Refiners discussed fluid catalytic cracking (FCC)-the workhorse of the modern refinery-in great detail at the most recent National Petroleum Refiners Association's annual question and answer session on refining and petrochemical technology.
Among the topics covered were the newest FCC refractory linings and particulate-control methods. The panel of experts (see box) also answered questions on the role of FCC in reducing gasoline benzene to meet reformulated gasoline specifications.
The 1992 NPRA Q&A Session was held Oct. 14-16, in Anaheim, Calif. For further details on this important meeting and its format, see OGJ, Feb. 22, p. 45.
REFRACTORIES
Several new refractories, some of which are vibracast, have been used to replace AA-22 and other refractories. Have you compared performance, ease of application, maintainability and dry out procedures for various services?
Quinn: There are two categories of refractories which are used in lieu of AA-22 for severe erosion service. The first category is castable-type refractory. The second category is premixed phosphate-bonded plastic.
The castable types were originally used and have been used because of the desire to convert hot-wall catalyst piping systems to cold-wall systems. This requires placement of enormous volumes of refractory, typically 4 in. thick versus the 1-in. thick AA-22. The castable refractories allow for easy placement, using cast-vibrated methods for installation.
To date, the performance of these linings has been very good. Maintenance experience is limited. Repair techniques include utilizing 4 in. of plastic-type refractory pneumatically rammed to full thickness.
After initial installation, the castable refractories are shop-fired and dried. We take added precautions to ensure adequate and thorough drying of the castable refractories. These linings must stay dry until they are placed in service. Precautionary start-up procedures are required to prevent excessive exposure to condensing steam, especially during cold weather start-ups.
Moving on to the phosphate-bonded plastics, we use 85-90% alumina content in thin layers in high-erosion areas. These refractories are anchored either in refractory grating or S-bar anchoring systems. These linings are typically used for both reactor and regenerator cyclones. They are shop-fired prior to shipment to the field. This type of refractory is also used for wear pads and target zones on FCC equipment. Refractories of the plastic type which are field installed require special consideration until fired and heated.
Depending on the location and quantity of refractory installed, we may elect to fire the plastic refractory prior to start-up. This task commonly is performed using electrical resistance heating coils or forced circulation of hot air. Condensing steam or water should not come in contact with the unfired plastic.
Considerable additional information on refractory systems can be found in the proceedings of the FCC maintenance seminar, which was held in August 1992 in Houston.
Warwick: AA-22 refractory still retains a strong following among refining companies, particularly for use in severe erosion applications, such as cyclones and air/catalyst risers where a thin (1/4-1 in. thick) erosion-resistant lining is required. For thin linings, similar refractory products such as Curas 90PF and plastics such as coral plastics have also been used.
These products are intended for installation by hand packing or ramming. All three are chemically bonded materials and have excellent erosion resistance compared to hydraulically bonded refractories. Compressive strength and modulus of rupture is very good. Heating to approximately 700 F. to set the ceramic bond is the only dryout required.
For applications requiring thicker refractory linings (4-5 in. thick), high density, erosion-resistant hydraulically setting refractories such as RS-88VC, LO-Abrade, AR-400, and Plicast Erozist, among others, have performed well. Linings of these materials are frequently found in risers, standpipes, and flue gas lines.
For the best quality installation, these products are usually installed by the vibracasting method. Vibracasting provides a very dense and uniform thickness quality lining that performs better than linings of the same materials applied by the gunning method. A complete and thorough dryout of these materials in accordance with the manufacturer's recommendations is required. Repair and removal of vibracast linings or those of similar materials can be very difficult and time-consuming, if not impossible.
Malik: In general, our experience at Corpus Christi with vibracast materials in the FCCU has been very positive. To date vibracast refractories have suffered minimal repairs and exhibit good on stream performance. Installation of these materials suffers only from the relative lack of availability of adequately trained installation technicians. Furthermore, installation of AA?? is time-consuming, labor intensive, and requires constant monitoring by inspection personnel.
