A message from the editor...

Transcription

1

2 A message from the editor... Dear Refiner: Synergy has been a major focus of both Grace Davison and Advanced Refining Technologies over the last year. This issue of the Catalagram features an article on our GENESIS catalyst blend, which draw on the performance of both our commercially proven MIDAS and IMPACT catalyst families. GENESIS catalysts have demonstrated their flexibility and superior yields in a number of commercial applications, proving that, when combined, the blended catalysts' performance exceeds that of each of the individual components. Our previous issue featured a lead article on synergy between hydroprocessing and FCC operations, concentrating on the dynamics between hydrotreated cat feed and its effect on FCC catalyst yields. Our leadership position with Grace Davison and Advanced Refining Technologies allows us to explore this area from both the DHT and FCCU angles. Another synergy can be found in ART's SmART TM and ApART TM catalyst systems for ULSD and CFH, respectively. These systems are designed with various components to meet your specific needs for yields, feed, operating severity, hydrogen use and other critical variables. As we continue to explore the synergy between our existing products, we are also committed to customer-driven research and development to deliver the refining catalysts for your next challenges. Explore the synergy of our catalysts and let us optimize your units. We stand ready to partner with you to give you the best operations possible. Regards, Joanne Deady Joanne Deady Vice President FCC Marketing/Research and Development Grace Davison Catalagram 12 Fall 27 1

3 NUMBER 12 Fall 27 IN THIS ISSUE The GENESIS TM Catalyst System Rosann K. Schiller, Denise Farmer, and Larry Langan Grace Davison Refining Technologies 3 The GENESIS TM Catalyst System Taking advantage of the synergy between catalyst technologies provides optimized yields and product selectivities for refiners. INSIDE > Recombination of Technologies > FCC Equilibrium Catalyst Trends > ULSD Catalyst Performance > Sulfur Reduction CATALAGRAM 12 Fall 27 Managing Editor: Joanne Deady Contributors: Wu-Cheng Cheng Denise Farmer Ruizhong Hu David Hunt Larry Langan Marilyn Moncrief Chuck Olsen Rosann K. Schiller Yuying Shu Kelly Stafford Brian Watkins Rick Wormsbecher A. E. Zieber Please address your comments to betsy.mettee@grace.com W. R. Grace & Co.-Conn. 75 Grace Drive Columbia, MD 2144 (41) GSR Products Maintain Excellent Performance When Used with Olefins Additives Wu-Cheng Cheng and Ruizong Hu Grace Davison Refining Technologies The simultaneous and optimized use of Davison Additives allows a refinery to reach its operating and economic objectives. Worldwide FCC Equilibrium Catalyst Trends: A Ten-Year Review Marilyn Moncrief, David Hunt and Kelly Stafford Grace Davison Refining Technologies Catalytic FCC Gasoline Sulfur Reduction: Mechanism of Sulfur Reduction Excerpted from the Journal of Catalysis Yuying Shu and Rick Wormsbecher Grace Davison Refining Technologies 12 This update demonstrates how Ecat activity, contaminants, and other properties have changed, both worldwide and geographically, over the last ten years. It also allows individual refiners to rank their own Ecat properties relative to the industry in several categories. The Effects of Treat Gas H 2 Purity on ULSD Catalyst Performance A. E. Zieber Process Engineer, Chevron, USA, Salt Lake City, UT Refinery Brian Watkins and Chuck Olsen Advanced Refining Technologies Hydrogen purity is a key component in extending run length and extending hydrotreating catalyst life Understanding the mechanism of gasoline sulfur species formation enables scientists to define catalytic solutions for gasoline sulfur reduction. 27 W. R. Grace & Co.-Conn. The information presented herein is derived from our testing and experience. It is offered, free of charge, for your consideration, investigation and verification. Since operating conditions vary significantly, and since they are not under our control, we disclaim any and all warranties on the results which might be obtained from the use of our products. You should make no assumption that all safety or environmental protection measures are indicated or that other measures may not be required. 2

4 The GENESIS TM Catalyst System Rosann K. Schiller Product Manager Denise Farmer Senior R&D Engineer Larry Langan R&D Engineer Grace Davison Refining Technologies Columbia, MD M ost refiners need flexible catalyst systems that allow them to take advantage of changing operating situations and feedstocks. Grace Davison delivers this flexibility with the GENESIS catalyst system. GENESIS catalysts provide a means to maximize yield potential through the optimization of discrete cracking catalyst functionality. GENESIS catalysts are a blend of two catalyst types in which one component is a MIDAS catalyst. The GENE- SIS blend concept requires that each catalyst component excel in specific performance categories in standalone use. The MIDAS component provides a high matrix input, maximizing conversion of bottoms and increasing LCO yield in many operating scenarios. Designed for refiners who are interested in maximum bottoms upgrading, MIDAS offers the ultimate in bottoms destruction and catalyst stability [1] without a coke or gas penalty, as often has been seen with competitive high matrix catalysts. The other GENESIS component is usually a high zeolite containing catalyst and is most often an IMPACT catalyst. The IMPACT family has been a breakthrough in catalytic cracking. [2,3,4,5] The novel, integral rare earth based vanadium-trapping technology in IMPACT delivers Catalagram 12 Fall 27 3

5 exceptional activity maintenance, coke and gas selectivity. The inclusion of IMPACT in the GENESIS system provides critical zeolite surface area and activity as well as superior coke and gas selectivity in a broad range of applications, from severely hydrotreated gasoils to heavy resid feeds. It is well known that the key to optimal FCC catalyst performance is the right balance between zeolite activity and matrix activity, or the Z/M ratio. [6] In pilot plant comparison testing, a high Z/M catalyst will tend to have better coke and gas selectivity than a low Z/M formulation. However, the low Z/M catalyst will achieve lower bottoms yield at constant conversion, often with a coke penalty. It is critical that zeolite and matrix activity are appropriately balanced for each unit, taking into consideration the type of feed that is processed along with the unit constraints and objectives. The GENE- SIS concept provides the ultimate in flexibility to optimize formulation Z/M for each specific application. MIDAS: The Key Component The superior bottoms destruction and catalytic stability obtained from its open particle morphology make MIDAS an excellent fit for application in all units wishing to minimize slurry. Moreover, the design of this catalyst maximizes the balance of pores in the 1-6 Å range with an optimized ratio of weak to strong acid sites in the matrix. This balance of mesoporosity and optimized acid strength greatly improves the selectivity of the bottoms cracking reactions that take place without the gas penalty often observed with other high matrix catalysts. [3] For short contact time FCC applications, it is critical that feed molecules have no limitations in getting to the catalyst acid sites, and the MIDAS catalyst architecture is ideal for eliminating diffusion constraints. [7] In addition to overcoming the challenges of diffusion limited systems, MIDAS catalysts are also 4 Hydrogen, wt.% Bottoms, wt.% Figure 1 MIDAS Commercial Performance Competitive Low Z/M Catalyst MIDAS Coke, wt.% Figure 2 MIDAS Commercial Performance Competitive Low Z/M Catalyst MIDAS Conversion, wt.% Table I Properties of IMPACT and MIDAS Catalysts IMPACT MIDAS Fresh Properties Al 2 O 3, wt.% Re 2 O 3, wt.% Surface Area, m 2 /g Zeolite, m 2 /g Matrix, m 2 /g Deactivated Properties Surface Area, m 2 /g Zeolite, m 2 /g Matrix, m 2 /g Unit Cell, Å Hg PSD, cm 3 /g , cm 3 /g , cm 3 /g , cm 3 /g

