Laserfiche WebLink
Revised Monitoring and F\_,.orting Program No. 93-100 g <br /> Lawrence Livermore National Laboratory Site 300— Pit 1 <br /> San Joaquin County <br /> Table 4: Pit 1 Constituents of Concern, Evaluation Monitoring Wells and <br /> Statistical Methods <br /> Constituent of Evaluation Statistical <br /> Concern Monitoring Wells Method <br /> Tritium (pCi/L) K1-06 Trend Analyses <br /> W-PIT1-2209 <br /> W-PIT1-02 <br /> _ W-865-2005 <br /> Perchlorate K1-06 Trend Analyses <br /> (Ng/L) W-PIT1-2209 <br /> W-PIT1-02 <br /> W-865-2005 <br /> STATISTICAL PROCEDURES FOR DETERMINING SIGNIFICANT INCREASES <br /> 14. The significance of increases for detection monitoring wells in constituents of <br /> concern over water quality protection standards shall be established through <br /> the use of the statistical procedures in California Code of Regulations, Title 23, <br /> Division 3, Chapter 15, Section 2550.7 as identified in the above tables. <br /> 15. LLNL uses an intrawell comparison for each analyte at each downgradient and <br /> cross-gradient well to test for a "statistically significant evident of a release." <br /> Intrawell tests compare each measurement at a well to past measurements at <br /> that well, in order to detect increases in concentration that exceed the radiation <br /> historically seen at that well. Intrawell tests are used when (1) there is spatial <br /> variation among the monitoring locations, due either to natural variation or <br /> impacts from an upgradient source; or (2) there is known contamination at a <br /> monitoring location and the monitoring program is required to detect additional <br /> increasing levels of contamination or new releases. <br /> 16. Where sufficient detections are available, LLNL uses a statistical prediction <br /> limit method to implement intrawell comparisons. The method uses the <br /> average and standard deviation of historical measurements to calculate a <br /> statistical limit (SL) value. The SL is calculated so that it will be exceeded by <br /> approximately one percent of individual measurements when there has not <br /> been a release, in accordance with CCR Title 23, Chapter 15, Section 2550.7. <br /> When too few detections are available, either the analytical reporting limit or <br /> maximum recent detection is used as an SL. Subsequently, each <br /> measurement is compared with the SL. <br /> 17. The SL comparison is augmented by a verification procedure containing two <br /> discrete retests, in accordance with CCR Title 23, Chapter 15, Section 2550.7. <br /> This protects against false positives due to other causes, such as analytical <br />