Rive Molecular Highway TM Catalyst Delivers Over $2.50/bbl Uplift at Alon s Big Spring, AM-13-03

Transcription

1 Annual Meeting March 17-19, 213 Marriott Rivercenter San Antonio, TX Rive Molecular Highway TM Catalyst Delivers Over $2.5/bbl Uplift at Alon s Big Spring, Texas Refinery Presented By: Mr. Larry Dight, Rive Technology, Inc. Monmouth Junction, NJ Mr. Gautham Krishnaiah Rive Technology, Inc. Monmouth Junction, NJ Dr. Barry Speronello Rive Technology, Inc. Monmouth Junction, NJ Mr. Allen Hansen Rive Technology, Inc. Monmouth Junction, NJ Mr. Jimmy Crosby Alon USA Big Spring, TX American Fuel & Petrochemical Manufacturers 1667 K Street, NW Suite 7 Washington, DC voice fax fpm.org

2 This paper has been reproduced for the author or authors as a courtesy by the American Fuel & Petrochemical Manufacturers. Publication of this paper does not signify that the contents necessarily reflect the opinions of the AFPM, its officers, directors, members, or staff. Requests for authorization to quote or use the contents should be addressed directly to the author(s)

3 Rive Molecular Highway TM Catalyst Delivers Over $2.5/bbl Uplift at Alon s Big Spring, Texas Refinery Mr. Gautham Krishnaiah, Director of Technical Service 1 Dr. Barry Speronello, Research Fellow Mr. Allen Hansen, Process Modeling Engineer Rive Technology, Inc. Mr. Jimmy Crosby, Vice President, Refining Alon USA Abstract: Refiners are continuously striving to expand margins by optimizing their operations. Flexible FCC catalyst technologies that offer better coke selectivity and bottoms upgrading can help refiners overcome operating constraints, make a more valuable product slate, and optimize profitability. FCC catalyst innovation the last decade has largely focused on achieving these results through improved matrices, binders and additives, rather than through improvements to the zeolite component of the catalyst. Rive has focused its research on the zeolite component of the FCC catalyst and has developed Molecular Highway mesoporous zeolite technology for improved mass transfer into and within zeolite crystals. The enhanced porosity of the zeolite, when incorporated into FCC catalysts, allows FCC feed molecules to more readily access the interior of the zeolite, undergo the desired reactions, and then quickly exit, leading to improved selectivity namely, better coke selectivity, lower bottoms yields and higher yields of products such as gasoline, diesel, and light olefins, depending on the refiner s objectives. Last year Rive successfully trialed the first generation of Molecular Highway technology on a paraffinic VGO feed in the CountryMark refinery FCCU. This year Rive has successfully demonstrated the second generation of its technology, on a resid feed, at the Alon USA FCCU in Big Spring, TX. This paper will discuss Molecular Highway technology and the results from the trial at Alon s Big Spring, TX refinery, where a W.R. Grace & Co.-manufactured catalyst containing Rive zeolite was used successfully to achieve over $2.5/bbl value uplift in the FCCU. 1 Currently with KBR Corp., Houston, TX, USA 1

