9326
0.9252
0.9358
0.9284
0.9382
0.9284
0.9431
0.9B33
0.9739
0.9678
0.9714
-0 .0650
0.9579
0.9595
0.9529
0 .0607
0.9326
0.9330
0.9307
0.9365
0.9389
0.8881
0.8845
0.8927
0.8964
0.8901
0.8906
0.8809
6.8828
0.8794
0.9193
0.9056
0.9022
0.9007
0.8974
0.9057
0.8963
0.8557
0.9009
0.8998
0.9144
0.8970
0.9000
0.9007
0.8807
0.8722
0.8687
0.8778
0.8769
0.8710
0.8664
0.8747
0.8705
0.8768
0.8683
0.880'J
0.8935
0.9017
0.8963
-0.0G56
-0.0653
0.8894
0.8936
0.8853
0.060G
0.8707
0.8714
0.8669
0.8722
0.87S1
0.990J
0.9943
O.90J4
6.9826
0.96C7
ft . 9 7 C 1
0.9b4b
0.9602
0.9 1ft 7
0.94C3
0.9220
O.92C9
0.92C3
0.9212
it . 9 1> <* $
M.985G
S.98C&
0.9773
0.97CS
tf.9624
0.97&7
0.9576
0.960(1
0 .01 1 2
1.6137
0.9541
0 .9891
0.9UI.J
a.9JL7
0.8&C6
0.89/9
0.88L8
0.901.9
0.8913
0.931.5
0.95:13
0.9b!,G
0.96UI
0.9DC6
-0.0112
0.9939
. 0 1 JU it
.00C5
. 6 1 1 U
.041.7
.CUF.G
.0413
. oc :' 4
. o '_ :.• a

-------
                       Table 11, cont.  NOX AND 03 CALIBRATIONS  FOR  1984
16-M.y-I985
•
-------
 HC Processing	Calibration

     Figure 9 shows the flowchart for HC calibration data processing with CALEIITRY
 events indicated. Table 12 shows the RU!!E!JTRY events flowchart.

 Identification of Calibration Chromatogram and Calibration Source

 If one or more forms of documentation indicate that a calibration(s) was performed,
 then the first step is to identify the data and calibration source(s). If the calibration
 was performed during an experiment, it was logged in CALENTRY when the experiment
 was logged in RUMENTRY. Otherwise, the  run calendars indicate a calibration event
 and the presence of a new calibration folder.  If the calibration occured during an
 experiment, the run folder INSTRUMENT INVENTORY CHECKLIST indicates the instruments
 calibrated, otherwise, the stripcharts present in the calibration folder indicate which
 instruments were calibrated. Calibration sources are identified on the stripcharts.

 Issue Calibration Pick Instructions

 Calibration processing (pick) instructions are  given by the PC via the CAL  PICK IN-
 STRUCTION form.  The form is stored in  the CAL PICK INSTRUCTION NOTEBOOK by date.
 The fact that pick instructions are given is entered into CALENTRY. Several QA steps
 occur at this point. Documentation is checked  for reasonableness and completeness
 (sources, chart settings, instrument settings). The chromatograms are inspected for
 reasonableness (match standard example chromatograms with source for conditions
 specified).

 Official Calibration Sources/ OCS Report
 The raw data can be processed (picked), but before final processing can occur, the
 calibration source data must be validated and entered into the OFFICIAL  CALIBRATION
 SOURCE (ocs) DATATRIEVE database  and the OCSFIL.DAT updated.  The  forms used for
 recording the OCS data were discussed above.

    Several sources are used throughout the season. These were listed above. The
 validation  procedure was also reported.  The first step is to check the report from
 the ocs database to see if the source has been entered. If the DCS does not exist then
 it must be created or at least validated. The details of processing single (used once)
 OCS are discussed next. The DCS report is listed  in Appendix C.

 Create OCS for Single Injection
 These ocs are created in the smog chamber. The amounts of HC that are injected
 are documented along with the chamber temperature in the calibration folders or
on the RUN  SHEET.  A LOTUS  123 database/spreadsheet ocsic (described above) is used

                                      59

-------
 Calibration
                                     HC Processing
                CAL

                FOLDER
                    V
                         A
                   RUNFqLDERJ
DTR (OCS)
                  DIGPIK
                                mmddyRiii
                  CALFAC
                     *
                   eUSf IjQ
                    DISK	J
                          mmddyQ.iii
                               DTR (HCCAL)

                                    I
     HCCAL84.DAT
                  CALANA
                   I

™ T^nr
_*



stat
rept
Figure 9. HC Autocal Processing Flowchart.
                       60

-------
HC Processing
                                                        Calibration
                                   Table 12.
        Processing System for Instrument Auto-Calibration Data
          tot inst    cal analysis
\
 !CPI!
     \
                      \
                       -
                                 \
                                 IOCS!
                                     \
                      \
                     ---(Q2V)--(DTR)—(C2L)----!CQA!
          Step
                     !sss! == stages determined by Project coordinator (PC)
                      == stages determined by Peak Pickers (PP)
                     (sss) == stages determined by Computer Techs (CT)
  CALEUTRY stages
Meaning
PC
PP
PC
PP
CT
CT
CT
PP
PC
PC
PC
CPI
DPR
DCS
CFR
Q2V
DTR
C2L
CLR
CQA
NTB
BAD
CAL pick Inst
Digpik run
DCS ok
CALFAC run
Qfile to vax
DTR file updated
Cals to LSI
CALAUA run
Cal QA
Not to be proc
Stop processing
Charts marked, cal status sheet in folder
Cal data digitized and R-file exists
Official Cal Source exists in DTR
Cal Factor computed, Q-file created
Q-file moved to VAX
Cal Data added to DTR data base
DTR CAL file loaded on LSI-11
Cal Factors analyzed
Cal Factor Quality Assurance
Not to be processed
Something is wrong with this run
                                      61

-------
 Calibration	HC Processing

 to process this  information.  The species id number is used to look up  all the
 necessary species specific information (density, molecular weight, carbon number,
 etc.) and the program calculates and records the ideal concentration. Table 4 lists
 the OCSIC report.  All DCS information  is recorded on DCS forms  and recorded in
 the DCS notebook as described above. The validated ocs concentrations are entered
 into the DATATRIEVE database CAL^SOURCESJMS. A database report procedure is used to
 update the OCSFIL.DAT which is transfered to the LSI to await final processing. The
 fact that the ocs exists is entered into the DATATRIEVE CALEIITRY database.

 Run DIGPIK on calibration Data to  Make R-files
 Once calibration processing instructions have been given, raw data are picked using
 a program on the LSI (DIGPIK) where  the  data are digitized and a response file
 (R-file, MMDDYR.iit) is produced. The files are backed-up  (copied) onto floppies.

 Run CALFAC to Calculate Calibration  Factors
 Once the response data has been picked and the ocs data has been updated to OCS-
 FIL.DAT, the calibration factor can be calculated. A program on the LSI CALFAC is run
 which reads the internal documentation in the response file R-f lie and compares the
 data with the appropriate ocs to generate the concentration per response unit cali-
 bration factor. A QA step is performed here.  The calibration factors are inspected
 and reviewed in context with the existing database with the program CALLOK which
 plots or lists the data vs. Julian  date with  a linear least-squares fit. The program
 identifies calibration data which varies  by more than  a specified percentage.  The
 generated calibration data  is output in files ("Q"-files) specific for the instrument
 and the day calibrated. These files  are backed-up on floppy,

    The Q-files are transferred and  updated to the DATATRIEVE HC_CAL database if the
 calibration data seems reasonable. A DATATRIEVE report procedure PRIHT-HC.CAL.FACTORS,
 is run to generate an  updated calibration list sorted  by  instrument, species, and
 date. All data are converted to a single specified attenuation  for easy inspection.
 Another DATATRIEVE procedure, PRIMT.FILE_CALA1IA,  is used to generate another file
which is then transfered to  the LSI-ll/23 computer to be used by a program called
CALAtIA which performs a statistical  and graphical  analysis of the calibration data-
base.

    We expect  a  certain amount of noise or variation based on experience.  Usu-
ally  the calibration factors are very  stable  with little or no trend demonstrated.
Variations are found, however,  and  some are real indications of instrument change.
Discrepancies that are more than a few percent are investigated. Figure 10  shows

                                      62                                           .

-------
 HC Processing	Calibration

 example CALANA plots and statistical reports. Appendix D contains a larger collection
 of these plots and statistics.

     The statistical report first  lists the data range searched in the database. The
 number of data points  found and included for the analysis are  listed. The slope
 and intercept of the straight line fit through the data are listed with the standard
 deviation and relative standard deviation. The correlation coefficient is also listed.
 The minimum, maximum and mean of the data are shown with the standard devi-
 ation and relative standard deviation. Lastly the equation of the straight line fit is
 shown. Table 13 shows  a compilation of the statistics from this set.

     This analysis shows  that little or no trend was observed in the data. If the line
 was perfectly horizontal, the correlation factor would be  0.0. Table  13 shows the
 means, standard deviations, and relative standard deviations for the more complete
 data set shown in Appendix D. The average relative standard deviation is just under
 10 percent with a standard deviation of 2.2 percent. This suggests the uncertainty
 in the data would be about 10% if the  calibration factors were  used  straight from
 the  initial processing and  no  additional  information and QA techniques were used
 to check and tune the calibration data. Because specific calibration factors are used
 and because the data are subjected  to further examination and adjustment, the
 uncertainty in the final data is less than 10%.

    Valid deviations can be caused by many factors, including variation in the in-
 strument operating characteristics.  Calibration factors  specific to the current in-
 strument status should be used rather than some average value.  Non-valid  sources
 of calibration data deviation include improper processing. Many QA measures are
 taken to prevent these errors. These measures will be discussed further.

    There are several explanations for variation in calibration data.  Insufficient
 sample flow will result in very low responses and consequently very large calibration
 factors. Sometimes the  instrument operating conditions change (e.g.  carrier flow)
 which changes the peak  width arid consequently the calibration factor. Incorrectly
 recorded attenuation will, of course, cause generated calibration data to be off by a
 factor of multiples of 2 or  10. The Perkin Elmer Sigma  10 integrator  which yields
 areas not affected by the  GC attenuation  settings, can be compared with areas
 of calibrations performed on other days to determine if this type of  problem has
 occurred; the  attenuation can be corrected in the "R" file and reprocessed.

Estimation of Calibration Factors for Infrequently Calibrated HC
The  technique discussed  above requires an explicit calibration source for the com-

                                      63

-------
      Calibration
                                                             HC Processing
       0.40
       0.30
           I  I I  I I I  I I  I I  i| I I  I I  I I  I I  i I I  I f I  I I I  I I  I I
                                  Carle  I     GC         !
                                  PROPYLENE            1
                                  PPMC/IN  Atten =   2.00
    £ s.2»k
       0.10
       0
                         ,

                         I   IV
                                    29-AUG
/  ^\
                           0.<
                         -J0.30
                         _j    g>
                                                               O
                           0.10
       ' IS0 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320'
                         Julian  Date, 1984
                         I*
      For the  species PROPYLENE  and the instrument  Carle I
                               GC:
      Beginning  Caldate:
                           Day:
      Ending     Caldate:
                           Day:
      Upper Cal Factor Limit
      Lower Cal Factor Limit
30-MAY-34
150
16-NOV-S4
320
   0. 40000
   0 . OOOOO
      The number  o-f  data points
      The number  included  was
      The slope o-f  this fit  is
               with  a std dev o-f
      The intercept  is
               with  a std dev o-f
      The std  dev of the fit is
      The correlation coeff is
   62
   62
1 . 3522654E-004
6.6911662E-OO5
1.6694184E-001
1.5341250E-002
1. 30799 19E-002
 0.25246
rel . dev  =  49.5 7.

rel . dev  =  9.2 7.
PE =  0.01205
      The min, mean,  max are        0.1304O,    0.19760,     0.25009
      The standard  deviation is  0.01353   rel. dev  =  9.4 7.

      Cal =  0.00014  X (Julian Date -  150)  +  0.18723

Figure 10.  CALANA Example Figures.
                                    64

-------
  HC Processing
                                                        Calibration
 O


LL.

 ITJ



•?.


0.

A


*'

0.

i i | i | i | i | i
i—
i
25J-
i
i
i
2»|-
h

I
lL '
'jJ3i-MAY
J-
i
<
| 1 | 1 j 1 j 1 | 1 | i | 1 :' 1 | 1 ! 1 |
Carie II GC
ETHYLENE
FFMC/IN Atten = 2.00


f*
M ' l'
._ .... • • II
iS-jLL « v\. .• ti
* ^a^^3>r_ _\R» ^ 	 	 ^.-> -^- - -





1 i '
-



-^

—


^^
^~
~
	
A *^A


?.25


.. ,-..
" Q|
•—
fl 1R "^
• iij
i j

-------
 Calibration
                                                        HC Processing
      I      I
                            Carle  III  GC
                            ISOPENTANE
                            FPMC/IN  Atten =  2.00
                                                     i    O
                                                     1    ID
                                                    1    S"

                                                     0.10
                                                    1
     L L_L I I  ! I  I !. I I  I M I  ! I I  I I  I I  I I  I I  ! ! !  I !  I I
     160 170 130 190 206 210 220 230 240 250 260 270 280 290 300 310 320'
                   Julian  Date, 1984-
                                                       00
For  the  species ISOPENTANE  and  the instrument Carle  III GO;
Beginning  Caldate:
                     Day:
Ending     Caldate:
                     Day:
Upper CalFactor Limit
Lower CalFactor Limit
                           3O-MAY-S4
                           150
                           16-NQV-84
                           320
                              0.40OOO
                              O.OOOOO
The number  o-f  data points     60
The number  included  was      60
The slope of  this fit  is -1.222656E-004
         with  a std dev of 5.6O03271E-005
The intercept  is          2.2209272E-001
         with  a std dev of 1.273873BE-002
The std  dev of the fit is 1.6698649E-002
The correlation coeff is  -0.27557
                                             rel. dev =  -45.3

                                             rel . dev =  5.7 "/.
                                             PE =  0.01113
The min, mean,  max  are
The standard  deviation is
                              0.15936,     0.19463,    0.23074
                            0.01722    rel.  dev = 3.3 '/.
Cal = -0.00012 X  (Julian Date -  150)  +  0.20375
Figure 10, cont.  CALANA Example Figures.
                               66

-------
   HC Processing
                           Calibration
      Table 13.  Summary of CALANA Statistics
Carle
Carle
Carle
I


II




III















propylene
isopentane

ethyl ene
ethane
propylene
propane

1-butene
n-butane
cis-2-butene
trans-2-butene
1 ,3-butadi ene
isopentane
n-pentane
benzene
toluene
m-xylene
o-xylene
methanol
ave
std dev
MIN
0. 130
0. 140

0.074
0.013
0.369
0.428

0.074
O.OS3
0.070
0.091
0.090
0. 160
0.228
0.080
0. 114
0. 187
0.256
O.EJ35


  MAX
0.250
0. 196
                                     0. 198
                                     0.231
                                     0.976
                                     1.059
 MEAN   STD DEV REL STD  DEV
0.198     0.019      9.47.
0.158     0.010      6.37.
          0.121
          0.147
          0.353
          0.607
         0.009
         0.007
         0.054
         0.060
7.67.
3.07.
9.8%
9.97.
0.074
0.053
0.070
0.091
0.090
0. 160
0.228
0.080
0. 114
0. 187
0.256
O.EJ35


0.130
0.131
0.127
0.132
0. 137
0.231
0.347
0. 106
0. 167
0.269
0.318
1.206


0.100
0.102
0.105
0. 115
0.116
0. 195
0.289
0.090
0. 137
0.225
0.281
1.050


0.011 .
0.016
0.015
0.011
0.0.12
0.017
0.028
0.009
0.014
0.027
0.024
0.143


1 1 . 57.
1 1 . 77.
14.07.
9.27.
9 . 97.
8.87.
9.87.
10.37.
10.07.
12.07.
8.77.
13.77.
9.97.
2. 27.
                                    67

-------
 Calibration	HC Processing

 pound being monitored to generate a calibration factor.  Sometimes calibration data
 is sparse or the species to be processed has no explicit calibration source. There are
 several techniques for establishing appropriate calibration factors. These are often
 used together.

     When sufficient  supporting calibration data exists, a plot of  the calibration
 factors of well calibrated compounds measured on the same column, as a function
 of  retention time, can be  least-squares fit to an exponential function.  A LOTUS
 123 spreadsheet called CALEXTR is used  for this.  Figure  11 shows plots for the two
 columns  in the Carle III  GC. Our experience has been that good fits are obtained.
 The calibration factor for the uncalibrated compound can be estimated by locating
 its  retention time on the curve.  The ratios of the frequently calibrated compounds
 to  the less frequently calibrated compounds can be determined for future needs. If
 integrator data is  available, a check is done using integrator  areas  ratioed  to the
 estimated concentrations of all the compounds.

    If the compound requiring  a calibration factor is part of  a well characterized
 mixture (the source), the concentrations  of the other compounds can be used  to
 determine the concentration of the compound in question by ratios. The calibration
 factor can then be calculated.

    If reliable integrator  data  is available, the  area count can be  calibrated with
 any known calibration sources,  and the corrected area used to determine the con-
 centration of the uncalibrated  compound.  The concentration  can then be used  to
 calculate the calibration factor for the picked data.

    If good calibration data exists for  all species at  some time, ratios of less fre-
 quently calibrated species to frequently  calibrated species can be determined.  These
 ratios can be used to calculate calibration factors for times when the calibration
 data are less complete.

 Final Calibration Factor Calculations
 To  process data, calibration factors can be taken from the HCCAL DATATRIEVE report
 for  explicit calibrations or  extrapolated  from calibration data for other species.
 Calibrations may not have been performed the day of an experiment for all species.
 Interpolation may be  needed between the calibrations for a species done right before
 and after the day in question.

    The discussion  to this point has been concerned with the processing of calibra-
tion data and using calibration data to  estimate calibration factors for compounds

                                      68

-------
       HC Processing
                                          Calibration
                 0.30 -
                 0.28 -
                 0.28 -
                 0.24 -
                 0.22 -
                 0.20 -
                 0.1B -
                 ••" -
                 0.14 -
                 0.12 -
                 0.10 -
                 o.oe -
                 0.06 -
                                 LIGHT COMPOUNDS
                                           Carlo m
                 0.7
                 0.8 -
                 0.3 -
                 0.4 -
                 0.3 -
                 0.2 -
                 0.1 -
                         a  eml d«t»
                                    20               40
                                        RETENTION TIME
                                          +  exp: r2 - 0.984
                                                                     SO
                           HEAVY COMPOUNDS  COLUMN
                                           Ctrla
                         O  eal 
RETENTION TIME
   o  «xp: r2 » 0.98
Figure 11.   Calextr Examples.
                                              69

-------
 Calibration	HC Processing

 not explicitly calibrated. Calibration factors need additional checking and perhaps
 adjustment sometimes before they are applied to process data. These checking pro-
 cedures are actually QA steps performed in experimental data processing but are
 definitely part of calibration data processing.

    Once calibration factors are calculated, they are applied to the raw experimental
 data. The resulting concentration data should fit with the known target conditions,
 the amounts injected, and the known resulting  conditions of other similar experi-
 ments. The data should also yield the same composition as determined by analytical
 data for other similar experiments. This should occur with calibration factors used
 which are at least  similar to  those used to process experiments conducted before
 and after the experiment in question (see discussion of RUN_CAL.USED). Ratios of ini-
tial conditions should be somewhat consistent with integrator areas when peaks are
reasonably well defined and integrateable. The calibrated integrator area should be
consistent with peak height derived concentration data for large peaks with good
resolution and  clean base lines.
                                     70

-------
 Experimental  Data
This chapter presents the procedures, tools, and techniques for processing the ex-
perimental data.  Sections  will discuss documentation, data processing tracking
systems, and how data are actually processed.

Experimental Results Documentation

Documentation of experimental conditions and results is performed at every level of
processing to assure that no data or other useful information is lost and that all data
can be processed. Methods have been developed to allow for redundant documen-
tation and to provide the smog chamber operators several different opportunities
to document the  experiment.

Data
Data are documented several ways reflecting the different types of data acquisition
techniques. Data is acquired and recorded by both computer and stripchart. Print-
outs are also produced by the computer during and after an experiment and provide
additional documentation.

Stripcharts
Almost every source of data is recorded on  stripchart, providing a backup for the
computer DA system in case of failure.  For instruments that produce data not
suitable for the computer DA system, stripchart recorders are the main data acqui-
sition device. All stripcharts are documented with rundate, instrument, time,  side,

                                 71

-------
 Experimental Data                                               .             Data

 attenuation, chart speed, full scale deflection, and calibration information.  Docu-
 mentation of stripcharts is requested by the setup checklist and performed by the
 setup operator.  Changes  of instrument settings  or conditions are noted directly
 on the stripchart during the experiment. Time and side is recorded several times
 throughout the experiment, as requested by all of the checklists discussed above.  A
 pack-up inventory checklist assures that all stripcharts are collected and stored  in
 the runfolder. Stripchart data is obtained for the following instruments:
  •  Os monitor
  •  NOX monitor
  •  Carle 1 GC-FID for HC
  •  Carle 2 GC-FID for HC            :
  •  Carle 3 GC-FID for HC
  •  Formaldehyde monitor
  •  Dewpoint monitor
  • Varian GC (PAN, tracers, nitrates, biacetyl)
  • PE900GC
  • TSR (total solar radiation)/UV sensors

 Printouts
 Printouts produced during and after an experiment also document the experimental
 data. These are retained in each runfolder:
  • PRINT PROFILE—computer data dump during run
  • Experiment COMMAND FILE listing—list of scheduled injections and chamber con-
    trol commands
  • Experiment BAKTRN FILE listing—list of actual  injections and chamber control
    commands (both scheduled and manually activated)
  • Operator's REMARKS FILE- comment file

 Computer Data
There are two  types of computer data generated by this data processing system:
computer  acquired data during the experiment and files produced from stripchart
recorded data after the experiment in data processing. Documentation for computer
data exists within the data files produced themselves. Time, instrument, side, and
any intermediate conversion factors used to store the data is recorded automatically.
The remarks file contains comments entered by the site operators before, during,

                                     72

-------
 Forms                                                            Experimental Data

 and after the experiment directly into the computer, which are automatically doc-
 umented with time entered. The data files produced from stripchart data after the
 experiment contain all of the documentation  recorded on the stripcharts and the
 name of the people involved in processing the data.

 Forms
 Several paper forms  are used to help document the experimental conditions and
 resulting data.  Several of  these  are filled out at the smog chamber site as  the
 experiment is set up  and executed.

 Run Form
 The run form is a major documenting tool. It  is used to record all the information
 needed to determine what experiment was attempted, what the chamber conditions
 were, time the experiment starts and ends, and important notes about the exper-
 iment actually conducted. The form is partially filled out before the experiment
 to officially record the desired target conditions including sources and amounts of
 chemical species. The morning operator takes instructions from this form to initiate
 the experiment, and records important comments about chamber conditions, devi-
 ations from target conditions, and any problems. This form is shown  in Figure 12.

 Instrument Checklist Form
 The instrument checklist is filled out the morning of the experiment and is attached
 to the run  form.  It is used to  indicate  which instruments  are operational and
 were  used,  any  calibrations performed, and  any  problems with the instruments
 themselves. This form is shown in Figure  13.

 Hun Folder Inventory Checksheet Form
 This form is used to inventory the run folders  contents when the runfolder is first
 logged into the data processing system. It is a primary QA step used in conjunc-
 tion with the instrument checklist form to determine  what data exists, including
 documentation.  Login and filling out of this form occurs before anyone is allowed
 to work with the contents. This form  is shown in Figure 14.

 Data Processing Instruction  Form
 The data processing instruction form is used to collect and convey important in-
formation about the data collected on stripcharts during an experiment.  It is an
 intermidiate documenting tool used to indicate start and stop times of valid data,
calibration  data (not  to be confused and picked as run data), species to pick and

                                     73

-------
        Experimental Data
                                                   Forms
         Run  Date
         Operator  Horning _

         Experiment  Number

         Command File(s)  	
      RUN SHEET

Project 	
Operator Set-up
            Operator Afternoon
       from Table
                      Tank File Date
        Run  Description and Purpose
        Special Tasks,  Modifications
             ACTUAL  INJECTION CONDITIONS  AND  ESTIMATED INITIAL CONCENTRATIONS

        Dilution Tracers ID    Blue,  ul                         Red,  ul
                                    Blue
                              Red
        Species ID stat  Concen.  secs,ul,g   dynara    stat    Concen.   sec,ul,g  dynan
                   dynam Concen.  ul,liter    ml/min   dynam   Concen.   ul,liter  ml/m;

        N0_
        N02
        Total.HC_
        Total.NOx

        Drying Performed?   Red

        Dew Point(sunrise)  Red
 	 Delta Time 	

 .F 	mv Time..
  Blue 	 Delta Time     .

 __ Blue 	F 	mv Time	
        Horning Chamber Condensation Observed ?  (Y/N)  Red:  inside
                                                       Blue: inside
        Vent Doors Checked (initials):  Horning
        CHARTS LABEL (initials):        Morning

        Special Problems, Concerns, or Comments
                        End of run
                        End of run
                                         outside 	
                                         outside 	
        DVM Data Recorded (TIMES) 	 to ;	.  Experiment starts  (TIME)
        Experimental Conditions established at (TIME) 	
        Run rating (circle one):  useless

Figure 12.  Runsheet
          borderline   fair   good   great   unknc
                                               74

-------
       forms	     Experimental Data



                      Instrument Checklist      Run Date 	
                                                                 Data
                                                Not               Acqui-   Note
   Instrument         Cal    Cal     Opera-    Opera-     Not     sition
                     Today  TIME(S)  tional    tional     Used    QA
                     (y/n)           (check)   (check)   (check)   (init.)   (num)
                     (type)

   ATC
   Carle  I
   Carle  II
   Carle  III
   CEA  Formaldehyde
   CO meter
   Dew  Point
   DNPB Aldehydes
   Nitric Acid
   NOx  Bendix
   Ozone Bendix
   PE 900  (chamber)
   Saltzman
   Temp
   TSR
   UV
   Varian aerosol
   Varian gas
     PAN
     Nitrates
     Tracer
   Other
   Other
   Other
   Other
  * Notes
Figure 13.  Instrument Checklist Form

                                        75

-------
        Experimental Data
                       Forms
         RUN FOLDER  INVENTOR*  CHECKSHEET     INIT:  	
         PROJECT:  AUTO    DODG   REAC    WOOD      DATA:   AERO
         PRESENT  ITEM                     COMMENTS
RUN
DATE: ._
   GAS
         WOOD
                Run  sheet
               . Instrument Checklist
               . ATC  stripchart*
               , Cade  I stripchart*
                Carle  II stripchart*
                Carle  III stripchart*
                CO meter stripchart*
                Dewpoint stripchart
                FORM stripchart*
                NO stripchart
                N02  stripchart
                03 stripchart
                PE900  stripchart*
                PRINT  PRO printout
                .R CMANDR printout
                .R DATRAN printout
                .R.MERGE printout
                Sigma  10 printout
                TSR stripchart
                DV stripchart
                Varian stripchart*
                DNPH aid stripchart*
                OTHER:	
                OTHER:	
                OTHER:	
                OTHER:	
Figure 14.   Run Folder Inventory Checksheet Form
                                              76

-------
 Plots	Experimental Data

 problems in general with the data. Calibration factors and any calculations or notes
 are indicated on this form. The information on this form eventually ends up in the
 final segmented file used for distribution. This form is shown in Figure 15.

 DVMFIX Data Sheet Form
 This form is similar to the Data Processing Instruction Form used with stripchart
 data.  It is  used to record and convey important information required to properly
 process the data recorded with the DVM computer acquisition system.  Calibration
 factors, periods of valid data, and general documentation are indicated here.  The
 information on this form  eventually ends up in  the final segmented file used for
 distribution. This form is shown in Appendix E.

 Plots
 Plots are a form of documentation as well as a method of presenting data. They
 indicate when  calibrations were performed, when injections were made, and when
 the chamber is vented. They often indicate power failures, computer failures, or in
 general problems, and general quality of data.