6 Bottoms, wt.% Conversion, wt.% designed to minimize the deleterious effects of high equilibrium catalyst (Ecat) contaminant metals levels associated with resid processing while delivering superior bottoms cracking to LCO and gasoline at low coke yield. The benefits of the MIDAS architecture have been demonstrated in commercial applications. Figures 1 and 2 compare the performance of MIDAS to a competitive high matrix catalyst technology. Not only does MIDAS demonstrate excellent bottoms cracking, it also reduced hydrogen production by 2% despite higher metals on Ecat. GENESIS: Combined Effectiveness Figure 3 DCR Evaluation on Resid Feedstock Bottoms Selectivity IMPACT MIDAS GENESIS exists and GENESIS catalysts demonstrate a superior coke to bottoms relationship than either component alone. The synergy is attributable to the unique matrix properties of MIDAS and its interaction with Grace Davison high zeolite catalyst technology. This synergy is the greatest when MIDAS is combined with the IMPACT catalyst family; however similar performance advantages are seen when MIDAS is combined with other high zeolite Davison catalysts. Grace Davison has done extensive R&D work to understand the synergy that exists between IMPACT and MIDAS. The summary of that data is presented here. Table II Properties of Feedstocks Research & Development Traditionally, in testing, the expectation of catalyst blend performance is that the resultant yields would be close to the linear average of each individual catalyst. In performance testing of GENESIS systems, the resultant yields exceed those of either blend component. A comprehensive study was designed to measure the selectivity advantage of these catalyst systems due to the synergy that exists between MIDAS and IMPACT. Properties of commercially produced IMPACT and MIDAS catalysts are shown in Table I. The catalysts were deactivated individually using the CPS-3 protocol [9] at 1465ºF, and the GENESIS blend was prepared after the steaming of each component. The two catalysts had similar unit cell sizes after steaming. MIDAS had lower zeolite and higher matrix surface area with higher pore volume. The catalysts were tested in the Davison Circulating Riser (DCR) on a resid, a VGO, and a hydrotreated VGO feedstock. The properties of the feedstocks are shown in Table II. The resid feed has 5.1 wt.% Con Carbon, 2.6 ºAPI and about 35% 1ºF+. The VGO feed is paraffinic (11.94 K-Factor) with about 15% 1ºF+. The hydrotreated feed has very low sulfur (.35 wt.%) and is highly naphthenic (37.2% Cn). Since MIDAS is the key component in the GENESIS blends, the expectation is that the bottoms cracking performance of the GENESIS system would be improved over IMPACT alone. Generally when bottoms cracking increases, there is a decline in coke selectivity and vice versa. It is expected that as matrix activity increases this is also true. [8] However, the relationship between bottoms cracking and coke selectivity of the GENESIS system is unique in that increased bottoms cracking is accompanied by improved coke selectivity. A synergistic effect Resid Feed VGO Hydrotreated VGO API F Sulfur, wt.% Conradson Carbon, wt.% Ni, ppm V, ppm K Factor Refractive Index Average Molecular Weight % Paraffinic Ring Carbons, Cp % Naphthenic Ring Carbons, Cn % Aromatic Ring Carbons, Ca Simulated Distillation, vol.%, F IBP % % % % % FBP Catalagram 12 Fall 27 5

7 Coke, wt.% Figure 4 DCR Evaluation on Feedstock Coke Selective Bottoms Cracking IMPACT MIDAS GENESIS Bottoms, wt.% The yields on the resid feedstock (Table III, Figure 3) show that at constant conversion, MIDAS produced about.7 wt.% lower bottoms than IMPACT. However, GENESIS had lower coke and produced only.25 wt.% higher bottoms than MIDAS. GENESIS has a bottoms yield lower than that of the weighted average of the MIDAS and IMPACT and a coke yield similar or better than IMPACT. Overall, GENESIS had the lowest bottoms at constant coke (Figure 4). Remarkably, this phenomenon is observed across a broad range of feeds, e.g. VGO feed (Table IV, Figures 5-6) and hydrotreated feedstock (Table V, Figure 7-8). 6 Table III DCR Evaluation Resid Feedstock Constant Conversion Comparison Since MIDAS and IMPACT crack different molecules in the feed, these analyses reveal the strength of the GENESIS approach. IMPACT is more effective at cracking the parafwww.e-catalysts.com IMPACT MIDAS GENESIS Cat to Oil Total Dry Gas, wt.% H 2 Yield, wt.% Total C 3, wt.% Total C 4, wt.% Gasolin e, wt.% LCO, wt.% Bottoms, wt.% Coke, wt.% Bottoms, wt.% Figure 5 DCR Evaluation Bottoms Selectivity VGO Feedstock IMPACT MIDAS GENESIS Conversion, wt.% Evidence of Synergy The detailed simulated distillation data of the bottoms (65ºF+) fractions from the DCR study are shown in Figures For the resid and VGO feedstocks, IMPACT excels at cracking the lighter (65-75ºF) bottoms, while MIDAS is superior in cracking the heavier (8-15ºF) bottoms. For the hydrotreated feed, both the IMPACT and MIDAS completely destroy the 9ºF+ portion of the bottoms. MIDAS was the most effective in cracking bottoms of all boiling ranges for this highly naphthenic feed. [1] The 65ºF+ yield from the DCR pilot plant was separated by the clay/gel method and analyzed by GC mass spectrometry [ASTM D and D ]. The analyses for the resid and hydrotreated feeds are shown in Figures 12 and 13. IMPACT is most efficient in cracking the paraffins and single ring naphthenes. MIDAS and GENESIS were more effective in cracking the 2-4 ring naphthenes. These data reflect the selectivity shown in Figures 9 and 11.

8 fins and the lower-boiling fraction of the feed, while MIDAS is more effective at cracking the naphthenes and the higher-boiling fraction of the feed. The GENESIS blend is tailored in all cases for the specific feedstock type. Why It Works GENESIS catalysts optimize the three-step bottoms cracking mechanism of Zhao et. al. [1], shown in Figure 14. Type I cracking involves vaporizing the feed. Since resid feeds may contain a significant amount of molecules boiling above the mix zone temperature of the riser (ca. 17ºF), pre-cracking, either catalytic or thermal, is necessary to achieve complete vaporization. Depending on the atomization efficiency, vaporization time of resid feed droplets may approach that of the riser contact time, which may lead to bypassing of the feed through the riser. [11] High porosity in the range of 1 to 6 Å, a signature of MIDAS (Table I), is particularly effective for adsorbing and facilitating the coke-selective cracking of large molecules during this stage. Type II cracking, involving dealkylation and paraffin cracking, is readily catalyzed by high zeolitic activity catalysts, such as IMPACT. The Type III mechanism involves cracking the naphthene rings of naphthenoaromatic compounds. These reactions are more demanding, as the competing aromatization reactions (hydrogen transfer and dehydrogenation) proceed readily. MIDAS selectively cracks the naphthene rings without any coke or gas penalty. Current Worldwide Use The synergy we observe in laboratory testing has easily translated to field performance. Figure 15 illustrates the GENESIS performance advantage observed in commercial testing. The data presented here are three commercially deactivated Ecats taken from an FCC unit pro- Coke, wt.% Figure 6 Coke Selective Bottoms Cracking DCR Evaluation on VGO Feedstock IMPACT MIDAS GENESIS Bottoms, wt.% Table IV DCR Evaluation VGO Feedstock Constant Conversion Comparison IMPACT MIDAS GENESIS Cat to Oil Total Dry Gas, wt.% H 2 Yield, wt.% Total C 3, wt.% Total C 4, wt.% Gasoline, wt.% LCO, wt.% Bottoms, wt.% Coke, wt.% Bottoms, wt.% Figure 7 Bottoms Selectivity DCR Evaluation on Hydrotreated Feed Conversion, wt.% IMPACT MIDAS GENESIS Catalagram 12 Fall 27 7

