|
Revised 12 September 1995
<br />GENERAL COMMENTS ON TRITIUM RESULTS
<br />Tritium Scale (New)
<br />Tritium concentrations are expressed in TU, where 1 TU indicates a T/H ratio
<br />of 10-18. The values refer to the new tritium scale of U.S. National Institute
<br />of Science and Technology (formerly NBS), and based on their tritium water
<br />standard rt4926 as measured on 1961/09/03 and again 1978/09/03, and age-corrected
<br />with the new half-life of 12.43 years, i.e., a - 5.5769 year 2. In this scale,
<br />1 TU Ss 7.088 dpm/kg H2O, or 3.193 pCi/kg H2O, or 0.1181 Bq/kg H-0 (Bq -
<br />disint/sec). TU values are calculated for date of sample collection, REFDATE in
<br />the table, as provided by the submitter. If no such date is available, date of
<br />sample arrival at our laboratory is used. The stated errors, eTU, are one
<br />standard deviation (1 sigma) including all conceivable contributions. In the
<br />table, QUANT is quantity of sample received, and ELYS is the amount of water
<br />taken for electrolytic enrichment. DIR means direct run (no enrichment).
<br />Through 31 December 1993, we reported tritium values in the "old" scale
<br />using the half-life 12.26 years, i.e., a - 5.659 year 1. In that old scale, 1
<br />TU(old) is 7.186 dpm/kg H2O, 3.237 pCi/kg H2O. To convert from the new scale
<br />back to the old at any given point in time, multiply the listed TU(new)-values
<br />by F, where
<br />F - 0.9645 - (year-1990) x 0.0008
<br />i.e. for 1994 the factor is 0.9613. The formula is correct within 0.029 between
<br />1962 and 1999. To convert data from the old scale to the new, divide by F.
<br />Verv low tritium values
<br />In some cases, negative TU values are listed. Such numbers can occur because
<br />the net tritium count rate is, in principle the difference between the count rate
<br />of the sample and that of a tritium-free sample (background count or blank
<br />sample). Given a set of "unknown" samples with no tritium, the distribution of
<br />net results should become symmetrical around 0 TU. The negative values are
<br />reported as such for the benefit of allowing the user unbiased statistical
<br />treatment of sets of the data. For other applications, 0 TU should be used.
<br />Reliability of results
<br />Refer to Services Rendered (Tritium), Section II.B, in the "Tritium Laboratory
<br />Price Schedule; Procedures and Standards; Advice on Sampling". Tritium
<br />efficiencies and background values are different in the nine counters and values
<br />are corrected for cosmic intensity, gas pressure and other parameters. For
<br />tritium, the efficiency !s typically 1.00 cpm per 100 TU (direct counting). At
<br />SOx enrichment, the efficiency is equivalent to 1.00 cpm per 2 TU. The background
<br />is about 0.3 cpm, known to about ± 0.02 cpm. Our reported results include not
<br />only the Poisson statistics, but also other experimental uncertainties such as
<br />enrichment error, etc.
<br />References
<br />Mann, V.B., M.P. Unterveger, and B.M. Coursey, Comments on the NBS
<br />tritiated-eater standards and their use, Int. J. Appl. Radiat. Isot., 33, 383-
<br />386, 1982.
<br />Taylor, C.B., and V. Roether, A uniform scale for reporting low-level
<br />tritium measurements in water, Int. J. App1. Radiat. Isot., 33, 377-382, 1982.
<br />
|