 Raw Data Plots ("P"- and "U"-file Plots)
 These are plots of raw uncorrected and at first unQAed data. "P"-file plots show raw
 "picked" (digitized data). "U" file plots show the  raw DVM data, often including the
 AUTOCAL calibrations and venting.  Both types of plots indicate completeness of data.
 ratios of conditions, profiles and general  reasonableness of data.  Discontinuities
 may indicate problems during picking.

 Processed Data (G- and C-files QA plots)
 These plots of the  final corrected data show immediately any problems or errors
 which might have occured in processing. Only valid experimental data should have
 been processed. Concentrations should be close to  target conditions.  Light, dew-
 point, and temperature data should be reasonable in magnitude.

 Tracer Data Plots (CSTAR)
 Inert tracers are usually injected into each chamber half to monitor dilution during
 the experiment. Raw  data are read and processed by  a computer program CSTAR
which fits the data  to an exponential function  with the least-squares method. The
slope is the dilution rate which should be used by modelers. This data is processed
for every run; chamber leaks can be detected quickly.

                                     77

-------
       Experimental Data         	      	       		Plots
                              DATA PROCESSING INSTRUCTION FORM        PS	of
        RUN DATE	          INSTRUMENT	

        RUN HCANAL <*/n>	  ON FILES BETUEEN  	:	AND
        COMPOUND       SIDE  CAL FAC X ATTN   CAL FACTOR CALCULATIONS/DOC       INITIAL/
                                (AND UNITS)                                       PATE
       Cal Source Reports Used	-	   Date _	...

       Cal Source Reports Used	   Date	

       Cal Source Reports Used __			   Date __	

       RECORD ANY SPECIAL NOTES OR CALCULATIONS ON BACK OF THIS FORM PLEASE  !
Figure 15.   Data Processing Instruction Form

                                              78

-------
 Run Data Processing Status Databases	Experimental Data

 Run Data Processing Status Databases
 Description of Lab/Site Calenders
 Calendars have been  discussed above. They are used as a primary documenting
 method.  Experiments, calibrations,  and some  instrument or  chamber  work are
 indicated on the calendars at the site and the data processing office first. These are
 used to guide the data processing effort.

 Description ofRUNEHTRY database
 A DATATRIEVE database maintained on the VAX is  used to  track all data processing
 and record  the  experimental conditions.  The processing status of a single exper-
 iment or a  group of experiments can be quickly assessed.  It is used with report
 procedures  to guide processing tasks. Tasks are listed by the priority given  by  a
 project coordinator.

    The RUNEi.'TRY system is a PL/I program  that runs on the VAX-n/780 computer
 in the SPH. It provides on-line, interactive access to the database of run processing
 status. The program uses the VAX "electronic" Forms Management System to  sup-
 port a number of  "forms" on a video display unit (terminal). The user can fill out
 the form, and the data are either called up from the database, or are stored in the
 database for future reference.

    Table 14 is the individual record description for an entry in the  database. In
 addition to  the general information about the  run  conditions, the status of the
 processing steps for  digital voltmeter data, for documentation preparation, for the
 processing of each instrument that had valid data, and  for the progress toward
 creating the final file are stored in a record for each run.
                                           t
    There are a possible total  of 110  checkpoint steps in processing a run.   The
 events that  are entered will be  listed in sections following. The record provides for
 nine instruments, accounting for most  of the  checkpoints.

    Single page printouts  (132 columns wide) can be created for each run. An ex-
 ample of a printout is shown in  Table 15.  This shows a nearly completely processed
 run that had five instruments.

    Figure 16, Figure 17, Figure 18, and Figure 19 are "hardcopies" of the terminal
screen showing the "electronic" forms. These are designed to be used on terminal
that have video attributes such as reverse fields (shown in the figures as dark areas),
bold letters  (not shown in figures), blinking areas and letters for attention. Each

                                     79

-------
Experimental Data                                        Run Data Processing Status Databases

form also has on-line HELP for each field, and if more help is needed, the screens
have full-screen help forms with such information as legal key values and HC species
numbers.

    Several report procedures have been developed to help guide the data processing.
These look up the status in RUMEHTRY for specific tasks for each "type" of personnel:
PCREPORTS. PPREPORTS. CTREPORTS. These were explained in Chapter 2. An example of
the PCREPORT is shown in Figure 20.
                                      80

-------
Run Data Processing Status Databases	Experimental Data


   Table 14. Record Description for RUNENTRY Database System.


      01  RDN_REC.
      02 GENERAL.
         03  RDN_DATE       USAGE  IS  DATE
            EDIT_STRING IS DD-MMM-YYYY.
         03  CONDITIONS     OCCURS 2  TIMES.
            06  SIDE        PIC X(4) .
            06  NOXCONC     PIC 9.999.
            06  HC          OCCURS 3  TIMES.
               09 HCCONC     PIC 99.99
                 MISSING VALUE IS 0.00
                 EDIT.STRING IS  Z9.00?"      ".
               09 SPECIES  PIC 999.
         03 PROJECT        PIC X(4) .
         03 QUALITY        PIC 9.
         03 PRIORITY       PIC 9.
      02 DETAIL.
         03 DVMDATA.
           06  ITEM        OCCURS 10 TIMES.
               09 DVEVENT    PIC X(3)
                 DEFAULT IS "non"
                 VALID  IP DVEVENT IN DVEVENT_TAB.
               09 DVEDATE    USAGE IS DATE
                 EDIT_STRING IS DD-MMM-YY.
               09 DVINIT     PIC X(3).
        03 DOCUMENTATION.
           06  ITEM        OCCURS 5 TIMES.
               09 DOEVENT    PIC X(3)
                 DEFAULT IS "non"
                 VALID  IF DOEVENT IN DOEVENT_TAB.
               09 DOEDATE    USAGE IS DATE
                 EDIT_STRING IS DD-MMM-YY.
               09 DOINIT     PIC X(3) .
        03 FINAL_FILE.
           06  ITEM        OCCURS 5  TIMES.
               09 FFEVENT    PIC X(3)
                 DEFAULT IS "non"
                 VALID  IF FFEVENT IN FFEVENT_TAB.
               09 FFDATE    USAGE IS DATE
                 EDIT_STRING  IS  DD-MMM-YY.
               09 FFINIT     PIC X(3) .
        03 PICKDATA.
           06  TOTAL.SPECIES   PIC 99.
           06  MUM          PIC 9
               MISSING VALUE IS  0.
           06  INST        OCCURS  0  TO 9 TIMES DEPENDING ON  NUM.
               09 ID        PIC X(3).
               09 ITEM      OCCURS  10  TIMES.
                 11 PIEVENT PIC X(3)
                    DEFAULT IS  "non"
                    VALID IF PIEVENT IN PIEVENTJTAB.
                 11 PIDATE USAGE IS  DATE
                    EDIT_STRING  IS  DD-MMM-YY.
                 11 PIINIT  PIC  X(3).

                              81

-------
                                                                      Table  15.
                                                                                                                                            7)
                                                                                                                                            3
                                                                                                                                            ft
                                                                                                                                            n

                                                                                                                                            »
                                                                                                                                            a
                                                                                                                                                          D
                                                                                                                                                          v
                                               IIUNENTRY Processing Summary for September 14, 1081
                                           Run Detei U-SEP-
  Priorityi  t

HCCONC   SPECIE
  Z.72    191
 ««.7i    i/a
 ii.ii    tni
HAHE
Uncut*
Ii2,4-trInethylbonienc
                      -COMPLETED  OVM  EVENTS-
oo
to
URU Unpick pan run  IZ-OCT-I98I  MOS
V2T At tip* era.ted             MOS
UPC Riw Plot cr««td             RFP
PDA Plot OA conpltd             KCS
CfS C.I f.ct deter* 29-OEC-I98I  KGS
0«U OVHriX pg* run   7-JAN-I98J  Kit
COA Cone OA conpltd'- 7-JAN-I983  KCS
R2M Reldy to nerge   J-JAH-I98J  HEJ
               	COMPLETED INSTRUMENT  EVENTS	
               lnitru*enti   CIG Ctrlo I  CC
               PID Pick Initrc d.t I8-MAY-I9B2  KCS
               F«U PICK pg* run    27-OCT-I982  Jill
               rPT Flit plot creet 27-OCT-I9B2  JIH
               POA fit pit  OA coup 2H-OCT-I982  KCS
               CfO C.I feci deter* 2S-HAY-I9B2  KGS
               CRU CAICOII pgn run  27-OCT-I982  JLH
               HRU HCANAL pg* run  27-OCT-I982  JIH
               COA Coop Chi.hi OA  27-OCT-I98Z  KCS
               f2V Flltl tr to VAX  7-JAH-I983  HFP
               V2U »«.dy to ».r9.   S-JAH-1983  H£J
               	COHPIETED INSTRUMENT EVENTS	
               Initrtoanti  V2C V.rl.n GC Igxl
               PID Pick Initrc det IS-MAY-1982 KGS
               Ml) PICK P9« run     7-APR-I982 JIH
               FPT f.il plot cre.t  7-APH-I982 JIH
               POA Fit pit OA coup            KCS
               CFD C.I f.c. deter* 2I-HAV-I982 KCS
               CRU CAICON P9« run  I6-OCI-I96Z JIH
               COA Coup Chklhl OA- 28-OCI-I962 KCS
               F2V rilei tr to VAX  7-JAH-I983 RFP
               VZU needy to Berge   S-JAH-1983 HEJ
                                                            	COMPLETED DOC EVENTS	
                                                            SUM Run Siinry conpl  29-OCf-l382  KGS
                                                            D2V Doe tr.n to VAX   7-JAN-I981  KGS
                                                            V2T AL tope ereited   9-JAN-I963  Off
                                                            T2D Tape trn to dik   9-JAH-I903  IIEJ
                                                            DOA Doe OA eonpletd   9-JAN-I983  HEJ
                                             	COHPIETED INSTRUMENT EVENTS	
                                             Inftruxenti  C2G C.rle II CC
                                             PID Pick  Initrc det I8-HAY-I982 KGS
                                             PRU TICK pg» run     S-APR-1982 JLH
                                             FPT F.It plot cre.t  S-APR-1982 JIH
                                             POA Fit pit OA comp 2B-OCT-I982 KCS
                                             CFD C«l f*c> deter* 2S-MAY-I982 KGS
                                             CRU CAICON P9« run   e-SEP-1982 COM
                                             HRU HCANAl pg* run   8-SEP-I982 CDH
                                             COA Comp Cnrlht OA  28-OCT-I992 KGS
                                             F2V Fllei tr to VAX  7-JAH-I983 RFP
                                             V2U Reidy to Kerge   9-JAN-I9B3 IIEJ

                                             	COHPIETED INSTRUMENT EVENTS	
                                             Initrunenti  COG Beckn.n C8B0 ICOI
                                             IAD Stop procenlng I8-MAY-I982 KGS
                                                                           	COMPLETED FINAL  FILE  EVENTS	
                                                                           MR) 3 lectl eierged   e.j/\N-l9B3  HEJ
                                                                           U2T NL tipe created  K-JAN-I9B]  HEJ
                                                                           	COMPLETED INSTRUMENT EVENTS	
                                                                           Initruvienti   FOG CEA Form. Idchyie
                                                                           PIO Pick Initrc det II-HAY-I9B2  KCS
                                                                           PRU PICK pg* run
                                                                           FPT Flit plot cre.t
                                                                           PpA Fit pit  OA conp
                                                                           CFD C.I Feel doter*
                                                                           CRU CAICOII pg* run
                                                                           COA Coup Chkiht OA
                                                                           F2V File! tr to VAX
                                                                           V2U R»dy to *erge
                                                                                       7-APR-I982 JIH
                                                                                       7-APR-I982 JLH
                                                                                                  KGS
                                                                                       l-JUl-1982 KCS
                                                                                       I-SEP-I9B2 KCS
                                                                                      28-OCI-I982 KCS
                                                                                       7-JAN-I9B3 RFP
                                                                                       (-JAN-I9B3 HEJ
                                                                                                          	COMPLETED  INSTRUMENT EVENTJ-
                                                                                                                            "0
                                                                                                                            •i
                                                                                                                            o
                                                                                                                                            (n
                                                                                                                                            c»
                                                                                                                                            V
                                                                                                                                            c

-------
      Run Data Processing Status Databases                               Experimental Data
                       SMOG   CHAMBER


               RUN     TRACIN6    PROGRAM


                 ENTER NAME:     Harvey



                 ENTER DATA YEAR:      85
                  MAIN      MENU


            1   search -For a  RUN using the  run date
            2   display run record
            3   print run record
            4   update information (but not run date)
            5   delete records
            ft   create run records
            7   change run date of existing run
            B   print -from list file
            9   ex i t

           ENTER CHOICE  (1-9):     AND PRESS 


           Press PF2 for Help.
Figure 16.  Runentry Welcome and Main Menu Screens

                                    83

-------
         Experimental Data	     Run Data Processing Status Databases
                                          SELECT   A   RUN

                            ENTER THE DESIRED RUN DATE:  05-aug-19a4 AND PRESS  (RETURN).
                             (A run did not hive to occur on this date, but  it helps.)
                           If the  fdialing Run.Date  is not the sate as the one entered,
                                then a run Kith the requested date does not exist.

                           Run.Date  5-AUS-1984             Project NETH
                           Red Hydrocarbon                  Blue Hydrocarbon
                             SYNTHETIC AUTO EIHAUST           SYNTHETIC AUTO EIHAUST
                             ARQHATIC HU                     AROMATIC HIS
                             FORMALDEHYDE                     FORMALDEHYDE
                                         IS THIS  THE CORRECT RUN ? .y

                                 (Y) yes, establish as current run.date
                                 (L) no, leave  search and return to tain tenu
                                              SELECTION OF DISPLAY

                                   1  shot GENERAL run  intonation
                                   2  sho» INSTRUMENT run information
                                   3  Shan DVN run intonation
                                   4  shod DOCUMENTATION and FINAL FILE intonation
                                   5       return  to ijin lenu

                             ENTER  CHOICE (1,2,3,4, or 5):    AND HIT (RETURN)
Figure 17.   RUNENTRY Date and Display Selection Screens

                                                       84

-------
          Run Data Processing Status Databases
                                                                                          Experimental Data
                                        KEY
                                              0 V N    DATA    STATUS
                                                Run Date:  5-AU6-1994
                   EVENT
DATE
INITIALS
URU
ACS
RFC
U2V
PQA
V2T
CFS
DRU
CQA
nan
Unpack pga run
Auto cal stripd
Ran Plot creatd
Data trn to VAX
Plot SA coapltd
Van to tape
Cal facs detera
DVI1FIX pgi run
Cone SA coapltd
Not yet coapltd
9-AU6-1984
9-AUB-1984
9-AU6-1984
9-AUB-1984
4-SEP-1984
9-AUB-1984
30-HAY-1985
30-HAY-1985
30-MAY-1985

JKJ
JKJ
JKJ
JKJ
K6S
JKJ
KGS
JLH
KBS

                                 press RETURN-key to continue
                                     PICKED  INSTRUMENT  STATUS
                                                Run Date:  5-AU6-1984
                                                            Instruoent  selected: C2B
                        INSTRUMENT ID AND DESCRIPTION    KEY
                                  EVENT
               DATE     INITIALS
                          AT6
                          C16
                          C2B
                          C3S
                          FOB
                          V26
ATC electron ca
Carle I BC
Carle II BC
Carle III GC
CEA Foraaldehyd
Varian 2 BC (ga
PID
PRU
FPT
PQA
CFD
CRU
non
non
non
non
Pick instrc det
PICK pga run
Fast plot creat
Fst pit QA coap
Cal facs detera
CALCON pga run
Not yet coapltd
Not yet coapltd
Not yet coapltd
Not yet coapltd
20-AUB-1984
20-AUB-1984
20-AUB-1984
5-JUN-1985
5-JUN-19B5
5-JUN-1985




KGS
JLH
JLH
KBS
KBS
JLH




                    Enter Y for  another instrument, ENTER-key to continue


Figure 18.   RUNENTRY General, Documentation, and File Status  Screens

                                                       85

-------
          Experimental Data
                                                                            Run Data Processing Status Databases
                                          GENERAL   INFORMATION

                            RUN DATE  3-AUB-1984     PROJECT HETH      QUALITY 9  PRIORITY  9

                            SIDE RED     NO! 0.350   HC  00.41  SP 403  SYNTHETIC AUTO EIHAU
                                                    HC  00.59  SP 698  ARQHATIC nil'
                                                    HC  00.02  SP 139  FQRHALDEHYDE
                           SIDE BLUE    NOX 0.350   HC 00.27  SP 603  SYNTHETIC  AUTO EIHAU
                                                   HC 00.33  SP 498  AROMATIC I1II
                                                   HC 00.01  SP 139  FORMALDEHYDE
                                            SUGARY OF PROCESSING STATUS

                            DVHDATA  9  PICKED DATA 4 of &   DOCUMENTATION  0   FINAL FILE 0


                   press  RETURN-key to continue
                     DOCUMENTATION   AND   FINAL   FILE    INFORMATION
                                            Run Date:  5-AU6-1984
                    DOCUMENTATION:
KEY
EVENT
DATE     INITIALS
                                        non   Not yet coipltd
                                        non   Not yet coipltd
                                        non   Not yet coipltd
                                        non   Not yet coipltd
                                        non   Not yet coipltd
                          FINAL FILE:    KEY
           press RETURN-key to continue
         EVENT
                                        nan   Not yet coipltd
                                        non   Not yet coipltd
                                        nan   Not yet caipltd
                                        non   Not yet coipltd
                                        non   Nat yet coipltd
               DATE     INITIALS
Figure 19.   RUNENTRY DVM and Picked Instrument Status Screens

                                                        86

-------
Run Data Processing Status Databases                                         Experimental Data

                        Figure 20.  PCreport Example
              Count  of  runs  that have reached  stated  stages of processing
                         for ALLRUN84
              Total  runs  Total  Inst
                 63        312
              id*******  DVM data ************
              URU RFC ACS MAC U2V V2T PQA CFS  DRU CQA R2M NTB BAD UDV
              58  58   58   0   57  57 54  38  37  37   0   2   1   3
                  Sum  of URU,  NTB. and NDV is   63
              ########  Inst  data ############
              PID PRU FPT PQA CFD CRU HRU CQA F2V V2U NTB BAD
              190 190 190 116 104 104  0  17   0   0 135   6
                  Sum  of PID and NTB is  325
              ########  Documentation  ############
              SUM D2V DQA
               772
              ######## Final File ############
              MRS ALT FQA
               000
              Number of P-files should be :  190
              Number of C-files should be :  104
              Number of K-files should be :  104
                                       87

-------
Experimental Data                                         Run Data Processing Status Databases

                    Figure  20(cont.).  PCreport Example

     ALLR runs/instuments needing Picking Instructions Determined  (PID).
           This is an  HIST first processing stage report.

                               PRIORITY >» 8  <«
              26-Jun-84  Humber of  inst =   8
                   PCG  PE 900  GC
                        (chamber)

                               PRIORITY >» 5  <«
              25-Jun-84  Number of  inst =   8
                   PCG  PE 900  GC
                        (chamber)
          ALLR runs/instuments  needing  Cal  Factors Determined  (CFD).
         The following instruments  have ( PID  ) Pick inst det completed
                               but not  ( CFD  ) Cal fac det completed.
                RU!!
                DATE         Instruments            Date done     Person

                              PRIORITY >» 9  <«
              25-Jul-84  Number of  inst =   5
                  V2G  Varian GC (gas)
                        PID    Pick inst det        6-Aug-84      KGS

                              PRIORITY »> 5 <«
              25-Jun-84  Number of inst =   8
                  COG  Beckman 6800 (CO)
                        PID    Pick inst det       31-Jul-84      JA
                  V2G  Varian GC (gas)
                        PID    Pick inst det       31-Jul-84      JA
                                       88

-------
 Use of Data Processing Status Database                                     Experimental Data

 Description  of Data Processing Status Database
 This section describes the data processing status database and how it is used.

 Description
 The main  parts of the data processing status database are the calendars and the
 RUNENTRY DATATRIEVE system.  RUUEHTRY maintains status information for each experi-
 ment by date. The general categories are:
  • GENERAL
      o  experimental chemical conditions for both sides,
      o  quality and priority index,
      o  number of experiments
  • INSTRUMENT
      o  pick instructions,
      o  cal factors given,
      o  qa  performed,
      o  final processing steps performed;
  . DVM
      o  unpacking performed,
      o  cal  factors given,
      o  QA steps  performed,
     o  final processing performed;
  • DOCUMENTATION
     o  final documentation processed and written,
     o  moved to  VAX for final segfile formation.

Use of Data Processing Status Database
This section describes how  the database  is  used from the time the  experiment is
conducted to final processing.

Log Experiment
When an experiment  is  conducted (attempted), the conditions and general qual-
ity are  indicated  on both calendars (site  and office).  This information is usually
obtained by telephone conversations between the project coordinators and site per-
                                     89

-------
 Experimental Data	Use of Data Processing Status Database

 sonnel within a few hours after the start of the experiment in order to plan for
 the next days activities.  Another late afternoon telephone conversation follows to
 determine the outcome of the experiment and to finalize the next, days plans. If the
 experiment is thought to  be successful or might provide any the useful information,
 it is considered to exist. The absence of major negative comments such as "failure"
 written on the calendar,  triggers the data processing procedure.  The experiment
 is packed-up and inventoried if it is not  considered a "failure" for transport to the
 dataprocessing office.
    From the-telephone conversation mentioned above, the experiments success is
 determined.   If the experiment is  considered worth  archiving, the actual experi-
 ment conditions and general quality are recorded on  the calendars. Otherwise the
 attempted experiment and the problems associated are noted on the calendar.
    Data processing personel expect both the runfolder and data floppies to be
 transported to the school  within a few days after the experiment. Special locations
 in the office are set aside for incoming  data.  The calendars and  these areas are
 checked daily.  If data is late, the site personel are contacted and questioned. When
 either the run  folder or data floppies arrive at the office, the experiment  is entered
 into the RUHEHTRY database.
    The contents of the runfolder are first inventoried before any data processing
 occurs.  The instrument  checklist completed during  the experiment is  compared
 with the folder contents. Missing items are investigated.

 Preliminary Review
 Preliminary review usually begins the day or day after an experiment is conducted
 by a PC. Conversations with the site operators indicate  resulting experimental condi-
 tions and outcome and any problems which might have occured.  The run folder  con-
 tents are inspected. Instrument operation behavior,  chamber background(proper
 venting, drying or absence of condensation), temperature and dewpoint, sunlight,
 initial  conditions, and reasonableness of results are determined. Basic quality and
 processing priority is determined. Completeness for data and documentation is de-
 termined. The calendars  and the RUUEHTRY database  are updated.  Besides a QA
 step for the experiment itself, the success of meeting the experimental plan goals is
 checked. A variation in the experiment might be  useful if documented but would
 result in rescheduling of the target experiment.

 Processing Instructions and Information

RUHENTRY is used to track data processing by recording of processing events. Auto-
matic status reports can be generated.  Report procedures are used to guide  and

                                     90

-------
 Digital Voltmeter Data Processing	Experimental Data

 assign processing tasks: PCREPORTS, PPREPORTS. and CTREPORTS. The reports are listed
 by priority (0-9) as determined by the PC.
    PCREPORTS  tells the  project coordinators  that  picking instructions are  needed,
 or picked  data exists and needs  QA and  calibration factors,  or final data exists
 and needs QA. CTREPORTS indicates to the computer technicians that DVM data needs
 unpacking and archival, and AUTOCAL data needs to be stripped. PPREPORTS tells the
 peakpickers that pick instructions have been given, or that picked data has been
 QAed and calibration factors exist for final processing.
    Each worker updates RUlJEHTRY as each step  is performed.  The presence of each
 new step completed generates instructions to the next appropriate worker to perform
 the next processing task when a new report is generated. When all processing steps
 are completed, the segmented file is assembled, and the final initial experimental
 conditions and quality  are updated in the  RUNENTRY GENERAL section. This data
 is input  to the RUNSUMMARY database which is used to sort, organize, and analyze the
 total smog chamber experimental  database.


 Digital Voltmeter Data Processing

 The Digital Voltmeter Data (DVM) from the Computer DA system is usually processed
 easily  once the calibration database is  established.  The  processing flowchart  is
 indicated in Figure 21.  The RUtiEHTRY events flowchart is shown in Table 16.
    The computer technician (CT) "unpacks" the data from the two floppies which
 arrive from the research site.  Several files generated by the site DA system are copied
 to the LSI-11 harddisk: the main  DVM file (mmddy.DVM), the remarks file (mmddy.REM),
 the command file (mmddy.CMD), the adjustments file (mmddy.ADJ), and the file of actual
 computer commands  (mmddy.OPS). Printouts of all of these files except the mmddy.DVM
 file are made and stored in the RUNFOLDER. A computer program (UNPACK) is run which
 produces a "U" file, which is the  same raw data but in ASCII  format and is the
 principle raw data file for the smog chamber experiment. This file contains 24 hours
 of data taken every minute for every instrument connected to the DVM. The "U" file
 is back-up  onto two sets of floppies,  and to  two types of magnetic tape used on
 different  computers: the VAX-780 and the University's IBM 4381.
    Several programs are used to  process  data from the "U-file".  FASTDV is  run
several times:  to produce "P" files for plotting the raw data ("fastplots") from the
experiment and to strip the AUTOCAL data (makes  "A" file). The processing of the
calibration data was discussed earlier. PLOPIC is used to make  the plots from the
 "P" files for the NOX/O3 plot and  the TSR/UV, temperature,  and dewpoint plot.
Theses plots are stored both in the RUUFOLDER and  the FASTDV notebook. All of the

                                      91

-------
    Experimental Data
                   Digital Voltmeter Data Processing
                           mmddy.DVM, REM,

                                     SPG,  CMD
                 UNPACK
                     i
DVMFKf-
                         mmddyU.DVM


                               FASTDV-
                             CAL

                            FACTORS
                        e^LSIlQ

                          DISK	I
       PLOCONf
          I
                  "OFFICIAL'
Figure 21.  DVM Processing Flowchart (Computers and Files)


                          92

-------
Digital Voltmeter Data Processing
                                                        Experimental Data
                                    Table 16.
                     Processing System for DVM Data
          cal processing
          /             \
        tot runs       (CFS)	
          \             /            \
          V        (ACS)              \
           \     /                    \
           (URU)	(RFC)	!PQA!---!CQA!--!R2M!
                \                     /
                 (UTV)               /
                      \             /
                       (V2T)	
                !sss! == stages determined by Project coordinator (PC)
                == stages determined by Peak Pickers (PP)
                (sss) == stages determined by Computer Techs (CT)
    Pers
     DVM data processing steps
Step                   Meaning
CT
CT
CT
CT
PC
CT
PC
CT
PC
CT
PC
PC
CT
URU
U2V
V2T
RFC
PQA
ACS
CFS
DRU
CQA
R2M
NTB
BAD
NDV
Unpack run
Data to VAX
to archive tape
Raw Plot done
Plot QA
AutoCals strip
Cals done
DVMFIX run
Cone QA
Ready 2 merge
Not to be proc
Stop processing
No DVM data
Site floppy data expanded to ASCII Ufiles
Ufiles moved from LSI disk to VAX
Data moved from VAX to tape
Plot of voltages to examine data
Initial quality check
Cals separated for processing
Calibration factors determined
Voltages changed to concentrations
Concentration Quality Assurance
Merge process on VAX can be run
Not to be processed
Something is wrong with this data

                                       93

-------
 Experimental Data	Digital Voltmeter Data Processing

 tasks discussed so far are done by the CT. The CT update RUNENTRY as these tasks are
 performed.
     When the FASTDV plots are made and RUIJEIITRY has been updated so, the PCREPORT
 will direct the PC to perform a QA on the raw data.  The plot is compared with
 the original stripcharts.  If data is missing,  the  documentation  (RUNSHEETS,
 REMARKS FILES,  INSTRUMENT CHECKLIST, CALENDARS) is investigated
 for mention of power or computer failures, or instrument malfunction. The data
 may need to be retransfered from the site. Missing data may have to digitized from
 the stripcharts. Once the  raw data is QAed, the PC updates RUliEMTRY that the raw
 data has been unpacked successfully  (PQA) and calibration factors can be applied
 (calibration factors need to be determined).
    Calibration factors are applied to raw DVM data with a program called DVMFIX
 which runs on the LSI computer. The program also controls which time periods of
 raw data are processed.  Bad data, or cal data is eliminated in this way. Calibration
 factors are obtained from a list resulting from calibration factor processing discussed
 above. Interpolation may  be required. Usually however NOX and Oz monitors are
 calibrated twice a day, before and after each experiment.  The calibration factors
 can be used directly.
    The PC fills out a DVMFIX sheet indicating the calibration factors to apply to the
 data, the periods of valid data to process,  and brief documentation. The calibration
 factors are obtained  from  the AUTOCAL report as described  above.  The  PC updates
 RUNEliTRY that calibration factors have been determined.  The  information on the
 DVMFIX sheet is transfered to a file on  the LSI-ll computer (mmddy.CRD)  usually by
 a PP or CT. The "U"  file is moved  back to the harddisk.  This might require the
 assistance of the CT if these files are not still on floppies and need to be copied off
 of one of the magnetic tapes.  The program DVMFIX is used to apply the  calibration
 data in the mmddy.CRD  file to the raw data, producing a new file of corrected data for
 only valid experimental data for every fourth minute (mmddyG.DVM "G-file"). DVMFIX is
 usally run by the person who creates the mmddy. CRD file.  The "G-files" are backed-up
 on floppies. RUNEMTRY is updated that DVMFIX has been run and final QA is need by the
 PC.
    The  final corrected data is listed and plotted.  Concentration data for NO and
 NC>2  must equal  NOX concentrations within a few ppb throughout the experiment.
 "Zeros"  are checked.   Initial conditions are compared with the target conditions.
 Dewpoint, and temperature data are checked  for consistency  and  reasonableness
 (dewpoint data cannot be higher than chamber temperature).  Profiles  and maxi-
mum values of light are compared with other seasons for resonableness. Data not
to be processed or missing  should have "-9.9999" indicated. If there are problems,
the PC will re-issue processing instructions, deleting from RUNEIITRY that  DVMFIX has

                                      94

-------
 Strip Chart Data Processing	Experimental Data

 been run. This will cause the PPREPORTS and CTREPQRTS to request that the data be
 reprocessed. The process is repeated until the QA criteria are satisfied.

