TECHNOLOGY Refiners focus on FCC catalysts at Q&A meeting

Moderator: Terrence L. Higgins, technical director, National Petroleum Refiners Association, Washington, D.C.

At the most recent National Petroleum Refiners Association question and answer session on refining and petrochemical technology, refiners, suppliers, and other industry participants discussed issues surrounding fluid catalytic cracking (FCC) catalysts.

Among the topics addressed in the just released transcript were:

• Monitoring techniques for fresh FCC catalyst

• Catalysts' effect on the sulfur content of FCC naphtha

• Disposal of FCC catalyst fines

• FCC SO_x-reduction techniques.

At this renowned meeting, held Oct. 4-6, 1995, in San Antonio, a panel of experts answered presubmitted questions they researched in their refining networks. These questions are on process, mechanical, catalytic, and support issues related to refinery and petrochemical plant operations. Audience members then related their own experiences, or asked further questions on a given topic.

The panelists are identified in the accompanying box and pictures. Following are the questions and answers.

Catalyst monitoring

How is the quality of fresh FCC catalyst monitored by the refiner? What is the frequency of testing? What are the properties monitored and what tests used? What is the recourse for quality problems?

Solis:

We have developed a very extensive process of selection before any changing of catalyst. This process is designed to choose the optimum catalyst for our feedstock and unit operating conditions. Once we have selected the vendor, we include a clause in the purchase order that we have the right to take samples of the fresh catalyst that they supply.

The quality control program to carry out with the samples has been previously agreed upon with the catalyst producer. If the sample does not meet specifications, we can reject the truck or ask for liquidation damages.

The test always includes microactivity test (MAT) activity (72%); Na2O (<0.45%); and attrition index (DI<8). Be sure to crosscorrelate, as well as possible, the analytical methods of the refiner and the catalyst supplier. If you do not, you will have serious problems comparing results.

One of the major concerns of this program is sampling. We have developed a procedure in which we take one sample from the truck for the supplier, one as reference, and one for testing. We have developed this program over 6 years, and we have only observed problems on two occasions. One was low attrition and the other was a case where they sent the truck to the wrong refinery. Generally speaking, we are very satisfied with the quality of our supplier.

Parker:

Generally, we monitor the fresh catalyst using the data sheets from the vendor. We receive most of the tests listed on the data sheets for each shipment, with the exception of attrition, MAT, and surface areas. Those are only done occasionally-maybe every third or fourth truck.

We do compare these numbers to the specifications we have for our catalyst. If there is a problem, we call the vendor and usually we are compensated on the next shipment or by making a price adjustment.

We have really not had many problems. One problem that we have had occasionally is that we may be getting material that is within specifications, but continuously on the low side of the range on a particular property. We call and usually get an adjustment toward the target value.

Once in a while we will put equilibrium catalyst (ECAT) in our resid cracker. We have had instances where we had to reject shipments based on samples, so we do test our ECAT when we receive it.

If we suspect a problem, or if there is a discrepancy with a vendor, we conduct testing at our research center. They can test all the properties, and, typically, before we run any new catalysts in the unit, they conduct a full range of fresh and deactivated catalyst property tests. This provides a reference point for specifications and vendor tests on the new catalyst because there are such discrepancies between the different vendor tests.

Ross:

Some refiners rely on the quality assurance programs and certifications of the catalyst vendors, whereas others are more proactive. Typically, fresh catalyst is sampled randomly or each batch received is sampled, with samples retained but tested randomly by the refiner's central laboratory or research and development facility.

Properties such as pore volume, surface area, chemical composition (zeolite, alumina, rare earth, and promoter/platinum content), attrition index, and particle-size analysis are performed on perhaps 20-30% of the samples. One refiner cautions that particle-size distribution can be tricky as sampling a large truck is not trivial due to settling/segregation during transit.

