This paper addresses the experimental validation of the optimum design of reverse-flow gas-cyclones, obtained through the solution of a numerical non-linear optimization problem, viz. maximizing cyclone collection. The simulation model is based on the predictive properties of a finite diffusivity model, where the particles' turbulent dispersion coefficient is estimated through an empirical correlation between the radial Peclet and Reynolds numbers. The optimizations were formulated with constrains on pressure drop, saltation velocity and geometrical considerations, such that feasible cyclones could always be obtained.
The optimum geometry, named RS_VHE, is different from available high-efficiency designs, and represents reverse-flow cyclones with a predicted significantly improved performance. An innovative partial recirculation system within a collector-first arrangement further reduces emissions with only a moderate increase in pressure drop. The generally observed unexpected high collection of submicron particles is attributed to capture by larger particles in the turbulent flow field due to turbulent dispersion, much like what occurs in recirculating fluidized beds.
Results obtained for the RS_VHE cyclones with partial recirculation at laboratory, pilot and industrial scales, for temperatures ranging from 300 to 600 K, gas flow rates from 1 to 10 4 m3/h and inlet loads from 15 to 10 4 mg/m3, show them to be significantly more performing than equivalent diameter HE cyclones or smaller diameter multi-cyclones. under certain circumstances, with recirculation the proposed system shows better performance than on-line pulse-jet bag-filters.
Overall, the results show that the numerically optimized RS_VHE cyclones, when coupled with a partial recirculation system, open the applicability of these simple devices for fine particle collection which is typical of more expensive devices, such as venturis and on-line pulse.jet bag-filters.