9 cessing resid that has used IMPACT and MIDAS separately as well as a GENESIS blend of both technologies. These Ecats were then tested in an ACE pilot plant over a single feed. GENESIS has the advantage in coke-selective bottoms cracking. For resid applications, GENESIS blends allow for maximization of carbon content of the feed or increased conversion up to the coke burn limit. Coke, wt.% Figure 8 DCR Evaluation on Hydrotreated Feedstock Coke Selective Bottoms Cracking IMPACT MIDAS GENESIS We've demonstrated how GENESIS performs relative to the individual components, but how does GENE- SIS compare to other technologies? A refiner processing hydrotreated feed switched from a state-of-theart Davison catalyst for hydrotreated feeds to a GENESIS system comprised of IMPACT and MIDAS. Figure 16 demonstrates the performance advantage of GENESIS relative to the hydrotreated benchmark technology. In ACE testing of unit Ecats, GENESIS results in an improved coke-to-bottoms relationship, providing the ultimate in operating flexibility for this refiner to optimize hydrotreating and FCC operations. In GENESIS catalysts, typically the other blend component is from the IMPACT family. The full range of Davison Al-Sol and Si-Sol catalysts have been used with success in GENESIS systems as shown in Table VI. These commercial applications demonstrate the viability of the GENESIS catalyst system across many different feed types. Conclusions Bottoms, wt.% Table V DCR Evaluation Hydrotreated Feedstock Constant Conversion Comparison IMPACT MIDAS GENESIS Cat to Oil Total Dry Gas, wt.% H 2 Yield, wt.% Total C 3, wt.% Total C 4, wt.% Gasoline, wt.% LCO, wt.% Bottoms, wt.% Coke, wt.% Figure 9 Bottoms Cracking Selectivity Resid Feedstock Constant Conversion DCR Comparison In the DCR study, we saw that the GENESIS system consistently performs with the coke selectivity of IMPACT and the bottoms selectivity of MIDAS, yielding the synergistic effect of a superior coke-to-bottoms relationship. This synergy exists because both catalysts crack specific feed species: IMPACT cracks paraffins more efficiently while MIDAS destroys more ringed naph- Yield, wt.% of Feed MIDAS is more selective in cracking the heaviest fraction of the feedstock Boiling range F IMPACT GENESIS MIDAS 8

10 Yield, wt.% Figure 1 Bottoms Cracking Selectivity VGO Feedstock Constant Conversion DCR Comparison IMPACT 8-85 Boiling range F GENESIS 85-9 MIDAS Figure 11 Bottoms Cracking Selectivity Hydrotreated VGO Feedstock Constant Conversion DCR Comparison Yield, wt.% IMPACT 8-85 Boiling range F GENESIS 85-9 MIDAS thenes. MIDAS improves unit coke selectivity by effectively eliminating and cracking coke precursors. The GENESIS system provides the utmost in formulation flexibility by custom tuning the blend ratio. We can optimize the catalyst Z/M ratio to match the specific unit feedstock and operating constraints. In addition to optimizing the blend ratio, the activity levels of the individual components are carefully selected to match the operating mode and feed types. This formulation flexibility can deliver a significant selectivity change, allowing the refiner to accommodate a seasonal operation, or to manage a swing feedstock, or even a hydrotreater outage. In high metals resid applications, GENESIS catalysts allow for maximization of carbon content of the feed or for increased conversion up to the coke burn limit. Not only does GENESIS reduce slurry yield, it does so without the expected coke and gas penalty often observed with competitive high matrix catalysts. In VGO and hydrotreated feed applications, GENESIS provides operational flexibility to react to seasonal economics and the opportunity to optimize hydrotreater and FCC operations together Figure 12 GC/MS Characterization of Bottoms Resid Feedstock Constant Conversion Comparison For more information on how GENE- SIS catalysts can improve your operation, please call your Grace Davison Sales Representative or Rosann Schiller at (41) , rosann.schiller@grace.com wt.% of Bottoms Paraffins 1-Ring Naphthenes 2-Ring Naphthenes 3-Ring Naphthenes 4-Ring Naphthenes IMPACT MIDAS GENESIS Catalagram 12 Fall 27 9

11 Figure 13 GC/MS Characterization of Bottoms Hydrotreated Feedstock Constant Conversion Comparison wt.% of Bottoms Paraffins 1-Ring Naphthenes 2-Ring Naphthenes 3-Ring Naphthenes 4-Ring Naphthenes IMPACT MIDAS GENESIS Figure 14 Bottoms Cracking Mechanism Catalytic Type II R Type III R Type I Coke Feed Thermal/Catalytic MIDAS is more effective for Type I and Type III. IMPACT is more effective for Type II. 9 Figure 15 Commercial Resid Example of GENESIS Synergy 8 IMPACT MIDAS GENESIS Bottoms, wt.% Coke, wt.%

12 Name Table VI GENESIS Applications Ni + V Feed Type Blend Component Refiner VGO SPECTRA Refiner 2 3 Hydrotreated IMPACT Refiner 3 4 Hydrotreated ORION Refiner 4 9 Hydrotreated IMPACT Refiner 5 12 VGO IMPACT Refiner 6 13 VGO IMPACT Refiner 7 18 VGO IMPACT Refiner 8 19 VGO IMPACT Refiner 9 2 Resid AURORA Refiner 1 2 Resid IMPACT Refiner Resid SPECTRA Refiner VGO ORION Refiner VGO IMPACT Refiner 14 3 Resid IMPACT Refiner 15 3 VGO AURORA Refiner 16 3 VGO IMPACT Refiner Resid IMPACT Refiner Resid IMPACT Refiner Resid AURORA Refiner 2 4 Resid ORION Refiner 21 6 Resid IMPACT Refiner 22 7 Resid IMPACT Refiner 23 7 Resid ADVANTA Refiner Resid IMPACT Refiner Resid IMPACT Figure 16 GENESIS Outperforms Benchmark Catalyst for Hydrotreated Feeds ECAT ACE TESTING 11 References 1. Catalagram 98, Maximizing Bottoms Upgrading: Give Resid the MIDAS TM Touch, Hunt, L., Fall Catalagram 93, IMPACT TM : A Breakthrough Technology for Resid Processing - Commercial Update, Purnell, S.K., Fall Catalagram 96, New Catalyst Technologies Based on Tunable Reactive Matrices: IMPACT TM, LIBRA TM and POLAR- IS TM, Cheng, W.C. & Nee, J.R., Fall Catalagram 97, Next Generation Al- Sol FCC Catalyst Technologies, Nee, J.R., Spring Catalagram 99, Recent Commercial Experience in Improving Refinery Profitability with Grace Davison Alumina-Sol Catalysts, Petti, N., Yaluris, G., Hunt, L., Spring Mott, R.W., Wear, C., FCC Catalyst Design for Optimal Performance, NPRA Annual Meeting 1988, AM Spry, J.C., Sawyer, W.H., 68th Annual AIChE Meeting, Los Angeles Young, G.W., Creighton, J., Wear, C., Ritter, R.E., Effect of Feed Properties on the Optimization of FCC Catalysts for Bottoms Reduction. NPRA 1987, Annual Meeting, AM Wallenstein, D, Harding, R. H., Nee, J. R. D., Boock, L. T.; Recent Advances in the Deactivation of FCC Catalysts by Cyclic Propylene Steaming (CPS) in the Presence and Absence of Metals; Appl. Catalysis A: General 24 (2) Zhao, X., Cheng, W-C., Rudesill, J. A.; FCC Bottoms Cracking Mechanisms and Implications for Catalyst Design for Resid Applications ; NPRA Annual Meeting, San Antonio, TX, Huang, Z., and Ho, T. C.; Effect of Thermolysis on Resid Droplet Vaporization in Fluid Catalytic Cracking; Chemical Engineering Journal, 91 (23) Bottoms, wt.% GENESIS Grace HT Benchmark Coke, wt.% Catalagram 12 Fall 27 11

