OCTANE CATALYST EFFECTS ON FCC UNIT HEAT BALANCE
Leuenberger, Executive Tech. Service Engineer
Linda J. Wilbert, Technical Service Engineer
Edison, New Jersey
Presented at the
March 29-31, 1987
San Antonio, Texas
An FCC catalyst that increases cracked gasoline octane by 2 research numbers can be expected to increase the endothermic heat of cracking by 50 BTU per pound of feed and to reduce the catalytic coke making tendency by 10 relative percent. Commercial and laboratory data indicate that the heat of cracking increase is due to the "recracking" of gasoline to C4 and lighter gas as well as the minimization of hydrogen transfer reactions. The coke reduction is also attributed to lower hydrogen transfer. These heat balance effects of octane catalysts can reduce regenerator temperature by 30°F in a commercial FCC unit.
Introduction and Theory
Catalytic cracking reactions are endothermic; they create products with higher heat contents than the reactants and they absorb heat from the environment. In the cracking of paraffins by the beta scission mechanism, a high molecular weight paraffin is cracked to form a lower molecular weight olefin and a paraffin. Table I uses the cracking of normal decane and normal heptane as examples of this beta scission mechanism. Lower molecular weight hydrocarbons and higher molecular weight hydrocarbons both require approximately equal BTU's of heat to crack a carbon-carbon bond by beta scission, but the energy on a per pound basis increases as the molecular weight of the feed decreases. Thus Table I shows a higher endothermic heat of cracking for heptane than for decanes.
Rare earth exchange in a zeolite catalyst promotes hydrogen transfer reactions in competition with beta scission. One effect of hydrogen transfer is to limit the production of C3 and C4 gases by beta scission. When the number of cracking reactions that form light gases is inhibited, the endothermic heat of reaction is reduced.
The hydrogen transfer reactions promoted by rare earth exchange also tend to reduce the endothermic heat of cracking because they are exothermic. Table II gives examples of two such exothermic hydrogen transfer reactions. In the first example, a napthene and lower molecular weight olefins react to form an aromatic and light paraffins. When the olefins saturated by hydrogen transfer are in the gasoline boiling range, this reaction is responsible for reducing gasoline octane. The second reaction shows how hydrogen transfer can form a carbonaceous deposit on the catalyst from a heavy aromatic. When this type of hydrogen transfer is eliminated, coke make is reduced.
FCC octane catalysts maximize the octane of the cracked gasoline by minimizing the hydrogen transfer reactions that saturate gasoline olefins. Mauleon(Ref. 1) et al observed that reducing hydrogen transfer must also increase the endothermic heat of cracking. The results of his work, which are presented in Table III, show that heat of cracking can be correlated with catalyst rare earth content. Rajagopalan and Peters(Ref. 2) observed that reduced hydrogen transfer reduces coke make. Table IV was extracted from their work to show the correlation between coke make and catalyst rare earth exchange.
In this report, we further examine the heat of cracking and coke make differences between rare earth exchanged catalysts and FCC octane catalysts. Our objective is to quantify the heat balance effects and to use the resulting correlation to model catalyst changes in commercial FCCU's.
Methods of Determining Heats of Cracking
Two different approaches to determining the heat of cracking are possible:
1. Determine the heat absorbed by the cracking reaction through heat balance methods.
2. Analyze the reaction products and assign each one a heat of combustion, then add up the heat of combustion of the products. For constant feedstock, the reaction with the highest product heat of combustion has the highest endothermic heat of reaction.
The first method was not attempted for laboratory heat of cracking measurements, but was found useful for commercial data analysis. The second method was found to be applicable for both laboratory and commercial data analysis.
The first technique, which we call the heat balance technique for determining heats of cracking, involves calculating the heat of combustion of coke in a commercial FCC operation, then subtracting all other heat requirements. The remaining heat is then assumed to be the endothermic heat of cracking. The accuracy of this measurement technique depends on the accuracy of the measurements that determine the coke heat of combustion and the completeness of the heat balance information recorded at the commercial plant.