 Strip Chart Data Processing

 Processing strip chart data is a little more complex than processing DVM data. More
 data is represented, more calibration factors are required, and there are more sources
 of error.  Basically the data are identified, the conditions  of the instruments are
 identified, the data is picked (digitized), plotted, and QAed, calibration factors are
 determined and applied, and the final data is listed and plotted for final QA. The
 processing flowchart is indicated  in Figure 22.  The RUMEKTRY events flowchart is
 shown in  Table  17.
    Data processing starts with RUNENTRY indicating to the PC in the PCREPORT that pick
 instructions are needed for the instruments used in an experiment. Data processing
 starts  with  the  documentation. The RUHSHEET and RUNENTRY indicate which species
 where injected for the experiment.  The instruments which require processing are
 identified. The stripcharts are investigated for proper documentation: attenuation,
 chart full scale  setting and chart speed, time,  and  chamber side  sampled.  The
 chromatograms  are inspected for proper appearance (operation). The background
 chromatograms  are inspected for contamination. Major problems are noted. If the
 data is valid but not useful the t.'TB is indicated  in RUllEliTRY. If the data  is bad, BAD
 is indicated.  No further action will be taken.
    If the instrument seems to be free of problems and data are considered useful,
 then pick instructions are given.  Calibration data are identified and noted as not
 to be processed by the peakpicker  (PP). The compounds monitored and their peaks
 are identified and labeled. The entire stripchart is  inspected for reasonable appear-
 ance and  operation and data.  Problems are identified and noted. The peakpicker
 can indicate in the data that data are missing (off-scale, power failure, instrument
 adjustment, etc.). The start and end times of valid data are identified. The com-
 pounds to be picked are  listed on the DATA PROCESSING  INSTRUCTION  FORM with any
 additional instructions. RUNENTRY is updated with PID (pick instructions determined)
 by the PC. This indicates in the PPCREPORT to the PP that data is ready to be picked.
    The PP obtains the run folder and determines from the DATA PROCESSING INSTRUC-
TION FORM which compounds are to be picked. The stripchart is removed and spread
out over the digitizer pad.  Using  the instructions and  peak identification on the
stripchart the data is digitized using the program DIGPIK run on the LSl-ll computer
to produce a file named mmddyP.iit ("P"  file).  The files are backed-up on floppies
and magnetic tape. The data is then plotted with a program called PLOPIC on the
LSI-ll with GIGI graphics.  A hardcopy is stored in the  RUtJFOLDER. The PP updates

                                     95

-------
     Experimental Data
             Strip Chart Data Processing
                    DP

                   inst
t
                              RUNFqLDER
                             stripchart
             PP

            Inat
                             DIGPIK
                                       mmddyP.iit
          HCANAL
             or
         LOTUS 123
                                       mmddyC.iit
                                       mmddyK.iit
                                           PLOCON
Figure 22. Instrument Processing Flowchart (Computers and Files)

                               96

-------
Strip Chart Data Processing
                                                               Experimental Data
                                   Table 17.
                  Processing System for Instrument Data
    ! tlTB !
   /
tot inst  __ cal processing.
   \     /                  \
    !PID!                    !CFD!
                                              __
                                 \
                                                     \
\
                                  
                                                      !CQA! ----  (F2V) ----  (V2U)
               \
                ----  --IPQA!
               !sss! == stages determined by Project  coordinator  (PC)
                == stages determined by Peak Pickers  (PP)
               (sss) == stages determined by Computer Techs  (CT)
    Pers

    PC
    PC

    PC
       RUHENTRY stages
                                Meaning
       PID  : Pick inst det
       NTB  : Not to be proc

       CFD  : Cal fac det
PP
PP
PC
PP
PP
PC
CT
CT
PRU :
FPT :
PQA :
CRU :
HRU :
CQA :
F2V :
V2U :
Pick run
FPlot done
FPlot QA
CALCOU run
HCAMAL run
Comp CS QA
File 2 VAX
Ready 2 merge
    PC
       BAD :  Stop processing
                     Charts marked, Inst status sheet in folder
                     Ho futher processing marked on Inst Status
                     sheet
                     Cal factors for each comp entered on Inst
                     status sheet in folder
                     Data has been digitized and P-file exists
                     P-file has been plotted and plot is in folder
                     Fast plot has been marked OK (no bad data
                     points in  P-file)
                     C and K files exist on floppies
                     HC analysis printout in run folder
                     Instrument QA completed on C and K files
                     Initial Conditions updated
                     C and K files moved to VAX
                     C files merged together,  K files merged
                     together
                     Something is wrong with this data
                                       97

-------
 Experimental Data	Strip Chart Data Processing

 RUIIENTRY that the data has been picked  (PRU: pick run) and a "fastplot"  has been
 made (FPT). This indicates in the PCREPORT to the PC that raw data has been picked
 and requires QA.
    The  "fastplot" is inspected and compared with the original stripchart. Appro-
 priate chemical behavior is noted. The stripchart data is compared to assure that
 the data is  picked correctly: time and side for chromatograms and amplitude of
 peaks.  When possible, picked data is inspected for consistency:  similar compounds
 on the different sides should show similar behavior, and the least reactive side should
 match the picked data side indicating less or less reactive material.  If additional or
 correctional picking is required, the PC deletes the PRU and FPT entries, and changes
 the date for  the PID instructions to the current date.  This will indicate to the PP that
 picking is required. Comments on the DATA PROCESSING  INSTRUCTION FORM will indicate
 the new work required with notes added by the PC during the last  QA procedure.
 Once the correction PP work is  performed, RUNENTRY will be updated with PRU again
 which will indicate raw data QA procedures are needed again to the PC in the next
 PCREPORT. If the data passes the QA tests, the PC enters PQA to RUNE1ITRY which will
 indicate to the PC in future PCREPORTs that calibration factors are.needed for further
 data processing.
    Once the picked  data is QAed, calibration factors need to  be  determined for
 final data processing: conversion .of raw picked data ("P  files") to concentration
 data  ("C" files).  The  techniques  for determining appropriate  calibration factors
 was discussed above. The calibration factors are indicated  on the DATA PROCESSING
 INSTRUCTION FORM. The PC indicates that calibration factors have been determined by
 entering into RUIIENTRY CFD (calibration  factors determined).  This will indicate  to
 the PP that raw data is  ready  to be converted to concentration data in  the  next.
 PPREPORT.
    A program (CALCOM) which  runs on  the  LSI-11 computer is  used to apply the
 calibration factors to the raw data ("P"-files), and to document which factors were
 used, who determined them, who is running  the program and when, the full names
 of the compounds (abbreviations are used to identify  data fields),  and the maxi-
 mum amplitudes of the  raw and final concentration data.  Two separate files are
 produced, the actual  concentration data (mmddyC.lit, "C" file) and the documenta-
 tion (mmddyK.lit, "K" file). The  resulting files are listed, and the  data is plotted for
 final QA with PLOCOH, a plotting program using GIGI graphics and the LSI-H. These
 are retained  in the run folder. The PP indicates that the CALCOH program was run by
entering CRU. This will indicate to the PC that concentration files  are ready for final
 QA in the next PCREPORT.
    The QA  procedure for the final concentration data varies with the complexity
of the experiment and the chemical species.  The first checks however  are to look

                                      98

-------
 Documentation Processing	Experimental Data

 at the documentation file ("K" file) and the DATA PROCESSING  INSTRUCTION FORM to be
 sure that the desired calibration factor(s) was actually applied. The PLOCON plot is
 inspected to determine if the resulting data is reasonable and consistent with the
 other experimental data and the indicated target conditions on the RUN SHEET. If the
 concentrations of the initial conditions are far from the target conditions, a reason-
 able explanation must be found. Often the  RUN SHEET, REMARKS FILE printout, or the
 INSTRUMENT CHECKLIST will indicate problems  which could explain the discrepancies.
 Usually resulting data is not more than 25 percent from the  target conditions.
     If the concentration data is for a standard HC mixture, a compositional analysis
 should conform well with the known composition.  An option  of the CALCON program
 is to produce files which are readable  by a computer program (HCAHAL) for com-
 positional analysis.  In some cases  where several experiments are performed  with
 a similar mixture, the concentration data for the initial conditions for all the ex-
 periments are entered into a LOTUS spreadsheet for relational  composition analysis.
 Calbration factors should result in data which when analyzed yields a consistent
 composition if indeed a similar mixture was used throughout the program. QA pro-
 cedures may indicate that the calibration factors are not reasonable or that the raw
 data is invalid.  See  the discussion  above for determination of calibration factors.
 Invalid data may require special processing, or a decision that the data is not pro-
 cessable (BAD). Once QA procedures are successfully performed. RUHEiiTRY is updated
 with CQA  (concentration data QA performed). The  "C"  and "K" files are backed-up
 on floppies.  If new factors are needed, the CRU is deleted and the CFD date is updated.
 New factors are indicated with explanations on the DATA PROCESSING  FORM.
    Once the final QA procedures are completed, the "C" and "K" files are trans-
 fered from the LSI-H to the VAX  (RUNEHTRY F2V) to await merging with the DVM data
 and  the documentation, to form the final segmented data file for distribution.  The
 "P", "C", and "K"  files are backed  up on VAX magnetic tape.  The RUM.CAL.USED
 database is  updated with the calibration factors used in data processing.  Initial
 conditions are updated in RUNENTRY in the general section and  RUHSUMMARY.

 Documentation Processing

 Documentation is associated with all phases of the experiment; recorded on many
sources.  Much is  needed only to process the data.  Some  documentation data
 however is processed  to be  included with the final data  files. Calibration  factor
 documentation is processed as the data is processed ("K" file, see discussion above).
 Some other  documentation however is  needed, which  increases the value of the
resulting final  data.
   A general documentation file is  produced in addition to the concentraion data

                                     99

-------
 Experimental Data	 General Documentation Form


 files  (mmddyy.DOC). It  becomes the first part of the  final  segmented data  file.  Its
 purpose is to convey all information with the data  file itself, required to  describe
 the experiment, and  data which might be useful to photochemical modelers.  The
 processing flowchart  is indicated in  Figure 23.  The RUllEiJTRY events flowchart is
 shown in Table 18.


                                      18

                           Documentation Steps

                   DOCUMENTATIO!! data processing steps
     Pers    Step                  Meaning

     PC       SUM :  Run sum done     DVM,  Instrument, and  Raw Quality  Combined.
     CT       D2V :  Doc to VAX      Documentation transferred to VAX
     PC       DQA :  Doc QA           Documentation Quality Assurance
General Documentation Form
A five page paper form called General Documentation is used throughout all stages of
the data processing. All documentation needed to be included in the final general
documentation file is recorded on this form as data are processed. In addition the
form includes checklists and formulas used to arrive at a general quality index that is
specified in the final documentation file. An example form is included in Appendix
E.

General Description and Purpose of Experiment
The first part of the general documentation file is the experimental run date which
is the primary identifier of an experiment.  A  general description of the experi-
ment follows. The purpose of the experiment is included if it is not clear from the
description. Results  are sometimes added if a specific comparison was being made.

Initial Conditions
The initial conditions of the experiment are listed  next.  Concentrations for each

                                     100

-------
     Initial Conditions
                                                 Experimental Data
                                  A
                        RUNFQLDER


'C'ffle

't
'G'file
                         EDIT

1
PLOTS

PLOTS
                                PLOCON   CSTAR
                   mmddyy.DOC
Figure 23.  Documentation Processing Flowchart (Computers and Files)



                              101

-------
 Experimental Data	Meteorological Conditions

 species for each side are listed. These values are obtained from the  "G" and "C"
 files, usually at sunrise. If the initial  conditions are from data  taken after sunrise
 then the data  time is reported.  Initial  conditions are updated in RUHEliTRY in the
 general section.

 Meteorological Conditions

 Next the general  meteorologial conditions  (solar radiation and temperature) are
 indicated. Detailed data from IIOAA reports and TSR stripcharts are the basis for a
 general ranking value computed for both before and afternoon.

 TSR data compared against RDLJ data

 Total solar radiation data  is compared with the HOAA data collected at the nearby
 RDU airport. The data profile for the experiment recorded at the site is inspected
 for "holes". A  1 to 10 grading scale is used  to rate the experiments solar radiaion.
 The holes are  counted and used to assess the grade value for both morning and
 afternoon portions (the morning sun is thought  to be more important).

 Temperature compared against RDU airport data
 The temperature at sunrise and at the maximum is compared with the RDU airport
 data for consistency and listed.

 Data Times,  Data Exceptions, Special Problems and  Concerns
 The times that the initial conditions are established, sunrise time and when the ex-
 periment ends are  given. Data exception  are listed. Special problems and concerns
 are listed.

 Quality Assessment
 An overall quality  assessment grading scale  has  been  devised to help  compare the
 large number of experiments.  The index ranges from 0 to 10 with 10 being a perfect
 experiment. Although  the  assessment of  the quality index is somewhat subjective,
 many characteristics can be quantified relatively.  There are  five categories which
 are each weighted equally.

 Solar Radiation
 For the earlier years of the dataset  sun light was the principle factor used for rep-
resenting run quality.  Starting in  1981  more formal  methods were developed:  a
formula which  would grade the sunlight quality more consistently but  with the

                                     102

-------
 Quality Assessment	Experimental Data

 same general techniques used earlier:  a) the percent of possible solar radiation
 measured for the day. b) the number "holes or dips" in the radiation profile, and
 c) morning fog or haze. The earlier sunlight profile of the day is usually considered
 more import than the post solar noon portion. Therefore the first half of the day is
 weighted 70of the total quality index. Meteorological data obtained at a local major
 airport (RDU)  lists percent of possible sun data which  has been found to correlate
 well  with  observed solar profiles and is used to establish the base index value:  90
 percent is converted to a quality index of "9".  Then "holes" (25 percent or more
 attenuation of expected  intensity) are divided and counted into three classes: 15,
 30 and 60 minute duration. Index units are subtracted for each "hole" weighed by
 duration class.
 Target Conditions Established

 There are three general  areas of concern: a) whether target conditions were ob-
 tained, b) whether the internal (one side consideration)  and external ratios (two
 sides compared) for each key condition parameter were obtained, and c) whether
 the experimental conditions are quantifiable or verifiable.
    The base sub-index  is  set at  "10".   One index unit  is deducted for each  20
 percent error from the absolute target condition desired.  Very often with the dual
 smog chambers it is the relative ratios of the initial component concentrations that
 is more important. There are two types of relative comparisons which are made: the
 ratios of components to each other within a mixture and the'ratios of components in
 two different mixtures  (side by side comparison). If a mixture is used, the relative
 fractions  of components  often needs to  be  carefully  established.  Very  often the
initial condition of interest is a matched HC to NOx ratio.  One index point is
 deleted for each 5 percent error from the target relative composition.
    Usually initial  conditions are verifiable analytically. When an analytical tech-
 niques fails or a sample is not taken before the beginning of an experiment, other
 techniques may have to be  used to establish or estimate the initial conditions. Di-
 rect injection  of neat liquids or gases are often used as a calibration technique and
 therefore can be used as a second best estimate of initial conditions. If a component
 of a commonly used mixture is not directly analytically monitored for a particular
 experiment, its concentration can be estimated from the concentrations of the other
 mixture components and  relative composition data. However these techniques are
 less than ideal and two index units are deleted from this quality category for each
compound not verified by an analytical technique.

                                      103

-------
 Experimental Data	Quality Assessment


 Analytical Support

 Analytical support is not always available for all species for  all runs:  techniques
 sometimes fail during an experiment. The base index is set at "10" and index points
 are deleted weighed by importance of the species or physical data not measured.
 Chamber Conditions

 Chamber conditions can have a major impact on experimental results.  Important
 parameters include:  a) condensation on walls,  b)  dilution rate, and c) hours of
 venting before the experiment.  The base index is set at "10".
    Condensation observed on chamber walls is noted by the operator starting the
 experiment. One index point is deleted if condensation is observed on both cham-
 bers. Chamber driers do not always operate well and  sometimes one side may be
 dried better than the  other resulting in condensation in one side only.  Two index
 units are deleted for this situation.
    Dilution rate is usually monitored by  use of inert tracers such as CC14.  A
 exponential least-squares fit analysis yields the dilution rate (necessary  for proper
 modeling) for each chamber. The quality  index is  deleted by  one if the  dilution
 rates differ  by more than 20 percent. Additional points are deleted proportionally
 for dilution rates exceeding the normal  dilution rate  of 1.5 * 10-4 min-1.  The index
 is deleted by two if the dilution rate is not  quantifiable (incomplete of poor tracer
 data). The quality index is deleted by one if the dilution rate is  high or irregular
 from windy conditions. Dilution rates are normally  well matched and less than 10
 percent per ten  hours.
    Two hours of venting will eliminate observable concentrations of  chamber pol-
 lutants. Regular experimental protocol requires that each run by preceded with at
 least six hours of venting.  Three index points are  deleted if less venting time is
 used.


 Overall Value—Need

 Other major aspects of an experiment are overall usefulness and value.  Many exper-
 iments are conducted to satisfy  questions requiring immediate attention  or special
 effort.  Often these experiments are difficult to conduct; the experimental quality
 may be less than desired but nevertheless  contributes usefull information which
may be otherwise impossible to obtain. These experiments may be supportive of
other experiments. This category is used to adjust the quality index to reflect these
experimental needs.

                                      104

-------
 Final Segmented File Production                                          Experimental Data

 Documentation File Production
 When all of the documentation information has been collected and the GEIiERAL DOC-
 UMENTATION form has been filled out, the PC updates RU!.'E!iTRY with the SUM event.  This
 will indicate to the CT that documentation  information is ready to be moved to the
 VAXll/780. The CT uses a template file and  a text editor.  When the documentation
 file is produced, the CT updates RUNEIITRY with the D2V event. This indicates to the
 PC in the next PCREPORT that the documentation file is ready for final QA. The file is
 listed and stored in the RUHFOLDER.
    The listed documentation file is compared with the GENERAL DOCUMENTATION form
 to be sure that at least the information was correctly transfered. The documentation
 information is checked one last time with general experimental results: plots of the
 experimental data are compared with the initial conditions for consistency of relative
 reactivity. The stated initial conditions and the data  plots should be checked for
 agreement. The quality indices for sunlight should be compared with the light plots
 for reasonableness.  The general run quality  index should  be checked for overall
 reasonableness having read the documentation  and looked at the data plots.  If
 it  is low,  the reasons should  be apparent in the documentation (e.g.missing data,
 chamber condensation present, or poor sunlight).
    When the  QA criteria are satisfied, the PC updates  RUHEI.'TRY with the DQA event.
 The GENERAL section of RUNEHTRY is updated with the  final  initial conditions  and
 the quality index  value. This  is usually the last step before producing the final
 segmented file.

Final  Segmented File Production

 When all data processing tasks are completed, the files are moved to the VAXll/780.
The files are merged together with a procedure called FIHALMRG which follows a pre-
scribed  order.  An ASCII labeled tape is made. Plots are made from the segmented
file as a QA step: if the format is incorrect the plotting programs will  probably
not be able to  read the file.  The file is listed and inspected.  The RUl.'El.'TRY event
flowchart is shown in Table 19.
                                     105

-------
Experimental Data
Final Segmented File Production
                                   Table 19.
                  Final Segmented  File Prodxiction Steps
    FINAL FILE data processing steps
    Pers     Step                  Meaning
CT
CT
CT
PC
PC
PC
MRS
.ALT
FPC
FQA
MTB
BAD
Merged 3 sect
Files to tape
Final plot done
Final QA
Not to be proc
Stop processing
Merge pieces to final file
Ascii labeled tape made
Correct cone plot, made
Final Quality Assurance

Something is wrong v.'ith this





data
                                      106

-------
 Backup of Data                                 •                     Experimental Data

 Security of Data

 There are three basic activities to enhance data security: inventory, retention  of
 original data to restricted areas, and maintaining several backup copies of all data
 in different locations. All data are under lock at night.

 Inventory
 Before anyone is allowed to work with the run folder contents, a run folder inventory
 is performed.which documents all data and forms. This helps assure that no data
 or documentation  is lost.

 Retention of Data
 All original data, both paper and electronic files are retained.  Paper formated data
 (all stripcharts, forms) are  not allowed out of the office.

 Backup of Data
 Multiple copies of data  are maintained in several  formats  in  several  locations.
 Backup copies of raw and final data on floppies on the IBM  PC are kept in several lo-
cations. Lotus worksheet files are backed up in two locations. Hardcopy (printouts)
of worksheet formulas using a spreadsheet AUDITOR is kept. Backup  copies of raw
and final data  on LSlli/23 computers are maintained in several locations, including:
raw data and processed data sets backed  up on two diskettes, processing programs
 (sources) backed up on two  diskettes, and whole image harddisk backup of LSlll/23.
Backup copies of final data are  maintained on the VAX  and further  backed-up on
tapes of VAX files.  Frequent printouts  of of data in all stages of processing ("P"-,
 "C"-, and "K"-files) and plots are kept in runfolders  and in notebooks.
                                     107

-------
 Segmented   Data   File
This section describes the final segmented data file used for distribution of the
experimental results.

General Description

All of the data for a run are stored in a single file.  This includes: documenta-
tion including the calibration information; the continuous instrument data, every
4-minutes, alternating on chamber sides in physical units; and the gas chromato-
graphic instrument data given in concentration units for each side of the chamber
for each instrument approximately every 30 minutes.
                             108

-------
File Formats
                                                                   Segmented Data File
File Formats

No record exceeds 80 columns (data is in "card image") form. The general lay out
of the file is:
     DOCUMENTATION segment 
          General  documentation section
     88888 Documentation section terminator>
          first instrument documentation ''K-file''  section
     88888 Documentation section terminator>
          second instrument documentation ''K-file''  section
     88888 Documentation section terminator>
     88888
         last instrument documentation section
     88888 
         DVH data documentation section
     99999    
         DVM Data, up to 3 ''cards1'  per  time
     99999    
         First instrument data,  up to 3 ''cards'
     99999    
         Second instrument data
     99999    
                                   per time
    99999

   An example segmented data file is shown in Table 20.
                                      109

-------
Segmented Data File	File Formats

                  Table 20. Example Segmented Data File
                         GE1JERAL DOCUMENTATION

    RUNDATE: OCTOBER 04, 1983

    RUNTYPE: AUTO

    RU1J DESCRIPTION: COMPARISON OF REACTIVITY OF EXHAUST
    FROM DIRECT INJECTION FROM DODGE CHARGER IN HIGH IDLE
    WITH SYNTHETIC AUTOEXHAUST.

    RESULTS: TV/0 SYSTEMS RESULTED IN SIMILAR REACTIVITY

    INITIAL CONDITIONS:           BLUE          RED
    DODGE CHARGER                 0.0           2.587
    SYNTHETIC EXHAUST NMHC        2.190         0.0
    NO                            0.214         0.215
    M02                           0.037         0.039
    88888
    04-OCT-83                                   GENERATED ON  19-MAR-84
    SY:OC043K.C1G       .C1G
    PICKED DATA ENTERED BY JEFFREY HOFFNER
    CALIBRATION FACTORS APPLIED BY CHARLES
    TOLUENE                           2.912000E-01    PPMC/IN       2     KGS
    ETHYLENE                          9.020000E-02    PPMC.IN       2     KGS
    NAME ABBREVIATIONS -  SAME ORDER AS IN DATA
    TOLUENE   IS  TOLUENE                           MAX A:   1.5610 MAX CON:   0.4546
    ETHYLENE  IS  ETHYLENE                         MAX A:   4.9164 MAX CON:   0.4435
                                      110

-------
File Formats
                                                                   Segmented Data File
               Table 20, cont.  Example  Segmented Data File
     88888
     04-OCT-83                                   GENERATED 0!i 13-MAR-84
     SY:OC043K.C2G        .C2G
     PICKED DATA EilTERED BY JEFFREY HOFFNER
     CALIBRATION FACTORS APPLIED  BY JEFFREY  HOFFNER
     ETHANE                             1.498000E-01    PPMC/IN       2
     PROPYLE1IE                          7.307000E-01    PPMC/I11       2
                                                                       KGS
                                                                       KGS
    NAME ABBREVIATIONS - SAME ORDER AS III DATA
    ETHANE    IS ETHANE
    PROPYLE1IE IS PROPYLEUE
                                               MAX A:   0.2014 MAX COIJ:    0.0302
                                               MAX A:   0.1311 MAX CON:    0.0958
    88888
    88888
    USER DOCUMENTATION FOR RUN 831004
    CALIBRATION FACTORS USED:
    SPECIE
     TIME     INTERVAL  BEGINNING    GAIN     ENDING  BEGINNING     OFFS     ENDING
                          GAIN       SLOPE     GAIN      OFFSET     SLOPE    OFFSET
                 HR                  HR-1                          HR-1
  HR
03G
 2.500    15.500
18.000     6.000
                           1.01380   0.00000   1.01380   0.00000   0.00000   0.00000
                           0.00000   0.00000   0.00000  -9.99999   0.00000  -9.99999
    END OF PROGRAM DOCUMENTATION
    99999
    YYMMDDHHMM    U    S     03G
        LTMP      TSR       UV
    8310040500    1   7R     0.0008
       58.6810   -0.0046    -0.2439
    8310040504    1   3B     0.0196
       21.6880   -0.0062    -0.3430
                                  HOG

                                  0.0864

                                 -0.0022

                                   111
 NOXG
H02G
DPG
CTMP
 0.1138    0.0237   50.0450   56.8300
-0.0007    0.0001   56.6496  -60.5560

-------
 Segmented Data File
ANSI Tape File Format
                Table 20.  cont. Example Segmented Data File
      8310041704    1   3B    0.5787   -0.0012    0.0513    0'.0507   72.7266  -60.5120
         20.5120    0.1592   11.7073
      8310041708    1   7R    0.5508   -0.0013    0.0578    0.0573   64.7658   83.3620
         81.3040    0.1376   10.2439
      99999
      YYMMDDHHMM    USER SIDE      GENERATED Oil  19-MAR-84
      2.4,4 TRI TOLUE11E   ETHYLEHE  N-BUTAHE  TRAIJS-2-B ISOPE1ITAII H-PE11TA11E ACETYLEliE
      8310040625    1    B
        0.2266    0.4546    0.4435    0.0877   0.0366    0.1339    0.0000    0.0905
      8310040655    1    R
        0.1868    0.2580    0.2248 -999.0000   0.0459    0.1195    0.0634    0.2332
     8310041455    1    R
        0.1439    0.1573    0.0890  -999.0000    0.0000    0.0828
     8310041525    1    B
        0.1821    0.2514    0.1300    0.0688    0.0000    0.0921
     99999
     YYMMDDHHMM    USER SIDE     GENERATED 01! 13-MAR-84
     ETHYLEHE  ETHAI.'E    PROPYLENE
     8310040625    1    B
        0.4522    0.0093    0.0952
     8310040655    1    R
        0.4097    0.0238    0.0958
 0.0340    0.1672
 0.0000    0.0000
     8310041525    1    B
        0.1603    0.0086    0.0000
     8310041555    1    R
        0.2675    0.0260    0.0000
     99999
ANSI Tape File Format
The segmented data files are written one file after  the  other onto 1/2 inch wide
industry standard magnetic tape at a density of 1600 bits per inch.
                                       112

-------
 Final Data Plots made from SegFile	Segmented Data File

    We have adopted the American National Institute of Standards (ANSI) standard
 labeled tape format.  This format is supported  by a wide variety of computers
 including most mini-computers, and IBM. An  advantage of a labeled tape is the
 "directories" of the tape contents can easily be created without  having to list the
 data.