4 Introduction: In 211, Rive successfully trialed its first generation mesoporous zeolite, called Molecular Highway technology, at the CountryMark Refining FCC unit in Mount Vernon, Indiana. The catalyst demonstrated good: - hydrothermal stability - activity maintenance - attrition resistance and fluidization - coke selectivity and bottoms cracking, with an overall increase in transportation fuels. At the AFPM 212 Annual Meeting, Rive reported the development and manufacturing scale-up of a second generation (Gen II) version of the Rive zeolite. Since then, Grace was granted a permit for the commercial production of Gen II Rive zeolite, and 125 tons were produced for a second refinery trial. Using this commercially-produced Rive zeolite and Grace matrix technology, Rive/Grace developed catalyst formulations for ACE testing with Alon s feedstock. The ACE unit is an industry accepted tool for evaluating FCC catalysts in the laboratory and then predicting commercial performance. Our results were used to build an economic model for the Alon Big Spring refinery that predicted a $2./FCC bbl value increase over the incumbent catalyst, substantial enough to easily justify a commercial trial. Formulation Development: Based on learnings from Rive s first successful trial at CountryMark Refining, a second generation of mesoporous zeolite was developed (Gen II) for production at Grace s Valleyfield, Canada catalyst plant. Nominally 125 tons of Rive zeolite were produced, and this commercial zeolite was used for all formulation test work. Testing conditions were designed to simulate Alon s Big Spring operation. All catalysts were impregnated with nickel and vanadium and cyclic propylene steam (CPS) deactivated for testing in an ACE unit. Table I, below contains the results of ACE testing comparing Alon s incumbent catalyst (which incorporated both state-of the-art matrix metals trapping and matrix bottoms upgrading technologies) with Rive s proposed formulation (referred to as the Rive MH-1 catalyst) on Alon s FCC unit feed (22 API gravity, 1.6% CCR, and 2% sulfur). They show a substantial improvement in coke selectivity and a substantial increase in gasoline yield with the Rive catalyst. The anticipated increase in bottoms cracking at constant conversion was smaller than typical, but was much larger when converted to a constant coke basis (as was evident in actual unit operation). 2

5 Table I: Rive MH-1 catalyst showed lower coke and increased gasoline in ACE testing Catalyst 2 (MH-1) C/O ratio Conversion Yields, wt% Dry Gas wt% LPG wt% Propane wt% Propylene wt% Butanes wt% Butenes wt% Gasoline wt% LCO wt% Bottoms wt% Coke wt% The ACE test results were used in modeling (Profimatics ) the impact of Rive s MH-1 catalyst on the Alon FCC unit operation. The yields predicted 3 with a 1% change-out of the catalyst to Rive s formulation would provide a value uplift of $2./FCC feed bbl using the refinery s constraints and product pricing. It should be noted that the MH-1 catalyst s improved coke selectivity also predicted that optimized operation would occur at both lower reactor and lower regenerator temperatures. Grace manufactured 328 tons of MH-1 catalyst for Alon. Rive MH-1 had an average zeolite surface area of 227 m²/g, a matrix 4 surface area of 1 m²/g, and a Grace Davison Attrition Index (DI) of 6. The incumbent catalyst had a similar fresh zeolite surface area and a slightly lower matrix surface area. Catalyst quantity was sufficient for a 19-day trial (at 3 tons/day) and was projected to yield an 8% change-out. Unit Description: The Alon Big Spring, TX FCC unit is a UOP stacked design, revamped to include an external vertical riser with state-of-the-art feed injection nozzles. The riser terminates into a pair of primary cyclones that discharge the spent catalyst via dip legs in the stripper. The spent catalyst is subjected to steam for stripping absorbed product vapors before it flows down the spent catalyst standpipe into the regenerator. The riser product vapors exiting the primary cyclone gas tubes are quenched by LCO sprays. The quenched product vapors and stripping steam exit the reactor via a pair of secondary cyclones which further separate out entrained catalyst. 2 Excludes Grace gasoline sulfur reduction additive D-PRISM 3 Attachment 1 - Yields and operation predicted for Alon (Big Spring, TX) with a 1% catalyst change-out to MH-1 4 Proxy for mesoporous surface area 3