Unfortunately, repairs of a section of vibracast refractory have proven more expensive than repairs to a section of AA-22, primarily due to increased demolition time. Lower failure rates for the vibracast refractory mitigates this issue somewhat. Dry-out procedures are comparable for both of these materials, each with no notable problems. As a word of caution, vibracast sections tend fo by 4 or 5 in. thick, vs. 1-in. thickness for AA-22, thus installation in an existing vessel/pipe can have unanticipated and/or undesirable effects on the inner diameter of the equipment.
Our refinery has also installed a dual material of heterogenous composition in the past, consisting of a softer thermal-refractory layer, about 3-4 in. thick, against the vessel wall, with 1-in. AA-22 installed on top of this material. The AA-22 hex-metal is welded to the flat plate on top of the standoffs required for a thermal layer.
This material combination has not worked well for us, with a majority of the problems consisting of the separation of AA-22 from the thermal layers (due to coke buildup), separation of the hex-metal from the standoff plates, and process fluid erosion of the soft thermal layer after loss/erosion/bypass of the AA-22 layer.
PARTICULATE CONTROL
Discuss the advantages, disadvantages, and relative cost-effectiveness of the following FCC regenerator particulate-control methods for opacity and total catalyst loss rate:
a. Third plus fourth-stage cyclones
b. Wet scrubber
c. Electrostatic precipitators (ESPs)
d. Baghouses.
Quinn: Third plus fourth-stage cyclones are not very effective if the first two stages are effective. The dust-loading is low, and the catalyst is very fine. As a result, effectiveness per stage for a third-stage cyclone is often about 50%, and lower for a fourth stage. Also the very low loadings and very fine dust have a tendency to bridge and plug the diplegs. The advantages of this type of system are that it will have a pressure drop of 10-20 in. of water and that it is the cheapest of the four methods to be discussed.
Wet scrubbing of the flue gas after two stages of cyclones will, depending on the design and operating conditions, effectively remove 90% of the remaining dust. The main disadvantage is it is the most expensive proven technology for particulate control. Handling the wastewater and the solid waste is also a problem.
The pressure drop across the scrubber, depending on design, will be about equal to cyclones. A further advantage of a scrubber is that SO,, control can also be accomplished in the equipment. There has been some work at developing new wet scrubber technologies that might be somewhat less costly, but these are not commercially proven.
ESPs are probably the most commonly employed technology for dust removal after the regenerator cyclones. ESPs are less than half as expensive as scrubbers, and have a lower operating cost. ESPs also have a very low pressure drop, usually only a few inches of water. Depending upon the dust loading and power density, an ESP can achieve up to 90% removal of catalyst from the FCC flue gas. The key to good ESP performance is maintenance, both routine and turnaround.
To our knowledge the use of baghouses on FCC flue gas has not been tried, except as test facilities on small side streams. Baghouse facilities would be expected to be expensive and employ ceramic or etched metal filters. Since dust removal would be by direct filtration, it would be expected that dust removal would be higher than 95%.
For the same reason, baghouses would be expected to have a high pressure drop of 50-70 in. of water. This would reduce FCC capacity and modify the unit pressure balance. When the baghouses would be switched for blowback to clean them, there would be a step change in the back pressure on the regenerator. If there was a failure of the valve switching between baghouses, or the filter elements were blinded with dust, back pressure might build, forcing air into the reactor or lifting the regenerator relief valves.
So, adding baghouses on the FCC flue gas will increase system complexity, increase the risk of overpressurizing the regenerator, increase the risk of sudden catalyst releases, cause bumps in the system pressure balance, and increase the risk of losing control of the pressure balance that provides the seal between hydrocarbon and air. If you get the idea that I would not really want to be inside the battery limits of an FCC with a baghouse on the FCC flue gas, you have understood my concerns.
