NPRA O&A-1 FCC CATALYSTS RECEIVE CONCENTRATED ATTENTION

As the dominant process in the refining industry, fluid catalytic cracking (FCC) and its catalysts received heavy attention at the most recent National Petroleum Refiners Association annual question and answer session on refining and petrochemical technology, held Oct. 3-5, 1990, in Philadelphia.

March 25, 1991

19 min read

At this renowned meeting, moderated by Terry Higgins, NPRA assistant technical director, representatives of the refining and petrochemical industries gather to discuss, in great detail, a wide range of topics. This meeting helps processors worldwide stay abreast of the latest technical developments and improve their operations.

Presubmitted questions are first answered by a select panel of refiners, petrochemical processors, and industry experts (see box). The audience is then invited to respond to those questions with comments or additional questions.

The first abstract from the recently released transcript of the Q & A session addresses the selection and use of FCC catalysts.

METALS PASSIVATION AGENTS

Relative to catalyst metals passivation please discuss:

The optimum antimony level compared to nickel content of equilibrium catalyst.
Experience with various types of antimonybased passivator formulations, comparing their relative effectiveness.
The relative effectiveness of nonantimony-based passivators currently on the market.

STEGELMAN: The antimony retained on equilibrium catalyst appears to be unit specific. The ratio of antimony to nickel on catalyst may, for example, be 0.25 in one unit and 0.55 in another unit. Yet both units can be achieving optimum benefits of passivation of metals with antimony.

The key guide to metals passivation is the concentration of metals in the oil feed. Antimony metals passivation agent is most effective when the addition rate is determined by the quantity of metals entering with the oil. All commercial antimony-containing metals passivation agents are economically effective. A Phillips pilot plant study of the relative effectiveness of antimony-containing metals passivation agents was presented at the 1982 NPRA Annual Meeting, paper AM82-50, Mar. 21-23, 1982, San Antonio. Antimony, tridipropyldithio-phosphate produced significantly higher benefits than other antimony compounds. Extensive information is included in the original paper.

Phillips refineries have only used antimony-containing metals passivation agents. Phillips R&D has evaluated various potential passivation agents. A ranking of 17 elements for reduction of hydrogen was published at the Washington meeting of the American Chemical Society in August 1990 (Preprints, Division of Petroleum Chemistry Inc., ACS Volume 35, No. 4, pp. 744-749).

WILSON: In answer to Part c, we have developed and currently are marketing a system in the catalyst that is designed to encapsulate and trap the nickel and reduce its dehydrogenation activity. This has been used quite extensively in Europe with a great deal of success. It is most effective at fairly high nickel levels. This is sold under the name Advanced Catalyst System.

RATERMAN: We have found the optimum antimony-tonickel ratio for most of our operations is in the range 0.3 to 0.5. This is usually dependent on catalyst type and age of metals on the catalyst. For the most part, we find that Phillips-recommended dosage rates are quite accurate. We have had experience with both water and hydrocarbonbased antimony additives, and have found both to be equally effective. With the hydrocarbon base, however, we have had some decomposition problems when using high feed preheats. These are generally above 500 F. Relocating the injection quill close to the feed nozzle, however, solved those problems.

We have tried commercially available rare earth-based additives in two commercial trials. The preliminary results have shown no real advantage to the rare earth additives over the antimony.

SCHAUB: Normally we see antimony optimum at a rate to attain antimony level on equilibrium equal to 0.25 to 0.4 nickel on the units we look at. I wanted to point out one interesting experience we have had with one of our foreign FCC units with a fairly high equilibrium nickel content at 7,000 ppm and a vanadium content that is relatively low compared to nickel, at 700 ppm.