    Individual records of the files are collected together into a "block" of records
 before being written to the tape. They can appear on the tape in one or two forms:
 fixed length, blocked records, or variable length, blocked records.  The default value
 is fixed length blocked records with a record size of 80 bytes and a block size of 2000
 bytes. For variable length  records, the records are between 5 bytes and 84 bytes,
 and the block size is 2048 bytes.

    Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29,  Figure 30,
 Figure 31, and Table 21, Table 22, Table 23, and Table 24 explain the details of the
 ANSI tape format as an aid in reading the tape.  Volumn identifiers on the ANSI
 tape (in the VOLl label field are UNC001 to UNC010 depending upon the release.

    File identifiers on the ANSI tape (in the HDRl label format) are YYMMDD.SEG
 for  the year, month, day of the run.

 Tape  Contents

 Final Data Plots made from SegFile

 Although  all processing of data and documentation  is subjected to many QA pro-
 cedures, the final SegFile is inspected and tested. The file is  listed and inspected
 for  order and completeness of the SegFile components.  This  is also a last chance
 look at  the general documentation and format of the data files.

    The final product is intended to be read by computer.  The  final QA test is
using a computer program to read the SegFile and produce plots of the data to
demonstrate that the file is correct  and to examine the data visually.
                                     113

-------
         Segmented Data File
                             Final Data Plots made from SejFile
                           Beginning-of-Tape
                           Marker (EOT)
                             Volume Label
                             (VOL1)
                          File Header Labels
                          (HDRt  HDR2. HDR3)
                           Tape Mark (TM)
                              File Section
                            Tape Mark (TM)
                          File Trailer Labels
                          (EOF1. EOF2. EOF3.
                          EOV1. EOV2. EOV3)
                              Tape Mark
                              Tape Mark
                             Scratch Tape
                             End of Tape
                             Marker (EOT)
 Marks beginning of writeable area 'on a volume
Identifies the volume
Describes and delimits each file
Separates header labels from file section
Contains user data. Data in volumes interchanged to
non-VAX/VMS systems must be ASCII "a" characters.
Data m volumes interchanged to VAX/VMS systems can
be in binary form.
Separates file section from trailer labels

Describes and delimits files. When a volume is contin-
ued. EOV labels are written instead of EOF labels. EOF3
or EOV3 labels are written only when a HDR3 label is
written.

Indicates the logical end-of-volume. Two conseculive
tape marks are always written after the trailer labels of
the last file on a volume.
Incficates tape that is blank or that has not yet been
overwritten. Scratch tape can exist between the logical
end of volume and the EOT marker.


Marks beginning of the end of the writeable area.
Figure  24.   Basic  Layout of a VAX/VMS ANSI Labeled Volume
                                                           114

-------
       Final Data Plots made from SegFile
                                              Segmented Data File
                            Labels and  Components Supported by VAX/VMS
          Symbol
      Meaning
Symbol
      Meaning
           BOT


           EOF1


           EOF2


           EOF3


           EOT

           EOV1


           EOV2
Beginning-of-tape
marker

First end-of-file
label

Second end-of-file
label

Third end-of-file
label

End-of-tape marker

First end-of-volume
label

Second end-of-volume
label
EOV3


HDR1


HDR2


HDR3


VOL1

TM


TM TM
Third end-of-volume
label

First header  label
Second header label


Third header label


Volume label

Tape mark
Double tape mark
indicates an empty
file section or
the logical
end-of-volume
• * •
BOT
VOL1
HDR1
HDR2
HDR3
TM
N
                                         First File
TM
EOF1
EOF2
EOF3
TM
TM
Scratch ^"
Tape
Figure 25.   Single File/Single Volume Configuration

                                         115

-------
        Segmented Data File
                                                         Final Data Plots made from SegFile
                            Continuation Volume

BOT
VOL1
HDR1
HDR2
HDR3
TM
^•N
                                          Second Section of Second File
TM
EOF1
EOF2
EOF3
TM
HDR1
HDR2
HDR3
*s
TM
                                                  Third File
TM
EOF1
EOF2
EOF3
TM
TM
Scratch
Tape
EOT
ScratcrA
Tape I
                               Multifile/Multivolume Configuration
                 Label Identifier   Volume Identifier
                       1   345   »   10 11 12
                        VOL
                              '/GEORGE-
       Label Number  Accessibility
    Owner Identifier
38        I        505152
                                                           Reserved
                                                           Reserved
                                                                                  37
                                                                                    7980
Figure  26.
                             DIGITAL Standard Version
                     User Accessible Field
                                                             Label Standard Version
                                                                         2K-3SO-8I
                                         VAX/VMS  ANSI  VOL1 Label Format
                                                     116

-------
          Final Data Plots made from SegFile
                                                                                             Segmented Data File
              Label Identifier
                 11345
                                  File Identifier
                                                     File-Set Identifier
                                                                      File-Sequence
                                                                        Number
                                      Creation Date
21 22
2728   31 32 i 35 36  39404142
                                              47
                  HDR
                                                        GEORGE
                                                                   0001
                                                                          0001
                  Label Number
            File-Section
             Number
               Generation
                 Number
                                                                                        \
Generation-Version
     Number
          Expiration Date  Block Count     System Code
           48   r  535455  i    6061
                                                            Reserved
                                                       73 74
                                                                   80
                          000000
                                        DECFILE11A
                 Accessibility
                User accessible field
Figure 28.   HDRl Label Format
                                                        117

-------
         Segmented Data File
                                       Final Data Plots made from SegFile
                           Label      Block   Record
                          Identifier    Length   Length
                           1  13456
                                          10 11
                                                  151617
                                                                                       36 37
                        Label
                       Number
Record
Format
System-Dependent
                                                                                Form Control
                      System-Dependent
                  38
                                      50 5152
                                                                 Reserved
                                                                                           80
                                         00
                                     Buffer Offset
                       User Accessible Field
                       User Accessible Subfield
Figure 29.   HDR2 Label Format
                                                       118

-------
        Final Data Plots made from SegFile
                                                    Segmented Data File
                 REC
               50 Bytes
  REC
50 Bytes
  REC
50 Bytes
  REC
50 Bytes
  REC
50 Bytes
                                               BLOCK
                                              300 Bytes
                                Blocked Fixed-Length Records
              Record Size = 54 Bytes
                       Record Size = 112 Bytes
                                   Variable-Length Records
Figure 30.   Blocked Fixed-Length Records
  REC
50 Bytes
                                                                                ZK-353-81
RCW
54

DATA
50 Bytes

RCW
112

DATA
108 Bytes

Pad
Characters
14 Bytes
BLOCK
180 Bytes
                                                                               ZK-354-81
                                                 119

-------
Segmented Data File
                                         Final Data Plots made from SegFile
    Table 21. VAX/VMS  ANSI VOLl Label
  Character
  Position
          Field
    (length  in bytes)
              Contents
   1-3

     4

   5-10


    11  .
  12 - 37

  38 - 50




    51


  52 - 79

    80
"Label  Identifier  (3)

 Label  Number  (1)

 Volume Identifier  (6)


 Accessibility (1)
Reserved  (26)

Owner  Identifier  (13)
DIGITAL  Standard
Version  1

Reserved  (28)

Label  Scandard
Version  1
Alphabetic characters VOL

Numeric character 1

Volume label consists of ASCII "a"
characters.

Volume  accessibility;    provides
compatibility       with      some
non-VAX/VMS systems.  A space, the
VAX/VMS   default,   indicates  no
restrictions.  To write any  ASCII
"a"  character  in this field, use
/LABEL=VOLUME_ACCESSIBILITY   wi th
the   INITIALIZE   command.    Any
character but  a  space  indicates
the  /OVERRIDE  qualifier  must be
used with the INITIALIZE and MOUNT
commands.

Spaces

Volume  ownership   set   by   the
INITIALIZE/PROTECTION     command.
The contents  of  this  field  are
used for volume protection
Numeric character 1

Spaces


Numeric character 3
                                    120

-------
 Final Data Plots made from SegFile
                                                                 Segmented Data File
     Table 22.  First File Header Label (HDRl) Fields
Character
Position
                     Field
               (length  in bytes)
               Contents
  1-3

    4

  5-21


22  - 27




28  - 31





32  - 35
36 - 39
40 - 41
42 - 47
48 - 53
  54
            Label  Identifier  (3)  •

            Label  Number  (1)

            File Identifier  (17)


            File-Set Identifier  (6)
           File-Section Number  (4)
           File-Sequence Number  (4)
           Generation Number (4)
           Generation-Version (2)
           Number
           Creation Date (6)
           Expiration Date (6)
           Accessibility (1)
55 - 60    Block Count (6)
 Alphabetic characters HDR

 Numeric character 1

 A user-supplied file name and file
 type

 Same as the file-set identifier of
 the   first   file  on  the  first
 volume,    whether    single    or
 multivolume configuration

 Numeric  characters  starting   at
 0001  and  incrementing  by  1 for
 each   additional   volume    with
 respect  to  the  first  volume on
 which the file begins

 File, number within the volume  set
 for  .this   file;    consists  of
 numeric  characters,  starting  at
 0001 that indicate the position of
 this  file  with  respect  to  the
 first file of  the set

 Numeric characters  that  indicate
 the unique generation of a file

 Numeric characters that   indicate
 the  version    of   a   particular
 generation of  a file

 System   stores  the   date  in  the
 Julian    format   (IVYDDD)1;    the
 default is the current date

 User   specified    Julian    date
 (SYYDDD)*  or   default   is  the
 creation  date,   indicating    file
'expires immediately

 File   accessibility;      provides
 compatibility        with      some
 non-DIGITAL systems.     A   space
 (used    by    DIGITAL    systems)
 indicates  no   restrictions.    Any
 character  but  a  space indicates
 the  /OVERRIDE   qualifier  must   be
 used at mount  time for  the user  to
 access  this file.

 Always  000000  for  the  HDRl label
1.  The number sign (I) in the Julian
                                     format represents a space.

                                                 (continued on next page)

-------
Segmented Data File
                                           Final Data Plots made from SegFile
         Table 22.  (Cone.):   First File Reader  Label (HDR1)  Fields
   Character
   Position
          Plaid
    (length in bytes)
                                                        Contents
   61 - 73
 System Code (13)
   74 - 80
 Reserved (7)
 Identifies the  file  system  that
 created   the   file.    DEC,  the
 3-character   constant,   occupies
 positions  61 through 63, followed
 by the name of  the  file  system;
 DECFILE112    indicates    VAX/VMS
 Version  1.6  and   earlier,   and
 DECFILE11A    indicates    VAX/VMS
 Version 2.0 and later.

 Spaces
            Table 23.  Second File Header Label (HDR2) Fields
   Character
   Position
         Field
    (length In bytes)
              Contents
              Label Identifier (3)

              Label Number (1)

              Record Format (1)
    6-10
   11  -  15
     16
Block Length (5)
Record Length (5)
System-Dependent (1)
Alphabetic characters HDR

Numeric character 2

Character definition:

  F  fixed-length
  D  variable-length

The S for spanned record format is
returned  as  an  undefined format
when processed by VAX/VMS 1

Five numeric characters specifying
the  maximum  number of characters
per block

Numeric characters indicating  the
record   length  for  fixed-length
records  or  the  maximum   record
length for variable-length records

In VAX/VMS Version 2.1  and  later
versions,  this  field  contains a
space indicating  the  VAX-11  RMS
attributes are in the HDR3 label

For  VAX/VMS   Version   2.0   and
previous versions, this field does
not contain a space  but  contains
the  first  byte of the VAX-11 RMS
attributes, Indicating the  VAX-11
RMS  attributes  are  in  the HDR2
label
   1.   To  process  undefined  records  properly,  the  user  must  know  what  the
   original   format   of  the  records  was.   Only logical  I/O can  be used  to
   process undefined  record  formats.
                                       122
                                                   (continued on next  page)

-------
 Final Data Plots made from SegFile
                                                     Segmented Data File
      Table 23.   (Cant.):  Second File  Header Label  (HDR2) Fields
 Character
 Position
          Field
    (length  in bytes)
                                                      Contents
 17  -  36
   37
38 -  50
 51 -  52

 53 -  80
 System-Dependent  (20)
 Form  Control  (1)
 System-Dependent  (13)
 Buffer  Offset  (2)

 Reserved  (28)
 Spaces available for   future   use.
 For    VAX/VMS   Version    2.0   and
 earlier,  this field   contains   the
 VAX-11 RMS   attribute   in binary
 format

 Defines   the   carriage   control
 applied  to   the  records  within  a
 file,  as  follows:

    A   First   byte   of   record
        contains  FORTRAN   control
        characters

    M   Record   contains  all  form
        control  information

 Space   'Line  feed/carriage return
        will  be  inserted between
        records  (default)

 Spaces available for   future   use.
 For    VAX/VMS   Version    2.0   and
 earlier,  this field   contains   the
 VAX-11 RMS   attributes  in binary
 format

 The numeric characters 00

 Spaces
            Table 24. Third File Header Label (HDR3) Fields
Character
Position
         Field
   (length in bytes)
              Contents
 1-3

   4

 5-68



69 - 80
Label Identifier (3)

Label Number (1)

VAX-11 RMS Attributes (64)



System-Dependent (12)
Alphabetic characters HDR

Numeric character 3

Files-11  record  attributes  that
override  information in fields of
the HDR2 label

Spaces
                                     123

-------
Summary  of  QA
Summary of Calibration and Data QA

 • 3 NBS traceable standards for NO were used in this project
 • the three standards agreed at 3 concentration levels to within an average of 2.3
   percent of each other
 • 8 HC tanks  and precision liquid chamber injections  were used  as calibration
   sources in this project; two of the HC tanks were NBS  tanks from an RTI multi-
   year comparison study
 • the four principle HC tanks, validated  with the RTI tanks and liquid chamber
   injections, agreed to within an average of 1.6 percent  of each other; and agreed
   to within an average of 3.7 (std dev 1.8 percent) percent to the manufacturers
   specified concentration values
 • HC calibration factors show very little or no trend  in instrument sensitivity
   during the experimental season; the relative standard deviation from the mean
   calibration factors for the principle species used in this project ranged from 5
   to 14 percent with an average of 9.9 percent (std dev 2.2 percent).
 • computer spreadsheets were used to adjust calibration factors (within observed
   "noise" range) to maintain consistency of experimental initial conditions com-
   position produced from fixed composition (pressurized gas tank and liquid mix-
   ture) HC sources used in this project.
 • two pressurized tanks of HC mixtures  were used to produce  the experimental
   initial conditions of the lower molecular weight compounds assuring consistent
   composition and initial  concentration throughout the experiment series

                                  124

-------
 After Run-Experiment Logging                                             Summary of QA

  • one of the principle HC source tanks for the experimental program (UNCMIX,
    a component of the synthetic urban mixture)  is certified to 2 percent
  • a liquid mixture of the higher molecular weight compounds was prepared  and
    used throughout the program permitting one large neat injection to be made
    with 1 percent syringes to establish the experimental initial conditions: assur-
    ing consistent composition and initial concentration throughout the experiment
    series
  • a multi-computer-based data acquisition and  data processing system incorpo-
    rated into a QA program which monitors every step of data generation  and
    processing is utilized to minimize risk of human e.rror and maximize accuracy
    and overall data quality.

List of QA Steps

This section lists all specific QA steps performed.

Before Runs
  • review progress and schedule of experiments from work plan:
     o do experiments to date make sense? are they consistent?
     c do they indicate new experiments or other immediate needs?
     o are data for these experiments good enough or must they be rescheduled?
  • check site status and weather report
  • match highest priority experiment with immediate site capabilities
  • fill out run sheets and instrument checklists
  • use site checkout lists
  • perform instrument  checkout and calibration

After  Run-Experiment Logging
  • PC phone call to site and update calendars:
     o verify general quality of run and conditions
     o reschedule?
  • use "pack-up" inventory checklist at site
  • check calendar for transfering data to floppies at site
                                     125

-------
Summary of QA                                                       During Calibration

  • perform runfolder inventory
     c verify  proper and complete documentation from site
     c verify  complete data from site
  • RUllEllTRY -  log experiment; store folder in office bookcase by date
  • calibration folders - store in office bookcase by date
  • CALEilTRY -  log any calibration performed
  • calendar, run folder/RUilEUTRY, fastplot in notebook check:
     c were.runfolder and computer data transfered to data processing office?
     o successfully ?

During  Calibration
  • planning:
     c what species will be measured: what calibration sources needed?
     c frequency?
     o technique?
     c documentation required?
     c procedure for transfering  calibration data to dpo?
  •  identify standards and perform cross comparison:
     o check standards consistency                            .
  •  CALEllTRY - calibration log and processing status database:
     o during run log-in: check instrument checklist for calibrations performed or
       check for presence of calibration folder:
     o log calibration data in calentry
  •  routinely check CALE.'iTRY:
     o check that all have been given processing instructions
  •  check documentation QA while giving processing (pick) instructions:
     o verify proper and complete documentation from site
  •  inspect calibration data stripchart for appropriate appearance
     o mark peak ids and chromatograms to pick
  •  routinely check CALENTRY:
     o check that all have been digpicked
                                     126

-------
During DVM Data Processing	Summary of QA

  • after DIGPIK inspect picked raw data on computer - compare with stripchart data
  • routinely check CALEliTRY:
      o check that all are run through CALFAC
  • run CALLOK:
      o check report lists and plots to display results for representative compounds
       for early QA to identify any problems
  • check DTR HCCAL database:
      o generate HCCAL report and scan entire report;
      o observe large variations and problems
  • run CALAMA: final check and statistical analysis
  • during determination of calibration factors for data processing: does calibration
    result in meaningful data?

During  DVM Data Processing

  • calendars and CTREPORT inspection:
     o are there data files missing?
     o new files to be unpacked?
  • documentation comparison:
     o run sheet, instrument checklist, remarks file: what were target conditions?
     o any problems  recorded?
  • inspect stripchart:
     o were target conditions achieved  (within calibration uncertainty)?
     o documentation complete? recorder settings, side, time for all pens?             -\
     o does data make sense? are there unexplained problems?  missing or noisy
       data?
  •  FASTDV plot and comparison with stripchart data
  •  AUTOCAL strip  (see cal)
  •  AUTOCAL report:
     o did AUTOCAL data get  processed?
     o any problems or large variation?
  •  run DVMFIX:
                                     127

-------
Summary of QA	During GC Data Processing

     o check that DVMFIX NOX concentration equals concentration of NO plus NOo all
       during experiment
     o check for good zeros
  •  QA PLOCOli plots:
     o can plot file (format good)
     o data plot agrees with original stripchart
     o was data processed  properly? correct times, sides, and general magnitude?
     c were holes  processed properly?
     o dewpoint data higher than temperature?
     o does temperature data check with RDU airport?

During GC Data Processing
  •  PCREPORT from RU!,'E!!TRY:
     o instruments needing pick instructions correct?
  •  documentation comparison:
     c inspect run sheet, instrument checklist, remarks file:
     o what were  target conditions?
     o any problems recorded?
  •  inspect stripchart
     o instrument documentation complete?  recorder settings, instrument atten.
       sides, time?
     o does data make sense? see run sheet:
     o is there  missing or noisy data?
     o is data good? worth processing? bad?
     o identify  calibrations and mark not to be processed as run data
     o identify  compounds, pick start and end time, bad data to be skipped, base-
       lines for peaks
  •  compare raw data stripchart and picked computer file:
     o is raw (picked) data picked correctly?
  •  PLOPic plots (of raw data)
     o does  data look reasonable? no side switching, attenuation switching prob-
       lems?
                                     128

-------
During Documentation Stages	Summary of QA

      c sides id correct?
  •  after CALCOI::
      o perform checks on C- and  K-files and plots
  •  calibration factors and RUHCALUSED:
      c are calibration factors used  consistent with others, or if not are they  ex-
       plainable?
  •  compare K-file with data processing instruction form
      o are the calibration factors used shown in K-file the same as specified in data
       processing instruction form?
      o are names of species correct?
  •  PLOC011 plots:
      o can plot be made (file format correct)?
      o do plots seem reasonable?
  •  HCANAL or LOTUS:
      o if HC  data for mixture, is composition consistent with known composition?
  •  run sheet target:
      c is resulting data consistent with target experimental conditions? or expected
       initial conditions?

During  Documentation Stages
  •  compare  documentation:
     o inspect run sheet,  instrument checklist, remarks file, calendars:
         >  what were target conditions? any inconsistencies?  may  need to consult
           site operators (see run  sheet)
         >  any problems recorded?
         o  special conditions, modifications, or operations performed?
         >  missing  data? species not measured?
         o  how many hours of venting?
         >  drying performed? condensation observed in morning?
  •  inspect stripcharts for notes of injection times or amounts
  •  check climate data, site temperature and light data:
  •  inspect data files and plots:
                                      129

-------
Summary of QA                                                      After SegFile Made

     o how complete is data?
     o explanations available for lost data?
     c reason to communicate this to modelers?
  • K-files check:
     o documentation complete?
     o any species co-eluting requiring documentation editing?
  • compare  original experiment list, general documentation  paper form and run
    ranking formulas, and computer file:
     o is information consistent or explainable?
     o was experiment close to target? if not why not?
     o is paper form with general documentation and formulas reduced and trans-
       fered  correctly to computer file (general doc for segfile)?

After  SegFile Made
  • verify file format and contents:
     o by listing file and performing a visual inspection of the contents
     c and by plotting data from file  and inspecting plots
                                     130

-------
                            References
Smog Chamber Background Reactivity, Parti: The Influence of NOX from Chamber Walls on
Kinetic Computer Models for Air Pollution, submitted for publicntion
Smog Chamber Background Reactivity, Part2: The Influence of Radicals from Chamber Walls
on Kinetic Computer Models for Air Pollution, submitted for publication
Code of Federal  Regulations, Title 40, part 50,  subchapter C, Appendix F, pages 568-572.
                                   131

-------
Site  Checklists
Checklist  For 0900 EDT Site Operator.
As you leave car walk by the red barn and pick up keys.  A note pad for listing
problems detected in checkout should be used.


Outside Check

 • Walk towards smog chamber on service-gas-tank-house-side of lab. Look at lab
   door. Are there any messages?
 • Listen for Air  generator pump running—pull back on metal plates protecting
   pump and inspect:
    o are pumps upright?
    o do they sound "right"?
 • Quick check five service gas tanks (assuming BECKMAN THC  is not opera-
   tional): He, H2  for Carles, 02 for NOX meter, ethylene for 03 meter, and Ar/CH4
   for Varian PAN instrument (generally need minimum of 100 Ibs in tank to get
   through the day). While in the service tank house, check flood lamps mounted
   on wood chamber platform; if they are on, turn them off.
 • Continue walking to BLUE chamber - walk to where manifold goes under cham-
   ber.  Check service tank N2 for AC  driers air circulation valve actuators: sec-
   ondary meter should show between 15 and 20 Ibs and NOT higher or it will
   destroy pneumatic valves; note total Ibs in tank  and replace (later in the morn-
   ing) if below 100 Ibs.

                                  132

-------
Outside Check	Site Checklists

 • Touch heat tapes on BLUE Glass manifold: they should be warm. Inspect for
   condensation. Look along length of manifold including under chamber for large
   breaks or condensation.  Visually check BLUE circulation fan over injection
   manifold in chamber: it should be running and turning smoothly.                ,
 • Walk to opposite side of chamber and observe venting doors (Make mental note
   of doors for comparison with run sheet which will be read shortly.
     o If there is no run then chamber is probably venting;  all  doors should be
       open and exhaust fan should be on.
     o If a "run" is on-going then in most cases all doors will be shut.  Check RED
       exhaust door and make sure it is closed completely and door motor arm is
       positioned exactly like BLUE door motor  arm.
     o If auto driers (AC system) is operational (one of the circulating fans will be
       on) then an intake door will be open.
     c If the  dehumidifiers were used for drying they should  have been  removed
       before injections. If they  have not. make  note on the runsheet  and call PC
       to determine if the experiment can be continued.
 •  Visually inspect for condensation both inside and outside the chamber on both
   sides (RED and BLUE). RuB! hand over a large area of one  of the side (not
   end) panels. Note any large drops of water or puddles on the floor of chamber.
   Note position of center teflon wall separating the chambers.  Does it seem to
   hang more to one side than the other?  Are all the fans running? Finally look
   for general problems such as mud (footprints), bugs, objects  (tools, tape, paper
   towels, etc), in the chamber.  Are the glass return and sample manifold intact?
   If the dehumidifiers were recently removed, were the service panels replaced and
   seated properly?
 •  Continue walking around the chamber to the other side back to the  injection
   house and complete inspection of glass (RED) manifold. Feel heat tapes and       x
   inspect for condensation, breaks along length including under chamber. Is the
   RED thermistor laying on the ground?
 •  Go to BLUE side door in injection tank house.  Observe if door is open or closed.
   It should be closed. Open door to inspect tanks, manifold, and solenoid plugs.
   Do not touch anything except the RED and BLUE injection  lines! Do  not turn
   off any tanks or touch any regulators. You will be coming  out  again shortly.
   Check that, the  RED  and BLUE 1/4  inch Teflon injection lines  are connected
   to  the glass manifold. Check to see that they  are tightly connected. Check to
   see if ALL computer controlled AC power boxes for solenoids have  a plug in it
   (not including the light brown plastic box used for the WOOD project). They

                                    133

-------
Site Checklist? _ Outside Check

   should also each have a light bulb.
 • Close door and go to the other side: open door. Do any of the wires leading over
   the door catch?  Quickly  inspect this side of the injection shed for problems.
   Don't  touch anything yet. Don't close any tanks yet. Is there a small wrench
   for turning off the NO; tank later? Close the door.
 • Go inside the lab on the same side of building.  Look up as you enter and notice
   if the outside floods are on. If they are turn them off with the switch just inside
   the door as you enter.
 • Turn to the right as you enter and head for the NOX and 03 strip chart. Look and
   observe if NO and N02 (or  possibly 03) have been injected.  Notice approximate
   concentrations. Are conditions matched (generally)?
 • Head for computer console (decwriter).  Is the computer and DATCOL up? Are
   there error messages? Is there a print profile? If not, can  you get the time of
   day from the computer? (Type TIME and return.)
   Are  ALL switches  in AUTO?  Or is there a note saying  some should be  in
   MANUAL? Is the 4 minute clock in AUTO?
 • Turn around and look for a filled out run sheet for today's run. What was the run
   today? Look  at the command file printout if there is one.  What was injected?
   What tanks were used? Did anything have to be injected or  otherwise done
   manually?  Was someone here this morning? See if morning operators name is
   filled in.  Are there any notes  on the desk?
 • Does NOX and Oa stripchart make sense  compared to information on run sheet?
   Walk towards Carle GCs.  Were injections made?  Make  quick inspection  of
   chromatograms. Did anything show?
 • Go outside  to the  injection shed (Carle side). Consider the run sheet, command
   file, NOX and 03 stripcharts, and GC stripcharts. If injections were scheduled but
   failed, inspect tanks to see  if they were turned on and connected to appropriate
   solenoids. Were they plugged in?  R/B injection  lines connected to the glass
   return manifold?  If  resulting injections  seemed low, check flow rate.  (Hastings
   mass fiowmeter reads 8.7 for 10.0 1pm; check with soap bubble  flowmeter with
   Once you are finished with the injection shed, turn off all tanks with the main
   valve. Do not turn off tanks with the regulator. Use a small wrench to close
   the NOo tank valve.
   Enter the  lab again on the NOX and Og side.  Go to the DVM and HP clock.
   Is the DVM scanning?  If not,  issue immediate command get time from the
   computer by typing TIME and return. Does the HP clock match? (If not set it

                                    134

-------
 TSR and UV zero check	Site Checklists

     later).  Check 4 minute side light. Go over quickly to NOX and Os instruments.
   •  Check  NOX and 03  pressure (20) and vacuum pump pressure (25.5).  Check
     Os ethylene  (15) and rotoball setting. Mark stripchart with current side and
     time. Does it look like the 4 minute R/B switching valve is operating? Perfect
     matched injections are almost impossible. If it looks like a perfect match, check
     the valve and  the computer 4 minute switch again.  Check that stripchart is
     marked properly for run date, instrument, stripchart settings, attenuation, etc.
   •  Is the HCHO instrument running? Should it be? Is the sample pump on? Is the
     rotoball at 4?  Is the solutions pump on? Is it drawing solution? Are bubbles
     running through mixing/reacting chamber (below solutions pump). Check de-
     gasser chamber (on right side).  It should not be  filling up with liquid (a little
     bit at the bottom is ok). Check each sample line.  Look at the HCHO stripchart.
     Does the  signal trace look "right"? Check strip chart zero by pushing down
     black zero button: it should not be negative. If it is adjust it to about 10%  of
     the chartscale. Adjust HCHO zero if necessary and make note on stripchart.
  •  Check waste container.  Is it too full? Check that stripchart is marked properly
     for run date, instrument, stripchart settings, attenuation  etc. Check dewpoint
     instrument stripchart.  Check dewpoint stripchart zero. Mark it as zero check,
     and mark time and side. Is DP pump on? Is signal trace reasonable?  IF there
     is  a difference in dewpoint between the two sides the signal should be  a square
    wave. If it shows a slow response, clean the detector and rebalance after letting
    it  operate for a couple of minutes.  If it looks 0.  K. , check  balance.  Walk to
    other side and  move switch  to test, wait for it to stabilize and adjust balance
    knob. Turn knob back to operate. Check  that stripchart is marked properly for
    run date, instrument, stripchart settings, attenuation, etc.