Sophisticated, high-volume users may request vendor process-control charts and trend manufacturing-control points and variables. Average values of variables can be monitored, as well as deviations from the manufacturer's set points, to track consistency and control of the process to identify irregular batches.

As for ECAT analyses, samples are taken either weekly or twice a week for analysis by the vendor. Response times of about 1 week are reported.

When unit performance changes unexpectedly, samples are sent to the research and development or central laboratory for independent analysis of MAT, metals level, chemical composition, etc., and perhaps operation in a pilot unit using the feedstock from the unit in question. Results of these tests are then discussed with the catalyst vendor, possibly resulting in compensation. Good sampling and sample retention procedures are essential.

Abrahams:

Our practices are similar to those of Mr. Parker. We rely heavily on our supplier's analysis and crosscheck periodically. We also have been very successful in dealing with the vendors directly for recourse on the rare occasions we have had problems.

There was one instance a few years back, though, in which a supplier was unable to deliver suitable quantity and quality of a new catalyst. While it was painful, we parted ways for a couple of years and went with another supplier.

Deady:

A handful of our customers continue to monitor fresh catalyst properties. But most of that has been discontinued due to refinery and research and development cutbacks.

The refiners who still do their own analyses analyze every shipment. The properties they monitor include chemical and physical properties, especially particle size and attrition, and catalytic testing to measure activity and coke selectivity. Some refiners without internal laboratory facilities have sent random samples for third-party testing.

One important caveat before starting this kind of testing would be to establish a correlation with your supplier's laboratory analysis. This would eliminate potential laboratory biases due to analytical differences, and allow you to establish a baseline, should discrepancies arise in the future. The refiner also needs to make sure they obtain a representative sample for testing.

Hansen:

We also generally rely on the quality control sheets provided by the catalyst suppliers. In our case, the items of major interest are the fresh activity, rare earth, particle-size distribution, and hardness.

We have found the need, on occasion, to send samples to third parties to confirm shipment quality, and, like most, we would change suppliers if we felt we had a problem.

William D. Henning (Conoco Inc.):

We monitor fresh catalyst quality in a central technical support group and laboratories. We collect both vendor lot samples and truck samples of the fresh catalyst at the refinery.

Typically, each truck sample is tested for composition by X-ray and for surface area. For some units, we also conduct attrition and particle-size distribution tests.

We have found in the past that some incidents of FCC performance loss have correlated to catalyst quality. We have worked closely with our catalyst vendors and have put in place quality-control agreements for various composition and physical properties.

There are specified ranges for each shipment and a narrower range for a six-truck moving average. There are economic penalties associated with falling outside of some of these ranges.

I might add that a refiner seeking a formal quality agreement with their catalyst vendor should understand not only what catalyst properties are important to their unit, but also the capabilities and driving economics of the FCC catalyst manufacturing process.

Steven K. Pavel (Coastal Catalyst Technology Inc.):

Previous responses illustrate the importance of testing refinery fresh-catalyst deliveries to avoid getting a rogue delivery of a catalyst meant for another refiner, or a catalyst that does not meet specifications. Most often problems are detected in the unit.

To eliminate fresh catalyst from the list of potential contributors, testing and analysis are essential. In any case, units with high addition rates relative to the size of the circulating catalyst inventory are most sensitive to fresh catalyst property changes.

Fresh catalyst shipments are routinely sampled and tested. Delivered fresh catalyst properties are compared to specifications, shipment inspection reports, and previous deliveries.

Minimum testing includes X-ray fluorescence (XRF) and X-ray diffraction (XRD), for complete elemental analysis, and crystallinity on each delivery. Profiles on new catalysts might include: XRF; XRD; surface areas, pore volumes and pore-size distributions by nitrogen and mercury; pore volume by water; bulk and skeletal density; particle-size distribution and attrition; toxicity characteristic leaching procedure; and full MAT-yield tests, unsteamed, and after steaming for 4 hr and 16 hr. If a fresh delivery shows a significant change in XRF or XRD results, more testing is performed.

