Light-gas selectivities

Calcined AFS and USY zeolites show significant differences in both selectivity ratios whereas steamed zeolites show similar light gas selectivities. However, steamed selectivities are dramatically different from those of calcined zeolites. These results are in good qualitative agreement with results obtained for AFS and USY zeolites by gas-oil cracking (17). 

Table III gives a range of the possible feedstocks that can be used to produce ethylene and the kinds and amounts of by-products that can be made from them. For our purposes we have selected a constant basis of 1 billion lbs/year ethylene production. The feedstocks illustrated in Table III include ethane, propane, n-butane, a full range naphtha, a light gas oil, and a heavy gas oil. The yields reflect high severity conditions with recycle cracking of ethane in all cases. For propane feed, propane recycle cracking has been included as well.

Selectivity results at constant 50% conversion are reported in Tables VI and VII for calcined and steamed zeolites, respectively. Product selectivities are divided into light gas (C1-C4), gasoline (C5-C12) and coke. The gasoline fraction is further divided into paraffin, olefin, naphthene and aromatic (PONA) components. 

Depending upon the needs of the refiner he may be limited by any one of these at various times as he moves from a maximum octane operation to a maximum gasoline operation or switches from a light gas oil to a heavier gas oil feed. The selection of the proper catalyst provides flexibility to address these requirements in a manner otherwise unavailable. 

The nitrogen compound in the light gas oil was classified with basic nitrogen compound and non-basic compound. For the practical condition, quinoline was used as basic nitrogen compound and carbazole was as non-basic nitrogen compound. And normal hexane was selected as model light gas oil. And, silica was used for the adsorbent of nitrogen compounds. 

Ahhou platinum alone or on a variety of neutral supports selectively converts n-hexane to benzene most of these catalysts deactivate rapidly due to coke formation. With the neutral zeolite KL as a support, however, much longer on-stream times are feasible and within a few years of Bernard s origind publication [130] 4e Aromax process had been developed by Chevron [131], Table 6 corrqrares the aromatic selectivity obtained with Pt-Ba KL and Pt Re Sn / AI2O3 - Cl reforming catalysts [58]. Associated with the much hi er aromatic selectivity is a lower amount of light gas production. 

Cracking of decalin, centane, cumene, light gas oil, heavy oil Al-R Good olefin production. Activity and stability > USY, resulting in high selectivity to isobuytlene and isoamylene. 48... 

When the molecular weight of high carbon number paraffins is to be lowered by hydrocracking, with a minimum of light gas production, there also are encountered two types of selectivity problems. Catalyst acidity and operating conditions are now to be chosen (see scheme XIV) to maximize the Y2 reactions as leading to the desired products, which generally implies that the Yi reactions are still faster (see Fig. 14) and tend to push the isomer distribution toward equilibrium. 

To this point the temperature required for 47% conversion has been described as a measure of catalyst "activity". However, since this was taken after 7 days on-line it is evident that it also contains information on catalyst lifetime (or deactivation). NU-87 is a 2-D channel system while EU-1 contains uni-dimenional channels. Hence it may be expected that there would be differences in deactivation rates. A separate, but related, topic concerns selectivity. The TDP reaction also generates light gas and C9+ aromatics ("heavy ends"). As with "activity" it may be expected that the difference in pore structure would be reflected in differences in the selectivity. Indeed there is evidence for differences in the xylenes and benzene balance in NU-85 samples (compared to EU-1). Both topics (deactivation and selectivity) are complex and beyond the scope of this contribution and will be addressed in a separate paper. 

Why is it that Knudsen diffusion typically cannot provide high selectivity for light gas separations ... 

---