FCC Unit Optimization Using The MagnaCat Process

by Terry L. Goolsby and Howard F. Moore Ashland Petroleum Company Ashland, Kentucky

Dennis C. Kowalczyk Refining Process Services Cheswick, Pennsylvania and Tiffin E. Johnson, Mario L. Zampieri, and B. Karl Bussey The M. W. Kellogg Technology Company Houston, Texas

Introduction The refining industry has experienced numerous changes in markets and refining operations. Refiners are looking hard at cutting costs, improving FCC operating efficiency and for processes which improve refining margins with the least amount of capital investment. Fluid catalytic cracking (FCC) in particular has a significant bottom line impact on both refining revenue generation and cost. FCC costs include fresh catalyst purchases as well as spent catalyst disposal which can be significant at times. Both of these can be offset to a degree by utilization of a new process called MagnaCat. A series of previously published technical articles 1-5 and patents 6-8 describes this technique for using high field strength permanent magnets to continuously remove the older, higher metals laden, less active fluid cracking catalyst from the equilibrium FCC inventory by dry magnetic separation techniques to produce a catalyst with lower metals and higher activity and selectivity. The first commercial application of this technology has been successfully operating since April of 1996 at Ashland Petroleum's Canton, Ohio refinery. This paper describes the implementation of the MagnaCat Process at the Canton refinery for upgrading FCC catalyst on a continuous basis. Since start-up, substantial improvements have been observed in the conversion level and product profile of the Canton FCC unit yielding significant economic benefits. The final section of the paper describes the design of the MagnaCat process unit developed for commercial licensing.

Page 1

Benefits of the MagnaCat Process As FCC catalyst particles age in the unit, they accumulate contaminant metals and become more susceptible to a high strength magnet. The unit catalyst inventory contains a continuous spectrum of catalyst particles ranging from very old catalyst with high metals and low activity through day-old fresh catalyst, Figure 19. As metals are deposited on the catalyst over a period of time, Figure 2, the magnetic susceptibility of the catalyst particles increases2. The MagnaCat Process provides a mechanism for separation of the older, more metals-laden particles in the FCC catalyst inventory from the newer, more active and more selective particles. Instead of discarding a full equilibrium catalyst mixture, the refiner can preferentially discard the poorest performing FCC catalyst particles and return the freshest, most active particles back to the unit. Figure 1

Distribution of Catalyst Activity as a Function of Age

Microactivity

Estimated Days in Unit

Figure 2

Page 2

Metals Distribution Correlates with Magnetic Susceptibility

5 4 3 2 1 0 0 10 20 30 Nickel 40 Vanadium 50 60 70 Magnetic Susceptibility Metals Content, ppm (Thousands)

It should be no surprise that FCC catalyst activity and selectivity is degraded as the particles age. As catalyst particles get older, they produce less gasoline and more coke and hydrogen. First, as zeolite deactivates, the crystallinity deteriorates and more nonframework alumina is formed. This leads to less selective cracking and the production of more coke and gas. Second, as contaminant metals concentrate on the catalyst surface, they catalyze undesirable dehydrogenation reactions and zeolite deactivation. With MagnaCat, the FCC catalyst particles removed from the circulating inventory will be those containing the highest level of metals contamination. This results in a reduction in the level of contaminant metals in the equilibrium catalyst inventory, if catalyst addition rate is held constant. It should be noted, however, that FCC metals reduction is not the primary objective of MagnaCat processing. The intended benefit of implementing MagnaCat is to upgrade the entire FCC catalyst inventory via removal of the oldest catalyst particles. This results in a significant upgrade in catalyst activity and selectivity performance, while at the same time achieves a moderate reduction in contaminant metals level. The MagnaCat process provides a safe and clean method of FCC catalyst management. The process requires very little power since the magnetic separation can be achieved with a permanent magnet, not an electromagnet. No chemicals are involved in the process and no wastewater is generated.

