SEPA
            United States      Office of Air Quality        EPA-450/4-90-008a
            Environmental Protection  Planning and Standards      May 1990
            Agency         Research Triangle Park NC 27711
            Air
               IMPROVE

           PROGRESS REPORT
               APPENDIX A

-------
                                EPA-450/4-90-008a
       IMPROVE
PROGRESS REPORT
     APPENDIX A
              By
          Marc Pitchford
Environmental Monitoring Systems Laboratory
  U. S. Environmental Protection Agency
        Las Vegas, NV 93478

             And

          David Joseph
         Air Quality Office
        National Park Service
         Denver, CO 80228
       U S EPTlronosntal Protection Agency
       p—iTi 5 I io".ry (£"FL-lo)
       Rue,lOn -J I J-j-.J-'-^J
       230 S. Dearborn Street. Room 1670
       Cliicago, IL  60604
Office Of Air Quality Planning And Standards
      Office Of Air And Radiation
  U. S. Environmental Protection Agency
    Research Triangle Park, NC  27711 '

            Mav 1990

-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U. S. Environmental
Protection Agency, and has been approved for publication. Any mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use.
                                     EPA-450/4-90-008a

-------
Standard Operating Procedures for

              IMPROVE

  Particulate Monitoring Network
   Analytical Services Division
    Crocker Nuclear Laboratory
  University of California,  Davis
         .Davis,  CA 95616

             July 1989

-------

-------
IMPROVE STANDARD OPERATING PROCEDURES                            JULY 1989
                                  Table of Contents
  I.  INTRODUCTION	4

      A.  Role of Procedures in the IMPROVE system	4
      B.  The IMPROVE Sampler	6
      C.  Responsibilities for Maintaining the Procedures	7
      D.  Overview of the Standard Operating Procedure Document	8

 II.  SAMPLE HANDLING BEFORE SHIPMENT TO THE SITE	10

      A.  Preparing the Clean Filters	11

          1.  Module A fine Teflon filters	11
          2.  Module B fine nylon filters	14
          3.  Module C fine quartz filters	14
          4.  Module D PM10 Teflon filters	14

      B.  Loading the Cassettes	14
      C.  Verifying the Loading Procedures and Leak-testing the Cassettes  .  .  18
      D.  Prepare the Blue Box for Shipment	  19

III.  SAMPLE CHANGING AT THE SITE	20

      A.  Preparing for the Weekly Sample Change	21
      B.  Removing the Exposed Cassettes	21
      C.  Inserting the Clean Cassettes	22
 IV.  SAMPLE HANDLING AFTER SHIPMENT FRCM THE SITE
23
      A.  Receiving the Blue Box	24
      B.  Reviewing the Field Logsheet	24
      C.  Unloading and Cleaning the Cassettes  	 26

          1.  Module A fine Teflon filters	26
          2.  Module B fine nylon filters 	 26
          3.  Module C fine quartz filters	27
          4.  Module D PM10 Teflon filters	27
          5.  IMPROVE Dot Chart	27

      D.  Checking and Entering the Field Data	28

          1.  Review the Field Logsheet for Quality Assurance	28
          2.  Entering the field data into the sample-handling database ... 28

      E.  Processing and Shipping the Exposed Filters	29

          1.  Module A fine Teflon filters	29
          2.  Module B fine nylon filters	30
          3.  Module C fine quartz filters	30
          4.  Module D PM10 Teflon filters  	 31

-------
  V.  SAMPLE ANALYSIS	32

      A.  Gravimetric Mass Analysis	32
      B.  Laser Integrating Plate Analysis (LIPM)	33
      C.  PIXE/PESA Analysis	36
      D.  Ion Oircmatograph. Analysis (1C)	37

          1.  Receipt of the filter	37
          2.  Filter extraction	,	38
          3.  Ion analysis	38
          4.  Data transfer to UCD	38

      E.  Thermal Optical Reflectance Analysis (TOR)  .,	39

 VI.  PROCEDURES FOR DATA PROCESSING	41

      A.  Introduction:  The Equations of Concentration and Uncertainty  ...  41

          1.  Volume	41
          2.  Gravimetric Mass	41
          3.  optical Absorption	42
          4.  PIXE Analysis	42
          5.  PESA Analysis	42
          6.  Ion Analysis	42
          7.  Carbon Analysis	42

      B.  Entering the Data into the Concentration Database	43
      C.  Validating the Data	44
      D.  Preparing Magnetic Tapes and Floppy Disks	44
      E.  Preparing the Seasonal Summaries	44

VII.  PROCEDURES FOR SAMPLER MAINTENANCE	45

      A.  Evaluating Sampler Modifications	45
      B.  Calibrating the Flow Audit Device	45
      C.  Preparation for Annual Site Visit	46
      D.  Annual Site Visit	46

-------
                       Standard Operating Procedures for
                                    IMPROVE
                        Particulate Monitoring Network
                                   July 1989


                          List of Tables and Figures



Table 1.  Filter used in standard IMPROVE sampler	7

Table 2.  Filter identification code for Teflon filters 	 13

Table 3.  Sample identification code and color code for all filters .... 15
Figure l.  Schematic of IMPROVE Sampler
Figure 2.  Flow diagram for the procedures for sample handling before
           the shipment of the clean filters to the sampling site.
           The starred procedures (*) are done by the carbon contractor . . 10

Figure 3.  Flow diagram for sample changing at the site	20

Figure 4.  Flow diagram for the procedures for sample handling after
           the receipt of the exposed filters from the sampling site ...  23

Figure 5.  Schematic of the LIPM system	33

-------

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
     I.   INTRODUCTION


         A.   Role of Procedures in the IMPROVE system

             The nature and extent of the particulate monitoring program based on the
             IMPROVE sampler require carefully designed operating procedures.  The
             procedures have developed out of our experience in operating monitoring
             networks since 1973.   Some of the major factors influencing the
             procedures are as follows.

             1.   There are presently 48 sites using the IMPROVE sampler (20 IMPROVE,
                 19 NFS, 7 NESCAUM, 2 Lake Tahoe),  requiring the processing of over
                 500 filters each week.  The large  flow of samples requires
                 sample-handling procedures in the  laboratory to be as efficient as
                 possible.  To handle this workload, the procedures in the laboratory
                 have been separated into well-defined tasks.

             2.   The system uses multiple filter media, with a given filter following
                 one of 6 different pathways.  It most of these pathways,  it is
                 necessary to identify the clean filter as well as the exposed
                 filter.  This complexity requires  a sample-identification system
                 that is simple and reliable.  The  procedures must include numerous
                 cross-checks, but cannot be allowed to become cumbersome.  The
                 procedures must incorporate a good inventory system so that all
                 filters can be accounted for at any time.

             3.   Most of the sites are located in pristine regions where the ambient
                 concentrations of fine particles are extremely low,  often 1 to 2
                 orders of magnitude below urban levels.   The low loadings require
                 procedures that minimize and monitor any contamination of the
                 samples.  For example, every box of clean filters is given a date on
                 which the filters are to be installed in the sampler, in order to
                 avoid filters remaining at the sites too long and picking up extra
                 contamination.  In addition, the filters remain in sampling
                 cassettes from Che time they leave the laboratory until the time
                 they return.

             4.   Since the samples are collected twice a week,  it is  essential that
                 the downtime of the samplers be kept to a minimum.   This  requires
                 procedures that ensure a steady flow of clean filters to  the site.
                 The procedures must also include internal checks of  the field
                 information to identify problems with the sampler and with sample
                 collection.  It is essential that  the problems be identified and
                 rectified rapidly.

             5.   The particulate data are to be used in various source-receptor
                 models to determine causes of visibility impairment.   The data must
                 be able to stand legal scrutiny.   The procedures must include
                 sufficient quality assurance checks to provide the most accurate and
                 dependable results possible.
                                       Page 4

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             6.   The changing of filters at the sampling site is handled by persons
                 who are involved in the program less  than an hour a week.
                 Therefore,  the field procedures must  be simple and self-evident.  By
                 making the  normal routine as simple as  possible,  the field operators
                 can concentrate on abnormalities.
                                       Page 5

-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
         B.   The IMPROVE Sampler

             The IMPROVE sampler  consists of four independent filter modules  and a
             common controller, as  shown in Figure 1.  Each module has  its  own inlet,
             PM2.5 or PM10 sizing device, flow rate measurement system,  flow
             controller, and pump.   The  4 pumps are housed in a separate unit to
             isolate their vibration from the rest of the sampler.
                                                                PM10 teflon
                                                                       sunshleld
                                        wires
                                        and
                                        vacuum hoses
                                        in conduit
             Figure 1.  Schematic of  IMPROVE Sampler
                                                                                    power
                                        Page  6

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             Each module in the standard sampler has two filter cassettes, with the
             first cassette collecting samples on Wednesday and the second on
             Saturday.   Each filter cassette is color-coded,  with one set of colors
             referring to the module (red,  yellow, green,  blue) and the second set to
             the first or second cassette within the module (black, white).   Table 1
             lists the major data for each module, including the filter media, the
             color codes,  and the analytical techniques.
             Table 1.  Filters used in standard IMPROVE sampler.

             module   particle     filter       color  .       analytical methods
                        size      material      code

               A       PM2.5       Teflon       red         mass,  LIPM,  PIXE, PESA
               B       PM2.5       nylon        yellow      1C
               C       PM2.5       quartz       green       TOR
               D       PM10        Teflon       blue        mass
             The filters are transported to and from the site in sealed cassettes
             using specially designed shipping containers.   The field operator
             removes the cassettes with exposed filters and inserts the new cassettes
             without directly handling the filters.   With this system it is possible
             to have multiple filters in a single cassette.   The C module has two
             quartz filters in tandem,  with the second filter to monitor the artifact
             caused by adsorption of organic gases;   this second filter is not
             routinely analyzed.   For NFS congressionally mandated sites, the D
             module has an impregnated filter following Che Teflon filter in order to
             measure sulfur dioxide gas.

         C.   Responsibilities for Maintaining the Procedures

             The overall responsibility for the procedures  belongs to the Project
             Manager.   He is responsible for appropriate coordination between
             personnel to maintain the procedures.   As a senior scientist, the
             Project Manager is responsible for characterizing and calculating the
             various artifacts and uncertainties in  the collection and analysis.
             This includes proposing QA tests using  the Davis Field Station.

             The Quality Assurance Manager is responsible for the details of the
             procedures.  He works closely with the  Field Support Manager to verify
             that the proper procedures are being followed in the. sample-handling
             laboratory and in the field.   He is responsible for validating the
             results of the various UCD analytical systems  arid the final data,
             including the data from the external contractors.
                                       Page 7

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             The Field Support Manager is responsible for the procedures in the
             sample-handling laboratory and for procedures in the field.  Included in
             the laboratory procedures are the procedures for measuring mass and
             absorption.   He coordinates with the external contractors to maintain a
             proper flow of samples and data.  He coordinates with the Field Engineer
             to correct sampler problems immediately and oversees the annual
             maintenance schedule.

             The Field Support Manager is also the sample-handling laboratory
             supervisor.   Assisting him are three full-time laboratory technicians
             and several student technicians.

             The maintenance of the samplers is the responsibility of the Field
             Engineer and the Field Technician.  This responsibility includes the
            •annual audit and routine maintenance.  These persons are also
             responsible for conducting QA tests at the Davis Field Station,
             including tests to consider the feasibility of sampler modifications.

             The PIXE/PESA Manager is responsible for the procedures to be followed
             in the analysis of the samples by PIXE and PESA.

             The Support Group is responsible for developing and maintaining the
             computer hardware and software used in the procedures.   The software
             includes several programs for sample handling, programs for the PIXE and
             PESA analytical methods, programs for data reduction and validation, and
             programs for data presentation and transmission.  The Head of the CNL
             Computer Support Division is the Computer Support Manager.   The other
             members of the group are the Quality Assurance Manager and the PIXE/PESA
             Manager.  All three members of the Computer Support Group have
             considerable experience in aerosol analysis.  The group coordinates with
             three additional staff members who develop and maintain software
             relating to operating procedures.

         D.   Overview of the Standard Operating Procedure Document

             The procedures are separated into 6 segments.   The first 3  segments
             concern sample handling:  before shipment to the site,  at the site, and
             after return from the  site.   These procedures  are designed to maintain a
             smooth flow of samples,  assure good inventory  control,  and minimize
             contamination.  The fourth segment treats the  several analytical
             techniques used in the program.   The purposes  of the analytical
             procedures are primarily to maintain the highest quality analyses and
             secondarily to maintain efficiency.   The fifth segment  examines the
             procedures to produce  the concentration database from the analytical
             results, to  validate the data,  and to generate the seasonal data
             summaries.   This section includes procedures to compare all the data in
             order to detect possible errors.The final segment deals with procedures
             for routine  and unscheduled maintenance of the samplers in  the field.
             At the beginning of each of these segments there is a flow  diagram or a
             summary of the procedures of the segment.
                                       Page  8

-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
             Supplemental information is included in appendices to these segments.
             Appendix 1 lists the various forms used in the three sample-handling
             segments.   Appendix 2 describes the IMPROVE sampler and the field
             procedures in detail.  Appendices 3 to 7 give more, detailed operating
             procedures of the various analytical procedures.   Appendix 8 provides
             additional lists for the annual inspection and maintenance of the
             samplers,  and forms for flow audits.
                                       Page 9

-------
IMPROVE STANDARD OPERATING PROCEDURES
                                                               JULY 1989
    II.
SAMPLE HANDLING BEFORE SHIPMENT  TO THE SITE

The handling procedures  in this  section cover all actions from  the  purchase
of the flltJrS up to  the shipment of the clean filters to the sampling site
Figure 2 gives the  flow  diagram  for this portion of the procedures.   UCD is
responsible for all actions in this section except for the purchase and
pre-firing of the quartz filters, the responsibility of the  carbon
contractor.
                    A
                fine Teflon
                            B
                        fine nylon
    C
fine quartz
    D
total Teflon
purchase
filters
i
'
measure
pre-LIPM
i
F
measure
pre-MASS
i
<

purchase
filters



purchase
filters
1

pre-fire
filters*
i
f
send to
UCD*
,
>

purchase
filters


measure
pre-MASS
i
•
load clean filters into cassettes
attach identification fags to cassettes
                                        leak-test all cassettes
                                 send cassettes with clean filters to site
          Figure 2.  Flow  diagram for the procedures for sample handling  before the
          shipment of  the  clean filters to the sampling site.  The starred procedures
          (*) are done by  the  carbon contractor.
                                         Page 10

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         A.   Preparing the Clean Filters

             This section includes the procedures  for preparing the four types of
             clean filters used at the IMPROVE sites:   fine Teflon, fine nylon, fine
             quartz,  and PM10 Teflon.   Because the procedures  differ significantly
             between the types,  each type will be  treated individually.   This section
             includes procedures from two analytical methods:   gravimetric mass (MAS)
             and the Laser Integrating Plate Method (LIPM).  The procedures are
             discussed in detail in sections V-A and V-B.

             1.   Module A fine Teflon filters

                 a.   Purchase Teflon filters (Gelman Teflo)  from commercial vendor.

                 b.   Prepare collection masks.

                     The collection mask fits underneath the fine Teflon filter in
                     the cassette and reduces the  area of collection.   The primary
                     purpose is  to improve the sensitivity of  the PIXE and PESA
                     analyses by concentrating the particles.   The mask also reduces
                     the mass artifact by isolating the filter from the 0-ring.  The
                     decrease in the area of collection is  limited by filter
                     clogging;  in regions of high concentrations the flow rate can
                     be reduced below the acceptable range.

                     The collection masks are prepared from the inert paper spacers
                     that Nuclepore Corporation provides between Nuclepore 47 mm 8 urn
                     polycarbonate filters.   The spacers have  a very light coating of
                     Apiezon-L grease.   Many years of experience have shown that this
                     paper will  not transfer mass  to the filters.   The spacers are
                     retained in their original factory containers when 47 mm
                     Nuclepore filters are processed for other studies.   These
                     containers  are labelled and sealed until  required as stock for
                     preparing the collection masks.

                     The masks are prepared using  specially machined double-action
                     cutter punches.   The cutter punches simultaneously cut the 25 mm
                     outer diameter and the desired collection diameter.   Both punch
                     and cutter  are concentric,  centering the  collection area in the
                     25 mm mask.  It is imperative that the cutter punch be handled
                     gently and the heat-treated and ground edges be protected.  The
                     edges are very sharp and can  cause injury.   If the  cutter punch
                     is dropped or not correctly used,  damage  to the cutter punch
                     will result and it will have  to be remachined.

                     The procedures for preparing  the  collection masks are as
                     follows.

                     i.  Obtain a sealed and marked container  of 100 Nuclepore
                         spacers from the laboratory supply.
                                       Page  11

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                    ii.  Obtain the correct punch for the desired collection area,
                         generally 2.2 cm .

                   iii.  Obtain a thick pad of clean, pure Teflon material from the
                         laboratory supply retained for this purpose.  This pad,
                         approximately 1/2-inch thick, provides a cutting surface
                         that will not dull or change the geometry of the cutter
                         punch faces.

                    iv.  Place the Teflon material on the base of the mandrel press.
                         Open the mandrel press enough to accommodate the cutter
                         punch assembly when resting on the Teflon.  Initially, place
                         no more than 5 spacer papers on the Teflon pad and center
                         them under the cutter punch.  Gently, but firmly, apply
                         pressure to the assemby with the press arm until a distinct
                         cutting sound is heard and felt.  Open the press and observe
                         that 25 mm masks with the desired hole have been produced.
                         Remove the completed masks and put them in a clean
                         container.  Discard the excess spacer material.

                     v.  Verify that the mask geometry is correct.  Using a vernier
                         inside/outside caliper from the laboratory supervisor,
                         measure the inside diameter of the fabricated masks.   Make
                         several measurements and determine the mean.  For a 2.2
                         cm^collection area,  the diameter should be 16.73 + 0.01 mm.
                         (The collection area of the actual particles on the filters
                         has been found to equal this measured area.)

                    vi.  With practice and experience, up to 10 masks can be produced
                         at a time.  If incomplete cutting occurs, reduce the number
                         of pieces of stock in the press back to 5.

                   vii.  When all 100  spacers in the original supply have been
                         converted to  masks,  seal the masks in their container.  Wrap
                         the cutter punch with tape and return it co che protective
                         foam box.  Return the Teflon pad to its storage.
                 c.   Process a group of 50 filters through the LIPM station.

                     At this station the filters are given filter identification
                     numbers and the precollection LIPM values (pre-LIPM)  are
                     measured.   The filter identification number is used for  the
                     Teflon filters (modules A and D)  to associate the precollection
                     values of LIPM and mass with a given filter.   The filter
                     identification codes are given in Table 2.   The filter
                     identification system is used in addition to the sample
                     identification system that specifies the sampling site and the
                     date of collection.   The correspondence between the two
                     numbering systems  is made after the exposed filter returns from
                     the site.
                                       Page  12

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989



                     Table 2.  Filter identification code for Teflon filters,

                                  module    collection area      code

                     fine Teflon    A          1.1 cm;?         TFnnnn.M2
                     fine Teflon    A          2.2 cm2,         TFnnnn.Ml
                     fine Teflon    A          3.8 cm2         TFnnnn.MO
                     PM10 Teflon    D          3.8 cm2         TFnnnn.U
                     i.   Prepare 50 Filter Identification Tags (red Avery dot labels)
                         with the next sequential filter identification number.
                         (Check last assigned number on the A-Preweight Logsheet and
                         increment by one.)  Obtain a petri box of 5o petri dishes and
                         place the Filter Identification Tags on the petri dishes.
                         Record the numbers  on the A-Preweight Logsheet.

                    ii.   Insert clean Teflon filters from the supply into the 50
                         Quickie Mounts of the LIPM system.

                   iii.   Measure the 50 pre-LIPM values following the procedures
                         described in section V-B.   (The pre-LIPM value is the
                         intensity of light  transmitted through the clean filter.)
                         Record each pre-LIPM value alongside the filter
                         identification number on the A-Preweight Logsheet.

                    iv.   Place each filter in its proper petri dish.   Place the petri
                         dish back in the petri box and label the box.
                 d.   Measure the tare masses of 50 filters (pre-MASS).

                     The filters are weighed in batches  of 50 filters.   Within this
                     batch the work is separated into  groups  of approximately six
                     filters.   Between each group of 6 filters,  the calibration of
                     the balance is checked.   If necessary,  the balance is
                     recalibrated.

                     Remove 6  petri dishes from the appropriate petri box and measure
                     the pre-MASS values  of the filters  following the procedures
                     described in section V-A.   Record the pre-MASS value of each
                     filter alongside the proper filter  identification number on the
                     A-Prew'eight Logsheet.   Return each  filter to its petri dish
                     after weighing and return the petri dish to the petri box.

                 &.   For each  of the 50 entries on the A-Preweight Logsheet, enter
                     the filter identification number,  the pre-MASS value, and the
                     pre-LIPM  value in the sample-handling database.
                                       Page 13

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             2.   Module B fine nylon filters

                 Purchase nylon filters (Gelman Nylasorb) from commercial vendor.  No
                 further processing is necessary.

             3.   Module C fine quartz filters

                 a.  (carbon contractor) Purchase Pallflex 2500QAT-UP quartz filters.

                 b.  (carbon contractor) Pre-fire the quartz filters at 900° for at
                     least 4 hours, following operating procedures developed by the
                     contractor.

                 c.  (carbon contractor) Ship the carbon filters to UCD.  These are
                     identified by lot number.

                 d.  (UCD) Receive the prefired quartz filters and store in the
                     freezer until required for the loading sequence.
             4.   Module D PM10 Teflon filters

                 a.   Purchase Teflon filters (Gelman Teflo) from commercial vendor.

                 b.   Prepare 50 Filter Identification Tags (blue Avery dot labels)
                     with the next sequential filter identification number.  (Check
                     last assigned number on D-Preweight Logsheet and increment by
                     one.) Get a petri box of 50 clean petri dishes and place the
                     Filter Identification, Tags on the petri dishes.  Record the
                     numbers on the D-Preweight Logsheet.

                 c.   Measure the 50 pre-MASS values following the procedures
                     described in section V-A.   (The pre-MASS value is the tare mass
                     of the filter.) Record each pre-MASS  value alongside the filter
                     identification number on the D-Preweight Logsheet.   Place each
                     filter in its proper petri dish after the weighing and return
                     the petri dish to the petri box.   Label the petri box when
                     finished.

                 d.   For each of the 50 filters on the D-Preweight Logsheet, enter
                     the filter identification number and  the pre-MASS value into the
                     sample-handling database.
         B.   Loading the Cassettes

             The procedures in this section are triggered by having a Blue Box with
             empty cassettes.   The Blue Box is a shipping container that was
             specifically designed to transport the weekly supply of filter cassettes
             between UCD and the sampling sites.   The blue case is molded ABS
             material with metal reinforcement.   The cassettes are vibrationally and
             thermally insulated by a soft polyurethane foam liner.   The case is


                                       Page 14

-------
IMPROVE STANDARD OPERATING PROCEDURES
                               JULY 1989
             water-tight and held closed with cam locks.   Molded on the front of the
             case is the message "KEEP OUT OF DIRECT SUNLIGHT."  On the front there is
             a flush-mounted mailer with a metal frame and an indented section to
             display the site-date tag.   The mailer has a reversible prepaid mailing
             label.   The UCD address is on one side of the mailer,  while the site
             address is on the other.   Each Blue Box is assigned to a given site.
             For each site there are 5 to 7 Blue Boxes in the system.

             1.   Prepare the Field Logsheet with the site and date sequence.

                 a.   The Blue Box will have the site and date for the desired sample
                     change listed on the outside label.   Verify that this date
                     agrees with those of the next tags on the sheet of preprinted
                     sample identification numbers for this site.   Update the IMPROVE
                     Dot Chart to indicate that sample preparation is in progress.

                 b.   Record the appropriate sample identification numbers
                     corresponding to each sample on the Field Logsheet.  Table 3
                     gives the code for the sample identification numbers and the
                     associated color codes for all of the modules.   (The color code
                     is used to mark the cassettes.)
                 Table 3.  Sample identification code  and color code for all filters.

                    module   filter   color code   sample  identification code
                      A
                      A
                      B
                      B
                      C
                      D
                      D
1      red/black
2      red/white
1      yellow/black
2      yellow/white
1      green/black

2      green/white

1      blue/black
2      blue/white
site-date-Al
site-date--A2
site-date-Bl
site-date-82
site-date-GlP
site-date-CIS
site-date-C2P
site-date-C2S
site-date-Dl
site-date-D2
(primary)
(secondary)
(primary)
(secondary)
                 The site code is  the standard four-letter NFS  code.  (e.g.  ROMO)
                 The date is that  for the  Tuesday of  the  sample change,  (e.g.  1/7/89)
             2.   Check the cleanliness  of the  module  A cassettes  and load them with
                 clean Teflon filters.

                 a.   Remove the protective red cap  from the  Al  cassette (red/black)
                     and verify its cleanliness.
                                       Page  15

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 b.  Remove the cassette top and the lock ring.  Check that the top
                     has a single flat gasket.  Recheck the condition and the
                     cleanliness of the collection mask, the drain disk, and the
                     ethylene propylene (EP) 0-ring.  Replace the drain disk, if  it
                     has discarded during cleaning or if it is dirty.  Replace the EP
                     0-ring, if it was removed during cleaning or if it deformed.  Be
                     careful to return the filter grate with the finer grids oriented
                     up and the large bar grids down..  If necessary, reclean the
                     filter grate and 0-ring by gently brushing with the brush
                     reserved for nonquartz cassettes.

                 c.  Verify that the correct collection mask is installed;  if the
                     mask is dirty or worn, replace it with a new one.

                 d.  Remove the next petri dish from the fine Teflon petri box.
                     Transfer the filter identification tag from the petri dish and
                     the sample identification tag from the pre-printed sheet to  the
                     top cap, locating them between the exposed ridges.  Center and
                     mount the clean drain disk, the collection mask, and the Teflon
                     filter on the grate.   The rough and extended edges of the filter
                     should mate with the mask.  This orientation will present an
                     open, smooth, and flat surface.  Gently place the lock ring on
                     the cassette, taking care to keep the filter centered and not
                     damage the filter support ring.  Place the cap with a single
                     flat gasket over the lock ring and tighten it.   Keep all
                     elements centered and make it finger tight.   Reinstall the
                     protective red cap designated "A"  over the cassette.

                 e.  Record the filter identification number on the Field Logsheet.

                 f.  Put the cassette in the Blue Box.

                 g.  Repeat procedure for the A2 casette (red/white).


             3.   Check the cleanliness of the 47 mm module B cassettes and load them
                 with clean nylon filters.

                 a.  Check the cleanliness  of the Bl cassette (yellow/black)
                     following the procedures for module A.

                 b.  Transfer the Sample Identification Tag from  the pre-printed
                     sheet to the cassette.   There is no filter identification number
                     for the nylon filter.

                 c.  Obtain a clean nylon filter from the supply  and load it into the
                     cassette directly on to the diffusion and support grid.

                 d.  Put the cassette in the Blue Box,

                 e.  Repeat for the second  nylon filter (B2,  yellow/white).
                                       Page  16

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             4.   Check the cleanliness of the module C cassettes and load them with
                 clean quartz filters.

                 The module C cassettes contain two sections,  each with a quartz
                 filter.   The first or primary filter collects all the particles and
                 adsorbs  organic gases, while the secondary filter only adsorbs
                 gases.

                 a.   Disassemble each section of the Cl cassette (green/black) and
                     check the cleanliness.   Assure that two flat silicone rubber
                     gaskets are used in the cassette cap of the first section.  Two
                     gaskets are needed to provide spacing so  that the hold down
                     retainer of the cyclone manifold will fit.   If there is still
                     any  glass filter debris, remove it with the brush reserved for
                     quartz cassettes.  Replace the drain disk if it was removed
                     during cleaning or if it is dirty.   Replace the 0-ring with a
                     Viton 0-ring,  if it was removed during cleaning or if it is
                     deformed.

                 b.   Load a clean quartz filter from the supply into each section of
                     the  cassette.

                 c.   Record the lot number for each filter on  the Field Logsheet.

                 d.   Transfer the Sample Identification Tags from the pre-printed
                     sheet to each section of the cassette.  The primary filter has
                     the  suffix "P",  while the secondary filter has the suffix "S."

                 e.   Put  the cassette in the Blue Box.

                 f.   Repeat for the second quartz cassette (C2,  green/white).
                 Check the cleanliness of the module D cassettes and load them with
                 clean Teflon filters.

                 a.   Check the cleanliness of the Dl cassette (blue/black) following
                     the procedures for module A.

                 b.   Insert a clean Teflon filter from the module D container box.
                     Do not use a collection mask with this filter.  Mount the filter
                     with the ribbed side down and the smooth side up.   ("Up" means
                     facing the incoming air.)

                 c.   Attach the Filter Identification Tag and the Sample
                     Identification Tag to the outside of the cassette.

                 d.   Record the filter identification number on the Field Logsheet.

                 e.   Put the cassette in the Blue Box.
                                       Page 17

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 f.   Repeat for the module D2 cassette (blue/white).
             6.  If a dynamic field blank (DFB) cassette is in the DFB queue, include
                 it in the current Blue Box.  (The DFB is identified by having a
                 yellow cap on the hose end.) Do not include more than one DFB per
                 Blue Box.  The type of module D cassette must match the normal type
                 for this site, with respect to the impregnated filter.  Repeat the
                 procedure for a cassette of the given module.  Add the suffix DFB to
                 the sample-identification number.  Include the standard letter
                 describing how to install the DFB.

             7.  This completes the loading of the cassettes.  Transfer the Blue Box
                 to the leak-test area.  Leave the Blue Box open.
         C.   Verifying the Loading Procedures and Leak-testing the Cassettes.

             At this point, all the cassettes have been loaded with clean filters,
             have their identification tags, and are in an open Blue Box, along with
             the accompanying Field Logsheet.

             1.  Verify that all cassettes are present and have been loaded with
                 clean filters.

             2.  Verify that the filter- and sample-identification numbers and che
                 quartz-lot numbers correspond to those recorded on the Field
                 Logsheet.

             3.  Record these numbers on the Mailer Record Logcard.

             4.  Mount a cassette on the leak-test element designated for the
                 specific module, following the standard color code.  Hand tighten
                 the nut.   Remove the protective red cap and place che appropriate
                 leak-cap gently over the open cassette face.   There are two
                 leak-caps, one for 25 nun and one for 47 mm.  Open the vacuum valve
                 and observe the flowmeter.  If the flow rate  is less than 1 L/min,
                 the system is considered to be airtight.   If  the flow rate exceeds
                 this value, find and remove the leaks (recenter components, replace
                 0-rings,  etc.).

                 Special care is required for the module C quartz cassettes.  Because
                 of the fragility of the filters, it is not possible to get as good a
                 seal;  the criterion for this module is 3.5 L/min.  Begin by
                 tightening the secondary filter, and then tighten the primary
                 filter.   Be careful to tighten each filter segment only to a "firm"
                 hand tightness.   An absolute leak seal cannot be guaranteed by
                 further tightening.  Excessive tightening will cause the lock ring
                 to sever the edges from the filter.  This will cause the cassette to
                 leak significantly, and the flow meter will rise sharply.   The
                 filter will appear to warp.  Replace all  filters which have been
                 overtightened.
                                       Page 18

-------
                                                     JUL.Y 1989
    5.  Replace the protective red cap designated for the module.

    6.  Repeat the leak- test for the remaining cassettes in the Blue Box.

    7.  Record the current date on the Mailer Record Logcard in the column
        marked "date shipped. "

    8.  File the Mailer Record Logcard in the Mailer File chronologically by
        site.  Verify that the this card follows the previous card by  1
        week.


D.   Prepare the Blue- Box for Shipment.

    1.  Make certain the reusable,  reversible mailer label on the Blue Box
        displays the site destination.

    2.  Update the IMPROVE Dot Chart by filling in the site/week box with
        yellow ink.

    3.  Place the Blue Box in the mailing tub for dispatch.
                              Page 19

-------
IMPROVE STANDARD OPERATING PROCEDURES
JULY 1989
   III.   SAMPLE CHANGING AT THE SITE

         Sample changing is performed every Tuesday by the  field  operator.   Each site
         has a manual (Appendix 2) with the procedures and  a  description of the
         program and the sampler.  The procedures are also  written  on the inside of
         the doors for the controller, module A, and module D.  Figure 3 gives the
         flow diagram for sample changing.
                             receive container of clean filters from UCD
                                  record date and temperatures
                                     start 30-minute timer
                                 measure and record flow rates
                                     record elapsed time
                               remove cassettes with exposed filters
                                              JL
                                 insert cassettes with clean filters
                                  measure and record flow rates
                              send container of exposed filters to UCD
         Figure 3.   Flow diagram for sample changing at the site.
                                       Page 20

-------
IMPROVE STANDARD OPERATING PROCEDURES                          JULY 1989
         A.   Preparing for the Weekly Sample Change

             1.   When the Blue Box is received at the site,  record on the Receipt
                 Log,  the date received and the date written on the label of the Blue
                 Box.   The Blue Box should be received 1 to  2 weeks before the date
                 for sample-changing.
                                         *
             2.   Keep the Blue Box in a clean, dry, and cool location.

             3.   For the Tuesday sample change, take two Blue Boxes to the site:  the
                 empty Blue Box with last Tuesday's date (Old Blue Box) and the full
                 Blue Box with the current Tuesday's date (New Blue Box).
         B.   Removing  the  Exposed  Cassettes

             1.   Remove the Field Logsheet from the Old Blue Box.   The front side
                 will  have been filled in last Tuesday.   On the reverse side, record
                 the present date and time,  your initials,  and the current, the
                 minimum,  and the maximum temperatures.

