-------
Energy Dispersive X-Ray Fluorescence  Analysis  of  Air Filters
     2.  Filter loading is done in  the  laminar  flow  hood.   Turn the hood
         blower and lights on and clean the work  surface with a Kaydry
         towel and methanol.  Lay out a Kaydry'towel in  the hood.   Clean a
         pair of forceps and the filter loading block with  Kimwipes and
         methanol and place on the  towel.  Clean  an  2RJ  sample case inside
         and out and place in the hood.  Obtain the  filters to be  analyzed
         from the sample storage cabinet.  Place  the filters to be loaded,
         a set of numbered filter holders, a box  of  Kimwipes, a methanol
         container, the XRF LOG and a pen in the  hood.   The filter containers
         should be labelled with their  ID numbers.

     3.  Clean the filter holders with  Kimwipes and  methanol.  Pay particular
         attention to surfaces which the filters  will contact.   Wear
         disposable gloves to avoid excess contact with  methanol.   When
         finished cleaning the filter holders,  remove gloves and rinse hands
         if gloves are of the talced variety.

     4.  Select filter holder I, place  on the filter loading block, and
         remove the retaining ring.  If the blank has already been loaded
         from a previous run, select filter holder 2 instead of 1.
     5.  Select the filter whose position number  as  indicated on the XRF LOG
         corresponds to the filter holder.  Note  the condition of  the filter
         and record any comments on the XRF LOG that may apply to  XRF analysis,
         such as uneven deposit, wrinkled filter, water  spot on filter,  etc.
         Filters with loose deposits should not be analyzed.
     6.  Remove the filter from its container and place  deposit side down
         in the filter holder.  Use forceps and handle the  filter  only around
         its perimeter.  If the deposit is touched by the forceps, clean them
         before proceeding.
     7.  Replace the retaining ring so  that the filter is held in  place
         without any wrinkles or other  misalignment.
     8.  Place the numbered filter holder into  the like  numbered location in
         the sample case and the filter container in the proper numbered
         location in the filter container tray.   The blank  filter  holder and
         container can both be placed in the large section  of the  filter
         container tray.
     9.  Repeat steps 4-8 for the remaining filters.   The quality  control
         standard is permanently loaded and is not put into the sample case.
    10.  Verify that the sample ID number to position number relationship is
         correct.  Each filter holder number must match  its position number
         in the sample case.  The ID number on each  filter  container must
         correspond to its position in  the filter container cray exactly as
         on the XRF LOG.
    11.  Place the filter container tray on top of the filter holders in the
         sample case, put the XRF LOG in the sample  case, and transfer to
         the' TEFA.
    12.  Clean uo the laminar flow hood.
                           (Rev. August,  1981)

-------
Energy Dispersive X-Ray rluorescence Analysis  of  Air  Filters                      13
5.3  Analyzing Filters
     1.  Verify that TUBE VOLTAGE is 5.0 Kv  and  TUBE  CURRENT  is  1  uamp.
         Press che SYSTEM CONTROL STANDBY buccon.   Make  cercain  che red
         X-RAYS ON lighcs are out and  che green  X-RAYS OFF  lighc is on.
         Never open che excicacion chamber cover unless  chis  is  che case.

     2.  Life che excicacion chamber cover co ics  open posicion.

     3.  Open che sample case and place che  filter  concainer  Cray  in che  lid.

     4.  Transfer each filter holder in sequence co che  like  numbered
         posicion in che TEFA sample Cray.  Make sure che seating  is
         proper by rotacing che filcer holder and  align  Che holder so
         chat ics indicacing line poincs coward  che center  of che  cray.
         Close che sample case.  If che standard is noc  already  in
         posicion, obtain ic from che  scandard cabinec and  place in posicion
         on che sample cray.

     5.  Verify chac all samples are correctly loaded in cheir designaced
         positions on che sample cray.

     6.  Make sure che excicacion chamber gaskec is properly  seaced and
         close che cover.

     7.  Push che X-RAY PATH PUMP buccon.  The pump will quiec down afcer
         10-20 seconds if che chamber  is properly sealed.

     8.  Press che SYSTEM CONTROL ON buccon.  The red X-RAYS  ON  lighcs
         will come on and che green X-RAYS OFF lighc will go  out.

     9.  Selecc a formacced, zeroed disk and wrice  the PROJECT NAME on che
         disk label.   Use a felc cipped pen.  Record both che data disk
         and program disk numbers on the XRF LOG, as well as  INITIAL, DATE,
         TIME, and ANALYSIS SEQUENCE.  Obtain che analysis  sequence number
         from che laboratory supervisor.  Insert che daca disk inco
         drive 1 (righc hand side).

    10.  Record che dace, projecc, inicial, and  daca disk number in che
         TEFA Log Book.

  '  11.  Press che DATA ENTRY X-RAYS buccon co recurn co speccrum  node, then
         press OPERATING MODE TERMINAL.  Type "C" while holding  down che
         CTRL key.  Type R ORACL .   Type RUN XR .   Encer daca
         as requested by che program and displayed  on che CRT. If  Chare are
         less than 12 samples,  simply press recurn  when asked for  che next
         sample.   Before verifying che particle  size codes,  check  co see
         chac che X-RAY PATH VACUUM READY lighc  is  on.  Sample analysis begins
         when che parcicle size codes are verified  and proceeds aucomacically
         from sample  co sample  for each excicacion  condicion.

    12.  When analysis is complete,  indicated by 5  beeps, press che DATA ENTRY
         X-RAYS butcon and verify chac cube voltage is 5.0  Kv and  cube current
         is 1 uamp.   Press che  X-RAY PATE OFF and SYSTEM CONTROL STANDBY
         buccons.   The green X-RAYS  OFF light should come on and the red
         X-RAYS ON lighcs should go  out.
                            (Rev. August,  1981)

-------
Energy Dispersive X-Ray Fluorescence  Analysis  of  Air  Filcers
    13.  Remove che data  disk,  and  place  in  storage.   Advance the printout
         to top of the page and give  printout  to  laboratory supervisor
         for reviev.
    14.  When excitation  chamber has  reached room pressure, check that red
         X-RAYS ON lights are  out  and green X-RAYS OFT light is on, and
         open chamber cover.

    15.  Transfer in sequence  each filter holder  from the  TEFA sample tray
         to its like numbered  position in the  sample  case.   If the next
         run will use the same blank  and/or quality control standard, they
         may be left in the sample tray.  Otherwise return the blank to
         the sample case  and the standard to its  container in the standards
         cabinet.
    16.  Close the excitation  chamber cover, place the XRF LOG in the
         sample case and  transfer  to  the laminar  flow hood for unloading.

5.4  Unloading Filters

     1.  Filter unloading is done  in  the laminar  flow hood.   Turn on the
         hood blower and  lights and clean the  work surface  with a Kaydry
         towel and methanol.   Lay  out a Kaydry. towel  in the hood.  Clean a
         pair of forceps and the filter loading block with  Kimwipes and
         methanol and place on the towel.   Place  the  sample case in the
         hood.  Record your initial on the  XRF LOG.

     2.  Select a filter holder and the corresponding filter container from
         the sample case.  Check that the sample  ID matches position number
         exactly as on the XRF LOG.   Place  the filter holder on the loading
         block and the filter  container next to it.
     3.  Remove the retaining ring and transfer the filter  to its container.
         Usually it is easiest to  hold the  filter  holder over the open filter
         container and carefully invert the filter holder  so that the filter
         drops deposit side up into its container.  Sometimes use of the
         forceps is necessary.  Replace the filter container lid and set
         aside.  Set che  filter holder aside.

    '4.  Repeat steps 2 and 3  for  the remaining filters.
     5.  Return the filters to the sample storage  cabinet,  the filter
         holders to the standards  cabinet and  the  XRF LOG  to the laboratory
         supervisor.
     6.  Clean up the laminar  flow hood.

5.5  T£~A Shutdown
     1.  Verify that tube voltage  is  5.0 Kv, and  tube  current is 1  jamp.

     2.  Press the SYSTEM CONTROL  STANDBY and  then SYSTEM  CONTROL OFT buttons.

     3.  Press HV1 on the ZEDS II  control panel to turn off  the high voltage.
                           (Rev. August,  1981)
14

-------
Energy Dispersive X-Ray Fluorescence Analysis  of Air  Filters
                                                                              15
     4.  Turn off the printer.
     5.  Press the PWR button on the EEDS  II control panel  to  cum  off
         power.
     6.  Press the SYSTEM POWES. OFT button.
     7.  Switch CONTROL POWER to OFF and place key in  the standards  cabinet.
     8.  Return che XR program-disk to storage, and close the  disk drive
         doors.
 6.   Calculations
         The count rate, I., of x-ray photon events from element i measured by
     the Si (Li)  detector is proportional  to the mass of element i,  M., tines an
     attenuation  factor, A., which takes into consideration the absorption
     of the primary excitation and characteristic x-rays in the sample.
     That is,
          I.  - K.M.A.
                                                                 (1)
     where  1L  is  a  proportionality constant.   An unknown mass, M. ,
     may  be determined  by comparing the count rate measured for the
     unknown with the count  rate,  Is,  measured with a known standard
           5                       ^
     mass M. .
          M.
            i
                                                             (•2)
where C. is a calibration factor,

       C. '
                 I5.
                                                                 (3)
    The uncertainties in M. and C. are given by
    ana
           aM.
           aC.
               M.
               C.
                        cC,
                         C.
 I.
ii
 .3
 i.
                                                            (4)
(5)
           aMS
            SlJ
                   0.05  per  manufacturers  specifications.
                           (Rev. August,  1981)

-------
   Energy Dispersive  X-Ray  Fluorescence Analysis of Air Filters





                                                  Tnble 3
rxr. l TAT T riM r
                                                                            r T OH:
rxrr i AT
•-r : r i RI
Ti'iN r.riNniTION: 1
ANODE '- MI '
JJU.TER: HO
CURRENT: 2OO
FNEROY:0- 10
IN RFriTi
iNS OF I.NTFRF
KpruriN rilANNFI
I

•":
1
r:"
A
7
!~!
O
1")
1
'2
3
/I
5
A
••»
o
•~t
2O
2 1
2 2
-. ~i







J 56-

J 97-
2l2~
227-
217-
25:r'.~
276-
?92~
310-
329™
31 ft—
•-i ^ .-;. 	
'"if-!'"'---
1 i 2-
52 t-
511-
r* *^ t-» 	
597 -
62:":--
A57--
790-
•-••27-*







JA4
191
205
220
235
251
269
28.1
302
320
339
ft5ft
379
;.';•;>•:>
126
537
5AO
P.!-: 4
6 1 3
63C'
AA4
808
815







A.
7.
7-
8.
9.
9.
10.
1 1.
1.
'2.
~-',.
•~' ,
1.
*%
16.
2O.
21.
'2?
.C. i, •
.Vi '
2*'-- .
31.








ST

FNFROY
21-
32- -
•"; f '• __
Ifi-
O8-
PiF:—
•T>'7- —
01-
68-
lo-
16
9 2 —
76-
'riA—
19-
81--
76-
7 ;•-.
/ ^_
•'.-; ".' -.
?R-
60 -
!">:•':-







A.
7.
ft.
ft
9.
ig.

j. ".
2.
2-
3-
1 .
n
5.
17.
21.
2 2 .
2 5
£.. . •
21 -
-ilr-
2^-.
32.








