NPRA Q&A—3 (Conclusion): Discussion turns to gasoline: alkylation, naphtha hydrotreating

Sept. 5, 2011
This is the final of three articles that present selections from the 2010 National Petrochemical and Refiners Association Q&A and Technology Forum (Oct. 10-13, Baltimore).
Ergon Refining Inc.'s 25,000-b/d Vicksburg, Miss., refinery, with addition of a second hydrotreater and ROSE propane deasphalting unit, has grown to be among the largest manufacturers of naphthenic process oil in the world, according to the company. It sits 340 river miles north of New Orleans in the Vicksburg Industrial Park just off the Mississippi River. Photograph from Ergon Refining.

This is the final of three articles that present selections from the 2010 National Petrochemical and Refiners Association Q&A and Technology Forum (Oct. 10-13, Baltimore). It highlights gasoline processes, especially dealing with safety, alkylation, and naphtha hydrotreating.

The first installment, based on edited transcripts from the 2010 event (OGJ, July 4, 2011, p. 80), dealt with safety and process operations. The second (OGJ, Aug. 1, 2011, p. 82) continued a discussion of safety and added discussions of coking and corrosion.

Each session employed five panelists (see accompanying box). The only disclaimer for the panelists was that they discuss their own experiences, their own views, and the views of their companies. What has worked for them in their plants or refineries might not apply to every situation but can provide sound guidelines for what would work to address specific issues.

Safety

What are the best practices for entering the vapor space above an internal floating roof in a gasoline tank?

Harbison: The primary hazard here is entering an air atmosphere that contains hydrocarbon vapor or toxics with liquid hydrocarbon beneath the floor, where wiper seals, pontoons, etc., create a barrier that prevents conditions within the confined space from changing to a potentially explosive atmosphere.

The panel…

• Greg Harbison, Marathon Petroleum Co. LLC

• Alec Klinghoffer, CVR Energy

• Randy Peterson, DuPont-Stratco

• Wayne Woodard, Valero Energy Corp.

• John Clower, Chevron USA Inc.

The types of work we may allow here include regulatory inspections, other noninvasive inspections, and minor cold work. Of course, there are permitting requirements for entry into a confined space. For this work, we typically require a level of supervision higher than the normal permit issuer to help us ensure that the risks are weighed against the benefits.

We also completely isolate the tank inflows and outflows and lock all tank mixers. This isolation helps eliminate the potential for tank disturbances that could lead to changes in the working atmosphere.

Atmospheric monitoring requirements are for oxygen in the 19.5% to 23.5% range, LEL [lower explosive limit] less than 10%, and other contaminants beneath the PELs [permissible exposure limits].

As a best practice, air-supplied respirators are typically used when someone is working inside this confined space. Also, continuous monitoring of oxygen content and LEL (particularly in the work area near seals, pontoons, or other roof penetrations) is a good practice. If ventilation is required, then we typically use air or steam-driven equipment to minimize the potential for ignition.

Rescue personnel are also a requirement, and the entrant's use of a full-body harness, lifeline, etc. are all good practices.

In general, we try to minimize the distance between the floating roof and the fixed roof to 10 ft with less being better. We also prohibit entry onto tanks where the material or the condition of the roof cannot be determined. These are roofs typically constructed of fiberglass and aluminum. We also factor in the inspection history of the tank, particularly any roof corrosion.

Finally, we do not allow any entry into this kind of confined space if product is on the roof or if there is lightning in the area. API [Recommended Practice; RP] 2026 has much more information than we are presenting today, and we refer you to that if you want more details.

Clower: At the Chevron Richmond [Calif.] refinery, we do not inspect internal floating roof tanks that are in service or have product in them. We take them out of service first. The normal API [RP] 653 inspections would apply; so we would wash the tank with cutter [light material to assist moving heavy or waxy material] for heavy material, water wash, and then prepare for entry.

Samit Roy (Saudi Aramco): Do you use chemical decontamination for this floating roof tank?

Clower: Yes, we do.

Samit Roy: Is it a special chemical? Or, is [decontamination] accomplished with chemical cleaning followed by potassium permanganate (KMnO4) treatment? Can you elaborate?