13 GSR Products Maintain Excellent Performance When Used with Olefins Additives Wu-Cheng Cheng Grace Davison Ruizhong Hu Grace Davison Introduction T o maximize profitability from the FCCU and comply with environmental regulations, many refineries have turned to the use of FCC additives. For example, a refiner may use D-PriSM or GSR -5 additives to lower gasoline sulfur, while using an olefins additive, such as OlefinsMax to increase propylene and gasoline octane. For units using multiple additives, it is logical to inquire whether the action of one additive may interfere with the effectiveness of another. This is particularly true with the increased focus on sulfur in fuels. It is well known that ZSM-5 cracks gasoline range olefins into LPG olefins, thus limiting the extent of hydrogen transfer reactions that form gasoline range paraffins. [1] As a consequence of decreased gasoline volume, the concentration of aromatics and sulfur compounds in gasoline are expected to increase. Others [3] have suggested that due to the increase in product olefinicity, there would also be a greater possibility of recombination reactions involving olefins and diolefins with H 2 S and sulfides to form thiophenic compounds, thus increasing the concentration of sulfur compounds in gasoline. In this article, we will show that although ZSM-5 increases gasoline sulfur by a concentration effect, it does not diminish the effectiveness of catalytic sulfur reduction technologies, either in laboratory testing or in commercial applications. 12

14 Figure 17 Gasoline Sulfur Formation through Recombination Historical Understanding H-S-R or R-S-R H-S-R or R-S-R The origin of sulfur molecules in FCC gasoline and their interaction with FCC catalyst have been studied extensively [1-4]. Harding et al. determined that only about 4% of feed sulfur ended up in gasoline boiling range (43 F-) products, while the majority of feed sulfur cracked into H 2 S. The gasolinerange sulfur compounds were primarily thiophene derivatives and benzothiophene. Thiophene compounds could be removed primarily via hydrogen transfer to tetrahydrothiophenes and subsequent ring opening to form H 2 S. Thus, increasing the hydrogen transfer activity of the base FCC catalyst or using D- PriSM additive, which enhances the rate of cracking of tetrahydrothiophene to H 2 S, can both effectively lower gasoline sulfur. S R S S S R R the rate of recombination reactions (such as those in Figure 17) due to the increase in light olefins present was also suggested as a possibility. The most likely explanation is that the addition of 1% ZSM-5 additive to the base catalyst, which had inherently low activity, diluted the cracking and hydrogen transfer activity. This reduced cracking and hydrogen transfer activity could very well have led to the modest increase in gasoline sulfur observed. In many commercial applications where higher levels of ZSM-5 additive are used, the base catalyst is typically reformulated to maintain constant cracking activity. 1 R An attempt to quantify the potential for recombination reactions in the FCC unit was conducted by Leflaive et al. [4] using pure olefin and diolefin compounds in the presence of H 2 S and FCC catalyst. The study showed that at very long contact times, the transformation of olefins or diolefins into thiophenic compounds was possible, and diolefins were shown to be more likely than olefins to transform. The proposed mechanism by Leflaive et al. involves first the formation of a thiol through a nucleophilic addition of H 2 S, then a cyclization into tetrahydrothiophene derivatives that can be dehydrogenated into thiophenic compounds. However, as pointed out by Harding et al. and Corma et al. [5], the thiol and tetrahydrothiophene sulfur compounds are among those most easily converted in the FCC process. Moreover, many gasoline sulfur reduction technologies (e.g. D-PriSM additive) are specifically designed to drive these easy to crack sulfur compounds to H 2 S, and make them unavailable for recombination reactions. Hence, though the recombination reactions are mechanistically possible, their contribution to gasoline sulfur formation in actual FCC commercial operation has not been quantified. Figure 18 Effect of ZSM-5 on Gasoline Olefins ZSM-5 does not Increase gasoline olefins concentration DCR Evaluation [3] Lappas et al. showed that the addition of a 1% ZSM-5 additive to a low metals (vanadium and nickel), low unit cell size Ecat increased the concentration of FCC gasoline sulfur. However, most of this increase could be attributed to the concentration effect caused by shifting of gasoline to LPG. Calculated on a feed basis, gasoline sulfur increased by only 4%. The authors attributed this modest gasoline sulfur increase to the change in relative rates of chain cracking, hydride transfer and cyclization. Increasing Gasoline Diolefins, wt.% wt.% OlefinsMax.2 2 wt.% OlefinsMax 4 wt.% OlefinsMax Conversion, wt.% Catalagram 12 Fall 27 13

15 Recent Laboratory Examples Based on our laboratory testing and commercial experience, we have found that the use of ZSM-5 additives does not increase gasoline sulfur beyond the above mentioned concentration effect for two significant reasons. First, while diolefins are much more reactive than olefins and are expected to be more likely to undergo recombination reactions, ZSM-5 does not increase the concentration of diolefins (Figure 18), which are mainly a product of thermal cracking (Figure 19). Next by examining the possible pathways of a gasoline range olefins molecule, it is evident that recombination reactions are not as favorable as alternative reactions. A gasoline olefin molecule can undergo hydrogen transfer to form a paraffin. It can further crack to form smaller olefins or it can recombine with H 2 S or sulfides to form thiophenes. The ability of ZSM-5 to alter FCC yields indicates that the rate at which Gasoline Diolefins, wt.% Figure 19 Effect of Temperature on Gasoline Olefins Gasoline olefins are a result of thermal cracking DCR Evaluation 97 F 11 F 15 F gasoline olefins crack on ZSM-5 is the same or faster than the rate at which gasoline olefins undergo hydrogen transfer reactions on base FCC catalyst. In other words, ZSM-5 does not Conversion, wt.% decrease the intrinsic rate of hydrogen transfer; it merely offers a faster alternative pathway for gasoline olefins. Thus, the relative rates of the three reactions are: Cracking Table VII Effect of ZSM-5 on 3 ppm Ni, 3 ppm V/CPS DCR Evaluation on Resid Feed - 75 wt.% Conversion Base Base 1% OlefinsMax Cat to Oil H 2 Yield, wt.% C 1 + C 2 s, wt.% Total C 3, wt.% C 3 =, wt.% Total C 4, wt.% Total C 4 =, wt.% Base 2% OlefinsMax Base 4% OlefinsMax Base 8% OlefinsMax Gasoline, wt.% G-Con P, wt% G-Con I, wt% G-Con A, wt% G-Con N, wt% G-Con O wt% G-Con RON EST G-Con MON EST LCO, wt.% Bottoms, wt.% Coke, wt.%