6 Coke on catalyst is burned off in the regenerator, which is operated in a partial combustion mode. Combustion air is supplied by three air blowers working in parallel. The flue gas flows through four pairs of cyclones which recover and return entrained catalyst to the regenerator bed. After exchanging heat in a flue gas steam generator cooler, the CO-rich flue gas is incinerated to CO 2 in a CO boiler. Cooled flue gas from the CO boiler flows through an Electro-Static Precipitator (ESP) before it is discharged to the atmosphere. The reactor vapors are fractionated into various products in the main fractionator and gas plant. The main fractionator operates with three side-draws (HCN, LCO and HCO 5 ) and a bottom draw (CSO 6 ) to provide improved column operation and fractionation. The LCO product (HCN + LCO) is routed to the diesel hydrotreater. Catalyst entrained with the CSO is recovered in a Gulftronic Separator, and the recovered catalyst is returned to the reactor riser. The overhead from the fractionator is routed to a wet gas compressor (WGC) and gas plant and fractionated into gasoline, LPG and fuel gas. The gasoline product is hydro-desulfurized to produce ultra-low sulfur gasoline. LPG is separated into C 3 s and C 4 s, and the C 4 olefins. Trial Description: The feed to the FCC unit is a mix of vacuum gas oils and mildly hydrotreated PDA 7 oil. Prior to the Rive trial, the unit typically operated with a riser temperature at or above 1 ⁰F, to maximize LCO minus conversion. Reactor product vapors were quenched with LCO to minimize dry gas rate. In addition the regenerator bed temperature was controlled to maintain a carbon on regenerated catalyst (CRC) below.3 wt%. The main constraints on the FCC unit are the air blower, wet gas compressor, catalyst circulation and main fractionator (in that order). Combined feed temperature was maintained to minimize coke yield. During summer (May September), the feed rate to the FCC unit cycles daily, varying by between 1, BPSD and 1,5 BPSD - inversely with ambient temperature due to the air blower limit. Rive MH-1 catalyst reached the refinery on June 28 th, and the trial began on June 3 th. Catalyst addition rate was 3 tons/day the same as the incumbent catalyst. One feed sample and three ecat samples were collected per week. All collected samples were shipped to Grace for analysis. As the Rive catalyst changed-out in the unit, coke selectivity improved and feed rate to the FCC unit steadily increased despite an increase in the ambient temperatures. Slurry yields decreased with an increase in slurry density due to improved bottoms upgrading. Based on the improvement in coke selectivity and improved bottoms upgrading, the refinery progressively increased feed rate and progressively lowered the riser and regenerator temperatures by a total of 2 degrees over a period of several weeks while maintaining less than.3% CRC. Towards the end of the trial, fresh feed to the unit 5 HCN = Heavy Cat Naphtha; LCO = Light Cycle Oil; HCO = Heavy Cycle Oil 6 CSO = Clarified Slurry Oil 7 PDA = Propane de-asphalted oil 4

7 had been increased by 2, BPSD, albeit part of the increase was due to a lowering ambient temperature. However as the data analysis shows, the improved coke selectivity of the Rive catalyst allowed the refiner to process an additional 7 BPSD over the prior operation at similar ambient temperatures. In general, the FCC unit operation was steady during the trial. Feed Properties: The FCC feed properties of API gravity and Conradson Carbon residue (CCR) are shown in Figures 1 and 2, respectively. The gravity of the feed processed during the trial ranged within the norms for this unit. However the CCR during the trial was higher than typical Feed API Gravity Feed CCR, wt. % Aug11 3Nov11 1Feb12 1May12 3Jul12 28Oct12 Figure 1 FCC Unit Feed API 1..8 Figure 2 FCC Unit Feed CCR 5Aug11 3Nov11 1Feb12 1May12 3Jul12 28Oct12 Equilibrium Catalyst Analyses: Ecat samples were analyzed by Grace to monitor the catalyst activity, zeolite & matrix surface areas, coke & gas factors, and physical & chemical properties. The catalyst change-out shown in Figure 3, is calculated on the basis of catalyst chemical composition. It rises normally with time until there is a dip after a unit shutdown towards the end of the trial when ecat from early in the trial was added to the unit. Following unit startup and Catalyst Change-Out, % May12 26Jun12 26Jul12 25Aug12 24Sep12 24Oct12 Figure 3 Catalyst Change-Out 5

