Counting errors due to overlapping particles in thermal precipitator samples.

A sample of airborne dust taken with the standard thermal precipitator consists of two deposits about 1 cm. long and 1 mm. wide obtained from either side of the heated wire (Green and Watson, 1935). These deposits are examined under a microscope and a count is made across each of them to estimate the number of particles collected. The particles commonly have an irregular shape and some consist of several smaller particles clumped together. It is usually assumed that these clumps or " aggregates" of particles were present when the dust was airborne and they are therefore counted as single particles. During the sampling some of the particles are deposited so that they cover or overlap others already deposited, producing spurious aggregates. When examined under the microscope this type of aggregate cannot normally be distinguished from one which existed as such in the air and these aggregates are also counted as single particles. Consequently the airborne dust concentration is underestimated. In the many previous tests of the thermal precipitator the magnitude of this error does not seem to have been investigated. If two instruments are operated close together and the volume of air sampled by one is greater than the volume of air sampled by the other, the deposits obtained with one instrument will be more crowded than the deposits obtained with the other and hence the airborne dust concentrations estimated from the two instruments will be different. Therefore it was thought that by comparing the results from pairs of samples taken in this way, with very different volumes, it would be possible to estimate the magnitude of the bias due to overcrowding that normally occurs. As air is drawn through the thermal precipitator at a constant flow rate the volume of air sampled can be varied by altering sampling time.