16 Table VIII Gasoline Sulfur Concentration 3 ppm Ni, 3 ppm V/CPS DCR Evaluation on Resid Feed - 75 wt.% Conversion ZSM-5 concentrates gasoline sulfur by cracking gasoline into LPG. Gasoline sulfur concentration on a feed basis is constant Base Base 1% OlefinsMax Base 2% OlefinsMax Base 4% OlefinsMax Base 8% OlefinsMax Sulfur Concentration ppm in Gasoline Mercaptans Thiophene MethylThiophenes TetrahydroThiophene C 2 - Thiophenes C 3 - Thiophenes BenzoThiophene AlkylBenzoThiophenes Sulfur Concentration ppm on Feed Basis Mercaptans Thiophene MethylThiophenes TetrahydroThiophene C 2 - Thiophenes C 3 - Thiophenes BenzoThiophene AlkylBenzoThiophenes on ZSM-5 > Hydrogen Transfer > Recombin-ation, and the addition of ZSM-5 should actually lower, not increase the relative contribution of recombination. To support this discussion, a pilot plant study was commissioned to show the effect on yields and gasoline sulfur of increasing levels of ZSM-5 additive combined with a high rare earth base FCC catalyst. Tables VII and VIII list hydrocarbon yields and the concentration of various gasoline sulfur species, respectively, at constant conversion resulting from cracking a resid feedstock over the various samples in the Davison Circulating Riser (DCR). The samples were deactivated with 3 ppm nickel and 3 ppm vanadium, using the Cyclic Propylene Steaming (CPS) protocol. Table VII exhibits the effect of OlefinsMax additions, at, 1, 2, 4 and 8%, where the hydrocarbon yields clearly show the expected trend when an increasing amount of ZSM-5 additive is used. The C/O at constant conversion was essentially unchanged for these blends with 1%, 2% and 4% OlefinsMax and increased slightly for the blend with 8% OlefinsMax due to a slight activity dilution effect. The C 3 and C 4 olefins and paraffins increased at the expense of gasoline olefins, and there was an overall decrease in gasoline yield. Table VIII shows the major gasoline-range sulfur species produced for the catalyst blends tested, and are listed first on a gasoline basis and then on a feed basis. It is evident that on a gasoline basis, the concentrations of most sulfur species increased with the increase in OlefinsMax additive concentration. However, on a feed basis, the yields of most sulfur species were unchanged (fluctuations were within the measurement error). It is interesting to note that the concentration of mercaptans (sum of all gasoline range sulfur species with a boiling point below thiophene) appears to decrease with increasing concentrations of OlefinsMax additive. This trend is consistent with the observations of Corma et al. Compared to the earlier example by Lappas et al., we believe that the higher activity of the base catalyst and the lower concentrations of ZSM-5 additive used in this study resulted in a lower impact on the overall cracking and hydrogen transfer activity, and therefore no gasoline sulfur increase was observed on a feed basis. In addition to this study where laboratory deactivated samples were used in the testing, we also commissioned a study using commercial equilibrium catalysts (Ecats) containing both ZSM-5 and gasoline sulfur reduction additives. Table IX lists three Ecats from one commercial unit that were collected at different times. The first two Ecats contained % and 2.% OlefinsMax, respectively, while the third Ecat contained 1.5% OlefinsMax but with a different base catalyst. The first two Ecat samples contained an alu- Catalagram 12 Fall 27 15

17 Table IX Properties of Ecat Samples Base Ecat No OlefinsMax Base Ecat 2.% OlefinsMax SuRCA Ecat 1.5% OlefinsMax Al 2 O 3, wt.% Re 2 O 3, wt.% Na 2 O, wt.% Fe, wt.% Ni, ppm V, ppm ABD, cc/gm APS, m Surface Area, m 2 /gm Zeolite Matrix Unit Cell, Å mina-sol catalyst with moderate rare earth, while the third Ecat sample is the SuRCA version of that catalyst, designed to provide similar yields as the base catalyst but to also reduce gasoline sulfur. The surface area and unit cell size of the three Ecats are essentially identical. All three Ecats were tested in an ACE [6] pilot plant unit using the refiner's hydrotreated feed. The hydrocarbon yields and sulfur in gasoline at constant conversion are listed in Table X. The increase in propylene yield and octane from the % OlefinsMax sample to either the 2.% or the 1.5% OlefinsMax samples is proportional to the increase in OlefinsMax in the Ecat. The decrease in gasoline yield is also proportional to the increasing ZSM-5 additive concentration. In spite of the concentration effect from the decreasing gasoline yield, the gasoline sulfur concentration of the SuRCA containing Ecat is about 3% lower than the two base Ecats. Table X ACE Evaluation of Base and SuRCA Ecat Samples SuRCA lowers gasoline sulfur, even in the presence of ZSM-5 additive Conversion 7 Base Ecat No OlefinsMax Base Ecat 2.% OlefinsMax SuRCA Ecat 1.5% OlefinsMax Catalyst to Oil Ratio Dry Gas Propylene Total C 3 s Total C 4 = s Total C 4 s Gasoline RON MON LCO Bottoms Coke Gasoline Sulfur, on gasoline basis Reduction Light Cut Sulfur, ppm % Heavy Cut Sulfur, ppm % Cut Gasoline Sulfur, ppm % Total Sulfur, ppm % 16

18 G - Con RON EST C 3 = wt.% Figure 2 Propylene Yield Trends for Ecat Sample Blends 4% OlefinsMax produces expected increase in propylene GSR-5 does not negatively impact ZSM-5 ZSM-5/GSR Interaction DCR Study Conversion, wt.% Base Ecat Al 2 O 3 (wt.%) 42.9 RE 2 O 3 (wt.%) 3.2 Na2O (wt.%).36 Fe 2 O 3 (wt.%).72 P 2 O 5 (wt.%).9 Ni (ppm) 4 V (ppm) 6 Surface Area (m 2 /g) Zeolite Base Ecat Base Ecat + 4% OlefinsMax Base Ecat + 25% GSR-5 Base Ecat + 4% OlefinsMax + 25% GSR-5 Table XI Properties of Base Ecat and Feed Matrix 37 Unit Cell (Å) Feed API 24.7 Aromatic Ring Carbons, Ca (wt.%) 2.5 Napthenic Ring Carbons, Cn (wt.%) 17.1 Paraffinic Carbons, Cp (wt.%) 62.4 Sulfur (wt.%).82 Concarbon (wt.%).77 Distillation ( F), IBP 31 1% ( F) 583 3% ( F) 723 5% ( F) 799 7% ( F) 878 9% ( F) 12 FBP ( F) 1177 Figure 21 Gasoline Octane Trends for Ecat Sample Blends 4% Olefins-Max produces expected increase in octane GSR-5 does not negatively impact ZSM-5 ZSM-5/GSR Interaction DCR Study Conversion, wt.% Base Ecat Base Ecat + 4% OlefinsMax Base Ecat + 25% GSR-5 Base Ecat + 4% OlefinsMax + 25% GSR-5 To further decouple the ZSM-5 and sulfur reduction effects, a DCR study with commercial Ecat was conducted. A high rare earth, low metals base Ecat was blended separately with 4% OlefinsMax and with 25% GSR-5, as well as with both 4% OlefinsMax and 25% GSR-5. GSR- 5 is an additive based on the sulfur reduction functionality of SuRCA that also provides base cracking functionality. The samples were lab deactivated to match typical commercial performance. A fairly paraffinic feedstock with.8 wt% sulfur was cracked over each of the samples. The properties of the base Ecat sample and the feed used in the study are shown in Table XI. The addition of 4% OlefinsMax provided similar shifts in propylene (Figure 2) and gasoline octane (Figure 21) as the same amount of the additive produced in Table VII. The addition of the GSR-5 additive did not cause any change in olefins or octane. Additionally, because GSR-5 imparts cracking activity while providing gasoline sulfur reduction, the addition of GSR-5 to the base Ecat did not significantly change the yields of other FCC products. Figure 22 shows the cut gasoline sulfur concentration on a gasoline basis for the samples tested. Cut gasoline sulfur is the sum of the sulfur species that boil through 428 ºF (mercaptans through C 4 thiophenes). GSR-5 by itself reduces gasoline sulfur by 3% at 78% conversion, and the sulfur reduction performance is similar with the addition of 4% OlefinsMax. This is consistent with the performance observed in Table X. The addition of 4% OlefinsMax yields 8% higher gasoline sulfur at 78% conversion on a gasoline basis, which is comparable to the 7% increase in gasoline sulfur observed for the commercial comparison of the base Ecat to the 2% OlefinsMax Ecat in Table X. Normalizing the results in Figure 2 to a feed basis, the increase in Catalagram 12 Fall 27 17