             2.   Open  the  control module and verify the date and time of the
                 controller clock.  Turn the bypass timer to 30 minutes.  This will
                 start all four pumps.

             3.   For each  module (A,  B,  C, D) in turn,  do the following:

                 a.  Record the small gauge  before pressing either filter switch.
                    This  should be between  16 and 25 "Hg,  depending on the
                    elevation.  If it drops below this,  the problem might be
                     internal pump failure or leaks in the  hose.   Call UCD.

                 b.  Press the filter 1 toggle switch and record the small gauge and
                     the magnehelic gauge.  Repeat for filter 2.   The readings should
                    be near uhe green line  marked on the gauges.   See manual or
                     instructions on door for troubleshooting help.

                 c.  Record both elapsed times in hundredths of hours and zero both
                     timers.  Times of 24.00 hours are expected.

                 d.  Remove each cassette from the holder by first unscrewing the
                    hose  from the solenoid.  Then unscrew  the knurled knob holding
                     the cassettes enough to lift the support bracket and remove the
                    cassette.  Do not remove the knob completely.  Get the
                    protective red cap marked for this module from the old box.  Put
                     the cap on the cassette and place the  cassette in the Old Blue
                    Box.   When finished with module D,  place the  completed Field
                    Logsheet in the Old Blue Box.
                                       Page 21

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         C.   Inserting the Clean Cassettes

             1.  Remove the Field Logsheet from the New Blue Box.  Check the date on
                 the box and on the Field Logsheet;  if this is not the present date
                 for the week, make a note on the Field Logsheet.  Fill in the
                 current date and time and your initials.

             2.  Insert the cassettes for each module in turn, selecting the two
                 cassettes for each module:'  red for A, yellow for B, green for C,
                 and blue for D.  Remove the red cap from the black (filter 1)
                 cassette and put the cap in the New Blue Box.  Insert the cassette
                 into the left side of the manifold (under the clamp) and attach the
                 hose onto the left (filter 1, black) solenoid.  Hand tighten the
                 nuts on the cassette.  Do not use a wrench or pliers to tighten
                 them.   Make certain both cassettes are firmly mounted on the
                 manifold and tighten the knurled knob.

             3.  Record the gauges for each module in turn.

                 a.  Check that the elapsed timers were reset.

                 b.  Record the small gauge before pressing the toggle switches.

                 c.  Press the filter 1 toggle switch and record both gauges.  If
                     either gauge is not near the green line, check the
                     troubleshooting guide inside the module A or D door.

                 d.  Press the filter 2 toggle switch and record both gauges.
             4.   Store the Field Logsheet in the New Blue Box.

             5.   Reverse the mailing label on the Old Blue Box with the exposed
                 filters,  so that Che label shows the UCD address.  Do not add
                 postage;   the shipment is prepaid,  First Class.  Send the Blue Box
                 to UCD.

             6.   If a dynamic field blank is sent with the Blue Box, it is
                 accompanied by a form letter describing the special steps required.
                 The dynamic field blank is needed to monitor any contamination in
                 the system.   It follows the same path as an actual sample, except
                 that no air is drawn through.   The  cassette with the dynamic field
                 blank has a yellow cap on the  end of the hose and has the suffix DFB
                 on the end of its identification number (e.g. GUMO 09/17/89
                 AlDFB.) Do not remove the yellow cap.   The hose will be tagged with
                 the color appropriate for that module.   The blank is inserted on the
                 maniford in place of one of the two manifold caps, so that there are
                 three cassettes on the manifold. When the cassettes are removed on
                 the following Tuesday, replace the  cap on the cyclone manifold.
                                       Page 22

-------
IMPROVE  STANDARD  OPERATING  PROCEDURES
                           JULY  1989
    IV.   SAMPLE HANDLING AFTER SHIPMENT FROM THE  SITE

          The handling procedures in  this section  cover all actions  from the  receipt
          of the Blue Box with exposed  filters up  to the  transfer of the filters to
          the external contractors or to the queue for cyclotron analysis.  Figure 4
          gives the  flow diagram for  this portion  of the  procedures.
                                receive cassettes with exposed fillers from site
                                      review field log slice!
                                     discard Invalid samples
                        enter info database
                            transfer filters and Identification lags from cassettes to petrl dlslies
                              A
                          fine Teflon
                                           send to
                                         Ion contractor
                 send to
             carbon contractor
   n
fine nylon
   C
fine quartz
    I)
total Teflon
          Figure  4.   Flow diagram for  the procedures for  sample handling after the
          receipt of the exposed filters from  the  sampling  site.
                                           Page 23

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         A.   Receiving the Blue Box

             1.   Immediately upon receipt of the Blue Box of exposed cassettes from
                 the site, update the IMPROVE Dot Chart by placing a blue dot in the
                 site/week box, on top of the yellow coloring.

             2.   Clean the outside of the Blue Box.
         B.   Reviewing the Field Logsheet

             The Field Logsheet is reviewed by the person who removes the exposed
             filters.   The review is conducted prior to the removal.   The Field
             Logsheets will be reviewed again by the laboratory supervisor prior to
             their entry into the sample-handling database.

             If correction or explanation is required,  draw a single line through the
             existing information using a fine red pen, and enter the correct
             information.   Comments should also be made using the fine red pen.

             1.  Verify that the actual date of the sample change corresponds to the
                 sample date of the sample-identification numbers.   (Actual dates
                 that are  1 or 2 days before the sample-identification date are
                 acceptable.)

                 If the dates do not agree,  and the samples were not collected over
                 multiple  periods, replace each Sample  Identification Tag with one
                 indicating the actual Tuesday sampling date.   The  discrepancy in
                 dates could occur if the field operator used the incorrect Blue Box.

             2.  Check for omissions.  If possible estimate the correct reading from
                 other readings on the Field Logsheet,  or from readings on the
                 previous  Field Logsheet.   If the solution is not evident,  check with
                 the laboratory supervisor and call the field operator.  Clf the
                 correct data cannot be determined the  sample will  have to be
                 invalidated.)

             3.  Review the comments by the field operator and take action
                 accordingly.  If an equipment or hardware malfunction is reported,
                 take action to correct the problem.

             4.  Verify that all closed-solenoid gauge  readings (system vacuum
                 without sampling air flow)  are greater than 16 inches mercury.

             5.  Verify that magnehelic values range from 0.2.0 to 0.60.

             6.  Verify that filter pressure drops measured on the  small gauge are in
                 the following approximate ranges.

                             A      < 1.5 inches mercury
                             B      < 2.0
                             C      < 5.0   "      "
                             D      < 1.5


                                       Page 24

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 Verify that the sampling durations are approximately 24.00 hours.
                 For sites running with small generators known to have variable
                 frequency (CANY,  ARCH, and ISRO),  if the times range from from 23.6
                 to 24.4 hours,  correct the times to 24.00 hours.  (The controllers
                 use quartz clocks and are not affected by the line frequency.) This
                 variation reflects a frequency variation of ±1 hertz.

                 If the samples  are not valid they are generally removed from the
                 normal system at this point and put in the Unusable Archive.  The
                 action must be  recorded on both the Unusable Archive Inventory and
                 the Field Logsheet.  Add the suffix, listed below, appropriate to
                 the action to the sample-identification number on both the Sample
                 Identification  Tag and the Field Logsheet (e.g.  ROMO 3/15/88A1XX).
                 Include a short explanation on the Field Logsheet of why the sample
                 is invalid, such as "Cassette loose and moved from sample port" or
                 "Tubing attaching support broken." Transfer the filter to a petri
                 dish and put the petri dish in the Unusable Archive.  If there are
                 questions concerning validity,  check with the laboratory supervisor.
                 If necessary, call the field operator.   The three acceptable
                 suffixes and the accompanying reasons for invalidating a sample and
                 archiving it in the Unusable Archive are:

                 a.  DB:  double sample.   The samples were exposed for longer than 24
                     hours, that is, over more than one period.

                 b.  PF:  pump failure.  The vacuum gauge readings with the solenoids
                     closed are  less than 15 inches of Hg.

                 c.  XX:  invalid for other reasons.   The sample ran for less than 12
                     hours, the  filter has a hole,  the filter has obvious problems
                     (not centered properly, grill  upside down),  the filter was
                     dropped or  contaminated by water,  the cassette is broken.
             9.   In two cases,  when the filter clogs  or when the filter is not
                 installed in the sampler,  the invalid samples  are given a suffix,
                 but are not removed from the  normal  system at  this point.  The
                 samples will be analyzed but  the data will be  discarded during che
                 data validation procedures.   Add the suffix appropriate to the
                 action to the sample-identification  number on  both the Sample
                 Identification Tag and the Field Logsheet.   If there are questions
                 concerning validity,  check with the  laboratory supervisor.   If
                 necessary,  call the field operator.   The  suffixes and the
                 accompanying reasons for including a sample in this category are:

                 a.   CG:   clogged filter.   The final  magnehelic reading is less than
                     1/2 of the initial reading,  resulting in unreliable flow rate
                     measurements.   (This sample is kept to document the factors
                     involved in clogging.)
                                       Page  25

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 b.   TB:   transport blank.   The filter was returned without having
                     been installed in the  sampler.   (The transport blank is used to
                     determine artifacts from all causes except adsorption of gases.)


            10.   Enter the date and your initials as reviewer at the end of the Field
                 Logsheet.


         C.   Unloading and Cleaning the Cassettes

             1.   Module A fine Teflon filters

                 a.   Remove the protective  red cap from the Al cassette (red/black)
                     and verify its cleanliness.   If required, wipe the inside of the
                     cap with a Kimwipe dampened with ethanol.

                 b.   Remove the cassette top and the lock ring.  Transfer the fine
                     Teflon filter to a petri dish and transfer the accompanying
                     Filter Identification  Tag and Sample Identification Tag from the
                     cassette to the petri  dish.

                 c.   Place the petri dish in the fine Teflon storage tube.

                 d.   If an incorrect collection mask was used, write the actual mask
                     used (e.g. "1.1 cm  mask") on the Field Logsheet with the red
                     pen.  Change the filter-identification number on the Field Log
                     Sheet and on the Filter Identification Tag to the correct mask
                     code.

                 e.   Check the condition and the cleanliness of the collection mask,
                     the drain disk, and the EP 0-ring.  Discard the drain disk if
                     dirty and the 0-ring if deformed.   Clean the filter grate and
                     0-ring by gently brushing with the camel-hair brush reserved for
                     nonquartz cassettes.  Discard the collection mask if the visible
                     aerosol is not uniformly circular.

                 f.   Reassemble the cassette.
             2.  Module B fine nylon filters

                 a.  Transfer each nylon filter to a petri dish and transfer the
                     accompanying Sample Identification Tag from the cassette to the
                     petri dish.   There will be no Filter Identification Tag.

                 b.  Transfer the petri dishes to the nylon petri box.

                 c.  Record the sample-identification number on the RTI-Ion
                     Contractor Inventory.
                                       Page 26

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 d.   Clean the cassettes following the procedures for module A.


             3.   Module C fine quartz filters

                 The module C cassettes contain two sections, each with a quartz
                 filter.  The first or primary filter collects all the particles and
                 adsorbs organic gases, while the secondary filter only adsorbs
                 gases.

                 a.   Transfer each fine quartz filter to a petri dish and transfer
                     the accompanying Sample Identification Tags from the cassette to
                     the petri dish.   Keep the primary and secondary filters
                     separate.

                 b.   Place the petri dishes in the quartz petri box.  When the box is
                     full, return the petri box to the freezer.

                 c.   Record the sample-identification number on the DRI-Carbon
                     Contractor Inventory.

                 d.   If two filters were inadvertently used for either the primary or
                     the secondary filter, process them as a single filter,  and write
                     a comment on the Field Logsheet.   (This occasionally occurs
                     because the quartz filters tend to stick together.)

                 e.   Make certain that two flat silicone rubber gaskets are used in
                     the cassette cap of the first section.  Disassemble each section
                     of the cassette and clean.  Gently brush away glass filter
                     debris with the brush reserved for quartz cassettes.   Make sure
                     no debris  is retained under the lip of the anti twist lock ring.
                     Brush the  debris into a waste receptacle.  Be careful not to
                     raise or breathe any filter debris dust during this operation.
                     The quartz debris is highly electrostatic and can contaminate
                     other samples.   Discard Che drain disk if dirty and the 0-ring
                     if deformed.
             4.   Module D PM10 Teflon filters

                 a.   Transfer each PM10 Teflon filter to a petri dish and attach the
                     accompanying Filter Identification Tag and Sample Identification
                     Tag to the petri dish.

                 b.   Place the petri dishes  in the PM10 Teflon storage tube.

                 c.   Clean the cassettes following the procedures for module  A.
             5.   Find the date for the next unprocessed week for this site on the
                 IMPROVE Dot Chart.   Make a new tag for this site-date and place the
                 tag in its  place  on the  Blue Box.   Transfer the Blue Box to the
                 queue for clean filter processing.


                                       Page 27

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         D.   Checking and Entering the Field Data

             This step consists in reviewing the Field Logsheet a second time and
             entering the data on the Field Logsheet into the sample-handling
             database.  It can be performed at any time after the Field Logsheets are
             initially reviewed and the filters are removed from the cassettes.  The
             second review is conducted by the laboratory supervisor or his
             representative prior to the entry of the field data into the
             sample-handling database.

             1.   Review the Field Logsheet for Quality Assurance

                 a.  Verify that there is an entry for every space on the sheet.

                 b.  Verify that the correct identification codes were made for all
                     samples.  Verify that the dates are all correct or have been
                     corrected on the Sample Identification Tags and the Field
                     Logsheet.

                 c.  If the data indicate any mechanical problems, verify that all
                     necessary corrective action has been initiated.  If necessary,
                     call the field operator.

                 d.  Check for consistency between the module B pressure drop and the
                     magnehelic readings.   Low magnehelic readings may indicate that
                     the denuder has caused condensation to accumulate in the high
                     pressure side of the magnehelic.   Call the field operator and
                     have .him remove the hose, drain the water, and re-attach it.
                     This operation requires less than a minute.

                 e.  Verify that the temperatures have been reported in Celsius;  if
                     not, convert them.

                 f.  Check for discrepancies in the elapsed time.  If one of the
                     elapsed times is less than the others for trie same sampling
                     period, find the reason for the difference.  Either correct the
                     value or discard the sample.

                 g.  Review the comments of the field operator.

                 h.  When you are satisfied that the Field Logsheet is complete and
                     correct, place your initials as reviewer above the date changed
                     on the front side of the Field Logsheet.


             2.   Entering the field data into the sample-handling database

                 The entry is performed through the computer program DOLOGS.   The
                 steps of this program are as'follows.
                                       Page 28

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 a.   To the question of which record is to be entered, enter the. site
                     code (4 characters) plus the subcode "1", and enter the date
                     encoded in the sample-identification number (e.g. "GLAC
                     05/02/89").  It is essential that this entry be correct.

                 b.   The program will now call up each entry individually, following
                     the layout of the Field Logsheet.

                 c.   If a character is typed incorrectly, use the backspace to delete
                     the character, and then type the correct character,

                 d.   If incorrect data has already been entered, use the displayed
                     menu to modify the data.

                 e.   When the last entry has been completed, the program will
                     automatically return to the site-date question for a new Field
                     Logsheet.

                 f.   When the data for a Field Logsheet have been completely entered
                     into the data base, use a date stamp and a red inkpad to stamp
                     the current input date in the upper left corner of the reverse
                     (final) side of the Field Logsheet.

                 g.   Place the  date-stamped Field Logsheet in the file folder for
                     that site  and season maintained in the IMPROVE file cabinet.
                     The Field  Logsheets should be arranged in sequential order.
         E.   Processing and Shipping the Exposed Filters

             At this point,  the exposed filters are in petri dishes with their
             identification tag(s),  and the dishes are in the appropriate storage
             container.   Sach filter type now follows a separate pathway.

             1.  Module A fine Teflon filters

                 a.   Determine the  post-MASS

                     In this segment,  the post-MASS is determined and the filter
                     transferred to  a slide mount.

                     i.   Get a petri dish from the fine Teflon storage tube.
                         Determine  the post-MASS following the procedures of section
                         V-A.

                    ii.   Record the  post-MASS on the Mounted Samples Logsheet.

                   iii.   Load the filter into a slide mount.

                         Mark the white side of the slide mount with a blue indelible
                         wide-tip felt pen in four places.   Write the
                         sample-identification number on the left,  write the


                                       Page 29

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                         filter-identification number on the top,  and write the site
                         code on the right.   Read the position number corresponding
                         to the sample date  on the position chart on the wall and
                         mark this number on the bottom edge of the mount.  Place
                         this bottom edge into the temporary slide tray first;  this
                         number will be hidden by the tray structure.

                    iv.   When the temporary  slide tray is full, transfer the tray to
                         the LIPM analysis queue.
                 b.   Measure the post-LIPM value for each sample in the temporary
                     tray,  following the procedures described in section V-B.  Record
                     the post-LIPM value for each slide on the Mounted Samples
                     Logsheet.

                 c.   Enter the  sample-identification number,  the
                     filter-identification number,  the post-MASS,  and the post-LIPM
                     into the sample-handling database, for each of the 50 lines on
                     the Mounted Samples Logsheet.

                     This action connects the filter-identification number and the
                     sample-identification number,  permitting the computer to get the
                     proper pre-MASS and pre-LIPM values.   At this point the computer
                     will determine the net mass and optical absorption of the
                     sample, check that the values  are in the acceptable range,  and
                     print a message to the technician.  If the values are not
                     acceptable, the technician will check the entries.

                 d.   Transfer the slides from the temporary tray into permanent
                     site-specific trays.  (Each site-specific tray will hold the
                     samples for one season.   The position in the tray corresponds to
                     the sampling period within the season.)  The number at the bottom
                     of the slide mount will give the position in the tray.  Place
                     the tray in the queue for PIXE analysis.
             2.   Module B fine nylon filters

                 When four petri boxes (200 nylon filters)  have accumulated, ship the
                 boxes to the ion contractor via First Class Mail,  along with the
                 Nylon Filter Inventory.

             3.   Module C fine quartz filters

                 Retain the petri boxes of exposed quartz filters in the freezer
                 until the large shipping container of 10 boxes (500 filters) is
                 filled.   Add blue ice to the shipping container and ship to the
                 carbon contractor by Federal Express or UPS Second Day so that the
                 filters will remain cold during transport.  Include the Quartz
                 Filter Inventory.
                                       Page 30

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             4.   Module D PM10 Teflon filters

                 a.   Get a petri dish from the PM10 storage tube.   Determine the
                     post-MASS following the procedures of section V-A.   Record the
                     post-MASS on the PM10 Archive Logsheet.   Enter the
                     sample-identification number, the filter-identification number,
                     and the post-MASS into the sample-handling database.

                 b.   Return the filter to the petri dish.   Place the petri dish in
                     the next position in the 50-position petri box for  permanent
                     archiving.  Record the sample and filter-identification numbers
                     on the PM10 Archive Logsheet.  When the  Archive petri box is
                     full,  label the box and archive it.
                                      Page 31

-------

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989



     V.  SAMPLE ANALYSIS

         A.   Gravimetric Mass Analysis

             For sample-handling procedures before and after the gravimetric mass
             analysis, see sections I-A-1, I-A-4, III-E-1, and III-E-4.  For the
             startup precedures every morning and afternoon, see Appendix 3 for
             Cleaning and Calibration of the Electrobalance, and IMPROVE Gravimetric
             Controls.

             Never turn the electrobalance off.

             1.  Remove the filter from the petri dish by gently slipping the flat
                 forcep under the outer polyolefin support ring.  IMPORTANT:  The
                 forceps should contact only the outer support ring of the filter.
                 Do not touch the deposit area.  Place the filter momentarily on the
                 antistatic strip, with the aerosol side up.  Discard the petri dish
                 if this is an A filter.  Repeat for approximately 5 more filters.

             2.  Enter the appropriate identification numbers on the appropriate
                 logsheet:

                       module A precollection 	 A-Preweight Logsheet
                       module D precollection 	 D-Preweight Logsheet
                       module A postcollection .... Mounted Samples Logsheet
                       module D postcollection .... PM10 Archive Logsheet


             3.  Remove the filter from the antistatic strip and place it aerosol
                 side up in the center of the weighing pan but offset toward the
                 right side of the pan by approximately 2 mm.  The offset is required
                 so that the filter can be removed from the pan after weighing
                 without putting stress on the balance.  If the filter is
                 inadvertantly centered on the pan, use a second pair of forceps to
                 nudge the filter very gently to a position where the flat forceps
                 can again grasp the filter support ring only.

             4.  Close the glass door and allow the electrobalance to stabilize.
                 This will require approximately 30-45 seconds.  Note that the
                 digital weight readout will slowly decrease until stabilized.

             5.  When the balance is stabilized,  record the mass value on the
                 logsheet.

             6.  Remove Che filter by opening the glass door and gently grasp the
                 filter support ring with forceps.   IMPORTANT:   Be very careful not
                 to grasp the edge of the weighing pan while removing the filter;
                 pulling the balance pan could damage the balance.   Slowly remove the
                 filter from the balance cavity and place it either in the same petri
                 dish (D filter) or on a prepared PIXE slide mount with the aerosol
                 side up (A filter).
                                       Page 32

-------
IMPROVE STANDARD  OPERATING PROCEDURES
JULY 1989
             7.  After every 6 measurements, allow the balance to stabilize without a
                filter  to observe the "zero." If this value exceeds ±2 micrograms,
                recalibrate the balance following the procedures of Appendix 3.

             8.  Whenever the balance is not in use, close the balance glass window.
         B.   Laser  Integrating Plate Analysis (LIPM)

             The  LIPM system  is used to measure the optical absorption of the
             particles on the fine Teflon filters.  The absorption by the particles
             on the filter is smaller than the absorption by particles in the
             atmosphere because of the layering of particles on the filter.  A
             correction to the measured value, based on the areal density of
             particles on the filter, is made at the time of data processing.  A
             schematic of the system is given in Figure 5.  Light of 633 nm
             wavelength from  a He(Ne) laser is diffused and collimated to provide a
             uniform beam of  light of approximately 0.7 car at the sample.  The light
             transmitted through the sample is collected with an Oriel 7022
             photodiode detection system.  The decrease in light intensity is
             produced by both absorption and large-angle scattering.  (Light
             undergoing small angle scattering will be collected by the
             detector.) The blank Teflon filter does not absorb light, but it does
             scatter light;   therefore, it is necessary to measure the transmission
             of the blank filter.  For the particles on the filter, the absorption is
             the  primary cause of decrease in light intensity, with only a small
             amount of scattering.
                                             LIGHT SHROUD
        -FILTER  „ OPAL
                   GLASS
                                                            MASK
                                                         (VARIABLE)
                                               RADIOMETER
             Figure  5.  Schematic of the LIPM system.
                                      Page 33

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             The UCD LIPM system is designed Co handle samples in 18x24 mm slide
             mounts arranged in linear trays.  Prior to the analysis of the
             pre-collection clean filters it is necessary to put the filters in slide
             mounts;  specially modified slide mounts (Quickie Mounts) have been
             machined to simplify the process.  These Quickie Mounts keep the filters
             centered but permit the filters to be inserted and removed easily.  The
             exposed filters are arranged in permanent slide mounts and linear slide
             trays prior to the post-LIPM analysis.

             The system is calibrated at the beginning of every analytical session,
             following the procedures of Appendix 4.   The system is adjusted to give
             a reading of 0.000 when then Beam Attenuation lever is closed and 0.750
             when no sample is present.   A set of standard samples are analyzed and
             the measurements are compared with the  standard values.   If the
             measurements differ by more than a preset value, approximately equal to
             the uncertainty of the measurement, the calibration is repeated.  If
             differences still exist, the system is  examined for problems.

             The steps for the pre-LIPM analysis is  as follows.  See section I-A-1
             for procedures before and after the pre-LIPM measurement.

             1.   Clean the flat metal forceps with ethanol and transfer 50 Teflon
                 filters into Quickie Mounts.  (The  "Quickie Mount" is a modified
                 slide mount used to hold the filters during the precollection LIPM
                 analysis.) Note that each filter support ring has two distinctive
                 sides, the smooth side and the lipped side.

                 Place the filter in the Quickie Mount with the smooth side facing
                 the black side of the mount.  This  orientation is followed in other
                 Air Quality Group protocols and assures that smooth side is always
                 the aerosol side,  and that the smooth side always faces LIPM and
                 PIXE detectors.  This standardized  procedure also assures that when
                 aerosols cannot be visibly seen,  the orientation on the filter is
                 known.

             2.   Place each Quickie Mount in the 40  position slide tray with the
                 black side to the front.  This will  assure the black face with
                 filter smooth aerosol side faces  the detector.  Note that the
                 position 1 corresponds  to the sample identification 1 or 51 in the
                 identification sequence.

             3.   Place the filled slide  tray to be prelasered in the  LIPM system.
                 Adjust the calibration level to 0.750 using the vernier multipler
                 extension arm.

             4.   Push the slide in position L fully  into the system.   Observe and
                 record the value on the A-Preweight  Logsheet under "Pre Laser".
                 Representative values should be between 0.350 and 0.500.

             5.   Periodically verify that the calibration level of 0.750 is
                 maintained when the slide changer is fully out.   Continue until all
                 filter prel'aser values  are obtained  and recorded.
                                       Page  34

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             6.   Remove the 40 position slide tray from the LIPM system and orient it
                 such that the openings of the Quickie Mounts are vertical.
                 Carefully remove the filters sequentially and place them in their
                 corresponding petri dishes.   After closing each petri dish, return
                 it to its petri box.

             7.   Repeat steps 4-6 for any remaining filters in the group of 50.

                 When the petri box is full (50 petri dishes),  close it and attach a
                 l-l/2"x3" white Avery label  to the container showing the sequential
                 filters enclosed,  the date of the pre-LIPM analysis, and your
                 initials.  Also include a space to enter the data for gravimetric
                 analysis.

                           Example:  TF 4551.Ml thru 4600.Ml.
                                     Prelasered      5/15/89 KW
                                     Preweighed      	 	

                 Place the completed container and the A-Preweight Logsheet next to
                 the electrobalance to facilitate the pre-Mass  analysis.
             The post-LIPM analysis follows the post-MASS analysis (cf.  section
             III-E-1).   At this point in the process,  the exposed filters have been
             mounted on permanent slide mounts arranged in temporary 40-position
             linear slide trays.   Each filter has been centered on the mount with the
             aerosol face towards the black side of slide mount,  and the sample date,
             the filter-identification number, the site code,  and an archive position
             number have been written on the white side of the mount.   Accompanying
             the tray i's a Mounted Samples Logsheet.   This logsheet identifies the
             position,  the sample-identification number,  the filter-identification
             number, the post-MASS and provides the location to record the post-LIPM
             values.  The procedures for the post-LIPM analasis are as follows.

             1.   Calibrate the LIPM system as in the  pre-LIPM analysis.

             2.   Place the 40-position slide tray in  the holder on the LIPM
                 apparatus.   Adjust the vernier multiplier extension lever to read
                 0.750 with the slide changer fully out.

             3.   Sequentially cycle slide mounts and  filters into the  system.  Record
                 the post-LIPM value for each sample  in the Postlaser  position on the
                 Mounted Samples Logsheet.

             4.   Periodically verify that the calibration level of 0.750 is
                 maintained when the slide changer is  fully out.

             5.   When all samples in the tray have been analyzed, transfer the slide
                 mounts to the permanent linear slide  trays organized  by site.

             6.   Leave the LIPM system on,  with the slide changer out  and the cover
                 closed.
                                       Page 35

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             7.   Enter all the data from the Mounted Samples Logsheet into the
                 sample-handling database and put the logsheet in the permanent file.
         C.   PIXE/PESA Analysis

             The UCD Particle Induced X-ray Emission Analysis (PIXE) system has been
             optimized to analyze lightly loaded aerosol samples.  The system uses a
             4.5 MeV proton beam from the Crocker Nuclear Laboratory cyclotron to
             excite x-rays in the sample.  The analysis program controls the
             cyclotron beam and the changing of the samples.  The x-rays are measured
             by two detectors.  The first detector measures the entire spectrum from
             Na to Pb, but is optimized for elements with x-rays below the main line
             of Fe.  The second detector is optimized for Fe and above.  With the
             two-detector system, it has been possible to reduce the minimum
             detectable limit of elements heavier than Fe by a factor of more than 2.

             The PESA system runs concurrently with the PIXE system, and determines
             the concentration of hydrogen in the sample.  The method is most
             effective when the filter material does not contain any hydrogen, as is
             the case with teflon.

             1.   Run the dBIII program to produce the instruction files for the
                 PIXE/PESA analysis.  The program will determine the flow rates,  the
                 elasped time, the volume/unit area (used in PIXE to calculate
                 concentration in mass/volume), the mass, the optical absorption and
                 other collection parameters and validity flags.  The program will
                 control which samples to analyze and will permit the calculation of
                 concentrations at the time of analysis.

             2.   Run the PIXE system, using a source to verify the correct operation
                 of the detector system.

             3.   Run Che PIXE standards dray of approximately 40 standards.  These
                 are single element, double element,  and multi-element standards.
                 Verify the calibration of the system and update renormalization
                 values to compensate for small shifts in the system.

             4.   Run the PESA standards tray immediately after the PIXE standards
                 tray.  This tray includes several mylar standards plus a series  of
                 blank Teflon filters.  Enter the calibration value from the mylar
                 standards into the analysis parameter file.  Record the hydrogen
                 values for the blanks.  Select the cleanest blank to be used as  a
                 system blank for PIXE to estimate the background of x-rays in the
                 spectra.

             5.   Reanalyze a tray of filters from an IMPROVE site analyzed during  the
                 previous  IMPROVE analysis session.   Compare the concentrations of
                 the two analyses for all major elements, including hydrogen.   If  the
                 Quality Assurance Manager and the PIXE/PESA Manager find the
                 calibration and reanalysis acceptable,  begin actual analysis.
                                       Page  36

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             6.   Run the actual IMPROVE samples.   Each tray contains all samples for
                 a given season from a single site.   Initialize the analysis of the
                 samples in the tray by entering  the identification code of the tray
                 in the acquisition program.   Initialize the spectral analysis
                 program by entering the run  number  of the analysis.  Check the
                 following features in the PIXE spectral analysis of the sample.  If
                 any problems are detected, consult  with the QA Manager, the
                 PIXE/PESA Manager, or a senior scientist.

                 a.   Is the background completely removed by the two background
                     steps?  That is,  do the  valleys between the peaks in the final
                     spectrum go to zero.   If not, determine if the problem is an
                     inappropriate blank.   Verify on the output log that the run
                     number of the blank is correct.   Do not revise the fitting
                     parameters without consulting with the QA Manager or PIXE
                     Manager.

                 b.   Periodically check the agreement of elements in the overlap
                     region of the two detectors.  Make a note on the PIXE Runsheet
                     if the ratio of the two  Fe's differs significantly from 1.0.

                 c.   Monitor the livetime  of  the  detectors and the cyclotron beam
                     current carefully.   If the livetime drops below 50% or if the
                     beam integrated charge begins to decrease,  ask the cyclotron
                     operator to either drop  or raise the beam current.

                 d.   After every three trays,  while  the next set is being loaded into
                     the system,  run the QA program  to generate the standard
                     correlation and time  plots for  the three sites just analyzed.
                     Note any anamolies for the QA Manager.
             7.   Rerun the standards  at the end of the  session and verify the
                 results.

             8.   Before ending the run,  wait until the  QA Manager is able to validate
                 the results for all  of the trays.   If  problems in the analysis are
                 determined, select the trays and samples involved and reanalyze
                 them.
         D.   Ion Chromatograph Analysis (1C)

             The procedures of the ion analysis  by the  ion contractor,  Research
             Triangle Institute,  are given in complete  detail  in Appendix 6.   This
             section will summarize the procedures.

             1.   Receipt of the filter

                 The exposed filters are shipped from UCD to RTI in batches of 200
                 filters.  Record the following  data on the Sample Log Form when the
                 shipment is received:  the sample indentification numbers, the date
                 of receipt, and any comments on the condition of the samples.


                                     . Page 37

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989



             2.   Filter extraction

                 a.   Desorb the sample in 15 mL of desorbing solution.

                 b.   Expose the sample and solution to ultrasonic energy for 30
                     minutes.

                 c.   Allow the sample and solution to sit overnight.


             3.   Ion analysis

                 The analyses  are performed on two ion chromatographs, Dionex Manual
                 Model 21201 and Model -14.  The following calibration and standards
                 are run separately on the two units.   The model used to analyze a
                 given sample  is included in the reported data,  although there does
                 not appear to be any bias between the models.   The steps below are
                 performed daily.

                 a.   Run a quality control standard (made at least weekly from
                     independently prepared stock solutions);   proceed only if the
                     value agrees with the predicted value to  within 10%.

                 b.   Inject 100 mL of sample and solution into the 1C unit and
                     analyze for chloride, nitrite,  nitrate, and sulfate.

                 c.   At the midpoint of the day,  determine the calibration of the
                     model by  running a series of calibration  solutions of varying
                     concentrations.

                 d.   During each day, run a quality assurance  sample  (EPA Acid
                     Precipitation Audit sample)  and reanalyze one sample (Duplicate
                     Sample).

                 e.   Periodically run a blank sample of only the desorbing solution.


             4.   Data transfer to UCD

                 The data are  transferred to UCD via ASCII files on floppy disk in
                 groups of at  least one batch of samples.   The following information
                 for chloride, nitride, nitrate,  and sulfate is  provided for each
                 batch.  All values are in units  of micrograms/filter.

                 a.   The values,  mean,  and standard deviation  of the  quality control
                     standard  for each 1C unit.

                 b.   The values,  mean,  and standard deviation  of the  quality
                     assurance standard for each  1C unit.

                 c.   The values for each sample.   This includes  dynamic field blanks,
                     travel blanks,  laboratory blank,  and other  quality assurance
                     samples provided by UCD.