TSR and UV zero check
Mark time.

    Check manifold on this side (RED). Start at the  NOX and O3 instrument end.
(While you  are at this end check zero air supply used for injections and be sure it  is
off.) Look for open fittings; missing caps or dangling sample lines. Feel heat tape;
it  should  be warm. Look for condensation. Check zero air supply near stripchart
for DP and HCHO. Move towards the other side of the bench. Check manifold fans
to see that they are operating as you go around to the other side. Perform manifold
check on this side walking  to the end where  the CARLEs are.  Come back to the
manifold fans.  Check zero air supply. 100 Ibs in and 60 Ibs out.  Check Tektronix
terminal and clear screen by hitting page.

                                     135

-------
 Site Checklists	GC Check

 GC  Check
 Check switch 6 on AUTO/MANUAL switch panel: is switch 6 on (Carles valve min-
 ders  power)? Walk down towards Carles. On the way check the Varian stripchart.
 Are there injection peaks?  H2O peaks?  Tracer peaks?  Is  the sample pump on?
 Check that stripchart is marked properly for run date, instrument, stripchart set-
 tings, attenuation, etc. Check bucking current.

    Are  Carle valve minders in auto/repeat and sequencing? Do chromatograms
 look  right? Air peaks? Look at example notebook if necessary. Are the GC sync?
 Do the chromatograms look right considering what was injected and the nature of
 today's run? Look at background chromatograms. Contamination? Ethylene? Was
 an autocal scheduled  and were the chromatograms made? Peaks on scale for initial
 injection? Are they still  on?  Does  it  look  like the baseline might drift off scale?
 Adjust with GC zero, not stripchart zero.  Look at the service gas regulators. Do
 the readings match the numbers  on the wall? Is the sample pump on? Is the 30
 min RED/BLUE clock on and working? Is  the flame lit?  Reasonable attenuation
 getting? Are Sigma integrator reports being made?

    Considering what was injected for the run:
  o Target match condition? Are  they matched?
  o If difference, are the sides switching? Are both sets of chromatograms on scale
    (no peaks off scale)?
  o If different amounts of same composition are ratios of peaks correct?
  o Mark chromatograms with time and side.  Try  to mark  time  of sample injec-
    tions.
  o Try to mark when 30 min clock changes  sides and whether B to R or R to B.
  o Check that stripcharts are marked properly for run date, instrument, stripchart
    settings, attenuation, etc.

    Check lab AC. Is  it set at  72?  Is fan in ON position (not AUTO)? When was
the filter changed last? (It should be changed every  three weeks.)

    Enter PRINT PRO command if one is not printing already. Don't wait until
later.
                                     136

-------
 Formaldehyde
 Checklist for Leaving the Site  Set Up for  a Run

 Computer
 Is the system up and ready?
  • There has to be a day created on the disk for today in order for the system to
    run tomorrow.
  • Create a day on the disk with tomorrow's date using the program "NEWDAY".
    If necessary, create one for today also.
  • Create a command file for tomorrow using the Program "CMANDR". Once you
    are satisfied, store the command file by hitting "W. Now run the command file
    thru "DATRAN" and  merge ("EMERGE") it into  tomorrow's DATCOL-OPS
    file.
                                                      4
  • Run DATCOL, wait 1 minute and you should see a "*". which will tell you that
    DATCOL is up; hit  and you should get another *. If there was
    a run today, key in "BACK". This will print out the commands for the current
    day and tell you if the restart commands to vent the chambers (Set 1,2,3,4)  at
    1900 are still in there; otherwise the chambers won't vent!!!  If there was not
    a run today, issue an immediate command IMM  and Set 1,2,3,4;  to
    open the chamber doors.
  • Are all front panel switches  in auto not  manual position; DVM on auto, rear,
    remote.
  • Is the 4 minute clock switched to auto and not locked on one side?
  • Check decwriter paper. Need at least 10  pages.
Formaldehyde

 • Run on water for 15 minutes by placing all three reagent tubes in water.
 • Loosen all four tubes on parastatic pump.
 • Shut off two switches on rear of instrument starting from the right as you face
   the switches, leave left most switch on.  This supplies charging "juice"  to the
   instrument batteries.
 • Make sure there is enough TCM stock reagent prepared (without sulfite) so that
   the person making the run in the morning just has to add 0.25 g sulfite  to 200
   ml TMC.
                                    137

-------
                                                           NOx and O3 Instruments
 NOx and O3 Instruments
  • Are the 3-way valves for the O3 and NOx meters plugged into the 4-minute
    power strip?
  • Is there a good connection  between the  NOx and O3 meter  intakes and the
    round Teflon filters?
  • Are the front panel settings  of the O3 and NOx meters correct?
     c The NOx meter should be in ambient mode. NO, NO2.  and NOx should be
       set to "1 ppm full scale.  The oxygen pressure should be set at 20, and the
       valve knob should be set to NO-NO2-NOx. Vacuum gauge on NOx pump
       should read 26" Hg.
     o The O3 meter should be set to 2 ppm  and full scale. The ethylene pressure
       should be on  15. The time constant on the instrument  should be set to 10
       sec. The top of the + ethylene  flow rotoball should be set at 30, and the
       mode selector should be  set on ambient.
     o Strip Chart Recorder:
        t>  Pens on; on "record"
        [>  Chart "on"
        >  Voltages 1, 1, 1/2 for NO, NO2, O3
        i>  Chart speed 3 cm/hr and "onr
     c Autocal Tank:
        >  Tank should be turned on, 5 psi showing on low side of regulator, and
           on/off needle valve on regulator turned on.
        i>  Flow should have been previously adjusted so that meter  will have
           enough  sample when  span tank is turned on in the morning (rotoball is
           positive showing excess flow when  meter is sampling from this appara-
           tus.

TSR and UV Chart
 • Have the TSR and UV  pens  been properly zeroed?

Charts
 • Check that sufficient paper remains Red line will appear on paper when the roll
   is ending.
                                   138

-------
 Carle II nnd Carle III
  • Record the date, type of injection, attenuation, chart speed and scale of each
    instrument on the charts.
  • Check that pens are full of ink and pens down.
  • Check that power to charts works from AC switch no. 16 on  blue computer
    AC switch panel. Activate no. 16 to manual and observe deflection on charts,
    return these to "auto" position.


Carle I

     •  Is the G.C. up and running?
         o  Check pressure gauges on the wall behind the instrument H2 =  28.5 psi.
           He = 63 psi, air = 13.5 psi.
         o  Is the flame  lit?  (check by placing mirror or chromed surface  over the
           FID. It should fog  H2 pressure to 35 psi. hit ignite button; you should
           see positive  response on strip chart (strip chart has  to  be on) after 5
           minutes. Return H2 to 28.5 psi. Watch the gauge go down and make
           sure it stabilizes at 28.5 psi.
         o  Has the proper attenuation been dialed  into Carle I?  (X4 for  mix. X8
          for propylene and ethylene. everything else determined by the nature of
          the cal and the injection.
     • Has the valve minder for Carle I been put on light no. 5 and a drum setting
       of 29?
         c To check this, activate switch no. 6 on the computer.
         o If the valveminder  is not on light 5, switch the valveminder knob from
           "standby" to "sequence" at 10 second intervals  until  light no.  5 is on.
          While the valve minder know is in standby, rotate the timing drum un-til
          the face setting is 29. CHECK FLAME AGAIN.  The valve minder knob
          should still be on standby at this time. Continue with Carle II and Carle
          III.


Carle II and Carle III

     •  Is GC up and running?
         o Check pressure gauges on wall.  Carle II, He =  38 psi. H2 = 27.5 psi,
          Air = 12.5 psi. Carle III, He = 64 psi, H2 = 28 psi, Air = 13.5  psi.
         o Check to see  if flame is lit (place mirror over the FID).

                                     139

-------
                                                               Sigma 10 Integrator
         c  Check for proper attenuation (normally X4).
     • Have the valve minders for  Carle II and III been placed on light 8 and a
       drum setting of 59?
         o  To check this activate switch 6 on the computer.
         o  If light 8 is not on. switch valve minder knob from  "repeat" to "manual"
           and activate "event advance'7, toggle switch until light 8 is on.  WAIT
           10 SEC. BETWEEN EACH ADVANCE SWITCH. Make sure that you
           see light no.  1 before you advance to light no.  8. Set the drum at 59.
           CHECK FLAME  AGAIN. Return to the computer blue switch panel
           and switch no. 6 to "auto". NOW GO BACK TO THE CARLES AND
           SWITCH CARLE I FROM STANDBY TO AUTO.
     . ALSO TURN  THE VALVE MINDER ON CARLE  II AND  III TO  RE-
       PEAT. DO NOT LEAVE IN SINGLE OR OFF!!!


Sigma 10  Integrator

 • If we want to collect integrator data on the  Sigma 10.  Carle I. II, III must be
   linked to the Sigma by a "set up" procedure.

         SIGMA       (your key strokes)
        Response      SET UP KEY EIJTER
    EXAMPLE:
    (1.2,4)   1 Enter
    method?
    1(2,4)      Enter
    mode       Enter
    Sample  ID  1 Enter
              Carle (no.  ) enter
    Ptr       1 Enter
   Note that:  Carle 1 - Inst 1, Method 1., Carle II = Inst 2. Method 2, Carle III
   = Inst 4, Method 4.  PE 900 = Inst 3, Method 41.
 • If PE 900 is to be used, is it set up? Instrument 3 Method 4.
 • Is there enough paper in the printer?
     c There are 199 sheets total, so if the printer has less than 20 sheets left.
       change  the paper.
     o Void plotter  unless you want to use it. it will otherwise use a lot  of paper.

                                    140

-------
 Air Generators
   •  Clear out old SIGMA files by using the following keystroke sequence:  Shift*F6
     enter 200 enter or Shift *F Enter Enter. Hit the "Stat" key after this is  done.
     You should have at least 1000 free blocks.
   •  Turn the Tekronics terminal  off (this is the second channel to the Sigma).
 Dew Point Meter
  • Has the dew point meter been balanced and returned to operate? Set to "Test"
    and balance indicator to center of scale.
  • Has the detector been cleaned within a week? Using isopropyl alcohol and Q-tip
    on mirror in detector in rear  of instrument.
  • Is the dew point meter chart  been properly zeroed?
  • Is the  3-way valve for the dew point meter plugged into the 4-minute power
    strip?
 ATC and Varian
 Check standing current in morning and afternoon.
  • Set the ATC  and Varian injection clock right before injection.
  • Is the ATC and Varian injection clock plugged into charts?
  • Check to see  if the ATC and Varian sampling 3-way valve is plugged into the
    30 min red/blue clock and if the Teflon lines from the 3-way valve are attached
    to the red and blue sampling manifolds.
  • Is the correct attenuation dialed into the Varian (atten x*)?
  • Is the correct attenuation dialed into the ATC (X26)?
CO Meter
  •  Is the sample line on the gas chambers?
  •  Turn off service air tank Back on in morning.
  •  Are front panel switches in correct  position?

Air Generators
Must have 100 psi on input gauge, and 60 psi on output gauge.
                                     141

-------
                                                                      General
General

(CHECK INJECTION TANK FLOW WITH BUBBLE METER EACH DAY AND
SET TO 10 L/M1N).
 •  Check tanks for tomorrow's injections.
     o Go to the injection house and remove the line for the tank to be used from
       its control outlet.
     o Plug this line into the white power cord and read the tank flow on the mass
       flow meter.
     c If the flow rate is low or high adjust it with the low side pressure regulator
       valve.  All tanks should have a flow of 10 1/min including NO.
     c Plug the tank line back into the control outlet for that tank.
 • Check the following tanks in the service gas house:  O3 ethylene NOx oxygen.
   Carle I He and H2, Carle II He and H2, CO3 and H2; Varian: ArCH4.
 • Liquid Injections
     c Determine the number of ul  that have to be injected from the computer
       listed table on the wall in back of the formaldehyde instrument. This table
       tells you the number of u] that you have to inject into the chamber to give
       1  ppmC. This has been computed for three different temperatures 60 F,
       70 F. and 80 F.  Selected the  number ul to be injected from the list which
       will approximate tomorrow's morning temperature. Clearly write down the
       number of ul to be injected on the run sheet.
     o  Locate the liquids to be injected and set them out in the work space of the
       hood. Find the liquid injection apparatus and clean the needed syringes so
       that the operation will not have to spend  time doing this tomorrow. Place
       these on a clean paper towel by the injection ports of the return manifold.
     o  If you have to calculate the no.  of  ul to  get 1 ppmC of a liquid into the
       chamber use the following formula:
 •  Has the run sheet been filled out?
 •  Have the charts  been marked with day.  month, year, instrument, chart speed.
   full scale chart voltage, and attenuation  and range where applicable?
 •  Check for loose or missing caps on the manifolds.
 •  Check the chamber recirculation fans.
 •  Check all the red blue 3 way valves connected to the GCs to make sure they are
   plugged into the 30-min power outlet.
 •  Final check

                                   142

-------
General
     o Check all auto/manual switches-computer and cal box-to make sure they
       are in auto position.
     o Check DVM to make sure it is in remote not local mode.  Especially 4 minute
       clock; it should be in auto.
     o Check the red exhaust door on the chambers to see if it is the same as the
       blue exhaust door.
     o Make  sure all charts are on and plugged into "charts" strip.
     o Check Carle 1 to be sure its valveminder is in auto and that Carle II and III
       valve minders are in repeat.
     o Check high Cone. NOx calibration tank to be sure it  is  off.
     o Check the Cal tank for the Carles to see that it is open.
     o Check injection tanks to seer that they are open  and that  the flow  is correct.
     o Check the reagent levels  for the HCHO instrument;  also check the waste
       flask to see that it can't overfill.
     o Shut off water  to toilet.
     c Make  sure the chambers  doors  will open at 19:00. and vent, by a Datcol
       Restart command, or by an immediate command to Datcol to set 1.2.3.4:
       before you leave the site.
                                    143

-------
                                                           Early Morning Checkout
Early Morning  Checkout

 • Did the command file run? If not. see HELP LIST on end of rack over the DEC
   writer. Booting instructions are at other end of the rack over the disc.
 • Is the DVM scanning? If not,  issue an IMM to stop GASDVM; then, another
   IMM to start  GASDVM.  If DVM still  is not scanning, switch from "Remote"
   to "Local". "Step" through to end of series (channel 60). return to "Remote"
   and issue an IMM to start GASDVM.
 • Did the chambers vent?  Check this by reading the 03  strip chart-background
   should be almost completely free  of O3. If the chambers have not vented and
   enough  time remains before sunrise-2.5 hours-manually vent chambers until
   just before sunrise. Close the chambers and then  make all injections manually.
 • Make CEA formaldehyde solutions.
 • Read the run sheet and DEC printout of the day's command file to determine
   target concentrations.
 • Find manual injection species and amounts and  all apparatus .necessary to make
   injections.
 • Check chamber condensation and all vent doors (especially  red exhaust  door).
 • Go to the injection house at the time of the injections to see if the lights confirm
   that the proper tanks are being injected. A list  of tank  concentrations, species,
   and switch numbers is on the end of the rack above the DEC writer.
 • Check all charts.
    o  Charts on at charts.
    o  Pens down and full.
    o  Charts fully labeled.
    o Sufficient chart paper for the day.                                           x
 •  Check manifolds to see that all caps are on and that each instrument is connected
   to both the red and blue manifolds.
 •  See that 4 minute computer clock is in AUTO and not locked manually to either
   side.
 •  Check status of SIGMA 10.
    o "/STAT"; all Carles and PE 900 (if used) should be set  up with the corre-
      sponding methods.
        t> Carle 1  Method 1
        > Carle 2  Method 2
                                   144

-------
Early Morning Checkout
        i>  Carle 3   Method 4
        >  PE 900   Method 31
     o Check that sufficient paper remains for the day (about 75 sheets).
 • Start CEA formaldehyde.
     o On  sample line
     o Advanced past auto 0
     o All  pump lines connected and useable and pulling solutions.
     o Sample pump on.
 • Turn on air tank to the Beckman 6800 and turn on the instrument if it is to be
   used.
 • Perform manual injections before 05:00 at the latest.
 • Check standing current on the Varian: it should read about 600-9 x64 amps.
 • Check injection concentrations and relate to target amounts by measuring peak
   heights  and using the provided concentration conversion factors. See attached
   chromatograms.
     o  NOX: additive value from strip chart or from DVM printout.
     o  Carle I,II.Ill: see attached chromatograms and respective  values.
     c  Adjust concentrations of NOX or HC species, if necessary, to achieve desired
       amounts,
 •  Change attenuation  on the  Carles  after the auto cal;  generally,  attenuation
   should be x2  or x4.
 .  Issue a PRINT  PROFILE of the DVM data.
 •  After successful injection turn off injection tanks and successful auto HC span
   turn off HCspan tank.
                                    145

-------
                                                              Run Pack-Up Checklist
 Run Pack-Up Checklist
 Before a run folder is taken to school, check to see if the following items are in
 it:
 1. Carle I and II stripchart
 2. Carle III stripchart
 3. Sigma 10 printout
 4. Varian stripchart
 5. Dewpoint stripchart
 6. Beckman stripchart
 7. PE900 stripchart
 8. DNPH integrator Printout and stripchart
 9. TSR and UV stripchart
10. Formaldehyde stripchart
11. NOX and 03 stripchart
12. Run Sheet
13. Instrument Checklist
14. Profile Printout
15.  Copy of command file used
16. Copy of BAKTRN of the day's OPS file

    If one of these is missing note reason on the instrument checklist.
                                     146

-------
HC Cal Sources Certification
             147

-------
                              APPENDIX B.

             COMPARISON OF HYDROCARBON CALIBRATION SOURCES
RTI TANKS
ave USED

regression estimate
                 Last Update:

PRIMARY STANDARD #1  AND #2

         Species manuf.
                  cone.
       propylene 14.820

       27-Apr-78 14.580
       22-May-78 14.820
       23-May-78 14.340
       08-Aug-78 14.940
       14-May-80 14.790
       03-Oan-84 14.400
                 14.645

       02-Aug-84 14.449
                                                     13-Oun-85
  Specles manuf.
           cone.
ethylene  20.000

27-Apr-78 19.400
25-May-78 19.200
10-Aug-78 19.800
0G-May-80 16.800
1'J-Oul-Gl 20.000
29-May-84 19.000
          19.480

02-Aug-84 19.304
Calibration of Carles  1,2,3 with RTI tanks » SIGMA  10  Integrator ppmC/area

Assumption - these species are representative of average carbon response efficiency.
Did use May 5 1980 ethylene value.
Choice between regression value or average analysis  value:  used average.
                              UNC  use:   14.645
                               UHC use:   19.480
Carle  I
Ave
al lave
8/2/84 Sigma 10
            44.6
            44.5
            44.5
            44.8
ppmc/area
0.328
0.329
0.329
0.327
slgma 10 ppmc/area
5G.3 0.346
57.1 0.341
57.2 0.341

                                        0.328
                                                                0.343
                                   0.335
                       8/1/84  slgma  10  ppmc/area
                                  44.8  0.327
                                        0.327

-------
Carle II
8/2/84 sigma 10  ppmc/area
            19.1  0.767
              19  0.771
            19.2  0.763
            19.6
            19.2
            19. 1
            18.9
                                        0.747
                                        0.763
                                        0.767
                                        0.775
slgma 10
     24
     24.4
     24.8
     24.8
     24.5
  ppmc/area
7  0.789
   0.798
   0.785
   0.785
   0.795
Ave
al lave
                                        0.765
ave
                      8/1/84 slgma 10  ppmc/area
                                  17.4  0.842
                                  19.2  0.763
                                  19.3  0.759
                                  20.5  0.714
                                  19.3  0.759

                                        0.767
                                   0.778

                               sigma 10
                                   25.31
                                    24.7
                                    24.7
                                    25.2
                                                         0.774
                                                                0.791
          ppmc/area
           0.770
           0.789
           0.789
           0.773
                                                                0.780
Carle III
8/2/84 slgma 10  ppmc/area
            4.15  3.529
            4.15  3.529
            4.14  3.537
Ave
                                     17
                                     14
                  3.512
                  3.537

                  3.529
                      8/1/84 slgma  10  ppmc/area
                                  3.95  3.708
                                  4.03  3.634
                                  4.44  3.298
Ave
                  3.547

-------
Atrco Tank - Determination of Concentration based
Date
Carle I
8/2 - RTI
ave
8/1 - RTI
ave
toluene ethylene
area cone. area
ave ppmC/are 0.335
28.90
29.90
29.40 9.86
ave ppmC/are 0.327
31.20
31 .50
31.35 10.25
on RTI tank - analyzed same day
trans-2-butene
area
21.30
21 .50
21 .40 7.18

22.70
22.70
22.70 7.42
1 sopentane
area
28.00
28.00
28.00 9.39

29.70
29.80
29.75 9.73
propane
area
. 2 2 . 00
22.00
22.00

23.60
23.60
23.60
propylene
area
24.80
24.90
7.38 24.85 8

26.20
25.90
7.71 2G.05 8

.34

.52
tn
O
     Carle  II
     8/2 -  RTI ave ppmC/are 0.778
8/1  - RTI  ave ppmC/are 0.774

                       16.20
                       14.50
                       14.20
                       14.70
ave                    14.90     11.53

Carle III
8/2 - RTI  ave ppmC/are 3.529

                        3.22
                        3.26
ave                     3.24'     11.43

8/1 - RTI  ave ppmC/are 3.547

                        3.30
                        3.23
                        3.24
ave                     3.26     11.55
                                               4.75
                                               5.08
                                               4.03
                                               4.75
                                               4.SI
                                               4.87
                                                3.69
3.77
                                                                1.94           2.52
                                                                1.96           2.55
                                                                1.95    6.88    2.54
                                                                1 .84
                                                                1 .90
                                                                1 .86
                                                                1 .87
                        2.42
                        2.47
                        2.44
                 6.62   2.44
                               8.95
                                                                                              9.74    7.57   10.80    8.40
10.60
10.30
 9.66
 9.64
10.05
                                                     11 ,
                                                     1 1 .
                                                     IS.
                                                     I 1
60
40
90
00
                                                                                                     7.78   11 .23
                                                             8.68
                                                                                     8.67
      THE  AVERAGE  -Alrco  tank           10.92           3.73
      THE  STD.  DEV                      0.72           0.04
      (these values  are listed below -  Alrco  Tank  Final)
                                                                 7.03
                                                                 0.30
                               9.18
                               0.41
        7.61
        0. 15
     8.48
     0. 13

-------
CALIBRATION OF CARLES WITH AIRCO FOR ANALYSIS OF  LOW AND HIGH MW HIGH CONC TANK - 7/18
"> <" Indicates values not used In calculations
Date
toluene
  area
       ethylene      trans-2-butene
  cone.   area  cone.      area  cone.
  10.92          3.73             7.03
             Isopentane    propane         propylene
               area   cone.      area   cone.   area  cone.
                      9.18             7.61          8.48
Carle I
7/18- Cal with Alrco standard


ave
Cal today
ave Carle
Carle II
7/18- Cal




ave
Cal today



ppmC/area
I ppmC/area

with Alrco
>




ppmC/area
26.7
32.1
29.40
0.372  23.7  15. 1 <
11.5
10.5
11.8
11 .27
0.676
26.6
26.5
2G.55
0.320



11.8
11.9
11.6
11 .9
11 .80
0.719
ave Carle II ppmC/area 0.693
Carle  III
7/18-  Cal with Alrco standard

                    >    4.17 <  NOT USED
                         3.31
                         3.26
                         3.25
                         3.27
 ave
 Cal  today ppmC/area
   3.27
  3.337  <
NOT USED
 1 .92
 1 .82
 1 .82
 1 .82
 1 .97
 1 .87
3.757
 3.94
 2.37
 2.41
 2.38
 2.40
 2.39
3.842
 4 . 26 propane &
 4 . 27 propy1ene
 4.27
 4.27
 4.39
4.292
3 .750

-------
      ave Carle  III ppmC/are  3.783

      Date                  toluene
                              area
      Carle  I
      8/14-  Cal  with Alrco  standard
      ethylene      trans-2-butene   tsopentane    propane         propylene
cone.    area  cone.       area  cone.   area  cone.      area  cone.   area  cone.
10.92           3.73              7.03          9.18             7.61          8.48

ave
Cal today ppmC/area
26.3
27.1
26.70
0.409 < NOT USED
20.6 27.2
20.9 27.4
20.75 27.30
0.339 0.336
21 .5
21 .3
21.40
0.356
24.5
23.8
24.15
0.351
      ave Carle I  ppmC/area  0.345
      Carle II
      8/14- Cal  with Alrco standard
in
to


ave
Cal today ppmC/area
13.5
13.9
13.70
0.797
4.7
4.62
4.66
0.800
9.58
9.31
9.45
0.806
10.5
10.3
10.40
0.816
      ave Carle II ppmC/area 0.805
      Carle III
      8/14- Cal with Alrco standard


ave
Cal today ppmC/area
3.29
3.26
3.28
3.336
1 .95
1 .97
1 .96
3.584
2.57
2.57
2.57
3.573
4.65
4.63
4.64
3.469
propane &



      ave Carle III ppmC/are 3.490

-------
Liquid Injection Cals same day!
SIGMA 10 AREA CALS
Date
tol uene
area




m-xylene o-xylene
cone. area cone.
2.99
2.98
area

pentane 2-me-pentane d litr 1 -me-pentane
cone. area
2.66

cone. area
2.99
cone. area cone.
3.16 5.98
Carle I
8/14



ave
Cal
ave

- Cal with LIQUID




today ppmC/area
Carle I ppmC/area

standards
8
8
8
8
0.
SS.

.06
.17
.06
. 10
369
367

8
7
7
7
0.
aro 0.