Analyses of feedstock elements usually do not include all the deactivating contaminant elements in the feedstock, which are deposited on the fresh catalyst either routinely or as a spiked anomaly. Fresh catalysts are a source of the spent-catalyst elemental constituents; without analysis of the fresh material it is difficult to calculate the deposition of contaminant elements.

Fresh catalyst elements are analyzed by XRF, and, in certain cases, by inductively coupled plasma (ICP). X-ray diffraction is used to measure unit cell size and crystallinity, and to provide a qualitative analysis of mineral composition or phase. In several instances, the X-ray diffraction analyst was the first to identify changes in the fresh catalyst composition.

It is generally accepted that "any system under observation changes." The occurrence of quality-control situations can only be determined by sampling and testing.

Peter G. Andrews (Consultant):

Most refiners have operated for a long time without testing fresh catalyst. Terry, it would be of interest to know how many refiners are now testing catalyst all the time.

Higgins:

Could we have a show of hands of who tests fresh catalysts all the time? [The audience indicates five or six.]

FCC naphtha sulfur

What level of sulfur reduction in FCC naphtha is possible by changing FCC catalyst? What is the impact of this change on other FCC yields and properties, especially the coke yield? Are there any other operational impacts (e.g., increased flue gas SO_x emissions)?

Deady:

This year we completed two refinery trials of our gasoline sulfur reduction (GSR) technology described in recent NPRA papers. This additive technology converts gasoline sulfur species to H₂S.

In the commercial trials, we observed up to a 25% reduction in full-range-gasoline sulfur. This reduction was corrected for changes in feed sulfur and gasoline end point. We saw no other changes in any of the other yields or the gasoline octanes. Further commercial testing is scheduled while work continues on improving the performance of GSR.

Regarding conventional FCC catalysts, we have seen that changing from a low-hydrogen-transfer catalyst to a high-hydrogen-transfer catalyst will reduce gasoline sulfur by about 6%, based on our circulating-riser pilot plant testing.

Matrix surface area also impacts gasoline sulfur. In MAT testing, we have seen a correlation in gasoline sulfur reduction with the steamed matrix surface area of catalysts. The range of matrix surface area tested was from about 20 to 125 sq m/g. The higher-matrix-surface-area catalysts gave a 15-25% sulfur reduction, compared to the low-matrix-surface-area catalysts.

From past experience, the percent reduction we have observed in microactivity testing has been somewhat higher than what we have observed in riser testing. We normally attribute this to the difference in contact time between the MAT and the riser. In commercial units, we do not always observe a significant reduction in gasoline sulfur when the matrix surface area is increased.

The yield shifts associated with increasing matrix surface area depend on the specific type of matrix surface area, but can include better bottoms cracking with the penalty of increased coke and dry gas, particularly with very-high-matrix-surface-area catalyst.

If the FCC unit is constrained by the air blower or gas compressor, increasing catalyst matrix surface area may not be a practical option to reduce gasoline sulfur. That is the major reason we concentrated on the additive approach to reducing gasoline sulfur. The refiner avoids having to deal with significant selectivity changes in the base catalyst.

Johns:

I would cite some pilot unit testing at our Port Arthur, Tex., research facility. We have seen, on a variety of catalysts, up to a 10% reduction in FCC naphtha sulfur. Usually, high-rare-earth and high-matrix-activity catalysts tend to reduce the sulfur the most.

The sulfur is observed to go to H₂S and LPG products, rather than to flue gas. We have also commercially tested a proprietary sulfur reducing additive in one of our units.

Ross:

Circulating-pilot-plant work by one refiner has confirmed the 20% target reduction in full-range-gasoline sulfur when using these additives. But there seems to be the following trade-off with constant catalyst-to-oil ratio and riser temperature: i.e., a 3% absolute reduction in conversion, 1.5% absolute reduction in light catalytic naphtha, with the heavy naphtha unchanged, and a 10% relative increase in the coke yield.

