STACK SAMPLING 1117
The following Measurement System Performance
Specification must be passed by the instrument before actual
environmental samples are analyzed:1) Analyzer Calibration Error must be less than 5%
of the span for the zero, mid-range, and high-range
calibration gases.
2) Zero Drift must be less than 3% of the span
over the period of each run.
3) Calibration Drift must be less than 5 of the span
over the period of each run.The analytical range must be selected based on the appli-
cable regulation and is usually 1.5 to 2.5 times the emission
limit. The EPA Test Method 25A write-up contains detailed
descriptions of the calibration gases required, calibration
procedures, sampling procedures in addition to a list of
references. It should be read in detail before the Method
is attempted. The manufacturer’s instructions will provide
instrument specific instructions. Testing should be per-
formed only by personnel trained and experienced with the
equipment being used.Test Method 27Test Method 27 is used for determining leaks from gasoline
delivery tank trucks or rail cars. It does not involve actual
measurements of gasoline emissions. Instead, it involves the
pressurization and/or evacuation of the tank and the subse-
quent measurement of the pressure and/or evacuation of the
tank and the subsequent measurement of the pressure and/or
vacuum after a given number of minutes. The pressure, time
span, and allowable pressure change are all specified in the
applicable regulation. Typically, the initial pressure is 450
mm water, with an allowable pressure loss of 75 mm water
after five minutes.
Prior to conducting this test the tank must be emptied of
both gasoline liquid and gasoline vapor and must be ther-
mally stabilized. This is best accomplished by flushing the
tank with a non-volatile liquid such as heating oil. In addi-
tion, care must be exercised to protect against fire by ensur-
ing proper electrical grounding of the tank.
A suitable pump is attached to the tank and pressure or
vacuum applied until the specified level is reached, as indi-
cated on a liquid manometer or other instrument capable of
reading up to 500 mm water gauge pressure, with 2.5 mm
precision. The valve is then shut and the reading taken again
after the specified period.
This test method should be performed only by people
experienced in dealing with gasoline delivery equipment and
operation.Test Method 101Test Method 101 is used to determine the emissions of gas-
eous and particulate mercury (Hg) when the gas stream is
predominantly air. It is used mainly at chloralkali facili-
ties that produce chlorine gas from salt in mercury-basedgalvanic cells. The gas stream is extracted from the stack iso-
kinetically using a sampling train like the standard Method 5
train.
The main differences are that the impingers contain an
iodine monochlorine solution (ICI), and no filter is employed.
The probe must be lined with borosilicate or quartz glass
tubing to prevent reactions with the mercury.
The method relies on the reaction of both particulate and
gaseous mercury with the ICI to form HgCl 2 , which remains
in the impinger solutions. During subsequent analysis, the
HgCl 2 is reduced to elemental mercury with a solution of
HCI and SnCl 2 , forming H 2 SnCl 6. The mercury is then aer-
ated into an optical cell where it is measured by atomic
absorption spectrophotometry (AA).
Sample train preparation is about the same as for Method 5,
with a few exceptions. First, care must be taken in select-
ing the nozzle size to allow the use of a single nozzle for
each entire test run. Second, the 0.1 M ICI solution must be
prepared, used to clean the impingers, and added to them.
Third, it may be necessary to break a 2 hour sampling run
into two or more subruns if high Hg or SO 2 concentrations
lead to liberation of free iodine (evidenced by reddening of
the first impinger solution). Finally, an empty impinger may
be used as a knock-out chamber prior to the silica gel to
remove excess moisture.
Calibration of the sampling train and actual sampling
proceed exactly as with Method 5, as do calculations of per-
cent isokinetic conditions. Sample recovery is essentially the
same as for method 5 except that 0.1 M ICI solution is used
as the rinse solution, to ensure capture of all mercury from
the probe walls, etc.
The analytical system of this Method is designed to free
the Hg from solution and to allow the Hg vapor to flow
into an optical cell connected to an AA spectrophotometer
that records the absorption at 253.7 nm light, the charac-
teristic wavelength of Hg. This is accomplished by mixing
a stannous chloride solution with aliquots of the recovered
sample in a closed container then aerating with nitrogen.
The nitrogen carrier-gas and the Hg then flow through the
optical cell.
The flow rate through the aeration cell is calibrated with
a bubble flowmeter or a wet test meter. The heating system
for the optical cell, needed to prevent condensation on the
cell windows, is calibrated with a variable transformer. The
spectrophotometer itself is calibrated by using six differ-
ent aliquots of a 200 ng/ml working solution of mercury,
repeating each analysis until consecutive pairs agree to
within 3% of their average. Either peak area or peak
height may be used. The aliquots are added to the aera-
tion cell and N 2 bubbled through as during cell calibration,
except without the flowmeter. A straight line is then fitted
through the five points; it must pass within 2 percent of
full scale of the origin.
Analysis follows the same procedures as calibration
except that aliquots of diluted sample are substituted for
the working standard. Again, the aerated sample from each
aliquot is analyzed until consecutive pairs agree to within
3%. The average is then compared to the calibration line andC019_003_r03.indd 1117C019_003_r03.indd 1117 11/18/2005 11:07:17 AM11/18/2005 11:07:17 AM