.39
.82
.37
.86
379
378 par
P/A=
7.46
6.8
6.4
6.89
#.387
0.356
0.94
8
8
8
8
0.
.82
.52
.51
.62
347
NOTE ARO CALS


8.47
8.3
9.14
8.64
0.366
HIGHER THAN PAR

17
17.1
16.6
16.90
0.354


Carle II
8/14- Cal with LIQUID standards


ave
Cal today ppmC/area
4.42
4.22
4.32
0.692
4.33
4.6
4.47
0.667
3.72
4.06
3.89
0.685
ave Carle II ppmC/area 0.681
Carle III
8/14- Cal with LIQUID standards



ave
Cal today ppmC/area
0.989
0.983

0.986
3.030
0.917
0.931
0.917 "
0.922
3.230
0.797
0.800
0.784
0.794
3.355
0.812
0.787
0.787
0.795
3.757
ave Carle III ppmC/are 3.343 NOTE DIFFERENCE BETWEEN AROS AND ALKANE
std dev               0.2654

-------
PEAK HEIGHT CALS
Date
                  toluene
                   ht  In
                m-xylene       o-xylene          pentane       2-me-pentane
          cone.   ht  In  cone.      ht  tn   cone.   ht  In   cone.     ht in  cone.
           2.99           2.98              2.66           2.99             3.16
Carle I
8/14- Cal  with LIQUID standards



ave
Cal today ppmC/ln
6.8
7
6.86
6.89
0.434
3.4
3.4
3.3
3.37
0.884
2.4
2.4
2.4
2.40
1 .110
8.4
8.4
8.36
8.39
0.356
6
5.8
5.8
5.87
0.539
Carle II
8/14- Cal with LIQUID standards
ave
Cal today ppmC/ln
                     6.1
                     6.1
                    S. 10
                   0.490
                   3.2
                   3.2
                  3.20
                 0.930
                   2.3
                   2.3
                  2.30
                 1 .158
Carle III
8/14- Cal
      with LIQUID standards
ave
Cal
today ppmC/f n
 13.2
 13.2
13.20
0.226
  7.3
  7.3
 7.30
0.40B
  5.2
  5.2
 5.20
0.512

-------
Ul
en
Low MW
Date
Carle
7/18-
ave
Date
Car le
7/18-
ave
Date
Carle
7/18-
ave
Date
Carle
7/18-
ave
Date
Carle
7/18-
ave
HI Cone Tank - Determination of
octane
area cone.
I
AIRCO ave ppmC/a 0.313
288.60
300.10
294.35 92.08
pentane
area cone.
I
AIRCO ave ppmC/a 0.313
63.80
63.80
63.80 19.96
octane
area cone.
II
AIRCO ave ppmC/a 0.693
129.30
129.30 89.56
octane
area cone.
Ill
AIRCO ave ppmC/a 3.783
28.20
28.30
28.25 106.87
1 sopentane
area cone.
Ill
AIRCO ave ppmC/a 3.783
4.88
5.98
5.43 20.54
Concentration
based on Alrco Cals - analyzed
ethy lene/ethanbutanes-butenes
area area
35.50
38.10
36.80 11.51
274.90
277. 10
276.00 86.34
tsopentane
area
56.50
56.80
56.65 17.72
same day
propane propylene
area area
66.00 66.90
67.00 67.30
66.50 20.80 67.10 20.99
2-me-l , 3-butad lene
area
132.40
133.90
133.15 41.65
ethyl ene
area
13.40
13.30
13.35 9.25
1-butene
area
4.91
5.10
5.01 18.93
pentane
area
7.12
8.55
7.84 29.64

ethane
area
13.30
14.40
13.85 9.59
n-butane
area
2.71
2.83
2.77 10.48


propane/al lenepropy lene
area area
47.60
47.10
47.35 32.80
c 1s-2-butene
area
4.98
5.09
5.04 19.05
27.20
27.20 18.84
trans-butene 1 , 3-butad lene
area area
- 5.12 4.51
5.21 4.63
5.17 19.54 4.57 17.29
propane/proplene/allene
area
14.80
16.50
15.65 59.20



-------
Low MW HI CONC Tank as source for cal for determination of LOMW
Date
7/13

Carle I


ave
Cal Today ppmC/area
Date
7/13

Carle I


ave
Cal Today ppmC/area
Carle I ave ppmC/area
Date
7/13
»_*
Cn
°> Carle II


ave
Cal Today ppmC/area
octane
area cone.
96.17

319.70
299.70
309.70
0.31 
-------
Low MW
Date
Carle
7/13-
ave
Date
Carle
7/13-
ave
Date
Carle
7/13-

ave
Date
Carle
7/13-
ave
Date
Carle
7/13-
ave
Lo Cone Tank - Determination of
octane
area cone.
I
LMWHI ave ppmC/a 0.315
19.62
19.78
19.70 6.21
pentane
area cone.
I
LMWHI ave ppmC/a 0.315
4.38
4.33
4!36 1.37
octane
area cone.
II
LMWHI ave ppmC/a 0.720

0.00 0.00
octane
area cone.
Ill
LMWHI ave ppmC/a 3.954
1 .81
1 .99
1.90 7.51
1 sopentane
area cone.
Ill
LMWHI ave ppmC/a 3.954
0.31
0.31
0.31 1.22
Concentration
based on HI cone
ethylene/ethanbutanes-butenes
area area
0 . 00 0 . 00
17.27
17.28
17.28 5.44
Low MW Cals - analyzed same
(sopentane propane
area area
3.88 ' 4.31
3.83 4.07
3.86 1.21 4.19 1.
day
propylene
area
4.41
4.85
32 4.63 1.46
2-me-l , 3-butad lene
area
8.21
8.21 2.59
ethylene
area
1.87
1.87 1.35
1-butene
area
0.33
0.33
0.33 1.30
pentane
area
0 . 00 0 . 00

ethane
area
1.19
1.19 0.86
n-butane
area
0. 18
0. 18-
0.18 0.72

propane/al lenepropylene
area area
3.45 2.67
3.45 2.48 2.67 1.
cls-2-butene trans-butene
area area
0.33 0.34
0.33 1.32 0.34 1.



92
1 , 3-butad lene
area
0.30
0.30
36 0.30 1.18
propane/proplene/al lene
area
0.99
0.99
0.99 3.91



-------
Low MW Tank as source
Date
5/30

Carle I


ave
Cal Today ppmC/area
Date
5/30

Carle I


ave
Cal Today ppmC/area
Carle I ave ppmC/area
Date
5/30

Carle II


ave
Cal Today ppmC/area
for cal
octane
area


329.00
336. 10
332.55
0.29
pentane
area


65.10
65.20
65.15
0.306
0.307
octane
area


129.70
129.70
129.70
0.741
for determination of
HMW Cal Tank
came day
et hy lone/ etna nbutenes-butenes Isopentane
cone . area
96.17 18.

56.70
56.50
56. G0
0.33
area
84

281 .40
280.50
200.95
0.30
area
05.29 17

58.30
57.70
58.00
0.31

propane propylene
area area
.72 20.80 20.99

70.30 68.40
','fS.ZS 60.20
70.25 e;-i.30
0.30 0.31
2-me-l ,3-butad)ene
cone. area
19.96 41.

131.90
131.80
131 -85
0.316

ethylene
cone. area
96. 17 9.

12.80
12.60
12.70
0.728

65






ethane
area
25

13.40
13.20
13.30
0.721








propane/a 1
area
9i59 35

48. 10








1 enepropy lene
area
.51 20.99

28.80
40.70 
-------
Cn
(O
i Date Isopentane pentane
j 5/30 area cone. area
17.72 19.96
Carle III
4.30 4.71
4.32 4.72
ave 4.31 4.72
Cal Today ppmC/area 4.11 4.23
Car.le III ave ppmC/are 4.09

High MW Tank
Date
Carle I
7/18- AIRCO

ave
Date
Carle II
7/18- AIRCO

ave
Date
Carle III
7/18- AIRCO

ave

- Determination of
toluene
area
ave ppmC/a 0.313
60.50
60.50
tol uene
area
ave ppmC/a 0.693
26.40
26.40
tol uene
area
ave ppmC/a 3.783
6. 12
6.12
propane/proplene/al lene
area
56.50
13.90
13.90
13.90
4.06
Concentration based on Alrco Ca.l
ethylbenzene
cone. area cone.
63.50
18.93 63.50 19.86
ethylbenzene
cone . area
25.80
18.29 25.80 17.87
ethylbenzene
cone. area cone.
6.42
23.15 6.42 24.29
m-xy lene
area
61 .90
61.90
m-xy lene
area
29.60
29. G0
m-xy lene
area
6.80
6.80
s - analyzed same day
o-xylene 1-pentene
cone. area cone. area cone.
60.80 49.60
1.9.36 60.80 19.02 49.60 15.52
o-xylene benzene
area area
27 .00 24 .00
20.50 27.00 18.70 24.00 16.62
o-xylene 1-pentene benzene
cone. area cone. area cone. area
6.37 7.21 5.74
25.72 6.37 24.10 7.21 27.27 5.74 21.71
                                                                                                            NOT  USED

-------
Date                  Isobutane
                        area
Carle III
7/18- AIRCO ave ppmC/a 3.783
                                       eyelohexane
                                 cone.   area

ave
Date
Carle I
8/14- AIRCO
ave
Date
Carle II
8/14- AIRCO
ave
3.03
3.03
tol uene
area
ave ppmC/a 0.345
54.84
55.94
55.39
tol uene
area
ave ppmC/a 0.805
24.42
24.63
24.53
9.G4
11.46 9.64 36.47
ethy Ibenzene
cone. area cone.

59.31
61 .65
19.14 60.48 20.89
ethylbenzene
cone. area
25.19
25.77
19.74 25.48 20.51


m-xylene
area

57.65
60.26
58.96
m-xy lene
area
30.66
31.12
30.89


o-xylene
cone. area cone.

55.37
58.22
20.37 56.80 19.62
o-xylene
area
25.87
26.31
24.86 26.09 21.00


1-pentene
area cone .

44.68
44.88
44.78 15.47
acetylene benzene
area area
20.61 23.00
21.70 23.10
21.16 17.03 23.05 18.55
Date                  toluene
                        area
Carle  III
8/14-  AIRCO ave ppmC/a  3.490
ave

Date
                         6.73
                         6.82
                         6.78
                      Isobutane
                        area
Carle III
8/14- AIRCO ave ppmC/a 3.490
 ave

 Date
                        3.37
                        3.38
                        3.37

                      toluene
                        area
                                       ethylbenzene  m-xylene
                                     o-xylene
1-pentene
benzene
                                                          7.29           7.08
                                                          7.45           7.08
                                                          7.37   25.73    7.08   24.70
        7.07
        7.20
23.65   7.14  24.91

      cyclohexane
cone.   area
11 .77

      ethylbenzene  m-xylene         o-xylene      1-pentene
cone.   area  cone.       area  cone.    area  cone.      area   cone.
    4.30          6.36
    4.31          6.43
    4.30  15.02   6.40  22.32

-------
Carle I
8/14- LIQUID ave ppmC/ 0.367

                       54.84
                       55.94
ave                    55.39
Date                  toluene
                        area
Carle II
8/14- LIQUID ave ppmC/ 0.681

                       24.42
                       24.63
ave                    24.53
                  59.31
                  61.65
           20.32  60.48   22.19
                 ethylbenzene   m-xylene
           cone.    area
                  25.19
                  25.77
           16.70  25.48   17.35
57.65
60.26
58.96
'lene
area
30.66
31. 12
30.09
55.37
58.22
21.63 56.80 20.84
o-xylene
area
25.87
26.31
21.04 26.09 17.77
44.68
44.88
44.78
acety 1 ene
area
20.61
21 .70
21.16
16.43
benzene
area
23.00
23. 10
14.41 23.05 15.70
Date                  toluene
                        area
Carle III
8/14- LIQUID ave ppmC/ 3.343

                        6.73
                        6.82
ave                     6.78

Date           '       Isobutane
                        area
Carle III
8/14- LIQUID ave ppmC/ 3.343

                        3.37
                        3.38
ave                     3.37
                 ethylbenzene  m-xylene         o-xylene      1-pentene       benzene
           cone.    area   cone.       area   cone.    area  cone.      area  cone.    area
                   7.07
                   7.20
           22.65   7.14  23.86

                 cyclohexane
           cone.   area
           11.27
                         7.29           7.08
                         7.45           7.08
                         7.37   24.65    7.08   23.66
4.30          G.36
4.31          6.43
4.30  14.38   C.40  21.38
Date
toluene
ph x64
Carle  I
8/14-LIQ av ppmC/1n x4 0.434
 attn  correction
 ave
    16

  2.95
  3.00
  2.98
      m-xylene      o-xylene
cone,  ph x64   cone.      ph In  cone.
                   0.88
                                          1.55
                                          1 .60
                                  20.65    1.58  22.28
                         1.11
                         1 .27
                         1 .30
                         1.29  22.81

-------
Date                  toluene
                       ph 1 n
Carle II
8/14-LIQ  av ppmC/ln x4 0.490
      m-xylene      o-xylene
cone.  ph in  cone.     ph in  cone.
        0.93
                         1 . 16
attn correction
                          16
ave
Date
Carle III
8/14-LIQ av
2.47
2.50
2.49
toluene
ph In
ppmC/ln x4 0.226
1 .80
19.48 1.80 26.79
m-xy lene
cone. ph In cone.
0.41
1.22
i.2b
1.24 22.88
o-xylene
ph in cone.
0.51
attn correction 16
ave
Date
Carle I
5/30- LMWHC
ave
Date
Carle I
5/30- LMWHC
ave
5.72
5.62
5.67
tol uene
area
ave ppmC/a 0.307
66.80
66.71
66.76
isobutane
area
ave ppmC/a 0.307
29.60
30.70
30. 15
3.72
3.80
20.54 3.76 24.53
ethylbenzene
cone. area cone.

66.50
67.70
20.50 67.10 20.61
cone .

9.26
2.80
2.85
2.83 23.15
m-xylene o-xylene 1-pentene acetylene
area cone. area cone. area cone. area cone.

38.20 NA 48.60 42.80
42.00 NA 48.60 43.70
40.10 12.31 NA NA 48.60 14.93 43.25 13.28




-------
Date                  toluene
                        area
Carle II
5/30- LMWHC ave ppmC/a 0.732

                       25.40
                       25.70
ave                    25.55
                 ethylbenzene   m-xylene
           cone.    area              area
                  26.00
                  26.40
           18.69  26.20  19.17
                                     o-xylene
                                       area
                        29.50          27.10
                        30.80          28.20
                        30.15   22.06   27.65
            acetylene
                area
      20.23
23.80
24.60
24.20
             benzene
               area
       23.40
       23.70
17.70  23.55  17.23
Date

Carle III
5/30- LMWHC ave ppmC/a
ave
toluene
  area

 4.094

  5.83
  5.80
  5.82
Date                  Isobutane
                        area
Carle III
5/30- LMWHC ave ppmC/a 4.094
      ethylbenzene   m-xylene          o-xylene     'l-pentene       benzene
cone.    area  cone.       area   cone.    area   cone.      area  cone.    area
        6.08
        6.17
23.81   6.13  25.07

      cyclohexane
cone.   area
                                                          6.24
                                                          G.32
                                                          G.28  25.71
5.97            3.63          5.48
6.05            3.61          5.47
6.01  24.60     3.62  14.82   5.48  22.41


ave
ethylene
propy lene
propane
tol uene
trans-2-butene
Isopentane

2
2
2
.84
.85
.85
Alrco Tank
tnanuf. UNC-
4.80
9
8
9
6
9

.15
.61
.94
.92
.85



11


.65
Dave
3.65
8
7
10
6
9

.34
.47
.72
.91
.03

4
5
5
DCS
UNC
3
8
7
10
7
9

.67
.53
. 10


20.


88
112 UNC/
- Catnanuf .
.73 0.78
.48
.61
.92
.03
.18

0.
0.
1 .
1.
0.
0.
93
88
10
02
93
94

-------
       Low MW Tank HC
     HIGH CONC                  DCS 109
manuf.  UNC-Dave U-Dave/Spread/   Spread
Carle I
Carle II
Carle ICarle III
ethylene
ethane
propy lene
propane
a 1 lene
1 -butene
butane
c is-2-butene
trans-2-butene
1 , 3-butad lene
10.0
10.JBT
21.9
21 .8
15.0
19.6
10.8
20.0
19.6
20.0
manur .
9.32 0.93
9.84 0.98
20.70
20.50
17.60
18.80
10.40
18.90
19.40
17.20
0.95
0.94
1 .17
0.96
0.96
0.95
0.99
0.86
Isopentane 18.1 17.40 0.96
pentane 20.0 19.70 0.99
2-methyl-l ,3-butadlene 17.1 41.40 2.42
octane 104.2 104.00 1.00
ave 0.97
Low MW Tank
HIGH CONC
manuf. UNC-Spd UNC/
manuf .
ethylene
ethane
propylene
propane
al lene
1 -butene
butane
c 1 s-2-butene
trans-2-butene
1 , 3-butad lene
t sopentane
pentane
2-methyl-l ,3-butadlene
octane

not Including 2-me-l,3
10.0
10.0
21 .9
21.8
15.0
19.6
10.8
20.0
19.6
20.0
18.1
20.0
17.1
104.2

-butad lene
9.25
9.59
20.99
20.80
14.70
18.93
10.48
19.05
19.54
17.29
17.72
19.96
41 .65
96.17

ave
0.92
0.96
0.96
0.95
0.98
0.97
0.97
0.95
1 .00
0.86
0.98
1.00
2.44
0.92

0.96
nianuT .
0.92
0.96
0.96
0.95
0.98
0.97
0.97
0.95
1.00
0.86
onoeT.
9.25
9.59
2^.99
2ii . iiJtl
! •' . 7.0'



20.99


20.80 sum >
1 G ii" }
10!48 >
19.05 sum )
19.54 >
17.29 >
0.98 17.72
1.00 19.96
2.44 41.65
0.92 96.17
0.96
LOW CONC
THC UNC
x 0.073
0.675
0.700
1 .532
• 1 .519
1.073
1 .382
0.765
1.390
1 . 426
1 .262
1 .294
1 .457
3.041
7.020

0.073
Dave
0.634
0.644
1 .350
1 .310
.240
.300
0.713
.310
.350
. 180
1 .210
1 .340
2.520
6 .000


low/hi

0.069
0.067
0.064
0.063
0.084
0.069
0.068
0.069
0.069
0.068
0.068
0.067
0.061
0.071

0.068
86.34
17.72
19.96
41 .65
92.08
DCS 111
UNC
Spread
0.629
0.628
1 .458
1 .320
. 148
.305
0.716
.317
.356
.184
1.218
1.372
2.586
6.859



9.25
9.59
18.84

}


32.80 sum > 59.20 58.1

89.56
low/hi low/hi/
{ 18.93 18.9
{ 15. <3 10.4
85.29 l-j.JJS 10.2
{ 19.54 19.3
{ 17.29 17.6

20.54 17.5
29.64 19.3
106.87
Carle II ph der
avg low/hi 5/30 7/13
0.068
0.065
0.069
0.063
0.078
0.069
0.063
0.069
0.069
0.068
0.069
0.069
0.062
0.071

0.069
0.99
0.95
1.01
0.93
.14
.01
.00
.01
.01
.00
1 .00
1.00
0.91
1 .04

1 .00
-0.01 0.667 0.592
-0.05 0.610 0.645
0.01
-0.07
0.14
0.01
.00
0.01
0.01
.00
.00
.00
-0.09
0.04
sd
0.05
avg
0.629
0.628
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000



-------
                            High MW Tank
                                             Carle  I
                                                                      Carle  II
Carle III
o>
      acetylene
      Isobutane
      1-pentene
      benzene*
      toluene*
      ethylbenzene*
      m-xylene*
      o-xylene*
      * AVE NOT USED
      toluene
      m-xylene
      o-xylene
benzene*
toluene*
ethylbenzene*
m-xylene*
o-xylene*
       acetylene
       t sobutane
       1-pentene
       benzene**
       toluene*
       ethylbenzene**
       m-xylene*
       o-xylene*
Sprd av
15.49
11.08
15.22
19.49
20.46
21 .38
22.88
21.29

manuf .
21.10
22.30
22.40
5/30 7/18 8/14-T
13.28
9.26
14.93 15.52 15.47

20.50 18.93 19.14
20.61 19.86 20.89
12.31 19.36 20.37
NA 19.02 19.62
ave/ Meas .
Dave manuf . ave
20.40 0.96 20.22
23.50 1.05 23.41
22.50 1.02 22.95
8/14-L 5/30 7/18 8/14-T 8/14-L 5/30 7/18
17.704
11.646 11.46
16.43 14.819
17.228 16.62 18.55 15.70 22.413 21.71
20.32 18.691 18.29 19.74 16.70 23.805 23.15
22.19 19.167 17.87 20.51' 17.35 25.074 24.29
21.63 22.057 20.50 24.86 21.04 25.709 25.72
20.84 20.228 18.70 21.00 17.77 24.603 24.10

Carle I Carle ICarle III
20.65 19.48 20.54
22.20 26.79 24.53 *CARLE II M-XYLENE NOT USED
22.81 22.88 23.15
8/14-T 8/14-L
11.77 11.27
15.02 14.38
22.32 21.38
23.65 22.65
24.91 23.86
25.73 24.65
24.70 23.66





Car IIIMeas. ave/Car III *
ave
21.96
23.32
24.53
25.45
24.27
ave res
High MW

manuf .
14.60
12.00
15.20
20.00
21.10
23.80
22.30
22.40

Car III ave res.
20.00
0.87
22.34
0.92
0.95
0.91
Tank
UNC RTI-DavSprd av
RTI-Dave manuf. manuf.
12.90 0.88 1.06
11.30 0.94 0.92
14.60 0.96 1.00
20.00 1.00 1.00
20.40 0.97 0.96
22.50 0.95 0.94
23.50 1.05 1.05
22.50 1.00 1.02
0.97 0.99






,

DCS 108 Carle III
Sprd ave Mo's 83
15.49
11.08 12.20
15.22 15.50
20.00 22.70
20.22 23.90
22.34 24.30
23.41 25.30
22.95. 23.50





















-------
                       80.45     24.71

                      1sopentane
                        area     cone.
                       41 .40
                       41 .60
                       41 .50
                                 12.74
Carle I cal 5/30
ppmC/area =
               0.3071
Carle I cal 5/30
ppmC/area =
               0.3071
Carle II cal 5/30
ppmC/area =
               0.7316
Carle II cal 5/30
ppmC/area =
               0.7316
Carle  III cal 5/30
ppmC/area =
               4.0938
                       Isopentane
                         area      cone.
 Carle  III  cal  5/30       3.03
 ppmC/area  =              3.01
                4.0938  	
                         3.02      12.36
                      D1 methylpentane  n-Octane      Ethylene
                        area     cone.   area  cone.      area
                       79.10           117.80             8.50
                       81.80           130.G0             8.52
      n-Butane      cls-butene
cone.   area  cone.     area  cone.
       26.30           23.90
       26.40           24.10
                                                                                               7.37
124. 15 38. 13
2-me-l -butene
area cone.
39.90
40.40
40.15 12.33
n-Octane
area cone .
45. G0
48.80
47.20 34,53

n-Octane
area cone.
10.40
11.20
10. C!0 44.21
2-mc- 1 -butene
area cone.
3.21
3.22
0.51
propane
a i- ea
19.40
19.70
19.55
Ethy lene
area
3.43
3.43
3.43
propane
area
5.98
6.00
5.99


2.61
cone .
6.00
cone .
2.51
cone .
' 4.38


26.35 8.09
propy lene
area cont.
18.20
18.40
18.30 5.62

propy lene
area cone.
6.06
6.02
6.04 4.42
n-Dutane
area cone .
1 .90
1 .89
1.90 7.76
24.00



c 1 s-butene
area
1 .78
1 .77
1 .78
propane/propylene
area cone.
2.85
2.83
                                                                                              cone.
                                                                                                7.27
                                         3.22   13.16
 2.84   11.63

-------
d 1 methyl pen tane
n-octane
ethylene
n-butane
c 1 s-2-butene
t sopentane
2 -me- 1-butene
propane
propy lene
propylene+propane


propylene- RTI











ethylene- RTI



















N
1
1
1
1
1
1
6




N
1
1
1
0
1
1
manuf .
28.700
27.680
2.680
8.000
6.800
1 9 . 500
15.450
6.060
5.430
11 .49



X
1
26
27
104
749
2078





X
1
29
106
740
1180
2225
DCS #1
19.190
20.223
2.718
7.979
8.000
12.971
16.300
6.119
6.120
12.239



Y
14.58
14.82
14.34
14.94
14.79
14.4





Y
19.40
19.20
19.80
16.00
20.00
1 9 . 00
Car I
24.706
38.127
2.613
8.092
7.370
12.745
12.330
6.004
5.620
11.623



N*X
1
26
27
104
749
2078
2985
YBAIi :
XBAIl:


N*X
1
29
106
0
1100
2225
Car II
34.530
2.509




4.382
4.419




N*Y
14.58
14.82
14.34
14.94
14.79
14 .40
87.87
14.645
497.5


N*Y
19.40
19.20
19.00
0.00
Zii . 00
19.00
Car III
44.213

7.758
7.267
12.363
13.162

-------
Official Calibration Sources
              168

-------
OFFICIAL CALIBRATION SOU1CSS
ID
NVM
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
37
37
37
37
37
37
37
37
37
33
39
40
41
42
43
44
45
46
47
43
49
50
51
52
53
53
53
53
53
53

OSSC
TOLUENEINO
TOLUENEINO


TOLUENEINO
MXYLENEINO'
TOLUENEINO
TOLUENEINO
124TMBINOB
124TMBINOR
MXYLENEINO
MXYLENEINJ
OXYLENEINO
OXYLENEINO
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
TOLUENEINO
ISOPENTIMO
124TM3INO
TOLUENEINO
ISOPENTIrlO
124TMBINO
TOLUENEINO
OXYLENEINO
METOLINO
224TMPINO
ISOPENINO
OXYLENEINO
TOLUENEINO
DETHKETINO
DETHKETINO
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS

SDATE
18-Auq-1933
2S-A;j!J-1983


22-Aug-1983
22-Aug-1983
6-Oct-1983
6-Oct-1983
6-Oct-1983
6-Oct-1983
6-Oct-1933
S-Oct-1933
6-Oct-1983
6-Oct-1983
23-Ou1-1983
23-Oul-1983
23-Oul-1933
23-Oul-1933
23-Ou 1-1-J83
23-Oul-1983
23-Ou1-1933
23-Oui-1983
23-Oui-1983
23-Oui-19Q3
25-Ou1-1983
25-Ou1-1933
25-Oul-1983
25-Oul-1983
25-Ou1-1933
25-Ou1-1983
8-Aug-1983
8-Aug-1983
9-Aug-1983
19-Aug-1983
19-Aug-1983
l-Aug-1983
l-Aug-1983
20-Oun-1983
20-Oun-1983
9-Sep-1983
9-Sep-1983
9-Sep-1983
9-Sep-1983
9-Sep-1983
9-Sep-1983
SER
MUM
NA
NA


NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA






                  TOLUENE
                  TOLUENE
                  TOLUENE
                  M-XYLENE
                  TOLUENE
                  TOLUENE
                  1,2,4-TRIMETHYLBENZE
                  NE
                  1 ,2,4-TRIMETHYLBENZE
                  NE
                  M-XYLENE
                  M-XYLENE
                  0-XYLENE
                  0-XYLENE
                  2-METHYLPENTANE
                  2 ,'3-DIMETMYLPENTANE
                  METHYLCYCLOHEXANE
                  TOLUENE
                  ETHYLBENZEME
                  M-XYLENE
                  N-PROPYLBEMZENE
                  2,2,4-TRIHETHYLPENTA
                  NE
                  TERT-BUTYLBENZENE
                  1 ,3,5-TRIMETMYLBENZG
                  NE
                  TOLUENE
                  ISOPENTANE
                  1 ,2,4-TRIMETHYLBENZE
                  NE
                  TOLUENE
                  ISOPENTANE
                  1 ,2,4-TRIMETHYLBENZE
                  NE
                  TOLUENE
                  0-XYLENE
                  M-ETHYLTOLUENE
                  2,2,4-TRIMETHYLPENTA
                  NE
                  ISOPENTANE
                  0-XYLENE
                  TOLUENE
                  3-PENTANONE
                  3-PENTANONE
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1,2,4-TRIMETHYLBENZE
                  NE
                  SEC-BUTYLBENZENE
                  M-ETHYLTOLUENE
25-?-
Pag-
STAT-1D
cone
0.795
0.788
ft cTf •- •;
•• .-• /••