                                       Page 38

-------
E.   Thermal Optical Reflectance Analysis (TOR)

    The procedures of the carbon analysis by the carbon contractor, Desert
    Research Institute,  are given in complete detail in Appendix 7.  This
    section will summarize the procedures.

    The measurements are made using an OGC/DRI  thermal/optical carbon
    analyzer.   The method is based on the preferential oxidation of organic
    and elemental carbon compounds at different temperatures.  It relies on
    the fact that organic compounds can be volatilized from the sample
    deposit in a helium atmosphere at a lower temperature than elemental
    carbon.  A small punch is removed from the  sample and placed in the
    analyzer.   The temperature and oxidizing environment are changed in
    time.   This volatilizes carbon compounds, which are first converted to
    carbon dioxide by passing the compounds over heated manganese dioxide;
    the carbon dioxide is next reduced to methane.   The amount of methane is
    then measured by a flame ionization detector.

    The principal function of the optical (laser reflectance) component of
    the analyzer is to correct for the pyrolysis of organic compounds to
    elemental carbon.  Without this correction,  the organic carbon fraction
    of the sample would be underreported and the elemental carbon fraction
    would include some pyrolyzed organic carbon.  The correction is made by
    continuously monitoring the optical reflectance of the filter and sample
    using a laser and photodetector.   This  reflectance, largely dominated by
    the presence of black elemental carbon, decreases as pyrolysis takes
    place and increases as elemental carbon is  liberated during the higher
    temperature stages.

    1.  Receipt of the filter

        The exposed filters are shipped from UCD to RTI in batches of 500
        filters.  Record the sample-identification number for each sample in
        the Air Analysis Logbook upon receipt.   Store tne samples in a
        freezer until analysis.

    2.  Carbon analysis

        a.  Remove a punch from the filter and  place the punch in the sample
            boat.

        b.  Place the sample boat in the analyzer and begin the automatic
            sequence.

        c.  At the end of the analysis,  examine the thermogram for proper
            laser response, temperature profiles,  realistic carbon peaks,
            and the presence of the calibration peak at the end of the
            analysis.  If a problem is found, analyze another punch from the
            sample.
                              Page 39

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
                 d.   Analyze replicate samples at a rate of one replicate per ten
                     samples.
             3.   Validate the analyses following the procedures described in Appendix
                 7.

             4.   Provide the results in micrograms/filter to UCD in ASCII files on
                 floppy disks.   Include both the original and the replicate analyses.
                                      Page 40

-------

-------
 IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989



VI.  PROCEDURES FOR DATA PROCESSING

      A.  Introduction:   The Equations of Concentration and Uncertainty

          1.   Volume

              The volume is the product of the average flow rate and the sample
              duration.   The average flow rate is calculated the the gauge reading
              recorded on the Field Logsheet and stored in the sample -handling
              database.   The average flow rate depends on the average temperature over
              the sampling period,  but not on the temperatures at the times of
              measurement.  The equations for the average flow rates from the
              magnehelic and small gauge readings are:

                  magnehelic:  AFM - a ( Mlb + M2b )  / 2 * (Tavg/Tcal) 1/2

                  small gauge:   AFG - [c - d(Gl+G2)/2] * (Tavg/Tcal)1/2 .

              The variables in these equations are:
                    Ml,  M2    initial and final magnehelic readings
                    Gl ,  G2    initial and final small gauge readings
                    a,b,c,d   calibration constants  during audit
                    Teal      absolute temperature at the time of audit
                    Tavg      average absolute temperature during sampling

              In Che following equations, V is the volume and fV is the fractional
              uncertainty in volume.  The fraction uncertainty in volume equals the
              fractional uncertainty in flow rate, since the duration is well defined.
              The uncertainty can be estimated from  internal and 3rd-party audits.
              The value  determined in this manner includes both precision and
              accuracy.   The difficulty in making the estimate for the IMPROVE sampler'
              is that the precision of the built-in  flow measurement system is as good
              as the precision of most audit devices.   At present,  the best estimate
              of internal precision of average flow  rate is that it is better than 1%.
              and the best estimate of total uncertainty is that it is better than 3%.
              All results from internal and 3rd-party audits are recorded in the audit:
              database.   All calculations are based  on a volume uncertainty of 3%.

          2.   Gravimetric Mass

              The equation for  the  mass concentration is

                   C - ( M - B  )  /  V,

              where M is the mass difference between  the post-MASS  and pre-MASS,  and 3
              is the mass artifact  determined from the mean of the  controls and the
              dynamic field blanks.   The uncertainty  in concentration is

                   aC -  [  (fV C)2 + (aFB / V)  j1/2,
              where aFZ  is  the  standard  deviation in the  controls  and field blanks.
              The  minimum detectable  limit  (mdl)  is  300 ng/nr .
                                        Page  41

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         3.   Optical Absorption

             The equation for the coefficient for the particles on the filter depends
             on the initial and final LIPM measurements,  the volume,  and the area.
             The values are multiplied by 0.97 to compensate for scattering by the
             particles.  To calculate to coefficient in the atmosphere, it is
             necessary to divide the measured coefficient by a factor R that depends
             on the areal density of the particles on the filter.   The coefficient in
             the atmosphere is calculated by the equation

                  b - (area/V) * 104 * ln(PRE/POST) * 0.97 / R .

         4.   PIXE Analysis

            •The artifact concentrations for the PIXE elements are zero.  PIXE
             determines the areal density of the given element;  to determine
             concentration, this areal density is multiplied by (area/V).   The
             uncertainty of the concentration is the quadratic sum of the volume
             uncertainty, the analytical uncertainty of calibration (proportional to
             measured value), and the statistical uncertainty associated with the
             number of counts in each spectral peak.  The 3% volume uncertainty and
             the 4% PIXE calibration uncertainty add together to give a total
             uncertainty of 5%.  The total uncertainty is thus 5%  plus statistics,
             added quadratically.

         5.   PESA Analysis

             The equations for PESA for hydrogen are the  same as for PIXE.   Again,
             there is no significant hydrogen artifact on dynamic  field blanks.

         6.   Ion Analysis

             Analysis of field blanks indicate that there is artifact formation
             during the period in the cassettes.  The standard deviation of the field
             blanks provides an estimate of the uncertainty of the artifact.   The
             data from replicate samples indicate that the analytical uncertainty is
             proportional to measured value, rather than  a constant.   The equations
             of concentration and uncertainty in concentration are

                  C - ( M - B ) / V,

                  oC - [(fV C)2+(<7FB/V)2+(fA*M/V)2 ]l/2,

             where M is the mass measured by the ion analysis and fA is the
             fractional uncertainty in the analysis.  The mdl is equal Co twice^ the
             uncertainty with no loading

                  mdl - 2 (aFB/V)  .

         7.   Carbon Analysis
                                       Page 42

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             Organic carbon artifact is caused by contamination in the filter
             material, by contact with the cassette, and by the adsorption of organic
             gases during collection.  Elemental carbon artifact is caused by
             contamination in the filter material and by contact with the cassette.
             At present, we have insufficient data to to define the various equations
             for organic and elemental carbon.


     B.   Entering the Data into the Concentration Database

         1.   Run the dBIII program FLOWS to calculate and print all the flow rates.
             This program will select the appropriate calibration constants,  obtain
             the data from the sample-handling database, and calculate the flow
             rates.  (The calibration database contains a record of every flow rate
             calibration with the constants and the date of measurement.) The flow
             rates will be calculated and printed.  Check all cases when the two flow
             rates differ significantly, find the cause, and correct the information
             in the sample-handling database.

         2.   Run the dBIII program TRAY to create the instruction files for PIXE/PESA
             and begin the process of transferring data from the sample-handling
             database to the concentrations database.  This program will include the
             following information in the instruction file, using the data from the
             sample-handling database:

             a.  The module A average flow rate from the magnehelic readings.

             b.  The module A sample duration (elapsed time).

             c.  The volume / area.

             d.  The fine mass concentration and uncertainty.

             e.  The PM10 mass concentration and uncertainty,

             f.  The coefficient of absorption and uncertainty.

             g.  The filter-identification number for module A.

             h.  Information on the start time and date, and the site name.

             i.  Instructions on which samples to analyze.


         3.   These data are transmitted from this instruction file  to the PIXE/PESA
             spectra files,  and from there to the PIXE/Pesa output  files.

         4.   When the data from the external contractors are available for a  complete
             season,  run the program IONS to generate similar output files for the
             ion and carbon concentrations.   This program combines  the collection
             data from the sample-handling database with the masses from the  external
             contractors to determine the concentrations and uncertainties.
                                       Page  43

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         5.   Run the program PROM to combine the output files from the last two steps
             to produce concentrations database files.   Each file contains all the
             data for a single site and season;  this information is encoded in the
             file name (e.g. ACAD1A1D.C88).


     C.   Validating the Data

         At  this point, the data from the external contractors have undergone an
         internal analytical validation process.  The gravimetric mass and absorption
         values have been check for consistency. • The PIXE/PESA analyses have been
         check for internal consistency.   In this segment of the procedures the
         different variables are intercompared, and the results are examined for
         anomalous variations with time.

         The procedures here are followed by the Quality Assurance Manager.

         1.   Run the programs CORFIL and CORCOF, using the data in the concentrations
             database, to generate desired correlation plots for all sites.  This
             should be done for the following pairs:   Si and Fe, S*3 (Teflon,PIXE)
             and S04-(nylon,1C),  mass and H, mass and reconstructed mass, OC and
             organic mass by hydrogen and sulfur.

         2.   Run the program IMPSUM to generate time  plots of major variables.

         3.   Run the program DBP to generate a wide variety of statistical
             comparisons.

         4.   Check the results for systematic variations.

         5.   Check individual anomalies;   look for errors in transcribing data.


     D.   Preparing Magnetic Tapes and Floppy Disks

         1.   Run program DBP to create ASCII versions of the concentrations database
             files.

         2.   PIP the files to the magnetic tape or to a floppy disk, as desired.


     E.   Preparing the Seasonal Summaries

         1.   Run the program IMPPG2 to prepare a table of concentrations and a table
             of averages for key physical variables at each site for the season.

         2.   Run the program IMPSUM to prepare a page of time plots of key physical
             variables for each site for three-month period.

         3.   Use the averages files from IMPPG2 to prepare contour maps of key
             physical variables for the network.
                                       Page 44

-------
  IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
VII.  PROCEDURES FOR SAMPLER MAINTENANCE

  The following procedures are to be followed by the Field Engineer and the Field
  Technician.   Sections A, B, and C are to be followed at Davis, while section D is
  done in the  field.

       A.  Evaluating Sampler Modifications

           If  a modification to the sampler is being considered, first set up an
           experiment at the Davis field site to test the modification and its impact
           on  the entire sampling process thoroughly.  (The modifications will normally
           be  made during the annual site visit.)

       B.  Calibrating the Flow Audit Device

           The flow audit device is used to audit the flow rate of the samplers.  The
           device is first calibrated at UCD using a spirometer and a dry gas meter.

           1.   Set up the calibration system to be the same as with a normal sampling
               system:  the air enters through the measuring orifice, then passes
               through the cyclone, a cassette/valve assembly to imitate a loaded
               filter, a critical orifice, and the pump.   Connect the survey spirometer
               to the exhaust side of the pump.  The spirometer measures the displaced
               volume over a given time.

           2.   Start the calibration by turning on the pump and adjusting the valve so
               that the reading on the audit magnehelic is 1.0 "H20.

           3.   Start the spirometer chart recorder rotating and mark the displacement
               in liters.  Repeat for two to four revolutions of the chart recorder, or
               until the measurements are consistent.

           4.   Compute the flow and record the flow on the Audit Calibration Logsheet.

           5.   Verify the flow rate using the dry gas meter.   Install the meter Co the
               intake of the cyclone and take a 5-minute  measurement.  Repeat the
               calibration procedure if the spirometer and dry gas meter measurements
               do not agree.

           6.   Repeat the calibration procedure for magnehelic readings of 1.0,  0.8,
               0.7,  0.6, and 0.4 "H20.

           7.   Calculate a best-fitting logarithmic equation,  following the procedures
               of Appendix 2.

           8.   Record the date,  temperature,  and atmospheric  pressure at the time of
               calibration.

           9.   Label the orifice meter with the calibration date,  temperature,
               pressure, and 5-point logarithmic equation.
                                         Page  45

-------
                                                         JUJ-Y
   10.  Enter the orifice meter calibration data into the audit database.
C.   Preparation for Annual Site Visit

    Each site will be visited at least once per year, normally in spring or
    summer.  The steps prior to the visit are

    1.   Notify the appropriate land management departments about one month prior
        to the scheduled site visit with information on the site, approximate
        date of the visit, and the name of the visiting technician.

    2.   Notify the site operator two weeks prior to scheduled visit.  Coordinate
        the exact date of arrival with the site operator.   Determine if the site
        operator is having any problems with the sampler and/or sample handling.

    3.   Conduct a Site Review

        Prepare a folder for each site to contain a site checklist, a
        Maintenance Checklist (Appendix 8),  a copy of the  previous flow rate
        calibration, and any other pertinent information.

        a.  Review the site summary for the previous year  and note any problems
            on the site checklist.

        b.  Review the description of site location and note any missing
            information, such as site pictures.

        c.  Review the Field Logsheets for the previous year and note any
            problems with flow rates and defective parts on the site checklist.

        d.  Consult with the Sample Handling Laboratory Supervisor about any
            special problems for the site that should be considered during the
            visit.

        e.  Review the concentration data with the Quality Assurance Manager,
            with special emphasis on (1) any inconsistencies that may be
            attributed" to faulty sampler operation, and (2) any change in
            elemental composition indicating changing emissions.  Record any
            significant results on.the site checklist.
D.  Annual Site Visit

    The site visit will include any necessary modifications to the sampler,
    replacement of the nitrate denuder,  a complete inspection of the entire
    sampler and replacement of any defective or deteriorating parts (such as
    hoses and 0-rings), an audit of the flow rate calibration, and optional
    operator training.  The steps are
                                  Page 46

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
         1.   Review the performance of the site with the field operator.  Review the
             site information summary sheet.   Obtain any missing information for the
             description of site locations, including taking site pictures.

         2.   Perform an initial inspection following the Maintenance Checklist
             (Appendix 8).

             a.   Take the final readings of any exposed filter and record the values
                 on the Field Logsheet.

             b.   Inspect the controller module.

                 Read the program of the control clock.   Check the control clock
                 override switches by turning the sampler off.  Turn the bypass timer
                 to thirty minutes and verify the delayed pump startup.   Record any
                 damage on the Maintenance Checklist.

             c.   Inspect the filter modules.

                 Record on the Maintenance Checklist any damage to solenoid values,
                 elapsed timers, relays,  toggle switches pressure guage  and
                 magnehelic.  Check the maximum vacuum of all of the pumps.  Verify
                 the proper installation of cassettes.   If they are improperly
                 installed, discuss the proper procedures with the site  operator.

             d.   Inspect the pumphouse and the pumps and record any damaged parts.
         3.   Conduct the initial flow rate audit.

             a.   Fill out sampler calibraton log information.   Include the data,
                 altitude,  temperature and orifice meter calibration constants.

             fa.   Install the calibration cassettes in filter 1 for each module.

             c.   Leak-test the system,

             d.   Return the bypass timer to thirty minutes.

             e.   Audit the four flow rates indicated on the  sampler calibration
                 logsheet for each module.

             f.   Calculate the orifice meter value corresponding to the indicated
                 flow rates.   Adjust for the ambient pressure  and temperature using
                 the calculated correction factor.

             g.   Press the filter 1 switch and record the vacuum ("Hg)  and magnehelic
                 values.

             h.   Remove the aluminum cap at the bottom of the  stack and insert the
                 orifice meter into the stack opening.
                                       Page  47

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
             i.   Press the filter 1 switch and record the value from the orifice
                 meter magnehelic.   (This value should correspond to 23.0 1pm.)

             j.   Install the valve inline between the filter and the critical
                 orifice.

             k.   With the  orifice meter in place,  press the filter 1 switch and
                 adjust the valve until the orifice meter magnehelic indicates the
                 desired flow rate.

             1.   Remove the orifice meter from the bottom of the stack and insert the
                 aluminum  cap.

             m.   Press the filter 1 switch and record the vacuum ("Hg) and magnehelic
                 values.

             n.   Repeat step c.

             o.   Repeat steps f, g, and h.

             p.   For auditing flows higher than the nominal flow rate, use an empty
                 cassette.

             q.   Repeat steps i and j  until you have finished audit.

             r.   For the PM-10 module audit, insert the orifice meter where the stack
                 attaches  to the cassette holder and leave it in throughout the audit
                 procedure.

             s.   Zero all  the magnehelics.
             Perform the scheduled maintenance following the Maintenance Checklist
             (Appendix 8) ,

             Final Calibration

             a.  Calibrate the fine filter modules to 2 . 5yum cut point by inserting a
                 23.0 1pm critical orifice.
             b.  Calibrate the PM-10 module to lO-O^m cut point by inserting an 18.9
                 1pm critical orifice.

             c.  Check all cyclone 0-rings to insure that they are seated.

             d.  Perform leak-test.

             e.  Perform the multipoint flow audit again.

             f.  Use a permanent pen to mark the maximum, nominal, and minimal flow
                 values on the faces of the magnehelic and vacuum gauges .
                                       Page 48

-------
IMPROVE STANDARD OPERATING PROCEDURES                         JULY 1989
            •g.   Replace the calibration cassettes with the proper monitoring
                 cassettes.

             h.   Record the initial readings on the Field Logsheet.

             i.   If the audit took place on Wednesday or Saturday, start the sampler
                 by pressing the appropriate override switch.   Record the length of
                 time the sampler was not operating on the Field Logsheet.

             j.   Calculate the logarithmic equations for all modules using the data
                 collected in the multipoint audit.
         6.   Enter the Calibration Equations

             After returning from the sites, verify the logarithmic equations for all
             modules at every site.   Input the site visit information into the
             calibration database system.   Record the logarithmic equations,  the site
             code, the temperature at the  time of the final multipoint audit in
             degrees Celsius,  the elevation factor,  and the date.
                                      Page 49

-------

-------
                                  APPENDICES
APPENDIX 1:  Logsheets Used in Sample Handling



APPENDIX 2:  IMPROVE Sampler Manual




APPENDIX 3:  Gravimetric Mass Startup Procedures




             A.  Cleaning and Calibration of the Electrobalance



             B.  IMPROVE Gravimetric Controls



APPENDIX 4:  LIPM Startup Procedures



APPENDIX 5:  PIXE/PESA Procedures



APPENDIX 6:  Ion Contractor Procedures (RTI)



APPENDIX 7:  Carbon Contractor Procedures (DRI)



APPENDIX 8:  Maintenance Checklist (Annual Site Visit)

-------
          Appendix A—1



Logsheets Used in Sample Handling

-------

-------
IMPROVE SOP APPEND                                        JULY 1989
               APPENDIX 1:  Logsheets Used in Sample Handling


     A.   Balance Log--Record of calibration date and significant events for
         each balance at UCD.

     B.   LIPM Calibration Logsheet--Record of calibration data for UCD LIPM
         system.

     C.   IMPROVE Dot Chart--Portion of large chart posted on the laboratory
         wall, indicating the status of all filters by the use of several
         colors.

     D.   A-Preweight Logsheet--Record of data for module A filters, before
         collection.

     E.   D-Preweight Logsheet--Record of data for module D filters, before
         collection.

     F.   Mailer Record Logcard--4"x6" card, record of date that each filter
         was shipped to sampling site.

     G.   Field Logsheet--Record of data for sample change.  The front side is
         for clean filters and the reverse side for exposed filters.

     H.   Mounted Samples Logsheet--Record of data for module A filters, after
         collection.

     I.   PM10 Archive Logsheet--Record of data for module D filters, after
         collection.  Accompanies each box of 50 filters in PM10 Archives.

     J.   External Contractor Inventory--Sample identification numbers of all
         exposed filters sent to the Desert Research Institute or to the
         Research Triangle Institute for analysis.   Accompanies each box of 50
         filters in shipment.

     K.   Unusable Archive Inventory--Record of sample identification numbers
         with problem code for samples that are damaged or otherwise invalid.
         Accompanies each box of 50 filters in archive.

-------
IMPROVE SOP APPEND
                                                                JULY 1989
     A.   Balance Log--Record of  calibration date and significant events for
          each balance  at UCD.
     !>.•»( o
             Temp
                    Tlmf
  7.pro
.003/
  .000
Cnl lbr.il Ion
199. 976/
200. OOO
 r>0 . OOO
mp, . si d.
50.011
 20.000
mg. std.
20.OOA
                                                                     1 n ( 1 1 .1 1

-------
IMPROVE SOP APPEND
                                                         JULY 1989
     B.  LIPM Calibration Logsheet--Record of calibration data for UCD  LIPM
         system.
UATE/Initials
                H    82   83
          c
               ,O5O
o'f'fe
7Y\
                    30?
                        .-*/!
 LASER CALIBRATION

      «5    116    «7   «8    «9  I  S10
                               .SZ8
                                         • 25'J
.!
-------
IMPROVE  SOP APPEND                                            JULY  1989
      C    TMPROVE Dot Chart--Portion of large  chart posted on  the laboratory

       '   wall,  indicating the status of all filters by the use  of several

          colors.
                                HBOCStBBS -OS OUT IW1£, JSBWZBIIwa.
           ia^K&ai i0101 '\ s\ i f\  •*!  i   i  i  i   i  i  ii  i  i  i   i  i  i   i  i  i  i   i  i  r
                      "\ "i  i  i   i  i  i   i  i  i   i  i  i  i   i  i  i   i  i  i  i   i  i  i
                                           i   i  i  i  i   i  i  i   i  i  i  i   i  i  i
                                           i   i  i  i  i   i  i  i   i  i  i  i   i  i  i
                                           _.  .  ,  .   ..
                                       1  1  1  1   1  1  1   1  1  1   1  1  1  1   1  1
                      t  f\  \  i   i  i  i
                    I   I  )  I  I   I  I  I   I  I  I  !   I  I  I   I  I  I  I  I  I  I   I  I  I

                    '\  '\ '\  '\

                    '\  f\ '\  r
                        'i 'l  1   1  1  1   1  1  1  1   1  1  1   1  1  1  1   1
                        'l  1  1   1  1  1   1  1  1  1   1  1  1   1  1  1  1   1  1  1   1  1  1
                    i   i  i  i  i   i  i  i   i  i  i  i   i  i  i   i  i  i  i   i  i  i   i  i
         DOM 01 '\ ^ r]  *T1
                                                1  1
             101. 1" I  I   I  I  I   I  I  I  I   I  I  I   I  I  I  I   I  I  I   I  I  I   I  I  I

             iu/i "r •*!  *r '\ '\  i   i  i  i  i   ii  i   i  ii  T   i  i  i   ii  i   i  i  i

                 'l *"l  'l   I  I  I   I  I  I  I   I  I  I   I  I  I  I   I  I  I   I  I  I   I  I  I

-------
IMPROVE SOP APPEND
                                                          JULY 1989
     D.   A-Preweleht Logsheet--Record of data for module A filters,  before
collection.
PREWEIGHTS PREWEIGHTS
IMPROVE "A" FILTERS, MASKED
Identfftcslton | Pre Wvlghl
iTF 0S&I. Ml
2TF ttTPZ.. Ml
3 TF &70 3 .H!
4 TF #S7>/. M/
s TF £^. Ml
7. TF &S£iJ:lJ
s. TF- ££rc?£Lk(j
9 TF f-?^-}, MI
10 TF f^/D, Ml
11 TF P)S/I . HI
12 TT P>5"/4?, HI
is TF ft£/3. M/
14 TF ^/^ Hi
V^, /$*?
<-//./ ^
y^< /^^
L/ 1 . {;££)










15 TF- h5/5" H)
16 TT «£r///, M/

17 TF fi5/?._MJ i
is TF b5/8. Ml
19 TF £6"/9 /U/
20 TF f->5=2D /-( |
21 TF fc «"2 I Ml
22 TF 0,^52.. HI




PreLMer
0. '^W2
0- 3fr 1
D. 4M-0
O , H 31
(p , Lf (f^
O ' fLfL-
0 . T^T?
O_. ^5/
Idwtlinaflon
26 TF AC4J/a.Ml
27 TF ^521- .Ml
28 TF £63B. M I
26 TT- P;^ ^cf )-/ /
30 TF ft5 30. M 1
31. TF flS4?!. Ml
32. TF ^32 M)
33 TF ft5^?^? Ml
^j, L/97 j 34 TF P>5 B'i , M 1
C <-P~\i.
D ^L
O •^"*~\
D . ^^&
f^ LL£~ *T
t_' ' J^ ^
/"7 , C/£j /—
D. ^^"9
0 . <^f. «?
D.^0
£> , 45;^
o,  V-T"?-
35 TF ££r*r. Ml
36 TF ^5 3^5 , M /
37 TF £537. M/
ss TF p-.53%. M/
Pre Weight












Pre Laser
O -H T I
O.^WD
o . f i T
/^ u ^ i
/"^i W-O^»
£ , 40?-
0 -395
D ^tob
O . 2,>0
O . 3T-/?
O , VC6"
n . ^.-6"
D. W4 !
39 TT- f?S39. H/ /O. ^D? i
40 Tr H5W, M/
41 TF £^74 / , U )
42 TF ^542. M/
43. TF J55"_
-------
IMPROVE SOP APPEND
                                                            JULY 1989
      E.   D-Preweight  Logsheet--Record of data for  module D filters,  before
          collection.
              PREWEIGHTS
                                                      PREWEIGHTS
TP fr I
            U
23.
                              IMPROVE TV FILTERS, UNMASKED
McntmatkNi
1.TF (0l5\. U
2.T? to/52 U
3. TF" fc(=;?. l(
4 TF faisH. U
5 TT" l/> IS5", U
e. 7T fc i 5 i, . U
7. TP It \S1.U
B. TF It IS& U
s TF (s\S9. 6/
20 7T kil-D, U
21 TF fo / ~i \ . I ,/
22 TF /r n 7 IJ
















24 TP u 1 74- ///
25 TP Ic 1 15 . tJ




MwtDflcaflon
26. TF
27 TF
28. TF
ki^b.U
h m u
/el TB. iJ
29 TF /e/ ? 9. L/
so TF
31. TF
/ol&D.ti
fa { ft 1 , L/
32. TF /oi&Z-.U
33. TF
34. TF
Lol %3. U
klft'i-. U
35 TF (el %. Li
36 -TF fc>l&tn.lJ
37. TF fr>iP>T-.IJ
38. TF Lf\r)n,\A
39 TF L"/?^. /x/
40 TF bl^D.U
41. TF /firtl.U
42 TF
/0/92.L/
43. TF Ce/93 M
44. TF 0/9f, L/
45. TF (tftS.U
46 TF
47 Tp
48 TF

-------
IMPROVE SOP APPEND
                                                         JULY 1989
     F.  Mailer Record Logcard--4"x6"  card,  record of date  that each filter
         was shipped to sampling site.
BOX f




DATE OUT





Channel
Module
Code
Al
A2
Bl
B2
Cl
C2
Dl
02
SI
S2
Teflon Filter
ID










Ounrtz Filter
Lot 1










S02 Filter
Lot I?











-------
 IMPROVE  SOP APPEND                                                          JULY  1989
        G.    Field  Logsheet--Record of  data  for  sample  change.    The  front side  is
              for  clean  filters  and  the  reverse  side for exposed  filters.
             Samlcr Pkrtieulae* sa^lin? mewocfc ri«ld
     1; (bl*di)
 rtlttr 2: (**UM)
    a on aaacltt
                                                                      Svaplvt Particulat* SM^linq Mtwork ritld
                                                                                                   Operatori
                                                         <*«*•      /_   ti«»:   	    dayi Sun nan Tut  Initial! 	
                                                         nodule A  («DI
                                                          FUtet 1: (black)
nodul* A («DI        	                                      ^ 2_ (¥||1
 filter i. (blacJt)         '	  .	
 fllttr 2: iMftiut       .	      __^_                      nedult •  (Ylxioil
                                                          Filter li (black)
                                                          Filter 2: 1«*iltel
 niter :• iwttiwi      	      	                      noduu c  (caosNt
                                                          riit*r 1  (black)
                 	                                   Filter 2: iwttitt)
                 	      	                      nodule D  (BLUE)
                                                          Filter I: (black)
   , , iiux.        	                                   rllMr Jt (-|tit|
   '— 1: (blao*t)      	      	
                                                             "  wsamlvr i

-------
IMPROVE SOP APPEND                                               JULY 1989
      H.  Mounted Samples Loesheet--Record of  data  for module  A filters,  after
          collection.
                            IMPROVE/CRITERIA MOUNTED SAMPLES
                .       SAMPLE ID             FILTER ID     POST WEIGHT    POST LASER
                (71 (MO Qs-.oq.RTi fit	  mi^MI    
-------
  IMPROVE  SOP APPEND                                                JULY 1989
        I.   PM10  Archive Logsheet--Record  of data for  module D  filters,  after
             collection.   Accompanies  each  box of  50 filters  in  PM10 Archives.
                                   IMPROVE/CRITERIA ARCHIVED m   SAMPLES
      SAMPLE ID         FILTER ID   POST WEIGHT            SAMPLE ID         FILTER ID  POST WEIGHT
                               I 42.2-01   |
         OS 
-------
IMPROVE SOP  APPEND                                              JULY 1989
      J.   External  Contractor  Inventory--Sample  identification numbers of all
          exposed filters  sent  to the  Desert Research Institute or  to the
          Research  Triangle Institute  for analysis.   Accompanies each box of 50
          filters in shipment.
                    F1I.TER INVENTORY FOR EXTERNAL ANALYSIS BY 1^(. X. ~ C1
                  Filter Iocnttfir.lt Ion                Fitter IHptit I f (c.illoii
                  Site  Date  Channel                Site   Date  Channel
           i-   f*iGu  og-pq. en   of      z»._
           i.  '&)km as -en. en  czs     27.
           3.   iDSt  C-Mfo'gf)   C.-IS      28.
                                           10.
           6.                               31.
           7.                               17.
           fl.	     13._

           9.	I	     3'.. _

           10.                              35.
           11.	     36. _

           12.	     37._


           13.	     38

           14.	     39._


           15.	     40._

           16.	     41._


           17.	     42.

           IB.	     43._

           19.	     44.


           2O-	     <••).


           21.	     66.


           22.	     47.


           23.	'.a._

           24.	49.

           7'j.	50.

                           InventorIctl liy

-------
IMPROVE SOP APPEND
                                                            JULY  1989
     K.  Unusable Archive Inventory--Record of sample identification numbers
         with problem code for samples  that are damaged or otherwise invalid.
         Accompanies  each box of 50  filters in archive.
                    ARCHIVE OF UN-USABLE IMPROVE/CRITERIA SAMPLES
PROBLEM
IDENTIFICATION FILTER ID CODE
1 SKMD /pV 
z7T)}JT'c»J:lt'?>cif\\
3. CCL.A ctf- /*.#n? i
4 ^F.LL t4'2b'ffir>\
5.
6.
7,
8
9.
10.
11.
12.
13.
14
15.
16
17.
18.
19.
20
21
22.
23
M.
25
TF- 7V9.MI
TF-?«f5l>U
__
TF «rt M





















yy
?H
XX
T>R





















PROBLEM
IDENTIFICATION FILTER ID CODE
26
27
28
29.
30.
31.
32
33
34
35
36.
37
38
39.
40
41.
42.
43.
44.
45
46.
47
46
49.
50


















































P-i
i
                                                            ARCHIVE

-------

-------
              IMPROVE    Sampler    Manual

                                Version 2
                              January 1988

                            Robert A. Eldred
              Air Quality Group, Crocker Nuclear Laboratory
                University of California, Davis CA 95616


                            Table of Contents

1.  Overview	    3

2.  General Description of the Sampler  	    6

3.  Site Location	    9
    3.1  Site Selection Criteria  	    9
    3.2  Power Requirements   	    9

4.  Sample Changing Procedures  	   10
    4.1  General Description  	   10
    4.2  Details of Sample Change	   12
    4.3  Trouble Shooting   	   16

5.  Control Module	   18
    5.1  General Description  	   18
    5.2  Operating the Time Clock    	   18
    5.3  Programing the Time Clock	   19
    5.4  Electrical Circuitry   	   21

6.  Filter Module	   23
    6.1  General Description  	   23
    6.2  Flow Rate Measurement	   24
    6.3  Flow Rate Audit	   26
    6.4  Flow Rate Calibration	   28
    6.5  Electrical Circuitry   	   32
    6.6  Nitrate Denuder	   33

7.  Pump House	   34
    7.1  General Description	   34
    7.2  Pump Specifications	   34
    7.3  Electrical Circuitry   	   35

8.  Sampler Stand	   37

9.  Acknowledgments	   39

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
                          Improve Sampler Manual
                                 Version 2
                               January 1988

                         List of Table and Figures
 Table  1.   List of IMPROVE and NPS Criteria sampling sites  ....    4
 Table  2.   Density factor at IMPROVE sites  	   30
 Table  3.   Density factor vs. altitude	   30




Figure  1.   Map of IMPROVE and NPS Criteria sampling sites 	    4
Figure  2.   Layout of IMPROVE sampler, without pump house  	    7
Figure  3.   IMPROVE sampler field log  	   11
Figure  4.   Reading the min/max thermometer  	   13
Figure  5.   Layout of fine filter module	   14
Figure  6.   Layout of PMlO filter module	   15
Figure  7.   Electronic time clock with cover plate removed 	   18
Figure  8.   Time clock programs for standard IMPROVE cycle 	   20
Figure  9.   Electrical schematic for control module  	   22
Figure 10.   50% cutpoint for cyclone vs. flow rate   	   23
Figure 11.   Flow rate audit log	   27
Figure 12.   Flow rate calibration log	   29
Figure 13.   Electrical schematic for filter module 	   32
Figure 14.   Nitrate denuder   	   33
Figure 15.   Layout of wall of pump house	   34
Figure 16.   Electrical schematic for pump house  	   35
Figure 17.   Electrical schematic for relay box   	   36
Figure 18.   Outdoor stand for IMPROVE sampler  	   38

-------

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
                               1.  OVERVIEW
Fine aerosol particles affect remote areas primarily by impairing
visibility and secondarily by producing acid precipitation.  These fine
particles are generally manmade, although some are produced by smoke and
windblown dust.  (Most naturally produced particles are coarse and have a
smaller effect on visibility and acid rain..) In the case of sulfur, fine
sulfate particles are produced by the transformation of sulfur dioxide gas
in the atmosphere.  Measurements of the concentration and composition of
these fine particles are necessary to determine the extent of the problem
and possible sources of the particles.