Page 3

The MagnaCat process can be employed in essentially two operating modes. The recommended mode of operation involves implementation of MagnaCat while holding the fresh catalyst addition rate constant. This will result in a moderate reduction in the metals on equilibrium catalyst, while at the same time significantly upgrading the catalytic properties of the inventory. Refiners can take advantage of the improved activity and selectivity through: Increasing unit throughput Increasing unit conversion Running poorer quality feedstock The second mode of operation involves reducing fresh catalyst addition rate until either constant metals level or catalyst activity is achieved. This would result primarily in a reduction in catalyst usage and disposal costs. All other operating modes are intermediate to these two modes. History of MagnaCat Process Development Magnetic separation techniques have been used for many years in the mining, food and other industries. These techniques include the utilization of electromagnets and permanent magnets for separation of magnetic from non magnetic material on a wet or dry basis. Electromagnets use electricity to induce a magnetic field in a metallic object by flowing electrons through a wire round core to induce a magnetic field in a metal object in the center of the core. These types of magnets are relatively expensive and operating costs are usually high due to the consumption of electricity. Permanent magnets are generally used in operations where the material being removed exhibits strong ferromagnetic and/or paramagnetic properties. Ashland and Nippon Oil independently applied High Gradient Magnetic Separation (HGMS) technology to cracking catalysts in the late 70's and early 80's and were granted patents, and Nippon Oil operated an HGMS carousel separator on their FCC unit in Japan for about one year. Due to high capital, power, and land costs, this technology was mothballed. In the late 80's and early 90's, advancements in permanent magnet technology enabled magnetic separation techniques to be applied to processes in the Petroleum Industry. Ashland demonstrated the applicability of this less expensive rare earth roll magnet, dry separation technology to cracking catalysts in the late 80's and performed small scale tests in 1991 and 19922. The Canton MagnaCat Project Ashland's Canton refinery FCCU was chosen as the site of the first commercial unit based on 1) the size of the FCC catalyst inventory, 2) the suitability of the catalyst for magnetic separation, and 3) the support of Canton refinery Management. This first MagnaCat was completed in December 1995. The unit's design reflects the accumulation of years of Research, Development, and Engineering.

Page 4

The Canton MagnaCat Process unit was placed close to the FCCU with the majority of the unit contained within a 4 level steel structure. The structure houses the catalyst hopper, filters, catalyst water cooler, magnetic separator, bins, and catalyst transfer hoppers on various levels. In the magnetic separator, the heart of the operation, catalyst is vibrated onto a belt in a thin layer as the belt rotates. The belt surrounds two rollers, one a non magnetic follower while the other magnetic roller is motor driven. This process begins at Canton with a slip stream of FCC equilibrium catalyst flowing from the regenerator through a boiler feed water catalyst cooler. A slide valve regulates the catalyst in a dilute phase transport to the MagnaCat catalyst hopper. The hopper catalyst is slightly fluidized, and flows through a cooling water catalyst cooler to further reduce the temperature of the catalyst from 260F to 100F. The catalyst then transfers into the magnetic separator feed hopper in a dense phase mode, Figure 3. Figure 3

The MagnaCat Separator

Magnet Feeder FCC E-Cat Belt

Vibrator

Lower Metals / High Activity Recycle Catalyst Magnet

Higher Metals Discard Catalyst Collection Chutes

The magnet separates the catalyst into least and most magnetic fractions; a third minor fraction comprised of catalyst clumps and non-catalyst material of greater than 100 mesh is also collected. The fractions then gravity flow into separate catalyst transporters. The

Page 5

least magnetic or recycle is returned to the regenerator while the most magnetic or reject is transferred to the spent hopper along with the >100 mesh size particles. Continuous smooth operation of the MagnaCat Process unit commenced in late March and early April 1996 as shown in Table I.

