-----Forwarded by Melanie King/RTP/USEPA/US on 05/13/2010 08:49PM -----
To: Gary McAlister/RTP/USEPA/US@EPA 
From: Thomas McGrath <tmcgrath.mea@gmail.com> 
Date: 05/13/2010 05:10PM 
cc: Melanie King/RTP/USEPA/US@EPA, RobertJ Wayland/RTP/USEPA/US@EPA, Lisa Beal 
<lbeal@ingaa.org>, Matt Todd <ToddM@api.org>, "Jeffrey A. Panek" 
<japanek@ameritech.net>, Jim McCarthy <jamesmccarthy@comcast.net> 
Subject: EPA Method 323 and EPA Method 320 Formadehyde Measurements Comparison: 1 
of 3

Gary – per your request, the attached report provides the results of side-by-side formaldehyde 
measurements by Proposed EPA Method 323 and EPA Method 320 (extractive FTIR).  The report was 
prepared by Innovative Environmental Solutions and the project was sponsored by the American 
Petroleum Institute (API) and the Interstate Natural Gas Association of America (INGAA).  
 
In summary, a total of five test series (3 or 4 test runs per series) were conducted on rich-burn and lean-
burn gas-fired reciprocating engines.  The Method 323 measurements were higher than the Method 320 
measurements for all of the test series with the average difference ranging from 12 to 47 percent.  For 
these tests, all of the Method 320 QA spike recoveries were less than 100% suggesting a low bias to the 
measurements.  When the Method 320 measurements are adjusted for the QA spike recoveries, there is 
very good agreement with the Method 323 measurements with the average difference ranging from -4 to 
8 percent, well within the uncertainties of these test methods.  These data suggest that a low bias in the 
Method 320 measurements, indicated by the low spike recoveries, contributed to the difference between 
the Method 323 and Method 320 measurements for this test program.
 
Based on accepted statistical evaluations to validate method performance, the results demonstrate 
acceptable performance for Method 323 and support the conclusion that Method 323 should be included 
in the RICE NESHAP (40 CFR 63, Subpart ZZZZ) as an approved test method for formaldehyde 
measurement.  In addition, anomalous Method 323 results reported by EPA in the March 2009 proposed 
RICE NESHAP amendments have been investigated and determined to be a result of testing errors (i.e., 
human error).  Therefore, Method 323 should be accepted for formaldehyde measurement from gas-fired 
reciprocating engines and promulgated as an accepted EPA test method when the final rule that amends 
40 CFR 63 Subpart ZZZZ is published, which is planned for late summer or fall 2010.
 
Please note that, because of the size of the files, some report attachments will be sent separately.
 
Please call me if you have any questions or would like additional information.
 
Thank you – 
 
Tom
 
--  
Thomas McGrath 
Innovative Environmental Solutions 
610 California Street 
Huntington Beach, CA 92648 
tmcgrath.mea@gmail.com 
714.315.4040

