ADVANCED FCC CATALYST MATRIX TECHNOLOGY FOR REDUCED COKE AND SLURRY YIELDS
Group Leader, Petroleum Catalyst
Charles F. Keweshan, Development Chemist
Rostam J. Madon, Research Associate
Iselin, New Jersey
Joseph B. McLean,
Executive Technology Specialist
Presented at the
March 20-22, 1994
San Antonio, Texas
BASF's new Reduxion line of FCC catalysts have been developed to offer 10 to 20% lower coke versus current catalysts, while maintaining high bottoms upgrading selectivity. Aspects of BASF's PyroChem zeolite technology featured in the Precision Catalyst line have been combined with a controlled matrix acidity distribution which maintains sites selective for bottoms cracking while reducing the strong sites which lead to coke and gas formation. Results from pilot unit testing are presented and projections for commercial operations are provided.
With the ever increasing demand on refiners to process heavier crudes and maximize gasoline and diesel yields, fluid catalytic crackers are continually pushed to the edge of their operating limits. Maximum gasoline plus diesel yields are often achieved through the use of high bottoms upgrading FCC catalysts that feature active matrices. However, conventional bottoms upgrading catalysts increase the yield of undesirable products such as dry gas and coke. This often pushes the FCCU up against gas compressor or metallurgical limits, not allowing the refiner to take full advantage of the increased gasoline and diesel yield potential.
Current catalysts attempt to address these operating limits by minimizing the activity of the matrix. However, this drastically reduces the bottoms upgrading capability and falls far short of the maximum conversion goals. To impart bottoms upgrading performance, many catalysts utilize high alumina matrices which are stable to deactivation conditions(Ref. 1). These matrices effectively pre-crack large feed molecules to small enough fragments so they can enter zeolite cages where they can further crack more selectively to desirable products. However, there has traditionally been some tradeoff in that the cracked molecules can continue to crack to undesirable end products such as hydrogen and coke.
BASF has previously been successful incorporating alumina in the matrix to upgrade bottoms, but with improved selectivity to minimize coke and gas formation(Ref. 2, 3). Taking this technology a step further, we have found a unique method to control matrix activity through an engineered matrix acidity distribution which effectively pre-cracks large feed molecules, but minimizes secondary cracking to coke and gas. The new matrix essentially shifts the secondary cracking burden to the more selective zeolite.
By combining this new matrix with BASF's proprietary PyroChem zeolite technology, the Reduxion series of FCC catalysts delivers slurry yields equivalent to the best bottoms upgrading catalysts, but with substantially reduced coke and hydrogen yields. PyroChem technology has been demonstrated in the industry in BASF's Precision series of FCC catalysts, currently in use in over 20 commercial FCC units. PyroChem zeolite ultrastabilization technology reduces the formation of defects compared to conventional stabilization and dealumination processes. PyroChem zeolites have enhanced zeolite stability and accessibility resulting in decreased catalytic coke and gas formation and increased gasoline selectivity. The proprietary Reduxion process makes it possible to combine the new matrix technology with the PyroChem process.
Reduxion Matrix Technology
The various reactions that take place during fluid catalytic cracking depend primarily on the distribution of acid sites on the matrix and within the Y zeolite. Bronsted sites are well known as being the important sites for the cracking process(Ref. 4, 7) and most of these sites reside within the zeolite pore structure. However, cracking of very large molecules, i.e., bottoms upgrading, cannot take place inside the small channels of the Y-zeolite but rather occur on the external surface of the zeolite and on the acid sites of the amorphous matrix. Large paraffinic and naphthenic compounds are relatively easier to crack than small paraffins. Thus weak Bronsted acid sites on the matrix are sufficient to crack large hydrocarbons. The initiation reaction to form carbenium ions from these large molecules is not known with certainty: electron acceptor sites(Ref. 8) and Lewis acid sites(Ref. 9) have been proposed to be important as well as the formation of penta-coordinated carbonium ions on Bronsted acid sites(Ref. 10, 11). However, the presence of Bronsted acid sites accessible to large molecules is essential since once the carbenium ion forms, cracking reactions can take place via beta scission and carbenium ions can be replenished via hydride ion transfer.
Lewis acid sites play an important role in the formation of coke(Ref. 12) and in initiating the cracking process(Ref. 8, 9). Coke formation reactions are quite different from scission reactions since coke requires bond making rather than bond breaking. Strong Lewis acid sites in particular can adsorb unsaturated hydrocarbons for sufficient time for oligomers to form which then end up as coke and hydrogen. Weak Lewis acid sites can help contribute to the initiation of the cracking process, but have a reduced tendency to oligomerize unsaturated hydrocarbons to coke. A matrix with reduced strong Lewis acidity in conjunction with Bronsted acidity will be more selective in precracking large hydrocarbons without coke formation. Such precracked molecules are then small enough to enter the zeolite cages for more selective cracking on Bronsted acid sites.
