NonDetects
Dealing with results below the detection limit can cause problems to say the least.
100% NonDetects
For parameters that are all non-detect in the background (i.e., organic compounds) the UG suggests using the Double Quantification (Section 6.2.2 of the UG).
A confirmed exceedance is registered if any well-constituent pair in the ‘100% non-detect’ group exhibits quantified measurements (i.e., at or above the reporting limit [RL]) in two consecutive sample and resample events.
Two big points here:
The result has to be above the Reporting Limit so “j” flagged data do not count.
This is not a formal statistical test and therefore does not count into the 5% site wide false positive discussed in section 4 of this document.
What about less than 100% NonDetects?
More to come but here is a link for now.