Effect of FCC Catalyst Physical Properties on Particulate Emissions
By Ronald G. McClung
With increasing attention to environmental issues within oil refineries, particulate emissions from the FCCU have received increased scrutiny over the years. This article is the first in a two-part series seeking to bring some quantification of the effects of FCC catalyst physical properties on particulate emissions. This article will deal specifically with the effect of ABD (apparent bulk density) on cyclone separation. The second article in this series will deal with attrition and bulk density effects on opacity.
For the purposes of this article, stack emissions will be assumed controlled by the regenerator cyclone system. Specifically, the measure of changes in losses will be based on secondary and tertiary cyclone operation.
The model for cyclone performance will be that illustrated in Figure 1(Ref. 1), which requires definitions of Dpc and Dp.
The definition of Dpc (the critical particle diameter), is the diameter of the smallest particle that is theoretically separated from the gas-solids mixture with a 50% efficiency. Dpc is defined in terms of gas, solids physical properties and cyclone geometry in Equation 1(Ref. 1).
Dp is the particle diameter of a selected fraction of solids in the cyclone inlet solid/gas mixture.
For purposes of this publication, Equation 1 can be expressed only as a function of , since is much less than . A fixed cyclone geometry and inlet gas volume is also assumed.
This expression implies that the more dense the catalyst particle, the lower the critical diameter becomes...and therefore the higher the cyclone recovery of the FCC catalyst back into the unit inventory. This formula does not compensate for less entrainment of solids into the cyclone inlet with higher density catalyst. In addition, there is no compensation for attrition resistance in this expression, even though, generally the higher density equilibrium catalyst (E-CAT), implies a more dense fresh catalyst and a more attrition resistant catalyst inventory. As a consequence, the estimates of increased recovery due to using a higher density catalyst will be somewhat understated.
The particle density used in Equation 2 is a direct function of ABD, and the void fraction ( ) between particles in that measurement. The voidage will depend on the laboratory measurement method use. For instance, an ABD (non-tamped) will have a higher voidage that a CBD (tamped or compacted bulk density). However, the particle density will remain the same, therefore, the voidages must be consistent. For purposes of calculations in this article, a voidage ( ) of 0.42 will be used consistent with the E-CAT data ABD's reported in the latter part of this article.
Listed in Table 1 are a range of E-CAT ABD's and their correponding particle density at a 0.42 void fraction
The Dpc's corresponding to these different particle densities is also given in Table 1.
Using an equation derived for Figure 1 and the data of Table 1, a number of cases were generated to quantify the effect of ABD on stack losses. The results are illustrated in Figure 2 for all cyclones stages. The shaded region indicates the range of change estimated for secondary and tertiary cyclone recovery, in many cases representing stack losses.
The information plotted in Figure 2 allows assessment of the differences in cyclone recovery that can be expected as a result of an E-CAT density change. As an example for application of this figure, actual commercial data is given in Table 2 for various E-CAT densities achieved using catalyst from the three U.S. suppliers. For the range of E-CAT densities shown, there is a relative loss reduction in moving from the lower densities of suppliers A and B to the higher density of BASF of approximately 7 relative percent.
Chemical Engineering Progress, "Understand Cyclone Design"; A.K. Coker, December 1993, Pp. 51-55.