0 . j ,: ..,
0.504
0.381
0.381
0.233
0.233
0. 180
0.180
0.111
0.111
1 .000
1 .000
1 .000
1 . 000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
0.601
0. 166
0.369
0.601
0. 166
0.369
9. 125
6.083
4.000
4.053
4.059
2.699
4.553
1 .010
1.010
1 .000
1 .000
1 .000
1 .000
1 .000
1.000
p-1985
?
ACT'JAL
COMC
0.795
0.788
U . 00.3
ft ~t rt <~t

'.' . '. '
Ili .SIS;
0.381
0.381
0.233
0.233
0. 180
0. 180
0. Ill
0.111
1 .000
1 .000
1 . 000
1 . 000
1 .000
1 . 000
1 . 000
1 . 000
1 .000
1 . 000
0.601
0. 1Gb
0.369
0.6:31
0. 135
0.369
9. 125
6.083
4.000
4.053
4.059
2.699
4.529
1 .010
1 .010
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
             170

-------
OFFICIAL CALIBRATION SOURCES
ID
NUM
53
53
53
54
54
55
55
55
55
55
55
55
55
55
56
56
56
56
56
56
56
56
56
57
57
57
57
57
57
57
57
57
58
58
58
58
58
58
58
58
58
59
59
59
DESC
PE900CLMXS
PE900CLMXS
PE900CLMXS
NEW83COCH4
NEW83COCH4-
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLi''XS
PE900CLMXS
PE900CLHXS
PE900CLMX5
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE90CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMX4
PE900CLMXS
PE900CLMXS
SDATE
9-Sep-1983
9-Sep-1983
9-Sep-1983
30-Sep-1933
30-Sep-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
3-Oct-1983
4-Sep-1983
4-Sep-1983
4-Sep-1983
4-Sep-1983
4-Sep-l983
4-Sep-1933
4-Sep-1983
4-Sep-1933
4-Sep-1333
24-Jun-1983
24-Oun-1983
24-Oun-1983
24-Oun-1983
24-Oun-1983
24-Jun-19S3
24-Jun-1983
24-Oun-1983
24-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Oun-1983
30-Jun-1983
3-Jun-1983
3-Oun-1983
3-Jun-1983
          SER
          NUM
                  ISOPROPYLBENZENE
                  1 ,2,3-TRIMETHYLBENZE
                  NE
                  BUTYLBENZENE
                  METHANE
                  CO
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1 ,2,4-TRIMETHYLBENZE
                  NE
                  SEC-BUTYLBENZENE
                  M-ETHYLTOLUENE
                  ISOPROPYLBENZENE
                  1 ,2,3-TRIMETHYLBENZE
                  N.E
                 ' .BUTYLBENZENE
                  BE-NZENE
                  P-XYLENE
                  0-XYLENE
                  1 ,2,4-TRIMETHYLBENZE
                  NE
                  SEC-3UTYLBENZENE
                  M-ETHYLTOLUENE
                  ISOPROPYLBEHZENE
                  1,2,3-TRIMETHYLBENZE
                  NE
                  BUTYLBENZENE
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1,2,4-TRIMETHYLBENZE
                  NE
                  SEC-BUTYLBENZENE
                  M-ETHYLTOLUENE
                  ISOPROPYLBENZENE
                  1,2,3-TRIMETHYLBENZE
                  NE
                  BUTYLBENZENE
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1,2,4-TRIMETHYLBENZE
                  NE
                  SEC-BUTYLBENZENE
                  M-ETHYLTOLUENE
                  ISOPROPYLBENZENE
                  1,2,3-TRIMETHYLBENZE
                  NE
                  BUTYLBENZENE
                  2-METHYLPENTANE
                  2,3-DIMETHYLPENTANE
                  METHYLCYCLOHEXANE
25-Sep
Page 3
STATED
CONC
1 .000
1 .000
1 .000
4.934
5.114
1 .000
1 .000
1 .000
1 .000
1.000
1.000
1.000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 . 000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1.000
1 .000
1.000
1 .000
1 .000
-1S35
ACTUAL
CONC
1 .000
1 .000
1 .000
4 .934
5.114
1 .000
1 .000
1 . 000
1 .000
1.000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 . 000
1 . 000
1 . 000
1 . 00;J
1 . 000
1 . 00:J
1 . 000
1 . 000
1 . 000
1 .000
1 .000
1 ..'J0'J
1 . 000
1 .000
1 .000
1 . 000
1 . 000
1 . 000
1 . 000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1.000
              171

-------
OFFICIAL CALIBRATION SOURCES
 25-Sep-1985
 Pa-jo 4
ID
NUM
59
59
59
59
59
59
59
60
60
50
60
60
60
60
60
60
61
61
61
61
1J1
61
61
61
61
62
62
62
62
62
62
62
62
62
62
63
63
63
63
63
63
63
63

DESC
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXC
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMXS
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMXS
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMXS
PE900CLMX5
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5

SDATE
3-Oun-l983
3-Oun-1983
3-Oun-1983
3-Oun-1933
3-Oun-1983
3-Jun-1983
3-Oun-1983
6-Ou1-1933
6-Oul-1933
6-Ou1-1983
6-Oul-1983
6-Oul-1983
6-Oul-1983
6-Ju1-1983
6-Oul-1933
6-OuT-1983
18-0ul-1933
18-Jul-1983
18-Jul-1983
13-Ou 1-1983
lS-Oul-1933
13-Ou 1-1983
13-Oul- 1983
13--0ul-1933
18-Jul-1983
18-Cul-1983
18-0ul-1983
lS-Oul-1983
18-0ul-1983
18-0ul-1983
13-0ul-1983
18-Jul-1983
18-0ul-1983
18-0ul-1983
18-0ul-1983
23-Oul-1983
23-Oul-1983
23-Ou1-1983
23-Ju1-1983
23-Oul-1983
23-Oul-1983
23-Oul-1983
23-Jul-1983









NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
MA
NA
MA
NA
NA


















          SER
          NUM
STATED  ACTUAL
 COMC    CONC
                  TOLUENE
                  ETHYLBENZENE
                  M-XYLENE
                  N-PROPYLBENZENE
                  2,2.4-TRIMETHYLPENTA
                  NE
                  TERT-BUTYLBENZENE
                  1 .3.S-TRIMETHYLBENZE
                  NE
                        BENZENE
                        P-XYLENE
                        0-XYLENE
                        1 ,2,4-TRIMETHYLBENZE
                  NE
                        SEC-BUTYLGENZENE
                        M-ETHYLTOLUENE
                        ISOPROPYLBEMZENE
                        1,2,3-TRIMETHYLBENZE
                  NE
                        BUTYLEENZEME
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1 , 2,4-TRIMETHYLBEHZE
                  NE
                  SEC-BUTYLGEMZENE
                  M-ETMYLTOLUGNE
                  ISOPROPYLBEMZENE
                  1 ,2,3-TRIMETHYLBENZE
                  NE
                  BUTYLBENZENH
                  2-METHYLPEMTANE
                  2 ,3-DIMETHYLPENTAMG
                  METHYLCYCLOHEXANE
                  TOLUENE
                  ETHYLBENZENE
                  M-XYLENE
                  N-PROPYLBENZENE
                  2,2,4-TRIMETHYLPENTA
                  NE
                  TERT-BUTYLBENZENE
                  1 ,3,5-TRIMETHYLBENZE
                  NE
                  BENZENE
                  P-XYLENE
                  0-XYLENE
                  1,2,4-TRIMETHYLBENZE
                  NE .
                  SEC-BUTYLBENZENE
                  M-ETHYLTOLUENE
                  ISOPROPYLBENZENE
                  1 ,2,3-TRIMETHYLBENZE
                  NE
1 .000
1 . 000
1 .000
1 .000
1 . 000
1 .000
1 .000



:NZE



:NZE

i .000
i .000
1 .000
1 . 000
i .000
i . 000
1 . 000
1 . 000
1 .000
1 . 000
1 . 000
1 .000
! . 000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 . 00,'J
1 . 000
1 .000
1 . 000
1 . 000
1 . 000
1 .000
1 .000
1 . 000
1 .000
1 .000
1.000
1 . 000 .
1 .000
1 . 000
1 . 005
1 . 000
1 . 00'J
1 . 00.7
1 . 00.'?
1 . 006'
1 . 00.'?
1 . 00.0
1 . 00/7
1 . 00,0
1 . 00/1
1 . 000
1 . 000
1 .000
1 .000
1 .000
1 .000
1 . 000
1 .005
1 . 000
1 . 000
1 .000
1 .000
1 . 000
1 .000
1 .000
1 .000







1 .000
1 .000
1 .000
1 .000
1 . 000
1 .000
1 .000
1 .000
1 . 000



























                 172

-------
OFFICIAL CALIBRATION SOURCES
                        25-Sep-1985
                        Page  5
ID
NUM
G3
64
64
64
64
64
64
64
64
64
64
65
65
65
65
65
65
65
65
55
66
57
63
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94

DESC
PE900CLMX5
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX4
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
PE900CLMX5
HCOHINO
HCOHIMO
HCOHINJ
HCOHINO
HCOHINO
HCOHINO
HCOHINO
HCOHINO
HCOHINO
HCOHINO
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
no2aspan
noZaspan
no2aspan
noZaspan
N02ASPAN
N046PPMCAL

SDATE
23-Oul-1983
23-Oul-1983
23-Oul-1983
23-Oul-1933
23-Oul-1983
23-Oul-1983
' 23-Oul-1983
23-Oul-1983
23-Oul-1983
23-OU1-1983
23-Oul-1983
2-Aug-1983
2-Aug-1983
2-Aug-1983
2-Aug-1983
2-Aug-1983
2-Aug-1933
2-Aug-1383
2-Aug-1983
2-Aug-1983
6-Oun-1983
lS-Oun-1982
24-Nov-1983
18-0un-1983
12-0u1-1983
12-Aug-1983
23-Oul-1983
29-Oul-1983
13-Aug-1983
19-Sep-1983
10-0un-1983
18-0un-1983
13-0ul-1983
17-0u1-1983
18-0ul-1983
20-Jul-1983
21-0ul-1983
26-Oul-1983
29-Oul-1983
31-0ul-1983
4-Aug-1983
13-Aug-1983
5-Sep-1983
7-Sep-1983
18-Sep-1983
23-Sep-1983
2-Oct-1983
14-0ct-1983
22-Oun-1934
          SER
          NUM
                      STATED  ACTUAL
                       CONC    CONC
      SX13759
BUTYLBENZENE
2-METHYIPGNTANE
2,3-DIMETHYLPENTANE
METHYLCVCLOHEXANE
TOLUENE
ETMYLBENZEME
M-XYLENE
N-PROPYLBENZENE
2,2,4-TRIMETHYLPENTA
NE
TERT-BUTYL8ENZENE
1 ,3,S-TRIMETHYLBENZE
NE
BENZENE
P-XYLENE
0-XYLENE
1,2,4-TRIMETHYLBENZE
NE
SEC-BUTYLBENZENE
M-ETHYLTOLUENE
ISOPROPYLEENZENE
1 ,2,3-TRIH.ETHYLBEHZE
NE
BUTYLBENZENG
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
FORMALDEHYDE
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
NO.
1 .000
1 .000
1 .000
1 . 000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1.000
1 .000
1 .000
1.000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .072
1 .000
1 .024
1 .041
1 .029
1 .032
1 .047
1 .026
1.016
1 .026
0.244
0.152
0.037
0.071
0.055
0.081
0.061
0.025
0.074
0.059
0.036
0.074
0. 105
0.096
0. 114
0. 100
0.200
0.110
46.000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1 .000
1.000
1 .000
1 .000
1 . 00.0
1 . 000
1 .072
1 .2133
1 .024
1 .04 1
1 .029
0.000
1 .047
1 .025
1 . :J 1 5
1 .02G
0.24J
0. 152
0.037
0.071
0.055
0.051
0.061
0.025
0.074
0.059
0.036
0.074
0.105
0.096
0. 114
0. 100
0.200
0. 1 10
52.600
              173

-------
ID
NUM
95
96
97
98
99
100
101
102
103
104
105
106
107
108
108
108
108
108
108
103
103
109
109
109
109
109
109
109
109
109
109
109
109
109
109
110
110
1 10
110
110
110
110
1 10
111
111
111
111
111
111
111
111
111

DESC
MXYLENEINO
TOLUENEIMO
TOLUENE
M-XYLENE
0-XYLENE
TOLUENE
M-XYLENE
FORMALDEHY
FORMALDEHY
HCHOINOB
HCHOINOR
HCHOINJRAF
HCHOINORMO
HIMWHICONC
HIMWHICONC
HIMWHICONC
HIMWHICONC
HIMWHICONC
HIMWHICONC
HIMWHICOMC
HiMWHICOMC
LOMWHICOMC
LCMWHICOMC
LOMWHICOMC
LOMWHICO;!-:
LOMWHICOfiO
LOMWHICO;IC
LOMWHICOMC
LOMWHICCfIC
LOMWHICOMC
LOMWHICOMC
LOMWHICOMC
LOMWHICOMC
LOMWHICOMC
LOMWHICOMC
SCOTT84
SCOTT84
SCOTT84
SCOTT84
SCOTT84
SCOTT84
SCOTT84
SCOTT84
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC
LOMWLOCONC



SDATE
25-Oun-
25-Jun-
27-Oun-
26-Jun-
26-Oun-
19-Jun-
19-Oun-
10-Oul-
10-Oul-
1 1-Oul-
11-Jul-
27-Jul-
27-Jul-







































1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
















































NA
NA
NA
NA







































OFFICIAL CALIBRATION SOURCES
          SER
          NUM
                  M-XYLENE
                  TOLUENE
                  TOLUENE
                  M-XYLENE
                  0-XYLENE
                  TOLUENE
                  M-XYLENE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  ACETYLENE
                  ISOBUTANE
                  1-PENTENE
                  BENZENE
                  TOLUENE
                  ETHYLBGNZENG
                  M-XYLENE
                  0-XYLENE
                  ETHYLENE
                  ETHANE
                  PROPYLEME
                  PROPANE
                  PROPADIENE
                  1-BUTEME
                  N-BUTANE
                  1 , 3-BUTADIElNE
                  TRAN3-2-BUTENE
                  CIS-2-BUTGME
                  ISOPENTANE
                  N-PENTANE
                  N-OCTANE
                  2-METHYL-l,3-BUTADIE
                  NE
                  ETHYLENE
                  PROPYLENE
                  PROPANE
                  CIS-2-BUTGHG
                  ISOPENTANE
                  N-PENTANE
                  TOLUENE
                  2-METHYL-2-BUTENE
                  ETHYLENE
                  ETHANE
                  PROPYLENE
                  PROPANE
                  PROPADIENG
                  1-BUTENE
                  N-BUTANE
                  1,3-BUTADIENE
                  TRANS-2-BUTENE
25-Sep
Pag* S
STATED
COMC
3.000
6.000
5 . ;700
3.000
3.000
4 . 000
3.000
0.500
1 .060
1 . 000
1 .000
1 .000
0.200
1 4 . 600
1 2 . 000
15. 175
19.956
21 .056
23.734
22.272
22.432
10.004
1 :3 . 000
21 .9.06
2! . 3'34
15. .000
19 .500
10.300
20 . 000
20 . 000
2.0 . 000
13. 135
19.956
* * ***;•!
17.055
0.600
0.900
0.900
1 .000
1 .250
1 .250
2.300
1 .250
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-1985

ACTUAL
CONC
2.963
5.895
4 .903
2.973
2.653
3.929
2.967
0.542
1 .035
0.946
0.946
0.946
0. 189
15.490
1 1 .080
1 5 . 200
19.910
20.300
22.250
23.59.'J
22.95.J
9.25/J
9.59/T
2 0.99. 'J
Z.V.SQ'J
1 4 . 7.0.'J
1 3 . 33.'J
15.48.-;
17.29/J
19 . 540
19 .050
17 . 7C/J
19 . 96,'j
96 . 17''
41 .65.J
0.711
1.123
0..972
1.11 .'!
1 . 143
1 .347
3.250
1 .390
0.629
0.628
1 .458
1 .320
1 . 148
1 .305
0.716
1 . 184
1 .356
             174

-------
OFFICIAL CALIBRATION SOURCES
ID
HUM
111
1 i 1
111
111
1 1 1
112
112
112
112
112
112
112
113
114
115
116
117
113
113
113
113
1 13
113
1 19
120
121
122
123
123
123
123
123
123
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139

DESC
LOMWLOCONC
LOMWLOCONC
LOMWLOCOMC
LOMWLOCONC
LOMWLOCOMC.
NEWAIRC084
NEWAIRC084
NEWAIRC084
NEWAIRC084
NEWAIRC084
NEWAIRC084
NEWAIRC034
PROPYL741
TOLUENE211
HCHOINO
TOLUENEINO
MXYLENEINO
AROMINO
AROMINO
AROMINO
AROMIMO
ARCH I HO
AROMIMJ
TMPENTiNO
BENZENE I NO
RTIETHYL
RTIPROPYL
HIMWLIQIMJ
HIMWLIQINO
HIMWLIQIMJ
HIMWLIQIMJ
HIMWLIQIilJ
HIMWLIQiriJ
HIMWLIQIMO
M-XYLENE
HCHOINO
HCHOINO
MEOHINO
MEOHINO
MEOHINO
MEOHINO
MEOHINO
MEOHINO
HCHOINO
HCHOINO
HCHOINO
HCHOINO
HCHOINO
HCHOINO
85COCH4CAL

SDATE


























SER
NUM












N415664

31-0u1-1984
28-Oun-1984
28-Oun-1934
16-0ul-1984
lS-Oui-1984
16-0ul-1984
lS-Oul-1984
16-0u1-1934
lG-Jul-1984
27-Ou 1-1980
27-Ou1-1984


14-Aug-1934
14-Aug-1934
14-Aug-1984
14-Aug-1984
14-Aug-19S4
14-Aug-1984
14-Aug-1984
27-Oun-1984
13-0ul-1984
13-0ul-1984
24-Oul-1984
3-Aug-1984
4-Aug-1984
7-Aug-1984
8-Aug-1984
9-Aug-1984
7-Oct-1984
7-Oct-1984
9-Oct-1984
9-Oct-1984
16-0ct-1984
16-0ct-1984
3-Apr-1985

NA
NA
NA
NA
NA
NA
NA
MA
NA
NA
NA


DFB
DFB
DFB
DFB
DFB
DFB
DFB
LMQ
LMQ
LMQ
OBS
DFB
ORA
DFB
ORA
LMQ
NA
NA
NA
NA
NA
NA
AAL-




































15140
                  CIS-2-BUTENE
                  ISOPENTANE
                  N-PENTANE
                  N-OCTANE
                  2-METHYL-l,3-BUTADIE
                  NE
                  ETHYLENE
                  PROPYLENE
                  PROPANE
                  TRANS-2-BUTENE
                  ISOPENTANE
                  TOLUENE
                  PROPYLENE/PROPANE
                  PROPYLENE
                  TOLUENE
                  FORMALDEHYDE
                  TOLUENE
                  M-XYLENE
                  BENZENE
                  TOLUENE
                  P-XYLENE
                  M-XYLENE
                  0-XYLENE
                  STYRENE
                  2,2.4-TRIMETHYLPENTA
                  NE
                  BENZENE
                  ETHYLENE
                  PROPYLENE
                  N-PENTANE
                  2-METHYLPENTANE
                  2,3-DIMETMYLPENTANE
                  TOLUENE
                  M-XYLENE
                  0-XYLENE
                  2,2,4-TRIMETHYLPENTA
                  NE
                  M-XYLENE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  METHANOL
                  METHANOL
                  METHANOL
                  METHANOL
                  METHANOL
                  METHANOL
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  FORMALDEHYDE
                  METHANE
2S-Sep-1905
Page 7
STATED
CONC
0.000
3 . 000
0 . 000
0.000
0.000
4.800
9. 150
8.610
6.320
9.850
9.940
17 .760
7.410
14.770
1 .000
6.000
3 . 000
2.000
2 . 000
2 . 000
2 . 000
2 . 000
2.000
1 . 000
1 . 000
1 9 . 000
1 4 . 400
3 . 000
3.000
3 . 000
3 . 000
3 . 000
2.600
3 . 000.
2.000
1 .000
1 .000
0.900
0.900
0.300
0.300
0.790
0.263
1 .000
0.500
1 .000
1 .000
0.250
0.500
1 .907

ACTUAL
CONC
1.317
1.213
1 .372
6.859
2.586
3. 730
8.480
7.610
7.030
9.180
10.920
16.090
7.410
14.770
0.946
5.977
3.009
2 . 000
2 . 000
2 . 00.J
2 . 00.'J
2 . 00.U
2 .00'J
1 . 00J
1 . 000
19.43,'J
14 .650
2.383
3. 161
.99:7
. 93 a
.975
.663
2.992
1 .973
1 !033
1 .033
0.891
0.397
0.321
0.300
0.787
0.263
0.918
0.464
0.920
0.920
0.232
0.465
1 .907
             175

-------
OFFICIAL CALIBRATION  30V.CES
                                          Pag ,  3
ID
NUM
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
150
151
152
152
162
152

163
163
163
163

164
165
165
165
165
165

165

166
166
166
166
166
166

167
168
169
170

DESC
85COCH4CAL
MEOHINO
ACETALDINO
ACETALDINO
MEOHINO
MEOHINO
ACETALDINO
ACETALDINO
MEOHINO
HCHOINO
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
N02ASPAN
M02ASPAM
M02ASPAM
HCHOINO
ACETINO
ACETINO
AROMINO
A KOMI NO
AROMINO
AROMINO

AROMINO
AROMINO
AROMINO
AROMINO

HCHOIMO
AROMINO
AROMINO
AROMINO
AROMINO
AROMINO

AORMINO


AROMINO
AROMINO
AROMINO
AROMINO
AROMINO

RTIPROPYLE
RTIETHYLEN
AIRCONO
SCOTTNO

SDATE
3-Apr-1935
25-Aug-1934
27-Aug-1984
27-Aug-1984
l-Sep-1984
3-Sep-1984
16-Sep-1984
16-Sep-1984
17-Sep-1984
25-Sep-1984
20-Aug-1984
20-Aug-1984
18-Sep-1934
18-Sep-1984
3-Oct-1984
3-Oct-1984
4-Oct-1984
1 l-Oct-1984
30-Aug-1984
6-Oul-1984
S-Sep-1985
lS-Oct-1984
15-0ct-1984
14-May-1985
14-Mc:y-1985
14-Mav-1935
14-Mciy-1385

17-May-1985
17-May-1985
17-May-1985
17-May-1935

20-May-1935
22-May-1985
22-May-1985
22-May-1935
22-May-1985-
22-May-1985

22-May-1985


24-May-1985
24-May-1985
24-May-1985
24-May-1985
24-May-1985

3-Oun-1985
3-Oun-1985


SER
NUM
AAL-15140



ORA
KGS
KGS
KGS
LMQ
DFB
NA
NA
NA
NA
NA
NA
NA
NA
NA
MA
MA
NA
NA
NA
NA
NA
NA

NA
NA
NA
NA

NA
NA
NA
NA
NA
NA

NA


NA
NA
NA
NA
NA

NA
NA
1122590
BAL177


CO
METHANOL
ACETALDEHYDE
ACETALDEHYDE
METHANOL
METHANOL
ACETALDEHYDE
ACETALDEHYDE
METHANOL
FORMALDEHYDE
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
PAN
FORMALDEHYDE
ACETALDEHYOE
ACETALDEHYOil
BENZENE
TOLUENE
M-XYLENE
Z ,2 ,4-TRIMETHYLPENTA
NE
BENZENE
TOLUENE
M-XYLENE
2, 2,4-TRIMETHYLPENTA
NE
FORMALDEHYDE
BENZENE
TOLUENE
M-XYLENE
0-XYLENE
1 ,2,4-TRIMETHYLBENZE
NE
2, 2,4-TRIMETHYLPENTA
NE

BENZENE
TOLUENE
M-XYLENE
0-XYLENE
2,2 ,4-TRIMETHYLP.ENTA
NE
PROPYLENE
ETHYLENE
NO
NO
STATED
CONC
3.950
0 . 300
1 . /J0.CT
i . .0'.:::;
1 .000
0.266
1 .000
1 .000
0.600
1 . 000
0. 160
S3. 153
0. 191
0. 131
S3. 166
S3. 165
0. 164
0. 166
0. 169
0.207
1 . 000 •
1 . 000
1 . 000 -
0.500
1 . 000
2 . 000
2 . 005

0.500
1 . 00.GT
2 . 000
3.000

1 .000
1 .500
2.000
4.000
5.000
5.000

1 .000

S3 .000
1 .500
2.000
4.000
5.000
1 .000

14.400
1 9 . 000
52.400
52.600
ACTUAL
CONC
3.950
0.292
! . 000
:. . 00.J
0.970
0.264
1 . 000
1 .000
0.572
0.932
0. 160
0. 153
0. 191
0. 131
0. 166
0.165
0. 164
0.166
0. 169
0.207
0 . 9 b '•-
1 . 000
1 . 00.-J
0. 45X»
1 .005
1.99 ':!
1 . 1 L "•

0 . 5 a 5
1 . J6 •»
2. 105
2.349

1 .046
1 .659
2. 161
4.274
5.543
5-379

0.795

0.000
1 .613
2. 101
4.155
5.389
0.773

14 .645
19.480
52.450
52.600
              176

-------
                        OFFICIAL CALIBRATION SOURCES             25-S<5p-1985
                                                                 Page 9


ID                                SER                           STATED  ACTUAL
NUM     DESC        SDATE         NUM                            CONC    CONC

171  LOWNOAUTO                            NO                     0.000   0.28"
172  LOWNOAUT02                           NO                     0.693   0.665
                                     177

-------
CALANA Plots and Reports
           178

-------
    . 40]
 V    i
l2 0.20^;

"TO    h
                                             GC
Carle  I
PROPYLENE
FPMC/IN  Mttan =  2.30
                      24-J1
                                                        0.30
   a OQI i I  i I  i I  i I  ! I  i I  i i  i I i  ! -.  I !  ! i  I i  ! i  I i  t i  1 i i ft *,
    100 160 170 180 190 200 210 220 230 240 250 260 270 280 2SO 300 310 320'
                     Julian  Date,  1984
  For  the  species F'RDPYLENL  and the instrument.  Carle I
                               GC:
  Beg i nn i nq Cai date s
                       Day::
  E n d i n g     C a 1 d a t a:
                       Day:
  Upper CalFactor Limit
  Lower CalFactor Limit
30-MAY-84
150
16-NOV-S4
3 2 0
   0.40000
   0.OOOOO
  The number  of data points     62
  The number  included  was      62
  The slope of  this fit  is  1.3522654E-004
           with  a std dev of  6.6911662E-005
  The intercept is           1.66941S4E-001
           with  a std dev of  1.5341250E-002
  The std  dev of the fit is  1.8079919E-002
  The correlation coeff is
 O.25246
                  rel.  dev = 49.5 X

                  rel .  dev = 9. 2 7.
                  PE =  0.01205
  The min,  mean,  max are        0.13040,    0.19760,     0.25009
  The standard  deviation is   O.01S53   rel. dev  =9.4 X

  Cal =  0.00014  X  (Julian Date - 150)  +  O. 18723

                        ...press RETURN-key when  ready...
                               179

-------




£_
-»•>
IB
Li_

9 •
10
0



""-! i M 1 i 1 ' I i i ' 1 i I i 1 i 1 M I I i ! i ! i ! i
h Carle I GC
«.27L ISOPENTANE
i PFMC/IN At ten = 2.50
^»
i
3.21r- 04-AUG
h X
^ . J 	 	 _^, j» ' ^| **"S a»^"»'^~1*X^
" <- ^^r«^B[jjf * — ^^ ^\JU "***•
j_ T^8»
j
i
L
1
0.051-
(-
ft -J i ! i ! : i i ! i ! i I i I i ! i ! '. i i ! ! ! i I i ! i
!-i i ! j
— i
I
i
i
i
-"i
i
i
i
~ -- — 	 !
•
i
i
i
— i
]
"|

ill:!
• • -•-
«,27


a i< O
3"-1aj
•—
O.ieT1
Hi
ft
c*-
0.11 °
^
0.05

rt Ht1!
      160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 32&
                    Julian  Date, 1354
 For the  species ISOPENTANE and the instrument Carle  I    GC;

•E-ieginning  Caldate:         30-MAY-84
                      Day:   150
 Ending     Caldate:     '    16-NOV--S4
                      Day:   320
 Up p e r C alFactcsr L i m i t         0.. 3 2 00 0
 L!.:; w a r C a 1 F a c t o r L i m i t         0. 0 0 0 0 0

 The number of  data points     61
 The number included  was      61
 The slope  of  this fit  is  -6.274S75E-O05
         with  a stc! dev of  3.64SOO4OE-OO5    rel .  dev = -53.1  :
 The' i intercept  is           1. 7246965E-001
         with  a std dev of  S.4102O93E-OO3    rel.  dev = 4.9  A
 The std dev of the fit is  9.72SS922E-OO3  ,  PE =   0.00649
 The correlation coeff is   -0.21852

 The min, mean,  max are        O.I3965,    O.I5816,     0.1962O
 The standard  deviation is   0.009S9   rel. dev -  6

 Cal  = -O.OOOO6 X (Julian Date  -  150)  -i-  0.16306
                             180

-------



t-
o
-l->
k-
m
"n i i i i i i
i
0.25J-
r
i
i
i—
t
• *l
i—
1
! I 1 i i I ) i i
m
• I

Ai- ,,..
£V~JUL


I | i. | l | , i i ! i | i i i |- ,
Carle II GC
ETHANE
FFMC/IH At ten = 2.00

a
f 26-AUft
JVA f^\
** / • 	 ^

' ' ' !
1
J
I
-|
— i
i
i
!
~|
V • OV
0.25

0.20°
^
-flU
<-*-
. . . <-t
   h
                                                       J.05
     l I  i !  l ! i  I !  ! i  ! i  ! i  ! i  ! !  ! i  ! i  ! i  ! l  i l  ; !  ! i  ift ^^
     160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320
                   Julian  Date,  1984
For  the species ETHANE  and the instrument Carle  II   SC:
Beginning Caldate:
                      Day:
E nd ing     Ca1da t e:
                      Day;
Li p p e r C a 1 F a c t o r Li m i t
Lower C a 1 F a c: t o r L i m i t
                            30-MAY-B4
                            150
                            16-NOV-84
                            320
                               0.18000
                               O.OOOOO
The number  of  data points     35
The number  included  was      32
The slope of  this fit   is  5.6237375E-005
         with  a std dev  of  4.3453789E-005
The i n ter c ep t  i s           1.34691S8E-OO1
         with  a std dev  of  9.2956360E-O03
The std  dev of the fit  is  7.19S2245E-QQ3
The correlation coeff is    0.22995
                                              rel .  dev  =  77.3 7.