5A
61
20
fio
10
01
76
21
Oft
80
56
• -•2
16
'-.'.'-.
01
18
10
•-',/.,
f-^. .-'-.
•f, A
5 A
32








V-'i i| 1 Attl : 1 '"•
f NFFTsY: i>- 1 O
sppr TIVI IM RpruriN---. np jMTFrJpsi
I 1 18-- 132 1 . 18 - J . •:;• Mil
HJ- 155 1.11- 1.'".5 AI
3 167- 15-si 1 .67 - l.ftl SI
... p|v.t-...-i 1 J9:*:-- '•{'>•-•• ] •":•--:- ~- <-.•-• [-•
i-i.hnc.Ni r. -.74.. ':.-..:; :-:". 21 - ^:" ::ft 1
PF 6 '!AS - :j:7A 3.65- 3.76 i"A
Nf 7 A,-:l- /.^o A. •-:!-- A./JO pr
CU
7.N
GA
HO
AS FVrtTATinN r.fiNrm Ji'iN: ":
Sp Af'|i"irifr : t|
DR FI1 1FR: i H
PB VHI fAriF". -:S
f"
-------
Energy Dispersive X-Ray  Fluorescence  Analysis  of  Air  Filters
       Finally,  che  corrected  counts  are divided by analysis live time, L,
to obtain  the count  rate,
               T.
       and  OI.  ,
              1        L
       Calculations for instrument  calibration  are  similar  except that  C.
is solved for in eq .  3.  Also,  since  single  element and  non- interfering
multi-element standards are used  for  calibration,  spectral  corrections  are
no t made and  T . • M    and   aT    =  aN  ,
               11           i      i
7 .  Quality Control
           The filters are loaded and unloaded  into specially machined
acrylic holders.  Filter loading  is done in  the  laminar  flow hood and the
filters themselves are handled with forceps, out of che  analysis  and
deposit area.  The loaded filters are transported  to  and from the TEFA  in
covered sample cases, and the filter  holders are cleaned between  each use.
These procedures prevent the possibility of  sample  contamination.
           To prevent confusion in  identification  of  the samples  when they
are out of their ID coded filter containers, a  sample position number versus
ID number relationship used.  This  relationship  is  established by the XRF LOG
before the filters are removed from their containers  and it is verified after
each handling step in che procedure.
           For each XRF analysis batch of 10 samples, one blank and a quality
control standard  is analyzed.  Measured concentrations  of  the quality control
standard, which contains several key  elements, are  compared with  actual
concentrations (Figure 4).   If che deviacion is more  chan T 2%, all samples
of that run must be reanalyzed.   The  results of  che quality control standard
over a number of runs provides a measure of  the XRF analysis precision.   If che
results show a trend in drift, recalibration is required.
       Several elements,  including K,  Ca, Fe , As, Br and ?b are measured
under more than one of the  three excitation conditions normally used for each
run.   Results of  these elements  are compared for each of the excitation
conditions under which they are  measured.  If agreement  is for outside the
calculated uncertainties,  che  sample must be reanalyzed.

                           (Rev. August, 1981)
-------
                                      ERROR
•p
r~
•IT
        J. it .
        44 !
        4- .
         .i
             L      MEA3! i

           •";,')   4J.  ~o.j __
           •-•'•'   4J." :-:=•, -r-
           --
                                      -  <'> . 4

T T
r.p
rr
MT
        rONCENTRATION
     ACTUAL      MEASURED
       4<'i - xi'j   4 5 . 25-i~  ? . O
       4 4.. •-;<•/   44. 92*-  2.2
       44.90   4?;. (•)•=>•(—  2.2
       4T. -;O   4/-..94T--  2 .:"»''
       44.xO   44.i::(")H —  •?.2
                                    V.  ERROR

                                          1. 1
                                   i       1.4
                                          0. 4
        CONCENTRATION'        •/  ERROR
     ACTUAL      MEASURED
       AT.. :=:i')   4O . •:>•:>.+._  •-•. 0,4.      r:,. =;
       44 . •-:<•>   44 . •-'•='•+—  2 . 2A.      1 . ••-•
       44. •:•(-,   45. .-;•:>+_  2. 23      i.l
       47. ":'.")   47. 03n—  l!. ::••-.   —  '"'. ?
       44'. ;5'"i   44. '^V4-*-—  2. 1)5
       4--'. 7'j   4'r'. 42-i—  2. 32
 pr  FA:;.T  ! 1
ift ^TANDAPD  iTl'?(")43
 |       r.riMf.ENTPATTON        V.  ERROR
     ACTUAL      riEAil.iPED
 T     41-,. ;=•{•.!   40.'"'1H—  '.";.'Il=:.      '"'.::
 =1     44 . .":'")   44 . ^7 -i—  *'. *"":      •"). .--.
 =•     4J.. •:•!•)   4!=:.'"•":-!—  2.27      <"'..!:
 T     4-.^('i   47 . ••"I'Tn—  I:. V-.   -  •">.=;
                                       RP''P
       J./'.
                 44,
                                               (Rev. August,  L9B1)
-------
Energy Dispersive X-Ray  Fluorescence Analysis  of  Air  Filters
8.  Quality Assurance
       Calibration of the instrument  is by  thin  film  standards  prepared
by Micromatter, Inc., Eastsound, WA. , and by solution deposited and
particle standards made by Columbia  Scientific Industries, Austin, TX.
Plots of instrument response versus  atomic  number  should yield  smooth
curves, and are used to validate standard concentrations.
       Interlaboratory comparisons are used Co check  accuracy of  the
instrument calibration.  Several techniques have been used in interlaboratory
comparisons in addition to energy dispersive x-ray fluorescence,  including
neutron activation analysis, optical spectroscopy, wavelength dispersive
x-ray fluorescence, and ion chromocography.
       In house intermethod comparison vith neutron activation  analysis is
done routinely.  This independent method provides comparison with SRF on
about 20 elements.
       Results from each analysis run are reviewed by  the laboratory super-
visor.  The inter-excitation and quality control standard results are checked
as described in section 7.   The laboratory supervisor must initial the
XRF LOG before results of that run are considered valid.  Periodic audits
of the TEFA startup,  filter loading and unloading,  analysis,  and TErA
shutdown procedures insure  compliance with the Standard Operating Procedure.
                          (Rev. August, 1981)
-------
-------
                        QUALITY ASSURANCE PLAN

                (ORGANIC AND  ELEMENTAL CARBON ANALYSIS)


1.  QA Objectives in Terras of Precision, Accuracy, Completeness,  Representa-

    tiveness ,  and Comparability.

    a.   Precision.  Replicate measurements of organic, elemental,  and  carbon-

        ate carbon should correspond to a standard deviation of ±15%.

    b.   Accuracy.   Analyses will be expected to be accurate to within  ±152.

    c.   Completeness.   Analytical results will be obtained for essentially

        all samples delivered to OGC.

    d.   Representativeness.  Not applicable,

    e.   Comparability.   All data will be reported in yg of carbon  per  m^ of

        air at 25*C and 760 mm.

2,  Sample Custody.

         Mr. Cliff Frazier will be responsible for  the custody of the

    samples.

                                                                   s,
3.  Calibration Procedures and Frequency.

        Calibration is  accomplished in  two ways.  At the  end of each run a

known amount of CH^ gas  is injected into the oven and  the response measured.

Thus, a single point calibration is included in every  run.  The second cali-

bration procedure involves the measurement of filter samples containing known

amounts of specific carbonaceous materials.  This will be done on  a monthly

basis.

  4   Analytical Procedures.

          The initial configuration of  the carbon analyzer was conceptually sin-

  pie.   An aerosol sample  on a glass or quartz fiber  filter segment was heated

  to 600"C in a He atmosphere for the purpose of volatilizing organic  carbon.
-------
The volatilized organic carbon was oxidized Co C02 in a MnC>2 bed at 900"C,




reduced to CHi» and measured by a flame ionization detector (FID).  Elemental




carbon was measured by introducing 02 into the sample oven and combusting the




carbon to C02 which was measured as above.  Early in the project a difficulty




in this method of speciation was discovered; namely, during the organic anal-




ysis a fraction cf the organic carbon was pyrolytically converted to elemen-




tal carbon.  This was manifested by an increase in the "blackness" of the fil-




ter at the end of the organic analysis.   The degree of pyrolytic conversion




was variable and could not be related in any simple manner to measurable prop-




erties of the sample.




        To overcome  this difficulty,  the  system which is now in use was de-




veloped.  It involves modifications in the combustion method itself, the ad-




dition of an optical system for the continuous monitoring of filter reflec-




tivity, and the automation of the analytical sequence by microprocessor tech-




nology.  The system  is shown schematically in Figures 1 and 2 .  As described




below, the optical system is used to  correct for the pyrolytic conversion of




organic to elemental carbon.




       In  th?  combustion method which is  currently in use,  the filter segments




to be  analyzed  are placed  in  thfe  quartz  sample boat,  the  combustion zone tem-




perature set  to  350°C,  and  the  oven  purged  with  an 02(2%)-He(98%)  mixture.   The




boat is  then  inserted  into  the  combustion zone  in  which  oxidation and volatili-




zation of  organic  carbon into  the flowing 02-He  stream occur.   The volatilized




carbon is  oxidized to  C02  in  tho  Mn02 bed,  reduced to CH»,,  and measured as de-




scribed above.   This step  typically  removes  about  2/3 of  the organic carbon. On




the basis  of  the reflectivity measurements  no  net  oxidation  of elemental carbon




occurs,  and  conversion  of  organic to  elemental  carbon is  minimized during this
-------
                                Vent
r— Loc
cv[xh —
GSV|Xh-- -
He
or
He/02
153 Val
iding Heating 0
JL
xiaation |2\J 1
T.. . A
i ivv 1
-i i 1
j u
[Temperature 2
Control ~{
i
L _ —

ve

micro- 	
—- nrnrps^or

LI
i
i 	
Methanator

c" T r\
r ID
l
Electrometer
i
Integrator
i
Recorder



Figure  1   Block diagran of the carbon analyzer.
-------
 Si photo cell

        Pin hole
                  Quartz light tube

         2 x 3mm oval mirror        \
Sample boat
                Lenses "L
632.8nm
interference
filter
                                                   Oven
                                           \
To
microprocessor
                                                Oxidation zone (Mn
                            He/Ne laser
                                                                 Gas linoi
                                                                          \
                           Figure  2. Combustion oven and optical system.
-------
step.  The latter is apparently inhibited by  the oxygen  in  the  carrier  gas.   The


sample oven is then purged with heliun such that all oxygen is  removed  from the


oven.  During this process the sample remains in the oven,  and  the  oven tempera-


ture is maintained at 350*C.  When purging is complete,  the combustion  zone tem-


perature is raised to 600*C, and the remaining organic carbon is  volatilized


into the helium carrier gas.  This carbon is measured as above.

       For the measurement of elemental carbon the  combustion zone  tecperature


is dropped to 400'C following which the carrier gas is changed  to 02(2%)-He


(98*).  The sample remains in the oven during this  process.  After  100  seconds


under these conditions, the temperature is raised to 500°C  where  it is  held
              •

for 120 seconds and finally to 600*C where it is held for 200 seconds.   During


this step-wise combustion the evolved C02 is measured  as  usual.   The purpose


of the step-wise combustion is to facilitate the pyrolysis  correction.


       Throughout the combustion process the reflectivity of the filter  sample


is continuously monitored by a 633 nm He-Ne laser system.  The  time  course of


the reflectivity is shown in Figure 3.   At 350°C in 02/He little change  in the


reflectivity is observed, indicating no net oxidation  of  elemental carbon or py-


rolytic conversion of organic to elemental carbon.   As  the  temperature is raised


to 600°C in He, however, the reflectivity decreases, indicating  an increase in


elemental carbon.  The step-wise combustion of elemental  carbon  in the third


phase of the analysis results in an increase in the reflectivity corresponding


to the oxidation of elemental carbon.   The pyrolysis correction  is determined


by measuring the amount of elener.tal carbon oxid-ation  necessary  to return the


filter reflectivity to the value it had before pyrolysis  occurred.  The  shaded


area in Figure 3b corresponds to the pyrolysis correction.
-------
Figure 3 (a)  Filter reflectivity  as  a function of  time.


         (b)  FID output as a function of time.


         The  cross-hatched section of the 400-500-600*C 02/He peak


         corresponds to the correction for the pyrolytic production


         of elemental carbon.
         143.9
                                                        1
                                                        -I
                                                        J
 U
 hi
 H
 a
 O
        29.63
         isee.
         1290.
         833.8
                              ^^—t
              350 *C
              02/H«
                    600*C/He
                             400"[C
                                    600*C
                                            Cilibrition
                                     12S3.
 1
 -i
 J
  I

 1
 -l
                                                        J
2960.
                               • Figure 3
-------
        The analytical system is under the control of a microcomputer built




 around a Motorola 6802 microprocessor.  All switching of gas flovs, timing,




 temperature control, pyrolysis correction, analog to digital conversion elec-




 tronics, electrometer functions, signal integration, data stqrage, and data




 outputs are controlled by the computer system.  All data (7ID output, reflec-




 tivity, combustion zone temperature, integrated peak areas, pyrolysis correc-




 tion,  and time)  are stored on a cassette tape which can be analyzed at a later




 date on the OGC  PRIME 350 computer.  The only operator interaction during the




 analysis is to enter the filter identification code into the microcomputer




 and load the sample into the system.




