Optical Particle Counter Errors and Problem Areas

Abstract

Although particle counters are capable of rapidly counting and sizing very small particles in extremely low concentrations, they are also capable of producing erroneous data with no indication to the operator when this situation arises. Particle sizing errors and/or particle counting errors are produced. Particle sizing errors happen when the particles passing through the OPC are defined as being either larger or smaller than their actual size. All OPCs define particle size in terms of the diameter of a sphere whose projected area is equivalent to that of the particle being observed. Very few real particles are spherical, and the reported dimension varies with the orientation of the irregular particle as it passes through the OPC viewing zone. As indicated previously, the OPC response varies with the refractive index of the particle as well as with its shape. OPCs are calibrated with transparent latex spheres of refractive index 1.6. Any particle measured by the OPC is defined as having the same size as the latex sphere that would produce the same scattered light flux as did that particle. The error types that might be found in an optical particle counter can be summarized (Makynen et al. 1982). It was shown that counting efficiency could be affected by OPC resolution, sample-handling system losses, electronic system bandwidth effects on pulse amplitudes, and internal flow system recirculating of particles. Because scattered light from particles does not disperse equally at all angles, OPCs with different optical systems may respond differently to the same particles, as shown in Figure 27-1. The data shown here were obtained by two OPCs with different optical systems; both were just previously calibrated with PSL spheres. Both OPCs were then used to sample from an ambient airstream. Not only is particle concentration at several size ranges reported differently by the two OPCs but also the indicated size distribution function is different. To minimize this error effect, either the OPC to be used is calibrated over its working size range with particles with similar properties to the particles to be measured, or the OPC response is varied so that the signal from the analyzed particles matches that of the calibration particles. A comparison has been made of actual measurements of particles with theoretical calculations for particle counters, one type with narrow-angle collection optics and one with wide-angle forward scattering collection optics (Szymanski and Liu 1986). It was concluded that response for transparent particles is not monotonic for the narrow-angle system, whereas for the wide-angle system, output is attenuated for light-absorbing particles. Care in interpreting OPC response is always required, especially when the optical nature of the particles being measured is not unequivocally identifiable.