Also, there is a reduction in the gasoline olefins, resulting in an octane loss of 0.6 RON and 0.4 MON. Obviously, the technology is new and developing, and these are only pilot-plant results reported by one refiner.

An important consideration is the conversion level at which the sulfur level reduction is attempted. At low conversion, with some aliphatic species containing sulfur in the gasoline, the ability to achieve the hydrogen transfer or react out the sulfur would seem to be much more likely than at high conversion, where most of the sulfur species would be aromatic.

When attempting to attack the aromatics, coke production will likely increase. Although the fine-tuning of sulfur levels may be possible via this route, at present, it would appear that feed treatment would be more effective.

As an interesting aside, in our short-residence-time cracking pilot-plant work, similar sulfur reductions were observed. A reduction in secondary reactions-possibly those creating stable aromatic sulfur species-has been theorized. However, there is also some dilution effect as the gasoline selectivity is increased at very short residence time.

Solis:

I will add new information on this subject based on recent pilot-plant work. Also, I have to mention that the European auto/oil program will result in a sulfur reduction in gasoline, and this reduction will very much affect the FCC naphtha as a component of the gasoline.

By changing the actual commercial catalysts (e.g., changes in rare earth [RE] content and/or different matrices), very little, if any, reduction of sulfur occurs. When using specific additives for sulfur removal, reductions up to 35% are achieved.

However, the extent of that reduction strongly depends on the feedstock (crude source, end boiling point [EBP], sulfur content, etc.). Generally speaking, we have observed a linear relationship between the FQP (feed quality parameter, which depends on feed density, volume average boiling point [VABP], S, and aniline point) and the feed sulfur content, and the naphtha sulfur reduction achieved (Table 1 [25651 bytes]).

Thus, in some cases, sulfur reduction is clearly below 15%. On the other hand, the present catalysts (additives) seem to simultaneously increase the coke and H2 yields. As a matter of fact, coke yield increases 0.5 wt % for a naphtha sulfur reduction of 15-20 wt %. Finally, although most of the sulfur is retained by the additive and hydrolyzed to H₂S in the stripper, some still goes to the regenerator, increasing the SO_x emissions.

G. Andrew Smith (Intercat Inc.):

In recent trials of Intercat's bottoms-upgrading additive, we have achieved reductions of gasoline sulfurs ranging from 20% to 30%, with a significant portion coming out of the heavy gasoline cut. Refiners directly attribute these reductions to the use of the additive, and not to cut point changes.

However, at this time, the results of the sulfur balances have not been released; most of the sulfur removed is assumed to be in the form of H₂S. The other benefits include improved bottoms upgrading and reduction of slurry viscosity. There were no changes in coke and C2- gas yield.

FCC fines disposal

What are the current dispositions for FCC catalyst fines? Are there any properties of the fines that restrict the method of disposal?

Deady:

The current disposition for FCC catalyst fines is similar to that for FCC equilibrium catalyst. You can either use it in cement or use it in asphalt as road base. For cement use, the cement manufacturers really do not care much about the particle-size distribution, since cement is fairly fine-even finer than FCC fines.

We currently have discussed this application with cement manufacturers both in the U.S. and in Europe. In Europe, there are some metals limits on the equilibrium catalyst. This would presumably be the same for the fines.

The metals levels are 4,000 ppm nickel, 8,000 ppm vanadium, and 1,500 ppm antimony. In the U.S., the major limit appears to be vanadium, at about 8,000-10,000 ppm. Cement manufacturers worry about the vanadium content because it could attack the brick in the calcines and cause deterioration.

As for using the fines in road base, we have experience with this application in Europe and over in the Asia/Pacific area. In Europe, they have restrictions on the particle-size distribution. What probably will cause a problem is that the fraction below 10 m must be minimized, so I am not sure how practical this use would be for FCC fines in Europe. There are metal limits in Europe as well: 4,000 ppm nickel, 7,000 ppm vanadium, and 1,500 ppm antimony.