19 Cut Gasoline Sulfur, ppm Figure 22 Gasoline Sulfur Trends for Ecat Sample Blends GSR-5 reduces gasoline sulfur by 3% alone or with ZSM-5 Higher gasoline sulfur is due to concentration effect of ZSM-5 ZSM-5 / GSR Interaction DCR Study Conversion, wt.% Base Ecat Base Ecat + 4% OlefinsMax Base Ecat + 25% GSR-5 Base Ecat + 4% OlefinsMax + 25% GSR-5 gasoline sulfur with 4% OlefinsMax in inventory is only 3%, again indicating that it is a concentration effect that causes higher apparent gasoline sulfur with ZSM-5 additive use. Commercial Example The gasoline sulfur reduction performance of SuRCA catalyst and the GSR-5 additive in the above studies are similar to what we have observed with Ecats that do not contain ZSM-5 additives. In 24, the Alon USA, Big Spring, TX refinery co-authored an article with Grace Davison summarizing their successful experience with the use of SuRCA catalyst to reduce their FCC gasoline sulfur by 2% [7]. The refinery also used OlefinsMax on an opportunity basis to make incremental refinery grade propylene. The performance of SuRCA was quantified both with and without the ZSM-5 additive. As Figure 23 indicates, the 2% reduction in gasoline sulfur was consistent whether or not ZSM-5 was present. Conclusion Recent pilot plant analysis and commercial data indicates that any increase in gasoline sulfur observed with the use of ZSM-5 additives is due to a concentration effect from the cracking of gasoline molecules into LPG, as opposed to recombination reactions. Furthermore, ZSM-5 additives used in combination with gasoline sulfur reduction technologies do not show increased gasoline sulfur from recombination reactions. Refiners have utilized both technologies simultaneously and achieved performance comparable to the use of each independently. Those refiners considering the use of both technologies can be confident in the ability of each product to achieve its individual product performance goals. Gasoline S lbs/feed S lbs.7 Figure 23 SuRCA Performance at Alon USA, Big Spring, Texas.65 2% Lower Gasoline Sulfur Selectivity % Point, F References 1. P.H. Schipper, F.G. Dwyer, P.T. Sparrell, S. Mizrahi, and J.A. Herbst, Fluid Catalytic Cracking Role in Modern Refining, M.L. Occelli (Ed.), ACS Symposium Series, Vol. 375, American Chemical Society, Washington D.C., 1988, p R.H. Harding. R.R. Gatte, J.A. Whitecavage, and R.F. Wormsbecher, Reaction Kinetics of Gasoline Sulfur Compounds, J.N. Armor (Ed.), ACS Symposium Series, Vol. 552, American Chemical Society, Washington D.C., 1994, p A.A. Lappas, J.A. Valla, I.A. Vasalos, C. Kuehler, J. Francis, P. O'Connor, and N.J. Gudde, The Effect of Catalyst Properties on the In Situ Reduction of Sulfur in FCC Gasoline, Applied Catalysis A: General 262 (24) p P. Leflaive, J.L. Lemberton, G. Pérot, C. Mirgain, J.Y. Carriat and J.M. Colin, On the Origin of Sulfur Impurities in Fluid Catalytic Cracking Gasoline - Reactivity of Thiophene Derivatives and of Their Possible Precursors Under FCC Conditions, Applied Catalysis A: General 227 (22) p A. Corma, C. Martínez, G. Ketley, and G. Blair, On the Mechanism of Sulfur Removal During Catalytic Cracking, Applied Catalysis A: General 28 (21) p J.C. Kayser, Versatile Fluidized Bed Reactor, U.S. Patent No. 6,69,12 (2). 7. M. Gwin (Alon), E.J. Udvari, and D.A. Hunt, SuRCA Catalyst Reduces FCC Gasoline Sulfur and More at the Alon USA, Big Spring Refinery, Catalagram 96 (24) p (published by Grace Davison, a business unit of W.R. Grace & Co.). Base Division Catalyst SuRCA SuRCA with OlefinsMax 18

20 Catalagram 12 Fall 27 19

21 Worldwide FCC Equilibrium Catalyst Trends - A Ten-Year Review Marilyn Moncrief David Hunt Kelly Stafford Grace Davison Refining Technologies G race Davison's laboratories test thousands of equilibrium fluid cracking catalyst (Ecat) samples each year. These samples provide important insights into FCC unit operations and are critical for unit optimization and troubleshooting. The purpose of this article is twofold. First, it will communicate how Ecat activity, contaminants and other properties have shifted over the past ten years, both worldwide and by geographic region. Second, it will allow the individual refiner to rank their own FCC Ecat properties relative to the industry in several key categories. The following data reflects ten years of Ecat sample analyses from 1997 through 26. The data represents over 117, individual Ecat samples from approximately 3 FCC units around the world. Figures 24 to 32 show a ten-year trend of average Ecat properties across all regions of the world: Asia Pacific (AP), European Union (EU), Latin America (LA) and North America (NA). EU includes Europe, Africa, the Middle East and Russia. North America includes the United States, Canada and the U.S. Virgin Islands. Data reflects Ecat samples 2

22 MAT, wt.% Figure 24 Average MAT Activity by Region Year Region AP EU LA NA WW that we have received from 1997 through 26, both from refiners using Grace Davison FCC catalyst as well as competitor products. Figure 24 identifies interesting trends in MAT activity. All regions experienced significant increases in activity from 1997 through 24, at which time activity stabilized. On a worldwide basis, average activity increased from 67.5 to 7.7 wt.% over the ten-year period. Additionally, North America has consistently reported the highest activity of the four regions, while Asia Pacific has seen the greatest overall gains, from 64.2 to 69.7 wt.%. RE23, wt.% Figure 25 Average Rare Earth by Region Year Figure 26 Average Unit Cell Size by Region Region AP EU LA NA WW Higher activity is consistent with increases in Ecat rare earth content (Figure 25) and Ecat Unit Cell Size, UCS (Figure 26). Worldwide, average rare earth has climbed more than 65%, from 1.54 to 2.56 wt.%, over the past ten years. Similar increases are seen in each geographic region. Consistent with rare earth, average Ecat UCS data has seen a steady rise from to 24.3Å. Unit cell size and rare earth data suggests that the increase in Ecat activity is largely due to higher levels of rare earth exchanged onto the catalyst zeolite. Figures 25 and 26 also confirm conditions in the FCC Ecat that can lead to higher gasoline selectivity as a result of the shift to higher UCS Ecat. For many catalyst systems, this shift in UCS also suggests improved catalyst coke selectivity. UCS, Angstroms Year Region AP EU LA NA WW FCC catalyst alumina content has experienced a steady upward trend from 38.9 to 44. wt.%, as seen in Figure 27. Higher Al 2 O 3 has been observed in all regions and confirms the industry's acceptance of the value of alumina content on FCC catalyst performance. Grace reported on the value of alumina-sol catalyst technologies in a recent Catalagram publication. (1) Ecat contaminant trends are seen on Figures 28 to 32. Nickel in par- Catalagram 12 Fall 27 21