Clower: We have a specialty chemical contractor on site that will use chemicals to ensure that all pyrophoric material is removed during the water wash.

Harbison: [Editor's note: Some of the following response repeats earlier comments but is on point and has been retained.]

Entering the vapor space above an internal floating roof tank creates a set of somewhat unique safety concerns that must be addressed in a facility's safe work procedures.

The primary hazard is entry into an air atmosphere containing some level of hydrocarbon vapor or toxics with liquid hydrocarbon (gasoline, for this discussion) beneath the floor, with wiper seals, pontoons, etc. creating a barrier to prevent conditions within the confined space from changing.

Some of the specific areas that must be addressed before entry into this confined space include permitting, atmospheric monitoring, PPE [personal protective equipment] requirements, rescue, tank design or operating status, etc.

• Permitting: A Confined Space Work Permit is required for entry into the space above an internal floating roof tank. As a best practice, our refineries require approval from a level of supervision above the normal facility permit issuer. This ensures that the risk associated with the entry is thoroughly reviewed with the expected benefits.

Typical activities requiring entry are regulatory inspections, other noninvasive inspections, and minor cold-work activities. Confined-space work requires an attendant at the point of entry able to communicate constantly with the entrants and rescue personnel.

Additionally, it is our practice to isolate the tank inputs and outputs completely and to shut down and lock out all mixers. This practice minimizes the potential for disturbances to the tank's liquid contents that could create a change to the atmosphere of the work area above. Hot work in covered, internal floating roof tanks is not allowed.

• Atmospheric monitoring: Confined-space entry requires atmospheric conditions of 19.5% to 23.5% oxygen, less than 10% LEL, and benzene and hydrogen sulfide levels below the permissible exposure limit. As a best practice, respiratory protection in the form of a supplied-air respirator is used. In addition, the confined-space entry attendant is required to use this level of respiratory protection.

Continuous monitoring for percent LEL and oxygen level in the work area of the confined space is also a best practice, particularly near seals, pontoons, or other roof penetrations where hydrocarbon vapors could escape to the work area above. In some cases, mechanical ventilation may also be required. We use air or steam-driven equipment to minimize the potential for ignition.

Alkylate T90*

• Feed type:

—HF: propylene T90./amylene T90-.

—Sulfuric: similar to HF if olefins segregated (otherwise: T90- if C3 = alkylated with C5=/C4=).

—Sulfuric: MTBE/TAME raffinates T90..

• I/O ratio: most significant operational variable for both catalysts if I/O- then T90..

• Acid strengths:

—Low strengths: T90- for both catalysts.

—High strengths (>92%): T90- for HF.

*Gasoline Processes Q&A session.

• Rescue: A rescue team is always required to be available during confined-space entry work. Best practices in this area include the entrant's use of full-body harnesses and lifelines, avoidance of entanglement hazards when in the confined space, knowledge of the availability of a winch or other rescue device, and awareness that the rescue team is stationed at the tank.

• Tank design or condition: Good ventilation is best accomplished when the vertical space between the floating roof and tank's fixed roof is minimized. Our experience is to limit the distance from the floating roof to the fixed roof to 10 ft, with less being preferred.

It is also our practice to prohibit entering onto a covered floater made of fiberglass, aluminum, plastic, or similar materials because the condition of the roof is difficult to determine. For roofs made of steel, the inspection and service history of the tank should be reviewed to identify any known areas of concern to be avoided.

Additionally, entrants are not allowed to descend to a floating roof that is resting on its legs unless the space beneath the roof has been ventilated and atmospheric testing has been completed and is acceptable. Likewise, entrance is prohibited if the roof has product on it.

Two final notes: It is our practice to prohibit confined-space entry during lightning storms.

For additional details, we recommend a review of API Publication 2026 "Safe Access/Egress Involving Floating Roofs of Storage Tanks in Petroleum Service."

Clower: The best practice for vapor space entry is to remove the IFR [internal floating roof] tank from service in preparation for normal API [RP] 653 inspections. Chevron will not inspect the vapor space during normal operation of IFR tanks.

