EPA Publishes Summaries on Detection, Quantitation and Calibration
The EPA has recently published several document summaries about method detection limits, quantitation limits and calibration curves. To create the summaries, EPA evaluated the current use of these concepts and future needs across EPA program lines. The resulting compilations will help practitioners better understand the existing terminology and how each EPA program handles data.
One key point addressed by nearly all EPA programs is the need for flexibility in managing method detection/quantitation limits and calibration curves rather than rigid requirements. Some programs prefer having a performance based approach, allowing flexibility for laboratories to modify analytical approaches as necessary and as makes sense for that particular instrumentation/compound.
Related to method detection limits (MDLs), OSWER1 and ORCR2 removed the need to perform MDLs as per 40 CFR Part 136. The latest revisions of the applicable analytical methods (SW846 revision IV) introduce the new terminology for lower limit of quantitation (LLOQ) instead. By definition, the LLOQ is most often equivalent to the lowest calibration standard. A LLOQ verification is recommended for each project application/matrix and concentrations should ideally be less than regulatory action levels.
A new concept related to calibrations is the Relative Standard Error (RSE)3 calculation. The RSE calculation is published in the Federal Register for consideration by the Office of Water, Office of Science and Technology in addition to OSWER and ORCR.
A summary table organizes and presents the current requirements for each EPA program, and an Environmental Measurement Glossary defines commonly used acronyms and lets the reader know which EPA programs use which acronyms.
Additional information about these documents can be found on the EPA’s website at http://www.epa.gov/osa/fem/calibration.htm
1. EPA Office of Solid Waste and Emergency Response
2. Office of Resource Conservation and Recovery
3. A measure of a statistical estimate’s reliability