8 resuming normal catalyst additions, the catalyst change-out curve was re-established, albeit with a break. A regression fit to this data prior to the disturbance (dashed red line) projects a catalyst change-out of 78% in the absence of the ecat addition. The change in metals (Ni and V) on ecat during the trial is shown in Figure 4. Nickel on ecat steadily increased from 12 to 16 ppm. Vanadium increased from about 19 ppm to 25 ppm over the same period. 3, 8 Ecat Metals, pmw 2,8 2,6 2,4 2,2 2, 1,8 1,6 1,4 1,2 Nickel Vanadium Delta Regenerator Temperature, o F Dilute Phase Dense Phase 1, 18Jan12 28Mar12 6Jun12 15Aug12 24Oct12 Figure 4 Ecat Metals (Nickel and Vanadium) Figure 5 shows the change in regenerator temperature during the trial and it shows the reduction which occurred during operation on the Rive MH-1 catalyst. During the trial, the Rive catalyst addition rate was maintained at the same level as the prior catalyst. Despite a greater than 3% increase in ecat contaminant metals (Ni & V) and periods of high temperature operation, the Rive catalyst Zeolite Surface Area, m 2 /gm Jan12 28Mar12 6Jun12 15Aug12 24Oct12 MAT Activity, wt% conversion -4 31May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct Figure 5 Regenerator Temperatures 18Jan12 28Mar12 6Jun12 15Aug12 24Oct12 Figure 6 Ecat Activity demonstrated good hydrothermal stability and maintained activity (Figure 6). With no significant differences in fresh catalyst zeolite and matrix surface areas between the incumbent and Rive catalyst, no changes were expected in the in-unit surface areas. The stability of the Rive catalyst zeolite and mesoporous (matrix) surface areas is shown in Figures 7 and 8. Figure 7 Ecat Zeolite Surface Area 6

9 The Rive catalyst also demonstrated a reduction in the coke 8 and gas 9 factors (Figures 9 and 1). Flue gas opacity (an indicator of catalyst losses) was steady as the Rive catalyst replaced the prior incumbent catalyst during the trial (Figure 11). The opacity meter was reset (calibrated) early in the trial and that prevented comparisons with prior opacity measurements. Spent catalyst withdrawals and disposal were normal during the trial leading to the conclusion that the Rive catalyst s in-unit retention and attrition resistance were similar to that of the prior incumbent catalyst. Matrix Surface Area, m 2 /gm Jan12 28Mar12 6Jun12 15Aug12 24Oct12 Figure 8 Ecat Matrix/ Mesopore Surface Area INCUMBENT CATALYST RIVE CATALYST 3. INCUMBENT CATALYST RIVE CATALYST Coke Factor Gas Factor Jan12 28Mar12 6Jun12 15Aug12 24Oct12 Figure 9 Ecat Coke Factor The in-unit catalyst fluidization characteristic, as measured by the ratio of the minimum bubbling velocity to the minimum fluidization velocity (U mb /U mf ) was constant during the transition from the incumbent to the Rive catalyst and during the trial (see Figure 12). The unit did not have any circulation issues, but issues with the regenerator flue gas analysis caused calculated circulation rate to appear constant when, in fact, it rose during the trial in response to improved coke selectivity (manifesting as constant conversion at increased feed rate and lower reactor temperature). Flue Gas Opcaity, % 1. 18Jan12 28Mar12 6Jun12 15Aug12 24Oct12 Figure 1 Ecat Gas Factor May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Figure 11 Flue Gas Opacity 8 Coke factor is defined as coke yield per unit of kinetic conversion 9 Gas factor is the ratio of Hydrogen to Methane in the Dry Gas yield 7

