PARTICULATE EMISSIONS 827
Gas analysis for CO 2 , CO, and O 2 is almost always done
by Orsat analysis. Moisture may be determined gravimetri-
cally by condensation from a measured volume of gas as
required by EPA.
Overall Flow Rate Total flow rate is calculated from duct
area and average gas velocity as determined by the pitot tube
measurements. Pitot tube traverse points are at the center of
equal areas so no weighting is necessary to determine aver-
age velocity. This gives flow at duct conditions which is usu-
ally converted to standard conditions.Measurement of Particulate Concentrations in StacksStandard methods for measuring particulation concentrations
in stacks depend on the principle of isokinetic sampling. Since
particles do not follow gas streamlines exactly but tend to
travel in straight lines, precautions must be taken that the gas
being sampled experiences no change in velocity or direction
in the vicinity of the sampling point. This is done by using a
thin walled tubular probe carefully aligned with the gas flow
and by withdrawing gas so that velocity just within the tip of
the probe equals that in the main gas stream. Several recent
studies 29 – 31 have measured effects of probe size, alignment,
and velocity on accuracy of sampling. The sampled gas is
drawn through a train of filters, impingers, and a gas meter
by means of a pump or ejector. Typical probes are shown
in Figure 7, and several types are commercially available.
With these probes the necessary gas sampling velocity must
be previously determined by pitot tube measurement, and
the gas flow rate at the flow meter is adjusted (taking into
account gas volume changes due to cooling and condensation
between stack and meter) to equal that velocity. An alternate
method is to use a null nozzle, which contains static pressure
taps to the outside and inside surfaces of the sample probe as
shown in Figure 8. Flow through the probe is adjusted so that
the static pressures are equal at which point the velocitiesinside and outside the probe should be the same. The null
nozzle greatly simplifies sampling, but null nozzles require
careful periodic calibration and are not generally used for
high precision work.
The sampling train of filters and impingers, which col-
lects the particles, is usually carefully specified in the test
method or governmental regulation in force. Differences
between sampling trains to some extent reflect different
technical solutions to the sampling problem but they also
reflect differences in the philosophy of what exactly should
be measured.
Perhaps the most widespread train will be that specified by
the EPA^14 for testing new emission sources, shown in Figure 9.
The original intent was to collect and measure not only par-
ticles which actually exist in the stack at stack conditions, but
also solids or droplets that can be condensed out of the stack
gas as it is cooled to ambient conditions. The filter is heated
to avoid condensation and plugging. The first two impingers
contain water to collect most of the condensables. The third
impinger is empty and serves as an additional droplet tray
while the fourth impinger is filled with silica gel to collect
residual water vapor. Although the impingers in the train col-
lect condensibles, present regulations are written only in terms
of the solid particulates which are collected in the filter.Slatic tapSlatic tap17 6 - No.60 holes(^212)
(^3436)
(^34)
(^58)
(^12)
58 1
(^118)
(^1) 5°
3
1
12 °
FIGURE 8 Null type nozzle for isokinetic sampling.
HEATED AREAFILTER HOLDERTHERMOMETERCHECK
VALVE
VACUUM
LINE
ICE BATH
VACUUM
GAUGE
MAIN VALVE
AIR-TIGHT
PUMP
DRY TEST METER
THERMOMETERS
STACK
WALL
PROBE
REVERSE-TYPE
PITOT TUBE
PITOT MANOMETER
ORIFICE BY-PASS VALVE
IMPINGERS
IMPINGER TRAIN OPTIONAL. MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
FIGURE 9 Environmental Protection Agency particulate sam-
pling train.
Smooth Bend
d
Angle
30° or less
Pipe thread connection
to thimble holder
d
R
R
R .2d
Knife-edge circular opening
with straight internal wall
A. Elbow Nozzle
B. Goose-neck Nozzle
FIGURE 7 Nozzles for particulate sampling.
C016_001_r03.indd 827C016_001_r03.indd 827 11/18/2005 1:15:35 PM11/18/2005 1:15:35 PM