The second technique, which we call the product analysis technique, was applied by Dart and Oblad(Ref. 3) in their classic work on measuring heats of cracking. Each product was assigned a heat of combustion from the literature. For the liquid products, a correlation from the API data book(Ref. 4) was used to determine the heat of cracking from the API gravity and the K factor. This was a modification of Dart's procedure, where these liquid heats of combustion were determined by calorimetry. The heats of combustion used in the product analysis technique are presented in Table V.
The procedure described in the preceding paragraph is also applicable to commercial data and was used to confirm the heat balance heat of reaction calculations.
Table VI summarizes the heat of cracking and delta coke results calculated from catalysts with identical matrices and different levels of rare earth exchange. Rare earth exchanged gasoline catalysts are labeled as R1 and R2. Hydrogen exchanged octane catalysts are labeled as H1, H2, and H3. The average heat of reaction increase in Table VI is 30 BTU per pound for each 1 RON increase in octane. The delta cokes were calculated by dividing the coke make by the pilot unit catalyst circulation. The octane catalysts reduced the delta coke by 6 relative percent for each 1 RON gain.
An example of the changes in product distribution that are responsible for the heat of reaction gain with octane catalysts is shown in Table VII. The hydrogen exchanged catalyst shows a lower gasoline yield and higher C3 and C4 yields. This shift is due to the "recracking" of gasoline by beta scission. The recracking also increases the heat of reaction because it is highly endothermic. Table V shows recracking a pound of gasoline to C3's increases the heat of combustion of the product by 1000 BTU's, so that a 2 weight percent loss in gasoline to LPG increases the heat of cracking by 20 BTU per pound. The remaining increase in the endothermic heat of cracking is caused by reducing the exothermic secondary hydrogen transfer reactions that form aromatics and coke.
The lower delta coke make in Table VII is due to the increased catalyst circulation more than the change in coke yield. Hydrogen exchanged catalysts usually hold lower MAT activities, so they require increased circulation to maintain constant conversion. Laboratory coke makes were usually the same for both rare earth and hydrogen exchanged catalysts, but the octane catalyst had the lower delta coke because its circulation was higher.
The four commercial octane catalyst trials summarized in Table VIII demonstrated agreement between the two methods of calculating the heat cracking difference to within 20 BTU per pound of feed. The examples show the effects of changing from three rare earth catalysts, R3, R4, and R5, to two partially hydrogen exchanged octane catalysts P1 and P2. Also included is a change from hydrogen exchanged octane catalyst H4 to a more effective octane catalyst H5. The catalysts used in these commercial trials were often made by different manufacturers, and thus their zeolite contents and matrix surface areas varied. Due to these manufacturing differences, the changes in octane and heat balance properties could not be predicted from the catalyst rare earth content. However, the changes in heat of cracking and delta coke make could be estimated from the octane difference. Both the 24 BTU per pound increase in heat of cracking and the 6 relative percent decline in delta coke per 1 number research octane increase agree closely with the results from laboratory testing.
Significance of Heat Cracking and Delta Coke Differences on Commercial FCCU Operation
When a catalyst change increases the heat of cracking in a commercial FCCU, the unit must either increase its heat generation or reduce its heat requirements to stay in heat balance. Since the catalyst changeout usually occurs over a period of several weeks, the changes will be gradual and the unit will have sufficient time to respond without a crisis. A unit with normal slide valve controls will respond to the increase in heat requirement by gradually increasing catalyst circulation. According to one published model(Ref. 5) the coke yield will increase proportional to the catalyst circulation raised to the .65 exponent. Thus higher circulation will increase the coke burned in the regenerator and bring the unit back into heat balance. If the unit does not have the air blower capacity to burn more coke, changes must be made in the unit operation to reduce the heat requirements. The simplified overall heat balance in Table IX lists these other requirements as the heat required to increase the feed temperature to the riser outlet temperature, the heat to vaporize the feed, and the sensible heat of the air. When no more coke can be burned, it is most economical to raise the feed preheat in order to reduce the feed sensible heat requirement. The last resort would be to lower the feed rate to reduce all three feed enthalpy terms in the heat balance. The air heat requirement cannot be lowered if the unit is at a coke burning limit.