 5.   Data Analysis,  Validation,  and  Reporting.




         Data will  be recorded  on both  cassette magnetic  tape and on the x-y




 plot generated at  the time of  analysis.  The  data  stored  on the tape are pro-




 cessed  by the  OGC  PRIME  350 computer.  The processing involves peak integra-




 tion, blank subtraction,  pyrolysis  correction, and conversion to scientific




 units.   The master  equation 1s:
     „/
   UgC/zn
(peak area)      / *ass of carbon \      .


?calibration\  X   ln calibration 1   x f -^-

            I    \ PfiSK          /
         a  /
                \peak area  /                            segment analyzed
            - blank  (ygC/c^)     x    total area of filter

                              '         volume of air sampled
In this equation "peak area" has already been corrected for pyrolysis as de-




scribed in Section  4.




        Several criteria will be used to validate an individual analysis.  The




'first concerns the shape of the calibration peak.  If injection of the CH^ cal-




ibration gas results in two peaks,  this is indicative of incomplete oxidation
-------
 in the Mn02 oxidation zone.  If this is observed, then the oven must be re-




 packed with fresh MnO^.  The second criterion concerns the magnitude of the cal-




' ibration peak.  If this is significantly low ( £ 15Z with respect to the aver-




 age)  or shows a decreasing trend,  a loss of efficiency in the methanation pro-




 cess  is indicated.  This can be corrected by re-packing the methanator with




 fresh Ni/firebrick.   The trend of  calibration peak areas will be checked at




 the end of each analysis day,  and  relevant comments entered in a QA log book.




 Finally, an optical absorption measurement of each filter will be made to val-




 idate the elemental carbon measurement.  Absorption will be plotted as a func-




 tion  of measured elemental carbon  concentration (ygC/cm2).  Obvious outliers




 will  be re-analyzed.




          The  data flow in this program is almost totally automated.   The out-




 put of the carbon analyzer is  recorded on a cassette magnetic tape which is




 processed on  the OGC  PRBffi 350 computer.   The only operator interaction is to




 enter filter  codes and air volumes into the computer file.   This information




 will  be checked for  accuracy from  a hard-copy printout.   Emily Heyerdahl will




 be primarily  responsible for handling this data flow.






 6. Internal  Quality  Control Checks.




    a.   Standard filter.  At  the  beginning of each day a "standard filter"




          will be analyzed.  The "standard filter" is an actual high volume fil-




          ter  sample  collected  in Portland, Oregon.  Many analyses have been made




          from this filter.  If the standard filter analysis differs from the av-




          erage by -z-z than 2  stizdard deviations, the standard filter will be




          re-analyzed.  If the re-analysis is within the above limit, then rou-




          tine analysis will proceed.   If not, the cause of the problem will be




          investigated and corrected.
-------
     b.    Calibration.  Each run will be individually calibrated.  As discussed

          in Section 5 ,  the shape and magnitude of the calibration peak relate

          to the.analytical efficiency.  In addition, at the Beginning of

          each day CHj, calibrations are performed in both He and He/02 car-

          rier gases.  If these do not agree to within 102, the source of
                          •>
          the discrepancy will be investigated and corrected.

     c.    External Standards.   Filter segments containing known amounts of or-

          ganic,  elemental,  or carbonate carbon will be analyzed once each month.

          These will be  prepared by Richard Johnson and submitted blind to Isiily

          Heyerdahl who  will analyze them.


 " ~^'  Performance  and  System  Audits.

          This  is  accomplished by  the  actions  of Sections  5 and &.


 8.  Preventive Maintenance.

          The primary  components requiring preventive maintenance are the oxi-

 dation and methanation ovens ard  the  optical system.  The diagnostic tests and

 maintenance procedures for  the i7ens were discussed in Section 5 .  in the op-

 tical system the  light pipe between the oxidation oven and the detector occa-

 sionally  deteriorates with  respect to light transmission.  This is manifested

by an apparent decrease in  reflected light from a blank filter.  When this oc-

 curs, the only remedy is to replace the light pipe.


 9. Specific Procedures  to Be Used to Routinely Assess Data Precision,  Accuracy,

    and completeness.

         Precision will  be assessed in two ways.   As discussed in Section 7 ,  a

"standard filter" is analyzed at the beginning of  each day.   These standard fil-

ter analyses are  tabulated and stored in a computer file.   The standard deviations
-------
                                                                      10
 for  organic,  elemental,  and carbonate carbon will be  calculated  in  the  usual man-




 ner.   In  addition,  replicate analyses are performed on approximately  one  out of




 ten  filters.   The  absolute differences (as vgC/cm2) between  the  analyses  will be




 tabulated and stored  in  a computer file.  Standard deviations will  also be  calcu-




 lated.




          Accuracy will be determined from the monthly measurement of  external




 standards.  In addition,  interlaL-oratory comparisons will be conducted  from time




 to time.







 10.  Corrective Action.




          This has  been discussed in  Section 5.






11•  Quality Assurance Reports to Management.




          The  results  of  the  analyses of the standard filters,  replicate



filters,  and external  standards will be transmitted to  Dr.  John A.  Cooper.
-------
                                                             Page 10
     21.  Plot a graph  of  Q,  vs P  on two-cycle, semi-log
graph paper to obtain the  hi  vol orifice calibration line
as illustrated in Figure 2.7.   Use a -straight edge to draw  a
best fit line through the  calibration points.
     22.  If any calibration  point does not fall within  ± 1
percent of the line, rerun that point, recalculate, and  replot.
The percent deviation can  be  calculated by taking the questionable
flowrate (Q ) and the calibration line flowrate  (Q )  for the
same P  reading.
            .   .  a    (Qo  *  Qc)
Percent Deviation  =
                                           x 100
             .»
      .»  V.O I'.l l'.2 TTT 1.4  l.J  1.6 1.7 1.8 1.9
           Flow  rate (Q,),  m /min
    Figure 2.7   Example of hi vol orifice
           calibration relationship
   anot  to exceed + 1 percent
                                                         (Rev. 7/14/81)
-------
                                                       Page 11
  4.   Sampler
      Samplers  must be calibrated when first purchased,
 after- major maintenance on the sampler (e.g., replacement of
 motor or motor brushes) ,  any time the flow-rate measuring
 device (i.e.,  rotameter or recorder)  has to be replaced or
 repaired,  or any  time a one-point audit check
 deviates more than +  77, from the calibration curve.
      In  using  the orifice calibration unit to calibrate a
 sampler,  corrections  must be made to  the indicated flow rate
 if the ambient barometric pressure or temperature is substan-
 tially different  from the pressure or temperature values
 recorded  when  the orifice unit vj^g calibrated.   Calculate the
 corrected flow rate as follows :

               /T P \  1/2
where
       Q2 = corrected flow rate at  sampling  conditions,  m /min;
       Q, = uncorrected flow rate read  from  the  orifice
            unit calibration curve  for  a  given pressure,
            in. H2O;
       T, = absolute temperature when orifice unit  was
            calibrated, °K;
       P, = barometric pressure when orifice unit was
            calibrated, iranHg;
       T- = absolute temperature while  calibrating  the
            sampler, °K, and
       ?2 = barometric pressure while calibrating sampler,
            mmHg.
For example, if P, = P2/ Figure 2.8 shows  the percentage  change
of Q versus temperature differences.  If  T2  is greater than
T., the percentage change is positive;  if T2 is less than T^,
the percent change is negative.  The same  procedure is used
to correct for pressure differences.  The  above  formula  corrects
for field conditions only.
                             (Rev.  7/14/81)
-------
                                                        Page 13
    5.   Calibration of Hi Vol Sampler with Flow Recorder -
 Connect the calibration equipment as shown in Figure 2-13.
     All data  during  the initial and final calibrations for
 one hi  vol  are recorded on  the same  page 'in the Hi Vol Cali-
 bration Log-.    if the sampler is being
 calibrated  for the first time or if  it has been serviced,
 record  the  data on the "Initial" section of the calibration
 log.  If the sampler  is being calibrated prior to scheduled
 maintenance, record the data under the "Final" section of
 the calibration log.
     The stepwise calibration procedure for this model of
 Hi Vol  sampler is presented below:
     1.  Assemble a hi vol  with a clean filter and operate
 it for  at least 5 minutes at 115 volts.   If a step-down trans-
 former  is used during normal operation,  then calibration
 should  be performed with the transformer in operation.
     2.  Record the flow recorder number and date on three
 gummed  labels.   Affix one gummed label  to  the'very top of  the
 metal face  on  the front of  the flow  recorder.   Affix another
 gummed  label to the middle  of the vacuum hose and affix the
 last gummed label to  the other side  of  the hi vol motor.
     3.  Turn  the motor OFF and attach  the flow recorder to
 the hi  vol motor.
     4.  Install  a  clean recorder chart  and  check the recorder
 for proper operation.   Zero the pen  if  necessary.
     5.  Remove  the filter  holder.
    .6.  Attach  the calibrated orifice with  one  of the  load
plates between  the  motor and  the orifice.
     7.  Turn  the motor  ON  and record the  water  manometer
and flow recorder readings  after they stabilize.
     8.   Turn  the motor  OFF.
     9.   Repeat Steps  6-8 for  each of the  other  load plates.

                             (Rev. 7/14/81)
-------
    70
    60
    50
3  40
    30

   20-
HT^
                                                                  Page 14
                                                                            f    r
                   i     i    l .... i
                                          i     !     i    l .  .   I     ii     i     i
      ^^r ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
     0   .8    .9   1.0  1.1  1.2  1.3   1.4  1,5  1.6   1.7  1.8  1.9  2.0  2.1   2.2   2.3
       FLOWRATE  IN M3 MINUTE
     Temperature
     Cali
     Date     m^n \.
                  J
           HI-VOL  CALIBRATION CURVE
Barometric  Pressure  ~7£>O rr>'
          Calibration  log page no.  \ Q _ Hi Vol. No.   5
                         \6. 1Q76 _ Analyst  6 .& .
                              Example  of hi vol calibration curve
                                  (Rev.  7/14/81)
-------
                                                     Page  15
Figure 2-13.   Hi  vol and orifice  assembled for
        calibration with flow recorder.
                  (Rev.  7/14/81)
-------
                                                       Page  16
    10.  Repeat Steps 6-9 once.
    11.  Determine and record the air  flow rate  as  read from
the hi vol orifice calibration curve for each  flow  recorder
reading.
    12.  Record the barometric pressure in mm  Hg and  temperature
in °C.
    13.  Determine the percent difference between the temperature
and barometric pressure recorded and the temperature  and baro-
metric pressure measured when the hi vol orifice was  calibrated.
         a.  If the recorded barometric pressure is within
             +15 percent range and the recorded  temperature
             is within +100 percent range proceed to  step 16.
         b.  If the barometric pressure exceeds  the +15
             percent .range or the temperature  exceeds the
             +100 percent range or both occur  proceed to
             step 14.
    14.  Convert the temperature measured when the  hi vol orifice
was calibrated (T,) and the temperature recorded in Step 12
above  (T2) to absolute temperature  (°K).
    15.  Determine the true flow rates corrected to the baro-
metric pressure and temperature recorded in  Step 12 above.
This is done by substituting each of the flow  rates determined
in Step 15 above for Q, in the following equation and solving
for §2-

         Q2 -
where
        ' Q  = flow rate determined in  Step 15  above
         Q_ = corrected flow rate
         P. = barometric pressure measured when  the hi vol
              orifice was calibrated
         P_ = barometric pressure recorded  in  Step  12  above
         T, = absolute temperature determined  at  the  time
              the hi vol orifice was calibrated,  °K
            = absolute temperature determined  from  Step 12, °K.
          '2

                            (Rev. 7/14/81)
-------
                                                       __ Page 17