Decommissioning steps for API [RP] 653 inspections include removing tank contents, cutter, and waterwashes to remove all sludge, isolation, and preparation for confined-space entry.

Alkylation

What process parameters can affect alkylate T90? What are the critical variables you monitor in both sulfuric and HF [hydrogen fluoride] units? Discuss processing schemes, feed impacts, and operating variables.

Peterson: Feed type is very significant for alkylate T90. With HF alkylation, propylene generally lowers the T90 and amylene generally raises the T90. Diene contaminants (butadiene and pentadiene) raise T90 for both catalyst types as they form heavier alkylate. Selective hydrogenation units that remove dienes are therefore helpful in reducing T90.

With sulfuric alkylation, it is a bit more complicated: Amylenes will raise the T90 as with HF; and if propylenes are alkylated all by themselves, the T90 will be lower. But if you have propylene and-or amylene mixed with butylene, then negative side reactions will occur that cause the T90 to go up.

This is the basis for segregated feed technology. In several DuPont-licensed units, the propylene is segregated from the amylene and butylenes and fed to separate reactors. By doing so, the operator realizes the benefits of lower T90 for the propylene portion. Sulfuric acid consumption is reduced as well.

Also, if MTBE [methyl tertiary butyl ether] or TAME [tertiary amyl methyl ether] raffinates are fed to a sulfuric alkylation unit, then the T90 will be lower. This is due to the isoolefins being removed as the normal olefins in sulfuric alkylation have lower T90s.

I/O [isobutane-to-olefin] ratio is, by far, the most significant operating variable for both catalysts, meaning that if you have better conditions (higher I/O), then the T90 will be lower.

In HF alkylation, as the acid strength is lowered below about 87%, T90 will increase. If you raise the acid strength much above 91% or 92%, then the catalyst will be more active. The T90 will also increase due to more polymerization.

With sulfuric alkylation, the acid is generally staged from high to low strength. The higher stages have lower T90, and the lower stages have higher T90. Therefore, acid staging should be designed to minimize the fraction of alkylate produced at the lowest acid strength.

Higher reactor feed nozzle ΔP [pressure differential] and-or increased reaction-zone mixing reduces T90 for both catalyst types (see accompanying box).

Klinghoffer: We have an HF unit at Coffeyville [Kan.] that is a Stratco contactor and a Phillips reactor. It is an odd bird looking thing. I think that one of the things that affect our T90 the greatest is the amount of internal regeneration we use. It obviously makes the alkylate a little heavier when we are pushing rates, and that is one issue I do not think we really mentioned.

I know that UOP tells you always to run internal for the most part; but many, many years ago, we found that this was a bad idea for us. Because we had a lot of plugging in our iso-stripper, we do it the other way.

As I said, one of the things that will be affected when running high rates is that we typically have to increase the internal regeneration to pressure limits. That occurs when we have issues with T90 sometimes running a little over what alkylate really should be.

Peterson: Good point.

Another situation is that if you do have a significant amount of HF acid carryover to the iso-stripper, then it can go overhead and mix with the olefin before the reactor. This can lead to polymerization before the olefin reaches the reactor and is another way to get higher T90s with an HF unit.

Clower: At the Richmond refinery, we segregate C3 and C5 olefin streams to avoid some of the issues that Randy was talking about earlier. I do agree that the I/O ratio is the key factor in limiting the T90 increase. The segregated olefins trains operate with different acid strengths as well.

This question is going to come up in the DIB [deisobutanizer] fouling section as well, and I think Alec may have been handling it a little bit. But as we push I/O ratio higher, the net effluent is getting larger and resulting in issues in the net-effluent treatment section.

At the Richmond refinery, we have difficulties with heavier alkylate and corresponding issues with blending California gasoline. The addition of an FCC gasoline hydrotreater has been beneficial in blending gasoline in California where we can trade blended gasoline distillation for total sulfur.

Harbison: Reaction temperature and I/O ratio are two of the more important variables we see that affect overall alkylate quality. We also keep a close eye on acid-to-hydrocarbon ratio on our UOP units that are a pumped-acid design.

