โ Scribed by Richar D. M. Felder; Enrique Arce-Medina
No coin nor oath required. For personal study only.
A radiotracer-based method of measuring local particle deposition profiles and reentrainment rates in an electrostatic precipitator has been developed and demonstrated. In tests performed on a pilot-scale precipitator, differences in local deposition profiles obtained with and without a predeposited layer of dust on the plate were observed. Rates of reentrainment due to gas friction (aerodynamic drag) and to particle impaction on the plate were measured separately, and correlated with a semiempirical formula. Both friction and impaction reentrainment rates were found to vary as powers of the difference between the gas velocity and a minimum velocity required for reentrainment. The impaction reentrainment was found to vary as a power of the inlet dust loading.
Electrostatic precipitators are generally designed by analogy with similar operating units. Unfortunately, precisely analogous conditions (gas flow, temperature and pressure, dust loading, dust resistivity, particle size distribution, etc.) in two different process streams are rarely encountered, so that some degree of extrapolation is always required. Several empirical correlations of collection efficiency with various experimental parameters exist (Oglesby and Nichols, 1978), but they are imprecise, and relatively conservative overdesign procedures must therefore be used to arrive at a final precipitator design.
A number of mathematical models of electrostatic precipitators have been developed in recent years. The intent of these models is to enable the prediction of particle collection efficiencies from specified values of feed parameters (air flow rate, dust loading, and physical properties of the dust) and precipitator operating conditions (plate area and spacing, corona wire spacing, applied voltage, temperature and pressure). An accurate model of this nature would serve as an ideal basis for design or scale-up of a precipitator to operate at conditions different from those for which data already exist.
In their present state of development, however, these models can do little more than provide qualitative estimates of collection efficiencies for given sets of conditions. The inadequacies could be due in part to violations of the model assumptions re-
๐ SIMILAR VOLUMES