The reduction in delta coke make caused by introducing an octane catalyst will also create an imbalance in the overall heat balance. Again the standard FCCU control system would raise catalyst circulation to increase the coke make back to its original level. This control strategy for handling lower delta coke will fail only if the catalyst circulation is at its maximum. Alternate strategies to increase coke make include introducing a feed to the FCCU with a higher carbon residue such as vacuum resid or slurry recycle. spraying torch oil into the regenerator can be used as a last resort.
The increases in catalyst circulation caused by lowering the delta coke and raising the heat of cracking tend to reduce the regenerator temperature. The simplified regenerator heat balance in Table X shows that the heat transferred to the reactor plus the air sensible heat is equal to the heat released by burning coke. The coke burned in the table is represented by the previously sited model(Ref. 5).
When the heat balance is solved for the regenerator temperature minus the reactor temperature and the air sensible heat is ignored, this temperature difference is inversely proportional to the catalyst circulation raised to the .35 exponent. At constant reactor temperature, the heat balance then requires the regenerator to cool as the catalyst circulation increases.
More complete versions of the simplified coke make kinetics and heat balances discussed above have been incorporated into an FCCU computer simulation program. This program was used to model the effects of a 50 BTU per pound increase in the heat of cracking and a 10 relative percent drop in delta coke make on a commercial FCCU. The results of one simulation are presented in Table XI. According to the model, if enough octane catalyst is added to the unit to maintain catalyst activity, then a 20°F loss in regenerator temperature, a 15 relative percent increase in cat to oil, a 7 relative percent increase in coke make, and a 2 volume percent increase in conversion would be expected. Correlations show this catalyst would also increase gasoline octane by 2 RON. A second projection has also been included where only enough octane catalyst is used to hold conversion constant. This case is more usual for commercial octane catalyst trials, since the octane catalysts often do not hold activity in the unit as well as rare earth exchanged catalysts. For this alternate case, a 35°F drop in regenerator temperature and a 20 relative percent increase in cat to oil ratio are predicted.
Table XII shows the heat balance results of an octane catalyst trial that parallels the simulation model's predictions. When catalyst P1 replaced a rare earth exchanged gasoline catalyst, the regenerator temperature equilibrated 70°F lower. The cracked gasoline octane increased by 2.5 RON. The octane catalyst was not able to hold MAT activity; it dropped by 6 numbers. Even with that activity loss, the conversion loss was minimized because of a 24 percent increase in cat to oil ratio. The refiner could have held conversion constant if he had the coke burning capacity; but he was forced to increase the feed preheat to keep the coke make constant.
Octane catalysts increase the endothermic heat of cracking by 25 BTU per pound of feed and decrease delta coke make by 5 relative percent for each 1 RON increase in octane they achieve. These results are due to the reduction of hydrogen transfer reactions when octane catalysts replace rare earth exchanged catalysts. Besides the octane gain, the benefits of using octane catalysts include lower regenerator temperatures and higher cat to oil ratios. These changes are especially beneficial when a refiner wishes to process more of a high carbon residue feedstock, such as resid.
The main disadvantage of octane catalysts is the recracking of gasoline to wet gas. This recracking reaction increases when octane catalysts are used because hydrogen exchange does not promote the hydrogen transfer reactions that compete with recracking. The higher heat of cracking characteristic of octane catalysts can also be a disadvantage when a unit without the ability to increase feed preheat is at a coke burning limit.
1. J. L. Mauleon and J.C. Courcelle, Oil and Gas Journal, October 21, 1985, page 64
2. K. Rajagopalan and A.W. Peters, Preprint of the ACS Division of Petroleum Chemistry, Vol. 30, No. 3, page 538 (1985)
3. J.C. Dart and A.G. Oblad, Chemical Engineering Progress, Vol. 45, page 110 to 118, (1949)
4. Technical Data Book - Petroleum Refining, American Petroleum Institute Division of Refining, 1271 Avenue of the Americas, New York, NY, page 14-5 (1966)
5. E.G. Wollaston, W.J. Haflin, W.D. Ford, and G.J. D'Souza, Oil and Gas Journal, Vol. 87, September 22 (1975)