Background
For More Information
EPA has found that many Method 323
emission tests are not performed in
accordance with the method. Inaccurate
measurements may be obtained due to
common test method deviations, irregu-
larities, and errors.
This brochure summarizes best prac-
tices to minimize human error during
preparation, sampling, and data analysis
in performing Method 323.This guidance
applies when using Method 323 to mea-
sure emissions from both engines with
post-combustion controls (e.g., oxidation
catalysts) and engines with no controls.
To download or print more
copies of this brochure:
www.epo.gov/nscep
To access a Technical Report that
provides additional information
on the recommendations
in this brochure:
Visit www.regulations.gov and
search for Document ID
"EPA-HQ-OAR-2008-0708-0467"
To view the full text of EPA
Method 323:
Visit www.epa.gov/ttn/emc/methods/
method323.html
Still have questions on this method?
Contact the EPA expert Ray Merrill at
merrill.raymond@epa.gov
United States
Environmental Protection
Agency
How to Avoid Common
Testing Errors when Using
US EPA Method 323 - Best
Practices for Best Results
305-F-I2-002
October 2012
-------
Preparation
Sampling Procedures
QA/QC and Data Analysis
I. Minimize the sample line length where
possible, and ensure the sample line
configuration directs flow downward
toward the first impinger.
Condensation may accumulate in sections where
the sample line configuration is parallel, sloped
upward, or looped. Sample losses may result
where condensate containing formaldehyde is
allowed to accumulate, leading to inaccurate
measurement.
2. Inspect and clean sampling lines before
conducting test.
Pre-test inspection and cleaning will remove
contaminants that may be present. Flushing the
sampling lines with air will remove any moisture.
3. Make sure to configure the sample train
with midget impingers.
Method 323 Section 6.3 specifies three midget
impingers are required for sample collection.You
should avoid using Greenburg-Smith impingers
or larger impingers since the added volume of
the impingers and associated recovery rinses
would raise the detection limit of the method.
4. Do not exceed the test method maximum
sampling rate of 0.4 Llmin.
It is critical that you do not sample at a flow rate
higher than 0.4 L/min. Sampling at higher flow rates
may reduce formaldehyde collection efficiency
resulting in measured formaldehyde concentrations
that are less than the actual concentrations.
5. Make sure to thoroughly rinse the probe and
sampling line during sample recovery, even if
more than 10 mL of water is required.
Method 323 Section 8.3 states:"ln general, com-
bined rinse volumes should not exceed 10 ml."
However, in cases where a long, flexible extension
line must be used to connect the sample probe to
the sample box, sufficient water must be used to
rinse the connecting line to recover any sample that
may have collected.The volume of the rinses during
sample recovery should not be excessive, as this
may result in your having to use a larger-than-40 mL
VOA bottle.
6. Be aware that accurately measuring moisture
with Method 323 may be challenging.
Make sure to thoroughly dry the outside of the
impingers before weighing. Avoid allowing conden-
sate to collect in the sampling line, as described in
recommendation #1 in this brochure. Moisture that
accumulates in the sampling line should be added
to the first (knock out) impinger and included in
the moisture determination. Since moisture may
collect in a common sampling line feeding dupli-
cate sampling trains, the duplicate trains must be
independent, including independent sample probes
and lines.
7. Ensure that you review and follow the
Quality Control requirements for Method
323.
Method 323 Section 9 lists QA/QC requirement
categories, acceptance criteria, and
corrective actions.
8. When calculating mass emissions of
formaldehyde using Method 323
Equation 323-5, ensure that Kc, the
spectrophotometer calibration factor,
has been correctly developed.
Method 323 Section 12.1 defines Kc as:"spectro-
photometer calibration factor, slope of the least
squares regression line, |jg/absorbance." It should
be noted that the value of micrograms, |jg, used
in developing the regression line to calculate
Kc, should be per 2-mL aliquot. Equation 323-5
includes a division factor of 2 to account for this;
if Kc is not developed on the basis of a 2-mL
aliquot, formaldehyde mass emissions will be
under-estimated by a factor of 2.
------- |