The National Park Service and the Environmental Protection Agency have been
monitoring particulate concentrations at national parks, monuments,
forests, wildlife refuges and other remote sites since 1979 using stacked
filter samplers.  Coarse particles are collected on nuclepore filters and
fine particles on teflon filters.  The coarse filters are analyzed for
mass, while the fine particles are analyzed for mass, optical absorption,
hydrogen and elements Na to Pb, including sulfur and the soil elements.
From the measured data we can calculate the concentration of organic
particles.


The particulate monitoring program has been expanded to include  other
federal agencies with the establishment of the IMPROVE (Interagency
Monitoring of PROtected Visual Environments) program, designed to determine
the extent and causes of visibility impairment at selected class I areas
throughout the United States.  The National Park Service maintains
additional sampling sites through the NFS Criteria Pollutant Monitoring
program.  The two programs use the same sampler and nearly identical
sampling protocols and are operated by the Air Quality Group of  Crocker
Nuclear Laboratory at the University of California at Davis.  The  sites
selected as of October 1987 are shown in Figure 1 and listed in  Table  1.


A new sampler was designed for these networks called the IMPROVE Modular
Aerosol Monitoring Sampler that collects three samples of  fine particles
(smaller than 2.5 tm) and one of  respirable particles  (smaller than  10 pm).
At  the NFS Criteria Pollutant Monitoring sites, a fifth  filter measures
gaseous S02.  The entire unit is modular, with four filter modules,  a
controller module and a pump house containing four pumps.  The modules are
mounted either on an outdoor wood stand or on a wall of  an air quality
building.  The IMPROVE sampler  retains the  simplicity of the stacked filter
sampler but adds several features, including additional  filters  for
measuring nitrates and carbon,  twice  the flow  rate to  improve  sensitivity,
an  improved flow rate measurement system, and fewer sample changes.


The samplers require a weekly change of  filter cassettes by  field  personnel
provided by the cooperating  federal agencies.  Each site receives  from
Davis a weekly box of 8 filter cassettes and a log sheet.  After the filter
change, the 8 cassettes of exposed filters  are  returned  to Davis for
analysis.
                                     -3-

-------
                                                  IMPROVE Sampler Manual
                                                         Version 2, 1988
           Figure 1.  Map of IMPROVE and NPS Criteria Sites
          Table Is  List of IMPROVE and NPS Criteria Sites
   IMPROVE Network

ACADIA National Park
BIG BEND National Park
BRIDGER Wilderness
BRYCE CANYON National Park
CANYONLANDS National Park
CHIRICAHUA National Monument
CRATER LAKE National Park
DENALI National Park and Preserve*
GLACIER National Park*
GRAND CANYON National Park
GREAT SMOKY MOUNTAINS National Park
JARBIDGE Wilderness
MESA VERDE National Park
MOUNT RAINIER National Park
ROCKY MOUNTAIN National Park
SAN GORGONIO Wilderness
SHENANDOAH National Park
SUPERSTITION Wilderness
WEMINUCHE Wilderness
YOSEMITE National  Park
   NPS Criteria Network

ARCHES National Park
BADLANDS National Park
BANDELIER National Monument
GREAT SAND DUNES National Monument
GUADALUPE MOUNTAINS National Park
HALEAKALA National Park
HAWAII VOLCANOES National Park
ISLE ROYALE National Park
LASSEN VOLCANIC National Park
PETRIFIED FOREST National Park
PINNACLES National Monument
POINT REYES National Seashore
REDWOOD National Park
VIRGIN ISLANDS National Park
VOYAGEURS National Park
YELLOWSTONE National Park
 * Also an NPS Criteria Site
                                   -4-

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
In order to obtain a complete signature of the composition of the
particles, a variety of analytical methods are used.

  A   fine teflon filter — gravimetric analysis (mass)
                            laser integrating plate (optical absorption)
                            Particle Induced X-ray Emission (Na to Pb)
                            Proton Elastic Scattering (hydrogen)
                            Forward Alpha Scattering (H, C, N and 0)
  B   fine nylon filter 	 ion chromatography (nitrate)
  C   fine quartz filter — combustion analyzer (organic carbon
                              and elemental carbon)
  D   respirable particle teflon filter — gravimetric analysis (mass)
  S   impregnated quartz filter — ion chromatography (S02)

  Filters A and D are analyzed by the Air Quality Group at Davis
  Filter B is analyzed by Research Triangle Institute
  Filters C and S are analyzed by Desert Research Institute
The analytical results are included in 3-month seasonal summaries, which
are distributed to the cooperating agencies and to the local sites.  In
addition, interpretative studies are performed by the participating
contractors, the cooperating agencies and by other research groups.  The
data are available to local resource managers on a variety of media.

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
                2.  CTNERAL DESCRIPTION OF HIE SAMPLER


The IMPROVE modular aerosol monitoring sampler was designed and built by
the Air Quality Group at Davis specifically for the IMPROVE program.  The
following design criteria were followed:


—Accuracy and precision in particulate collection including air volumes,
  particle size cuts and time intervals.


—Reliability and ruggedness under extreme environmental conditions,
  without requiring air conditioned shelters.


—Ease of siting and operation, with protocols to eliminate potential field
  errors and to minimize the time demands on the field personnel.


—Flexibility in terms of present and  future sampling requirements, with
  ability to increase or decrease the  number of simultaneous samples and  to
  match the system sensitivity to the  specific site.


—Attractive and professional appearance, to permit inclusion  in public
  displays, if desired.


—Reasonable costs.
 The  sampler  is  composed of six modular units:   a control  module,  four
 filter  modules  and a pump house,  as shown in Figure 2.  The control and
 filter  modules  are all  contained  in identical gray fiberglass enclosures,
 20 inches high  by 17 inches wide  by 11 inches deep.  These are mounted
 either  outdoors on a wood stand with sunshield and work table, or on the
 walls of an  air quality building.  The aluminum pump house measuring 3 feet
 wide by 2 feet  deep by  4 feet high will contain four pumps, a cooling fan
 and two heaters.  The hoses and wires are enclosed in teflon coated metal
 conduit.
 The control  module contains a 7-day,  4-channel electronic time clock, a
 30-minute bypass timer for sample changing plus appropriate relays.  The
 module has a cooling fan that turns on at 85°F and a neater that turns on
 at 20°F.
                                     -fi-

-------
                                                    IMPROVE Sampler Manual
                                                           Version 2, 1988
                                                         sec.
      Figure 2.  Layout of IMPROVE modular Aerosol Monitoring Sampler,
                without pumphouse
The three fine particle filter modules are almost identical.  The inlet
protects against rain and eliminates bugs and particles larger than around
15 urn.  One of the units has a nitrate denuder to remove gaseous oxides of
nitrogen from the air.  The air passes through identical cyclones, designed
to remove particles larger than 2.5 im at a flow rate of 21.7 I/fain.  The
airstream passes through a filter, which collects all the fine particles.
In the standard configuration, each module contains two solenoids and two
elapsed time indicators to handle two filter cassettes.  The unit is
designed to be able to accomodate up to four cassettes by adding additional
solenoids and elapsed time indicators.  Each unit also contains two gauges
to measure the flow rate and two toggle switches for use during sample
changing.  Each module is connected to its own pump in the pump house.

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988


The PM10 module differs from the fine modules in that the fine inlet and
cyclone are replaced by a commercial PMlO inlet similar to those used for
standard dichotomous samplers (virtual impactors), but designed to operate
at 18.9 1/min.  All particles smaller than 10 j/m are collected on the
filter.  The filter cassettes are mounted vertically up rather than down as
in the fine modules.


The flow rate of 21.7 or 18.9 1/min is regulated by a critical orifice
calibrated for the intended filter medium.  The flow rate is measured
before and after the collection by two independent methods.  The first
measures the pressure drop across the filter and uses the equation for flow
rate through a critical orifice.  The second method calculates the flow
rate by measuring the pressure drop across a fixed orifice.


The filters are transported and installed in cassettes;  the filters are
handled only under controlled laboratory conditions.  This is essential in
maintaining quality control and in simplifying field protocols.  The
nitrate module uses 47mm cassettes while the others use 25mm cassettes.
The active area of the sample collection on the fine teflon is reduced from
3.8 mm to 2.2 cm by using masks in the cassette behind the filter, in order
to improve the analytical minimum detectable limit for elemental analysis  .
The use of cassettes permits adding extra filters to the system without
revising the field protocols.  For example, at some sites, the PMlO  filter
will be followed by an impregnated quartz filter  to capture gaseous  S02.
The following filter media are to be used in the  standard system, along
with the corresponding analytical measurements:

  module A   25mm teflon    fine   mass, absorption, elemental  (H, Na-Pb)
  module B   47mm nylasorb  fine   nitrate
  module C   25mm quartz    fine   organic and elemental carbon
  module D   25mm teflon    PMlO   mass
                                     -8-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988


                                  3.  SITE LOCATION

3.1  Site Selection Criteria


     The specific sampling site should be selected with the following criteria:


     —The site must be away front local combustion sources, such as diesel
       generators, automobiles, chimneys, and dumps.


     —The site must not have obstructions that would hinder sampling
       representative aerosols, such as trees or buildings.


     —The site should not be located in small valleys subject to nonregional
       conditions.


     —The site must have 25A of current at 120V.


     —The site must be accessible  for filter changes in all weather conditions.


     —The site should be located near existing particulate monitoring stations
       in order to provide continuity.

3.2  Power Requirements


     The maximum current for the standard system is 23 amps and lasts for  a  few
     seconds during the startup of  the pumps.  The normal  current with the pumps
     in operation is 14 amps.  The  heater in the controller module adds  0.7  amps
     and a 60W lamp in the pumphouse adds 0.5 amps.  The annual energy
     consumption is 4,000 kWh.
                                          -9-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988


                           4.  SAMPLE CHANGING PROCEDURES

4.1  General Description


     This section summarizes the protocols to be followed for the weekly sample
     change.  Section 4.2 will provide more details on the procedures and
     section 4.3 will discuss trouble shooting.  The instructions are posted
     inside the doors of the control module, module A and module D.  You will
     need last week's shipping box to put in the cassettes currently in the
     sampler and this week's shipping box with the new cassettes.


     The filters.are to be changed once a week, either on Sunday, Monday or
     Tuesday.  Because of other air quality instruments, most operators will
     make the change on Tuesday.  The shipping boxes containing the clean
     filters will be identified by the date for Tuesday.  The box should be
     received 1 to 2 weeks before the specified date.  A receipt log is provided
     to document the receipt of these boxes.  If the box is not present at the
     time it is to be used, please call Davis to discuss procedures.  We will
     have you use another box if available.  In any case, you must remove  the
     exposed cassettes before Wednesday:  if any filter is run for two periods
     it will be invalid.  You will need to have at the site the empty shipping
     box for the exposed filters in the sampler and the full shipping box  with
     clean  filters.


     Inside the shipping box will be eight cassettes and a log sheet.  Each
     cassette will be identified by both colored tape and a coded label.   The
     two systems are redundant;  if you follow the colored tape you can  ignore
     the code.  If "date" is the Tuesday date on the shipping box, the cassettes
     will be as follows:

            MODULE FILTER  COLORED TAPES     LABEL         DESCRIPTION

              A     1      red, black       date-Al       single  25mm cassette
              A     2      red, white       date-A2            (teflon)

              B     1      yellow, black    date-Bl       single  47mm cassette
              B     2      yellow, white    date-B2            (nylasorb)

              C     1      green, black     date-Cl       double  25mm cassette
              C     2      green, white     date-C2            (quartz)

              D     1      blue, black      date-Dl       single  25mm cassette
              D     2      blue, white      date-D2            (teflon)*

            *If S02 is to be  measured  at NPS Criteria  Pollutant Monitoring
              site, the  D  filter will be  a  double  25mm cassette.


     The  log sheet is two-sided, with the initial  data on one side and  the final
     data on the other.  These  are  shown  in Figure 3.   The  initial data is to be
     entered after inserting the  clean  filters.   The final  data is to be entered
     the  following week  before  removing  the exposed filters.   The log sheets
     correspond  to the  8 cassettes  in the shipping box.   The log sheets must
     remain with the  shipping box.
                                         -10-

-------
                                                                         IMPROVE  Sampler  Manual
                                                                                   Version  2,  1988
         nvKNI _»f>l*t rntle-UU (Mpllnf Wtwk M«M la*
DATUM
            IMUf tn_»rtlnt elna -_u*ttMl
                                                  Initial!
                   •Mil }«lfl
                      -
Hfnchtlle
 f«•(•!>
NBrfuK » OBI
  riu«t It (MxM
  MlUt It t
                                                                                        »i«w
                                                   HffDttta (•>(«» tnmrinf i)q fton Tu»  InltUll
                                                                                                Operator!
                                                                             r»"tel
                          Figure 3.   IMPROVE sampler  field log.
                                                  -11-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2,  1988


     The  protocols  to be  followed at the sampler for the weekly sample change
     involve  5  steps.

       step 1:   Record the following on the reverse side of last week's log
                sheet: date;   current,  min and max temperatures;  your
                initials;  and any comments on sampler problems, construction
                near sampler,  etc.
       step 2:   Check the time and date on the controller clock and turn timer
                to  30 minutes.
       step 3:   For each  module (A,B,C,D) record 5 gauge readings and 2 elapsed
                times.
       step 4:   For each  module (A,B,C,D) remove the exposed cassettes, placing
                in  last week's shipping box along with the log sheet.
       step 5:   Get the cassettes and the log sheet in this week's shipping box.
                For each  module (A,B,C,D) insert the new cassettes, matching
                colored tapes  (A-red, B-yellow, Ogreen, D-blue;  black on left,
                white on  right).
       step 6:   For each module (A,B,C,D) record 5 gauge readings.


4.2  Details  of Sample Change


     This section provides more details on the five steps in the sample change.

       step 1:   Remove the log sheet from last week's shipping box.  The front
                side should have been filled in the previous week.  On the
                reverse side,  record the present date and time and your
                initials.  Record the current temperature, the minimum
                temperature (since the last change) and the maximum temperature
                on  the min/max thermometer.  Figure 4 shows how to read the
                thermometer.  Use the Celsius scale.  Press the reset button
                until the min/max indicators drop to the mercury.  (The three
                temperatures are needed for precise flow rate measurements.)

       step 2:   Open the control module.  Verify the time of the time clock.
                Turn the bypass timer to 30 minutes.  This switch  turns on the
                four pumps, and disconnects the auto control of the solenoids.
                Thus the solenoids will all be closed.


                If the change takes longer than 30 minutes, you must  reset the
                switch.  It is best to do this before the 30 minutes  expires.
                If the pumps  turn off while there is a good vacuum behind the
                solenoids and if  the solenoids are closed, the pumps  will
                generally not restart.  Therefore if the bypass timer  turns  off,
                briefly open  one  toggle switch in each module to destroy the
                vacuum, before restarting the timer.
                                        -12-

-------
                                                      IMPROVE Sampler Manual
                                                             Version 2,  1988
                        WIN
MAX
read minimum
temperature at
bottom of blue
indicator 	
                                                read maximum temperature
                                                at bottom of
                                                blue indicator
                                                read current temperature
                                                at top of mercury
                                                (either side)
Figure 4.  Reading the min/max thermometer.  Note that the left side increases
as one reads down.  Use bottom of blue indicators to read minimum  (left side)
and maximum (right side) and top of mercury to read current temperature.  When
finished recording values, hold down reset button until both blue  indicators
drop to mercury.
    step 3:  For each module (Ar B, C, D) in turn, do the following.  Layouts
             of the fine and PM10 modules are shown in Figures 5 and 6.

                  a.  Before pressing either filter switch, record the small
                      gauge.  This should be 15 to 25 "Hg (depending on the
                      elevation) and indicates the pump is working properly.
                      If it drops below the indicated yellow line, look for
                      leaks in hose and hardware between the solenoid and
                      pump and call Davis.
                  b.  Press the filter 1 toggle switch and record the small
                      gauge and the magnehelic gauge.  The insert in Figure 5
                      shows how to read the small gauge below 5  "Hg.  Section
                      4.3 discusses the significance of the gauge readings
                      and how to use them for troubleshooting.
                  c.  Repeat substep b for filter 2.
                  d.  Record both elapsed times (in hundredths of hours).
                      Times of 24.00 hours are expected.  Zero both elapsed
                      timers.  In extreme cold the reset buttons may not
                      operate easily.  In this case it will be necessary  to
                      keep a running total and determine the durations by
                      subtraction.
                                      -13-

-------
                                                  IMPROVE Sampler Manual
                                                         Version 2, 1988
  INLET
   TACK
•bottom
cop for
 stock
             I   r^
I	1
,	j  PUMP VAC.

I	1
   O
                Toggle
                •0FILTER 2   \\

                             l'
Dwyer  2001 LT

Magnehellc   n
  0-I.O"H20   '

  AIR FLOW

        ^
         (-• -• i
               ocuum line
               to pump
               •Cup  for
               coarse particles
               ff
                Figure 5.  Layout of fine filter module
                                  -14-

-------
                                 IMPROVE Sampler Manual
                                       Version 2, 1988
       4   x- INLET STACK
      "^V (HAS PM-IOHEAD)
                                        CRfTfCAL
                                        ORIFICE
 Vocuum line
  to pump
Figure 6.  Layout of PM10 filter module
                 -15-

-------
                                                          IMPROVE Sampler  Manual
                                                                 Version 2,  1988


      step 4:  Remove each cassette by first unscrewing  the  hose from the
               solenoid.  Then unscrew the  knurled knob  holding the cassette
               enough to  lift the  support bracket  and remove the cassette.  Do
               not  remove the knob completely.   Put the  red  cap on the cassette
               and  place  it  in last week's  shipping container.   When finished
               with module D, place the completed  logsheet in also.

      step 5:  Remove the blank  log sheet from this week's shipping box.   Check
               the  date on the box and on the  log  sheet.  If this is not  the
               proper date for the current  week, make a  large note on the log
               sheet.   (Also call  Davis.) Fill in  current date and time and
               your initials.  If  this is change does not follow the removal  of
               the  exposed cassettes,  record the current temperature.  Insert
               the  cassettes for each  module in turn, selecting the two
               cassettes  for the module:  red for  A, yellow  for B, green  for  C,
               and  blue for  D..  Remove the red cap from the black (filter 1)
               cassette,  insert  the cassette into  the left side of the manifold
                (under  the clamp) and attach the hose and nut onto the left
                (filter  1, black) solenoid.   Place  the red cap in the empty
               shipping box. Repeat  for the cassette marked with white tape
                (filter  2) inserting on right side  of manifold and right
               solenoid.  Hand  tighten the  nuts on the hoses.  Do not use a
               wrench  or  pliers  to tighten.  Make  certain both cassettes  are
                firmly mounted on the manifold and  tighten the knurled knob.

       step 6:  Record  gauges for each  module in turn, in the same way as  with
                the  exposed  filters:

                     a.   Before pressing the toggle switches record the small
                         gauge.
                    b.   Press  the  filter 1  toggle  switch and record both
                         gauges.
                     c.   Press  the  filter 2  toggle  switch and record both
                         gauges.
                    d.   Check  that elapsed timers  were  reset.   If  the elapsed
                         timer will not reset because of cold, record the
                         initial  time  on the right side oth the  log sheet.


               You  could turn off  the bypass timer, although it will
               automatically turn off after the 30 minutes have elapsed.

4.3  Trouble Shooting


     This section discusses  how to interpret the readings of the two gauges when
     a solenoid is open.  The gauges provide two independent measurements of  the
     flow rate.  The equations  are given in section 9.  The magnehelic measures
     the pressure drop across the cyclone or a large orifice;  as  the flow  rate
     decreases, the  magnehelic reading will also decrease.  The  small gauge
     gives the pressure drop across the filter;  an increase in  this pressure
     drop (due to loading of the filter or a nonstandard filter) will cause the
     flow rate to decrease.
                                        -16-

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988


If the needles of both gauges are on the green line, then the sampler is
operating at the standard flow rate. " If there are no leaks in the system
and the only problem is a nonstandard drop across the filter, the two
gauges will shift in the opposite direction.  For example, a heavily loaded
filter would cause the small gauge to increase and the magnehelic to
decrease.  As long as the needles are within the red lines, the flow rate
is acceptable (10%).


Several problems in the vacuum system can be quickly seen by comparing the
two gauge readings.  Note that the fine modules (A, B, C) differ from
module D.

     1.  The cassette or cassette plug is not placed firmly on the cassette
         manifold.  (The plug is a problem for module B.)  Modules ABC:
         the small gauge will read in the acceptable range, but the
         magnehelic will read low.  In module D, the gauges will not
         indicate when the cassette is not seated firmly, so special care
         must be taken.
     2.  The. hose is not be connected properly to the solenoid.  Modules
         ABC:  both the magnehelic and the small gauge for one filter will
         read low.  Module D:  the magnehelic will be acceptable, but the
         small gauge will read low.
     3.  Foreign matter is blocking the critical orifice.  Modules ABC and
         module D:  both the magnehelic and the small gauge  for both
         filters will read low.  Call Davis for instructions.
     4.  Module B only.  The denuder has slipped and is  partially blocking
         the inlet.  The magnehelic will read high.  Call Davis for
         instructions.
 The  following  table  summarizes  the  possible  errors:

  module   vac gauge  magnehelic   possible problem
   ABC      normal        low       cassette or plug not seated
   ABC      low          low       hose not connected properly at solenoid
                                     critical orifice clogged *
                                     leak between solenoid and pump **
     B       normal        high       denuder  in inlet has dropped
     D       low          normal    hose not connected properly at solenoid
     D       low          low       critical orifice clogged *
                                     leak between solenoid and pump **

        *   should occur for both filters in module
        ** the  vac gauge  will also be lower than normal with both soleniods
           closed—should occur  for both filters in module

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2,  1988
                                     CONTROL MODULE
5.1  General Description
     The control module consists of a 4-channel 7-day time  clock,  a 30-minute
     bypass timer, five control relays, a  fan with  thermal  switch  set for 85°F,
     a heater with thermal switch set for  20eF, three time  delay relays,  a 200
     VA transformer and a terminal strip.  The electrical circuit  is discussed
     in section 8.  The time clock is discussed in  section  5.


     The time delay relays and thermal  switches are located under  the blue
     panel.  The time delay can be adjusted  using the DIP switches.  The thermal
     switch can be adjusted by removing the  small cap on the front of the switch
     and rotating the set screw.

5.2  Operating the Time Clock


     The time clock in the control module  operates  the pumps and the filter
     solenoids.  Each of the four channels is  independently programmable on a
     weekly cycle.  The clock has 4  to  7 day battery backup?  if this is
     exceeded so  that the clock  is either  blank  or  always reads 12:00 AM, the
     time and memory will have to be reset.  The clock has  a specified operating
     range of -4° to +1228F.  Figure 7  shows the clock face and the program
     buttons.  These buttons are normally covered by a plate.
           Auto   Auto    Aufo    Auto
          On | Off On  I Oft On ) Off On I Off
          w   xir
  Chan. 1 Chan.2 Chan.3  Chan. 4
   10   10    10    10
    NN  \  1	i
IMP
AM ^
PM Zl
*lh -
, no-no
' U U'U U
        Mo Tu  We  Th  Fr  Sa  Su
        GRflSSLIN digi 128-45/4
                           Reset    Override
                                                      Override
                                          ±lll    I/O 1
                                          o
                                         I/O  2  |/o 3  i/o  4
                                         Cancel  Wrff«  Read    h+    fTlH*
                                         Set time  Ix   .  IMP    h—    m —
      Figure 7.  Time clock with cover plate  removed,  showing clock face and
      program buttons.  The status shown is Tuesday, 8:00 AM with all four
      channels off.

-------
                                                         IMPROVE Sampler Manual
                                                                Version 2,  1988


    The  clock uses 12-hour notation, with the hour and minute  indicated
    directly and the AM/PM shown by a black  rectangle at the left.


    The  day of  the week is indicated by a black  rectangle  on the bottom.


    The  status  of each of the  four output channels is  indicated by black
    rectangles  at the top, with I-in  (on) and O-out  (off). When a channel  n is
    on,  the solenoid for filter n is open and a  sample is  being collected  for
    all  four modules.


    IPMr +lh and Ix are not  used and  should  have no  black  rectangles.  The  four
    screwqs at  the top should  all be  set to  auto.


    To set or adjust the time  of the  clock,  the  "set time" button  must be  held
    down while  pressing one  or more other buttons.   To set the time,  it  is
    necessary to difine:  the  day of  the week, the hour  (using the "h+"  and
     "h-" buttons) and  the minute  (using the  "mf" and "m-").  Be sure the AM/PM
     setting is  correct.  To  adjust the  minutes,  hold down  the "set time" button
    and press the  "rot" or  "m-n button.   To adjust for daylight savings time,
    hold down the  "set time" button and press  the "h+"  button in spring
     (beginning  daylight  savings) and  "h-" button in  fall (end of daylight
     savings).

5.3 Programming the  Clock


     Each program consists  of the  following three elements:

          1.  1  to 7  days
          2.  time of day
          3.  on/off command for 1  to 4 channels  (open/close solenoids 1 to  4)


     To enter a  program,  enter the above 3 elements,  and then press the "write"
     button.  Continue until all programs have been entered.  If you hesitate
     longer than 15 seconds between button-pushing,  the clock reverts back  to
     the time output mode and the program is not  entered.  To turn a channel on,
     press the I/O button once.  To turn a channel off, press the  I/O button
     twice.  Every command is registered immediately with  a black  rectangle  or
     as hour:minute.
     For the standard IMPROVE network,  four programs must  be entered, as shown
     in  Figure 8.  For a typical entry  press the  day, "h+" once, "nvf" once  and
     the appropriate channel I/O either once or  twice.  Note that  channels  3 and
     4 are unused.  The order of programs does not matter.  An  additional
     program can be entered at any time.
     You may cancel a program completely or modify one or  more  elements.  Press
     the "read" button until the desired program is  reached.  To cancel
     completely, press "cancel".  To change one  or more elements,  press the
     appropriate buttons and then press "write".
                                         -19-

-------
                                                      IMPROVE Sampler Manual
                                                             Version 2,  1988
The override buttons at the  top can be  used  to reverse the status of any
channel (on to off, off to on).  This continues in effect until cancelled
by another override or by being reset at  the next program time.
The override buttons may be  used during installation when calibrating the
magnehelic.  For example, pressing the  "override" button for channel 1 will
turn on the pumps and open all solenoids  number 1.
The override button may also be used if the  programs are entered after a
desired change.  Suppose a program was  entered at 8:10 that was to have
activated channel 1 at 8:00. Since the clock checks only for programs for
the present minute, channel  1 will not  turn  on.  By pressing the override
button it will then turn on. It will turn off at the programmed time.
                   — lx
            Chan. 1 Chan.3 Chan. 3 Chan. 4
             10   10   10   10
                 _      I
          IMP
          AM
          PM
Mo Tu We Th  Fr  Scr  5u
  Chan. I Chan.3 Chan.3 Chan. 4
   10   10   10  10

        \ \   II
IMP

AM
                                                Mo Tu We Th  Fr  So  Su
        program 1:  Wed,  12:00 AM, 1 on     program 2: Thu, 12:00 AM, 1 off
            Chan. I Chan.3 CSan.3 Chan. 4
             '  0   10   |0   10
          Mo Tu We Th  Fr  So  Su
                                         Chan. I Chan.3 Chan.3 Chan.4
                                          10   10   10   10
                                           NN n   /  /
                                       IMP
                                       AM ]
                                       PM __
                                       • HI —
      ^r-D'H ni\
      '^•CAHJi'J
                                                                    — lx
                                       Mo Tu We Th  Fr  Sa Su
        program 3: Sat, 12:00 AM, 2 on     program 4: Sun, 12:00 AM,  2  off
               Figure 8. Clock face for four standard programs
                                    -20-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
5.4  Electrical Circuitry
     The electrical schematic for the control module is shown in Figure 9.  The
     four outputs from the time clock (TC-1 to TC-4) each go to a relay.  One
     output from each relay goes to a terminal which is then connected to a
     solenoid and an elapsed time indicator in each filter module.  A second
     terminal from each relay goes to all of the pumps.  Thus when any clock
     output goes on then all pumps are turned on.


     The 30-minute bypass timer goes to a fifth relay.  When this goes on, all
     clock outputs are disconnected so that all solenoids will be closed.  A
     second output from the bypass relay turns on all pumps.


     Three of the four pumps have time delay relays (TDR) to produce a small
     delay between startups to prevent overload at sites where power is
     marginal.


     The numbers in square indicate the terminal number.  The colors refer to
     the wires on the six wire cables between control and filter modules and
     between control module and pump house.

-------
                                                  IMPROVE Sampler Manual
                                                         Version 2, 1988





















to
filter
modules








to
relays
•
in pump
house


110 v power
r!T~rl



f '" " Grasslin dioi 128

„ thermal /f~\
11 switch (85°F) ^)
,, thermal >• A A A
11 switch (20°F) hYavter
llOv
i i


24 v
II 30min (r-^] manual
bypass timer V8y override
ip , ii TC~I M-"i
it ii I q /
ro.o xrX
v*r\ o -r f •» x~x
II (rR)
,,TC-3 /W\
UiV
l,TC-4 /fp\
	 i,(5rVi 	 ^^^ 	
• •CR-IO t prn nrnnnn


jj^n-ic [4.^ |j|ue
' 	 4-81
j | ^
't"U_fe.__, black
1 	 2:E1
co a , — , Pump
l|OR-a . 	 	 |3l red *l
f n o TDR 	 ; _ , .
i|CR-9 i-.-ori /i nranop *? fl white
n^^*'^ |r cnrl ... _, i =; 1 nrnnn *T
f*Q II i
ii^n-ii ,jo _nr| 	 ..,,.._ 1 rl hinn *A
,,CR-I2
1 — '• 	 TTI hlnrk






'









*o
*2 tlm«»H
f /circuits


filters
C 7 nf
T c or
modules
A,B,C*D





~_* ^HIK^BAV






              Schematic for  IMPROVE Controller
Figure 9.   Electrical schematic for  control module.  TC-n is time clock
output for  channel n, CR's are control  relays, TDK's are time delay relays.
The other numbers refer to terminal  strip.  The colors refer to wire on
6-wire cable.
                                  -22-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
                                      FILTER MOtWLE
6.1  General Description
     Both fine and PMlO modules contain identical control panels with  two
     elapsed time indicators, two toggle switches, a 0-30" Hg vacuum gauge and a
     0-1" water magnehelic gauge.  They both also contain 2  solenoid valves and
     a critical orifice.  The layout of the fine filter module  is  shown  in
     Figure 5, and that of the PMlO module in Figure 6.


     The airstream enters the fine module through an inlet designed to keep out
     rain, insects and particles larger than around 15 (M.   The airstream than
     passes through a 3.66 cm metal cyclone that is identical  internally to  the
     well-tested cyclone design of Walter John  and Georg Reischl of the  Air
     Industrial Hygiene Laboratory used by the  California Air  Resources  Board.
     It has a 50% efficiency effective cutoff at 2.5 /um aerodynamic diameter  at
     a flow rate of 21.7 1/min.  The variation  of 50%  cutpoint versus  flow  rate
     is shown in Figure 10.  A 10% decrease in  flow  rate  from  21.7 1/min will
     produce a 10% increase  in cutpoint diameter.  For example, a  30%  decrease
     in flow rate to 15.3 1/min will increase  the cutpoint to  3.5  tm.
                  S  3-5
                  o
                  u
                  •H
                  e  3.0
                  I  '•'
                  o
                     2.0
                             \
                                  \
                         10            15        20
                                 Flow Rate (Ilters/min)
25
30
       Figure  10.   Diameter  of 50%  cutpoint vs.  flow rate for cyclone, from
       W.  John and G.  Reischl.
                                          -23-

-------
                                                         IMPROVE Sampler Manual
                                                                Version 2,  1988


    The PM10 module contains a standard PMlO inlet designed and built  by
    Wedding and Associates, that has a 50% efficiciency of 10 tm at  a  flow  rate
    of 18.9 1/min.  The variation in outpoint diameter with a change in flow
    rate  from 18.9 1/min is not known.  The PMlO  inlet is attached to  a filter
    cassette manifold identical to that used in the  fine module, except it  is
    mounted facing down.


    The flow rate is regulated by a critical orifice located downstream of  the
    solenoids.  The flow rate depends on  the area of the critical  orifice and
    on the pressure and absolute temperature of the  air as it enters the
    orifice. .At flow rates near 20 1/min, the pressure drop across  the filter
    can be a significant fraction of the  ambient  pressure, which can procude
    small changes in the flow rate.  This affects the system in  two  ways.
    First, since the different filter media have  different pressure  drops,  it
    is necessary to use critical orifices of different diameter  in each of  the
    four  modules.  It is not possible to  interchange filter media  from one
    module to another and  maintain  the desired  flow rate.  Second, the flow
    rate  can decrease when the filters become heavily loaded.   Based on data
    from  the SFU network this should  rarely be  a  problem.