The vanadium-to-nickel ratio in this unit, at 0.1, is roughly one tenth of what you would typically see. In this unit, it would seem as though higher antimony-to-nickel ratios are warranted to be optimum. Along these same lines, a paper was published in Industrial Chemical Engineering Research in 1988, entitled "Nickel-Vanadium Interactions on Cracking Catalyst," by Tatterson and Mieville of Amoco. Essentially what is being said in this paper, which this unit would tend to agree with on operating experience, is that there is some interaction between nickel and vanadium such that vanadium would inhibit some of the dehydrogenation activity of nickel.

With reference to the last part of the question, see Oil & Gas Journal, June 4, 1990, p. 81. This article discusses the performance of bismuth and would tend to confirm the experience of at least one of our clients. He claims that their bismuth is equally as effective as antimony.

BRANHAM:Antimony content of the equilibrium catalyst is not necessarily of primary importance since equilibrium nickel can be significantly deactivated by the process. Look at the nickel in the feed. In general, benefits can be obtained up to stoichiometric addition of antimony compared to fresh nickel in the feed. We find that optimum antimony utilization is in the range 0.5 antimony to nickel, or less. The Catlettsburg refinery RCC unit, when using an antimony additive, was found to require only a 0.2 antimony-to-nickel ratio on the equilibrium catalyst.

Research results have indicated that the type of antimony compound used is relatively unimportant. The most important parameters are accurate metering and distribution. These factors may be impacted by the specific formulations being used. Laboratory evaluation of several nonantimony passivators has shown them to have only limited effectiveness in our mode of operation, and these results have been confirmed in one commercial trial. Antimony passivation is generally easier to obtain. The alternatives require closer attention to operating and utilization techniques.

CARON: We target for an antimony-to-nickel ratio of 0.3 and adjust the antimony injection based on unit performance-hydrogen production, standard cu ft/bbl of feed in particular. Our experience indicates that at high nickel levels on equilibrium catalyst, say above 3,500 ppm, our antimony-to-nickel ratio is closer to 0.35.

We have used hydrocarbon-based and water-based antimony compounds. We did not see any performance difference. Water-based is less expensive.

We have only used antimony for nickel passivation. In our economic evaluation on choice of nickel passivator, antimony provides the most attractive economics.

GESICK:The Meraux refinery operates a solvent deasphalting unit, and consequently produces an FCC feed that contains typically 4.5 ppm nickel and 1.7 ppm vanadium. Hydrogen production is presently controlled with the addition of an antimony pentoxide solution into the raw oil. The dry gas hydrogen-to-methane ratio is the control variable used to control the passivation additions. The target ratio is 1.0 hydrogen to methane, and this produces about 120 standard cu ft of lean gas per barrel of feed. The nickel-to-antimony ratio on the equilibrium catalyst is held at 2.5.

We have run tests with other passivating materials: bismuth and Betz's proprietary chemical DM1152. The DM1152 suppressed the hydrogen production to the same 120 standard cu ft/bbl, but at a hydrogen-to-methane ratio of 1.44. The nickel in the feed was lower during this test also, about 2.5-3 instead of the 4.5 that is typical. Also the response time to the dosage change was a period of days for the DM1152, rather than hours for the antimony.

They also ran some bismuth passivating agent. It suppressed the hydrogen production to 140 standard cu ft/bbl, and averaged about 1.28 hydrogen-to-methane ratio. The required nickel-to-bismuth ratio was 2.0. In that case, they found that the bismuth formulation was very important for the program. The first blend they were using had a laydown efficiency of less than 5% and caused a lot of fouling in the slurry heat exchangers. A diluent change increased the laydown to about 45% and eased on the fouling problem.

RONGWe are processing residuum which is 75% yield on the crude, and contains 160 ppm nickel, 190 ppm iron, 3 ppm vanadium, and 1.2% asphaltene.

Even if we use antimony as a passivating agent in FCC catalyst and allow buildup of nickel content up to 9,000 ppm in equilibrium catalyst, replenishment of catalyst to contain nickel contamination would be huge. So we thought that we would coke the short residue and send the coker gas oil and vacuum gas oil to a once-through hydrocracker, followed by fluid catalytic cracking of the hydrocracked distillate. In this process, the only snag is that the coke would contain a lot of metals. Is this feasible?

