STACK SAMPLING 1097
the sampling duration might be changed. First, if the source
is known to vary over a longer or shorter cycle, such that
a different sampling period would be likely to yield more
meaningful results. Second, with some state operating per-
mits establishing very low emission limits, sampling times
are increased to provide adequate quantification limits.
At this point, the sample gas stream should be as repre-
sentative as can be collected using these methods. However,
there is still the inherent variability of the methods them-
selves. To account for this, most testing programs include a
series of three or more test runs in a single test. Depending
on the underlying regulations, the results of the test runs can
be used independently or averaged to reach a final emission
measurement.
Once an appropriate sample gas stream has been
extracted from an exhaust stream, it must be collected in
some way for subsequent analysis. Most particulate sam-
pling trains separate the particulate matter from the rest of
the gas stream, saving the particulates and exhausting the
particulate-free gas sample stream. This separation can be
accomplished physically, chemically, or both. Examples of
physical separation are filtration, inertial separation (using a
“cyclone”), or condensation. Examples of chemical separa-
tion include dissolving into solution, adsorption, or chemical
reaction into a solution.
One or more of these separation techniques may be
appropriate for a given particulate material or under given
process and/or stack conditions. The specific technique to
be used and the limiting conditions may be found under the
specific Test Methods described later in this Section. The
essence of each is to ensure that all of the particulate matter
flowing through the nozzle and probe is actually collected
for subsequent analysis.
The following is a brief list of these particulate sampling
principles, rearranged in an order that might be followed for
an actual test.1) Select a sampling location far from disturbances.
2) Determine traverse point locations.
3) Select appropriate nozzle, probe, heater, and so on.
4) Determine appropriate time for testing (worst
case, etc.)
5) Determine isokinetic sampling rates.
6) Select sample train configuration.
7) Perform test runs.Gaseous SamplingWhen the pollutant of interest is a gas at ambient pressure,
sampling is much easier and more straightforward than for
particulate samples. This is because gas streams are almost
always well mixed across the stack and are not subject to
the internal considerations at the nozzle or in the probe. As
a result, there is usually no need to worry about the sample
location, traversing the stack cross-section, or isokinetic sam-
pling. An exception is the measurement of nitrogen oxides
from gas turbines using method 20. This method requires a
stratification study.The only real concern is that the probe be constructed
of materials that will not react with or adsorb contaminants
from the sample stream.
Sampling is usually performed by inserting a probe at
a convenient location and sampling at the centroid of the
cross-section. As with particulate sampling, it is important
to measure the gas flow rate in the stack to allow calcula-
tions of emission rate. This is done in the same way as for
particulate sampling using Test Method 2.Protocol and Final Report PreparationA stack sampling project, like most other investigative
work, is not likely to succeed unless it is well planned and
documented. The Stack Test Protocol, or Quality Assurance
Project Plan, is the means used to document the planning
of the project. The more detail that can be included in the
protocol, the better the likelihood that the test will suc-
ceed the first time. Most EPA Regions and State Agencies
have specific protocol formats and require specific types of
information. Therefore, the project manager should contact
the Agency well before the projected test date to obtain the
format and to discuss any special conditions that need to be
included, such as audit samples.
The final Stack Test Report is just as important as the
Protocol. It is the means by which the testing team documents
what they did and their results. If the Report does not fairly
report what actually happened on the stack, in the laboratory
and in the calculations, the entire test might well be wasted.
Again, it is advisable for the project manager to contact the
regulatory agency well before the test to obtain information
of acceptable report formats and special information that
might be needed. It is likely, though, that the Agency will
require copies of all field data sheets, lab data sheets and
print-outs, calculation procedures, examples, and results,
diagrams, etc. A fourth advance has been made in the dis-
semination of stack testing information through EPA’s elec-
tronic bulletin board. The Emission Measurement Technical
Information Center (EMTIC) bulletin board is available as
part of the Technology Transfer Network Bulletin Board
Service (TTNBBS). The EMTIC bulletin board includes
promulgated methods, proposed methods, some state test
methods, papers on stack analysis, a data base on validated
methods for various compounds, etc. It can be used to get
answers to specific questions. Access to the TTNBBS is
available through the Internet at “http://www.epa.gov//ttn”
Stack Test Guidance is available at http://www.epa.gov/
Compliance/assistance/air/index.html
This his document does not address test methods.
However, it does provide a good discussion of regulatory
requirements for stack testing, including notifications, time
frames, observation by regulatory agencies, and reporting.TEST METHOD DESCRIPTIONSThe main body of this chapter includes brief descriptions
of selected current U.S. EPA Test Methods. These are theC019_003_r03.indd 1097C019_003_r03.indd 1097 11/18/2005 11:07:14 AM11/18/2005 11:07:14 AM