Feedstock When the feedstock quality changes, the delta coke will usually change. The important feed properties are its com- position and its boiling range. The three families of hydrocar- bons are paraffins, naphthenes, and aromatics. These tend to increase the coke laydown on the catalyst inversely pro- portional to their hydrogen content. Paraffins produce low delta cokes, while polynuclear aromatics (PNAs) are coke precursors. The single-ring aromatics end up in gasoline, and the two-ring aromatics go to diesel. The larger aromat- ics can undergo alkylation and isomerisation reactions that lead to condensation reactions and increased delta coke. Delta coke will normally change with feedstocks. This is more likely to happen when changing crude oils, which is common in large refineries. Stratification of feeds can occur in tanks if they are not circulated. Both the quality of the feed can change, and the concentration of contaminants like sodium can increase by a factor of 10 from the top to the bottom of the tank. When other streams make up part of the feed going to the FCC unit, such as coker gasoil, feed quality is going to vary due to the changes in the coker gas- oil composition and the amount processed. Even if a feed pretreater is in front of the cracker, the feed to the unit will not be uniform. The boiling range can greatly alter the delta coke. When some vacuum resid is in the feed, at least a portion of the PNAs will lay down as coke, and the nitrogen content will increase. Many nitrogen compounds are basic and form bonds with the acidic sites of the catalyst, reducing its activity and selectivity. More coke is formed since they do not desorb from the acid sites as quickly as other hydro- carbons. Conradson Carbon (Concarbon) is a test used to measure the coking tendencies of the feed. If the residue is mostly aromatics, the increase in delta coke will be the weight per cent Concarbon divided by the catalyst-to-oil ratio. If the residue is mostly paraffinic, then only a portion (possibly less than half) of the Conradson Carbon goes to coke. The coke-making tendencies of the feed can increase if the vacuum tower operation deteriorates. There are a num- ber of reasons this can happen, but pushing the distillation column beyond its design limits is a frequent cause. The darker colour of the gasoil is a sure sign that the FCC feed has more coke precursors. Gasoline was the main product from early FCC units, and diesel was often included in the feed stream. The front end of the straight-run cat feed does not produce as much delta coke as the rest of the feed. The diesel range (430-650ºF) molecules went through the units largely unconverted when amorphous cracking catalysts were used and would lower the regenerator temperature. They were recycled to increase conversion, and this stream was called light cycle oil (LCO). Zeolite catalysts will readily crack these molecules. Recycling LCO may lower the coke-making tendencies of the feed. However, if this stream is very aromatic (di-aromatics), the recycle may lead to increased bottoms yield and have only a marginal impact on the regenerator temperature. Recycle of decant oil, the material taken from the bottom of the main fractionator, is used to increase the regenerator
temperature in many units. This is common in cracking units with feed hydrotreaters or those that process high API crudes. The high concentration of multi-ring aromatics in this stream always makes it a variable that can be used to increase the delta coke and regenerator temperature. Low regenerator temperatures may leave more coke on the regenerated catalyst or prevent the complete burning of CO to CO 2 . Operating variables Several variables can be tried to reduce the regenerator temperature. At the same coke make, any reduction in delta coke will increase the catalyst-to-oil ratio and drop the regenerator bed temperature. Lowering the feed tempera- ture at the same reactor temperature will increase the coke yield, but the higher cat circulation rate will lower the delta coke and the regenerator bed temperature. This assumes the higher catalyst circulation rate will not reduce the strip- ping efficiency even though the residence of the catalyst in the stripper declines. If a feed containing resid is run, a reduction in reactor temperature will drop the stripper temperature, and the desorption of heavy molecules will be reduced. Regenerator temperature may increase. At some point, the regenerator temperature will no longer drop when the feed temperature is lowered because the viscosity of the feed will increase and atomisation will decline. Larger oil droplets take longer to vaporise, and delta coke will increase and push up the regenerator temperature. Increasing the stripping steam may bring down the regenerator temperature by reducing the hydrogen in coke. The actual weight per cent coke may increase slightly because more carbon is being burned in place of the hydro- gen. Since the heat of combustion is reduced, the regener - ator bed temperature will drop. It is a common practice for a refiner to adjust the stripping steam rate by lowering it until the regenerator temperature starts to rise. The rate may then be increased by about 10% to ensure good stripping. If large changes are made to the FCC operation, another check on the steam rate may be in order. Hydrogen in coke data shows a lot of scatter, but closer examination may show correlations with feed proper- ties and/or operating conditions, such as the steam rate and stripping temperature. Pressure is also important since high pressure inhibits the vaporisation of absorbed hydrocarbons. Dispersion steam can affect the feed vaporisation. It reduces the hydrocarbon partial pressure, and the feed will vaporise faster and more completely. Vacuum gasoils will normally vaporise in the riser reactor. If the oil droplets are very large or feed-catalyst contacting is poor, liquid can make it through the riser and cause problems. More dispersion steam might help in this situation. If heavier molecules (BP> 1,100ºF+) are present in the feed, a sub- stantial amount of them may not vaporise, and the liquid on both the catalyst and equipment surfaces can form coke. Catalytic cracking is supposed to be a vapour-solid reaction scheme. A liquid-solid system would inhibit con- version due to the poorer diffusivity of the reactants. More
58
Catalysis 2024
www.digitalrefining.com
Powered by FlippingBook