WILSON: I guess it is certainly feasible. You would have to look at the overall economics, certainly, given that high level of very poor quality coke. What are you going to do with the coke?

HEROS DERGREGORIAN (Giant Refining Co.): The panel has been addressing the effectiveness of antimony on passivation. Could somebody address the question of personal hazard on entry when shutting down the reactor for maintenance purposes?

STEGELMAN: We have not, at Borger, used very high levels of antimony because right now, we are running a relatively low metal feed since we have our HDS unit. I would suggest that you talk to some of the Phillips' licensing people.

ANTHONY WITOSHKIN (Engelhard Corp.): A wide range of antimony-to-nickel ratios has been given by the panelists as being the optimum for their units. It is very easy for any refiner to determine his own optimum based on the data at his disposal. A plot of hydrogen or hydrogen/methane versus antimony/nickel ratios will aid in locating that point. The optimum occurs at the sharp break in the curve,

FREDERICK A. PETTERSEN: (Chevron Research and Technology Co.): Chevron has converted all of its own FCC metal passivation applications from antimony to bismuth-based passivator. Whereas antimony is on the EPA list of toxic chemicals, bismuth is not and it is very unlikely to be added in the future. All of our pilot plant data indicate that the process performance of bismuth and antimony are comparable. This has been confirmed in all commercial applications where the user has followed the use recommendations provided by the licensor.

DE-SOX ADDITIVE EFFICIENCY

What is the efficiency range for de-SO,, additives in partial combustion FCCU regenerators and full combustion FCCU regenerators? What unit factors contribute to improved performance?

WILSON: The actual issue affecting these de-SO,, additives is not so much whether you are in partial combustion or full combustion, but the effect on the reaction kinetics, chemical equilibrium, and the ability of the additive to regenerate in the riser-stripper system. De-SO,, additives remove sulfur oxides by absorbing and chemically bonding with sulfur trioxide. Thus the goal is to maximize production of this species in the regenerator. Normally, chemical equilibrium is not a factor. A well designed additive will remove the S03 from the gas and prevent equilibrium from becoming limiting.

Reaction rate on the other hand does affect the amount Of S03 available. De-SOx agents work best in an atmosphere of excess oxygen, and at temperatures that favor fairly rapid reaction of the S02 to S03. This generally accounts for the fact that additives in complete combustion regenerators range from 70 to 90% recovery or trapping of the SOx whereas in the partial combustion, they usually average about 50%.

Also the additive must be able to regenerate in the riser. This is important, since if it fails to do so, it will rapidly become saturated with the metal sulfides and simply cease to work. That will require additional addition to maintain the SOx removal.

Finally, you need to have an additive that is physically compatible with the circulating catalyst inventory, as it will do very little good if it is quickly lost out the stack.

RATERMAN: Our commercial experience with de-SO additives is limited to our units which operate in complete combustion. For the most part, the expected performance has been as advertised by the vendors. We have obtained SOx removal efficiencies for partial combustion operations in our pilot plant. These data also confirm the vendor's claims for reduced SOx removal efficiency when operating in partial burn.

SCHAUB: Data on SOx capture efficiencies are all over the map, due to the fact that many of the units using De-SOx are older units trying to meet new regulations. Catalysts in these units operate in a wide variety of regenerator environments that will significantly affect the capture efficiency. Some older units with inefficient reaction systems will also not regenerate De-SOx agents well, and will result either in lower SOx capture or an increase in the amount of additive required. I guess there is no doubt that typically the efficiencies are lower on partial combustion regenerators, ranging from 50 to 70%, whereas on the complete combustion units, efficiencies are from 50 to as high as 90%.