10 Yields: A hallmark of Rive s mesoporous Molecular Highway zeolite is that it provides improved coke selectivity over current FCC catalyst technologies employing conventional zeolites. Refiners with FCC units constrained by air blower rate 1 take advantage of improved coke selectivity in a variety of ways (singly or in combination) through the ability to: Catalyst Circulation, TPM Fluidization Parameter, Umb/Umf - increase FCC unit feed rate - raise conversion severity - lower regenerator and riser temperatures - process heavier feedstocks Jan12 17Apr12 16Jul12 14Oct12 Figure 12 Catalyst Fluidization Parameter and Circulation The FCC unit at Alon Big Spring is constrained by air blower capacity, particularly in summer. The constraint is severe enough that during summer, the FCC feed rate cycles daily in sync with the ambient temperature, as air density impacts the air blower rate. Consequently the refinery builds up an inventory of unprocessed FCC feedstock over the summer. Alon Operations was able to consistently and continuously increase feed throughput, even at the height of summer. Figure 13 shows that the unit processed a higher feed rate while on the Rive catalyst at a given ambient temperature. On average while on the Rive catalyst, the FCCU was processing 2,5 over 7 BPSD of additional feed. Post 5% Catalyst Change-Out At the start of the trial with the higher riser 1,5 temperature used with the incumbent catalyst, CSO (bottoms) yield and gravity decreased to 1, near the gravity limit. Alon Operations was able 5 to take advantage of the improved bottoms cracking of the Rive catalyst and lower the riser temperature without exceeding pre-trial -5 bottoms yields. Simultaneously, regenerator Mean Ambient temperature, ⁰F temperature fell without affecting the carbon on Figure 13 FCCU Feed Rate Increased by 7 BPSD regenerated catalyst. As a result of these changes CSO gravity remained within limits while bottoms selectivity remained low and the split between gasoline and LPG was shifted to favor more valuable gasoline. Delta Feed Rate, BPSD 2, An informal refining survey suggests that about 6% of FCC units are constrained by air blower rate 8

11 Eventually, riser temperature was lowered by 2 ⁰F while the regenerator bed temperature decreased 3 ⁰F in partial burn at <.3% CRC (Figures 14 and 15). Typically a 1 ⁰F drop in riser temperature lowers regenerator bed temperature.8-1. ⁰F. During the trial the regenerator bed temperature clearly dropped by more than this typical value, indicating that it was due to the improved coke selectivity of the Rive catalyst Delta Riser Top Temperature, degf Delta Regenerator Temperature, degf Dense Phase Dilute Phase May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Figure 14 Riser Top Temperatures -4 31May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Figure 15 Delta Regenerator Temperatures As the trial progressed, the combined impact of increased feed rate from better coke selectivity and improved bottoms cracking, and despite the lowering of riser temperature, the refiner observed a significant increase (~ +1, BPSD) in gasoline production, and an increase (~ +4 BPSD) in LCO production (Figures 16a and 16b). The LPG/gasoline split shifted towards more gasoline (~ -4 BPSD LPG) and CSO production was constant (~ +5 BPSD). Delta Gasoline Rate, BPSD 2, 1,6 1, May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Gasoline Octane: 1,2 1, With the goal of developing data for an optimized operation on the Rive catalyst, the riser and regenerator temperatures were lowered and unit responses observed during the trial. A key response from lowering the riser temperature in a FCC unit is a drop in gasoline octane. During the Rive catalyst trial, the riser temperature was reduced by 2 ⁰F, and Figure 17 shows the impact of riser temperature on gasoline octane during Delta LCO Rate, BPSD May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Figures 16a & 16b Rising Gasoline and LCO Rates with Rive MH-1 Catalyst

12 normal operation in the Alon Big Spring FCC unit. Octane shows a linear increase with riser temperature. However the magnitude of the change is smaller than typical. One would expect a.4 (R+M)/2 octane response for a 1 ⁰F change in riser temperature. The observed response with Rive MH-1 catalyst is only.15 (R+M)/2; less than half the typical response. One possible explanation is that Rive catalyst produces more olefinic gasoline which partially offsets the effect of a reduction in riser temperature. More olefinic gasoline has been observed in ACE testing with Rive zeolite, but Alon FCC gasoline was not tested for this change. FCC Gasoline Octane (R+M)/ Jul -Oct '11 - Jun '12 Octane =.1x(RiserTopTemp) Delta Riser Top Temperature, o F Figure 17 Octane Response to Riser Temperature Figure 18 examines the impact of sulfur levels of gasoline feed to the ultra-low sulfur gasoline unit on the low sulfur gasoline product. It shows that catalyst type had no discernible impact on the low sulfur gasoline product octane. Butylene Yields: ULSG Octane (R+M)/ Jul -Oct '11 - Jun '12 LPG yields and composition are another key response to riser temperature an increase in riser temperature increases the LPG yield and olefinicity of the LPG. While the propane/propylene market value in the US is depressed due to the current excess of natural gas 11 in North America; butylenes are required as alkylation unit feed. Since alkylate is a prime high-octane gasoline blend component it is important to maintain adequate butylenes feed to the alkylation unit from the FCC plant. Butylenes production rates during the trial were similar to the pre-trial rates despite operation at a lower riser temperature. Figure 19 provides the impact of riser temperature on butylenes Delta C₄ Olefins, BPSD FCC Gasoline Sulfur, wt% Figure 18 Gasoline Sulfur and Low Sulfur Gasoline Octane 2, 1,5 1, 5 Jul-Oct '11 Jun ' , Delta Riser Temperature, o F Figure 19 Riser Temperature versus Butylenes Production 11 Excess due to Shale Gas production 1