Warwick: Opacity is a relative measurement of the degree of light scattering caused by particles in the gas stream being measured. The degree of scattering is a function of the mass of particles in the stream, and the ratio of the particle size to the wave length of the light source. Because of this, opacity meters tend to measure the quantity of small particles in the flue gas and as long as the particle-size distribution does not change, they can be calibrated to track the total quantity of solids. Understanding this is important to understanding which of the emission control devices will be most effective in reducing both opacity and losses.
a. Third state separators (TSS) are a cost-effective way to reduce overall emissions by 80-90%, depending on the type employed. However, they are most effective on particles larger than 10 ii, so they will not be as effective in reducing opacity. They are best used to protect expanders.
b. Wet scrubbers are between 90 and 95% efficient in removing solids and can incorporate other functions into their operation, such as SOx removal. These units are more expensive (Desox) than the other types listed and will create a very visible white plume if reheat of the flue gas is not employed.
c. ESPs are 95 + % effective in removing particulates. They occupy a plot space approximately equal to that required by third and fourth-stage systems together. Because they are more efficient in removing most particles down to the 1u size, they will effectively reduce opacity. They are also more expensive to operate than either a baghouse or a TSS system.
d. Baghouses are the most effective in removing particulates, being virtually 100% efficient. However, they require a large plot space and are prone to bag failure from a variety of causes including the repeated flexing of the bag and sulfide condensation, which causes the bag to stiffen. Because of their reliability they almost need to be 100% spared. This, in our opinion, along with the fact that they cannot take a flow reversal, makes them a poor choice for primary cleanup of the FCC flue gas.
Mobil uses baghouses as fourth-stage separators on the underflow from third-stage separators. We also have ESPs and TSS units as primary control devices singly and in connection with each other, and an Exxon scrubber for SOx and particulate control after a TSS.
Lavergne: Concerning third and fourth-stage cyclones, we currently operate a third-stage separator and are installing a fourth-stage separator this fall on our West Coast FCCU. The third-stage separator currently has a 70% dust removal efficiency. We will not be able to meet our opacity requirements without also running our ESP. The addition of the fourth-stage separator will drop the load on the ESP and as a result will improve our control. The separators require a small plot space, have no routine operating costs, and require little operator attention.
George: I would like to add the following information to what has been mentioned. Relative to a base fines emission (700-800 mg/normal cu m [Ncu m]) at the outlet of the regenerator, the following emission levels can be obtained:
- 3rd + 4th-stage cyclones: 60-80 mg/Ncu m
- ESP:
- Wet scrubber:
- Baghouse:
The selection depends on the local environmental legislation with respect to solid wastes, allowed emissions, and cost. Also, the installation of a power recovery unit and aspects like temperature and pressure drop will affect the selection of the cleaning system.
GASOLINE FEEDS
With the pending reformulated gasoline regulations, we are considering feeding high-sulfur and olefinic gasolines to the FCCU reactor. How should the gasoline be injected into the riser and what yield or property changes can be expected? Has anyone fed heavy naphtha when using ZSM-5? What were the yields?
Crisler: We have included naphtha from our delayed coker in the feed to our FCCU from time to time when we did not have any other destination for it. Our coker naphtha is high in sulfur, so we saw an increase in gasoline sulfur and a decrease in the octane, but not as much as if we had just tried to blend it off. Our preferred route for this is to hydrotreat the coker naphtha and include it in our feed to the reformer. I think a similar step for other refinery naphthas would be in order.
Davis: I have seen some commercial data for two units recycling FCC gasoline, with the amount of naphtha recycle being 23% of total feed in one case and 60% of total FCC feed in another. Both utilized a segregated riser for the naphtha, and observed an approximate 25% relative decrease in gasoline, a 50% relative increase in LPG, and increased motor octane of 23 numbers.
Gasoline-range aromatics increased by 30-50% and olefins decreased 20-30% on a relative basis. Additional coke yield was observed in one of the cases and was not reported in the other.
In the laboratory we compared the yield effects of (1) passing gas oil over an FCC catalyst, (2) passing gas oil over a 4 wt % ZSM-5 additive FCC catalyst blend (net 1% ZSM-5), and (3) passing FCC naphtha produced from the catalyst only over a 1 wt % ZSM-5-containing additive.
The effects of adding the ZSM-5 additive to the catalyst were those typically observed, with the yield effects limited to an increase in C3/C4 components at the expense Of C5+ gasoline. There was an approximate 20% decrease in gasoline-fraction paraffins with an increase in olefins and aromatics.