Some of the factors that affect this have already been mentioned. Lower temperatures in the regenerator would favor the capture as well as effective air and spent catalyst mixing and distribution, and oxygen level. This has become a perennial topic. Every year there seems to be at least one voice in the wilderness saying that some partial combustion regenerators have shown fairly high de-SO, efficiency; that it is more than just a question of excess oxygen. This is due to some of the reasons that have just been stated. I would like to refer you specifically in that regard to last year's discussion, especially from Petro-Canada.

STEGELMAN: All the information I have is on two tests on the same regenerator. In one case we were in partial burn, and the best we could do was 15% reduction of SO,; but in total burn we obtained about 33% reduction.

TIEMAN: We agree with most of what has been said. The main things that favorably affect efficiency are higher SO,, concentrations, increased oxygen, and a lower regenerator temperature. We have commercial operations on de-SOx that are in a complete combustion operation. De-SOx usage in a highefficiency FCC regenerator ranges between 8 and 25 lb of SOx per daily lb of de-SOx addition. In a partial burn regenerator, commercially we have observed efficiencies between 4 and 20 lb of SOx per lb of de-SOx addition. These efficiencies, again, are greatly dependent on the base SOx emission level and the percent reduction required.

MORGAN: We do not use de-SOx additives on a continuous basis, but we have conducted de-SOx test runs at two different refineries. Both units operated in a fully promoted combustion mode. One unit achieved about a 60% reduction in flue gas SOx from 250 ppm to 100 ppm. De-SOx additive efficiency was 8.2 lb of SOx removed per lb of de-SOx.

The other unit achieved a 77% reduction in stack SOx from 750 ppm to 170 ppm. Removal efficiency in this case was 15.8 lb of SOx removed per lb of de-SOx. In both cases the SOx removal efficiency was within the vendor's predictions.

UTLEY: In addition to what has been said, I do not think the percent reduction of SOx is a good measure of the effectiveness of SOx additives. I think a more accurate way to look at it is what concentration you can get down to in the flue gas, and that appears to be somewhere between 50 and 100 ppm.

One factor that has not been mentioned that contributes to good performance is adequate steam flow through the reactor stripper to hydrolyze the metal sulfides.

BRANHAM: Ashland has had very good results with Katalistics de-SOx, catalyst in a full combustion mode of operation. We have typically shown a 65-70% reduction in S02 emissions at current de-SOx, addition rates. The higher-than-expected efficiency of the de-SOx material in Canton's FCC unit has enabled us to reduce addition rates by nearly 65% from the original addition rate predicted by the process program.

I would agree with the factors that contribute to improved performance and note that Canton's regenerator typically operates at 1,310-1,315 F., and 1-2% oxygen. We do utilize a pipe grid air distributor.

CARON: We have tried de-SOx in a partial combustion unit. Our results have confirmed the estimate of approximately one third the performance achievable in a full burn regenerator. There was a paper presented by Joe Powell, et al., entitled, "Advanced FCC Flue Gas Desulfurization Technology," at the 1988 NPRA Annual Meeting. It provides interesting information on process variable effects.

MACKEY: I just wanted to comment on the low range of SOx removal efficiency. We operate our unit under 100 ppm SOx limits. We see 20 lb of SOx removal per lb of SOx additive. We use the Katalistics SOx additive, and that is pretty much in line with the vendor claims.

SCOTT ANDERSON (Koch Refining Co.): We recently ran a trial on a partial combustion unit and found that the de-SOx additive also promoted CO combustion. Has anybody else noted that effect?

DELBERT TOLEN (Akzo Chemical Co.): I think you should expect that if you have a good de-SOx additive that you are going to see increased CO combustion, because generally the limiting reaction in SOx shift is the S02 to S03 reaction. So, if you are going to get oxidation Of S02 to S03, YOU probably are going to get higher oxidation of CO to CO2.