Table I - Summary of Canton Refinery MagnaCat Process Unit

Tons Magnetic Catalyst Separation FCC Eq. Least Mag Most Mag 63.24 38.7 22.5 167.55 118.4 49.15 149.5 107.65 41.85 224.85 196.9 27.95 233.05 192.25 40.8 838 118 654 182 Wt.% Most Mag 36.76 29.33 27.99 12.43 17.51 22

Month March 28 to April 30, 1996 April 30 to May 30, 1996 May 30 to June 30, 1996 June 30 to July 30, 1996 July 30 to August 30, 1996 Total Catalyst Processed Days On Stream from April to August

Wt.% Least Mag 63.24 70.67 72.01 87.57 82.49 78

Ashland Petroleum's Research, Applications, and Development and Engineering Departments evaluated the MagnaCat Process unit from April 1 to August 30, 1996. During that period 838 tons of equilibrium catalyst were processed at an average split of 78 weight percent least magnetic recycle and 22 weight percent most magnetic reject with the amount of most magnetic reject decreasing over the evaluation period. This also coincides with a noticeable decrease in the fresh catalyst addition to the FCC unit as shown in Figure 4. FCCU performance tests were performed on May 1 and July 12, 1996 to document the MagnaCat performance at estimated 40% and 100% catalyst inventory treated. Monitoring of unit magnetic separation performance with Refinery Operations included evaluation of the FCC ECAT, least magnetic recycle and most magnetic reject catalyst fractions to set unit operation parameters. Analyses performed on the FCC ECAT, least magnetic recycle and most magnetic reject catalyst fractions included micro activity testing (MAT), metals (nickel and vanadium), and catalyst physical and chemical properties.

Page 6

Figure 4

FCC Catalyst Optimization With MagnaCat

Average Monthly Fresh Catalyst Addition
Pounds/BBL 0.35

0.3

0.25

0.2 4 10 95 11 12 1 2 3 4 5 96 6 7 8 9

Time

Commercial Performance Results with MagnaCat Ashland personnel performed FCCU test runs on April 12, 1995 and October 13, 1995, to establish a base case prior to start-up of the MagnaCat process. For reasons unrelated to the project, start-up of the MagnaCat process was delayed into the second quarter of 1996. An FCC test run was performed on May 1, 1996, with an estimated 40% of the equilibrium catalyst inventory consisting of particles processed through and recycled from the MagnaCat unit. For the July 12, 1996 test, it was determined that the entire FCCU catalyst inventory consisted of particles recycled from the MagnaCat process. Table II presents four sets of mass-balanced Canton FCCU yields corresponding to the test runs described above. The feed rates for all of the test runs are relatively similar at approximately 23,000 BPD. The conversion increased significantly as the MagnaCat unit came on-line and the entire catalyst inventory had been processed through the magnetic separation loop. The data further indicate that the increase in conversion was accompanied by a proportionate increase in gasoline yield and strong increase in total yield of C3 plus C4 hydrocarbon.

Page 7

Table II - Commercial Canton MagnaCat Results

Case Date % of Inventory Separated Riser Temperature, F Regen Bed Temp, F Feed Temperature, F Feed Rate, MBPD Vacuum Bottoms, BPD H2 Production, SCF/BBL Conversion: Vol% FF Yields, Wt% FF H2 Methane Ethane Ethylene Yields, Vol% FF Propane Propylene Isobutane N-Butane C4 Olefins Gasoline LCO Slurry Coke, Wt.%/FF Base 4/12/95 0 975 1310 632 23340 2000 56 75.4 Base 10/13/95 0 975 1300 598 22980 2500 50 76.4 MagnaCat 5/1/96 40 972 1300 620 23430 1600 47 81.7 MagnaCat 7/12/96 100 972 1285 610 22810 1200 34 82.6