Kurt Detrick (UOP LLC): Certainly, I/O ratio is important, but probably more important is the ratio of alkylate to isobutane in your isorecycle stream. And for a given unit running at normal conditions, the two correspond to each other.

But if you have, for instance, some alkylate or some liquid carryover in the top of your isostripper into your isorecycle stream and some poor fractionation at the top of the isostripper, then you can get some alkylate in the isorecycle, which will raise T90.

Some of the feed schemes in the heritage UOP units, like the split feed series recycle scheme, will have a little more alkylate in that second stage. If you are looking at just the I/O ratio in that second stage compared with one that is a single stage, you might get a little higher T90.

The SOFT [split olefin feed technology] feed scheme in the Phillips design [now licensed by UOP] has the same issue. You are injecting some olefin when some alkylate has already been made; so that can directly raise T90. Those schemes do tend to increase I/O ratio for a given isorecycle rate, so that is a benefit there.

The important thing to look at in all of those schemes is alkylate-to-isobutane ratio because the alkylate can react with the olefin just as the isobutane does. It is that tertiary carbon that reacts; so that is the alkylate you have to watch out for.

Clower: Alkylate T90 can be affected by a number of different schemes, feeds, and operating variables within a sulfuric acid alkylation plant. An increase in T90 signifies heavier, lower octane product and is normally a result of polymer formation.

The critical operating variable to monitor for alkylate T90 is the iso-to-olefin ratio. Polymerization becomes a favorable reaction at I/O ratios of less than 5:1. At these ratios, olefins can react with other olefins in the acid-continuous emulsion. At Chevron, we monitor alkylate endpoint to track polymerization as a check for reaction conditions.

Increased contactor temperatures can also increase polymerization, but they likely would not increase T90 as the polymer would not be a large percentage of the total alkylate product.

Olefin feed segregation is one means of controlling alkylate quality and acid spending strengths. Segregation of C3 olefins allows for their operation at higher acid strength contactors. At Chevron, we segregate C3/C4 olefins to high acid-strength contactors and C4/C5 olefins to low acid-strength contactors.

C3 olefins tend to make conjunct polymers at low acid strengths and will also form polymers with C5 olefins at low strengths. If a plant feeds less than 10% C5 olefins, then an increase in that percentage will result in increased T90.

Harbison: Reaction temperature and I/O ratio are two of the most important variables we monitor for alkylate quality.

Acid-to-hydrocarbon ratio is also an important variable and is routinely monitored for our UOP units that have a pumped-acid design. For [ConocoPhillips] units, this is not a variable that can be changed.

Naphtha hydrotreating

What has been the experience of refiners operating selective hydrotreating of fluid catalytic cracking naphtha regarding gum formation potential of the low-sulfur gasoline? Is gum inhibitor addition a recommended practice?

Clower: We installed a gasoline hydrotreating unit in Richmond a couple of years ago. As a precautionary measure, we decided that we would inject inhibitors, regardless of the fact that diolefins had been removed.

We made this decision particularly because our research group was concerned that gum potential existed in the presence of olefins, organic sulfurs, and organic nitrogen.

In my Answer Book response, I wrote that some of what we are doing with our gasoline hydrotreating units, in terms of diolefin saturation, is complete but that olefin saturation and organic nitrogen are both mild; but in sulfur, it is about 90% conversion of sulfur species.

We segregate FCC-treated gasoline from all the other tanks.

To date, we have not had issues with gum or gum potential in any of our gasoline blends. We inject antioxidant at about 20 ppm on that product stream. Shortly, we will begin testing on reducing the antioxidant injection to determine any negative effects on gum potential in our gasoline blends.

Harbison: We have two Axens Prime-G+ FCC gasoline hydrotreaters that use selective hydrogenation on the light-naphtha stream. Before installation of the units, we used stability additives to manage gum formation in cat gasoline. When the units went online, we discontinued the use of stability additives and did not have any issues with gums at all.

At some point after start-up, we installed some projects to blend the light naphtha separately from the heavy hydrotreated cat gasoline stream. At one refinery, we run down this light naphtha with isomerate and have not had any issues with gums.

At our other refinery, we run it down with high octane platformate and have had issues with gums. We conducted a fair amount of testing at this refinery, both in the lab and in the field, and we could not really find any direct correlation between the two. So we did a field test.

