standardizing, and distributing the standard samples was assigned to Western
Oregon State College (WOSC).
In this way the primary responsibility for
evaluating the analytical integrity of the data was vested in a disinterested
party.
In some cases, small variances from these criteria were allowed to facilitate
processing the data, but the deviations were never great enough to compromise the
integrity of the data. Specifically, the error requirements shown in Table | were relaxed
by 20% for some of the early DCDs to accomodate problems some of the contractors had
in reducing counting errors.
For example, for samples where a 10% relative standard
deviation was required, a 12% relative standard deviation was allowed.
Another variance was in the number of duplicates and standards that had to be in
compliance. In general, for a DCD to be considered acceptable, 80% of the duplicate
pairs and 100% of the standards had to be in compliance with the QC criteria. When the
number of duplicate pairs did not permit exactly 80% compliance (for example, 3 of 4
pairs would give 75% compliance), a fraction of duplicates in compliance slightly less than
but near 80% was still considered to be acceptable.
Less than 100% compliance on
standards was allowed occasionally if the radiochemical analysis on the standard was near
the accepted activity and if the laboratory had established
a record of accurate
radiochemical analysis on other subsamples of the standard in question.
PARTICIPATING LABORATORIES
Laboratories participating in the radiochemical analyses of samples from the NMIRS
were
Environmental
Analysis
Laboratory
(EAL),
Richmond,
California;
Eberline
Instrument Corporation (EIC), Albuquerque, New Mexico; and Laboratory of Radiation
Ecology (LRE), University of Washington, Seattle, Washington.
The radionuclides measured were 905, (beta counting), 13765 (beta and gamma
239
d 240
), 23942405, and 238
Pu (mass
counting
Pu (alpha pulse-height analysis),
Pu an
spectrometry), atl Dy (mass spectrometry and liquid scintillation counting), and
(alpha pulse-height analysis).
241
There were 16,282 analyses, including the duplicate and
standard samples of the QC program, requested of the three participating laboratories.
The largest fraction of the analyses was performed by EAL: 65.6% (42.3% terrestrial and
23.3% marine). Slightly over one-fourth of the analyses, 25.5%, was performed by EIC.
The balance, 8.9%, was analyzed by LRE.
Table 2 summarizes the duplicate and standard analyses associated with the 15,745
analyses of soil, vegetation, terrestrial animal, marine organisms, and marine sediment
4