Reduxion catalyst's new matrix maintains sufficient Bronsted acidity and weak Lewis acidity, but with reduced strong Lewis acidity. The net result is a high bottoms upgrading catalyst with reduced coke and hydrogen yields.
Acid sites and their relative strengths are often measured via infra-red (IR) spectroscopy using pyridine as the adsorbed base. In this evaluation, all spectra were collected at 40°C in the diffuse reflectance mode using a Spectra Tech controlled environment chamber in a Perkin-Elmer 1750 IR spectrometer. Band areas at 1450 and 1550 (1/cm) were used to estimate the amounts of pyridine adsorbed on Lewis and Bronsted sites respectively(Ref. 13) after purges at 200°C and 450°C. Measurement after the lower temperature purge corresponds to total acidity whereas bands measured after the higher temperature purge reflect strong acid sites. The weak sites are determined by difference. The Bronsted component of the matrix is difficult to quantify because of the overwhelming Bronsted character of the zeolite. Conversely, most of the Lewis acidity is associated with the matrix and effectively quantified by this technique.
To simulate FCCU catalyst deactivation, fresh FCC catalysts were steamed at 1450°F for four hours in a 100% steam atmosphere, followed by IR analysis. Figure 1 shows the relative weak and strong Lewis acid site density for five commercial FCC catalysts. The Precision catalyst is well established as an effective bottoms upgrading catalyst. Dimension and competitive catalysts A and D are low coke selective catalysts attributable to the low total Lewis acidity. Reduxion has about half of the total Lewis acidity and 4 times fewer number of strong Lewis acid sites than Precision. This imparts a lower coke selectivity while maintaining effective bottoms upgrading.
A variety of gas oils were employed in order to accurately determine the catalytic performance of Reduxion. BASF's standard CTSGO-175 oil is a light, clean oil and affords the benefit of a large internal database. However, the absolute coke yields with this oil are relatively low compared to feeds commonly used in today's refineries. Therefore, heavier gas-oil samples were also utilized in this evaluation, which are more representative of feeds used in current refinery applications. These oils have low Ramsbottom Carbon and basic nitrogen content but higher boiling point distributions. Sulfur content for all of the heavier feeds were similar to each other but roughly four times that of the CTSGO-175 oil. All of the feeds chosen have low metals content. A summary of the gas-oil characteristics used in the study is shown in Table I.
Microactivity Test Evaluation
Catalyst selectivity was determined using conventional microactivity testing (MAT). Relative MAT activities (RMA) were determined on steam deactivated samples (1500°F/4 hours/100% steam) at 910°F at catalyst/oil ratio (C/0) = 5, using CTSGO-175 gas oil and referenced against an internal standard. MAT selectivity data was collected on steam deactivated samples (1450°F/4 hours/100% steam) and selectivity dependence on activity was determined by varying C/0, usually in the range from 3 to 7.
Coke performance was determined on a range of feeds as described above. Table II shows the selectivities for a C/0 comparison of Reduxion and Precision regressed to 70% conversion. This data indicates approximately 11% lower coke relative to Precision, while maintaining comparable bottoms, LC0, and gasoline, and only a slight increase in LPG. In a similar study, using the heavier CTSG0-2255 oil (Table III), the coke benefit is more pronounced at 13%, while again maintaining bottoms upgrading, gasoline and LC0 yields. Additionally, a marked decrease (>50%), is seen in the hydrogen yield for Reduxion. Benefits of the reduced Reduxion coke become more dramatic when the selectivities are calculated at constant coke yield. For the CTSGO-175 oil, calculated to 3% coke (Table IV), a marked difference in conversion is noted with a 2 point increase in gasoline, while at reduced hydrogen and bottoms for Reduxion. In identical fashion, data on CTSGO-2256 oil regressed to 6% coke (Table V), shows similar increases in gasoline and LPG, with reduced hydrogen and bottoms.
Pilot Unit Testing
Additional testing in BASF's circulating FCCU pilot unit were conducted which more closely models commercial operation. Catalyst samples were prepared to comparable RMA (conditions previously described) to simulate typical commercial fresh catalyst activity. The fresh catalysts were then steam deactivated at 1450°F for 7-8 hours to achieve a standard BASF MAT activity between 65-70% conversion. This was done to allow comparison of catalysts at typical equilibrium catalyst activity.
The BASF circulating pilot unit is a small scale modified Arco unit featuring a vertical lean phase adiabatic riser. The unit is charged with 2500 gms of catalyst and has an oil feedrate of 10 gm/min. The unit operates at I atmosphere of pressure with a riser top temperature of 970°F and regenerator temperature of 1300°F. Unit conversion is varied by changing catalyst to oil ratio.
In addition to Reduxion and Precision, the pilot study included a competitive coke selective catalyst shown earlier to have low matrix acidity. Data obtained from the pilot unit on the heavier feedstock are shown in Figures 2 and 3. The bottoms upgrading performance of Reduxion and Precision are high and comparable. Reduxion demonstrates the lowest coke yield, approaching 15% below the other catalysts.
