An experimental protocol to evaluate FCC stripper performance in terms of coke yield and composition

Abstract

Tests have been conducted in a microactivity test (MAT) and a fluidized bed reactor to develop an experimental protocol to determine how the yield and composition of coke and the associated catalyst surface area vary as a function of stripper conditions in fluid catalytic cracking (FCC). In both reactors, the use of rapid quenching has allowed the relatively short stripping times encountered in FCC units to be simulated. Low sulphur vacuum gas oils (VGO) with a low metal equilibrium catalyst (E-cat) were used for stripping periods of up to 20 minutes. Significant variations occur in the structure of both hard and soft coke during stripping. Although the hard coke becomes more highly condensed with prolonged stripping, the surface area reduction by the hard coke remains fairly constant for stripping periods in excess of ca. 5-10 minutes and is small (10m2 g-1) in relation to the loss of surface area from the soft coke. The use of about 70 g of catalyst in the fluidized bed provides sufficient sample for demineralization to recover sufficient of the hard coke for 13C NMR analysis after the initial extraction of the soft coke. Indeed, it has been found that a further pool of soft (chloroform soluble) coke is physically entrapped within the catalyst pore structure and is only released after demineralization. In fact, this second soft coke fraction is much more highly aromatic than the first and ultimately controls the final coke yield. For the combination of E-cat and VGOs investigated here, typically about half of the final hard coke content of nearly 1% w/w catalyst is derived from this second soft coke fraction by carbonization. The structural information obtained has been used to formulate a model for the stripping process where the soft coke II fraction undergoes cracking in competition with coke formation and evaporative removal from the catalyst.