In the Asia/Pacific area, we are not aware of any restrictions on the particle size, so fines are probably appropriate for this use. Again, there is a limit of about 7,000 ppm nickel on the ECAT or the fines.

Frondorf:

Current disposition of FCC fines from both our refineries is to landfill facilities. Periodically we explore disposition to cement kilns when economical and available.

FCC fines are classified on a characteristic basis per normal EPA guidelines such as metals, organics, etc., and, fortunately, not on a definition basis such as sewer sludges or waste water sludges. The characteristic classification has been nonhazardous industrial waste for typical fine properties of today.

Hansen:

We currently dispose of the fines caught in the underflow system with the equilibrium catalyst, which is sold to a local landfill. We have also disposed of this as a nonhazardous waste and have used it in cement manufacture.

We also collect fines through our wet flue gas scrubber system. These are collected separately and sent out of the plant as a nonhazardous waste.

Keller:

Our dry FCC fines are transported by pneumatic truck to a local cement manufacturer and "blown" into a dry hopper for later mixing. Wet FCC fines are transported to a different local cement manufacturer and bottom-dumped into a liquid holding tank equipped with a mechanical mixer.

Wet fines are segregated from dry fines, otherwise they would form a cake, which would be unusable at either facility. There are no other restrictions at this time.

SO_x reduction

Several SO_x-reduction technologies are available to the refiner (catalyst additives, scrubbers). How do refiners determine which technology to use? When using the catalyst additive approach are there any significant shifts in sulfur content of the various liquid products?

Solis:

Selection of flue gas sulfur-control technology is driven by the cost of the wet-gas scrubber and the cost of the SO_x additive. If the flue gas SO_x level is less than 700-900 ppm, assuming a requirement to reduce it to less than 300 ppm, the cost difference is about break-even.

If the SO_x level is higher than that, the lower operating cost of the scrubber will easily offset the higher initial capital investment. The scrubber investment is used as a credit for the electrostatic precipitator, which would not be necessary with the scrubber. Local economics may shift the break-even point, so it should be reviewed on a case-by-case basis.

For existing units, I feel that, unless a scrubber is in place, catalyst additives are the convenient solution. However, it should be realized that additives are not very active in partial-combustion-mode units. Again, additive efficiency and prices are highly variable, and cost-effectiveness tests in pilot plants are recommended.

Regarding the second part of the question, the sulfur that is removed is SO_x from the regenerator, and is converted to H₂S in the reactor. The catalyst additive has no significant effect on the sulfur distribution of the liquid products, but we have observed that some SO_x catalyst additives can slightly reduce the sulfur in the FCC naphtha.

Deady:

I agree with Mr. Solis, especially on his point that the break-even point is highly unit-specific. We found that it is unit-specific because the SO_x capture efficiency of SO_x-reducing additives such as DeSO_x is extremely unit-specific.

Based upon our experience, we see a slightly higher break-even point with DeSO_x technology. We see that the SO_x-reducing additives are most cost-effective if you have SO_x emissions below 2,500 ppm base.

Above about 3,000 ppm base SO_x, the flue gas scrubbing economics can become much more attractive. In general, because this is so unit-specific, we would recommend a commercial trial of SO_x-reducing additive just to quantify that performance.

As an example, the cost for the additive approach to reduce SO_x from about 600 ppm to about 300 ppm is about $4,000/day for a typical 50,000 b/d unit, with an additional capital cost of about $50,000 for a loading system. Based on feedback from our customers for the scrubber scenario, the operating cost would be about $9,000/day, with a capital cost of about $20 million.

This is in order to achieve the new source performance standard requirement of 90% SO2-emissions reduction, or less than 50 ppm SO2 emissions. The standards with the flue gas scrubber are a little different, because of the new source performance standard requirements.