Performance with Metals
Processing of heavier feedstocks and the drive toward increased usage of resid feedstocks demands that FCC catalysts perform well in the presence of metals. In the laboratory, catalyst samples were metalated by cracking a metals loaded gas oil in cyclic cracking/oxidation cycles using the BASF Fixed Fluidized Bed reactor (FFB) as described previously(Ref. 2). This cracking-on of metals closely simulates the metals deposition on a catalyst which occurs under refinery conditions. MAT data on FFB metalated samples indicated that Reduxion behaved similarly to BASF's Dimension (a recognized metals-tolerant catalyst) with metals at levels of 1000/2000 and 2000/4000 ppm Ni/V. Over that range, both Dimension and Reduxion had 30-35% lower specific coke (coke/activity) and about 1/2 of the specific hydrogen make (hydrogen/activity) of Precision (Figures 4 & 5).
MAT comparison of Reduxion with a competitive metals-tolerant catalyst sample (Catalyst R) shows that Reduxion has similar low specific coke at low metals loading and superior coke performance at higher metals loading (Figure 6) in Mitchell method preparations (impregnation of catalyst with nickel and vanadium naphthanate solution followed by "burnoff" in air and steam deactivation).
Commercial Unit Performance
BASF has manufactured Reduxion at commercial scale and it is currently being trialed in operating FCC units. Based on the laboratory test results presented here, projections for commercial performance were generated using a heat-balanced computer model for three catalyst systems, Reduxion, Precision, and a low matrix catalyst. Table VI depicts a hypothetical example assuming an existing base case of Precision. Projected results using Reduxion and the low matrix catalyst are then compared at constant operating conditions. The lower delta coke properties of Reduxion result in a 20°F drop in regenerator temperature. This leads to a higher cat/oil ratio at constant preheat and reactor temperature, as well as slightly higher equilibrium MAT activity due to reduced hydrothermal deactivation. As a result, higher conversion is achieved at constant coke yield (within constraint of total air rate). As noted, the additional conversion gives increased yields of gasoline and LPG, since a constant rare earth comparison was assumed here. Slurry yield is reduced relative to the base case due to the higher conversion level and equivalent bottoms upgrading selectivity. Total dry gas is slightly lower, and octane is slightly higher.
Also shown is a comparison for a low matrix activity catalyst. The same delta coke and cat/oil changes are projected as for Reduxion, but at the expense of considerable tradeoffs for the low matrix catalyst. Catalyst makeup rate is increased to maintain the base case MAT activity, and higher slurry yield results from the poorer bottoms upgrading selectivity. Additional drops in C4 olefinicity and gasoline octane also result. It may be possible to offset these drops by lowering rare earth for this case, but the additional zeolite required would increase catalyst consumption and/or cost further.
This case study clearly shows the potential advantage for a selective matrix catalyst like Reduxion in an air-constrained case. Comparisons for regenerator temperature and/or wet gas constrained units would show similar advantages in potential conversion and/or throughput as a result of Reduxion's lower delta coke, while minimizing the types of tradeoffs typically associated with low matrix activity catalysts.
Reduxion, Precision, Dimension and PyroChem are tradmarks of BASF Corporation.
1. Otterstedt, et al. J. Applied Cat, (1988).
2. Dight, L., Leskowicz, M., and Deeba, M., "New Matrix Improves FCC Catalyst Selectivity." 1991 NPRA paper AM-91-53.
3. Silverman, L. D., et al. "Matrix Effects in Catalytic Cracking." 1986 NPRA paper AM-86-62.
4. Ward, J. W., J. Catal. 10, 34 (1968).
5. Bolton, A. P., and Bujalski, R. L., J. Catal. 23, 331 (1971).
6. Brower, D. M., in "Chemistry and Chemical Engineering of Catalytic Processes" p. 137 (Ed. R. Prins and G. C. A. Schuit) Sijthoff and Noordhoff, The Netherlands, 1980.
7. Zhao, Y., Bamwenda, G. R., Groten, W. A., and Wojciechowski, B. W., J. Catal. 140, 243 (1993).
8. McVicker, G. B., Kramer, G. M., and Ziemiak, J. J., J. Catal. 83, 286 (1983).
9. Hattori, H., Takahashi, O., Tagaki, M., and Tanabe, K., J. Catal. 68, 132 (1981).
10. Haag, W. D., and Dessau, R. M., in "Proceedings, 8th Int. Cong. on Catal., Berlin," Vol. 2, p. 305, Dechema, Frankfurt-am-main, 1984.
11. Corma, A., Planelles, J., Sanchez., and Tomas, F., J. Catal. 92, 284 (1985).
12. Mizuno, K., Ikeda, M., Imokawa, T., Take, J., and Yoneda, Y., Bull. Chem Soc. Japan, 49, 1788 (1976). 13. Parry, E. P., J. Catal. 2, 371 (1963).