When the naphtha obtained from cracking gas oil was passed over the ZSM-5 additive, total C3/C4 COMPOnents increased; however, there was also an incremental C2 yield of about 1.5 wt %. Also there was only a 10% reduction in the gasoline-range paraffins, indicating that the effect on the FCC gasoline of ZSM-5 cracking is not as pronounced as that of cracking FCC feed over a faujasite/ZSM-5 additive blend.
Lavergne: Our pilot plant testing has demonstrated that naphthas will not crack significantly under normal conditions with or without ZSM-5. If you hope to desulferize them or upgrade their octane they must be injected into the riser before the feed is injected.
For example, in a Model 2 unit, they can be injected in the 'J' bend below the feed nozzle in place of fluidization steam. The octane upgrade depends on the temperature and contact time. As an extreme example, we have seen a 24 road octane upgrade on a 60 road octane naphtha at 1,200' F. and 8-sec contact time. This kind of contact time and temperature are not likely in a conventional riser. These severe conditions also converted half the naphtha to C4-minus gas.
Turpin: Naphtha fed to the FCCU will crack based on its composition. Full-range paraffinic naphthas will probably crack to a greater extent. Heavy, narrow-boiling-range aromatic cuts will probably not crack at all. We would expect that recycling high-sulfur, aromatic, heavy-naphtha cuts will no! in itself improve overall FCC naphtha quality very much.
Naphtha injection will have one major indirect side effect that may improve FCC yields. That is, injecting naphtha will raise the catalyst-to-oil ratio and lower the bed temperature as the unit heat-balances around a naphtha feed that produces little or no coke.
The unit will have to raise catalyst-to-oil ratio to make more coke from the other feedstocks to heat-balance the unit. The higher catalyst-to-oil will raise overall conversion. Also, if the unit is constrained on bed temperature, naphtha injection will tend to relieve this constraint and allow more heavy feed to a higher-severity operation.
Warwick: The most significant impact of feeding a high-sulfur olefinic gasoline like coker naphtha is a pronounced drop in FCC gasoline octane.
We have used FCC heavy naphtha (HN) as both a recycle to the base of the riser and as a quench material. In either case, an increase in FCC gasoline yield and octane was observed. The following results were obtained during a commercial test run:
- 5% HN recycle: +0.9 vol % gasoline, +0.3 RON
- 5% HN quench: +1.7 vol % gasoline, +0.6 RON.
Benefits will increase for larger quench volumes.
Recently, we have also commercially tested HN quench and ZSM-5 simultaneously. We have found that the benefits from this combination of technologies are greater than for the technologies employed separately and are currently in the process of quantifying this effect.
Hildo Francisco Henz (Petroleo Brasileiro S.A.): We performed several trials cracking naphthas in pilot plants and commercial units. To convert the smaller naphtha molecules, high cracking severity is required. Then the unit should operate in the high catalyst-to-naphtha ratio, high reaction temperature, and high catalyst activity. The naphtha must be added to a part of the riser where these conditions are achieved. Or you can build a special riser for this purpose. The properties and yield changes are directly related to the conversion level.
Generally speaking, we have seen a great conversion to olefins and an increase in aromatic concentration in the final gasoline product, which resulted in an octane increase. In one experience cracking 12 vol % high-sulfur coker naphtha relative to the feed rate in a segregated riser, we observed about 40% conversion and 50% sulfur reduction. We solved the problem of the coker naphtha stability and we did not see any decrease in the final octane of the gasoline pool.
BENZENE REDUCTION
What is commercial experience with benzene levels in FCC gasoline? What can be done to reduce benzene levels, such as changes in feedstock, FCC catalyst, hydrotreatment of the feed, additional downstream processing, and changes in FCC unit operating conditions?
Malik: The benzene level of our FCC gasoline averages 0. 6-0.7 vol %. We do not adjust unit operating conditions or upstream hydroprocessing to compensate for variability in benzene levels. We do monitor the level of aromatic saturation in our gas oil hydrotreater.
The absolute level depends heavily on the feedstock quality and unit severity. Additionally, to a lesser degree, it also depends on hydrogen-transfer qualities of the catalyst. In order to minimize benzene content the following steps should help:
- Reduce unit conversion level-Reduction in conversion reduces total aromatics level in gasoline, including benzene. However, the reduction has to be relatively severe (5-15 vol %) in order to observe benzene movement of 0.1-0.2 vol %. The degree of reduction depends very heavily on the base conversion and feed aromaticity.