    The flow  rate  for a given sample  is measured  before  and  after  each sampling
    period using  two methods, as described  in  the next section.   The effect of
    deviations  in  flow  rate from  the  nominal value of 21.7 or  18 1/min is to
    change  the  cutpoint  from 2.3  or 10 im rather  than produce  errors in volume.

6.2  Flow Bate Measurement


    The  flow rate  is  measured  in  two ways as  a weekly quality assurance check.
     The  first measurement is based on the equation for flow through a critical
     orifice,  and  the  second on  the pressure drop across a fixed orifice.   Both
     equations depend  on pressure  and absolute  temperature.   The pressure is
     assumed to depend only on the altitude.   (Meteorological variations in
    pressure  are  much smaller than those due to altitude;  the average
     variation is  less than 1%.)  The ambient temperature is measured at the  time
     of the weekly flow rate measurements.  In addition,  the minimum and maximum
     temperatures  for each week are measured,  allowing a slightly more accurate
     determination of average flow rate over the 24 hour sampling .period.


     The flow rate through the critical orifice depends on the temperature  of
     the air and the pressure drop in the filter and  in other parts  of the
     system up to the critical orifice.   When the solenoid is open,  the value on
     the vacuum gauge, AP, (in "Hg) measures this pressure drop.   (For the  PMlO
     module there is a relative large calibration orifice between  the
     measurement point and the critical orifice, but  this decreases  the pressure
     by around 0.2%.)  If T is the absolute temperature (°C +273) and P is the
     ambient pressure in "Hg, then the flow rate by  this first method  is


               Q,  - Q0 (1- ^f)


     where Q9 is the flow  rate for no filter, corrected to 280°K,  and  depends
     only on the area of the critical orifice.  The  temperature of 280°K
     corresponds to 458F,  a typical mean  temperature for the network.  The
     temperature term is generally very close to  unity.   (It differs from 1.00
     by less than 5% for temperatures between -5%"F  to 96BF.) the  critical

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988


orifice is chosen to give a flow rate of 21.7 1/min (fine modules) or 19.8
1/min (PM10) for a typical filter.  The constant Q0 is measured during the
system calibration at the site.


The constant Q. is designed to give the appropriate flow rate  for a given
filter type.  The desired value depends solely on  the average  temperature
of the air at the critical orifice and is independent of the altitude of
the site.  If primes indicate the values at another site (such as at
Davis), then
          Q0 - V (r)   ,

In order to account for variations from the ideal equation,  the  equation
for Qx can be written in  the more general form
where at and b  are determined by varying  fiP  in  the  calibration protocol
and measuring Q1 •


The second measurement of the flow  rate uses  a magnehelic  gauge to measure
the pressure drop across a  fixed orifice that is large compared to the
critical orifice.  For the  fine modules, the  fixed orifice is provided by
the cyclone, while for the  PM10 module, it is provided by  an orifice built
into the system just before the critical orifice.  If 5P is the pressure
drop across the orifice and the pressure at the  front of the orifice is at
atmosphere then the flow rate is


          Q2- C(5P)b  (|)4

where C is a constant depending on  the geometry  of the orifice and b is 4
if the  flow is laminar.  Note that  the pressure  drop across this orifice  of
SP .is less than 1% of the AP across the filter.   Assuming  the sampler is  to
remain  at the same altitude, the equation  can be rewritten in terms of two
parameters
Again  the  parameters  a2  and b2  are measured during the system calibration.
The value  of  a2  is  around 40 and b2  slightly less than 0.5.


For the  PM10  orifice, the equations are slightly different, because the
pressure at the  front of the measuring orifice is reduced by the drop
across the filter,  so


           Q2-C(«P)b  (^(l-^)4

Eliminating AP,  the pressure drop across the measuring orifice can be
written  in a  parameterized form as

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2,  1988
     The value of a2  is around 40 and b2  slightly less than 1.


     The average flow rate for the sampling period will be a weighted average of
     Qx  and Q  for the initial and final readings.  The flow rate will be
     corrected for changes in temperature using the mean temperature for the
     week, calculated from the minimum and maximum.

6.3  Flow Rate Audit


     The flow rate audit is a check of the flow rate using an audit device
     placed between a filter cassette and a solenoid, for each of the four
     filter modules.  The audit may be conducted by a third party auditor, using
     any flow rate meter that has a low pressure drop, by field personnel using
     an orifice meter supplied by Davis and mailed to and from the site, or by
     Davis personnel using either an orifice meter or a dry test meter.  A mass
     flow meter is an acceptable device.  The orifice meter will be calibrated
     at Davis using a spirometer and dry test meter.  The system magnehelic and
     vacuum gauge will be read simultaneously in order to compare the flow rate
     measurements.  There are two purposes to the audit:  (1) to compare the
     flow rate measured by the module with the audit flow rate, in order to
     determine the accuracy in the volume of air collected, and (2) to compare
     the audit flow rate with the nominal flow rate  (21.7 or  18.9 1/min), in
     order to determine the accuracy in the particle sizing of the cyclone or
     PH10 inlet.  For mail audits the actual flow rates will  be calculated at
     Davis using the information on the logsheet.


     The equipment needed for the audit will be a flow rate meter with  1/2 inch
     standard compression fittings at either end  (male fitting at upstream end
     and female fitting at downstream end), and four filter cassettes with hoses
     containing standard  filters.  If the standard  filter cassettes are  not
     available, it is possible to use the filters at the site provided  for  the
     normal sampling.  A  logsheet will be provided  for the audit, as  shown  in
     Figure 11.  Record the information at the top  of  the sheet.


     The  following procedure  is  to be used for the  audit.   First, using the
     override button  for  channel  1 on the time clock (see  section  5.3),  turn on
     all  the pumps and open all  first solenoids.


     For  each of  the  four modules do the  following steps.

     — Attach the audit device  to solenoid  1  of module.

     — Attach the  filter  cassette with  the  appropriate filter for that module to
        other  side of  the  audit device.

     — Record the  reading for the audit device and for both module gauges.


     Any audit device added to the  system will decrease the pressure at the
      front of the critical  orifice  and thus decrease the flow rate.   It is
      important that the pressure drop produced by this device is small enough to


                                         -26-

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
produce no significant effect.  Suppose the pressure drop for the total
system up to the critical orifice without the device is AP and the device
produces an additional SP.  The equation for the fractional decrease in
flow rate is

          AQ   (P-AP) - (P-AP-&P) m ,    &P
           Q         (P-AP)             P-6P

That is, the drop produced by the device must be small compared  to the
pressure at the front of  the  orifice.  The worse case (smallest  P-AP) would
occur at high altitudes and large pressure drop across the filter.  For the
network this would be with P-AP approximately 15 "Hg - 200 "H20.  A
pressure drop of 2 "H,0 would cause a  1% decrease.  The standard
calibration devices have  a pressure drop of 0.5 "H20 at 22 1/min, causing
an error of 0.25%.
                   1HPROVI   I A H F L t *   AUDIT  LOO
ntt mm
BVTB
SAMPLER SERIAL NUMBER
(UDITD) BT
•nnPCRATORB C + Z73 • K
M.TITODB
ttJDIT NETCT NUMBER
ttJDIT coerrst •
b -
M.TITUDB f*CTOR (A)

DEHSITI FACTDRl (T/398P * A
flow rate at «ite - deiutty (actor * 10s(«plb
racxjuf A
s-Xvic.

put? v
•sonehellc
"•HjO

VCUUB

small
9*sr

flow tat* by audit devic* -
flow r«t» by tyiten na^nehelle *
flow tat* by trail gauqw •

HDDUUE C
devic*

pump v»euu» _
gyste*
•agnehalie
*v

imall
"ST

flow cat* by audit d*vle* -
flow tat* by cysten ragnctolie -
flow tat* by email 98119* •







HCCULK B pmv vacuum

audit system inwll
d«vlc* nBTJiehellc ^ 9fuj»
H*0 H^

flow rat* by audit devic* -
flow rat* by *y*teit magnehellc -
flow rat* by caall gauge •

HOCULS D pimp vacuui 	
audl t system small
device BemeHellc 9»uoe
"VH20 EH|

flow rat* by audit devic* •
flow rat* by tyiten BBgnehellc •
flow rat* by mall 9*09* •
                       Figure  11.   Flow rate  audit log.
                                     -27-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
6.4  Flow Rate Calibration
     The flow rate calibration is similar to the flow rate audit, except that
     the flow rate is adjusted to several values using a valve to imitate a
     variable filter.  The calibration will be performed by Davis personnel and
     includes calculation of the flow rates on site.  The only additional
     equipment is the valve assembly.  The calibration logsheet is shown in
     Figure 12.


     At the top of the calibration logsheet, record the site name, sampler
     serial number (located inside door of controller module), date,
     temperature, your name, altitude, and altitude (pressure) factor relative
     to Davis.  These last two numbers should be read from Table 2 or Table 3.
     Using the temperature (°K) and the altitude factor, calculate the air
     density factor relative to Davis.


               C - f)1* * (altitude factor)
     and record on the sheet.  This factor is to be used in several places  to
     convert flow rates from Davis values to site values.  Record the
     calibration coefficients for the meter on the sheet.  The coefficients
     relate the flow rate at the site and the measured pressure  from the
     calibration magnehelic:

               Q - C 10a (Sp)b

     where C is the density factor.


     Using the override button for channel 1 on the  time clock  (see  section
     5.3), turn on all the pumps and open all first  solenoids.


     Using the coefficients for the calibration orifice/magnehelic  calculate the
     APc's for the pre-set flow rates on the log using  the relationship
                SPC  -


      Record the desired meter settings on the calibration log sheet.
                                         -28-

-------
                                                            IMPROVE Sampler Manual
                                                                    Version 2,  1988
             IMPROVE   SAMPLER   CALIBRATION   LOG
SITE NAME
                 DATE
SAMPLER SERIAL NUMBER
CALIBRATED BY
                 TEMPERATURE
                 ALTITUDE
                   C + 273 -
CALIB METER NUMBER
CALIB COEFFS!  a -
                 ALTITUDE FACTOR (A)
b -
DENSITY FACTOR:  (T/296P * A
    flow rate at site - density factor * 10a(«p)b
MODULE A pump vacuum 	
flow rate
1/mln
max
21.7
19.5
16.3


calib
magnehelic
H2°






system
magnehelic
H-0






small
gauge
*Hg






magnehelic (log fit) r -
log flow- 4 *log(8P)
small gauge (linear fit) r -
flow- - *(AP)
MODULE C pump vacuum
flow rate
1/min
max
21.7
19.5
16.3


calib
magnehelic
"H2°






system
magnehelic
H2°






small
^^






magnehelic (log fit) r -
log flow- + *log(SP)
small gauge (linear fit) r -
flow- - MAP)
   max - readings with no filter
MODULE B pump vacuum
flow rate
1/mln
max
21.7
19.5
16.3


callb
magnehelic






system
magnehelic
H2°






small
V9






magnehelic (log fit) r -
log flow- + *log(«P)
small gauge (linear fit) r -
flow* - MAP)
MODULE D pump vacuum
flow rate
1/min
max
18.9
17.0
14.2


calib
magnehelic
H2°






system
magnehelic
H2°






small
gauge
*Hg






magnehelic (log fit) r -
log flow- + *log(8P)
small gauge (linear fit) r -
flow- - MAP)
                         Figure 12.   Flow rate calibration log.
                                           -29-

-------
                                                    IMPROVE Sampler Manual
                                                           Version 2, 1988
            Table 2. Density factors at IMPROVE sites
                      (factor relative to Davis)

                factor - I P(Davis) / P(site) ] *
 site
alt    P    factor
feet  "Hg
site
alt    P    factor
feet  "Hg
Acadia
Arches
Big Bend
Bridger
Bryce Canyon
Canyonlands
Chiricahua
Crater Lake
Denali
Glacier
Grand Canyon
Great Smoky
470
5500
3460
8000
7950
5950
5400
6479
2100
3200
7100
2700
29.4
24.4
26.4
22.2
22.2
24.0
24.4
23.5
27.7
26.6
22.9
27.1
1.007
1.106
1.063
1.159
1.159
1.116
1.104
1.126
1.037
1.059
1.140
1.049
Guadalupe Mtns
Jarbidge
Mesa Verde
Mount Rainier
Petrified For
Pinnacles
Rocky Mountain
San Gorgonio
Shenandoah
Weminuche
Yellowstone
Yosenrite
5446
6200
7200
1400
5500
1040
7910
5618
3515
9140
7750
5250
24.4
23.7
22.8
28.4
24.4
28.8
22.3
24.3
26.3
21.2
22.4
24.6
1.105
1.121
1.142
1.025
1.106
1.018
1.157
1.108
1.065
1.186
1.154
1.101
                 Table 3.  Density factor  vs.  altitude
                          (factor relative  to Davis)
                 factor - {  P(Davis)  / P(site)  ]
alt    P    factor
feet  "Hg
0
200
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2600
2800
3000
3200
3400
29.9
29.7
29.2
29.0
28.8
28.6
28.4
28.2
28.0
27.8
27.6
27.4
27.2
27.0
26.8
26.6
26.4
0.998
1.002
1.010
1.013
1.017
1.021
1.025
1.028
1.032
1.036
1.039
1.043
1.047
1.051
1.055
1.059
1.063
alt
feet
3600
3800
4000
4200
4400
4600
4800
5000
5200
5400
5600
5800
6000
6200
6400
6600
6800
P
"Hg
26.2
26.0
25.8
25.6
25.4
25.2
25.0
24.8
24.6
24.4
24.3
24.1
23.9
23.7
23.5
23.4
23.2
factor

1.067
1.071
1.075
1.079
1.083
1.088
1.092
1.096
1.100
1.104
1.108
1.113
1.117
1.121
1.125
1.130
1.134
                                    alt     P     factor
                                    feet   "Hg
7000
7200
7400
7600
7800
8000
8200
8400
8600
8800
9000
9200
9400
9600
9800
23.0
22.8
22.7
22.5
22.4
22.2
22.0
21.8
21.7
21.5
21.3
21.1
21.0
20.8
20.7
1.138
1.142
1.146
1.151
1.155
1.159
1.164
1.168
1.173
1.178
1.183
1.187
1.192
1.197
1.201

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988


For each of the three fine modules do the following steps.

—Attach the calibration meter to solenoid 1 of module.

—Attach the filter cassette with the appropriate filter for that module to
  other side of the calibration meter, in the same manner as for the audit.

—Calculate the flow rate using the calibration magnehelic using the
  relationship for the calibration magnehelic system

            Q - C 10a («p)b


—If the flow differs from 21.7 or 18.9 by more than a  few percent, replace
  the critical orifice until a better value is obtained.

—Replace the filter cassette with the valve assembly  (without  filter)
  designed to act as a variable filter.

—For each of the four points adjust the valve to give  the desired  SPc on
  the calibration magnehelic and record both module gauges.  If desired do
  other points.

—For modules ABC, perform a linear regression fit for  the magnehelic using
  the log relationship

            log A - a2 + b2 log (SP).


—For module D, perform a regression fit  for the magnehelic  gauge using  the
  linear relationship

            Q - a2 + b2 5P

  Check that r is greater than 0.990.  Record  the  coefficients.  For
  modules ABC, b2 is slightly less than 0.5 and a   is  around 1.5.   For
  module D, the coefficient a2 should be  much  smaller  than b2.

—Perform a regression fit for the vacuum gauge using  the  relationship

            Q - at + bx AP

  Check that -r is greater than 0.990.  Record the coefficients.   The
  coef f icient "bj will be negative, and a:  is much  larger than bt.

—For modules ABC, mark the 21.7  readings on both  gauges in  green ink and
  the 23.9 and 19.5  readings on both gauges  in red ink.  For module D,  mark
  the respective readings are 18.9, 20.8  and 17.0.
                                    -31-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
6.5  Electrical Circuitry
     The electrical schematic for a filter module is shown in Figure 13.  The
     connection with the control module is through a six wire cable.  The
     solenoid valve (SV) and elapsed time indicator (ETI) will be activated if
     either the appropriate clock output is on or if the appropriate toggle
     switch is turned on.  When the toggle switch is released if will always
     return to the normally open position, giving control to the time clock.
                   I  I Controller terminol

                   /\  Filter module terminal

                   TS  Toggle switch
       Figure 13.   Electrical schematic for filter module.
       ETI - elasped time indicator, SV - solenoid valves.
       the wire on the 6-wire cable.
TS - toggle switch,
The colors refer to
                                          -32-

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
6.6  Nitrate Denuder
     The nitrate denuder is installed in the inlet tube of module B.  As shown
     in Figure 14, it consists of a series of four concentric aluminum cylinders
     coated with NaC0.  The total surface area is 0.3m3.  The nitrate gases
(N02 ,  N03 )  will diffuse to a wall and be captured, while the particles pass
                   ,3.
                  ll diff
     through, because of the much smaller diffusion rate for particles.  At -a
     regular interval (perhaps a year) the entire cylinder must be  replaced and
     returned to Davis for recoating.  Normally this change will be performed by
     Davis personnel.
     The denuder will fit inside the inlet tube.  The operator need  only open
     the cap at the base of the inlet, pull out the old denuder,"  replace it with
     a new denuder and recap the inlet.  More detailed procedures will be
     available when the system is developed.


     To replace the denuder, it is necessary to remove the  cap at the base of
     the inlet and remove the screw-pin holding the denuder in the inlet.  A
     handle will be screwed into the base of the denuder  (from below) and the
     denuder pulled down.  The new denuder will be pushed up using the same
     handle and the screw-pin inserted.  The handle will  then be  removed and the
     cap replaced.  The spent denuder will be recharged at  Davis.
                                               Tutinj f

                                                alum!


                                               M.l«: Inner
                                    fbread
                             Figure  14.   Nitrate denuder.
                                         -33-

-------

-------
                                                          IMPROVE Sampler Manual
                                                                 Version 2, 1988
                                   7.  PUMP HOUSE
7.1  General Description
     The pump house is an all-weather house measuring 3'x2'x4'.  Inside there
     are four pumps with surge tanks, a box of 4 switched 120V outlets, a box of
     4 unswitched 120V outlets, a fan with thermal switch, a lamp socket with
     manual switch and heat tape with thermal switch (where necessary).  The
     wall of the pump house is shown in Figure 15.  Attached to the outside of
     the house is a weather-tight electrical box contain four power relays.


     The external 120V line is attached to the terminal strip inside the relay
     box.  120V lines are then carried to the control module, to the unswitched
     outlet box and through the power relays to the switched outlet box.  A
     6-wire cable from the control module connects to the control terminals of
     the power relays.

7.2  Pump Specifications


     The pump is a 0.5 HP oilless diaphragm vacuum pump, Cast model DAA-V132-GB.
     The pump can draw 38 1/min at 15 "Hg and 17 1/min at 20 "Hg.
                                    unswitched
                                    120V outlet
                                               switched
                                               120V outlet
O                                            pass-through
                                            to relay box
                      Figure 15.  Layout of wall of pump house.
                                        -34-

-------
                                                        IMPROVE Sampler Manual
                                                               Version 2, 1988
7.3  Electrical Circuitry
     The  electrical schematics for  the pump house are shown in Figure  16.  The
     output  from the control module activates a 20A power relay which  provides
     120V to a switch outlet used by one of the pumps.  For safety,  both the hot
     and  neutral 120V lines are switched.  In addition to the four switched
     outlets, the pump house has four unswitched outlets (not shown).
               CONTROL
               CIRCUIT
                                 red
                                orange
                                green  ^^
                                blue
                             61	U
                     Controller   lerminals
white

POWER
CIRCUIT

no v
[[thermal switch *Tfo"n)
i, thermal switch he?te/
II
umanual
n
II CR"A
U
i, CR-B
II
1 1 CR ~C
II
" V V V ~
switch >~--X
.CR-AH
ii
,iCR-Bz
II '
pCR-Cz
II '
,,CR-C3
II
        Figure  16.  Electrical schematic for pump house.  CR - control relays

-------
                                                   IMPROVE Sampler Manual
                                                          Version 2,  1988
The relays are located in a weathertight box attached to the outside of the
pump house.   Figure 17 shows a schematic for the  relay box.  The power
cable enters  the box from the bottom and is attached to the terminal strip
as shown.
     PUMP  OUTLETS
    (inside of pump house)
                      POWER
                                  RELAYS IN CONTACTOR  BOX
              Figure 17.   Electrical  schematic for relay box
                                   -36-

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
                             8.  SAMPLER STAND
The five fiberglass enclosures must be mounted on a stand.  Because of the
difficulty in transportation, and because individual sites may have special
needs, the stand is generally built locally.  At some sites the sampler is
located outdoors, so that a freestanding wooden stand is appropriate.  At
some sites where an air quality building already exists, the stand consists
of support boards attached to the walls of the building.


The stand shown in Figure 18, often built of redwood, is used at most sites
that do not have an air quality building.  To insure the stability of the
unit, the legs are fastened to concrete piers and imbedded in the ground.
In situations of uneven ground, it is desireable to have legs of different
length, in order to keep the horizontal beams level.  Holes with 12 inch
spacing must be drilled in the horizontal beams in order to mount the
enclosures.


The stand is also to have sunshields above the enclosures.  A short shield,
2 feet long, goes above the control module.  A long shield, 104 1/2 inches
long, goes above the four filter modules, with holes cut for the inlets.
The shields are attached to the stand using angle brackets.  It is
important to orient the stand so that the fronts of the enclosures do not
face the afternoon sun (west).


If the sampler is mounted inside a shelter, it is usually necessary to use
longer inlet stacks.  The four holes in the support board for each module
must be 12 inches apart horizontally and 18.75 inches apart vertically.
Adjacent modules should not have the mounting holes closer than 8 inches
apart.
                                    -37-

-------
                                                    IMPROVE Sampler  Manual
                                                             Version 2,  1988

Provide ? iddttlonil redwood el«nkt to t>* ui«i)
M sunshlotdt (to be «tt«ched by br*ek«tfl)t
I" « 17" « J4" «nd 1" > 17" i 104V
                                                  END VIEW
     Figure 18.  Outdoor stand for IMPROVE sampler
                             -38-

-------
                                                     IMPROVE Sampler Manual
                                                            Version 2, 1988
                            9.  Acknowledgments
The following members of the Air Quality Group or Crocker Nuclear
Laboratory were responsible for the development of the IMPROVE sampler.


Sampler design and evaluation;  C. Goodart, R.A. Eldred, T.A. Cahill,
P.J. Feeney, L.K. Wilkinson, P. Beveridge, T. Essert (electrical), 0. Raabe
(flow rate, denuder), S. Teague (flow rate, denuder), B. Pedersen
(denuder), K. Bowers (S02 collection)


Sampler construction and installation:  R.A. Eldred, C. Goodart,
P. Beveridge, I. Wittmeyer, P. Wakabayashi, D. Everitt, J. Cordova,
B. Matsumura, S. Eldred, P. Dyer, M. VandeWater, E. Steen,
D. Cecil (purchasing).


Special field studies (WHITEX, SCAQS):  L.K. Wilkinson, I. Wittmeyer,
P. Beveridge, M. Surovik, P. Wakabayashi, D. Everitt, J. Cordova,
B. Matsumura, H. Miyake, B. Perley, S. Eldred, T.A. Cahill, R.A. Eldred,
T. Tanada, D. Orr, C. Semantis, J. Cooper, E. Steen, H. Miyake,
K. Mitchell, 0. Beckmann, F. terVeer, W. Reeves, B. Nicolet, K. Bowers,
C. Cahill
Documentation:  R.A. Eldred, J. Hancock, C. Goodart, C. Baro.


The controller modules were fabricated by ENDECO Controls of Rio Linda.
Most of the machine work was done by Streetman Precision of Cameron Park
and G's Machine of Placervilie.
We would like to thank the UC Davis purchasing department  for  their
considerable help in reducing costs, especially Mari Harrod.


We would like to thank Walter John of Air Industrial Hygiene laboratory  for
the plans for the cyclone.


We would like to thank Marc Pitchford, EPA-Las Vegas and a member  of  the
IMPROVE committee for the many helpful discussions  concerning  design.

-------
IMPROVE SOP APPEND                                        JULY 1989
        APPENDIX 3:   Gravimetric Mass Startup Procedures






                     A.   Cleaning and Calibration of the Electrobalance






                     B.   IMPROVE Gravimetric Controls

-------
IMPROVE SOP APPEND                                        JULY 1989
        APPENDIX 3:  Gravimetric Mass Startup Procedures


                     A.  CLEANING AND CALIBRATION OF THE ELECTROBALANCE
The first work action at the beginning of each day and immediately after lunch
period is to clean and calibrate the electrobalance.   It will also be
recalibrated if the balance fails a "zero" test that is performed
periodically.  The procedures described here are outlined in the Cahn 31
Electrobalance Instruction Manual in Section 4, Operations.

A Balance Log is maintained for each balance.  All of the calibration results
are to be recorded in the Balance Log.  In addition,  all significant events
concerning the balance and any balance maintenance other than routine are to
be recorded in the Balance Log.

The steps for cleaning and calibration are as follows.

     1.  Clean the metal and plastic forceps with ethanol and a Kimwipe.

     2.  Carefully remove each balance pan by inserting closed forceps below
         the wire yoke of the pan bail and lifting the hook from the eye.   To
         protect the hangdown and hangdown loop, do not grasp the bail or hook
         assembly with the forceps.  Gently rest the pans on a. fresh clean
         Kimwipe.

     3.  Remove the antistatic ionizing strip from the weighing cavity.

     A.  Clean and deionize the inside of the balance cavity using a
         cotton-tipped applicator wetted with ammonia.  Gently brush the
         exposed surfaces of the cavity, taking care not to disturb the
         hangdowns extending from the top surface.   Gently close the glass
         slide door and gently brush the outer surface with the ammonia tip to
         control static change.

     5.  Clean the top surfaces of the balance pans with a cotton-tipped
         applicator wetted with ethanol.  The entire surface lust be clean and
         dry.  Do not use any solvent other than ethanol on the pans.

     6.  Clean the top surface and the strips of the antistatic ionizing units
         by gently rubbing with a cotton-tipped applicator wetted with
         ethanol.

     7.  Replace the clean ionizing unit in the center back of the balance
         cavity.

     8.  Gently return the balance pans to their hang down loops.   Use the
         bail- lifting procedure previously described.  Do not put any stress
         on the bail or hang down loop.

-------
IMPROVE SOP APPEND                                        JULY 1989
     9.  Record the time and the balance temperature from the top surface
         thermometer.   Read the mass when there is nothing in the pan;  this
         is the "zero" mass.  It should be within 0.010 mg of 0.000.  This
         value may be exceeded if the ethanol has not completely evaporated;
         in this case, the reading will drift toward 0 as the ethanol
         vaporizes.  Record the "zero" mass.

    10.  Depress the tare button.  This forces the "zero" mass to be exactly
         0.000.  Record this 0.000 value.

    11.  Momentarily ground yourself by touching the electrostatic mat.  Use
         plastic forceps to remove the 200.000 mg class 1.1 (Class M)
         calibration mass from its container.  Gently place it in the center
         of the balance pan and allow the mass reading to stabilize and stop
         decreasing.  This will require from 30 to 60 seconds depending on how
         smoothly the mass was released onto the pan.  When the reading
         stabilizes, record the mass.  Next,  press the calibrate button,
         forcing the balance to indicate 200.000 mg.  Record this value.
         Gently remove the calibration mass and replace it in its container.

    12.  Remove the standard 50.000 mg mass from its container using plastic
         forceps.  Gently place it in the center of the balance pan.  After
         the reading stabilizes, record the mass in the Balance Log.  Record
         your initials to indicate your calibration.

    13.  Allow the balance to return to "zero".   Compare the zero value and
         the value determined for the 50.000 mg mass to previous values.  If
         they exceed ±2 micrograms,  repeat the procedure.  If greater
         variations are observed, see the laboratory supervisor.

    14.  After cleaning and calibration, the electrobalance is available for
         routine determination of mass.

    15.  Clean the work surface of the antistatic mat with a clean brush or a
         Kimwipe dampened with water.  Do not use ethanol as it will damage
         the mat surface.

    16.  Clean the metal forceps and the antistatic ionizing strips by wiping
         them with ethanol and a clean Kimwipe.

    17.  The next step is to process the control filters, following the
         procedures described under IMPROVE Gravimetric Controls.

    18.  On a random basis, but at least semiannually, the laboratory
         supervisor will request a comparison of the normal calibration
         standards with a second set of reference standard masses maintained
         by the laboratory supervisor.  After calibration, measure these
         200.000, 50.000,  and 20.000 mg standards and report their masses to
         the supervisor.  These results are used to verify the integrity of
         the electrobalance and the standard masses used in daily
         calibrations.

-------
IMPROVE SOP APPEND                                        JULY 1989
                       B.  IMPROVE GRAVIMETRIC CONTROLS

The gravimetric control program is used to determine the precision of the
gravimetric analysis and the mass artifact associated with the storage of
teflon filters in cassettes.
                                                                          *
The procedure for a control filter is .to measure its mass twice, allow it to
remain in an IMPROVE cassette for approximately 35 days, and then remeasure
the mass twice.  The filter is then put in a permanent slide mount and
archived in a slide tray.  A computer program determines the precision of two
pairs of mass measurements and calculates the mass artifact associated with
the cassette by subtracting the mean initial mass from the mean final mass.

The procedures for the beginning of the day, immediately after cleaning and
calibrating the balance, are as follows.


     1.  Obtain a clean 25 mm teflon filter from the prepared stock maintained
         for SFU samplers.  This filter is in a petri dish identified with an
         additional "S".  This filter is known as the "PRE" filter.

     2.  Select the next premade Control Identification Tag kept at the front
         of the Balance Log (e.g.  1-465) and attach it to the petri dish.

     3.  Locate the oldest IMPROVE Control cassette, indicated by the lowest
         control number and place it in the electrobalance work area.  This
         filter is known as the "POST" filter.

     4.  Download the POST cassette and place both the PRE and POST filters on
         antistatic strips.

     5.  Prepare an IMPROVE Control tag for recording all the values required
         by the Control database.   This tag should have three columns and have
         the following form.

                 Pre         "new" control ID        RePre

                 Post        "old" control ID        RePost

                 Date 	 Technician 	 entered 	


     6.  Measure the PRE filter,  record its mass, and place the filter into a
         clean petri dish.  Affix the identification tag for this filter on
         the petri dish.

     7.  Measure the POST filter,  record its mass,  and place the filter into a
         clean petri dish.  Remove the identification tag from the cassette
         and affix it on the petri dish.

     8.   Store both petri dishes,  the control tag,  and the empty cassette
         until required for the afternoon measurements.

-------
IMPROVE SOP APPEND                                        JULY 1989
The procedures for the beginning of the afternoon, immediately after cleaning
and calibrating the electrobalance,  are as follows.

     1.  Remove the PRE and POST filters from their petri dishes and place
         them on antistatic strips.

     2.  Prepare an 18x24 mm slide mount to receive the POST filter.  Write
         the following information on the white side of the mount using a
         felt-tip pen.

         a.  IMPROVE control number (e.g. 1-430),
         b.  The corresponding S number (e.g. S-1391),
         c.  Today's date (e.g.  5/17/89),
         d.  The number of the next empty position in the 40-position archive
             slide tray (e.g.  13).


     3.  Remeasure the PRE filter, record its mass (REPRE),  and mount the PRE
         filter in the empty IMPROVE control cassette, following the standard
         protocol.
                                             n
         Center the drain disk and the 2.2 cm^ mask.  Place the filter on the
         disk and mask with the smooth side up.  Replace the lock ring and
         cassette cup and protective red cover.  Affix the PRE identification
         tags to the cassette  and place it at the end of the queue of IMPROVE
         control cassettes.

     4.  Measure mass of the POST filter and record its value (REPOST).   Mount
         the filter on the slide mount with the smooth side facing the black
         half of the mount.   Place the mount in the next empty slide tray
         position.   Replace the tray cover, and return the tray to its storage
         place.

     5.  Input all the control data into the database using the program
         IMPCONT.  Enter the morning masses, PRE and POST, followed by the
         afternoon masses, REPRE and REPOST.  The program calculates and
         presents the precision of both pairs and the 35-day mass gain.

     6.  Precision values should not exceed ±5 micrograms.  If they do,  select
         the no-entry function of IMPCONT and correct the gravimetric problem.
         Precision of ±5 to 10 micrograms generally indicates a calibration
         problem.  Greater values indicate incorrect identification problems.

     7.  When verified, input  the data into the database permanently by
         selecting the "yes" input.   This completes the daily control actions.

     8.  Every month, the laboratory supervisor will activate a feature of
         IMPCONT that calculates and summarizes all precision and mass
         artifact values for the selected month.  These results are appended
         to the data for previous months and printed out in hard copy.

-------
IMPROVE SOP APPEND                                        JULY 1989
                      APPENDIX 4:   LIPM Startup Procedures
       CALIBRATION OF THE LASER INTEGRATING PLATE METHOD (LIPM) SYSTEM


     1.   Do not begin the calibration unless the LIPM system has been on for
         at least two hours,  allowing the laser to stabilize.  The LIPM system
         is programmed to be  on daily between 0600 and 1800.  Verify that the
         timer is indicating  the correct time.   If necessary, set the timer to
         the correct time by  rotating ahead clockwise.

     2.   Routine operation of the system requires no adjustments except a
         vernier multiplier setting.   Verify that the Oriel Detection System
         Model 7022 has the following startup settings:

         a.  Multiplier is set on the xlO~" level
         b.  The 45 volt switch is ON
         c.  Ambient supress  is OFF
         d.  Response is set  to MEDIUM
         e.  Zero Set is OFF
         f.  Verify that the  collimator aperture is installed.
     3.   Blank off the input laser light by moving the Beam Attenuation lever
         to "closed." Verify that the detector digital readout is 0.000
         (zero).   If not,  adjust the Zero Set knob until the detector
         indicates 0.000.

