A winter chill has floated into Research Triangle, NC, and will likely transition to a cold shiver as we move through the winter months. Meanwhile, our stack-tester friends are still going strong out in the field knocking out one test after another. No matter how close to the end of the year we might be, however, the tips herein will stand the test of time and prove helpful throughout every season of the year.
Maybe you had a sampling event just wrap? Have a compliance test coming up? Perhaps you are onsite right now and the facility folks are stressed, regulators are a challenge, your confidence in the approach is a bit shaky and you are generally feeling like nothing is working out for you. Well, you have one thing on your side…
The Performance-based Test Method!
The self-validating, multiple-option, hold-time determining, defensible-data producing gem of an approach is ALWAYS there for you. Shout out to EPA Method 18, Method 308, and NCASI 105 to highlight a few.
These types of test methods are also referred to as “self-validating” approaches in that the data actually defends itself. I’ll explain…
What Does “Self-Validating” Really Mean in This Love Equation?
We call these approaches “self-validating” because they are designed to demonstrate recovery of the reported compounds, or “analytes of interest” in each specific source matrix by using a field spike and recovery technique. During the collection of gaseous samples, we draw the sample (gas from the source we are testing) through a “sampling train”. This is our “native sample” or “un-spiked sample”. We also purposefully introduce a known quantity of whatever the analytes(s) of interest are into a second sampling train, which is collecting gas side-by-side of the un-spiked train. We then compare results from this “spiked train” to the results of the native source gas, or “un-spiked train”. The results are compared and recovery calculations are presented. The data will then, in theory, show that the spiked analyte was recovered within the method-prescribed acceptance range, thus demonstrating that the method was applicable to the analyte of interest and the source. BOOM!
As long as we demonstrate recovery of the spiked sampling trains, we can be confident that the results for the native samples are representative of the source. Yes, I know, it is indeed like magic.
Why Performance-based Test Methods Aren’t Just a Whirlwind Romance
So why are we totally smitten with these performance-based test methods, and think you should be too?
For starters, they are incredibly versatile. Performance-based methods have the potential to be applied to a variety of different compounds provided the performance quality control (QC) criteria described in the applicable method are met. Even though there may be a list of target analytes referenced in the method, applicability can usually be extended beyond the named compounds. As long as QC is successfully demonstrated and spiked compounds recover, you’ve got yourself some defensible data! Not to mention, the Environmental Protection Agency has been documenting a push toward performance-based measurement for well over a decade now (see links below for additional reading).
Method 18, for example, allows for several different sampling options: direct interface or dilution interface, tedlar bags, or adsorption tubes. The tester determines which of the sampling options to use after considering the source and associated conditions. As long as the recovery criteria described in EPA M18 Sec 8.4 are met, the test approach is supported.
In cases where sample hold-times are vague, the source will actually determine the hold time. We demonstrate that samples are still within hold time by reporting a passing recovery study and once again, TA-DAA! We’ve got strong data and proven sample stability.
While we love these methods oh-so-very much, they can also be a bit intimidating, what with all the required QC and spiked sampling trains and all. You might find yourself holding your breath while skimming the spike recovery page in your freshly-delivered lab report. No? Well, I have certainly broken a light sweat while reviewing a report here and there.
So, what can be done prior to test time to set ourselves up for some passing recoveries and subsequent happy facilities and regulatory representatives?
Source Knowledge: The Foundation of a Happy
Relationship Performance-based Testing Method
Know. Your. Source.
I mean REALLY get to know it. Moisture levels, temperature, flow rate, previous data, any other known conditions may be useful in determining what approach to use. It is great to have this information to share with your contact on the laboratory side so that they can help with appropriate spiking levels for compliance testing in the field. Trust me, the lab definitely wants to deliver a report with passing QC just as much as you want to see it.
Any information on the expected concentration of target analytes at the source is going to be necessary to pass on to the laboratory preparing the spike media. If spikes are prepared at a low level and then native concentrations are very high, the native catch weights are going to be so heavy that the amount of analyte spiked will not recover.
For instance, let’s say we spiked 50 ug of Benzene on a charcoal tube for EPA M18. We assume that if the native samples are non-detect for Benzene, then we should show 50 ug Benzene in the final results for the spiked samples. However, if you have 5,000 ug of Benzene catch on your charcoal tube in the un-spiked sampling train, then we have totally under-spiked the tube media at only 50 ug. The slightest variation between the un-spiked run and spiked run could not only produce a failing recovery result; you may not see any of that 50 ug of spiked Benzene since there is so much Benzene in the source. In this case, a more appropriate spike amount would have been ~ 2,500 ug of Benzene on the tube media.
If there is no previous analytical data on the source, and the amount of analyte of interest is unknown, it is advisable to conduct a presurvey if at all possible. This helps get numbers that can be used to plan for compliance testing and determine appropriate spike amounts for the QC sample runs. For example, in NCASI Method A 105.01, section 1.2 strongly suggests that “sampling contractors and laboratories conduct train spikes and conduct full trial runs of all sample train configurations prior to use of this method in the field. Failure to do so will greatly increase the probability that quality assurance criteria will not be achieved.” We all want a successful compliance test with passing spikes, right? So when no historical data is available, vote “yes” for presurvey testing!
Lovers, Not Fighters
To recap, the performance-based test method is friend-not foe-when used appropriately. Having a strong profile of information on the source to be tested, and sharing that information with the analytical laboratory, is going to help in choosing the right test approach and set the foundation for a beautiful test event that will generate passing spike recoveries, defensible data, and demonstrate method QC with a snazzy red bow on top.
For More Information
For more information on EPA’s push toward performance-based measurement, please check out the following resources:
We Can Help!
Have questions about this topic? Need an air testing expert, or general help with your laboratory needs? Click here to start a conversation.