CHARLES WEAR (W.R. Grace & Co.): Just a practical comment here in terms of performance. It was mentioned that air distribution was important. If a unit is going through operating changes in order to maximize SOx pickup using an additive, and you notice S03 increases, this is a very good indication, a good diagnostic that there is an air distribution problem. This, of course, is very important to those units that have S03 or sulfate emission regulations, as opposed to total SOx.

CATALYST QUALITY CONTROL

What steps are refiners taking to assure the quality of fresh and equilibrium catalyst receipts? Are statistical quality control charts required? Other methods?

RATERMAN: We currently track catalyst quality by sampling each of our units on a near-weekly basis, submitting duplicate samples to the vendor and our Paulsboro laboratory. New catalysts are first tested in our laboratory to establish performance characteristics. Based on this information, we will place manufacturing specifications on the catalyst. After that it is more or less up to the refinery to submit fresh samples for testing. Otherwise we rely on aging characteristics of a particular catalyst and the equilibrium samples to track fresh catalyst performance. A six-point running average of key equilibrium catalyst quality factors is used to keep statistical track of the catalyst quality.

BRANHAM: Ashland's procedure on fresh catalyst has been to require the manufacturer to collect and ship to us a sample of every truck and rail car as it is loaded. We randomly pick from these samples and analyze and compare them to the manufacturer's analysis to confirm that we are getting the specifications we asked for. At least one sample a month is examined. Statistical quality charts are not required, but we do have quarterly review meetings with the vendors to review the shipment analysis and control targets. We have sent round robin samples to the catalyst manufacturers for particle size distribution to ensure that the laboratory equipment is properly calibrated.

Equilibrium catalyst is usually purchased from a reputable vendor, and his desire to remain as our broker is the main control method.

CARON: Our R&D group coordinates the quality monitoring program for fresh catalyst shipments delivered to our FCC unit.

Testing is conducted on a random monthly sample from each location.

Physical and chemical properties are compared with vendor data and purchase specifications.

We maintain SPC [Statistical Process Control] charts on our testing, but we do not require them from the vendor.

One other thing, our FCCU operators check before they unload catalyst cars or trucks. Their visual inspection picked up bad problems on more than one occasion. So have some faith in your operators.

We have no program in place for equilibrium catalyst.

GESICK: We routinely sample our equilibrium catalyst purchases and send it to the various catalyst manufacturers for analysis. Constant surveillance of catalyst losses from the unit and watching the catalyst activity and unit performance is probably the best way to monitor the equilibrium catalyst on a daily basis.

MACKEY: We have the catalyst manufacturer report the actual physical properties of each shipment they send us. We then sample each fresh catalyst shipment right off the truck, and we also have a separate sample taken by the operator on duty, which we retain.

These samples are retained for 1-2 months. If we suspect a performance problem, we pull the retained sample we have kept and send it to Refinery Process Services (RPS) for verification of the catalyst properties. We have had a data base established with RPS for comparison. This has proven for us to be an effective manner of monitoring catalyst quality changes. We do not use SPC at this point.

WARREN LETZSCH: (Katalistiks International Inc.): We have an ongoing SPC program, like a lot of people do, and it is backed up by an extensive sampling program. All of our shipments to customers are accompanied by certificates of analysis, listing the results of the tests that are important to that particular customer. These analyses, by the way, are actual shipment analyses taken from the material being loaded into the individual truck or the individual rail car. It is not a hopper composite or something like that.

For the refiners who have more sophisticated SPC programs, we do provide control charts of the runs for that particular customer. In some cases, quality control teams from various refineries come in, visit the plant, and review the results. Otherwise, monthly or quarterly meetings are held at the refiner's convenience.

A couple of comments may be helpful with regard to equilibrium catalysts. If you are buying it directly from a refiner, it would be helpful to have an analysis of that material, but it would also be nice to have analysis of what is actually in his unit at the current time and see if that in fact is what you are actually getting. It is hard to tell what catalyst really is in an equilibrium hopper and how long it has been there.