     4.   Open the Beam Attenuation lever and push the slide changer in all the
         way.   Make sure the tray-covering flap is closed.   The detector
         should indicate 0.000.   If the measurement exceeds +0.005, check for
         an open cover flap or for a physical malfunction of the changer.
         Notify the laboratory supervisor.

     5.   Pull the slide changer all the way out.   Use the vernier multiplier
         extension arm to set the detector reading to the protocol calibration
         level of 0.750.

     6.   Mount the LIPM standards tray of 10 standards in the system.   These
         standards test all parameters of slide orientation and include a
         spectrum of actual filters and aerosols.

     7.   Record the date,  your initials, and the  values obtained for each of
         the standards slides on the LIPM calibration logsheet in the LIPM
         protocol folder.   Record the values for  standards  1 thru 10.
         Periodically verify between slides that  the detector reference level
         remains  at 0.750  with the slide changer  fully out.

     8.   Compare  the values obtained with previous values reported in the
         logbook  and the highlighted standard values.   If values differ by
         more  than ±0.003,  repeat the procedure.   If deviations still exist,
         consult  with the  laboratory supervisor.

     9.   The LIPM system is now ready for routine  analysis  of filters.

-------
      IMPROVE SOP APPEND                                     JULY 1989
                    APPENDIX 5:  PIXE/PESA Procedures
                             PIXE RUN DIRECTIONS


1.   CONTROL ROOM OPERATIONS

    a.   Before going into CAVE (close target)

        — Hit 'CLOSE' button on the lower left hand side of control panel.
        — Hold down 'TARGET' button (in the middle of the control panel)
           until 'CLOSE' button turns amber. (Takes 5 seconds)
        — Hit 'CLOSE' Button to Cave (on top right hand side of control
           panel)
        — Remove key labeled North Cave Door No. 3.
        — Close pumps (Move toggle from HIVAC to middle position)

    b.   After coming out of CAVE (open target)

        — TURN ON roughing pump (toggle to rough). When vacuum reaches 80
           microns, turn toggle to HIVAC.
        — Put back key
        — Hold down 'OPEN' button to cave, in the center section of the
           control panel next to the 'TARGET' button. (Takes 30 seconds)
        — Hit 'OPEN' button on upper right hand side of panel until button
           lights up
        — Hit 'OPEN' button in lower left hand side of panel after you check
           with operator to make sure vacuum pressure is low enough.


2.  TO MOVE TRAY OR STRIP FRAME AT DISPLAY PANEL

        — Switch to manual
        — If slide is still showing on TV screen (and "slide in" light is
           on), hit black button next to 'Slide Drive' and slide will move
           back into tray, (and "slide out" light will be on).
        — Hit toggle switch next to 'Tray Drive' toward 'F' (forward) or
           'R' (reverse), depending on the direction you wish to move to
           tray or strip.
        — Hit black button next to tray drive and tray or slide will move
           one position at a time. (Remember, only move the tray when the
           slide is "out".)
        — Hit black button next to 'Slide Drive' and a slide should be
           showing on the TV screen. ("Slide in" light will be on)
        — Switch back to automatic and hit 'Enter Auto' on the display panel.
        — Enter master tray #, hit return, enter position # hit return, and
           analysis will begin again.


3.   TO CHANGE TRAYS OR STRIP FRAMES

    1)    Press the green button on the metal box located on the beam line
         before the slide changer. This is just to make sure the gate valve
         is really closed.

    2)   Turn blue handle on slide changer slowly to let air into tube.

    3)   a) FOR TRAYS

        — Open bottom of tube and pull out the 3 trays slowly, making sure
           the slides do not fall out of trays.

-------
    — Replace bottom
    — Place an "X" on tray boxes that were just analyzed
    — Open top of tube and slide new trays in with position # facing up.
       Make sure trays attach in tube and are not at an angle- this may
       cause the slide changer to jam.
    — Replace top

    b) FOR STRIPS

    — Move strip back to position #1 by holding both black buttons next
       to 'DRIVE' and 'DIRECTION' on side of slide changer
    — Remove plastic cover over stripper
    — Take out strip, put an 'X' on white side of frame and put into
       frame box
    — Put next strip to be analyzed into stripper with the blue side of
       the frame facing" down and the green dot on the left hand side.
    — Replace plastic cover-don't make screws too tight

4)  Use red button in back of slide changer to move 1st slide into
    place- you know you are in the right position when the #1 slot in the
    first tray is covered by the slide changer.

5)  Turn blue handle over slide changer to seal vacuum

6)  Push white button in back of cave and close the cave

-------
IMPROVE  SOP APPEND                                                JULY  1989
                  APPENDIX 6:   Ion  Contractor Procedures  (RTI)
      RESEARCH   TRIANGLE  INSTITUTE


                                STANDARD OPERATING  PROCEDURE:

                                   ION ANALYSIS OF  FILTERS


                                            for


                                  THE NATIONAL PARK SERVICE
                                        Prepared by

                                 Research Triangle Institute
                         Research Triangle Park,  North Carolina  27709
                              NPS Contract No.:  CX-0001-6-0007
                                        Prepared for

                               U.S. Department of the  Interior
                                    National P?rk Service
                                   Washington, D.C.  20013
       POST OFFICE BOX 1219.4  RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709-2194

-------
                        STANDARD OPERATING PROCEDURE
                           FOR NPS FILTER ANALYSIS


     The procedures followed by the Research Triangle Institute (RTI) in the
analysis of National Park Service (NPS) samples under Contract No. CX-0001-6-

0007 are summarized below:

     1)   RTI receives filters in lots of approximately 100 to 200
          each.  Sample information is entered onto sample custody
          logsheets upon receipt.

     2)   Upon receipt, the samples are stored in a locked labora-
          tory.  After the samples have been desorbed, the extract
          is stored in a refrigerator in a locked sample custody
          room.

     3)   Within 30 days after sample receipt, Cl", N02, NOg, and
          S0$~ analyses are performed on each sample by ion
          chromatography.

     4)   In addition to analyzing the field samples, RTI partici-
          pates in EPA-sponsored analytical performance audits which
          include the analysis of reference precipitation samples.

     5)   The data are reported, by filter lot, to the National Park
          Service in hard copy and on floppy disk in ASCII format.

Details of these procedures are presented in the following sections.


1.0  SAMPLE STORAGE AND TRANSPORTATION PROCEDURES

     Sample collection and transportation of the samples to the RTI Ion Ana-

lysis Laboratory are the responsibility of the particulate monitoring coordin-
ation contractor (PMCC)/NPS.
     Samples are shipped to RTI in lots of approximately 100 to 200.


2.0  SAMPLE CHECK-IN AND HANDLING IN THE LABORATORY

     Samples are mailed to RTI in secure containers.  Upon receipt, RTI does
the following:.

     1)   Open the shipping box and remove the samples.

     2)   Record the PMCC/NPS Sample ID, date of receipt, and com-
          ments (broken package, contamination, etc.) on the Sample
          Log Form (Figure 1).

-------
Date Received
       INKS Fi iter Samples

                     Lot
     Sample ID
Extraction
   Date
Comments
                                                     Analysis
Date
1C
Rerun
                         Figure  1.  Sample  Log  Form.

-------
     3)   Store the samples 1n a locked laboratory for future de-
          sorption.
3.0  SAMPLE ANALYSIS METHODOLOGY
     The analysis procedures for the filter samples are given below.  The Ion
chromatographic procedures Initially were taken from "Operations and Mainten-
ance Manual for Precipitation Chemistry Measurement Systems" prepared by Rock-
well International.  Some improvements have been made..
3.1  Filter Extraction
     Using tweezers, place each filter in a Nalgene Monovette and add 15 mL of
extraction solution (0.0017M NaHC03/0.0018M Na2C03) using a repipet.  Push up
the Monovette plunger so that no air space remains above the filter.  Expose
the Monovette and solution to ultrasonic energy for 30 minutes and then allow
to sit overnight.  This process releases greater than 97 percent of the chlo-
ride, nitrite, nitrate, and sulfate into the solution.  Record the date of
extraction on the Sample Log Form.
3.2  Determination of Chloride, Nitrite, Nitrate,  and Sulfate Using Ion
     Chromatography
3.2.1  Scope and Application —
     This method covers the determination of chloride, nitrite, nitrate, and
sulfate in filter extracts.
3.2.2  Summary of Method —
     The anions are separated when passed through a resin consisting of poly-
mer beads coated with quaternary ammonium active sites.  The separation is due
to the different affinities of the anions for ^he active resin sites.
     After separation, the anions pass through a suppressor column which ex-
changes all cations for H+ ions.  Species are detected as their acids by a
conductivity meter.  (An eluent which yields a low conducting acid is used.)
3.2.3  Interferences —
     Large amounts of anions eluting close to those anions of interest will
result in an interference.  No interferences have been observed in NPS filter
samples analyzed to date.

-------
3.2.4  Apparatus --
     Ion chromatograph (Dionex Manual Model 21201 or Model 14) with anion

guard column, anion fast run separator column, and anion micromembrane sup-

pressor column.

     Four-liter collapsible bags.

     Pi pets - an assortment of sizes.

     Volumetric flasks - an assortment of sizes.

     Disposable syringes - 5 ml capacity

     Disposable filters - Acrodisc 0.45 /jm, or equivalent

     IBM/PC-based Dynamic Solutions Chromatographic Data Acquisition System.

3.2.5  Reagents —

     Use ACS reagent grade chemicals for the preparation of all solutions.
Dry chemicals used for the preparation of calibration standards at 105°C  for  2
hours and cool  in a dessicator immediately before use.
     Eluent, 0.0017M NaHCC»3/0.0018M NaeCOs:  Dissolve 2.8562 g NaHCC-3 and
3.8156 g Na2C03 in 20 liters deionized water.
     Regenerant, 0..025N ^SO^:  Add 500 mi IN ^$04 to a Nalgene Carboy and
dilute to 20L with deionized water.

     Mixed Stock Solution, 1000 mg/l NO?, NO-j, and SO^2, and 200 mg/L C1-;
Dissolve 1.4998 g NaN02, 1.6305 g KNC-3, 1.8142 g K2S04, and 0.3297 g NaCl  in  1
liter deionized water.

     Standard Solution A:  Dilute 10 ml mixed stock solution to 100 mL with
eluent (100 mg/L NOj NO}, and SO??, and 20 mg/L C1-).

     Standard Solution 8:  Dilute 10 ml standard solution A to 100 ml with
eluent (10 mg/L NC>2, NOj, S0?2, and 2 mg/L C1-).

     Using standard solutions A and B, prepare standards with eluent in 100 mL
volumetric flasks as shown in Table 1.  Preparation of standards in eluent
eliminates the water dip which interferes with chloride quantisation.  Prepare
fresh standards weekly.

3.2.6  Procedure —

     1)    Begin the flow of regenerant through the anion micromem-
          brane suppressor column.

     2)    Set up the range for maximum sensitivity (usually 10 /imho
          full scale for NPS samples).

-------
           TABLE 1.   PREPARATION OF ANION CALIBRATION STANDARDS
Standard

1
2
3
4

5
6
7
8
N02, NOg, S(
mg/L

10
5
3
2

10
5
2
1
D$ Cl- mL
mg/L
STANDARD SOLUTION A
2.0
1.0
0.6
0.4
STANDARD SOLUTION B
0.2
0.1
0.04
0.02
mL Standard
Solution/100 mL

10.0
5.0
3.0
2.0

1.0
0.5
0.2
0.1
NOTE:  Higher concentration standards can be prepared from Standard
       Solution A or the mixed stock solution if needed.
        TABLE 2.   RECOMMENDATIONS FOR OPTIMUM INSTRUMENT SENSITIVITY
                                   DIONEX

1.  0.0017M NaHC03/0.0018M Na2C03 eluent with fast run separator and sup-
    pressor.

2.  Detector output range 10 pmho full  scale.

3.  100 mL injection loop

4.  Flow rate 2.3 mL/min.

-------
     3)    Begin to pump the eluent through the columns and allow
          baseline to stabilize.

     4)    Arrange calibration standards and samples to be analyzed.
          If the conditions suggested in Table 2 are used, an aver-
          age of 5 samples can be analyzed per hour.  Analyze a
          quality control  sample  first, using a single standard, to
          verify that the  instrument is operating properly.  If the
          observed value of any anion differs by more than 10% from
          the known value, find and correct the problem before ana-
          lyzing any samples.

     5)    Using a disposable Nalgene syringe fitted with a 0.45 /jm
          disposable filter, begin to inject the field samples.
          Record on the Analysis  Logsheet (Figure 2) the computer
          directory name for storing the chromatograms, the NFS
          sample ID, and the filename assigned to the chromatogram
          for that sample.  Record the date of analysis and the 1C
          Model Number on  the Sample Log Form.  Analyze one randomly
          selected sample  in duplicate.  If any sample produces an
          anion peak that  is offscale, indicate this on the Analysis
          Logsheet and the Sample Log Form.  Place the sample in a
          separate rack for later analysis using a higher detector
          output range (usually 100 /jmho full scale).

     6)    When approximately half of the day's samples have been
         "run, perform the daily  calibration beginning with the
          standard of highest concentration.  Record the standard ID
          and filename on  the Analysis Logsheet.

     7)    Complete the analysis of the field samples, including an
          EPA Quality Assurance sample and a quality control  sample.

3.2.7  Calculations Using  A Linear Least Squares Fit --

     Peak heights are entered into the computer where linear least squares

calculations are performed.  The  linear least squares fit yields the following

parameters:  slope (s), intercept (I), and correlation coefficient (r).  The
slope and intercept define a relationship between the concentration and the

instrument response of the form.

                    yi = sxi + I                      (1)

     where:

               yi is the predicted instrument response

               Xj is the concentration of standard i.

               s  is the response slope

               I  is the intercept

-------
Directory: MAX\DA1
Method name:
)ata Copied (date) :
Mskette No.:
)ata Deleted (date)
Sample ID
























"A \ Date:
Analyst:
1C Model No,.:
Project No.:
•

Fi 1 ename

























Comments

























Figure 2.   Analysis Logsheet.

-------
     Rearrangement of Equation 1 yields the concentration corresponding to an
instrumental measurement:
                    xj * (yj - I)/s                   (2)
     where:
               Xj is the calculated concentration for a sample
               yj is the actual instrument response for a sample
               s  is the calculated slope from the calibration above
               I  is the calculated intercept from the calibration above
3.2.8  Quality Control —
     Compare the regression parameters for the standard curve with those ob-
tained in the past.  If they exceed the control  limits,  stop analysis and look
for the problem.
     Analyze a quality control sample against a  single calibration standard at
the beginning of every analytical  run.  Compare  the results with those ob-
tained in the past.  If the observed concentration of any anion differs from
the known value by greater than 10%,  stop the analysis until the problem is
found.  Analyze a duplicate sample and an EPA Quality Assurance Sample each
day.
     When a new stock standard solution is prepared,  dilute calibration stan-
dards from both the old and the new stock.  Analyze the old and new standards
and compare the calibration curves.
3.2.9  Troubleshooting —
     Refer to the Dionex Model 212CH or Model 14 Operators' Manual for any
instrumental problems.

4.0  QUALITY ASSURANCE PROCEDURES
     The role of any analytical laboratory is to provide qualitative and quan-
titative data which accurately describe the characteristics and/or concentra-
tions of the constituents in the samples submitted.  The laboratory data must
be backed up by an adequate program to document  the proper control of all  the
factors which affect the final result.  RTI is committed to the implementation

-------
of a thorough and dependable quality assurance/quality control program which
is understood and followed by all operating personnel and supported by manage-
ment.
4.1  Project Organization and Responsibility
     QA project organization is shown in Figure 3.  Dr. W. F. Gutknecht, Pro-
ject Manager, provides overall  supervision of the project.  Dr. E. D. Estes,
Project Leader, is responsible for the analyses conducted during the study and
for data reporting.  Dr. W. C.  Eaton serves as the program's Quality Assurance
Officer.  Ann R. Turner is the sample custodian and analyst.
4.2  QA Objectives for Measurement Data in Terms of Precision, Accuracy^
     Completeness, Representativeness, and Comparability
     The filter samples collected in this study are analyzed for the parame-
ters summarized in Table 3.  The methods used for these measurements also are
summarized in the table.
4.2.1  Precision and Accuracy --
     Precision and accuracy objectives for each of the analyses are listed in
Table 1.
4.2.2  Completeness --
     Analytical results will be obtained for at least 98 percent of the fil-
ters.
4.2.3  Representativeness --
     All filter samples are collected by the PMCC/NPS which is solely respons-
ible for the representativeness of the samples.
4.2.4  Comparability --
     All analyses for a given parameter are reported in the same units and are
directly comparable.
4.3.  Sampling Procedures
     Sampling is the responsibility of PMCC/NPS which is expected to provide
representative filter samples.

-------
      NPS

Project Officer
                      RESEARCH TRIANGLE INSTITUTE

                           J. B. Tommerdahl
                            Vice President
                      Environmental Sciences and
                             Engineering

                             C. E. Decker
                               Director
                       Center for Environmental
                             Measurements
PROJECT OFFICE

W. F. Gutknecht
Project Manager

  E. D. Estes
Project Leader
                        KeyTechnical  Personnel
                            ION ANALYSIS

                            E.  D. Estes
                            A.  R. Turner
    QA Office

   W.. C. Eaton
Quality Assurance
     Officer
                Figure 3.   Project Management Structure

-------
       TABLE 3.   PHYSICAL/CHEMICAL PARAMETERS AND METHODS FOR ANALYSIS
Unit of
Parameter Report
CI-
NQ?
N03
S0,2
mg/L
mg/L
mg/L
mg/L
Recommended Range of
Methoda Parameter
1C
1C
1C
1C
0.03
0.02
0.002
0.02
- 7.5
- 1.0
- 27
- 22
Estimated
Precision
i 10%
. + 10%
i 10%
i 10%
Estimated1*
Accuracy
+ 10%
+ 10%
+ 10%
+ 10%
a  1C - 1on chromatography

b  These values assume concentrations which equal  or exceed 100 times the
   minimum detectable limit (MDL).   For concentrations at the MDL, measure-
   ments are accurate to within +  100%.  At concentrations equal to 10 times
   the MDL, accuracy is within + 20%.

-------
4.4  Sample Custody
     Ann R. Turner is the RTI sample custodian and analyst and is responsible
for maintenance of a laboratory sample custody log.  As samples are received
in the laboratory and processed, the following information is entered onto the
Sample Log Form.
     1)   Date samples received in laboratory
     2)   PMCC/NPS sample identification number
     3)   Date of desorption
     4)   Date of ion analysis and instrument used
4.5  Calibration Procedures and Frequency
     Calibration procedures and frequency for individual measurements are in-
cluded in Section 3.0 Sample Analysis Methodology.
4.6  Analytical Procedures
     Analytical procedures are found in Section 3.0 Sample Analysis Method-
ology.
4.7  Data Reduction,  Validation, and Reporting
     Data reduction schemes are contained in the methods presented in Section
3.0.
     Data entries are reviewed by the Quality Assurance Officer who checks
calculations and verify completeness of data.  Relative standard deviations
and relative errors are calculated from analyses of replicates and quality
control/quality assurance samples, respectively.  If precision and accuracy
fall outside the data quality objectives summarized in Table 3,  analyses are
repeated.
     Laboratory notebooks are checked and signed by the Quality Assurance
Officer and then reviewed and signed by the Project Leader.  Notebooks and
data sheets are available for review by the Quality Assurance Officer.  Ana-
lytical results for each filter lot,  including QC/QA data, are compiled by the
Project Leader and presented to the Quality Assurance Officer.  Upon approval
of the Quality Assurance Officer,  the data are submitted to the NPS.

-------
4.8  Internal Quality Control  Checks
     Each day of analysis,  a quality control  sample,  a duplicate sample, and a
quality assurance sample are analyzed.  Filter blanks are analyzed as they are
supplied by the PMCC.
4.8.1  Filter Blank ~
     This is a blank filter supplied by the PMCC which is subjected to the
same preparation procedure as are the field samples being analyzed.  This
sample will be used to check for contamination which  may occur in sample prep-
aration or analysis.
4.8.2  Duplicate Sample --
     This is a sample chosen at random from a day's run.  This sample provides
information on analytical precision.
4.8.3  Quality Control Sample --
     This sample is prepared by the Project Leader from stock solutions inde-
pendent of those used to prepare the calibration standards.  It is analyzed at
the beginning of each run to ensure that the  chromatograph is operating pro-
perly.  If the observed value of any anion deviates from the known value by
more than 10%, no samples are analyzed until  th-e problem is corrected.
4.8.4  Quality Assurance Sample —
     This sample is an EPA Quality Assurance  sample with known values of the
constituents.  The data are reported to the Project Leader who calculates
percentage recovery and reports the data to the analyst and to the Project
Manager.  This sample provides an assessment  of data  quality independent of
analyst's judgement.
4.9  Preventive Maintenance
     Analytical instrumentation used in this  project  will be carried through
preventive maintenance procedures and schedules as recommended by the manufac-
turer.
     Dionex Ion Chromatographs require a minimum of maintenance if operated at
pressures less than 800 psi.  During the following periods, perform the main-
tenance listed to eliminate unnecessary troubleshooting.

-------
     Dally (all operating days):

          -  Check for leaks at all valves and column fittings (at
             normal operating pressure).
          -  Cycle the injection valve by repeated switching between
             INJECT and LOAD and rinse with DI H20.
          -  Check the meter Zero and Cal adjustments.
          -  Wipe-up liquid spills and salted-out chemicals.
          -  Check the drip trays.
     Weekly:
          -  Compare standard chromatograms to check that no signif-
             icant changes in column efficiency have occurred.
          -  Check all  air and liquid lines for crimping or dis-
             color.

     Monthly:

          -  Check column resolution by measuring percent resolution
             of the N03 and 50$ peaks.
          -  Oil each pump with 2 drops SAE No. 10 oil.


5.0  DATA REPORTING PROCEDURES

     The data from the  analysis of each lot of filters are reported to the

National Park Service both in hard copy and on floppy disk in ASCII format.

The data are grouped according to the day of analysis and the instrument

(Model  2120i or 14) on  which they were run.  With each filter lot,  quality

control  (QC) and quality assurance (QA) data are reported for each instrument,

for each day that samples were analyzed.  Reporting of QC/QA data in this

manner facilitates the  detection and correction of any instrument-specific

problems.  Data from the analysis of daily duplicate samples are compiled and

periodically reported to the NPS so that precision estimates can be made.

-------
 IMPROVE SOP APPEND                                          JULY 1989
                APPENDIX  7:   Carbon Contractor Procedures (DRI)
Titles   Thermal/Optical  Carbon Analysis  of Aerosol             Page 1  of 	
        Filter  Samples

Numbers  2-204.1                                               Datet   5/2/89
1.0    GENERAL  DISCUSSION

1.1    Purpose  of  Procedure

      This  standard  operating procedure  is  intendeds
         -  to  provide a  basic understanding of  the principles  behind carbon
           analyzer  operation;
         -  to  describe routine determination of organic,  elemental,  and
           carbonate carbon from ambient and source filter samples  using  the
           OGC/DRI  thermal/optical reflectance  carbon  analyzer;
         -  to  detail the concerns and procedures which will insure  a
           state-of-the-art carbon analysis measurement process.

      This  procedure will be followed by all analysts  at the Environmental
      Analysis Facility  of the Energy and Environmental Engineering Center
      of the Desert  Research Institute.


1.2    Measurement Principle

      The OGC/DRI thermal/optical carbon analyzer is based on the preferential
      oxidation of  organic and elemental carbon compounds at different
      temperatures.   It  relies on the fact that organic compounds can be
      volatilized from the sample deposit in a  helium (He) atmosphere at Jow
      temperatures  while elemental carbon is not oxidized and removed.  The
      analyzer operates  by 1) liberating carbon compounds under different
      temperature and oxidation environments from a small punch taken from
      a quart?: fibnr filter; 2) converting these compounds to carbon dioxide?
      (C02) by passing the volatilized compounds through an oxidizer (heated
      manganese dioxide, MnO2); 3) reduction of C02 to methane (CH4)
      by passing the flow through a methanator (hydrogen-enriched nickel
      catalyst); and 4)  quantification of CH4 equivalents by a flame
      ionization detector (FID).

      The principal function of the optical (laser reflectance) component of
      the analyzer is correction for pyrolysis of organic carbon compounds
      to elemental  carbon.  Without this correction, the organic carbon
      fraction of the sample would be underreported and the elemental carbon
      fraction would include some pyrolyzed organic carbon.  The correction
      for pyrolysis is made by continuously monitoring the filter reflectance
      (via a helium-neon laser and photodetector) throughout an analysis
      cycle.  This  reflectance, largely dominated by the presence of black
      elemental carbon,  decreases as pyrolysis takes place and increases as
      elemental carbon is liberated during the latter part of the analysis.
      By monitoring the reflectance, the portion of the elemental carbon
      peak corresponding to pyrolyzed organic carbon can be accurately
      assigned to the organic fraction.   The correction for pyrolytic
      conversion of organic to elemental carbon is essential for an unbiased
      measurement of both carbon fractions, as discussed in Johnson et al.

-------
      (1981).

      Carbonate  carbon  may be  determined  by measuring  the  C02  evolved
      upon  acidification of the  sample  punch  before  the  normal carbon analysis
      procedure.

1.3    Measurement  Interferences  and  Their Minimization

      Carbonate  carbon  presents  significant interference in  carbon analysis
      if  it consititutes more  than 5% of  total  carbon  in the ambient or source
      sample,  as it  is  measured  as both organic and  elemental  carbon during
      thermal/optical carbon analysis.  Acid  pretreatment  of the  filter samples
      can eliminate  the carbonate interference.

      The presence of certain  minerals  in some  soils can affect the laser
      correction for pyrolysis.  These  minerals change color as the sample
      punch is heated,  generally resulting in a sample which is darker.   For
      samples  which  contain large fractions of  resuspended soils,  the split
      between  organic and elemental  carbon may  have  to be  estimated manually.

      Some  minerals, again predominantly  in soil samples or  soil  dominated
      samples, may affect the  laser  reflectance by temporarily changing color
      or  changing  the surface  texture of  the  deposit residue.   Unlike the
      effect described  above,  these  changes are reversible and highly
      temperature  dependent.

      Some  colored organic compounds can  affect the  laser  correction
      as  well, causing  increased reflectance  as these  compounds are removed.
      This  effect  is readily ascertained  by examining  the  laser response
      during the organic portion of  the analysis.  Again,  the  split between
      organic  and  elemental carbon may  have to  be estimated  manually if the
      effect is  large.

      Finally, the presence of certain  elements (Na, K,  Pb,  Mn, V, Cu,  Ni,
      Co, and  Cr)  existing either as contaminents on the filter or as part of
      the deposit  material has been  shown to  catalyze  the  removal  of
      elemental  carbon  at lower  temperatures  (Lin and  Friedlander, 1988).
      Such  catalysis would affect the distribution of  carbon peaks during the
      analysis.

1.4    Ranges and Typical Values  of Measurements

      A wide range of aerosol  concentrations  can be  measured with  this
      method,  the  limiting factor being the concentration  of the  carbon
      compounds  on the  filter  on a ug/cm2 basis. Dirtier  environments
      may be sampled and still be analyzed within the  range  of the carbon
      analyzer by  increasing the filter deposit area or  by decreasing the
    .  flow  through the  filter  media.

      The carbon analyzer can  effectively measure between.  0.2  and  750 ug
      carbon/cm2.   The  upper range is  somewhat  arbitrary,  depending on the
      particular compounds on  the filter  and  the temperatures  at  which they
      evolve.  This  upper range  may  be  extended by taking  special  precautions,
      such  as  reducing  the punch size  or  by special  temperature programming,
      to  avoid an  over-range FID signal.

      Typical  carbon values range between 10  and 100 ug  carbon/cm2 for ambient
      samples.

-------
      The lower  quantifiable limits  (LQLs)  of  carbon combustion methods
      depend upon  the  variable carbon content  of  the blank quartz filters as
      well as the  analysis  method.   For better LQLs , the unexposed filters
      should be  pre-fired in an oven at high temperatures for several hours
      to remove  any residual carbon  contamination (Fung, 1986;  Huntzicker,
      1986;  Rau, 1986).   All quartz  filters originating from DRI are pre-fired
      for a minimum of four hours  at 900 C  and are tested for blank levels
      before use.   For well-cleaned  quartz  filters,  the standard deviation of
      the blanks for organic and elemental  carbon is on the order of 0.5 and
      0.2 ug/cm2,  respectively (Fung, 1986).   Typical pre-fired blank levels
      at DRI are 0.5 - 1.0  ug organic carbon/cm2  and 0.0 - 0.2  ug elemental
      carbon/cm2 .   Because  even pre-fired filters can absorb organic vapors
      during shipping  and storage, the LQL  of  analysis on a particular set of
      filters depends  on the number  of field blanks  analyzed and the
      variability  in the results front those blanks.

      Acid-evolved carbonate levels  in pre-fired  quartz filters have been
      shown in several informal tests at DRI to be quite variable over
      time.   Part  of this phenomenon is apparently due to the reaction of
      ambient C02  with alkaline sites on the quartz  fibers .   Acceptance
      testing for  carbonate is not routinely performed at DRI .

      The precision of this analysis has been  reported to range from 2 to
      4% (Johnson,  1981).  For analysis of  actual ambient and source filters,
      homogeniety  of the deposit is  most important for reproducible results.
      For evenly loaded filters, precision  is  generally 5% or less; for poorly
      loaded filters,  replicates may deviate by as much as 30%.  The
      precision  of carbonate analysis results  is  approximately  10%.

      The precision of the  laser-dependent  split  between organic and elemental
      carbon fractions depends upon  how rapidly the  laser is increasing at
      the time of  the  split and whether the split is made in the middle of
      a  carbon peak or not.  Typically, relative  laser split times are
      reproducible within 10 seconds and deviations  in calculated splits are
      less than  5% of  the total measured carbon.

      The accuracy of  the thermal/optical reflectance method for total carbon
      determined by analyzing a known amount of carbon is between 2 to 6%
      (Rau,  1986).   Accuracy of the  organic/elemental carbon split is between
      5  and
1.6    Personnel  Responsibilities

      All  analysts  in  the laboratory should  read and  understand the entire
      standard operating procedure prior to  performing carbon analysis,  which
      includes routine system calibration, actual analysis,  and immediate
      review of  the data as  it is  produced to  correct system problems.

      It is  the  responsibility of  the laboratory manager or  supervisor
      to ensure  the carbon analyses procedures are properly  followed, to
      examine and document all replicate, standard, and blank performance test
      data,  to designate samples for reanalysis,  to arrange  for maintenance
      and  repair, to maintain the  supplies and gases  necessary to insure
      uninterrupted analysis,  and  to deliver the analysis results in dBase
      format to  the project  manager within the specified time period.

      The  quality assurance  (QA) officer of  DRI ' s Energy and Environmental

-------
      Engineering Center is  responsible to determine the extent and methods of
      quality assurance to be applied to each project,  to estimate the level of
      effort involved in this quality assurance,  to update this procedure
      periodically,  and to ascertain that these tasks are budgeted and carried
      out as part of the performance on each contract.


1.7    Definitions

      The following  terms are used in this document:

        Calibration  injection -  the injection of  calibration gases (methane in
        helium [CH4/He] or carbon  dioxide in helium [C02/He]) into the sample
        stream to check instrument performance.

        Calibration  peak - the FID peak resulting from the automatic injection
        of methane calibration gas (CH4/He)  at the end of each analysis run.
        All integrated peak  areas  are divided by  the calibration peak area
        and multiplied by an instrument-specific  calibration factor to obtain
        ug carbon.

        Elemental carbon - carbon  evolved from the filter punch in a helium/
        oxygen (He/02) atmosphere  at 550, 700, and 800 C minus pyrolyzed
        organic carbon.

        Laser split  - the time at  which the laser-measured reflectance of the
        filter punch reaches its initial value, indicating that all pyrolyzed
        organic carbon has been  removed and original elemental carbon is
        beginning to evolve.

        Lower split  time - the time at which the  laser-measured reflectance of
        the filter punch reaches its initial value minus the precision of the
        laser signal (currently  defined as 10 counts).

        Organic carbon - carbon  evolved from the  filter punch in a He
        atmosphere at 120, 250,  450, and 550 C plus pyrolyzed organic carbon.

        Pyrolysis -  the conversion of organic carbon compounds to elemental
        carbon due to incomplete combustion/oxidation;  may be envisioned
        as "charring".

        Pyrolyzed carbon - the carbon evolved from the time that the carrier
        gas flow is  changed  from He to He/02 at 550 C to the time that the
        laser-measured filter reflectance reaches its initial value.

        Upper split  time - the time at which the  laser-measured reflectance of
        the filter punch reaches its initial value plus the precision of the
        laser signal (currently  defined as 10 counts).


1.8    Related Procedures

        SOP's related to carbon  analysis activities which should be reviewed
        in conjunction with  this document are:

          DRI SOP t6-001.1  Shipping and Mailing  Procedures.

          DRI SOP #6-009.1  Field and Laboratory  Safety Procedures.

          any SOP's  dealing  with filter handling  and shipping in conjunction

-------
          DRI SOP #4-001.1  Creation,  Revision,  Distribution,  and Archiving
          of Standard Operating Procedures.

          DRI SOP #	  Pre-Firing of Quartz Filters for Carbon Analysis

          The maintenance and troubleshooting guide for the DRI/OGC carbon
          analyzer.

          The appropriate MS-DOS or PC-DOS manual for the computer used with
          the carbon analyzer.