13 production for the Rive catalyst and the incumbent catalyst. The data suggest that at a given riser temperature, the Rive catalyst improved butylenes production compared with the incumbent catalyst. Value Uplift: The unit operating data and production rates were heat and mass balanced with Profimatics. From the yields and selectivities discussed in the sections above, it is clear that the product rate changes are a combined result of both increased feed rates and yield selectivity improvements. The incremental FCC unit value uplift on a daily basis due to the change to Rive catalyst is shown in Figure 2. The value uplift was calculated by applying the April $4. 12 product pricing to the mass balanced product rates from the $3. $2.74/bbl Profimatics model and subtracting the average FCC value $2. uplift on the incumbent catalyst during the month prior to the $1. trial. The FCC value uplift $. progressively increased as the Rive catalyst changed out in the -$1. unit. Incremental Value Uplift, $/bbl Trending the value uplift, it is estimated that at the end of the trial the incremental uplift due to the Rive catalyst was over $2.5/FCC bbl. -$2. 31May12 3Jun12 3Jul12 29Aug12 28Sep12 28Oct12 Figure 2 Rive Value Uplift Conclusions: 1. FCC catalyst combining Rive mesoporous Molecular Highway zeolite and Grace matrix technology delivered over $2.5/FCC bbl of additional value in mild resid operation 2. Performance benefits resulting from Rive s unique zeolite were substantially improved coke selectivity, improved bottoms upgrading, and increased olefinicity of the cracked products 3. No capital investment was needed to realize this value, and operational changes were within the normal range of practice. 11

14 Acknowledgement: The authors wish to acknowledge the work and support of the Alon (Big Spring, TX) refinery staff in making the trial a success. In particular, Gordon Leaman, Ted Tarbet, Clarence Palmer, Jeff Brorman, Eric Selden, Manoj Katak and the FCC Operators all made valuable contributions to the trial. The author also wishes to acknowledge with sincere appreciation the entire team from W.R. Grace for their collaboration, expertise, and support of the research and development described in this paper, and Allen Hansen, Steve McGovern, Ken Peccatiello, and the other advisors to Rive Technology for their contributions. 12

15 Attachment 1 Yields and Operation predicted 12 for Alon (Big Spring, TX) Catalyst 13 Incumbent Formulation (MT-4) Rive Formulation (MH-1) Key Op. Conditions Riser Temp, F Feed Temp, F C/O, wt/wt Regen Bed Temp, F Carbon on Regenerated Catalyst, wt% Commercial Yields Dry Gas, wt% Total C3s+C4s, wt% C3+C4 non-olefins, wt% C3+C4 olefins, wt% Cat Gasoline, wt% LCO, wt% Btms, wt% Coke, wt% Commercial Yields, Vol% Total C3s+C4s, vol% C3+C4 non-olefins, vol% C3+C4 olefins, vol% Cat Gasoline, vol% LCO, vol% Btms, vol% Total Liquid Vol% Yield: Conversion, vol% LCO Conversion, vol% Rive Uplift, $/bbl FCC feed NA $2. Product Properties: Cat Gasoline: SG RON MON (RON+MON)/ Sulfur, wt%.41.4 LCO: SG Sulfur, wt% Btms: SG Sulfur, wt% Based on a 1% catalyst change-out 13 Excludes Grace gasoline sulfur reduction additive D-PRISM 13