|Gossman Consulting, Inc.|
GCI TECH NOTES©
Quantitation Limits and Stack Testing
David Gossman and Greg Rigo
Greg Rigo is a principal of Rigo & Rigo Associates, Inc. of Berea, Ohio and coauthor of the paper referenced in this issue of GCI Tech Notes.
Method 5 - Determination of Particulate Emissions from Stationary Sources
Method 23 - Determination of Polychlorinated Dibenzo-P-Dioxins and Polychlorinated Dibenzofurans from Stationary Sources
Method 26 - Determination of Hydrogen Chloride Emissions from Stationary Sources
Method 29 - Determination of Metals Emission from Stationary Sources
EPA promulgates stack testing methods as part of regulations. For example, the methods listed above are found in 40 CFR Part 60, Appendix A.
EPA requires that these methods be used to determine compliance and in the last few years, EPA is using this data to set MACT limits under the Clean Air Act. There is a significant problem in this situation. In many cases, EPA is using test results from methods, for compliance and MACT standard setting purposes, on sources for which the method(s) have not been validated. Some of these methods do not even have precision established on any source.
While many groups and representatives have expressed concern about this state of affairs and even commented along these lines on draft MACT standards, EPA has not yet scientifically addressed the issue. There has been no action even though the data in the few multi-source type validation studies clearly show that a method's precision is source specific. The validation study for Method 26 even warns that the validation study results for MSW incinerators are not applicable to HWCs or other sources!
On October 1, 1997, a meeting was held by The American Society of Mechanical Engineers (ASME) to discuss the formation of a Reference Method Accuracy and Precision (ReMAP) Consortium. The impetus for this meeting was a study by Greg Rigo and John Chandler which analyzed data from a series of duplicate train tests conducted at a municipal waste combustor to evaluate the precision of all four of the above methods. The data were analyzed to determine the quantitation limits using EPA Method 301 precision criteria, not the promulgated computational procedure. The EPA criteria establishes that the quantitation limit is the lowest concentration at which 99% of the data will likely be within +or-30% of the true value. An alternative way to look at the quantitation limit is that the uncertainty (imprecision) associated with each test run is 30% of that limit. The Reference Method Quantitation Limits (RMQLs) they found are in many cases higher than expected and even higher than some existing and proposed MACT standards. When emissions cannot be measured with sufficient precision, no one can tell if a source is in or out of compliance with a regulatory standard. Also, legal precedent establishing the unenforceability of water permit limits below the practical quantitation limit indicates there is a potential compliance problem if EPA MACT limits or permit limits are below the RMQL for the source category.
The Rigo and Chandler paper, accepted for publication in the Journal of Air and Waste Management Association, found that the RMQL for Method 23 TEQ PCDD/Fs (dioxins) may be as high as 0.51 ng/dscm for municipal waste combustors. The particulate (Method 5) RMQL may be as high as 0.39 gr/dscf. Both of these RMQL s are higher than proposed MACT standards for hazardous waste combustors. Similar problems exist with precision determined for Methods 26 and 29.
The ASME s Center for Research and Technology Development is working to establish the ReMAP Consortium to study the precision and accuracy of existing EPA methods and to help develop any necessary improved reference test methods. These products benefit all stakeholders. If you are interested in joining this consortium and want further information regarding the scope, schedule and funding needs, please contact Greg Barthold (ASME) at 202-785-3756.
(GCI performs emission test data reviews and manages test programs to ensure the highest quality data are obtained and properly utilized. Please call 847-683-4188 if we can be of assistance.)