2.0  Apparatus,  Instrumentation, Reagents, and Forms

2.1  Apparatus and Instrumentation

2.1.1  Description

       The components of the DRI/OGC thermal/optical carbon analyzer are
       depicted in Figures 1 and 2; the complete gas flow schematic is shown in
       Figure 3.   The programmable combustion oven is the heart of the carbon
       analyzer and includes loading,  combustion,, and oxidation zones in
       a single quartz "oven" as depicted in Figure 4.

       In addition to the DRI/OGC thermal/optical analyzer connected to a
       IBM-PC or compatible computer,  the following items are needed for
       routine carbon analysis:

         - Punchs 0.503 cm2 area for removing small sample punches from
                  quartz filters.  This punch needs to be kept clean and
                  sharp.  If the punch is sharpened, the punch area must
                  be reverified.

         - Syringes: a gas-tight 1000  or 2500 ul syringe for calibration
                  injections; 25 or 50 ul syringe for carbonate analysis
                  and for analyzer calibration.

         - Quartz filters: Pallflex 2500QAT-UP or equivalent.

         - Tweezers.

         - Glass petri dish.

         - Log book/notebook.

         - Transparent tape.

         - Kimwipes.

         - Small styrofoam cooler.

         - Blue ice.

         - A copy of Carbon.EXE (the analysis program),  version P2.1 or
           later, and Carbon.DAT (the  analysis parameter file), version
           D2.0  or later.

2.1.2  Instrument Characterization

-------
 Vl
 o
 M
 a)
5

 o
•g
 lit
u
u
 CU
o
 10

 M
 
-------
§•

g





1
o
a
c
o
o
3









J
•5 <*•
* O
0 &
I x
. °
JJ 1
• o 3
: Is 4
z u > C
» 1 NISI
s
T3
1
j
OTD
||j|
? ? 1 ;
1335
a i fl >
^ 
-------
                   o
                   c
                   o

                  ^-1
                   0)
CL  0)
  Q
                         1°
                          L J
                     	_

                     asuopayay
                          c
                        en1 o
                        I*
                        o .*_/
 0)
.4_*
 D
 a.

 S
 o
 u
 o
 L.
 u
                                   u
                                         r
   0)

   X
                                    (U

                                  'X
                                               OT
             Q.
             O


            O
                                                 Q-
                                                _O

                                                Q.
                                                W

                                                Q
                                         L .

(O
VI
o
2
0)
en
O
I—
O
-*— '
                         B
                         a
                         VI
                         00
                         to
a.
o

-------
o

 c
 o
•H
u
 w
 3


 O
U

 1-4
 0)
 M
 0.
o
OS
Q
O
o

-------
                    	 	_,—^  is  program-vanven ana aata is stored
       automatically to disk via an IBM-PC compatible computer.   Response
       times and signal lag times are  built into the parameter file which
       is loaded when the analysis  program begins.   The program is event-
       driven;  that is, when the FID signal returns  to its baseline
       after a  minimum of 80 seconds at one condition, the program will advance
       to the next temperature or carrier gas  mixture.  A maximum time limit
       per condition is also established to prevent  a slight baseline drift
       from holding the analyzer in one condition  indefinitely.

       This method requires no sample  pretreatment,  requires between 15 and
       70 minutes per sample, and destroys the sample.

       Operator concerns for correct routine operation of the instrument
       should be the following (refer  to section 4 for more details):

        - Remember to push to sample in when the tone sounds; DO NOT leave
          the room until the analysis  begins.

        - Insure that the thermocouple is physically decoupled from the
          sample boat after pushing in the sample  to prevent oven temperature
          from  influencing the laser reflectance signal..

        - Insure that the sample port  is tight after loading a sample punch.

        - Check the graphical printout after each  analysis run to insure that
          the FID, temperature,  and laser signals  are behaving as expected.

        - The analyzer's quartz oven is susceptible  to breakage, especially
          at the sample port.  Care should be  taken  to avoid exerting
          tangential pressure on the oven when manipulating the sample port
          fitting.
2.1.3  Maintenance
       Regular maintenance for the analyzer involves daily checking of
       compressed gas supplies,  cleaning the punch and tweezers between each
       sample, and backing up data files on a regular basis.   Checks
       of laser adjustments (physical and electrical) are made at least
       monthly; analyzer calibrations are performed every six months.  All
       calibrations and repairs  must be recorded in the log book.

       Refer to the maintenance  and troubleshooting guide for additional
       information.
2.1.4  Spare Parts
       It is strongly recommended that the following spare parts be kept on
       hand to insure minimal interruptions in analysis:

         - Quartz rods:  3 mm nomimal diameter, Homosil optical quality rod,
           available from GM Associates (Oakland)  cut to  9 3/4" lengths and
           polished on both ends.

         - Quartz ovens: specially built ovens by the Oregon Graduate Center
           glass blower.

         - Quartz boats: made in-house from scraps of broken ovens.

         - Thermocouple rods: 18" length by 1/8" OD, type K ground isolated

-------
           with J14 stainless steel sheaths (March! Associates,  #SDH175).

         - FID flame tips:  for Gow-Mac #12-800 FIDs (Gow-Mac,  #132-117).

         - Fuses:  15 A,  MDL 15, slow-blow.

         - Punches: smaller than the 0.503  cm2 punch normally used,  for
           excessively heavily loaded samples.

         - Septa:  1/4" and  1/8", for injection ports.

         - Replacement needles for syringes.

         - Replacement scrubber tube: Supelco oxygen scrubber (Supelco,
           #2-2396).

         - Stainless steel  wire: for forming  "ears" to hold the  sample boat
           in position and  for wrapping the "ears"  onto the thermocouple
           push rod (Rocky  Mountain Orthodontics,  #RMO E-19,  0.914 mm).

         - Quartz  wool:  for repacking the oxygen oven (Alltech Associates,
           #4033).

         - Teflon  ferrules: Parker or Swagelok style, 1/2"  ID, for the sample
           port fitting.

         - Teflon  ferrules: 1/2" OD by 1/8" ID,  for the thermocouple rod
           at the  back of the oven.

         - Heating element  for oven: custom made 650 W coiled heater (Marchi
           Associates #SDH175).

         - FID batter: 300  VDC (EDCO, EverReady #495).

         - Printer paper.

         - Printer ribbons.

         - Computer disks,  double side, double density.

2.2   Reagents

       The following chemicals should be reagent grade or better:

         - Potassium hydrogen phthalate fKHP), for  calibration use  (Fisher,
           SP-243).

         - Sucrose, for calibration use (EM Science,  ISX1075-1).

         - Manganese dioxide (Mn02), crystalline, as an oxidizer in  the
           oxygen  oven (Nurnberg Scientific,  #RM200).

         - Hydrochloric  acid (HC1),  0.4 molar solution,  for use  in cleaning
           punch and quartz ovens,  and for  use in carbonate analysis.

         - Distilled dionized water (DDW):  total carbon background should be
           6  ppm or less.


2.3   Gases

-------
       The following compressed gases should be zero grade or better:

         - Helium for a carrier gas,  regulated to 35 psi with a metal
           diaphragm regulator.  The  higher pressure is required due to
           the pressure drop across the Supelco oxygen scrubber.

         - 5% methane by volume in helium for calibration injections and
           calibration peaks; regulated to 10 psi by a metal diaphragm
           regulator.

         - 5% carbon dioxide by volume in helium for calibration injections;
           regulated to 10 psi by a metal diaphragm regulator.

         - 10% oxygen by volume in helium as a carrier gas, regulated to
           10 psi by a metal diaphragm regulator.

       In addition,  the following gases are required:

         - Hydrogen  for the FID flame, regulated to 10 psi with a metal
           diaphragm regulator.

         - Compressed air to supply oxygen to the FID, regulated to 10 psi
           by a metal diaphragm regulator.

       At least one  backup cylinder per gas type should be kept on hand
       at all times.  The calibration gases typically last for  one year.
       The hydrogen, helium, and 02/He mixture are typically replaced
       every three to five weeks.  The compressed air is replaced every
       4  to 5 days.   All gases are replaced when the cylinder pressure
       drops below 500 psi.

2.4   Forms and Paperwork

     All  samples are logged into the  "Air Analysis Logbook" upon receipt at
     the  laboratory.  Refer to Figure 5 for the format of this  logbook.
     A sample analysis list will be prepared by the laboratory manager
     indicating which samples will be analyzed and any special  instructions.
     Samples designated for carbon analysis are logged into the "Carbon
     Analysis Logbook" prior to analysis; Figure 6 provides a sample of
     entries in this logbook.

     As individual samples are analyzed,  entries are made in the "Carbon
     Analyzer Logbook", as shown in Figure 7.  As each analysis run is
     completed, the  sample analysis list is marked with the date and
     analyzer number,  as in Figure 8.

3.0   Calibration Standards

3.1   Preparation, Ranges, and Traceability of Standards

     Four standards  are used in calibrating the carbon analyzers: 5% nominal
     CH4  in He, 5% nominal C02 in He, KHP,  and sucrose.  Only the calibration
     gases are used  on a daily basis  as analyzer performance monitors.  KHP
     and  sucrose are used in conjunction with the two gases semiannually to
     establish the calibration curve  of each analyzer.

     The  calibration gases are assayed for exact concentrations by the gas
     supplier; the assay value is obtained from the tag on the cylinders and
     is typically determined by gas chromotography (GC) or gravimetry.

-------
               G
               Cj r-
               ^- rv
                                                                                                    u
                                                                                                    a
                                                                                                     00
                                                                                                     O
                                                                                                     •3
                                                                                                     1O
               j-
               ti
                        U(
V
      •x
               .r

               "A

-------
                                             I  I
—w
      \
            °
             .0
              o
                                I  ' I
                 T
             «M+
1'  i    '
-/iH '-!"
                                  _
                                  _      _
                               '~l"  t~M  i'  ,
                               :-j-rrrrf
               ~r
          \-vp\-1
          Vv\n\r
          4±j
          -!f i  '
          _LJ_U_LJ_
t
            o 10
              b
               i!f jr

                            i  i
                                  ,
        MM
          !  i i
                £>;_ !_   (T)
                                             ».{
                                      -i—)_..
                                      i  i
          Figure 6.  Example of Carbon Analysis Logbook Formac

-------
                                            -^H~
                 7 i'"i M  M  F IT 7
                 7—•>— —; -i  • - -I -7—1—•—,•

         i
                   nni
 3
I  I !  i !
-D
j  i-44-|..j.|..
                                                     o
                                                     b
                                                     o
                                                     o
                                                     ^3
                                                     m
                                                     c
                                                     <
                                                     o

                                                     a)
                                                     r-1
                                                     Q.

-------
                         IMPROVE:  Lot G
                              >*******************

ate: 4/15/89
rom: L.Pritchett
'o   : J.Chow
     J.Watson
     C.Frazier .
     Carbon Room

'otal nvunber of samples:   418

ipecies to be analyzed:
  OC/EC      by carbon analyzer

.nstmctions:

 1.  This is the sixth analysis  list for  IMPROVE carbon analysis.   We
     are processing  filters in order of lot  number,  starting with the
     oldest lots first.

 2.  Note that we are using analysis IDs  on  these samples,  given that
     the actual filter IDs are too  complicated and too long for the
     current carbon  program.  BE  VERY CAREFUL THAT THE FILTER ID MATCHES
     THE ANALYSIS ID ON  THIS LIST.   IF THERE ARE ANY DISCREPANCIES, SEE
     LYLE BEFORE PROCEEDING WITH  THE ANALYSIS.

 3.  Carbon analysis on  lot G will  begin  the week of April  17,  1989,
     immediately following the Santa Barbara,  Phoenix Pilot Study, and
     IMPROVE Lot F samples.

 4.  Deposit area for the quartz  filters  is  3.8 cm2.

 5.  dBase III files will be named:

        carbon data  : IMCarG.DBF
     Sample ID
Filter ID
OC/EC
P00701
P00702
P00703
P00704
P00705
P00706
P00707
P00708
P00709
P00710
YOSE020288C1
YOSE020288C2
SAG0020288C1
ARCH020288C2
ARCH020288C1
REDW020288C1
PORE012688C2
GRCA020288C1
MORA020288C2
BRCA020288C1
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
                       Figure 8.   Example of Carbon Analysis List

-------
     The KHP is  dried at 110  C  for two hours before dispensing.   Transfer
     0.3826  g of KHP  into a glass  100  ml  volumetric flask.   Dilute to
     volume  with 0.4  ml  concentrated hydrochloric  acid  (HC1)  and dionized
     distilled water  (DDW). Mix the KHP thoroughly.   Store  this  solution
     in a refrigerator until  it is used for calibration purposes.   This
     solution is good for about 40 days.  Label  the flask with the chemical
     name, the date of preparation, the name of  the chemist preparing the
     solution, and the exact  concentration.  The concentration,  nominally
     1800 ppm carbon,  is calculated by

      /actual g  KHPj/8*12 g  Carbon \f 10"-3 ml  j/10"6  ug J    ug Carbon

      \  100 ml    J ^204.23  g  KHpy\   ul     J\     g   J    ul solution


     The nominal 1800 ppm sucrose  solution is prepared  by transferring
     0.428 g of  sucrose  into  a  glass 100  ml volumetric  flask.  Dilute to
     volume  with DDW.  Mix the  sucrose thoroughly.   Store this solution
     in a refrigerator until  it is used for calibration purposes.   This
     solution is good for about 40 days.  Label  the flask with the chemical
     name, the date of preparation, the name of  the chemist preparing the
     solution, and the exact  concentration.  The concentration is calculated
     by
       actual  g  Suc\/12*12  g Carbon j /10~-3 ml  \ /10~6  ug\     ug Carbon

         100 ml    /I  342.31 g KHP / I    ul     /(     g    /     ul solution
     To prepare  a  blank solution,  add  0.4 ml  of  concentrated  HC1  to a glass
     100 ml  volumetric  flask  and dilute  to volume with  DDW.   This acidified
     DDW is  made fresh  each time a 1800  ppm KHP  stock solution is prepared.

     No primary  standards  currently exist for carbon  analysis.   Ideally,
     such standards  should include a range of organic compounds from low
     to high molecular  weights  and with  varying  degrees of  susceptibility
     to pyrolysis, as well as elemental  carbon and  carbonate  compounds.
     Currently,  KHP,  sucrose, and  the  two calibration gases are used at  DRI
     for calibration and system audit  purposes.


3.2    Use of Standards

      The calibration slopes  derived from the two gases and the KHP- and
      sucrose-spiked filter punches are  averaged together to  yield a
      single calibration slope  for a given analyzer.  This  slope represents
      the response of the  entire analyzer to  generic  carbon compounds
      and includes the  efficiencies of the oxidation  and methanator zones
      and the sensitivity  of  the FID.  Note that the  current  calibration
      procedure  is based only on the total carbon;  currently  no routine
      procedure  exists  to  check the accuracy  of  the OC/EC split.


3.3    Typical Accuracy  of  Calibration  Standards

      The accuracy of the  calibration  standards  is  primarily  limited by  the
      accuracy of  the calibration  gas  assays  and by the care  taken during
      preparation  of the KHP  and sucrose solutions.  The calibration slopes
      determined by  these  four  compounds historically differ  by less than 5%

-------
      on a given analyzer if sufficient care is taken during the calibration
      procedure (Section 5.1).

4.0   Procedures

4.1  General Flow Diagram

     The typical flow of samples and data for carbon analysis is depicted in
     Figure 9.

4.2  Analyzer Start-Up

     The following steps outline analyzer start-up:

       - Check  all gas cylinders'  pressures; cylinders with gas pressures less
         than 500  psi should be replaced before beginning the day's analysis.

       - Check  that all gas delivery pressures are correct:
           Hydrogen       — 10 psi
           Helium         — 35 psi
           Compressed air — 10 psi
           02/He mix      — 10 psi
           CH4/He mix     — 10 psi
           C02/He mix     ~ 10 psi

        -  Check that all FIDs  are lit by holding a  pair of tweezers over
           the  FID exhaust stack and watching for condensation.  If the FID
           is not lit (as immediately after the hydrogen or compressed air
           cylinders are changed), relight the flame by turning the H2
           rotameter to 50 and  holding a butane lighter or match over the
           FID  stack.  A light  pop indicates that the flame is lit.  Verify
           that the flame remains lit by the tweezer test.  Often the flame
           will not stay lit the first time, especially after the hydrogen
           cylinder is changed  and air gets into the gas lines.  If the FID
           is cold, allow at least 30 minutes at the high gas flow to pass
           before turning the H2 rotameter to its correct setting.

        -  Check and readjust if necessary all gas flows at the analyzer.  The
           correct readings are posted on each rotameter.  Read through the
           center of the ball.   If drastic adjustments are required on one
           analyzer, recheck that flows on the other two analyzers  have
           not  been affected.

           Turn on the computer monitor.   Note: the  computers are generally
           left on at all times; only the monitors are turned off at night
           to avoid "phosphor burn".

           If the computers have not been used for more than one day, reset
           the  date by typing "DATE"  and answering the question or  by
           rebooting the computer (  ).

           If the  computers on  carbon analyzers #2 or 13 are rebooted, turbo
           mode must be set. For the computer on analyzer #2, press
              <+>, using the plus key on the far right side of the
           keyboard.  For the computer on analyzer #3, press   <->,
           using the minus key  on the far right side of the keyboard.  In
           both cases, the screen cursor will change size as turbo  mode is
           initiated.

        -  At the  C> prompt, type "CARBON" to begin  the carbon program.

-------
 Insure that the sample port fitting is tight and that the thermo-
 couple push rod is reasonably snug at the back fitting.  If the
 push rod is loose, tighten the rear fitting NO MORE than 1/16 of
 a turn.  Do not overtighten this fitting: a push rod that is too
 tight is not only difficult to operate smoothly, but will cause
 excessive wear of the Teflon ferrule.

 From the opening menu, select option 4; see Figure 10.  After
 insuring that the thermocouple push rod is pushed into the
 combustion zone, type "Y" to begin baking the oven.  The oven will
 be baked at 800 C for 10 minutes to insure that the system is clean
 before beginning analysis.  This option is self-timed and will turn
 off the oven after 10 minutes has elapsed.

 After the baking cycle is complete and while the oven is cooling,
 backup the previous day's analysis data by ending the carbon
 program (press ), changing to the appropriate directory, and
 typing "BACKUP C: A: /M".  The system will prompt for a formatted
 disk to be placed in drive A:.  NOTE: if data must be archived from
 multiple directories, begin backing up the data with the above
 command; after changing to the second and subsequent directories,
 type "BACKUP C: A: /M /A".  If the "/A" is left off the command,
 subsequent backup attempts will erase the first set of data.  After
 all backups are complete, label the disk in the following format:

     \directoryname    Analyzer*    BACKUP
     dateofanalysis         disk  n of n

 For example:

     \IMPROVE\LOTC       C/A fl     BACKUP
     880714                 disk  1 of 1

 Wipe the sample tweezers, petri dishes, and sample punch with
 clean KimWipes, taking care not to contact the cleaned surfaces
 with fingers or other dirty items.

 Begin the daily entry in the carbon analyzer logbook.  Entries
 should follow the format in Figure 7.

 Insure that the printers have enough paper for the day and that the
 ribbon is producing legible printing.

After the ovens have cooled to less than 100 C, perform a leak test
on the system by flipping off the "From Oven" toggle valve.  After
the He-1 and He-2 rotameters settle to zero (if they don't reach
zero in 2 minutes, see leak fixing procedures below), flip off the
"To Oven" toggle valve.  This process pressurizes the oven and
connecting tubing and then isolates the oven.  After 30 seconds,
flip on the  "To Oven" toggle valve.  If the He-1 rotameter float
jumps more than 5 units, the system has an unacceptable leak.
Correct the  leak by checking the following items:

     - Check that the sample port fitting is tight.

     - Check that the push rod is snug.

     - If the system still does not leak check, disassemble the
       sample port fitting, wipe all threads and ferrules clean

-------
.7
j
7
                      ** CARBON  ANALYST; PROGRAM **
Desert Research Institute/EEEC/LCP
                                                       Ver. P2. I   02/10/8*
                   /./MMMMMMMMMMMMMMMMMMMMMMMMMMMMS
                   J Current Directory:  C:\ODEQ .?
                   TMMMMMMMMMMMMMMMMMMMMMMMMMMMM,S
Program options:
   1 ... Normal DC/EC run
   2 ... CO3 + OC/EC run
   3 ... Calibration injection  run

   A ... Oven baking cycle
   5 ... Manual mode

   6 ... Recall previous data

   7 ... Chanqe current directory
   8 ... Print daily summary files

•'Esc>. .. End program

Input option:
                       Figure 10.  Carbon Program Main Menu

-------
         with a clean,  dry KimWipe,  reassemble, and retry.

       - If the system continues to  leak,  check the integrity of
         all tubing and of the quartz oven.   Refer to the carbon
         analyzer troubleshooting manual for additional tips and
         procedures.

   When the system leak checks satisfactorily,  from the main menu
   of the Carbon program select option 5.   This will result in the
   screen shown in Figure 11.   While watching the He-1 and Cal
   Gas rotameters, select option 6 (calibration gas load) .  The
   He-1 rotameter should not change  from zero,  and the Cal Gas
   rotameter should momentarily dip  down.   While watching
   the same rotameters, select option 5 (calibration gas inject).
   The He-1 rotameter should jump up momentarily and the Cal Gas
   reading should jump slightly.  Behavior different than this
   indicates a leak in the calibration gas injection system which
   must be corrected before beginning any analyses.  Refer to the
   carbon analyzer troubleshooting manual  for additional information.

   Because calibration gas has been  injected into the system by the
   above step, the system must be purged before continuing.  Open
   the "From Oven" toggle valve to restore flow through the oven and
   wait at least two minutes to insure all calibration gas has
   progressed through the system.

   Press  to return the Carbon  program to the main menu.  Select
   option 3 to begin the morning calibration injection.  Select
   He/02 carrier gas (option 2).  Select either C02 or CH4
   calibration gas type as the same  gas used the previous afternoon.
   For any given day, one gas will be used in the morning and the
   other in the afternoon.  By using the same gas in the morning as
   was used the previous afternoon,  the calibration gas used in the
   morning will be rotated on a regular schedule.

   The computer will create a sample ID based on the gas type, current
   date, and run number.  This ID should be entered in the analyzer
   logbook (see Figures 12 and 7).  Press   in response to
   the purge option to begin the calibration run.

   Insure that the printer is on-line.

-  When the elapsed time reaches 90  seconds (Figure 13), flush the
   1000 or 2500 ul syringe with the  appropriate calibration gas three
   times.  A low pitch warning tone  will sound at 114 seconds (the
   number of beeps corresponds to the carbon analyzer number).  When
   the analysis start tone sounds at 120 seconds, inject 1000 ul of
   the calibration gas into the injection port before the oven.  The
   rest of the analysis is automatic.

-  When the analysis is complete, a  tabular and graphical printout
   similar to Figures 14 and 15 will be generated.  From the
   tabular printout locate the calibration peak counts and the
   calculated ug C/filter.  Record these values in the logbook as
   in Figure 7.  The calibration peak counts should be in the
   following ranges:

       Analyzer #1 : 25000 counts to 27000 counts
       Analyzer #2 : 23000 counts to 25000 counts
       Analyzer #3 : 21000 counts to 23000 counts

-------
                    /.'MMMMMMMMMMMMMMMMMMMMMMMMMMMMM6'
                    J  Current Directory: C:\CftLIB J
                    TMMMMMMMMMMMMMMMMMMMMMMMMMMMMM v

Carrier gas options:

  1 ... He on 1x
  2 ... He/02

Input option: 1

DODODDDDDDDDDODODODDDCfODDDOOODDDDDODDOODDDDDODDODDOODDODODDDOODDDDOODOOOCDODOi
Calibration gas options:

  1 ... methane
  2 ... C02

Input option: 1

DDDODDDODDDDODDODDDC€>DODOODOOOOODOOOOOODDDDDDDDODDDDDDOOODDDDOOODDOODDDODOODOl
Data will be written  to  MI0224-2
Does the oven need  to be  purged'"' N

Ready to begin run; press any key when ready ...
             Figure  11.  Carbon Program Before Starting Calibration Run

-------
C.O
                 I

               §
                                                                                                                                          o
                                                                                                                                          o
                                                                                                                                          Q)
                                                                                                                                          on

                                                                                                                                          o
                                                                                                                                          CT-

                                                                                                                                          
-------
Analyzer
   CARBON CALCULATION RE'rULT'3
Carbon. FAS: P2. 1  02/10/8'=
                                   Carbon.DAT:  02.1  02/lO/ij
Analysis ID
'Sample ID
Carrier gas
Anal ysis
Calculation
Anal program ver
Calib. slope
Calib. intercept
Baseline window
MI 022 A- 2
MI0224
He 1 i urn on 1 y
02/24/8'3 14:41
02/24/8* 14:52
P2. 1 C02/10/8*.>





Parm
23.64 ug C/peak-to-calibration
0 . 00 ug C
1 counts
Sample transit : 2b sec

Laser
Calib





file ver :
peak ratio

prec i s i on :
transit :





C>2. 1 (.02/1 0/8'=-)


10 counts
•50 sec
Calibration peak area:  26i;i9fe counts
Initial  FID baseline  :    107 counts
  DC Peak  ttl
  Peak Area
24573 counts
                    Carbon
                21.52  ug  C/punch
* •«••** •*•«••*•»••<*•*-»# -i» •»• •*•»
  Calculated Carbon
         21.5 ug C/cm2
         21.5 uq C/filts-r
                       Figure 13.   Example of Tabular Printout

-------
                                                                                                   _
                                                                                                  •X-
                                         O



                                         01

                                        H


                                         O

                                         CD
                                        I—I
                                         O.
                                                                                                            9)
                                                                                                            U
                                                                                                            3
                                                                                                            Ml
C.O
                                        s
£3
                  <0

-------
                    UMMMMMMMMMMMMMMMMMMMMMMMMMMMM8
                    J Current. Directory: C:\QOEQ J
                    TMMMMMMMMMMMMMMMMMMMMMMMMMMMM .>

Program option  :  Normal  OC/EC run


Enter full sample ID:  'BQ7176

Sample ID is "SQ7176".   Is  this okay? Y

Enter run/punch number for  this sample <1-9):  1

OOOODDDDDDDDDDOODDDDDDDDODOODDDODODODOODODODDOODOODDOODDOODDODODODODODODDOO:
Punch size options:

  1 ... 0.503 cm2
  2 ... 0.ASA cm2
  3 ... 1     cm2
  A ... other

Input option:
  1 ... 0.503 cm2
  2 ... 0.A8A cm2
  A ... other-

Input option: 1
  1 ...  1   cm2
  2 ...  7.8 cm2
  3 ...  6. A cm2
  A ...  8 . ~ cm 2
  5 ... 13.8 cm2
  b . .  . other-

Input  option: 5
Data will be written  to  'i07l7t-l

Ready to begin run: press  any  Key  when ready . ..

           Figure 15.  Carbon Program Before Beginning  Regular Analysis

-------
                                                j.n tne roj.xowing ranges:
               CH4  gas  :  20 to 22.0 ug C
               C02  gas  :  19.0 to 21 ug C

           If results different from these are obtained,  rerun the calibration
           injection run.   If results are still out of the above ranges,
           locate and correct the problem;  refer to the carbon analyzer
           troubleshooting manual.

           From the main  menu of the Carbon program,  select option 7 to
           change to the  appropriate subdirectory for the samples to be
           analyzed.  The new subdirectory name, if valid, will be displayed
           at the top on  the screen.

           Based on the analysis list for the day,  retrieve the samples to
           be analyzed  from the sample freezer and place  in a styrofoam cooler
           with blue ice.   Place the cooler in the instrument room.

4.3  Routine Operation

     Routine analysis procedures depend on whether or not carbonate carbon
     will  be determined before OC/EC analysis.  The procedures are different
     for these two  options.

4.3.1  Routine OC/EC Analysis

        -   Pull the push  rod back to the idle zone of the quartz oven
           (approximately half way between the sample port and the heating
           elements).  Allow the boat/push rod to cool until the reading
           on the front of the analyzer reaches 50 C or less.  Do not
           pull the boat  into the sample loading zone when the boat is still
           hot as the heat will affect the Teflon ferrules of the sample
           port fitting.

           Insure that  the petri dish, tweezers, and punch are thoroughly
           wiped clean  with a dry KimWipe.

           Based on the analysis list, remove the sample  to be analyzed from
           the styrofoam  cooler.

           Remove the filter from the PetriSlide or petri dish with tweezers,
           taking care  to handle the filter only by the edge.  Place the filter
           on the glass petri dish and remove a sample punch by pushing down
           gently on the  punch.  Rocking the punch slightly will insure that
           the punch is complete severed.  Try to remove  the punch from the
           edge of  the  deposit to avoid wasting the filter, but try to avoid
           areas of non-uniform deposits.  Leaving the sample punch in the
           punch, place the punch under a KimWipe.   Replace the filter in the
           PetriSlide or  petri dish.

           Record  the filter ID in the analyzer log book (Figure 7).

           After the boat has cooled to 50 C or less, loosen the sample port
           fitting  carefully with a wrench.  NOTE: avoid exerting any sideways
           pressure on  the quartz oven.  Try to confine the wrench pressure
           to only  rotational torque.  Loosen the front fitting before
           attempting the rear fitting.  Slide the sample port fitting forward.

-------
Pull the boat back until it is centered in the sample port, taking
care that the small stainless steel "ear" holding the boat to the
push rod does not catch on the sample port opening and bend.  If the
"ear" does bend, carefully bend it back into position with tweezers
or small clean pliers.

Using the tweezers, push the bottom of the punch in the boat forward
so that the top of the punch can be accessed.  Remove the punch
and place it on the top of the analyzer.

Pushing the bottom of the sample punch in, remove the sample
from the punch and place it in the sample boat.  Generally, the
punch must be inserted sideways into the boat and then turned
so the punch wedges itself facing forward.  Push the punch forward
until it is seated against the front of the slot in the boat.

Push the push rod forward until the boat is located in the idle
zone of the quartz oven.  Slide the sample port fitting back
until it is centered over the sample port and tighten firmly by
hand.  DO NOT tighten with a wrench.  As before, avoid exerting
any sideways pressure on the quartz oven.

Select option 1 of the main menu of the Carbon program.  Input the
full sample ID.  NOTE: the program will automatically place the
computer into Caps Lock mode.  After verifying the sample ID,
enter the run number (1-9).  The run number must correspond to the
number of punches removed from the filter.  Replicate runs are
designated simply by the appropriate punch number (usually "2").
Note: the program creates a file name using the last six characters
of the sample ID plus the run number; if the program finds -another
file with the same name, it will request that a new run number
be input so the existing file will not be overwritten.

Input the appropriate punch size (normally 0.503 cm2) and filter
deposit area.  Note that pressing the return key is not necessary.
Also note that if a mistake is made during input of analysis
data pressing  will allow the option to be aborted and
restarted.

When all data is correct, press any key except  to start the
analysis program.  See Figure 16.

Using a small piece of clear tape, attach the previous sample punch
to its thermogram, insuring that the deposit side is up.

Replace the PetriSlide or petri dish containing the filter into
the styrofoam cooler.

The program will purge the oven with He for 2 minutes, after which
data collection will begin.  Readings are collected for 2 minutes
to establish baselines.  At 114 seconds, a warning tone will sound
(the number of beeps corresponds to the analyzer number).  At 120
seconds the analysis start tone will sound.  At that time, push
the push rod in until the stop is against the back fitting.  While
watching the punch, pull the push rod back 1-2 mm to physically
decouple the pushrod from the boat.  If the boat slides back,
immediately push the thermocouple back in and try again.  The boat
cannot be physically attached to the push rod during analysis, as
expansion of the thermocouple as the sample is heated will push
the punch closer to the laser rod and cause erroneous laser signals.

-------
o
I-*

o
o o
c. oo  i
V5 CT"
UJ *H
UJ  '-
z:  at

-------
           All physical adjustments must be made within 10 seconds: between
           130 and 140 seconds analysis time the laser baseline is calculated.
           If the sample is not correctly positioned at the end of 10
           seconds,  press  to abort the program, pull the boat back to
           the idle  zone, and restart the program.   Decoupling the boat is
           most important for a meaningful laser signal.

           The program will proceed automatically from this point without
           further operator intervention.  At the end of the program,
           data is saved to disk, split times are calculated, carbon peaks
           are integrated, and tabular and graphical printouts are produced.
           When the  printer begins, the push rod may be pulled back to the
           idle zone to begin cooling.
                                                  *
           Examine the thermogram for proper laser response, temperature
           profiles, realistic carbon peaks, and the presence of the calibra-
           tion peak at the end of the analysis.  Examine the tabular printout
           to insure the calibration peak counts are within specifications
           (see Section 4.2).  Finally, examine the laser signal at the end
           of the run.  Drooping of the laser signal as the temperature
           is dropping is an indication that the boat was coupled to the
           push rod  and that the sample should be rerun.  If all aspects of
           the analysis appears correct.  Select the appropriate analysis
           flag from the screen that appears at the end of the run
           (Figure 17).  Mark the analysis date after the sample on the
           sample analysis list.  If a problem is found, indicate the problem
           in the analyzer log book and rerun the sample.

        -  Repeat the above steps for additional samples.

4.3.2  System Blanks

       System blanks are run each Monday.  Follow the steps outlined in
       Section 4.3.1 with the following exceptions:

           Use option 7 from the main menu to change to the \SYSBLK
           subdirectory.

        -  Go through all the steps for a normal analysis, with the exception
           that the  punch from the previous analysis is not removed.  Open  .
           the sample port, pull the boat .back into the loading zone, and
           without touching the existing punch push the boat forward into
           the idle  zone, seal the sample port, and proceed with the analysis.

           Use an ID number derived from the current date: e.g., SB0718.

           Calculated carbon concentrations should not be more that 0.2 ug
           carbon.  Values greater than this warrant a second system blank.

