Equal velocity in sampletaking of a gas for dust measurement.
To obtain a representative sample independent of particle size, it is necessary to remove the sample stream isokinetically, i.e. with the same velocity as the main stream. The following Figure shows the pattern of the flow lines in the vicinity of a thin-walled sampling probe.
In the isokinetic case (w=v), all particles flowing toward the intake opening are equally collected. If the sampletaking velocity is too low (w>v), heavy particles can enter the probe even if the flow line on which they were located passes by the probe. Thus too many large particles are collected. If the sampletaking velocity is too high (w<v), heavy particles fail to adhere to the flow lines and end up bypassing the probe. So too few large particles are collected. The error occurring in the case of under-isokinetic sampling (center) is many times larger than in the reverse case.
The following Figure shows the relative dust content e as a function of the velocity ratio of sample stream/main stream (WA/W0) and the factor B, which includes the particles' rate of fall and the probe diameter.
Obviously the error at equal velocities is zero, and it rises sharply for lower sampling velocities. In the range of higher sample stream velocities, however, the error is per and, even more important, is virtually constant from WA/W0 = 1.5 upwards.
Sigrist dust detection installations take advantage of this fact to avoid the need to achieve very expensive isokinetic sampletaking. In these installations, the sampling stream always has at least 1.5 times the velocity of the main stream. The absolute error is taken into consideration in calibration of the instrument, which is carried out with the actual gas at site.
The rate of fall (B) also includes the particle size of the dust particles. The more uniform the particle size distribution, the per the resulting error. The equipment used for flue gas treatment (electrostatic filters, gas scrubbers, etc.) narrow down the size distribution, which enhances the reproducibility of both calibration (solids concentration vs. scattered light) and sampletaking.