     16.  Plot the  flow  recorder  readings vs.  the air flow rates
     17.  Use a French curve  or a curve fitting technique such
as the least squares  fit,  to  draw a  best fit smooth curve
through the calibration  points.
     18.  If any calibration  point does not fall within ± 5
percent of the curve, or causes the  curve to be S-shaped or
have a sharp turn,  rerun that point,  recalculate,  and replot.
The percent deviation can  be  calculated by taking the question-
able flow rate (Q ) and  the calibration curve  flow rate (Q )
for the same flow meter  reading.
                               (Q   -  Q )
          Percent deviation =           x 100
should fall within ± 5 percent.
    Table 2.1  ACTIVITY  MATRIX  FOR CALIBRATION OF EQUIPMENT
Equipment
Analytical
balance
talative
humidity
Indicator
Ti»r
Zlapaed tiaa
Mter
Orifice
calibration
unit
Sampler
Acceptance limits
3 to 5 standard weights
covering normal range of
filter weight, all with
indicated weight • true
weight + 0.0005 g
Indicator reading <*
psychrometer reading
± 6*
^15 minutes/24 hours
+2 minutes/2 4 hours
Indicated flow rate -
actual flow rate +_ 4*
!Q0 - Qel
, ° . ' 0 05
we
QQ » observed flow rate
Q^ - flov rate from cali-
bration curve
Frequency and
method of
measurement
Gravimetric test
weighing at purchase
and when performing
periodic calibration
checks
Comparison with
reading of wet-bulb
dry-bulb psychro-
»«ter on receipt
and at 6 -month
intervals
Check at purchase
and quarterly
against elapsed time
meter
Standard timepiece
(known accuracy) at
receipt and at 6-
nonth intervals
Flow rate primary
standard at receipt
and 1-year interval
Calibration orifice
unit, on receipt and
after major mainten-
ance on sampler
Action if
requirements
not met
Have balance main-
tained and calibra-
ted by manufac-
turer's representa-
tive
Adjust or replace
to attain accept-
ance limits
Adjust and repeat
test
Adjust or replace
time indicator to
attain acceptance
limits
1) Adopt new cali-
bration curve if
evidence of orifice
damage is absent
2) Replace orifice
unit if evidence
of da^cage is
present
Rerun points for
which
|Q0 - Qcl
. ,r > n 01
Qc "
until acceptance
Limit attained
                           (Rev. 7/14/81)
-------
-------
                            STANDARD OPEEATING PROCEDURE

                     HIGH VOLUME .TSP FILTER HANDLING AND STORAGE
1.  General Discussion

    Filters used for High Volume TSP particulate collection are 8" x 10" Sierra
    glass fiber filters.

    Sample ID codes have a two letter prefix, EH for High Volume TSP filters.
    A three digit-number follows the prefix, starting with 001.

    All filter handling procedures are to be done in a clean area.  Clean the
    work area with a methanol dampened Kaydry towel, then lay out a clean Kaydry
    to use as a work surface before all handling procedures.

    Filters are stored in individual file folders, which, in turn, are placed
    inside manila envelopes.


2.  Materials and Equipment

       • Sierra Model C-305-GF glass fiber filters
       • file folders
       • manila envelopes
       • Kaydry towels
       • methanol
       • 4-digit number stamp
       • Vinyl medical gloves


3.  Filter Coding

    1.   Cut file folders such that they will fit inside 9" x 12" manila envelopes.

    2.   Place vinyl medical gloves on hands and clean with methanol dampened Kaydry
         towel.

    3.   Select filter to be coded and inspect both sides for foreign particles,
         pin holes, discolorations, or other imperfections.  Some particles may be
         blown off.  Discard the filter if defective; otherwise, open cut file
         folder and place filter inside.

    4.   Select ID # to be printed on filter using 4-digit number stamp.  Stamp
         ID // on upper right hand corner of top face of filter such that ink will
         not interfere with sampling area.

    5.   Close file folder and stamp on upper right hand corner.  Stamp manila
         envelope with same ID // and place file folder with filter inside.
-------
6.   Repeat steps 1 - 5 as necessary.

7.   Record ID # and lot number of all filters coded in the RDDP sample
     log book.

8.   Place loaded filters in sample storage cabinet.
-------
                           STANDARD  OPERATING PROCEDURE
                  MAINTENANCE FOR SIERRA DICHOTOMOUS SAMPLERS
1. Maintenance  Schedule
   Maintenance  will be performed  on a monthly basis.
2. Cleaning the Sampling Module
   1.  Particulate Deposits
       The Sampling Module is disassembled as shown in Figure 1.  All parts
       are sealed with "0" rings. Particulate internal loss deposits accumu-
       late primarily on the outer and inner surfaces of the tip of the receiv-
       ing tube in the virtual impactor head. (Figure 2)  The remaining internal
       surfaces may have slight  particulate deposits.  These deposits should be
       cleaned with alcohol or water using a. camel's hair "brush.
   2.  Bug Screen
       The bug screen is exposed for cleaning by disassembling the aerosol inlet.
       Brush all loose material  off  the bug screen and out of the aerosol inlet.
   3.  "0" Rings
       The "0" rings  in the aerosol  inlet  and the flow splitting chambers should
       be conditioned with a light coating of vacuum grease.
3. Battery Beplaceinent
       If AC power has not failed during operation, the dry cell battery for
       stand-by power need  be replaced only every  2-3 years.  If AC power
       has failed during operation,  as  indicated by the dot light on the Timer
       Module,  the battery must  be replaced at  the  next regularly scheduled
       maintenance.
       To replace the battery, unplug the line  power cord, and remove the
       front  panel of the  Control  Module by removing the six  screws.  The bat-
       tery is  located in  a bracket  on the  lower  left hand side of the back
-------
  Standard Operating Procedure                                Page 2
  Maintenance for Sierra  Dichotomous Samplers


vail of the enclosure.

Leak Testing

The Control Module is leak-checked by installing needle valves on the

coarse-and fine-particle "bulkhead fittings on the side of the enclosure.

Close the valves and turn on the pump.   Pressures P, and P.  on the vacuum

gages should be -2k to -25 in. Hg-if leaks are negligible.  Then gently

close the flov selector valve for the total flow and quickly shut off the

pump.  Important Note: .If the pump is  allowed to run with the total flow

selector valve shut, the  pump will act  as a compressor and can cause leaks.

If leaks are negligible,  these pressures will increase to about -12 to -15

in. Hg in approximately -20-30 seconds  and then will increase to zero in

about 2 minutes.  The initial increase  to -12 to -15 in- Hg is not caused

by external leaks but rather by a small flow through the pnuematic feed-

back line from the compressor to the vacuum side of the system.


If leaks exist, they most probably are  in the filter jars, which should be

tightened firmly.  The next leak possibility is the tubing fittings or a

tube improperly in contact with the hot pump and has melted through.  Most

leaks can be quickly isolated by use of the inlet valves and the two vacuum

gauges.
-------
                                             FROM  AEROSOL  INLET
   FINS PARTICLES,
   LESS THAN  2.5 MICRONS


.OARSE PARTICLES,
GREATER THAN  2.5 MICRONS
  ASSETTE

     FINE
     PARTICLE
     FILTER
                                                                              INLET TUB
                            0 . 9., CMH
                                                    0.1 CMH
VIRTUAL
IMPACTOR
•NOZZLE

•VIRTUAL
IMPACTOR'
RECEIVER  T
 FILTER-
 CASSETTE

 COARSE
 PARTICLE
 FILTER,
 37mm  Dia.
                                                                              FILTER
                                                                              HOLDER
                                ""Jr ~ -CC5. .TO  CONTROL  MCOULZ
-------
M
o
-------
                                                                  Page  1
           HI VOLUME - GLASS FIBER FILTER WEIGHING USING THE
                  TORBAL EA-1 AP ANALYTICAL BALANCE
                     Standard Operating Procedure


1.   General Discussion

    1.   Analytical Problem:

        The mass of air particulates collected on glass fiber filters
        must be determined to an accuracy of better than ± 10% relative.
        Minimum net weight is expected to be about 25 rag.

    2.   Interferences:

        Humidity changes may affect the weight of the filter and of the
        deposit.   Filters are equilibrated before and after sampling inside
        a Constant Humidity Chamber (CHC) 24 hours prior to weighing.

    3.   Readability:   0.1 mg.

    4.   Precision and Accuracy:

        The balance is  calibrated  with class S weight.   Precision expressed
        as standard deviation is 0.2 mg.

    5.   The technician  should read the Torbal EA-l-AP Instruction Manual
        entirely  before using the  balance,  paying close attention to
        section 3,  "weighing procedure".   The balance is always left ON to
        provide maximum stability.

    6.   Every  tenth filter is a  control filter and is labelled accordingly.

    7.   Keep the  calibration weight  clean and store in its  proper container.
        Handle  the  calibration weight  with  the cleaned weight  forceps  only
        (non-serrated stainless  steel)  and  do not use the weight forceps
        for any other job.
                            (Rev. 6/11/81)
-------
      Standard Operating Procedure                                 Page 2


2.  Equipment and Materials

    • Torbal EA-1 AP Analytical balance  with  easel

    • Constant Humidity Chamber

    • Calibration weight, Class S

    • Calibration weight forceps
    • Kaydry towels

    • Kimwipes

    • Vinyl medical gloves

    • Methanol

3.  Operating Procedure

3.1 Start Up:

    L.  Remove filters to be weighed from their manila envelopes.  Place
        file folder with filter inside into the constant humidity chamber.
        Allow 24 hours to equilibrate before weighing.

    2.  Remove hi-vol filter easel from weighing chamber.  Fiance vinyl
        medical gloves on both hands.  With methanol dampened Kaydry
        towel,  clean:   weighing chamber,  easel, tare weight, balance
        corttrol knobs/ left front  corner of CHC and hands.

    3.  Replace easel to weighing  chamber.   Clean area around balance and
        lay out a clean Kaydry towel on table.  Clean forceps using a
        methanol dampened Kimwipe.

    4.  Record  technician,  date, CHC humidity, room humidity and filter
        ID's on "hi  volume  weigh sheet".   Record whether the filters to
        be weighed  are unexposed or exposed samples.

    5.  Set weight  control  knobs and counter control knob to zero.  Use the
        coarse  tare  control and zero adjust knob to bring null indicator to
        center.

    6.  Open door and  place the calibration weight on easel.   Close door and
        adjust  weight  control knobs such  that  null indicator reads zero.
        Record  the  total  weight in  the  "calib" boxes provided by the data
        sheet.   Open door and remove calibration weight from easel to
        appropriate  container.

    7.   Adjust weight  control knobs to  zero.   Null indicator  should read zero;
        if  not,  repeat  steps  5-7.   When null indicator  complies  with this
        stipulation, check  the  box  to the right of the  first  filters ID //.


                             (Rev.   6/11/81)
-------
      Standard Operating Procedure                                 Page  3
 3.2  Weighing Filters:

      1.   Remove the first filter to be Weighed from the CHC and.check  to
          make sure that its ID // corresponds to the ID if on the data
        •  sheet.  Place filter on easel inside weighing chamber and weigh
          within 30 seconds.  Record weight and note any defects (torn,
          •particles falling off,  etc.) on the data sheet.   Remove filter
          from easel and replace  inside CHC.

      2.   Remove another filter,  following the same procedure as
          described above.

      3.   After every other filter,  the balance must be zeroed.  Note any
          large deviations  from the  null zero in comments  column.

      4.   Repeat steps 1-3  until  weighing of  the set is complete.

      5.   Place calibration weight on easel and record the calibration
          weight.

      6.   Place calibration and tare  weight in their proper containers.

3.3   Reweights:

      1.   Select at  random  three  filters to be reweighed,  preferably by
          another technician.

      2.   Use  the same start-up and filter  weighing procedures  with  one
          difference,  filters from up  to four  sets  can  be  reweighed  at the
          same  time.

3.4   Calculations:

     W '  »  (W   - W   -  AC) x 1000
      n       g     t

     Where:  W  = net deposit in mg
              n

             W  = gross weight in g
              &

             W  = tare weight in  g

        n
        £
       i .  (Gross weight Control.  - Tare weight Control.
AL> —   "~~"  ^^——^—^^-^——^^———^^———*—^^——-^
                                 n

       Where:   n = the number of  control filters in weigh set.

1.   Record W  in the  comments column of the appropriate field  data  sheet.
            n
                             (Rev. 6/11/81)
-------
     Standard Operating Procedure                               Page 4
3.5  Quality Control:

     1.   The calibration weight check at the end of a weighing set must
         be within ± .  0002 g of their correct values or the entire set
         must be reweighed.