23 Al23, wt.% Figure 27 Average Alumina by Region Year Figure 28 Average Nickel by Region Region AP EU LA NA WW ticular (Figure 28) provides insight into the differing FCC feedstocks processed in the Pacific Rim units as compared to the rest of the world. Vanadium (Figure 29) has been on the rise in the Asia Pacific region, increasing from 19 to 24 ppm and reflecting a 26% increase with a 14% increase in nickel. This trend is reversed on a worldwide basis, where vanadium has increased less than 6% while nickel is up almost 2% over the ten-year period, from 1475 to 175 ppm. Nickel, and to a lesser extent vanadium, acts as dehydrogenation catalyst that increases the yields of the unwanted products hydrogen and coke. (2) Vanadium is also mobile under FCC regenerator conditions and reduces catalyst activity by destroying zeolite framework. (3) V, ppm Ni, ppm Year Figure 29 Average Vanadium by Region Year Region AP EU LA NA WW Region AP EU LA NA WW Increased Ecat activity together with higher nickel and vanadium levels suggest that today's catalysts have improved coke selectivity due to enhanced metals trapping and improved zeolite and matrix design. Iron presents a mixed picture (Figure 3). Iron levels have dropped by 11% in Europe, 1% in Asia Pacific, and almost 5% in Latin America over the last ten years. Iron in North America, on the other hand, dropped significantly in the late 199's, but has been increasing steadily since 2. Today the FCC's with the highest Ecat iron levels are located in North America. Organic based iron deposited on the catalyst during the cracking reactions can have a serious adverse effect on activity and bottoms cracking. (4) Calcium (Figure 31) had been stable for several years, but has climbed substantially since 22. Worldwide CaO levels have increased 74% overall from.83 to.144 wt.%. Asia Pacific has the highest levels, while North and Latin America have seen the highest percentage increases. Europe has the lowest average calcium by weight percent and has also experienced the lowest percentage increase 22

24 over the time period. Ca is often found on the surface of the Ecat together with Fe and may be involved in the mechanism by which Fe poisons the Ecat. (4) Figure 3 Average Iron by Region Sodium (Figure 32) has been trending downward on a worldwide basis, decreasing by over 12% since Asia Pacific and North America have contributed to the overall decrease by dropping almost 25% and 18%, respectively. Over the last ten years FCC units in Europe have the lowest levels of sodium, while Latin America has the highest. Sodium on Ecat comes both from the raw materials used to manufacture the catalyst as well as salt contamination in the feedstock. Sodium can deactivate the catalyst by poisoning the acid sites on the matrix and zeolite and by surface area sintering. (5) Fe, wt.% Year Figure 31 Average Calcium by Region Region AP EU LA NA WW Figures 33 through 42 present normal distributions of worldwide 26 Ecat data. These plots can be used as a quick reference to determine where an individual FCC unit falls versus the industry. The numbers atop each bar represent the number of FCC units within that data range. CaO, wt.% Region AP EU LA NA WW As can be seen in Figure 33, MAT activity reflects a range from 57 to 81 with a mean of 7.7 wt.%. Most FCC units operating at an activity level greater than 77% likely process deeply hydrotreated feedstock, while units operating at lower activity could be processing residual based feedstocks and targeting lower conversion levels Year Figure 32 Average Sodium by Region Figure 34 presents a rare earth range from.2 to 5.48 wt.% with a mean of 2.56 wt.%. Many units operating at higher rare earth levels are taking advantage of Grace Davison's IMPACT catalyst technology, which incorporates an integral rare earth vanadium trap. Figure 34 also confirms the limited number of units using a zero rare earth catalyst. Na, wt.% Region AP EU LA NA WW Year Catalagram 12 Fall 27 23

25 Alumina has a fairly broad range from 26.5 to 56.5 wt.%, with a mean of 44. wt.%, as can be seen on Figure 35. Nickel and vanadium distributions are shown on Figures 36 and 37. While average Ecat vanadium levels are higher than nickel, nickel levels show a much wider distribution at the high end. Nickel ranges from a low of 22 ppm to several units with nickel levels greater than 12, ppm. The average nickel level worldwide is approximately 175 ppm. Vanadium ranges from 4 to 726 ppm, with a mean of about 188 ppm. Frequency Figure 33 MAT Activity Distribution MAT Activity, wt.% Mean Figure 38 shows the worldwide iron distribution. Like nickel, several units operate with a very high Iron content. The range is from.19 to 2.29 wt.% with a mean of.53 wt.%. Calcium reflects a range of.2 to 1.34 wt.% and a mean of.14 wt.%, as can be seen in Figure 39. Sodium distribution ranges from.9 to.95 wt.% (Figure 4). The mean is.3 wt.%. A normal distribution for total surface area is presented in Figure 41. Surface area indicates a range from 72 to 254 m 2 /g and a mean of 148 m 2 /g. Figure 42 shows the distribution of the Ecat -4 micron content, which ranges from to 28 wt.% with a mean of 8 wt.%. The high end of the range indicates units, which can hold a tremendous amount of -4 material and perhaps generate additional fines from catalyst attrition. Several units at the low end of the distribution likely have cyclone problems, which limit their ability to hold particles less than 4 microns. Frequency Figure 34 Rare Earth Distribution RE 2 3, wt.% 26 Figure 35 Alumina Distribution Mean Mean Data presented in this article confirms that the FCC industry values high activity catalysts for units that process hydrotreated feedstocks, are constrained by catalyst circulation and strive to operate at high conversion levels. Ecat contaminant levels continue to increase, particularly CaO and Fe. As a result, the industry will continue to demand Frequency Al 2 3, wt.%

26 Figure 36 Nickel Distribution 2 26 Mean Frequency Ni, ppm > 6 55 Figure 37 Vanadium Distribution 5 Mean 1881 Frequency V, ppm Figure 38 Iron Distribution Mean Frequency Fe, wt.% > Catalagram 12 Fall 27 25

27 Figure 39 Calcium Distribution 12 Mean.1441 Frequency CaO, wt.% > Figure 4 Sodium Distribution Mean.33 Frequency Na, wt.% Figure 41 Total Surface Area Distribution Mean Frequency Total Surface Area, m2/g 26

28 Frequency FCC catalysts that provide excellent coke selectivity and high liquid yields through enhanced metals tolerance. This data also confirms that 14 2 Figure 42-4 Micron Particle Distribution Micron, wt.% Mean > there is a wide distribution of contaminant metals and that each catalyst application must be designed for the specific application. References 1. Petti, Yaluris and Hunt, Recent Commercial Experience in Improving Refining Profitability with Grace Davison Alumina-Sol Catalysts, Catalagram No. 99, pg. 2-11, Petti, Tomczak, Pereira, Cheng, Investigation of nickel species on commercial FCC equilibrium catalysts-implications on catalysts performance and laboratory evaluation, Applied Catalysis General 169 (1998), pg Wormsbecher, Cheng, Kim, Harding, Deactivation and Testing of Hydrocarbon Processing Catalysts, ACS Symposium Series 634, 1996 American Chemical Society 1996, pg Yaluris, et al, The Effects of Fe Poisoning on FCC Catalysts, NPRA Annual Meeting 21, New Orleans, LA, AM Zhao, Cheng, Deactivation and Testing of Hydrocarbon Processing Catalysts, ACS Symposium Series 634, 1996 American Chemical Society 1996, pg Redesign - Coming Soon G race Davison is pleased to announce that we are working on the redesign of the e- Catalysts.com web site. The new site will maintain a lot of the familiar placement and navigational features of the old site with a new graphic style and appearance that is more focused on the needs of our customers. e-catalysts.com celebrates its 6th anniversary this year and continues to remain focused on being the premier internet source of customized technical services for refining processes. We invite you to visit e-catalysts today. If not already a member and interested in joining, we invite you to apply at Catalysts.com or contact Phyl Strawley at ( ). Catalagram 12 Fall 27 27

