ARCHIVED REPORTS_2012 - Laserfiche WebLink

Home Browse Search

Nonparametric Tolerance Limit When the background data set contains greater than 5O percentbut |eSsUl@n1OOpe[ceOtOOO- detectVo|ueoand/oritodicthbutioniSOD1Oon0a| (OrtransforDedDonna|), Sanitaoapp|ieathe nonparametric tolerance limit method. Hovvever, this method naquinaa a large number ofsamples to achieve afa|oe positive rate of 1 percent or less, which is required by CCR Title 23, Chapter 15. Article 5guidelines. Thus, when the alpha level was higher than 1percent, the concentration limit was not established and data was compared to established limits in the Waste Discharge Requirements. Non-statistical Approach When the background data set contains 100 percent non-detect values, Sanitas applies the non- statistical approach. This approach takes into consideration that if historically a constituent has not been detected, any detection of that constituent would be considered an exceedance. Thuo, in these situations (100 percent non-detected values), the tolerance limit is set at "detect", meaning that if a value is detected above its method detection limit, this value is an exceedance mfthe concentration limit. Trend Analysis Sen's Slope measures the change in constituent concentrations per unit time. Sen'amethod io not greatly affected byoutliers, and the slope can becomputed when data are missing. SeO'G estimator is closely related tothe Mann-Kendall test, which is m nonparametric rank correlation test for trend. The test uses only the relative magnitudes of the data rather than their actual values; therefore, missing values are allowed, and constituents that are reported asless than the detection limit can beused byassigning them values equal tDhalf the detection limit. Gen'o Slope and the Mann-Kendall tests are described in "Statistical Methods for Environmental Pollution K88OitO[iOQ." Richard 0. Gi\bert, Van No3t[aOd Reinhold, New York, 1987. Mann- Kendall test is recommended in "An Evaluation of Trend Detection Techniques for use in Water Clu8|\tV K8OOitO[iDg PnDgrB[OS.^` Jim C. LOftiS, et a|. (USEPA). 1989. For the Harney Lane Sanitary Landfill, the OU|| hypothesis is ''no statistically significant trend in constituent concentrations." The null hypothesis is rejectediftrends inthe data are significant at alpha = 0.05. which is the probability of rejecting the null hypothesis ofnotrend when it is true (i.e., aTypo 1 error). 8anitaa is used to calculate both the Sen'o Slope and the Mann-Kendall test for trends. Harney Lane Sanitary Landfill u-2 Department mPublic Works/Solid Waste