0.1 0.9 0.8 0.8 1.7 7.7 4.3 1.5 8.2 62.5 18.5 6.2 4.5

0.1 0.9 0.7 0.7 0.9 7.7 5.5 1.3 9.9 61.9 17.3 6.3 4.6

0.1 0.8 0.5 0.8 2.9 10.6 6.1 1.4 11.1 63.2 13.7 4.6 4.4

0.1 0.7 0.3 0.7 2.8 8.8 6.9 3.1 9.8 65.9 13.1 4.3 4.1

As is common with commercial data, all of these yield changes are not attributable to MagnaCat. Changes in black oil injection rate and a hydrotreater catalyst change also effect these yields. We have attempted to account for these changes in the following discussion. The improved conversion and liquid yield profile was accompanied by a reduction in coke yield, hydrogen-to-methane ratio and hydrogen yield. This provides further evidence that processing metals-laden FCC catalyst through the MagnaCat unit will remove the oldest particles and reduce the degree to which dehydrogenation reactions occur. In order to determine the impact of feed quality differences on FCC yield structure, FCC feed quality information was measured and is reported in Table III. API gravity was lowest at 26.5 for the feedstock samples collected in April 1995 and July 1996, with

Page 8

higher gravities in October 1995 and May 1996. The two feed samples collected in 1995 contained slightly higher sulfur and more carbon residue than the two feedstocks used for the partial and full MagnaCat test runs.

Table III - Canton FCC Feed Properties

Feed Properties Date of Balance API Gravity UOPK Sulfur, Wt.% Ramsbottom Carbon, Wt.% Nickel, ppm Vanadium,ppm Base Base 04/12/95 10/13/95 26.5 28.8 12.09 12.18 0.2 0.17 1.43 1.34 1 <1 1 <1 MagnaCat 05/01/96 29.3 12.18 0.1 0.98 1 1 MagnaCat 07/12/96 26.5 12.0 0.1 0.7 <1 <1

D1160, Distillation Degree F 5 50 95 HPLC, Wt.% Saturates Mono-Aromatics Di-Aromatics >Di-Aromatics

592 803 1075

563 784 1043

550 770 1055

549 785 1035

65.3 18.6 7.1 7.6

79.2 12.5 2.8 5.0

73.8 14.8 5.5 4.8

68.6 17.3 6.5 6.6

The boiling range for the four feeds is similar, with the exception of the highest boiling materials as evidenced by the differences in the 95% points. The differences in high boiling material correlate well with the measured carbon residue content. The level of saturates in the April 1995 and July 1996 feedstocks indicate lower cracking quality. Based upon this review, varying feedstock quality cannot explain the improved performance of the Canton FCCU which was achieved after full implementation of MagnaCat processing. Differences of this magnitude can only be explained by significant improvement in catalyst properties and performance. Table IV presents the results of an off-line analysis of equilibrium FCC catalyst from the commercial operation. Catalyst performance differences were obtained by Microactivity testing with a constant feed. MAT testing eliminates differences in feed quality and unit operating variables. The results in Table IV show that the FCC catalyst samples collected during the two MagnaCat test runs exhibited significant advantages in the areas of increased gasoline selectivity and lower coke and hydrogen selectivity. These results clearly demonstrate the powerful impact of removing the oldest catalyst particles in a magnetic separation unit.

Page 9

Table IV

FCC Catalyst Optimization With MagnaCat

MagnaCat Improves Equilibrium Catalyst Selectivity
Date % MagnaCat In Inventory MAT Conversion (Vol. %) H2 Yield (Wt.%) H2/C1 Ratio (Wt/Wt) Gasoline Yield (Vol.%) Gasoline Selectivity (V/V) Coke Yield (Wt.%) Coke Selectivity (Coke/2nd Order Conv.) 4/12/95 0 70.7 0.21 0.62 60.9 87 3.4 1.47 10/13/95 0 71.0 0.22 0.56 59.41 85 3. 9 1.67 5/1/96 40 70.7 0.2 0.5 62.05 90 2.7 1.21 7/12/96 100 70.3 0.16 0.38 63.31 91 2.7 1.19