     2.   The reweight must all be within ± .0010 g or the entire set must
         be reweighed.
-------
                                                                Page 1
                        Standard Operating Procedure
                      EPA High Volume TSP Sampler Audit"
    fl.O' -"'"AUDITING; PROCEDURE "•"       -     ',' .">  -\t
                                                '
                    .        .                          .
      ~- 'Audit, as used- here, implies an  independent: 'assessment of
 -   the" quality of data obtained by the . ambient'-.air monitoring
. .r-^ methods .   Independence .can-, be achieved by .having the -audit made
    by" "a .•different operator/ analyst tHan the one conducting the
                  > .         w'            ''   •    •«'*•• •
,.f   fouti-Bfe field -measurements,  -Th,e audit should- be-, a true assess-
  -;.•.„   '  .''    "          "      '          •-.'": A*. " '
', .  . ment oif" the,- i&eassurfiment- process under normal ..operation, i.e.,
.,  .withdht v,any special preparation or  adjus-tmenf of^th'e sys^tem.
    Routine- quality assurance  checks -conducted .by  the operator/
    analyst- are necess'ary for  obtaining and reporting good quality
    data, "but 'they are-'not to  be considered as part 'of the auditing
    procedure-.          .-.>•• :-          ..    "        •        ,-. 'j"7 .   ;
        'Two  types of audits' are recommended herein, performance
    and  system audits. '.V Four1; performance audits  and, a systems  ".
    audi£: are 'described.-.-in detail "in the'-'follQ'wing-"sections~. .An\
        .•"•-.,'     , .       •,..-,-.    .  •    '0   '     -. ' '"  ''.;>•'• •' -'V
•  '..alternative to p'erfonning  four 'individual '-''audits' .'is  -t6 '.audit
          - *            •           ,               ,*•-."    "*%"   *
  -  the  entire me-asurement process  by comparing  the final  results^.
    from the  field sampler -to  those  obtained -with  a collocated "..-.
           %        v   .  "    * '   *                   '         *•..
:    sampler".   This alternative  is described, in Section- U. 2 .v^.'.-A \ •   v . ..
  '  summary 'of these audits is -'given in Table^ 8*2  
-------
                                                      Page 2

by the total measurement system  (sample collection, sample
analysis, and data processing).  Performance audits are normally
a quantitative appraisal of quality.
     Four performance audits of individual variables are
recommended:
     1.  Audit of clean filter weighing.
     2.  Audit of flow rate calibration.
     3.  Audit of exposed filter weighing.
     4.  Audit of data processing.
     An auditing level of 7 out of 100 sampling periods is
suggested here as a starting frequency.  For the case where.one
sample is collected every sixth day, an auditing level of one
per month is recommended.  This would result in an auditing
level of approximately 3 for a lot size of 15 for data reported
quarterly.  If the number of sampling periods is greater than
15 but less than 50, four (4) audits are recommended.  These
frequencies are suggested starting frequencies and they are
to be altered based on experience and data quality.  The audit
frequency should be reduced if past experience indicates the
data are of good quality, or increased if data are of poor
quality.  In determining the number of audits to be made for
a large lot size, it is more important to make sure that the
sample is representative of the various conditions that may
influence data quality than to adhere to a fixed frequency.
The supervisor/quality assurance coordinator will specify
the audits and auditing level to be used according to monitoring
requirements.
8.1.1  Clean Filter Weighing Audits - Weighing audits are made
as soon as practical before or after the regular weighing.
Clean filters are normally weighed in batches.  This allows
for the sampling to be performed and corrections to be made
before the filters are used.
     1.  Divide into lot sizes of 100 or less and weigh.
   .. 2.  Randomly select and reweigh 7 filters .from each lot
of 100.  See Appendix I of Volume I of this Handbook for

                          (Rev. 7/14/81)
-------
                                                      Page 3
 recommending  sampling  procedures.
      3.   If any  one  of the  7  check weights differs more than
 2.8 mg from the  original  weight,  reweigh all the filters in
 that  lot.  Record  results in  a  laboratory log book.
 8.1.2  Flow Rate Calibration  Audits -  Independent flow  rate
 calibration audits should be  made  on site.   Portable audit
 equipment is  used.a  Perform  the  flow  rate calibration  audits
 according to  the following  procedure:
      1.  Set  up  equipment.
      2.  Select  one  of the  resistance  plates and obtain the
 actual flow rate,  Q  ,  and the rotameter  reading,  following the
                   a
 calibration procedures, Section 2.6.
     3.  Convert rotameter  reading to  flow rate,  0 ,  using
 the calibration  curve  and making corrections for ambient
 temperature and  pressure.
     4.  Compute the percent  accuracy
               Q_  -  Q
         % A = -2-	- x  100.                 Equation 8-1
                 ^a
      5.   If the  percent accuracy  is greater than ± 7 percent
 for any  one check, a complete recalibration should be performed
 before sampling  is resumed.•
     6.   Report 0  , Q  , and the percent  accuracy,  % A,  on
 an X and R chart, under measurement result,  items  1  and 2.
 Record .% A in the cells preceded by "Range,  R"  as  indicated
 in Figure 8.1.   The  value,  %  A,  can be positive or negative
 and the range is always positive.   The sign  of  the difference
 should be retained to  determine the existence of trends and/or
consistent biases.    The steps involved in the construction
of a quality control chart and in  the  interpretation  of the
results are described  in  Appendix  H, Volume  I of this Handbook. '
aUS EPA uses a reference flow device  (called  an  ReF  device) whic:
 is an orifice (with 5 different orifice plates)  that  mount
 onto the faceplate of the hi vol adaptor.  An ReF device
 may be purchased from Dexco Co., Inc., 630 Chapel Hill  Blvd.,
 Burlington, NC  27215.
                         (Rev.  7/14/81)
-------
                                       ^•("si"1.)""
S,
                                                                              Page 4
O £>,
*". '
Ifl
c
1-1
(D

CO
O
C
0)
I— '
H-
rr
O
O
3
rr
i-l
O
 O
 3"
 0)
 i-i
 Ml

 O
 1-1
 O


 1-1
 Q>
 rr
 CO

 O
 03
 cr
 1-1
 OJ
 rr
 H-
 O
 3

 OJ
 C
 CU
 H-
 rr
Comments
Action .
ate. )














RcryiWbcn
d«*


































RANGES , R
1
























1

















































j j
o
r






-4— C?
















||
II
AVERAGES, X
(In O tn
II !!ll






,+-
^
N



















< J



J I
k











%
H
i ^
•ui r
If



i >
(
s
V

.--'

s •
V











«.
k
^_
j
P
^3
n

i
-f


V

-(-

















\
*



\
'



























1









fr
a
h-U

1 11
O
1



















,i

•4
5
3


























i

























I



















M
It*



M

C
i
i
i
>
3
0<
CO
1
M^
£
-Lj
5
LU
— J
£
CjJ
tj\
O















VERAGE,X

























C/5
3

























1 '
MEASUREMENT
RESULT
w
CODE
o
-5
-n
Kl )k- Ul j K» In- jj
|CD G
!o\ cn
jo: (cn
u\ m
P^
P P
CD O
$P
W^ J^5
Ijl /^^


^^


P- 3





































































M_
5"
INI
10
1

JC
if



















IH1


















*n
a
^
_u
1















•z-s
§2
nn
£
cr.
_ PERFO
Z
•n
n
III
HI » *^
I? *• **
1
Di
a
i
i
c
z
M
3
*»
K
a1
;( H
^
3
rj
05
C
a
2
s
z
t
7
                                        (Rev.  7/14/81)
-------
                                                       Page 5
      7.  Repeat the above for each flow rate calibration  audit
 plotting all points on the chart and connecting the points by
 a straight line.
      8.  Tentative warning and control lines are given as + 4.7
 percent (warning lines) and +_ 7 percent ("out of control" lines).
 After 15 to 20 points are plotted, new control and warning lines
 may be derived as described in Appendix H of Volume I of this
          4
 Handbook.
      9.  Out of control points are an indication of calibration
 errors, instrument damage, etc.  Recalibrate the sampler prior
 to further sampling when out of control.
     10.  Forward the X and R chart to the supervisor for review.
 8.1.3  Exposed Filter Weighing Audits - In order to allow for
 possible data corrections, it is necessary to weigh exposed
 filters immediately after a 24-hour conditioning period.  Thus,
 it may be  impossible to have lot sizes greater than 10 or 20.
      1.  Randomly select and reweigh 4 out of every lot size
 of 50 or less (this would mean 100 percent checking if 4 or
 less exposed filters are weighed at one time).   For lot sizes
 of 50 or greater,  reweigh 7 from each lot.
i
      2.  Reweigh all filters in a lot if any audit weight differs
 by more than +5.0  mg from the original weight.
      3. Accept the lot with no change if  all audi-ts are within
 +_ 5.0 mg of the originals.
      4. Record the original and audit weights  on an X and R
 chart as illustrated in Figure 8.2.   Follow the procedure outlined
 in steps 6  through  1C in Section 8.1.2.  The accuracy A, in this
 case would  be defined as
            A = Original Weight - Audit Weight (mg).    Equation 8-2
 Tentative warning and control limits  of +  3.3 mg a"nd + 5.0 mg,
 respectively,  are recommended until  sufficient  data  are obtained

                           (Rev.  7/14/81)
-------
                                              Page 5
Table  7.1  ACTIVITY  MATRIX FOR  MAINTENANCE
XqoipaMt
Sopler Motor
faceplate gasket
kotaaeter
•star Baskets
Sampling head
Acceptance limits
-400 hours operation of
setor brushes
-Absence of malfunction
Absence- of leeks *c filter
•••1 -
-Ab'cnea of foreign at*teri*la
-Subl* op«r»cion»
LaaJe t±7ht fie
Xb*«nea of leaks
frequency andl
method of
a«««ur«m«nt
Vi«u«lly check upon
r«c«ipt *nd after each
400 hours of operation
Visually check after
•ach sampling period
Visually check for each
sample
Visually check each 400
hours of operation
Visually check each 200
hours of operation
Action i.:
requirements
not met
-Replace motor
brushes
-Other staintenJ
as indicated
Replace Basket
Clean: replace
if damaged
Replace gaskets
Replace 5aipljj
head
                 Rev. 7/14/81)
-------
                                                X anu K CIIAiti
        PROJECT I
         NAMB  I
         DATE 74.
         6
Cuidil
       SUM

      AVERAGE
t/l
               JII17
   J7I0.7
                       SOW
l/Zi
           3NI.O
    T11I.5
    Jlbtt
sin*
WJil
                                              Lot
                               4/L
                                                                                       MEASUUEMKNT
                                                                                   	   UNITS
                              4/0
Jftftf
3143
JJHM
Iht.7
                                                                 3:.;
ItTS.'l

1ZT4?
                   5/I.3 5/ZO
                                                                                           3TKI
                                                                                    37H?
                                                                                                    - _\-	—   oo
                                                                                                                 (U
              Figure  8.2   Quality control chart  for  audit of weights of  exposed filters
-------
                                                     Page 7
to  support  an  alteration  of  these limits.   Do not increase the
•limits  unless  so  authorized  by  the Administrator.
      5.  Forward  the  X  and R chart to the  supervisor for review.
      6.  Out of control points  indicate  the need for recalibra-
tion  of  the balance and/or improved operator technique.   Reweigh
all of  the  remaining  exposed filters in  the lot.
8.1.4   Data Processing  Check -  In auditing data processing
procedures, it is convenient and  allows  for correction to be
made  immediately  if audits are  made soon after the original
calculations have been  performed.   In particular,  this allows
for possible retrieval  of additional explanatory data from
field personnel when  necessary.
      1.  Use the  same audit  rate  as step 1 of 8.1.3.
      2.  The check is made starting with the raw data on the
data  sheet  or flow rate recorder  chart and continuing through
the recording of  the  concentration in yg/m  on the SAROAD form.
      3.  If the mass  concentration of suspended particulates
commuted by the audit check,  (yg  TSP/m )  ,  differs from the
                          3
original value, (yg TSP/m )   by as much  as  + 3 percent,  all
samples in  that lot are checked and corrected as  necessary.
The audit value is always given as the correct value under
the assumption that a discrepancy between  the two values is
always double checked by  the auditor.
      4.  Audit values are recorded in the  data log and reported,
along with  the original values, to the supervisor for review.
8.2  Auditing with a  Collocated Sampler
     An alternate method of  auditing the high volume method,
which in certain  situations  might be feasible, is to use two
collocated  samplers.  A network operating  several samplers
in a  reasonably small area  (e.g.,  city or  county)  might find
this method more  convenient  than  auditing  individual variables.
The field sampler and the audit sampler  will be denoted
respectively, as  the  first and  second sampler-throughout
the section.
                          (Rev. 7/14/81)
-------
                                                       Page 8