                                              rel .  dev  =  6.9 7.
                                              PE -  O.OG480
The min,  mean,  max are
The standard deviation  is
                               0„13445,     0.14661,    0.23050
                             0.00728   rel.  dev - 5.0 7.
Cal =  O.OG006 X (Julian  Date - 150) +   0.14313
                           181

-------

1.3 P
1.2 j-
^ < r~
£_ "'* j-
o im9 LT
+^ 0.9 '(-
CJ i —
OJ 0.3 {—
L1- 0.7 h-

rj, 0.5 gl^JA
0.4 r^
0.3 P
9.2 h-
0.1 r—
« * r\ \
: \ i | 1 | i i 1 | 1 | 1 j i | ! j i j 1 | ! j 1 1 1 j . 1 |
Carle II GC
PROPANE
FFMC/IN Atten = 2.00

Af
4j
/ir
»t!
- ^_^ ; jj
.' ~ ^s^"^ ~^~~^~~~~—*-—^. T i.
- -^J^*-^-=^^*=*- - - ~«- 	 ^» ^
₯^



i 1 ! ! 1 1 I 1 1 1 ! ! I 1 1 1 ! 1 ! ! ! ! 1 ! ! i i i ! i
yi.4
"3 i-3
-jl. 2
	 i 4 .1
-H
— i < ,!>

-10.3
—!0. 3
^1.7
— ^ J A ^
^ w. b
•q 0.5
— i 0.4
— i 0.3
j

-d a a




O
il>
•—
Tl
IU
C+-
~i




     160 170 180 190 200 210 220 230 240 250 260 270 260 230 300 310 32?
                   Julian  Data,  1984
For the  species PROF'ANE  and the instrument Carle  II   GC:

Beginning Caldate:

Ending     Caldate:
Day:
                     Day;
Up per  C a1F a c t or Limit
Lower  CalFactor Limit
30-MAY-84
150
16-NQV-84
320
   0.70000
   0.00000
The number  of data points     24
The number  included  was      19
The slope of  this fit   is  9.3254754E-005
         with  a std dev  of  3.484381&E-G04
The intercept is           5.B478932E-001
         with  a std dev  of  8.2456137E-002
The std  dev of the fit  is  6.1473938E-OG2
The correlation coeff is    0.06478
                        rel.  dev = 373.6

                        rel.  dev = 14.1 "
                        F'E  =   0.04O98
The min,  mean,  max are        0.42750,     0.60653,     1.O5930
The standard deviation  is   0.05987   rel.  dev = 9.9 "/.

Cal =  0.00009  X  (Julian Date - 150) +   0'. 59878
                           182

-------
-'- |_ I j I i I
i>4 E"
1.3 ]-
i.2 ' —
i.iP
^! H~
^"~ 1 A 1— —
o l-B tr
•** 0.3 h~
V t—
nj ».8 £•
^ 9.7 P
^^ 9.S r- ^^^

^ ' jpl-MAY
v » *r i~"
0.1 P
fl A I~! ! ! ! !
| 1 j 1 | 1 i i j I | i | 1 1 t ] 1 i I ; 1 i 1 j ;
Carle II GC
PROPYLENE
FrnC/Ii'i ntten = 2.50


*
A
,'!
''«
t Tl }
k ] -^ «i\^? _*•"' k*,S *^»« \ J! _ 'ii 	 ^
fc»^g -^ ••^•i yy ~§^^ — ~ *~= — - — *af^ }
•'•* * *


! ! ! ! i 1 ! ! ! ! ! ! ! ! i 1 ! 1 ! ! ! ! ! ! ! !
i ' — i
•H
— j
~>

•
— •
—1

H
—1
— i
H
H
i i~i
1.4
1.3
i.'^3
1 1 -
i .••>.
•••'
C.'3

0.8
e!s
«.fi

J'.3
O.i

                                                         tui

                   Julian   Date, 1984
For the  species PROPYLENE: and the  instrument. Carls II  GC;
B e g i n n i n g  C a 1 d a t. e 3
                     Day;
E n d i n g     C a 1 d a t e:
                     Day:
Upper CalFactor Limit
Lower CalFactor Limit
3O-l"!AY-34
150
16-NGV-84
320
   0.70000
   0.00000
The number  o-f  data points    63
The number  included  was     37
The slope o-f  this -fit  is 3. 2677424E-OQ4
        with  a std dev o-f 2.03396S5E-004
The intercept  is          4.8OBG472E-OOI
        with  a std dev of 4.5272312E-002
The std dev of the fit is 5.36267Q6E-OG2
The correlation coeff is   O.21172
                  rel.  dev = 62.2 ",

                  rel .  dev = 9.4 "/.
                  PE  =   0.03575
The min, mean,  max  are?
The standard  deviation i
   0.36388,    0.55264,     0.9756'
 0.05433   rel. dev  =  9.3 7.
Cal =  0.00033  X  (Julian Date - 150) +   0.52982

                      ...press RETURN-key when  ready...
                           183

-------
    I                        f^    t    ^* -^
    j-                       L-arie  il
   4-                       ETHYLENE


;;-
a^
v^
(v
o_
•—•
L?






i
i
a . 21?'' —
i
9.15; —
i
t- A^-jf^d
t — ^s-^r-*i
"'"'fci-MAY
i
J^j /Xgfv
~ * •
i-
fl .UJ i ! ! ! ! ! i !
FPHC/IN' Attan = 2.?0 i
i
rr-jj A '^i '*(
•j! ~|
ia-JUL » i\. • » — : a IR
» '^ -^ . ' \ ' ; 'j ' 	
T « • »^"^> * / > ) ( 1
— ! 9.1'i1
—i'
i
— - >
-------
v« ^U



A Tf,

S-
O
• %
T~
\J
iTJ
I j » • £ *
ID
O

0.19


A flrt
i i i ( j i j i | • j i j i 1 i | i i l | i j i i i i i ! i i , | .
i Carie III GC
i ISOPENTANE
i FPMC/IN At ten = 2.00
r~ ~
1
t
h 12-3EP
i
fljp */^__
[i 	 	 As-W^JMSjjW^ ,/*»'' v » -
r-


— —

_
1 1 ! 1 1 '! i 1 ! ! 1 i I I 1 ! 1 1 1 ! 1 1 ! ! 1 1 ! i ! J ! ! !
j w. +t»
i
]
1
i ii(-r.-5,

rw"o
j m
i i •
i
1 0.2'jZ1
1
-J

0.10


A eiti
  150 I6d 170 130 19U 20e 210 220 230 240 250 260 270 280 290 300 310 320
                   Julian  Data, 1984
For  the species ISOPENTANE and the instrument Carle  III  GC;

Beginning C a 1 d a t e:         3 O - M A Y -- 8 4
                      Day:   150
Ending     C a 1 d a t. e:         16 - N 0 V - S 4
                      Day:   32O
Up p er C a1F ac tar Limit         0„40000
Lower C a1F ac t or Li mi t         0.00O0O

The number of data paints
The  number included  was
The slope  of this fit   is  -1.222&56E-004
         with a std dev  of  5.6003271E-G05    rel .  dev = -45.3  7.
The  intercept is           2.2209272E-O01
         with a std dev  of  1.273873SE-O02    rel.  dev = 5.7  7.
The std  dev of the fit  is  1.6698649E-002    PE =   0.01113
The correlation coeff is   -0.27557

The min,  mean, max are        0.15986,    O.19468,     0.23O74
The standard deviation  is   0.01722   rel.  dev =  8.8 7.

Cal = -0.00012 X (Julian Date  -  150)  +  0.20375

                      ...press  RETURN-key when ready...
                              185

-------
      p- ;
i
r~
i
                         FFHC/IN   At tan =2.00
                                   T .A. * 1 1 —
                                   i MINC.
                                                      — A 4K
                                                       i v" *"w
                  ^
                           "*-*-*•*»-&*               ~. A <*  '
                           ^  -^Sr^-<:- — --.__   r-"ui
                                                           o
   i                                                    ••
   '.  I I  I I  I I  I I  I I  . I  I I  I !  I I  • I  i |  i I  ) i  i  1 I  I I  I !  I
A #141 - < - 1— 1 - 1— I - > - 1 - 1— I - 1— I - 1 - 1— I - 1— i - 1 - 1 - 1 - 1  I I - 1 - i_j - , - 1 - 1 - ; - 1 - 1 - 1 - 1 - 1
  150 1M 170 180 130 200 210 220 230 240 250 260 270 280 290 300 310 32
                   Julian  Date, 13S4
For  the species  N-BUTANE and  the instrument Carle  III  GC:

Beginning Caldate;         30-~lv!AY-84
                      Day:  ISO
Ending     Caldate:         16-NfJV-S4
                      Day:  320
U p p e r C a 11- a c t Q r  L. i m i t.         0. 2 0 0 0 0
L o w e r Ca1F actor  Limit         0.0O000

The  number of data  points     40
The  number included  was      40
Th e  slop e of this  f i t  is -1.50192 5E-004
         with a std  dev of 5.7705246E-005    rel.  dev  =  -3S.4 :
The  intercept is           1. 3697557E--001
         with a std  dev of 1. 24O1 192E-002    rel.  dev  =  9.1 7.
The std dev of the  fit is 1.145S466E-002    PE =  0.00764-
The correlation  coeff is  -0.38897

The min,  mean, max  are        0.OS147,    0.10504,    0.13079
The standard deviation is   0.0122S   rel.  dev = 11.7 7.

Cal = -O.OO015 X (Julian Date - 150)  +  0.11445
                             186

-------
   r
0.20r-
               i  l I  i i  i i i  i i  i i  i i  i i  I i i  i i  i
                            Carle  III  GC
                            1,3-BUTADIENE
                            FPMC/IN  At ten =  2.00
                                                . 20







O
01
3.15 —
"H
tu
f)
O
                                                     i
                 208 210 220 230 240 250 260 279 230 290 300 310 320'
                   Julian  Date, 1984
                                                     .0.00
For the  species 1,3-BUTADIENE  and the instrument  Carle III GC

B e q i n n i n g  C a 1 d a i: e:
                     Day:
E n d i n g     C a 1 cl a t s s
                     Day:
U p p e r C a 1 F a c t G r L. i m i t
Lower C a1Factor Limit
                     30-MAY-84
                     150
                     16-NDV-S4
                     320
                        0.. 2500O
                        O„OOOOO
The number  o-f  data points    38
The number  included  was     38
The slope of  this -fit  is -1.54757OE-O04
        with  a std dev o-f 5. 7985O70E-005
The intercept  is          1.4937310E-001
        with  a std dev o-f 1.2600050E-002
The std dev of the fit is 1.0673171E-OG2
The correlation coeff is  —0.40642
                                      rel .  dev = ~-37.b

                                      rel .  dev = 8.4 "/.
                                      PE  •-   O. 00712
The min, mean,  may, are       0.09043,     0.11606,     0.13717
The standard  deviation is  0.01152   rel. dev = 9.9  "/.
Cal
-0.00015 X (Julian Date
                                 150) H-  0. 12616
                           187

-------





iL.
O
+*
V
flj
LL.
fU
O






w"-vj ' 1 i I • i i i i
i
i
1
1
A {» —
•a, 1-1
I
i 	
;
j
I"1 — •% — -* _ ^>w_m_ -ft
/» < ftl 	 ! ~ "^ *>
• f
j
r
i
i
i
9.35|-
I

F •
., .J 1 ! ! ! ! 1 i i !
i 1 j ! i 1 | i i 1 i i { 1 j 1 i i 1
Carle III GC
1-BUTENE
FFHC/IN At ten = 2.00

28-AUG
*
A
; ^
V*»-» > '
i*TT\ 1 \ o
i---ii---f^%-x^..._
Y\r^* .\ ---
* "%






! i 1 i ! ' ! i ! ! ! ! i i ! i ! i 1
i 1 i ! ' j
1
J
~~i
i
i
— i
I

~j
j
i
j
—4
1

i
i
1
1
H
i
,:,!,!
V.L.V



•5.15





«.ie



•3.'?5



A rtrt

                                                        o
                  Julian  Date, 1984
For the  species 1-BUTENE and the. instrument Carle III GC:

Beginning  Caidats:         30—MAY—84
                     Day:  150
Ending     Caldate:         16--NOV-34
                     Day:  320
Upper Gal Factor Limit        0.20OOO
Lower CalFactor Limit        O.OOOOO

The number  of  data  paints    38
The number  included  was     38
The slope  of this fit  is -1.132031E-004
         with a std  dev of 5.9781O57E-005    rel.  dev = -50.6  "
The intercept  is          1.2497382E-001
         with a std  dev of 1.2990315E-002    rel.  dev = 10.4 "/.
The std  dev of the  fit is 1. 10O3754E--O02    PE =   0.00734 .
The correlation coeff is  -0.31299

The min, mean,  max  are       0.07366,     0.09953,     0.1299S
The standard deviation is  O.O1143   rel.  dev =  11.5 7.

Cal = -0.00012 X (Julian Date - 150) +  O.10724

                      ...press RETURN-key  when ready...
                            188

-------
c.\)





1 S
1 . w
O
i i : I i 1 i i i 1 i 1 i 1 i i ; j : ;
Carl
_
1 i i i • i i i 1 j 1
e III GC

METHANOL

FFMC/IN
7,.
"*--»,.

Atten = 2.00


"S r ^^ -^-«i

^ 
-------
 0.49J
 i ! ''  I i  i i  ! '  I 
-------
  9.3«r
  9.25f
o *-2«r
                   i  | i  i i  i i  , i i  i |  ! !  i |  i ;
                            Carle  III  GC
                            TOLUENE
                            PPMC/IN   Atten =  2.50
                                                        A ia

                                                      -! $ 15  '
                                                       i Uli3lD
     	i I  i !  ! !  i ! i  i i  ! !  ! I  ! i  ! I  '
    ?£d  160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320'
                    Julian  Data, 1984
For the  species TOLUENE and  the  instrument. Carle  III  GCs

B e a i n n i n g  C a 1 d a t e s

Ending     Caldata:
                     Days
                     Day:
Upper Cal Factor Limit
L a w e r C a 1 F a c t o r L i m i t
                             3 0 - M A Y - 8 4
                             150
                             16-NOV-34
                             320
                                0.30000
                                0 . 0 0 O 0 0
The n Limber  of  data points     25
The number  included  was      25
The slope of  this fit  is 9. 45263OSE-007
        with  a std dev of 7. 34S4704E-005
The intercept  is          1 . 3672349E-001
        with  a std dev of 1 . 65254 08E--OO2
The std dev of the fit is 1 . 39802 13E-002
The correlation coeff is    0.00268
                                               rel .  dev  = 7774.0

                                               rel.  dev  = 12.1 %
                                               PE  =  0.00932
  The min, mean,  max  are       0.11350,     0.13693,     0.16721
  The standard  deviation is  0.01369    rel.  dev =  10.0  7.

  Cal =  0.00000  X  (Julian Date -  150)  +  0.13687

                        ... press 'RETURN-key when ready...
                            191

-------
VltlV
0.15
O
U
• •
iTj
0
i ( 1 | 1 | ! | 1 | I | 1 | 1 | ! ! 1 i 1 i i i 1 | ! i 1 ! ! i ! |
Carie III GC \
BENZENE i
FFMC/IN Attan =2.00 !
~i
i
_ _j
i
1
i
i
•A. 15
O
01
-•"HI
b
~5
    -
  ,\/fcl_ i	I ___i _t_. ^		   			 __  ^ ^ _ . .  .	^ ^1
  fls0 160 170 180  190 200 210 220 230 240 250 260 270 260 29* 300 3l0 3£%'V
                    Julian  Data,  1984
For  the species  BENZENE and  the instrument. Carle  III  GC:
B e g i n n i n g C a 1 d a t e:
                      Day:
Endi ng    Caldate:
                      Day:
Li p p e i" Cal F a c t c r L i m i t
L o w e r C a 1 F a c t o r • Li rn i t
30-MAY-84
150
16-NQV-84
320
   0.20000
   0.00000
The  number of data  points      7
The  number included  was       7
The  slope of this fit  is 1.05&4688E-O05
         with a std  dev of 9. 76I59602E-O05
The  intercept is           8.7323S74E-O02
         with a std  dev of 2.150454IE-002
The  std  dev of the  fit is 1.OOS1S72E-OQ2
The  correlation coeff is    O,04832
                  ral .  dev .-- 924."

                  rel.  dev = 24.6
                  PE  =   0.00672
The min,  mean, max  are       0.08022,    0.08961,    O
The standard deviation is  0.00921    rel. dev  = 10.3 7.

Cal =   0.00001 X  (Julian Date -  150)  H-  0.08891

                       ...press RETURN-key when  ready...
                            192

-------
 «.e«r
                    i j  i
 i  , i !  i i  i ;  i I  ; i  i
 Carle III  GC
 N-PENTANE
 PFMC/IN   Attan =  2.00
                                                    I  !
                           •>>.S3
                       i
                                                      0.50
                                                         Hi
                                                       •3»J
                                                         <-*-


                                                         '-}
  ttfli ; i  i i  i i  i ;  ; i  ; i  i i  i i !  i I  ! I  i !  ! !  ! i !  I I  ! !  M ft
  'io0 160 170 180 190 200 210 220 220 240 250 260 270 280 290 300 310 320'
                   Julian  Date,  1984
For  the species N-PENTANE  and the instrument  Carle III GC:
Beginning Caldate:
                     Day:
Ending     Caldate:
                     Day:
Upper CaiFactor Limit
Lower CalFactor Limit
30-I1AY-B4
150
16-NDV-84
320
   0» & 0 0 0 0
   0.OOOOO
The number  o-f  data points     45
The number  included  was      45
The slope o-f  this -fit  is -2.873216E-O04
         with  a std dev of 1.0329445E-004
The intercept  is          3.5367520E-001
         with  a std dev of 2.372S269E-O02
The std  dev of the fit is 2.63080S6E-O02
The correlation coeff is  -0.39051
                  rel„  dev — —36.O

                  rel .  dev = 6.7 7.
                  F'E =   O. 01754
The rnin,  mean,  max are       0.22760,     0.28S5S,     0.34680
The standard  deviation is  0.02325   rel. dev = 9.S  "/.

Cal ~ -0.00029  X (Julian Date -  ISO)  •+•  0.31053

                      ...press RETURN-key when ready...
                           193

-------
                    i 1  i
! !  I
I  I i i   i  I I  I i  I
                            Carle  III  GC

0.20


t-
o
tl °-15
(J

1 1
'— « w.iw
_ -v _ _ f^. •A^ y%"^^*
j^__— ^ •>^1% — Wv^/^r^v
• •ft







• 1 r 1 i ! ; ! i ! ! 1 ) I ! i r ! I i I ! ! !

= 2.00 -JO. 20


~"

—

ito-GCT
n — 1
f
{
N^lT-- 	
^^4



!
H
!
! ! I ! i i '. ! I !


0
Ul
0.15 —

~n
Ul
- ,-rt
tf.iw <-<-
0
~s


0,05



ti l^iA
     160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310
                   Julian   Date, 1984
For  the  species CIS-2-BUTENE  and  the instrument Carle  III BC:
Beginning  Caldate:
                     Day:
Ending     Caldate:
                     Day:
Upper CalFactor Limit
Lowe r C a IF a c: t o r Li m i i:

The number of data points
The number included  was
  30-MAY-S4
  150
  16-NOV--84
  320
     0.35000
     0.00000

     45
     45
The slope  of  this -fit  is -2. 1BB447E-Q04
         with  a std dev o-f 4. S220399E-005
The intercept is          1.5340966E-001
         with  a std dev of 1.03059SOE-OO2
The std  dev of the fit is 1.2219133E-OQ2
The correlation coeff is  -0.56910
                   rel .  dev = -22. 0 7.

                   rel .  dev = 7.O 7.
                   F'E  =   0.00815
The min, mean,  max  are       O.O7010,     O.105O7,    0.12707
The standard  deviation is  0.01469    rel.  dev = 14.0 7.

Cal = -0,00022  X  (Julian Date -  150)  +   0.12058

                      ...press RETURN-key when ready...
                           194

-------
  0.25r

«J
                             Carle III  GC
                             TRANS-2-BUTENE
0.15
0.05
Oi OlOl
— PPMC/IN Attan =2.00 H
i
— -H
i
— -1
1
J
1
1
1
H
< 1 I 1 l ! I 1 < i i 1 ! ! i ! i ! i 1 i ! ! I ! ! ! ! i ! ! ! t i
0.20
O
111
0.15 —
UJ
b
0.05
rt.ftw
      160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320'
                    Julian  Date, 1984
  For the  species  TRANS-2-BUTENE and the instrument  Carle  III  GC:
  Beginning  Caldate:
                       Day:
  Ending     Caldate:
                       Day:
  Upper C a1F actor  Limit
  L o w e r C a 1 F a c t or-  L i m i t
30-!VIAY~84
150
16-NQV-84
320
   0.30000
   0.OOOOO
  The number of  data  points    bO
  The number included  was     50
  The slope of this fit  is -1.273S05E-004
          with a std  dev of 4.6032130E-005
  The intercept  is           1.4263S30E-001
          with a std  dev of 1.0106748E-OO2
  The std dev of  the  fit is 9.9038982E-003
  The correlation coeff is  -0.37O92
                 rel. dev  =  —36.1

                 rel . dev  =  7. 1  7.
                 PE  =  0.00660
  The min, mean, max  are        0.09145,    O.I 1494,    O.I3217
  The standard deviation  is  0.01056   rel. dev = 9.2 7.

  Cal = -O.OOO13 X  (Julian  Date - 150)  +  0.12353

                        ...press RETURN-key when ready...
                            195

-------




£_
O
flj
|,

OJ
o
Wl/0j i i ! I ; i i i
(—
i
0.66f~
h
i
0.50r~
h
0.40f-
f—
i
w.3wi ^^^
i

i ! ; * i > i ; : '• i i i
Carl
0-XY
FFMC/IN





•A __^^~^


i i i s i ; i i . i
e III GC
LENE
Atten = 2.30





"*~ 	
"^«

1 i ' j
~i
»-H
i
	 1
i
— i
-J
i
!
i
— i
1
— 1
t

                                                      i
     ! !  1
1
                          t
       1 1
i i

                                                     i
                                                           ~n
                                                           OJ

     160 170 180 190 200 210 220 230 240 250 260 270 280 290 3'<)0 310 32$'
                   Julian  Data, 1984
For  the species  0-XYLENE and  the instrument.  Carle III  GC:

Beginning Caldate:

End ing     Caldate:
Day:
                      Day:
L) p p a r  C a 1 F a c t o r Limit
Lciwer  Cal Factor Limit
3O-MAY-84
150
16-NOV-84
320
   0.. 70000
   O,. 00000
The number of data  points     7
The number included  was      7
The slope o-F this fit  is 1.3699588E-004
         with a std  dev of 2.3682240E-004
The intercept is           2.5194653E-001
         with a std  dev of 5.121S12SE-OO2
The std  dev of the  fit is 2.5398701E-OO2
The correlation coeff is   0.25046
                         r e 1 .  dev -- 172. 9

                         rel .  dev = 20.3 7.
                         F'E  =   0.01727
The min,  mean, max  are
          0.25603,
                0.231O3,
         0. 317
The standard deviation is  0.02442   rel. dev  ~ 8.7 7.

Cal -   0.00014 X  (Julian Date -  150)  +  0.27250
                            196

-------





f
o
+>
CJ
nj
1 1
LJ_

i—4
01
U»






j
h-
i
I
A.50J —
s
l~
i
3.43J-
i
i_
i
a. 3a|-

.
i
3.20J-
i

0.10f—
1
i
lA flAl — ! — 1
j ; j i ; i j i j i i i i i i ; ] i | i • j i i i i i i i i i ;
CEA Formal deny cie
FORMALDEHYDE
FPMC/IN At ten = i.00


1
(\ _

|.\ ^' ^
	 , ' \ ^***^^^~^. J* \

? !^ ^ \
\
f
. *



t ! j 1 1 i 1 ! 1 1 i ! •! ! 1 1 1 1 ! ! 1 1 ! i 1 1 1 ! 1 ! I
i i
— i
i

i
i
—i
i
— ^
i
i
i
i
i
i
i
i
1
i
~1



3.51?


.-"•
0.40^
j—

9.30JT

j^i
^r*-
0.20^


S.10


A Oil*
  <5d 16» 170 180 199 200 210 220 230 240 250 260 270 280 290 300 310 320'
                   Julian  Date, 19S4
Far  the species  FORMALDEHYDE  and the instrument CEA  Formal d eh yd,;-

Beginning Caldate:         30-MAY--S4
                      Day:  150
Ending     Caldate;:         16--NQV--84
                      Day:  32O
Upper-  Ca 1 Fractor'  Li mi t         0.6000O
Lower  C a 1 F a c: t o r  L. i m i i:         0. 0 000 0

The  number o-f data  points     17
'T h e  n u m b e r i n c 1 u dad  w a s      17
The  slope o-f this  fit  is -2.2B2711E-004   •
         with a std  dev of 4. 7236958E-004    rel .  clev  =  -2O6.9 "/.
The  intercept is           3. 1092736E--001
         with a std  dev o-f 1. 132S004E-001    rel.  dev  =  36-. 4 7.
The  std dev of the  fit is 7.9710921E-002    PE  =  0.05314        N
The  correlation coe-ff is  -0.12381

The  min,  mean, max  are        0.12160,     0.25699,     0.39930
The  standard deviation is  0.07773   rel.  dev  = 30.3 7.

Cal  = -0.00023 X (Julian Date -  150)  +  0.27669

                       ...press RETURN-key when  ready...
                            197

-------