At this refinery, we ultimately concluded that the stability additive was needed to take care of the gums.

Woodward: In the Valero system, we have five FCC gasoline treaters licensed by CDTECH and four Prime-G treaters licensed by ExxonMobil. The majority of the response within my system was that we did not add gum inhibitors, but there are one or two injecting gum inhibitors. I would invite the licensers to comment, if they are present in the room.

Kerry Rock (CDTECH): I would like to confirm that CDTECH does not require the addition of gum inhibitors to the product from our gasoline hydrotreating units. We have not observed any gum formation reported by our licensees in that application.

It is really not surprising because you are not only saturating diolefins, you are reducing the content of olefins, to some extent, especially the olefins that are likely to react in the form of oligomers.

Cheparthy Shankar (Indian Oil Corp. Ltd.): We have about three brand new units operating at our refineries. All three units are adding antioxidants to take care of the gum problems. This is recommended by the licenser and is part of the procedure we are following.

Clower: Gum potential exists with unsaturated hydrocarbons (olefins or diolefins) stored in refinery operations with oxygen as a free radical. Organic sulfur and nitrogen compounds remaining in FCC gasoline product can form gums in the presence of olefins and oxygen.

Temperature, residence time, and oxygen ingress are key process variables in the formation of gums in gasoline component storage. Best practice requires antioxidant injection of all FCC gasoline tanks to tie up any oxygen that would otherwise play a role in the formation of gums.

Typical gasoline hydrotreating operations are mild, resulting in the following typical species conversions:

• 100% diolefin saturation.

• 90% organic sulfur reduction.

• 20% olefin saturation.

• 10% to 20% organic nitrogen reduction.

Current best practice is to inject antioxidant into the product and measure both existent and potential gum in the component tank:

• Antioxidant is injected at 20 ppm on FCC gasoline product.

• The target for existent gum via ASTM D381-04 <4 mg/100 ml.

• The target for potential gum via ASTM D 525 <10 mg/100 ml.

Harbison: Marathon has Axens Prime G+ FCCU gasoline hydrotreaters at its Garyville, La., and Robinson, Ill., refineries. Both units use a selective hydrogenation process on the light naphtha streams (roughly 200° F. to 210° F. end point). Before installation of these two units, the gasoline streams were conventionally sweetened, and stability additives were used at both locations.

When the FCCU gasoline desulfurization units were commissioned, the stability additives were discontinued on the produced-gasoline streams. There were no reports of gum problems associated with the products.

After initial start-up of the units, projects were completed to allow the light portion of the naphtha to run down to tankage separately from the heavy naphtha. This facilitated additional blending flexibility because the light naphtha cut blend properties that differ from similar refinery streams in use at the time.

The light naphtha products from our two desulfurization units are managed differently at each refinery. One plant routes the light naphtha product to storage after blending with isomerate, and the other blends with the light naphtha a rundown with high octane reformate.

Our refinery that blends the light-naphtha product with high octane reformate has experienced high gums. Following several gasoline blends outside of specification limits for gums in which tanks had to be treated for stability, a more comprehensive lab study was completed. Although the study was unable to predict the off-specification finished product based on blend components and their ratios, tank circulation times, etc., we concluded that a field test was required.

The trial was conducted with the stability additive injected into the light-naphtha product when it was segregated for gasoline blending. Following a successful trial, the program has since been made permanent, and all gasoline blends have met stability specifications when the additive was used. The additive has been successful at moderately low doses.

The other refinery stores the light naphtha combined with isomerate and has had no incidents of high gums.

Gregg McAteer (Nalco Co.): We often see straight-run naphtha hydrotreaters that traditionally did not have a fouling problem start to foul as FCC or coker naphtha is introduced to the feed.

Gum studies are performed to determine if chemistry can reduce the fouling rate (gum formation/deposition rate). Gum-inhibitor applications have been able to give the refiner the long run it was looking for.

More Oil & Gas Journal Current Issue Articles
More Oil & Gas Journal Archives Issue Articles
View Oil and Gas Articles on PennEnergy.com