Appen dix B: Gra vimetric An alysis Proc edure
Airmetrics
77
petri dish label. The blanks serve several functions:
a.
As a qu ality contro l on the gra vimetric p rocedure , each bla nk is
reweighed at least one day after the initial weighing. At least 50% of
reweighs should be performed by someone other than the original
technician. The weight variations in a statistically significant number
of such audits are then used to establish control limits for the variation
at three stan dard dev iations (see Quality A ssurance Handb ook for A ir
Pollution M easurem ent System s, Vol. 1, A ppendix H). Co ntrol limits
should be recalculated every six months from the most recent blank
values. If a blank shows variation between weighings beyond the
control limits, the weighing process must be rechecked and, if the
discrepancy is not found to be associated with the second weighing,
all the filters tared with the initial blank weighing must be rechecked.
b.
Blanks are selected with samples of the same lot as analysis blanks for
ion chroma tography, X-ra y spectrome try, or other tests.
c.
Blanks ma y occasionally b e needed to rep lace damag ed control filters
for temperature/humidity adjustments.
d.
Unuse d blanks or portions thereof are stored in n umeric al order w ith
their samples for at least two years before discarding.
2.
Since it is known that humidity and temperature introduce most of the
variability in filter weights, close control must be maintained on these
factors. A ny chan ge of 1% in hum idity or of on e degree celsius in
temperature during a weighing session necessitates the reweighing of the
controls and recalculation of the adjusted tare.
3.
The three control filters furnish another quality check on the weighing
precision and on the accuracy of the tare adjustment. The variations in
)
C
between the controls are compared to control limits established from the
past performa nce of the filters--such limits are set at three standard
deviations as in 1a, a bove. W hen excess v ariation indicates an ou t-of-
control co ndition, w eighing m ust be sus pended until the bala nce is
checked and recalibrated and the control filter reweighed.
4.
Zero the balance after every twenty filte r weights . Reset the zero if
needed. (
Caution:
Large position zero displacements may be caused by
Airmetrics MiniVol Users Guide
78
Airmetrics
dirt on the p an. Rem ove the p an, clean it, and then recheck the zero in
such cas es. If the zero drift at any ch eck is ove r one m icrogram , reset it
and then recheck the calibration before continuing. If the zero drift is over
five micrograms, correct the balance and reweigh filters back to the point
where the differen ce is two micro grams or be low.)
5.
Special sh ort-term sa mpling projects us ually requ ire more rigorous Q A to
generate better P&A statistics over a shorter time period. In these studies
it is common to perform filter reweigh checks on both clean and exposed
filters at the 10 -15% level.
