REMOVAL_1999 - Laserfiche WebLink

Home Browse Search

�.. SECTION 1 91 1. Trace 2. ND(not detected) 3. Numerical value of MDL 4. A"less than"the numerical value of the MDL(BDL) 5. Zero 6. Some value between zero and the MDL,for example,half the MDL Y Actual measured value,even if below L 8. Actual measured value, MD parentheses Y followed b 9. Actual measured value with an estimate of precision,for example 2 t 4µg/L, where the estimate of error can be the MDL- Additional reporting forms are: 10. A number followed by Estimated Maximum Possible Concentration(EMPC) 11. Detected but less than MDL 12. Detected but less than PQ L ults in a real problem for some of the end The plethora of reporting techniques res users of the data. Those who wish to know whether their effluent is above or below a regulatory limit do not really care about the exact value of a pazazneter when it's on the I safe side;however other users,especially those who input the r el into mathematical 1 tter i I models for site assessments, are veryexactin in their data req applications"Not Detected" is quite different from a"Detected,but less than the MDL" or M "Below Detection Limits:' It is somewhat imperative for the laboratory to develop a dialog with the end user to determine the exact requirements of the data. Method detection limits(MDL)are the minimum level of an analyte that can be f ! determined with 99%confidence. Once an IDL has been determined,it is readily appar- ent that for samples that have target analyte present at the IDL, at least half the analyses S of the sample will result in responses below the IDL(Figure 1-21). The objective of the P' ' analyst becomes a determination of how high on the calibration curve one must go so that all signals will be above the IDL,in other words,no false negatives will be reported. n r The MDL attempts to answer this question. The procedure that follows b again drawn from EPA 40 CFR Part sw Appendix B. The analyst should consult the 40 CFR refer- ence,as th treoc dare a brief the mrelthod°sras follows:hoices that must be made during the MDL p I. Prepare a spike of the analyte into laboratory water that is very close(2 to 5 times)to the IDL obtained above. 2. Take seven aliquotes of this spiked solution and process each through the sample clean-up and preparation procedure. 3. Analyze each of the prepared aliquots in the exact same manner as Prescribed in it ! the Method used. 4. Calculate the standard deviation(S%R)of the aliquot results as described above. led t-statistic at a 99%v confidence level for the it $. The MDL is equal to the one-tin performed number of samples times the standard deviation: zc MDL=to.99,n x S%R or in this case MDL--3.143 S%R f a For other numbers of repetitions in the MDL study,the appropriate value of tp,99 can be found in Table 1-31. Genium Publishing Corporation