Table V contains analytical property data for each of the equilibrium catalyst samples Table V. FCCU Balance Catalyst Analysis
Catalyst Properties Date of Balance Carbon on Regen, Wt.% Total Surface Area, m2/g Matrix Surface Area, m2/g Pore Volume, cc/g Average Particle Size, Microns Apparent Bulk Density, g/cc Relative Zeolite Metals Analysis Nickel, ppm Vanadium, ppm Iron, ppm Base MagnaCat MagnaCat 04/12/95 10/13/95 05/01/96 07/12/96 0.12 0.14 0.14 0.13 135 144 146 146 72 75 72 70 0.4 0.44 0.42 0.42 89 92 88 88 0.69 0.69 0.67 0.67 5.5 7.2 8.1 8.0 Base

900 2000 8100

1000 2100 7400

1100 1900 7700

1000 1700 7000

Page 10

from the series of FCC test runs. These results show that the metals content of these samples is relatively constant across the series. It is important to note that the fresh catalyst addition rate remained relatively constant in the range of 3.9 to 4.4 TPD between April 1995 and May 1996. Catalyst addition rate was calculated as an average for the thirty days prior to each test run. In May of 1996, it was necessary to reduce fresh catalyst addition at Canton to 2.8 TPD as recovery section constraints were being exceeded. Figure 5 illustrates that the observed trend in gasoline yield for the equilibrium catalyst samples is confirmed by the gasoline yield trend on the commercial unit. Figure 5

FCC Catalyst Optimization With MagnaCat

Gasoline Comparison
67 66 65 64 63 62 61 60 59 58 57 4/95 10/95 5/96 7/96 MAT FCC

Figure 6 illustrates the trend toward reduced commercial FCCU hydrogen make as the full FCC catalyst inventory is upgraded via magnetic separation. The spike in hydrogen yield during the mid-May to early June time period is attributed to increased processing of residual bottoms and not to catalyst properties. Figure 7 illustrates the improvement in gasoline selectivity (vol/vol) as the MagnaCat process is integrated in the Canton FCCU operation.

Page 11

FCC Catalyst Optimization With MagnaCat R

FCC Catalyst Optimization With MagnaCat

Seven Day Average of Daily FCC Data

30-May-96

23-May-96

16-May-96

09-May-96

02-May-96

25-Apr-96

19-Apr-96

13-Apr-96

07-Apr-96

80

60

40

120

100

20

01-Apr-96

90

85

80

75

70

Seven Day Average of Daily FCC Data

31-Aug-96 29-Aug-96 22-Aug-96 15-Aug-96 08-Aug-96 01-Aug-96 25-Jul-96 18-Jul-96 11-Jul-96 04-Jul-96 27-Jun-96 20-Jun-96 13-Jun-96 06-Jun-96 30-May-96 23-May-96 16-May-96 09-May-96 02-May-96 25-Apr-96 19-Apr-96 13-Apr-96 07-Apr-96 01-Apr-96

31-Aug-96

29-Aug-96

22-Aug-96

15-Aug-96

08-Aug-96

01-Aug-96

Hydrogen Make (Daily Data)

25-Jul-96

18-Jul-96

Gasoline Selectivity

11-Jul-96

04-Jul-96

27-Jun-96

Date

Figure 6

Figure 7

13-Jun-96

06-Jun-96

Hydrogen, SCF/BBL

Page 12

20-Jun-96

Figure 8 presents a plot of research octane number (RON) during the MagnaCat implementation period. Although there is fluctuation resulting from frequent feedstock changes, gasoline octane is clearly increased over this time period. Figure 8

FCC Catalyst Optimization With MagnaCat

Effect Upon Octane
95 94 93 92 91 90 04/01/96 11-Apr-96 21-Apr-96 01-May-96 11-May-96 21-May-96 31-May-96 10-Jun-96 20-Jun-96 30-Jun-96 10-Jul-96 20-Jul-96 30-Jul-96 09-Aug-96 19-Aug-96 29-Aug-96 08-Sep-96 18-Sep-96 28-Sep-96

Daily Data

The economic benefit for MagnaCat at Canton consists of a combination of improved FCC product value and FCC catalyst savings. The results indicate that both the partial (40%) and full (100%) MagnaCat cases show significant economic benefits over the conventional FCCU yields measured during 1995. Based upon a comparison of the 1995 and 1996 cases, the economic benefits of MagnaCat are determined to be in the range of $0.30-0.60 per barrel of feed. MagnaCat Process Unit Modular Design The licensed MagnaCat process unit, based on the commercial Canton unit, is designed to be easily integrated into an operating FCC unit. MagnaCats modular concept maximizes installation of the unit on pre-turnaround basis with a minimum amount of tieins and other construction activity required during the turnaround period. The unit has virtually 100% turndown flexibility and can be taken off line for maintenance or at the discretion of the operators while the FCC unit is operating.