29 THE Effects of Treat Gas H 2 Purity on ULSD Catalyst Performance A. E. Zieber Process Engineer, Chevron USA, Salt Lake City, UT Refinery B.R.Watkins C. W. Olsen Advanced Refining Technologies C hevron USA s Salt Lake City Refinery loaded their recently revamped ULSD unit in April 26 with approximately 22, lbs of ART's SmART catalyst system consisting of 75% ART CDXi and 25% ART AT55. The unit commenced operation in May 26. At the time of startup, the unit received hydrogen from several sources, and estimates put the hydrogen purity between 65% and 75% at the reactor inlet. Shortly after start up the unit began deactivating at a higher than expected rate, and it appeared that a catalyst change-out would be required by the end of the year. A summary of the performance experienced during this period is shown in Figure 43. The figure shows that the start of run WABT was F, and the very high rate of deactivation is readily apparent. To help understand the performance of this ULSD unit the refinery provided feedstock for use in the pilot plant program. The goal of the study was to evaluate the performance of the SmART system using Chevron, Salt Lake City feed and operating conditions, and to explore 28

30 WABT, F the impact of hydrogen partial pressure on catalyst activity and stability. The initial phase of the pilot plant program utilized 1% hydrogen at unit conditions using lot retains of the catalyst loaded in the commercial unit. This showed that 628 F was required to achieve 7 ppm product sulfur. The Salt Lake City ULSD unit is not equipped with a recycle gas scrubber, and measurements indicated that there was about 2 mole% H 2 S in the recycle gas stream. Simulating 2 mole% H 2 S in the treatgas to the pilot plant resulted in a decrease in HDS activity, and it then required about 635 F to achieve 7 ppm product sulfur. The next phase of the testing was aimed at investigating the effects of hydrogen purity on performance. Chevron was interested in comparing the impact of 9% H 2 purity and 7% H 2 purity. As mentioned above, the purity at the start of the cycle was believed to be between 65-75%, and 9% represented the expected case after start up of a new hydrogen plant at the refinery. Since the pilot plant uses 1% hydrogen, the hydrogen purity effects were simulated by adjusting the pressure to match the hydrogen partial pressure corresponding to Figure 43 Chevron, Salt Lake ULSD Operation May 26 Through September Days the desired hydrogen purities. Flash calculations were completed using the commercial unit feed and conditions to determine the H 2 pressure for both 9% and 7% H 2 purity cases. The 9% purity case operated at a hydrogen partial pressure of 612 psia while the 7% purity case was at a partial pressure of 48 psia. The effects of lowering the hydrogen partial pressures representing the 9% and 7% purities had a significant effect on catalyst activity in ULSD service. Operating at 612 psia hydrogen resulted in an increase of WABT to achieve 7 ppm from 628 F to 642 F while the 48 psia hydrogen case required an increase in temperature to 657 F for 7 ppm sulfur. These results 7% PURITY 9% PURITY 2% H 2 S 1% PURITY are summarized along with the 1% H 2 purity and H 2 S effect in Figure 44. The pilot plant protocol then called for 1 hours of operation while maintaining 1 ppm sulfur in order to compare the effects of decreasing hydrogen purity on catalyst stability. The results of this are shown in Figure 45. The 9% purity system had a deactivation rate of about 6 F/month while the 7% system deactivated at about 17 F/month. These pilot plant runs are summarized Figure 45. In September 26 the Salt Lake City refinery brought their new hydrogen plant on line and this, combined with several other operational changes, increased the hydrogen purity of the treat gas to the ULSD unit to >95%. This resulted in a significant decrease in the rate of deactivation. The unit was experiencing 15 F/month deactivation on average, with periods as high as 3 F/month prior to starting the hydrogen plant. The fouling rate decreased to about 5 F/month with the improved hydrogen purity. This part of the unit cycle is shown in Figure 46. These changes allowed the unit to run a full 12-month cycle before turning around. Not surprisingly, the hydrogen purity of the treat gas also has a significant impact on the product color. Figure 44 Hydrogen Purity Effects on Catalyst Activity Required WABT Increase, F Catalagram 12 Fall 27 29

31 695 Figure 45 Relative Deactivation Rates With Hydrogen Purity 685 9% 7% 675 WABT, F Hours The color of the product degraded steadily as the catalyst deactivated, and after the aging step the ASTM color was significantly worst for the 7% H 2 case at <3. ASTM compared to 1.5 ASTM for the 9% hydrogen case. This compares with an ASTM color of around 1. for both cases prior to aging. Similarly, the HDN activity is severely affected. The product nitrogen increased from 1-2 ppm to 2 ppm during the aging test for the 7% purity case compared to a change of only 1 to 3 ppm for the 9% purity case. This demonstrates that lower H 2 purity results in a loss of HDS and HDN activity, degrades product color and increases the deactivation rate. Chevron, Salt Lake turned the ULSD unit around in June 27 and loaded a fresh SmART catalyst system. The unit has been operated at full capacity for several months without the same drastic start-of-run fouling rate observed on the previous run, due to operation with >95% pure hydrogen. The current deactivation rate is showing <5 F/month. Advanced Refining Technologies and Chevron worked closely together to troubleshoot the poor performance experienced in their newly revamped ULSD unit charged with a SmART Catalyst System. The pilot plant testing was able to prove that the high deactivation rate observed early in the cycle was completely the result of the low hydrogen partial pressure and not the catalyst. Based on the confidence developed in the pilot data, Salt Lake again chose to load the ART catalyst for the second run. This experience is a good demonstra- WABT, F tion of the importance of hydrogen partial pressure in ULSD applications. Maintaining good hydrogen partial pressure is extremely important for lower pressure ULSD units, and the data suggests that there is a hydrogen partial pressure "cliff where lower purities cause a much greater than expected increase in fouling rate. Figure 46 Chevron, Salt Lake City ULSD Operation With Improved H 2 Purity Days 3

32 Catalytic FCC Gasoline: Sulfur Reduction Mechanism Excerpted from The Journal of Catalysis Yuying Shu Rick Wormsbecher Grace Davison Refining Technologies E nvironmental regulations have caused refiners to lower specifications on the sulfur content of motor fuels. Because FCC gasoline contributes up to 9 % of the sulfur to the gasoline pool, this has highlighted the importance of reducing sulfur directly in the FCC unit. Grace Davison has been providing catalysts and additives that reduce FCC gasoline sulfur by up to 45% to the refining industry for over 1 years. These technologies have been proven in over 8 units worldwide and include both additive technologies D-PriSM and GSR -5, and complete catalyst systems, such as SuRCA and the newest catalyst system Neptune TM. In 1992, well before the industry realized the need for sulfur reduction in FCC gasoline, Grace Davison began its extensive research and development efforts to reduce sulfur in FCC products. The first patents, appearing in 1994 and 1996, describe a Lewis acid on alumina based additive technology, with the primary Lewis acid being Zn. Nearly all of the sulfur species in gasoline are thiophene and alkylthiophenes, which are Lewis bases. It is the realization that Lewis acid centers on the catalyst surface play a key role in the direct removal of sulfur species that led to the development of the D-PriSM additive, however, little was known about the details of the surface chemistry and reaction pathways. In 1994, Grace Davison scientists proposed that the sulfur reducing additive increased the cracking of tetrahydrothiophenes as shown in Figure 47. Catalagram 12 Fall 27 31