EPA contract labs

[VJP]

I work in GMP Lab so not familiar with EPA regulations.
1. Are EPA test methods validated (linearity, accuracy, specificity and robustness etc.)?
One of our contract lab is testing samples using EPA METHOD 8260B (GC MS), but we don't know if they are using the same parameters every time they analyze the samples. In several GC MS runs, I noticed that their std and check std % agreement is ranging from 92 to 115%. That's a big difference and their spec is 80 to 120%.
2. Does EPA requires to test wastewater samples has to by EPA lab within the state.
Example: if wastewater sampled from OH then samples needs to be tested by EPA lab in OH.

Thanks......

[Steve Reimer]

The standards for validation and how methods are developed by EPA depend upon the program. Wastewater and Drinking Water programs have promulgated methods for compliance monitoring. Those have an intended purpose and are required to be used for specified monitoring. SW-846 methods are intended for RCRA, Superfund and other programs and are no longer promulgated, except for method defined parameters. They are written to include inherent flexibility so that they can achieve data quality goals across a variety of uses. CLP uses the SW-846 methods but has contractual requirements specifying data quality objectives.
Compliance samples have to be run by a certified lab, there is no requirement to run them in any location. In fact large commercial labs run samples from all over the country. That is why TNI was created.

Overall it is not important that they use the same parameters every time, otherwise they should always use the same instrument, instead they should meet the same quality specifications. 92 to 115% looks like it is well within the 80 to 120% spec.

[James_Ball]

One note on Ohio, I remember in the past that they would not certify a laboratory that did not at least have an office in the state. Some large labs have at least a sales office in the state so that they can get certification to run samples there.

Method 8260 has Calibration Check criteria set to 80%-120% recovery and initial calibration limit is 15%RSD to use average response factor, if a compound exceeds that then either Linear or Quadratic fit must be used.

Method 524 for Drinking Water Volatiles requires 70%-130% Calibration Check criteria and 30%RSD to use average response factor calibration, otherwise if a compound exceeds that you must use linear or quadratic fit.

Method 624 for Waste Water Volatiles is a nightmare when it comes to Calibration Check criteria since you have to use the table provided in the method which has a different range for each compound. A 20ppb check standard must pass Vinyl Chloride at 4%-196% recovery, Ethylbenzene is 59%-141%, Chloromethane is 0%-204% recovery, .....

Method 624 was also based on a Packed Column GC/MS instrument and I think has finally been updated this past year to include verbiage to include capillary columns. I think the prior update was maybe 1986 or early 1990's. But regulating agencies still require you to follow it to the letter just because lawyers are writing the regulations instead of chemists.