Page 13

Modular Capacity and Dimensions The standard MagnaCat process unit with one module has a nominal processing capacity of up to 40 tons per day of equilibrium catalyst. Depending on the magnetic susceptibility of the equilibrium catalyst and the desired operation, the catalyst discard rate from the MagnaCat unit will vary from 4 to 12 tons per day with the remainder recycled to the regenerator. Included in the discard will be pieces of refractory, clumped catalyst, and other non-catalytic material typically found in the circulating catalyst stream of any FCC unit. The presence of these undesirable components in the equilibrium catalyst will appreciably decrease after the MagnaCat unit completes the initial processing of the FCC units entire catalyst inventory. The shop-fabricated module contains the majority of the MagnaCat process equipment and instrumentation, is readily transportable and quickly field installed. The modules structural framework has a footprint of 12 feet (3.7 meters) by 12 feet (3.7 meters) with a height of 40 feet (12.2 meters). A separate stair tower module with a footprint of 12 feet (3.7 meters) by 10 feet (3 meters) and a height of 10.5 feet (3.2 meters) is located adjacent to the main module and provides operating and maintenance access to its first platform level. When plot area is limited, ladder access can be provided in lieu of the stair tower. A three dimensional module layout is shown in Figure 9 for reference.

Modular Process Description Hot equilibrium catalyst withdrawn from the regenerator is cooled by boiler feed water in a proprietary design catalyst cooler, aerated, and then flow regulated by a slide valve. The catalyst cooler and slide valve are typically located a short distance from the regenerator withdrawal nozzle; they can be incorporated into the module design when necessary. Downstream of the slide valve, dry plant air is used to transport the catalyst to the catalyst hopper located above the second platform level of the module. The catalyst hopper serves as a feed buffer for the magnetic separators. Fluidization air supplied to the hopper provides minimal aeration to maintain an even catalyst temperature profile in the hopper and into the magnetic separators. A small bag house mounted on the top of the hopper removes catalyst fines from air discharging to the atmosphere. A choke feed system maintains steady catalyst flow from the catalyst hopper to the two parallel magnetic separators where the equilibrium catalyst is split into two streams. The less magnetic stream, which is higher activity and lower metals, is recycled to the regenerator; the more magnetic, lower activity, higher metals fraction is discarded to the FCC units spent catalyst hopper or an alternate location. A third minor stream consisting of catalyst clumps and non-catalyst material of greater than 100 mesh is also collected and combined with the more magnetic stream for disposal. The magnetic separators are enclosed within easily accessible panels to limit catalyst dust. The magnetic separators,

Page 14

along with process valves, instrumentation, and electrical motors are located on the first platform level. Figure 9