     The  second  sampler  should  be  operated in strict accordance
with the  procedures  given  in  this  section of the Handbook.   A
record  should be maintained of  the checks performed on the  second
sampler and  reported with  the data as  requested by the manager.
An audit  would be  to place the  second  sampler adjacent to
(but no closer than  6 feet) the field  sampler (see Reference 1
for discussion on positioning the  sampler)  and  sample  simul-
taneously.
     The  percent difference in the  concentration of suspended
particulates as measured by the field  sampler and the  second
sampler is computed  by
                      (ygTSP/m3)   -  (ygTSP/m3)_
Percent    = %D = 	.	—±	=-2-T x 100.   Equation 8-3
 difference       0. 5 I (ygTSP/m3) 1  +  (ygTSP/m^J
Based on  the results of a  collaborative test  a  defect would
be defined as
              |D | >_  15.*
     The  auditing level for collocated samplers  would  be  the
same as that given in the  previous  section,  i.e.,  n =  7,  N = 100
as the  initial rate.
     Values of (yg TSP/m'),,  (yg TSP/m )2/  and  D,  and  the
auditing  level would be reported on the quality  control
X and R chart along  with standard  identification information.
     The  data may be analyzed and  reported  as described in
steps 6 through  10 of Section 8.1.2,' above, with the exception
that the  difference, D,  is obtained by equation  8-3.   Tentative
warning and control  limits are given-as 10.0  and 15.0  percent
respectively.  Out of control values may  imply  errors  in  any
one of the several sample  collection and  analysis  steps.
Audits of specific steps will aid  in determining the likely
*If a = 3.5 percent of the concentration of TSP  for  each  sampler,
 then the percent difference, D, would have a  standard deviation
 of 5.0 percent of the mean value.  This gives a  3o  value  of  15
 percent.
                                (Rev. 7/14/81)
-------
                                                     Page 9
cause of the large deviations,  e.g.   calibration,  weighing,
and/or data processing errors.   See Figure  8.3  for an  example
quality control chart.
8.3  Systems Audit
     A systems audit is an on-site inspection and  review of   ^
the quality assurance system used for the total measurement
system (sample collection, sample analysis, data processing,
etc.).  System audits are normally a qualitative appraisal
of system quality.
     A systems audit should be  conducted at the beginning of
a new monitoring system and as  appropriate thereafter  to audit
significant changes in system operation.
     A preliminary form for use  in a system audit  is given
in Table 8.1.  These questions  should be checked for the
applicability to the particular  local, state, or Federal agency,
                        (Rev. 7/14/81)
-------
    PROJECT
     NAME
     ™J£!fc

       u
   SUM
   AVERAGE,X
7»D RANGE, n
Lofc
  *17
20 —




10 —
     a
    ' 13
        -10-
        -ZO-
      13
    U}J <
    §K§~
    !§«"
    5S^«
     !/e
       97
          ii^l[
            -3
                        lot
                        IMS..

                        J3&
              -6
2_r
                   JkL

                   I5o
   ZJ12
                         12J.
                              Lot
                            JL5L
           8
                                                                  MEASUREMENT      .,
                                                                   UNITS     Ibcizntdiffacnde
              J52,
                                          31
154
                 JHS
                                     3J
   _ys.
                                       u
                         u
           31
                                ii  it   if-^-r
n  »i  1,1
iT—li
                                                                                     — U
                                                                                     — f;r
                                                                                         L=3
                                                                                   1(5$
                                                                                              CO


                                                                                              -a-
             Figure  8.3  Quality control  chart for audit  with collocated samplers
-------
                                                      Page  11



   Table  8.1   CHECK LIST  FOR USE  BY AUDITOR

                   FOR  HI  VOL METHOD


 1.   What type of hi vol samplers are utilized in the network?


 2.   Hov often are the samplers  run?  (a)  daily,  (b)  once every
     six days, (c)  once  every 12 days,  (d)  other.

 3.   What type of filter and number is  being utilized? 	
 4.   Are there any pre-exposure checks for pin holes or imper-
     fections run on the filters? 	
 5.   What is the collection efficiency for your filters?
 6.   What is the calibration procedure for the hi vol sampler?
 7.  Which statement most closely estimates the frequency of
     flow rate calibration?  (a)  once when purchased,  (b) once
     when purchased, then after every sampler modification, or
     (c)  once when purchased, then at regular intervals there-
     after.

 8.  Are flow rates measured before and after sampling period?


 9.  If the answer to t8 is yea,  using the equation below, what
     is the estimated average percent of change in the flow
     rates?  Use the equation
                          100(Q.  - Q.)
                          	-r	 - percent change

     (a)  less than 10%,  (b) 10-20%, or (c)  greater than 20%.

10.  Is there a log book at each sampler to record flows and
     times? 	

11.  Are filters conditioned' before initial and final  weighings?
     	.  If so, how long?          .   And at what percent
     humidity? 	.

12.  Is the balance cheeked on a periodic basis? 	.   If so,
     how often? .With which standard weights? 	
13.   How often are the hi-vol filters weighed?
     How is the data from these weighings handled?
14.   Are all weighings and serials numbers of filters kept in a
     log book at the laboratory?  	

15.   What is the approximate time delay between sample collection
     and final weighing?  '  '  days
                          (Rev.  7/14/81)
-------
                                                         Page 12
Table  8.2   ACTIVITY MATRIX FOR AUDITING PROCEDURE
Audit
Weighing audits
Clean filters
Exposed filters
Calibration audit

Data processing
audit

Collocated sampler
audit
(ystvm*
audit
Acceptance limit*
Audit vt. - original wt. * 2.8 wj
Audit vt. " original vt. * 5.0 ag
flow rate
°'" ^ rlov rat.a i X-07
Flow rat«a • routine »««»ur«d r*t%
flow rrnte^ - Audited rate
iuc TSP/m^

(ug TSP/m-i)a
(ug TSP/»^) • routine neasured con-
centration
(119 TSP/n*) • audited conc«ntration
100[(u9 TSf/a^)i - (»q TSP/«J)j] -
Av*ra?« lug TSP/» ) ^ and 2
i.«., th« difference in concentration
do«> not exceed 15% of the average value
obtained by the field saapler 1 and the
collocated sampler 2.
These liaits are for repeatability (cane
operator) . Larger limits would be ex-
pected for reproducibility (between
two operators ) .
Method •• described 'In
this *«ctioa of ui«
Handbook.
Frequency and
awthod of
measurement
Freguancv - 7 audits
per 100 filters. For
a lot size of SO or
less filters aaJce 4
audits .
Method - Ose analytical
balance . Condition
filters for 24 hours
before weighing
rreourncy - See above.
Method" — Sane as for
calibration procedure
for one flow rate.
Frecruenc-y - See above.
Method - Re4o all cal-
culations > readings of
charts, etc. through
recording results.
Preouenc^ - S*« above.
Bonitor calibrated and
naintained in strict
accordance to method
description.
yrgguTncv - At the
b*vliULiaf of »_ new
•onitorinf cystesi
and periodically as
appropriate
I*-"""! "S-~.i£*en of
r-^*--*vr>Dts
not "Mt ;
Reveigtt all filters
in the batch
Perform calibra-
tion before
sampling is
resuned.
Recheck all calcnt-
lationa.

Data invalid for

Initiate iav10*"*1
•ethods aad/or train-la,
pctxjruu
                          (Rev.  7/14/81)
-------
             Dichotomous  Filter  Handling  and  Data  Procedures


I.   General Discussion

     Dichotomous  filters  best  suited  for  the  CMB analysis  of  ambient
     particulate  material consist  of  a  polyolefin  supporting  ring bonded
     to the top  (particle collection) surface of a teflon  membrane.   The
     filters are  37 -mm in diameter.

     Sample ID codes  should  consist of  a  letter prefix  indicating the sample
     site and whether the filter is the fine  or coarse  fraction.   Sequential
     three digit  numbers  should  then  be used  for the  sample number,  e.g.,
     BFSC 012 and BFSF 012 might represent  the twelfth  sample (both  coarse
     and fine) collected  in  Bakersfield (B) at the fire station  (FS)  site.
     The most important point  to emphasize  is that both the coarse and fine
     fraction filters should have  the same  numerical  code  to  facilitate data
     interpretation and the  letter portion  should  reflect  the sample site
     and whether  it is the coarse  or  fine fraction (C or F).

     Maintain a sample log book.   Record  sample sites,  filter numbers,  dates,
     comments, volumes* weights  and filter  lot numbers.

     All filter handling  procedures should  be done in a clean area.   Clean
     the work area with a methanol dampened Kaydry towel,  then lay out a
     clean Kaydry towel to use as  a work  surface before all handling procedures.
     The filters  must be  handled only by  the  supporting ring  with clean forceps.
     A laminar flow clean hood is  the most  appropriate  work area.  Disposable
     elastic gloves should be  worn when working with  filters.

      Petri dishes,  slides and cassettes can be reused but  must be
      wiped clean with methanol before reuse.


II.    Materials  and  Equipment

              Sierra Model FH-240-P filter cassettes
           •   Millipore plastic Petri  dishes  (PD 10 047 00)
           •   Millipore plastic Petri  slides  (PD 15 047 00)
           •   Ghia Corp.  2y PTFE 37 mm filters with polyolefin rings
              (R2P503700)

           •   field  sample transport boxes

              shipping boxes

              forceps
           •   Kaydry towels  (VWR Cat.  //  21903-008)
              methanol (polyethylene wash  bottle)
              waterproof  fine point marker  (e.g.  Pilot  Corp.  extra fine
              point  permanent markers)
-------
               adhesive labels

               disposable plastic gloves  (e.g., Tru-Touch  Vinyl Medical
               Gloves, VWR cat tf 32901-060)

               Tip-n-Tell shipping indicators
III.   Laboratory and Field Procedures

       1.      Lay out a series of 6 - 10 Petri dishes on the work  surface,
               label with the sample ID numbers, recalling that there
               should be coarse and fine pairs.

       2.      Select a new filter and inspect both sides for foreign
               particles, pin holes, discolorations, or other
               imperfections.   Some particles may be carefully blown off.
               Discard the filter if defective.

       3.      Steady the filter by pressing down on the plastic ring of
               the filter with forceps.   Write the ID number as recorded
               on the Petri dish on the plastic ring of the filter  using
               a fine point waterproof marker.

       4.      Weigh the numbered filter and record in log book.  Also
               record the filter lot number.

       5.      Place the bottom half (female part)  of a clean Sierra
               filter cassette on the work surface and place the weighed
               and numbered filter into it.   The filter should be placed
               top (numbered)  side up.   Place the top half (male part)
               of the clean Sierra filter cassette in position over the
               bottom half and press down so that it snaps into proper
               alignment.

       6.      Label the edge  of the cassette with an adhesive label.
               The number should correspond  to the  number on the Petri
               dish and polyolefin ring.

       7.       Place the loaded cassettes into the  corresponding Petri
               dishes.

       8.       Place the Petri dishes  into the field transport box  (in
               coarse-fine sets).

       9.   '    Once  at  the  sampling  site, place the  cassettes  into the
               dichotomous  samplers  (bevelled side  down)  again making  sure
               that  each coarse-fine  set  is  placed  into  one  dichotomous
               sampler.   Retain  the  numbered Petri  dishes  in the field
               transport  box.
-------
      10.   Record the. sample site, sampler serial number, filter  ID
            number, start and stop flow rate, start and stop time  and
            volume, as well as the initials of the operator and
            comments onto data sheets  (sample attached).  Note:  Total  (T)
            flow rate and volume and coarse (c) flow rate and volume are
            measured and recorded.

      11.   After a 24 hour sample has been collected, the loaded  filter
            cassette should be placed into its corresponding numbered
            Petri dish and.returned to the laboratory.

      12.   Label clean Petri dish slides  with the sample ID numbers.

      13.   Select a Petri dish containing an exposed filter from  the
            sample transport box and place on the work surface.  Obtain
            the Petri slide with the matching ID and place on the  work
            surface.

