Method Detection Limit (MDL) or Limit of Detection (LoD)

At Chemex, Method Detection Limit (MDL) is the estimated minimum concentration of the analyte in a sample that can be detected with a prescribed level of confidence of distinguishing that sample concentration from a method background signal.  The MDL should be referred to by data users when considering the significance of a reported low level concentration in a sample.  MDL values will be established for all standard analytes.

For UKAS accredited and any non-accredited tests the MDL will be determined and calculated according to the adopted USEPA protocol as defined in the US Federal Register, CFR Title 40- Protection of Environment, Chapter I- Environmental Protection Agency, Volume 22, Part 136- Guidelines Establishing Test Procedures for the Analysis of Pollutants, Appendix B- Definition and Procedure of Determination of the Method Detection Limit, revision 1.11, July 2006.

MDL =  student 'T' value  x  the standard deviation of the replicate measured sample concentrations

where the student 'T' value used will depend on the number of sample replicates (n) with n - 1 degrees of freedom (d.f) and represents the 99% confidence level, e.g. for seven replicates the student 'T' value used is 3.143 with 6 d.f.  The method detection limit will be expressed in the appropriate concentration units for the sample matrix of interest and for soil samples will be expressed in terms of dry weight, i.e. theoretical 0% moisture.

The MDL will be statistically derived by analysis of a batch of minimum of seven spiked replicates using an ideal matrix, that is, a matrix known to be absent of the analyte.  The replicates will be spiked with concentrations one to five times the estimated MDL.  By analysing the replicates using all steps included in the analysis of real samples, any likely errors in the method will be encompassed and thus the later comparison of the MDL to real sample data will be more valid.

For MCERTS accredited parameters the MDLs will be determined and calculated according to the EA Performance Standard for Laboratories Undertaking Chemical Testing of Soil, MCERTS, Version 3, March 2006.

The MDL for all analytes will be quoted in all final analytical reports for cross reference.  It should be noted that the method detection limit has been produced using an ideal matrix and that unknown interferences from real sample matrices may adversely affect the ability of the method to detect a concentration at this level.  The final report should note that any change to the protocol method e.g. different sample volumes taken or dilution of sample extract will result in the method detection limit quoted not being directly applicable.  Therefore a recalculated estimate of the method detection limit should be supplied, or where this is not practical, guidance in making the recalculation should be provided.