Page 15

The recycle and discard catalyst streams from the separators gravity flow to the recycle and discard transport hoppers located at the modules grade level. A programmable logic controller (PLC) provided with the module controls the operation of the two pneumatic transporters, including sequencing of the associated valves, along with additional module control functions. The PLC can be located on the module or installed in the FCC units control room. A lower capacity MagnaCat process unit capable of processing up to 20 tons per day of equilibrium catalyst is designed with a single magnetic separator on the module. All other equipment for this design is sized to process 40 tons per day of equilibrium catalyst, giving the refiner the flexibility of increasing the MagnaCat units capacity to the maximum of 40 tons per day at minimal cost by simply adding the second magnetic separator. Additional parallel modules are supplied for equilibrium catalyst processing requirements exceeding 40 tons per day. Conclusions The MagnaCat process has been proven to upgrade the FCC catalyst inventory through the magnetic separation and removal of the oldest, most metals-laden, least active and least selective FCC catalyst particles. Removal of these particles leads to improvements in the areas of increased gasoline selectivity as well as lower coke and gas selectivity. Implementation of the MagnaCat process at the Ashland Canton FCC unit has been a clear commercial success on both an economic and technical basis. Based upon a combination of increased FCC product value and FCC catalyst savings, MagnaCat is providing Canton with a $0.30-0.60 per barrel improvement in FCC operating economics. Based upon extensive studies conducted over the past five years, it is clear that MagnaCat can provide similar benefits for other FCC units processing even small amounts of residual material. The commercially licensed MagnaCat process unit developed from the original Canton unit utilizes a modular design concept to minimize installation downtime. The unit does not interfere with FCC converter operation, has virtually 100% turndown flexibility, and consumes minimum utilities. With reasonable initial investment, a short pay back period, low operating expense, and proven technology, the MagnaCat process is a desirable addition to any FCC unit processing a metals contaminated feedstock. Acknowledgments The authors wish to express their appreciation to Ashland Petroleum Company, Refining Process Services, and M. W. Kellogg. We particularly wish to acknowledge the support of the Ashland Research, Applications, and Development and Engineering Departments for their involvement in the development of the MagnaCat technology. Personnel directly involved include Rick L. Good, Danny E. Newman, Steven A. Barnett, Hai Dang, Judith A. Raimondo, David W. Shafer, Larry D. Tullock, John L. Reese, Harrell L. Duff,

Page 16

Gene M. Quodala, Robert J. Campagna, Jack Wilcox, Charles E. Mink, Mark A. Waldie, William Patrie, Robert C. Mihaly, William D. Espenschied, Edward E. Mitchell, Richard M. Hickman, Harold Rocketto, Tiff Johnson, Ray Herbrich, Gordon Chow, Dick Stebel, and Barry Moore. References 1. Kowalczyk, Dennis C., Robert J. Campagna, William P. Hettinger, Jr., and Shinji Takase, "Magnetic Separation Enhances FCC Unit Profitability," NPRA Paper AM 91-51, San Antonio, TX (March 17-19, 1991). 2. Goolsby, Terry L., Howard F. Moore, Maurice M. Mitchell, Jr., Dennis C. Kowalczyk, Warren S. Letzsch, and Robert J. Campagna, "Improved FCC Performance by Catalyst Upgrading with Magnetic Separation," 1993 Spring National AIChE Meeting, paper 64e, Houston, TX (March 1993). 3. Miller, Rik B. and Tiffin E. Johnson, "Advances in Resid FCC Technology," CHEMTECH Triple Expo '94, Bombay, India (November 1994). 4. Johnson, Tiffin E. and Rik B. Miller, "Resid FCC Regenerator Design," M. W. Kellogg Refining Technology Seminar, Houston, TX (February 9-10, 1995). 5. Goolsby, Terry L. and Howard F. Moore, "Development of FCC Catalyst Magnetic Separation," Ninth Symposium on Separation Science and Technology for Energy Applications," Gatlinburg, TN (October 1995). 6. Hettinger, W. P. and R. M. Benslay, "Magnetic Separation of High Activity Catalyst from Low Activity Catalyst," US Patent 4,406,773 (September 27, 1983). 7. Hettinger, W. P., "Magnetic Separation of High Metals Containing Catalysts Into Low, Intermediate, and High Metals and Activity Catalysts." US Patent 5,147,527 (September 15, 1992). 8. Hettinger, W. P. and R. M. Benslay, "Addition of Magnetically Active Moieties for Magnetic Beneficiation of Particulates in Fluid Bed Hydrocarbon Processing," US Patent 5,230,869 (July 27, 1993). 9. Palmer, J. L., and E. B. Cornelius, Applied Catalysis, 35 (1987), p. 217-235.

Page 17