      14.   Remove the Petri dish lid and, being careful not to touch the
            filter, remove the filter cassette assembly from the Petri
            dish.  Remove the side label and carefully separate the
            cassette assembly.  Then using a pair of forceps, transfer the
            filter to its Petri slide.

      15.   Record any unusual appearance or condition of the filter in the
            sample log book.

      16.   Weigh the loaded filter and record the weight in the sample log
            book.  Replace the filter into its numbered Petri slide.  Pinch
            the edge of the filter in the Petri slide when closing the slide
            to hold the filter in- place.  Note:  Less than 100 Petri slides
            per tray can be held when they are loaded due to the increased
            thickness caused by pinching the filter.

      17.   Place the Petri slides into the plastic trays.  Put the
            cardboard boxes containing the trays into a cabinet or other
            protected location until enough are collected for shipment.
IV .   Shipment
      1.      Place the cardboard boxes containing the plastic trays of
             loaded Petri slides into the flanged shipping box in such a
             manner so that all filters are held upright during shipment.
             Include filters from each filter lot number for blanks.

      2.      Photocopy the data sheets and appropriate pages from the
             sample log book for enclosure into the shipping box.  Include
             the uncorrected mass of particulate material collected on the
             filters in micrograms.

      3.      Place a "Tip-n-Tell" indicator on the shipping box with the
             accompanying label.  Activate the "Tip-n-Tell" indicator.

      4.      Ship directly to NEA Laboratories, Inc., 8310 S.W. Nimbus Avenue,
             Beaverton, Oregon, 97005.
-------
LOW-VULUML inCIIOTUMOUS KllilJ)  DATA Slltbl1
OX LC

Serial^

ID//
F
C

f
^

F
C


Sampled

Start
T
c


Stop


Start

•




T
c.



F
C




T
C-








Stop



Net











F
t
f
r
f
y

F
C
•
/

F
C


'•
T
c









T
C






-i- — , — . ., i _ - - .



i-IIJ-l. J.U1.D






T
C

T
c


T
c







T
c

T
c






















•




T
C
i







I




-------
                         Soil and Road Dust Sample Sieving

11.1   General Discussion
      The bulk soil and road dust samples collected according  to  SOP  #12  and  #13
      must be segregated before analysis.  This procedure describes the  isolation
      of two size fractions, <75 ym and  <38 urn.
      Special care must be  taken to prevent the spread  of dust to any other sample
      handling or analysis  area of the lab, and to prevent  cross-contamination of •
      the soil samples.  Clean' the work  area  and  spatulas with methanol  dampened
      Kaydry towels between each sample  handling.
11.2  Materials and Equipment
      1.  Set of Tyler sieves, as indicated               USA  No.      Opening
                                                               	um
          plus top cover and bottom pan.                  	
                                                              18         1000
                                                             40          425
                                                             80          180
                                                             200           75
                                                             400           38
      2.  Tyler sieve shaker,  Model RX-24
      3.  Polyvials, 26 ml.
      4.  Kaydry disposable towels.
      5.  Compressed air.
      6.  Spatulas, large  and  small sizes.
      7.  Methanol.
      8.  Soft bristle brush.
-------
       Standard Operating Procedure
                                           Page  2
11.3  Flow Diagram
                                Remove bulk
                                sample from storage
                                Transfer to 18, 40, 80
                                and 200 mesh  sieve
                                assembly
                                Shake one hour
                                                              Clean 18, 40, 80,
                                                              and 200 mesh sieves
      Place an aliquot
      of the <75 urn
      fraction in labeled
      vial
Place rest of the <75
ym fraction in 400 mesh
sieve assembly

Return the >75 ym
fraction to sample
storage bag
                                Shake one hour
                              Clean'400 mesh sieve
      Place the <38 ym
      fraction in labeled
      vial(s)
                              Return the <38 ym
                              fraction to sample
                              storage bag
                                                                  I
      Place      sample
      vials in storage
                              Return remaining bulk)
                              sample to storage
-------
        Standard Operating Procedure                                      Page 3

11.4  Sieve Cleaning
      1.  This procedure should be done  in  a  room  dedicated  for  sieve  cleaning
          or outside to prevent the spread  of  dust to  sample handling  areas of
          the lab.
      2.  Use compressed air to blow away all  loose dust.  Pay particular attention
          to the screen, the screen-to-frame joint, and  overlapping  surfaces on the
          frame.
                        t
      3.  Carefully brush all loose dust from  the  wire mesh  screen with a soft
          bristled brush.
      4.  Wipe all loose dirt from the frame with  a. brush or Kaydry  towel.   Clean
          both inside and outside and below as well as above the screen.
      5.  Repeat steps 2-4 until the sieve is  clean.
11.5  	Sieving
      1.  Clean the bottom pan using a methanol dampened Kaydry towel  and assemble
          the cleaned Tyler sieves in the following order, from the  bottom to  top:
          bottom pan, 200 mesh screen,  80 mesh screen, 40 mesh screen,  18 tnesh
          screen.
      2.  Select a      sample from the sample storage cabinet for processing.
      3.  Remove the inner bag and sample data card from the  outer bag.   Set  them
          aside for later use.
      4.  Transfer the contents of the sample  bag  to the sieve assembly;  minimize
          the loss  of fine dust by dumping the sample into the sieve slowly and
          carefully.   Set  the sample bag aside for later use.
      5.  Clean the top cover using a methanol dampened Kaydry towel and  compressed
          air,  and  put the top cover into position on the sieve assembly.
      6.  Place the sieve  assembly into position on the shaker.   Press  Che  locking
          tab down  firmly onto the sieve assembly with one hand and  tighten the
          locking  tab screw with the other hand.   Repeat for  the other  locking tab.
      7.  Start Che shaker.   If dust escapes from the sieve  assembly, either the
          assembly  is not  mounted tightly enough, or there is foreign material be-
          tween the mating surfaces of the sieves.   Correct  and restart if  neces-
          sary.   Using the built-in timer,  run the shaker for one hour.
      8.  While the sample is shaking,  clean the sieve(s) not in use according to
          the procedure in section 11.4.
-------
   Standard Operating Procedure                                         Page *

 9.   Remove the sieve assembly from the shaker and  set  on  a  clean Kaydry
     towel.  Remove the sieves and top cover as a unit  from  the  bottom pan
     and set on a clean Kaydry towel.
10.   Using a clean spatula, transfer dust from the  bottom  pan  to a new 26  ml
     Polyvial.  Fill the Polyvial about half full.   Close  the  lid and  cut  off
     the hinge.  Recore "I.D.	^ (from the sample data card)  and "size
     <75 urn" on an adhesive label.  Stick the label on  the Polyvial and secure
     in place with a piece of transparent tape.
11.   Place the clean 400 mesh sieve on a clean Kaydry towel.   Carefully pour
     the sample remaining in the bottom pan into the 400 mesh  sieve;  tapping
     lightly on the pan will remove most of the material.
12.   Transfer the material remaining in the 18, 40,  80  and 200 mesh sieves
     carefully into the bottom pan; tapping lightly on  the sieve frame will
     remove most of the material.
13.   Transfer the material from the bottom pan to the sample bag.
14.   Use a methanol dampened Kaydry towel to clean  the  bottom  pan.
15.   Place the 400 mesh sieve on the bottom pan and the top  cover over the
     400 mesh sieve.            ^
16.   Repeat steps 6 thru 9.
17.   Using a clean spatula, transfer dust from the  bottom  pan  to a new 26  ml
     Polyvial.  Use more than one Polyvial if necessary.   Close  the lid and
     cut off the hinge.  Record "ID f?	" (from the  sample data  card)
     and-"Size <38 urn" on an adhesive label.   Stick the label  on the Polyvial
     and secure with a piece of transparent tape.   Record  "ID  if	" and
     "Vial 1 of 2" or "Vial 3 of 3",  etc.  on the lids of the Polyvials from
     both-size fractions.
18.   Transfer material from the 400 mesh -sieve to the bottom pan,  and  from
     the bottom pan to the sample bag.  Close the bag and  seal with tape.
19.   Record the date of sieving and initials on the  sample data  card.   Place
     the inner sample bag and data card into the outer  bag.
20.   Return bulk      sample and Polyvials to sample storage cabinet.
-------
            TEFLON FILTER WEIGHING USING THE CAHN 27 ELECTROBALANCE

                         Standard Operating Procedure
1.   General Discussion

    1.    Analytical Problem:

         The mass of air particles collected on teflon membrane filters must
         be determined to an accuracy of better than ± 10% relative.  Two
         sizes of filters are used, 37 and 47 mm diameter, with tare weights
         of about 75 ± 20 mg and 120 ± 20 mg, respectively.  Minimum net
         weight is expected to be about 120 yg.

    2.    Interferences:

         Electrostatic charge on the filter is one possible source of inter-
         ference.  This  is eliminated by use of a beta source charge neutralizer.
         Humidity changes may affect the weight of deposit, but not the weight
         of filters, since they are non-hygroscopic.  Humidity changes are
         minimized by performing the weighing procedures in an air conditioned
         laboratory.

    3.    Minimum Detectable Quantity:  15 yg (three sigma).

    4.    Precision and Accuracy:

         The balance is  calibrated with class M weights.  Repeatability
         for the immediate reweighing of a filter is ± 1 yg; long term repeat-
         ability ±£ ± 5  yg.

    5.    The technician  should read the Cahn 27 Instruction Manual entirely
         before using the balance, paying close attention to Section 3, Balance
         Operation.  The balance is always left ON to provide maximum stability.

    6.    Keep  the  calibration weights clean and store in  their proper
         container.  Handle them only with the cleaned weight forceps  (non-
         serrated  stainless steel) and do not use the weight forceps for  any
         other job.
                                (Rev. 9/25/81)
-------
2.   Equipment  and  Materials

           • Cahn  27 Electrobalance with custom 60 mm open stirrups
           • Po210 charge neutralizer

           • Calibration weights,  10, 20,  50 and 100 mg Class M

           • Tare  weights,  20 50 and 100 mg Class C

           • weight  forceps

           • filter  forceps
           • Kaydry  towels

           • Kimwipes

           • Methanol


3.   Operating Procedure

3.1  Start Up;

     1.   Remove  the stirrups  from the  weighing chamber as in Section 3.1.2
          of the  Cahn 27 Instruction Manual and set aside on a clean Kimwipe.
          Clean the weighing chamber with  a methanol dampened Kimwipe.
          Replace the stirrups, one  on  the A side  and one on the TARE side.
          Place the charge neutralizer  in  the center of the weighing chamber.
          Clean the area around the  balance and lay out a clean Kaydry towel
          in front of the weighing chamber.   Clean the weight forceps (non-
          serrated stainless steel)  and filter forceps using a methanol
          dampened Kimwipe.

     2.    Obtain  the calibration weights and the filters to be weighed
           from  the  sample storage  cabinet.

     3.    Record  technician, date  and filter lot number on the data sheet.

     4.   Determine approximate weight of  one of the filters by weighing
           on the A200 range.  Select the 10, 20, 50 and 100 mg calibration
          weights in combination such that  their combined weight equals
          within + 8 and - 12 mg of the  filter weight.  (When doing gross
          weights or reweights, use the  same weights as for the tare
          weighing).

     5.   Set the balance controls to RANGE:  A 20  and RESPONSE:   0.   Make
          sure CALIBRATE, COARSE and FINE  ZERO controls are locked.   Press
          TARE twice to untare  the balance.

     6.   Place  the  selected calibration weights on the A pan and balance
          with appropriate  tare weights  on  the TARE pan so  that  the  display
          reads  0  • +  20  mg.
                             (Rev.  9/25/81)
-------
     7.   When  the  display stabilizes (after at least 30  seconds),  press
          TARE  to zero  the balance.

     8.   Add the 20  mg calibration weight to the A pan  (or  remove  it if
          already there).   Display should read + 20.000 ± 0.001
          (or - 20.000  ± 0.001)  when it stabilizes.  If not,  adjust CALIBRATE
          accordingly.

     9.   Replace calibration weight and note that 0.000  ± 0.001  is displayed.
          If not, repeat steps 7 and 8.  Record tare weight  as  the  sum of the
          calibration weights ± display reading, and the  calibration reading
          on the data sheet.

    10.   Remove calibration  weights to their container.'


3.2  Weighing Filters:

     1.   Up to  25  filters per set can be weighed.  When weighing  the
          filters,  note  any defects (torn, particles falling off, etc.)
          on the data sheet.