- Process less-aromatic feeds-Total aromatics and benzene levels in the gasoline are directly related to the aromaticity of the feed and the unit conversion. Doubling feed aromatics increases gasoline aromatics by approximately 50%. Benzene level will also increase.
- Use low-hydrogen-transfer catalyst - Using low-hydrogen-transfer catalyst (reduced rare earth oxide level) will reduce formation of aromatics from corresponding naphthenes. This will result in reduced aromatics and benzene.
- Severely hydrotreat feed gas oil Severe hydrotreatment (some ring saturation) of gas oil will result in increased gasoline and reduced aromatics yields. The benzene level will also be reduced due to lower yield as well as due to dilution effect of increased gasoline volume.
- Fractionate/extract benzene-Downstream processing of FCC gasoline (distillation/extraction) will lead to reduced benzene level.
Some of the options discussed are difficult and expensive. In order to meet the "pool" benzene/aromatics level, it probably is much easier to accomplish this by changing reformer severity.
Lavergne: The benzene level in our West Coast FCC gasoline is currently 0.8 vol %. Any reduction :in FCC severity, such as reduced catalyst-to-oil or lower catalyst activity, can reduce benzene levels, albeit with lower conversion. On our typical feeds we have to reduce conversion 5 vol % to lower FCC gasoline benzene content by 0.1 vol %. In summary, however, there is not much that can be done to significantly reduce FCC gasoline benzene content.
Cabrera: Our data show that the range of benzene content in full-boiling-range FCC gasoline is on the order of 0.4-1.0 vol %. The FCC gasoline in general, if you look at a full-boiling-range gasoline, has a constant ratio of benzene, toluene, and xylene. So any known process variable that you manipulate to increase the production of aromatics could have a potential effect on the benzene.
The most important process parameters that will affect the overall production of aromatics, and therefore could potentially affect benzene, are the aromatic nature of the feed, meaning feedstocks with lower UOP K factors, which have more propensity to make the aromatics, and therefore benzene.
Any process variable, be it catalyst or unit-related, that would increase the severity of operation will result in higher benzene. What I mean here is anything that would raise severity such that you begin to destroy the most easily crackable components in the gasoline. So any time you destroy paraffins, etc., the aromatics concentrate and the result would be higher benzene. This would include, of course, catalyst selection, reactor temperature, and catalyst circulation.
Another element which will have a profound effect on the amount of benzene is recycling of any liquid, especially unconverted liquid in the light cycle oil boiling range or hydrotreated light cycle oil.
Our belief is that the accommodation of benzene produced in the FCC will come from dilution by adjusting operations in other units and/or the additional production of alkylate and oxygenates.
Crisler: The strongest correlation we see is with feed type. In our refinery that runs sweet crudes and virgin feeds to the FCCU, the benzene levels in the gasoline are around 0.5%. In our other refinery, where we process sour crudes, including coker heavy gas oil, our benzene levels run 0.6-0.8%.
Davis: Commercial FCC gasolines which we have analyzed have a benzene content of less than 1 liquid vol %. I agree with the other comments made by the panelists on the impact of feedstock and operating variables on the reduction in benzene levels. A less aromatic feed would directionally have less benzene in the gasoline. Small directional moves in benzene content could be achieved by lowering reactor temperature, as well as with the use of low-unit-cell catalysts.
A caution in looking at laboratory results is that if you look at gasoline composition at constant conversion, USY catalysts have a 56% lower aromatics content than REY catalysts. At constant coke, the aromatic components, including benzene, tend to nearly equal out.
Another note is that ZSM-5 additives tend to concentrate everything except those components that the ZSM-5 additives cracked out of the gasoline. Therefore, the gasoline will become a little bit more aromatic with the use of ZSM-5 additives.
RFG-5, which is a reformulated gasoline catalyst that we developed as a potential future option for lowering aromatics contents in gasolines, reduces the benzene levels of FCC naphtha in pilot plant testing.
Copyright 1993 Oil & Gas Journal. All Rights Reserved.