We have not observed increased sulfur (or decreased sulfur, for that matter) in any of the liquid products with the SO_x additive approach. All the sulfur recovered from the flue gas leaves the unit as H₂S.

Frondorf:

Our Corpus Christi, Tex., refinery, operating on hydrotreated feed, does not have a flue gas scrubbing system. We use SO_x-reduction additive on an as-needed basis to meet final stack specifications.

The Lake Charles, La., refinery, operating on nonhydrotreated feed, uses SO_x-reduction additive on a regular basis to meet final stack specifications. The Lake Charles refinery is in the process of installing an FCC feed hydrotreater, and our plans at this time would be to continue the SO_x additive on an as-needed basis. We have no plans at this time to install a flue gas scrubber.

A spot reading from the Lake Charles plant for our SO_x additive reduction would be about 35/lb SO2 removed. Neither refinery has seen a significant shift in sulfur content of the liquid products following SO_x additive.

Keller:

We are currently using an SO_x-reducing additive, which is effective in reducing the emissions to below our limit of 99 ppm. We currently add a 40-lb dosage (110-ton catalyst inventory), which lowers the SO_x ppm about 20%.

This reduction typically lasts from 4 to 16 hr. We have not seen any significant shifts in the sulfur content of the liquid products with the use of the additive.

Ross:

I understand there is a broad difference of opinion on the breakpoint between the SO_x level and when the catalyst additives are useful. In some of the analyses we have done, it suggested somewhat less than 1,500 ppm, and perhaps closer to the 1,000 number we heard before.

Regarding its effectiveness in partial combustion, we have seen trials measuring an effectiveness of 50-70% in the first-stage, partial combustion regenerator of a two-stage regeneration resid unit.

Each SO_x-reduction technology has advantages and disadvantages which are site-specific. Issues to be addressed include:

• Required reduction in SO_x emissions

• Capital cost

• Operating and maintenance cost

• Compatibility with FCC process and ability to cope with catalyst fines

• Adaptability for dealing with other pollutants such as NOx and particulates

• Discharge issues such as water discharge, solids disposal, and opacity

• Plot space requirements.

Paul Sestili (M.W. Kellogg Co.):

To determine the most applicable technology for pollution control in a refinery, an overall analysis of the benefits of each system must be made. Kellogg recently completed a study comparing a wet-gas scrubber to a SO_x-reduction catalyst plus electrostatic precipitator, or ESP (Table 2 [40068 bytes]).

The wet-gas scrubber tends to have a higher capital investment, but lower maintenance costs and operator attention. The overall plot space requirement for the scrubber is higher; however, the purge treatment unit can be placed off-site. As a result, the on-site plot space requirement is much lower for the scrubber.

The catalyst-plus-precipitator option has a lower capital cost, but requires more maintenance and operator attention due to the ESP. The waste disposal for the scrubber is both solid and liquid, whereas the ESP waste is a solid dust only. In summary, it is important to look at the big picture when determining the most applicable pollution control system for a given refinery.

G. Andrew Smith (Intercat Inc.):

With the recent improvements in SO_x additives, refineries have a choice of purchasing an additive with equivalent performance to current deSO_x additives that costs 40% less, or purchasing ones with 40% greater pickup efficiencies at similar pricing.

We would also advise refiners that are considering their options to run a short test with a SO_x additive, like the NO-SO_x family, to determine their economics. It is my opinion that the economics will continue to favor the SO_x additives over mechanical scrubbers for years to come in many cases.

Morgan:

Once a scrubber has been approved by a state, your operating permit requires that the scrubber be in service. For that reason, if your scrubber has a malfunction, you may find it necessary to shut down your FCCU. From an operating standpoint, that, to me, greatly favors the deSO_x-type operation.

James D. Weith (Unocal Corp.):

Regarding flue gas scrubbers being part of the operating permit, per Mr. Morgan's comment, ESPs are part of the operating permit where we live, too.

Copyright 1996 Oil & Gas Journal. All Rights Reserved.