     2.   Place  the first filter to be weighed on the charge neutralizer.
          After a few seconds, transfer the filter to the A platform  and
          place  the next filter to.be weighed on the charge neutralizer.

     3.   When  the  display stabilizes (after at least 30 seconds), record
          the filter ID  number and weight reading including + or - sign on
          the data  sheet.

     A.   Transfer  the filter  from the A platform to its container, transfer
          the filter on  the charge neutralizer to the A platform, and place
          the next  filter in sequence on the charge neutralizer.

     5.   Repeat steps 3 and 4 until weighing of the sec is complete.

     6.   Place the  calibration weight(s)  on the A pan.   Record tare  weight
          as the sum of  the calibration weights ± display reading.   Press
          TARE if the  reading  is not 0.000.

     7.   Add the 20 mg  calibration weight to the A pan  (or remove  it if
          already there).  Record calibration reading.

     8.   Place calibration and tare weights in their containers.


3.3  Replicates:

     1.   Select at  random three  filters to  be reweighed,  preferably  by
          another technician.
                            (Rev.  9/25/81)
-------
      2.    Weigh che  filters using the procedures in sections 3.1 and 3.2.
           Indicate that  the weighings are replicates on the data sheet.
           Filters  from up  to 8 sets can be reweighed at the same time.

      3.    Calculate  the  difference, W  ,  between the original and replicate
           weighings.
3.4   Shutdown;

      1.   Replace  the  filters  that were  weighed  in  the sample  storage
          cabinet.

      2.   Clean  the  balance  as  in  3.1.1.

      3,   Record the date of weighing  in the  appropriate  Sample  Log Book.

      4.   Calculate  net weights if appropriate.

      5.   Return the data sheets to the  laboratory  supervisor.


3.5   Calculations:

      W   -   (W  - WJ X 1000,
      n       g    t

      where,   W   =  net deposit in yg
                                                                 •
              W   =  gross weight reading in mg
               o
              W   =  tare weight reading in mg
     W,  =  (W  - W ) X 1000
      d       g    r
         or
     Wj  »  (W  - W ) X 1000,
      d       t    r

     where,   W   =  difference between original and replicate weights in ug

              W   =  replicate weight reading in mg.


3.6  Quality Control:

     1.    The  tare and calibration weight checks at the end of a weighing set
          must be within ± 0.004 mg- of their correct values or the entire
          set  must be reweighed.

     2.    Replicate weighing differences, W ,  must be _< ± 10 ug or <_ 2% of W  or
          the  entire set must be reweighed.


                           (Rev.  9/25/81)
-------
3.7  Quality Assurance:

     1.   Class M weights are used for calibration.  According to NBS
          circular 547, they are "...designed for use as reference standards,
          for work of the highest precision, and for investigations demanding
          a high degree of constancy over a period of time."

     2.   12% replicates ensure that weighing precision is within limits.

     3.   Calibration and tare operations are performed before each set
          of 25 filters and are checked for accuracy after each set.

     4.   Results of replicate weighings and calibration and tare weight
          checks are reviewed by the laboratory supervisor.  He must initial
          the data sheet before the data is considered valid.
                           (Rev. 9/25/81)
-------
                                                       Page 1
                   Standard Operating Procedure
               EPA High Volume TSP Sampler Calibration

  1-  Elapsed  Time Meter
     The elapsed  time meter (synchronous motor type  60  hertz)
should be checked on site or in the laboratory every six
months against a  timepiece of known accuracy.  If the indica-
tor shows any signs  of being temperature sensitive,  it  should
be checked on site during each season of the year.
     A gain or loss  of more than 2 minutes in a 24-hour period
warrants an adjustment or replacement of the indicator.
     Record results  of these checks in the calibration  log
book.
  2.  Timer
     For those samplers that are equipped with an on-off timer,
the timer should  be  calibrated and adjusted using a calibrated
elapsed time  meter as the reference.   An example of this type
calibration procedure is presented below.   Figure 2.2  depicts
the wiring diagram for use in this calibration.
     The timer calibration procedure  should be performed on a
quarterly basis.   Calibration data are recorded in the  Timer
Calibration Log.   See Figure 2.3 for  an example.  The steos
in the calibration procedure are:
     1.  Plug a correctly wired timer into an electrical outlet.
     2.  Set  the  timer to the correct time.
     3.  Set  the  ON  and .OFF  time trippers for a 24-hour test
period.
     4.  Plug the test light into one of the output plugs and
an elapsed time meter into the other.
     5.  Check the system by manually operating the switch ON
and OFF.

                         (Rev.  7/14/81)
-------
                                                               Page 2
       6.   Allow the system to operate for the 24-hour test
  period and determine the elapsed time from the elapsed time
  meter.
           a.  If the elapsed time is 24 hours  + 15
              minutes,  the timer is acceptable for
              field  use.
           b.  If the elapsed time is not 24  hours
              +  15 minutes,  adjust the  tripper
              switches  and repeat the test.
 Indicator lamp
ON-OFF Timer
(± 15 min/24 hours)
                                                     Elapsed  time meter
                                                      (± 2 min/24 hours)
       Figure  2.2   Diagram of  a  timer  calibration system

  3.   Orifice Calibration Unit
      The orifice calibration unit should be calibrated against a
 secondary standard, for example  a Rootsmeter, upon receipt and at
 one-year intervals  thereafter.   The manufacturer's average calibration
 curve  can be used unless the calibration deviates from it by more
 than  ± 4  percent at any one point along the curve.   When deviations
 from the  manufacturer's curve  are larger than ±  4 percent and
 there are  no visible  signs of damage to the orifice,  the cali-
bration should be repeated by another  operator.   if  the  large
deviations persist  (after  the secondary standard  has  been checked
                              (Rev. 7/14/81)
-------


114 1














Timor
No.
1321














Oa«-e.
Start
do/75














Stop
Jlllfe














Elapsed ti.-ce
Indlcdtor
serial No.
0 \J() 0 %.0 (f>














Test
period
44hr5.














Elapsed time
indicator reading
A3 hrs. 53 rv),n.














Accuracy
97. 5














By (signature)
£. ty&ru^
/













Figure 2.3 Example of a timer calibration log 21
ID
-------
                                                      Page 4
and found satisfactory) a new average calibration curve  is
constructed using the results from at least five sets of
calibration data.
     Orifice units should be visually inspected for visible
signs of damage to the orifice before each use.  A calibra-
tion check should be made if the orifice appears to have any
nicks or dents.
     The following stepwise orifice calibration procedure is
adapted from Reference 3.
     1.  Assemble the parts as shown in Figure 2.4.
     2.  Zero the water and mercury manometers by sliding
their scales until the zero on the scale is level with the
meniscus as illustrated in Figure 2.5.
     3.  Check the level of the positive displacement meter
table.  Adjust the legs if necessary.
     4.  Install Load Plate 18 between the orifice and the
positive displacement meter.
     5.  Turn Hi Vol motor ON, and let the system operate 5
minutes.  While the unit equilibrates, continue with steps
6-9 below.
     6.  Write Plate #18 under the Plate # Column in the Hi
Vol Orifice Calibration Log.  See Figure 2.6  for an example.
     7.  Record the date, time, orifice number, name of
primary standard  (positive displacement meter) and the
serial number of the primary standard in the  appropriate
spaces in the log.
     8.  Record the temperature in °C.
                         (Rev. 7/14/81)
-------
                                                               Page 5
  MEROJR'
  MANOMETER
                                                      THERMOMETER
                                                             BAROMETER
                                                     POSTIYE
                                                     DISPLACEMENT
                                                     METER
                    IAL I 5
                   UNCOMPENSATED
HI-VOL
MOTOR
                      * «f  nositive  displacement
Figure  2.4   Diagram of  positive


                              (Rev.  7/14/81)
     meter  system
-------
00
H
c 	
-3-
-2-
-1-
3
H
J
vc 	 _ — p/
MERCURY
MANOMETER
ZEROED


K^i
:::!:::
-2-
-1-
-1-
-2-
^J
^

^^***iMU>"*^^
WATER
MANOMETER
ZEROED

70mm



y 	
'ME
MAf^
RE
P =
- 4-
-3-
- 2-
-1 -
- 1-
_ o
t
-4-
±~A
RCUR\
OMETE
AD INC
70nn


=7
r
:R
k
>


w
MA
R
P,
-3-
-2-
	 i _
-1-
-2-
WATER
NOMET
EADIN
- 3.0

i /
ER
G
In.
j ,
Pt =

                                                                                           W
                                                                                           UQ
                                                                                           (T)
              Figure 2.5  How to read mercury and water manometers
-------
Description
Derivation
^X^ymbol
Plate^v.
Number ^"X^
&
~T
13
1^
18





Ft.3 of
Air
\
100
100
100
100
100
100
100
100
100
100
3
m of
Air
\
2.83
2.83.
2.83
2.83
2.83
2.83
2.83
2.83
2.83
2.83
Barometric
Pressure
mm Hg
Pa
7^7
1(0-1
767
7k7
767





Vacuum
in Standard
in nun Hg
Pm
7o
TO
TO
70
7£>





Absolute
volui.w in m
v (Pa"Pra)
•V- Pa
V
2.57
Z.S7
2.B7
2L57
•2.S-7






Time in
Minutes
t
7.73
1.64
(.43
i.zq
I.2Z





Flow rate
ra-Vniin
V
a
t
'I
o.Q^
1.4-
I.B
z,o
2.1





Pressure Drop
Across the
Orifice in
inches of water
Pt
2.5
6.S
II. O
IS.5
18.3





00
Orifice  Number 	
Temperature  ZiS.Q
Date  (C\-~A
                                                              Manufacturer
                                         C  Barometric Pressure
                                        	  Primary Standard
                      Calibration voltage
                                                              Verified By
 Time JQ.-45 d.m
Serial Number  (Q42.3
                                                                                                                OQ
                                                                                                                (B
                Figure 2.6   Example  of hi  vol orifice calibration log
-------
                                                          Page 8
Definitions "for use with Figure 2. 6
V  = Actual volume  of  air  measured in cubic meters.
 a
V  = Volume measured by  the  positive  displacement meter in
 c   cubic feet
V  = Volume measured by  the  positive  displacement meter in
 m   cubic meters as calculated  from  Vc.
P  = Atmospheric pressure  in mm  Hg.
 3.
P  = Vacuum at the  inlet of  the  positive  displacement meter
 m   in mm Hg.
 t -= Minutes  of time elapsed during run.
Q1  = Flow rate in cubic  meters per minute'at prevailing
     atmospheric pressure  and temperature (uncorrected).
P . = Pressure drop  across  the orifice in  inches of water.
Equations given  in Figure  2.6
V • = V  x  0.0283
 m    c
         
-------
                                                          Page 9
     9.  Record the barometric, pressure  in mm Hg.
    10.  After the 5-minute equilibration  period,  read the
mercury manometer and record  this  value  in column  P .   The
example given in Figure  2.5 shows  a  reading of 70  millimeters
(mm) .
    11.  Read the water  manometer  and  record this  in column
P .  The example given in Figure 2.5 shows a reading of 3.0
inches.
    12.  Wind the stopwatch and set  it in   the horizontal
position with the dial facing  up.
    13.  Locate the uncompensated  dial on  the left end of  the
positive displacement meter.   The  location is shown in Figure
2.4.  Note:  this dial must be viewed  from the end.   One
revolution of Dial #5 equals  10 cubic  feet of air  passed through
the positive displacement meter.
    14.  Use a stopwatch to measure  the  time in minutes and
hundredths of minutes for exactly  10 revolutions of Dial #5
(i.e.,  for 100 ft  of air to pass  through  the positive dis-
placement meter).  Record 100  under  column V  and  the  elapsed
time under column t.
    15.  Record 2.83 under column  V  to  convert cubic  feet to
                               - - —  m
cubic meters.
                                                              •
     V  = V  x o.0283
      me           ^
           •3          m   _ _  _ _   3
     100 ft3 x 0.0283 ~3- 2'83 m

    16.  Turn the motor OFF.
    17.  Repeat this procedure with each of  the other  load
plates  in the set.
    18.  Repeat Steps 1-17 one time.
    19.  Calculate and record  V  for each  run.
                               a.
                  (3?* ~ PJ
         V  = v     a	m_
          am      p
                      a
    20.  Calculate  and record  Q, for each  run.
-------