4.3.3  Carbonate Analysis

        -  Follow the steps under Section 4.3.1 until the sample punch is
           loaded into the boat.  Pull the boat BACK until the punch is
           centered  under the acid injection port,  taking care that the
           "ears" holding the boat to the push rod are not bent in the process.

           Select option 2 from the main menu.  Enter the sample ID, run
           number, punch size,  and filter size.  Select the purge option and
           start the analysis program.

-------





31
O
O
LU
*— <
LU
tst CO
ZOO
O  a




-------
        -  At 90 seconds elapsed analysis time flush the 25 ul syringe with
           0.4 M hydrochloric acid (HC1).  When the start tone sounds at
           120 seconds elapsed time,  eject 20 ul HC1 onto the filter punch,
           insuring that the needle bevel is turned toward the punch and
           that the needle tip is touching the top of the punch.

        -  When all analyses are underway, flush the syringe with distilled
           water to prevent corrosion of the syringe plunger.

        -  After the carbonate analysis is completed, a tabular summary and
           a. copy of the graph will be printed.   The progam will  automatically
           cycle into the normal OC/EC analysis, using the same sample
           ID.  Push the sample boat into the punch drying area (about 1 cm
           from the first coil of the sample oven).  If the sample punch has
           tipped over during the carbonate analysis, open the sample port,
           reorient the punch, close the port, and proceed with drying the
           punch.   Heat from the oxidation oven will dry the sample in this
           position without, prematurely baking carbon from the sample; the
           sample temperature should not exceed 42  C.  When the  punch appears
           to be dry (wait at least 5 minutes),  start the OC/EC analysis.

4.4   Analyzer Shut-Down

      After the final sample for the day is analyzed, shut down the analyzers
      by:

           Leave the last analyzed punch in the boat and the boat positioned
           in the heating zone.

           Select option 3 to begin the calibration gas injection routine.
           Follow the injection procedures outlined in Section 4.2, with the
           exception that a He only atmosphere is used during the afternoon
           check.

        -  When the analysis is complete, record the calibration  peak counts
           and calculated injection calibration in the logbook.  Any values
           outside the ranges defined in Section 4.2 should be investigated
           and rerun.  Because low values from the end-of-day calibration
           could potentially invalidate the entire day's runs, any deviations
           from the accepted ranges must be noted and the cause defined.

        -  After a satisfactory injection concentration is obtained, use
           option 7 to change to the directories in which analysis results
           were saved during the day.  Use option 8 to print a single page
           summary of the day's analyses in each directory.

           Press  to end the Carbon program.  This is necessary because
           when the Carbon program ends, it sets the analyzer valves such
           that oxygen is flowing through the Mn02 catalyst, allowing some
           regeneration of the catalyst overnight.

           Remove the printouts and attach them to a manila folder labeled
           with the date and analyzer number.  Place on the lab supervisor's
           desk for Level I validation.

        -  Leave the computers and analyzers on overnight unless  the potential
           for power outages or surges exists.  Turn off the monitors overnight
           to reduce the possibility of phosphor burn.

        -  Make a final check of the gas cylinder pressures to insure that

-------
           will  be available to check them again.

        -  Put the samples  and blue  ice in the styrofoam cooler back into
           the sample storage freezer and lock the freezer.

           If the 25  ul syringe was  used for carbonate analysis, thoroughly
           rinse the  syringe with distilled water  and tightly cap all
           solutions.

        -  Lock  the carbon  analysis  room.

4.5  Abbreviated Operational Checklist

     Start-Up:

        - Check  all gas cylinders' pressures and delivery pressures.

        -  Check that all FIDs are lit by holding  a pair of  tweezers over
           the FID exhaust  stack and watching for  condensation.  Relight if
           necessary.

           Check and  readjust if necessary all gas flows at  the analyzer.

        -  Turn  on the computer monitor.

           Insure that the  date on the computer is current.

        -  At the C>  prompt, type "CARBON" to begin the carbon program.

           Insure that the  sample port fitting is  tight and  that the thermo-
           couple push rod  is reasonably snug at the back fitting.

           From  the opening menu, select option 4  to bake the oven for 10
           minutes.

           Backup the previous day's analysis data.

        -  Wipe  the sample  tweezers, petri dishes, and sample punch.

           Begin the  daily  entry in  the carbon analyzer logbook.

           Insure that the  printers  have enough paper for the day and that the
           ribbon is  producing legible printing.

           After the  ovens  have cooled to less than 100 C, perform a leak test,
           involving  isolating the oven and operating the Carle valve.

           Purge the  system of calibration gas injected by the above step.

           Perform the morning calibration injection by selecting option 3 and
           He/02 carrier gas (option 1).  When the analysis  is complete,
           record the calibration peak counts and injection  concentration.
           Insure that these values  are within their proper ranges.

           Change to  the appropriate subdirectory for the samples to be
           analyzed.

           Retrieve the samples to be analyzed from the sample freezer.

-------
Routine OC/EC Analysis

     -  Pull the push rod back to the idle zone of the quartz oven
        and allow the boat/push rod to cool to 50 C or less.

        Insure that the petri dish, tweezers, and punch are wiped clean.

        Based on the analysis list, remove the sample to be analyzed from
        the styrofoam cooler.

     -  Remove a sample punch from the filter.

     -  Record the filter ID in the analyzer log book, along with any
        comments on the condition of the deposit or any other conditions
        which might affect analysis results.

     -  After the boat has cooled to 50 C or less, remove the previous and
        load the current punch.

        Begin the analysis by selecting option 1 from the main menu of the
        Carbon program and inputting the sample ID, run number, punch size,
        and filter deposit area.

        Push the sample into the heated zone at 120 seconds, insuring that
        the boat is not physically coupled to the push rod.

     -  Using a small piece of clear tape, attach the previous sample punch
        to its thermogram, insuring that the deposit side is up.

     -  Replace the PetriSlide or petri dish containing the filter into
        the styrofoam cooler.

        At the end of the analysis, the push rod may be pulled back to the
        idle zone to begin cooling.

        Examine the thermogram for proper laser response, temperature
        profiles, realistic carbon peaks, and the presence of the calibra-
        tion at the end of the analysis.  Examine the tabular printout
        to insure the calibration peak counts are within specifications
        (see Section 4.2).  Finally, examine the laser signal at the end
        of the run.  Rerun any deviants immediately.  Indicate successful
        analyses on the sample analysis list.

        Repeat the above steps for additional samples.

 System Blanks (run first each Monday):

        Change to the \SYSBLK subdirectory.

        Go through all the steps for a normal analysis, with the exception
        that the punch from the previous analysis is not removed.  Open
        the sample port, pull the boat back into the loading zone, and
        without touching the existing punch push the boat forward into
        the idle zone, seal the sample port, and proceed with the analysis.

     -  Use an ID number derived from the current date: e.g., SB0718.

        Calculated carbon concentrations should not be more that 0.2 ug
        carbon.  Values greater than this warrant a second system blank.

-------
   Carbonate Analysis:
           Follow the  steps  under  Routine  Analysis  until  the sample punch is
           loaded into the boat.   Pull  the boat  BACK until  the punch is
           centered  under  the acid injection  port.

           Select option 2 from the main menu.   Enter the sample ID, run
           number, punch size,  and filter  size.   Select the purge option and
           start  the analysis program.

        -   At  120 seconds  elapsed  time, eject 15 ul HC1 onto the filter punch.

           Flush  the syringe with  distilled water between samples.

        -   Continue  the normal  OC/EC analysis when  the carbonate cycle is
           complete.

  Analyzer Shut-Down:

        -   Leave  the last  analyzed punch in the  boat and  the boat positioned
           in  the heating  zone.

           Select option 3 to begin the calibration gas injection routine.
           Follow the  injection procedures outlined in the  Start Up section
           with the  exception that a He only  atmosphere is  used.

           When the  analysis is complete,  record the calibration peak counts
           and calculated  injection calibration  in  the logbook.   Any values
           outside "the ranges defined in Section 4.2 should be investigated
           and rerun.

        -   Print  summaries of the  day's analyses.

           Remove the  printouts and attach them  to  a manila folder labeled
           with the  date and analyzer number. Place on the lab supervisor's
           desk.

        -   Turn off  the computer monitors

        -   Make a final check of the gas cylinder pressures.

        -   Put the samples and blue ice in the styrofoam  cooler back into
           the sample  storage freezer and  lock the  freezer.

           If  the 25 ul syringe was used  for  carbonate analysis, thoroughly
           rinse  the syringe with  distilled water and tightly cap all
           solutions.

        -   Lock the  carbon analysis room.

5.0  Quantification

5.1  Calibration  procedures

        The calibration procedures for  the carbon analyzers are of two types:
        the end-of-run automatic injection and the  manual calibration using
        KHP and the  two calibration gases. The  end-of-run  calibration
        consists  of  a  set  quantity of CH4  calibration which is automatically
        injected  by the Carbon program. All  FID readings during the analysis
        run are normalized to this peak to minimize the effects of FID

-------
performance and electronic drift over time.  The manual calibration,
performed twice a year or when a new calibration gas cylinder is
started, establishes the calibration slope used in converting counts
to ug of carbon, as explained in the next section.

The end-of-run calibration occurs automatically at the end of the
analysis run and requires no operator intervention.  The integrated
calibration peak counts should be checked by the operator immediately
after each run to insure that the analyzer is operating satisfactorily.

The manual calibration involves spiking prefired quartz punches with
various amounts of the 1800 ppm KHP and sucrose solutions
(Section 3.1) and injecting various volumes of the C02 and CH4 gases.

A clean blank quartz punch is baked in the oven at 800 C for 10
minutes using option 4 from the main menu of the carbon program.
After the punch has cooled to less than 50 C, the solution is injected
onto the punch using a 20 ul syringe.  The following volumes are
used:

       5 ul KHP or sucrose solution
      10 ul KHP or sucrose solution
      15 ul KHP or sucrose solution (do twice)
      20 ul KHP or sucrose solution
      no injection (as a system blank)
      20 ul acidified DDW only (check of background level of DDW)

The sample port is sealed and the punch is pushed to 1 cm from the
sample oven.  In this position the punch will be about 39 C due to
the heat from the methanator oven.  Allow the punch to dry thoroughly;
the punch will turn from translucent to opaque as it dries.  The
punch must be dry to avoid water vapor effects on the FID.  The
carbonate option from the main menu is selected and started.  This
yields a two-peak thermogram, including the normal calibration peak
at the end of the run.  The integrated peak counts for both the
sample peak and the calibration peak are recorded.

The C02 and CH4 calibrations are also run by using the carbonate
option.  The following volumes are injected:

      100 ml C02 or CH4 gas
      250 ml C02 or CH4 gas
      500 ml C02 or CH4 gas
     1000 ml C02 or CH4 gas
     2000 ml C02 or CH4 gas

Again, the integrated peak counts are extracted manually from the
tabular printouts.

Calibration values are plotted as actual ug carbon vs. the ratio of
the integrated sample peak counts to the calibration peak counts.
Obvious outliers are identified and rerun.  Linear regression is
performed on each set of calibration data individually.  The
calibration slope derived from the C02 injections typically has a
slightly different slope and does not fit as well.  The slope is
calculated from
        m

-------
         and the standard deviation is  calculated  by


                
-------
         the  data  is  merged into the final database (Section 7.3).


6.0  Quality  Control

6.1  Performance Testing

         System blanks  are performed at the beginning of each week to
         insure the system is not introducing bias in the carbon results and
         to insure that the laser signal is not temperature dependent.

         Contamination  is potentially due to:

               Operator practices,  such as improper cleaning of tweezers and
              jpunch.

           -  Teflon particles on the push rod are getting into the heated
               zone of  the quartz oven.

           -  The sample boat is contaminated.

               The carrier gases are contaminated.

         A temperature-dependent laser signal is potentially due to:

               Physical coupling of the push rod to the boat during the run.

           -  A quartz rod ready for replacement.  Microscopic cracks in the
               quartz rod will increase internal reflectance of the laser;
               as  the number of these cracks multiply, the effect of
               temperature on these cracks, and thus on the reflectance,
               becomes  an interference in the laser signal.

         As described in Section 5.1, the calibration peak at the end of each
         analysis  run serves as a regular standard; the integrated area under
         the  calibration peak serves as a measure of analyzer performance.  In
         addition, the  injection of two calibration gases daily further serves
         as standards.   Primary standards in the form of spiked filter punches
         do not yet exist.

6 . 2  Reproducibility  Testing

         Replicates of  analyzed samples are performed at the rate of one per
         group  of  ten samples.  The replicate is selected randomly and run
         immediately  after a group of ten is completed.  The ug/cm2 values
         for  OC, EC,  and TC are compared with the original run.  The values
         should fall  into the following criteria:

             Range              Criteria

          <=  10 ug/cm2         <= +-1.0 ug/cm2
          >  10 ug/cm2         <= 10 % of average of the 2 values

         Notice that  the criteria merge at 10 ug/cm2.   Replicates which do
         not  fall  within the above criteria must be investigated for analyzer
         or sample anomalies.  Analyzer anomalies include poor response (as
         reflected in the calibration peak areas) or poor laser splits.
         Typical sample anomalies include inhomogeneous deposits or
         contamination  during analysis.  Inconsistent replicates for which
         a reason  cannot be found must be rerun again.

-------
6.3  Control  Charts  and Procedures

         Three  types of control  charts  are  updated  at  the beginning of each
         week.   These charts  include a  month  of  data and are posted in the
         carbon room until the month is complete, after which they are filed
         with the  raw analysis results.

         The  first is a plot  of  calibration peak counts as percent deviation
         from a historical mean  versus  date (Figure 16).   Instances where
         the  calibration peak area  deviates by more than 10% from the
         historical  mean must be investigated and the  cause corrected.  The
         historical  mean covers  the previous  3 month's results and is updated
         quarterly,  when the  CH4 calibration  gas is changed, or when extensive
         repairs are performed.

         The  charts  are created  by  running  a  program called CarleQA.EXE.  This
         program requires a data file called  CalibSum.HIS, which contains the
         historical  means for each  carbon analyzer. When the historical means
         are  updated, the values in CalibSum.HIS are altered using an ASCII
         word editor.

         The  second  type of control chart is  a plot of calibration gas
         calculated  concentration versus date (Figure  17).  Separate charts
         are  generated for CH4 and  C02  gases.  Instances  where the calibration
         gas  concentrations deviate from the  historical mean by more than 10%
         must be investigated.  Most frequently, low calibration gas
         concentrations are due  to  poor injection practices, such as failure
         to flush  the syringe with  the  gas  prior to withdrawing a sample and
         slippage  of the plunger during injection into the analyzer.

         These  charts are created by running  a program called SyrinjQA.EXE.
         Again, this program  requires the CalibSum.HIS data file.  The
         discussion  above applies to the calibration gas  historical means as
         well.

         The  third type of control  chart is a plot  of  replicate results.  These
         charts are  generated on a  project-by-project  basis for the purpose
         of visual inspection of analysis precision.   Because the precision
         of the carbon analyzers is strongly  influenced by deposit homogeneity
         (Section  1.5), comparison  of precision  results across projects should
         be attempted only with  caution.

6.4  Analysis Flags

         During Level I validation  (see Section  6.5),  unusual conditions of
         the  deposit or analysis problems are noted on the analysis printouts.
         Errors in pre-analysis  data entry  (e.g., in filter ID, punch area,
         deposit area) are also  noted.

         Flags  are applied to the dBase file  created from the analysis results
         ASCII  file  (see Section 6.5).   The following  flags are used:

               bl   ...  field blank
               b2   ...  lab blank
               fl   ...  filter damaged, outside  of  analysis area
               il   ...  inhomogeneous filter  deposit
               i3   ...  deposit  falling off
               i4   ...  abnormal deposit area, possible air leakage during

-------
                        sampling
               15  ...   non-white sample punch after analysis
               rl  ...   first replicate on same analyzer
               r2  ...   second replicate on same analyzer
               r3  ...   third replicate on same analyzer
               r5  ...   replicate on different analyzer
               v   ...   sample void

         Note that all  results flagged with "v" must include a description
         of the reason  for invalidating the sample in the remarks field.

6.5  Data Validation and Feedback

6.5.1  Daily Validation

         Level I validation is performed by manually checking the tabular and
         thermogram printouts the day after the analysis is performed.   The
         following items are checked on the tabular data (Figure 18):

          -  the filter ID is correct

             for calibration injection runs, the gas type is He/02 in the
             morning and He only in the afternoon

             the analysis date is correct

             the punch  area is correct; errors in entry require that the
             calculated carbon concentrations be recalculated by hand.

             the deposit area is correct; errors in entry require that the
             calculated carbon concentrations be recalculated by hand.

             the calibration peak area is in the correct range (Section 4.2)

             the initial and final FID baseline are within 3 counts of  each
             other;  excessive FID baseline drift is cause for reanalysis

             the lower  laser split time and the upper laser split time are
             within 10  seconds of each other.   If the times differ by more
             than 10 seconds, check that the lower split OC and upper split
             OC differ  by no more than 5%.   OC values which differ by more
             than 5%, unless due to a small change in laser signal resulting
             from an extremely clean or very dark sample, requires reanalysis.

             calculated carbon values for calibration injection runs are
             within 10% of the current mean value for that gas type and that
             analyser.

             filter acceptance runs result in <1.5 ug/cm2 OC, <0.5 ug/cm2 EC,
             and <2.0 ug/cm2 TC.  Filters which exceed these levels must be
             refired.

         Items which are found to be okay are underlined in red.   Items which
         have problems  are circled in red.

         The thermograms are checked for the following (Figure 19):

              the initial FID baseline is flat, indicating that the analyzer
              has been  thoroughly purged before analysis began.

-------
Analyser ttl
                        CARBON  CALCULATION RESULTS
                     Carbon. PAS: P2. 1   02 .'10/89
                                   Carbon.DAT: D2. 1   02/1O/
Analysis ID
Sample ID
Anal ys is
Calculation

307176
02/10/89
02/24/89

15:37
15:01
Anal program  ver
Calib. slope
C a 1i b. i ntercept
Baseline window
'Sample transit

Punch area
Deposit area
                  P2.1   <02/lO/89)          Farm file ver   :
                  23.64  ug C/peak-to-cal ibration peak ratio
                   0.00  uq C
                      1  counts              Laser precision:
                     26  sec                 Calib transit   :

                  0.503  cm2
                  13.8U  cm2
                                            02. 1   (02/10/8'
                                              10 counts
                                              50 sec
Calibration  peak  area:  26jiJ_0__coLints
Initial FID  baseline :    _iiI7_-_ counts
Final FID baseline   :    107 counts
Initial laser         :   1790 counts
Laser Split Time
Lower split : l_523__3.ec
Regular split: 1532_ sec
Upper split : 153t sec
OC Peak 81
OC Peak #2
OC Peak #3
Q<. Peak #4
Lower pyro'd OC
Reg. pyro'd OC
Upper pyro'd OC
EC Peak ttl
E< P*^k #2
•E' Peak ft 3
Peak Area
801 counts
1225 counts
5lu2 counts
2701 counts
943 counts
990 counts
1034 counts
20t.2 counts
154t- counts
75 counts
vac ac
Lower spl it : 1.4 r:'. U
Regular split: 1.4 I'-' . 1
I'5. 5 263.7
Upper split : 1.4 1*.2
1-3.5 264.8
Laser FID Split Time
1781 counts 1554 sec
1793 counts 1558 sec
1804 counts 1562 sec
Carbon
0.71 uq C /punch
1.09 uq C /punch
4.53 ug C/ punch
2.40 uq C /punch
0 . 84 uq
0 . 88 ug
0.92 uq
1 . 83 ug
1 . 37 uq
'.>.07 uq
EC
4.8
4. 8
65. fc.
4.7
64.6
C /punch
C/ punch
C/ punch
C/ punch
C/ punch
C /punch
TC
23. '3 Ug C/cm2
329.3 uq C/f liter
23. -3 ug C/cm2
32'3.3 ug C/f ilter
23.9 ug C/cm2
329.3 ug C/f liter
QC/TC:
£'. /TO:
QC/EC:
          0.80
          0.20
          4.02
Figure 18.  Example Level I Validated Tabular Printout

-------
 E
 a
 u
 no
 o

 w
 a)
j:
H
 (0
•o
 11
 >
 
-------
              the  final  FID baseline  prior  to  the  calibration  peak is within
              3  pixels of  the  calculated  FID baseline;  excessive drift is
              cause  for  reanalysis.

              the  laser  signal during the first  2  minutes  appears near the
              bottom of  the graph  (no reflectance);  an  excessively high
              initial laser is an  indication that  the internal reflectance
              of the quartz rod is too high, either  due to too many internal
              cracks or  a  complete fracture of the rod.  High  initial lasers
              should result in a physical inspection of the analyzer.

              the  calculated initial  laser  line  matches the laser signal
              immediately  after the rod La  pushed  in.   A laser line which
              is too low is an indication that the sample  was  not pushed into
              the  oven in  time; a  laser signal which exceeds the calculated
              initial laser is a symptom  of physical coupling  between the
              sample boat  and  the  push rod, although some  automobile emission
              samples also show this  characteristic; a  spike or a number of
              jumps  in the laser signal indicates  that  the operator had
              difficulty in decoupling the  boat  from the push  rod.   All of
              these  problems are grounds  for reanalysis if severe.

           -   the  laser  signal should dip below  the  initial laser line until
              oxygen is  introduced at 550 C, at  which the  signal should rise
              steeply.

              the  laser  at the end of the analysis is flat; if the laser signal
              dips as the  oven begins to  cool, the boat is physically coupled
              to the push  rod  and  the laser signal during  the  rest of the
              analysis is  suspect.

              the  temperature  readings reflect stable temperatures at each
              level  and  smooth, quick transitions  between  levels.

         Problems  or deviations from  normal should be circled  in red.  If the
         sample  punch taped to the thermogram  is not white, it is also circled.

         If examination  of the tabular and  thermogram printouts result in
         a  decision  that a sample  should  be reanalyzed,  write  "Rerun" in
         red  on  the  printouts  and  prepare a reanalysis  list.  This list should
         be posted immediately after  the  validation  is  complete, and those
         samples should  be rerun as soon  as they can be conveniently fit into
         the  current day's analyses.

         Evidence  of persistent analyzer  problems  must  be  resolved, either by
         physically  examining  the  analyzer  or  reviewing the problems with
         the  analyzer operator.

6.5.2  Validation  of Final Data File

         The  following steps are followed to create  and do Level "I 1/2"
         validation  on carbon  data:

         Obtain  copies of  the  latest  version of  the  summary file from the
         directory corresponding to the desired  project.  These files are
         called  CPeaks.n,  where n  is  the  carbon  analyzer number.  These files
         may  be  either restored from  the  backup  files or copied directly from
         the  carbon  analyzer computers.  The  latter  method is  recommended,
         from the  standpoints  of ease of  use and of  a guarantee that the
         summary files retrieved are  the  latest  versions.   These files are

-------
 updated at the end of each analytical run, so the latest version is
 necessary to insure that all of the analyses are included.

 Copy the files together using the DOS COPY command:

    COPY CPEAKS.* TEMP
    REN TEMP CPEAKS.ALL

 The first command copies all of the available summary files into one
 file called TEMP.  The second command renames the TEMP file to
 CPEAKS.ALL, the designated name for the combined files.  Note that
 the two commands cannot be combined; i.e., don't try

    COPY CPEAKS.* CPEAKS.ALL

 This will not work, because at some point the computer will be trying
 to copy CPEAKS.ALL into itself.

 The ASCII CPEAKS.ALL file is reformatted, sorted, and placed into a
 dBase III file by the following:

   DBASE
   DO INPCAR

 The INPCAR program will prompt the user for the ASCII input file name
 (CPEAKS.ALL).  It will also ask for an output file name.  The
 dBASE file naming convention calls for a name in the following format:

   xxOEnnt.DBF, where    xx is the 2 character project identifier
                         OE stands for organic/elemental carbon
                         nn is the 2 digit batch number (generally
                            used to distinguish between different
                            projects for the same client or between
                            sampling quarters for an extended project)
                         t  stands for the sample type:
                              W = woodstove
                              A = ambient
                              P = generic point source
                              D = diesel emissions
                              G = gasoline emissions
                              E = mixed vehicle emissions
                              F = field/ag burn emissions
                              X = mixed types

 The final dBase file name is specified in the analysis list posted in
 the carbon room.

After the INPCAR program produces the dBase output file, the program
will alert the operator that it is ready to print the contents of that
file.  NOTE: wide carriage printer is strongly recommended.

After the printout is produced, immediately label the top of the
printout with the file name and printout date.

Begin validation by matching the filters listed on the analysis list
with the filters listed on the dBase printout.  There must be at least
one entry on the printout for every filter listed on the analysis list.
Flag field and lab blanks as the list is reviewed by placing "bl" or
"b2" in the second column of the printout.  Because the dBase printout
is sorted by ID number, replicates and reruns will be grouped together.

-------
Indicate missing data by writing the missing filter ID in the margin
with an arrow drawn to the appropriate place of insertion.  Scan the
printout for unusual IDs which may have been mistyped during analysis.
Generally these will appear at the beginning or end of the printout
due to the sorting process.   Make sure also that any samples listed
on a rerun list appear on the printout.

Resolve all missing data.  If a large amount of data is missing because
of analysis in the incorrect subdirectory, it is generally easier
to retrieve the summary file from that incorrect subdirectory,
trim the unnecessary data from that file using a word processor,
combine the remaining data with the CPEAKS.ALL file, and rerun the
INPCAR program.  If this is done, take care that the errant summary
file, also called CPEAKS.n,  does not overwrite the CPEAKS.n that
already exists.  If only a few data points are missing, it is
generally not too much trouble to simply write the correct values on
the printout and add those values manually to the dBase files at the
same time the flags and other corrections are made.

Scan the deposit areas for incorrect entries.  Circle the incorrect
entries to insure that corrected values replace those currently in
the database.

Scan the filter IDs for multiple entries of ID numbers.  Under normal
conditions, the only times multiple entries should occur are reruns
and replicates.  All multiple entries must be flagged to indicate
the reason for their existence.  If no flags appear, draw an underline
to act as a reminder to look for the reason as the files are reviewed.

Scan for missing runs.  The most common example is the first run being
aborted or lost for some reason, and the only entry in the dBase file
is the second run.  An entry for the first run must be inserted,
flagged as invalid, and labelled as to the reason it was invalid.  All
punches taken from the filters MUST be accounted for.

Scan the OC and EC columns looking for unusually high or low values.
At this time make sure that the field blanks and/or lab blanks are
all close to one another.  Circle any possible outliers for further
investigation.

Finally, pull the analysis files and go through the analysis summaries
and thermograms one by one.   At this time, resolve all circled items
and all missing flags.  Determine if analyses flagged by the operator
as "SL" or other such flags are legitimate.  If not, draw a line
through the flag to indicate it should be removed.  If the sample
should be rerun, add it to a rerun list.  If the analysis has some
anomaly but still appears to be legitimate, either flag or add notes
to the comments field as appropriate.  Analysis flags are defined in
Section 6.4.  All samples flagged as invalid must have an entry in
the comments field to describe why the sample is invalid. The
following notes and comments are commonly used:


    Comments                      Description

     "Anomalous laser"          Despite good initial laser, laser signal
                               drifted above initial laser signal
                               before dropping  (typical of auto
                               emissions)
     "Operator error"           Used with  "v" flag; operator missed

-------
                              pushing boat in, pushed abort button,
                              pushed manual advance button at an
                              inappropriate time, etc.
   "Analyzer malfunction"     Used with "v" flag; analyzer
                              malfunction or problem beyond the
                              control of the operator such as plugged
                              FID, broken oven heater, etc.
   "Poor replicate"           Replicate is outside the normal
                              criteria, but no reason can be found
                              for the discrepancy.
   "Poor initial laser"       Used with "v" flag; severe coupling or
                              boat not pushed in time for calculation
                              of initial laser signal.
   "Sample contaminated"      Used with "v" flag} rerun of sample
                              yields lower values or different peaks.
                              Typically used with blanks or reruns of
                              replicates.

All flags generated during the analysis must be either converted to
the flags and/or comments listed above or removed.  These flags are
temporary flags only and are not recognized as legitimate analysis
flags at DRI.

After all thermograms have been reviewed and all possible reruns have
been identified, post the rerun list in the carbon room and have the
reruns done as soon as possible.  Review the data from the reruns,
looking for inconsistencies.  Insure that the reasons for the rerun
have been addressed.  Mark the printout with the new values for manual
insertion into the dBase file.  Previous runs must be flagged as
invalid or the reruns flagged as replicates.

Finally, all comments, flags, insertions, and other changes made to
the printout are entered into the dBase file.  After all changes are
made, generate a new printout.  Label the new printout with the file
name and printout date.  Forward a copy of the printout and the dBase
file on disk to the person putting the final report together.

-------
IMPROVE SOP APPEND                                        JULY 1989
            APPENDIX 8:   Maintenance Checklist  (Annual  Site  Visit)

-------
IMPROVE SOP APPEND                                        JULY 1989
                      Maintenance Checklist.

                             clean   repair  replace     comments
                             always    if       if
                                   defective defective

            PUMPHOUSE
Enclosure  	 X  	 	
Temperature switch  	 X  	
Contractor box relays 	 X  	
Contractor box wiring 	 X  	
Cooling fans  	 X  	
Pump:   capacitor  	 X  	
Pump:   diaphram	 X  	
Pump:   piston 	 X  	
Pump:   flapper valve 	 X  	
Pump:   gasket 	 X  	
Pump:   sponge filters 	 X  	
              CONTROLLER
Enclosure  	 X  	
Cooling fans 	 X
Control clock 	 X
Fuse 	 X
Relays 	 X
Delay switch 	 X
Transformer 	 X
30-Minute bypass switch 	 X
Wiring 	 X 	

              FILTER MODULE A
Enclosure  	 X  	
Solenoid valves 	 X
Magnehelic 	 X
Vacuum gauge 	 X
Hoses 	 X 	 X
Fittings 	 X
Elapsed time meters 	 X
Toggle switches 	 X
Inlet 	 X  	
Cyclone 	 X  	
Electrical connections 	 X 	
Wiring 	 X 	
Cyclone 0-rings 	 X  	 X
Nuts and bolts  (tighten)  	 X

-------
IMPROVE SOP APPEND                                        JULY 1989
                             clean   repair  replace     comments
                             always    if       if
                                   defective defective

              FILTER MODULE B
Enclosure  	 X 	 	
Solenoid valves 	 X  	
Magnehelic 	 X  	
Vacuum gauge 	 X  	
Hoses 	 X	X	
Fittings 	 X  	
Elapsed time meters 	 X  	
Toggle switches 	 X  	
Inlet 	:	 X 		
Denuder (Module B) 	 X  	
Cyclone 	 X 	 	
Electrical connections 	 X 	 	
Wiring 	 X	 	
Cyclone 0-rings 	 X 	 X	
Nuts and bolts (tighten)  	 X	
              FILTER MODULE C
Enclosure 	 X 	
Solenoid valves	 X
Magnehelic 	 X
Vacuum gauge 	 X
Hoses 	 X	X
Fittings 	 X
Elapsed time meters 	 X
Toggle switches 	 X
Inlet 	 X 	
Cyclone 	 X 	
Electrical connections 	 X 	
Wiring •	 X 	
Cyclone 0-rings 	 X 	 X
Nuts and bolts (tighten) 	 X

              FILTER MODULE D
Enclosure 	 X 	
Solenoid valves 	 X
Magnehelic 	 X
Vacuum gauge 	 X
Hoses 	 X 	 X
Fittings 	 X
Elapsed time meters 	 X
Toggle switches 	 X
PM10 Inlet 	 X 	
Cassette manifold 	 X 	
Electrical connections 	 X 	
Wiring 	 X 	
0-rings 	 X 	 X
Nuts and bolts (tighten) 	 X

              STAND
Structure 	 X 	
Sunshield and bench 	 X 	

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing}
1. REPORT NO.
 EPA-450/4-9Q-008a
                              2.
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
  IMPROVE Progress  Report
       Appendix A
                                                            5. REPORT DATE
                                                              May  1990
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
  Marc Pitchford
  David Joseph
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Environmental Monitoring Systems Laboratory
  U. S. Environmental  Protection Agency
  Las Vegas, Nevada   93478
                                                            10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
                                                            13. TYPE OF REPORT AND PERIOD COVERED
  Office of Air Quality  Planning and Standards
  U.S. Environmental  Protection Agency
  Research Triangle  Park,  North Carolina  27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
       In Section  169A  of the Clean Air Act as  amended August 1977,  Congress  declared
  as a national goal  "the prevention of any future,  and the remedying  of any  existing,
  impairment of visibility in mandatory class  I  Federal  areas which  impairment results
  from manmade air  pollution."1   Mandatory class  I  Federal areas are national  parks
  greater in size  than  6000 acres, wilderness  areas  greater in size  than 5000  acres and
  international parks that were  in existence on  August 7, 1977.2  This  section required
  the Environmental Protection Agency (EPA) to  promulgate regulations  requiring States
  to develop programs in  their State Implementation  Plans (SIPs) providing  for visi-
  bility protection in  these areas.  EPA promulgated these regulations  on December 2,
  1980.3

       This report  summarizes the progress made  to  date in developing  and implementing
  the interagency monitoring network which supports  the effort, Interagency Monitoring
  of Protected Visual Environments (IMPROVE).
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Group
     Visibility  monitoring
     State Implementation Plans  (SIP)
     Class I  Federal  Areas
18. DISTRIBUTION STATEMENT
 EPA rorm 2220-i rRev. 4-77)
                                               19. SECURITY CLASS (This Report/
                                                                          21. NO. OF PAGES
                                               20. SECURITY CLASS (Tins page/
                                                                          22 PRICE
                       PREVIOUS S1D1TION IS OBSOLETE

-------

-------