NPRA Q&A-3 HIGH-OCTANE GASOLINE REQUIRES HIGH-QUALITY ALKYLATE

Steady demand for high-octane gasoline requires a good supply of high-quality alkylate, an important high octane gasoline blendstock.

Maintaining high-quality alkylate requires close attention to the process and operating conditions of both sulfuric acid and hydrofluoric acid (HF) alkylation units. Particular attention must be paid to the types of feeds, and the contaminants in them, to maintain alkylate-quality specifications.

At the most recent National Petroleum Refiners Association question and answer session on refining and petrochemical technology, held in New Orleans, Oct, 46, 1989, refiners discussed the process conditions necessary to alkylate several types of feeds, as this excerpt from the meeting indicates.

Details of this important annual meeting and its format can be found in OGJ, Feb. 26, 1990, p. 83.

SULFURIC ACID ALKYLATION

Is the hydrogen (or other light ends) contained in olefin feed that has been hydrogenated for diolefin removal a problem in the alkylation unit? How is it handled?

LIOLIOS: Any excess hydrogen or more importantly light ends that enter a sulfuric acid alkylation unit from an upstream selective hydrogenation unit will concentrate in the refrigeration section. The accumulation of these light ends will cause higher than normal compressor discharge pressure. This higher pressure will effectively reduce the compressor capacity giving higher reaction temperatures on those units which are compressor limited.

The best solution to this problem is to strip the light ends from the treated feed before it enters the alkylation unit. If this is not an option, the light ends must be vented from the refrigerant accumulator and the depropanizer overhead accumulator. The penalty for venting is increase in overall isobutane usage with any alkylation unit.

COMEAU: Hydrogen or light ends that are left in the olefin feed to either an HF or a sulfuric acid alkylation unit will eventually concentrate in the overheads of the de isobutanizers (DIB's) or in the depropanizers. These light ends act as non-condensables and will pressure up these columns unless vented to the relief gas scrubbing systems.

Proper control of the unsaturated gas plant de-ethanizing stripper will minimize this problem. Most selective hydrogenation unit designs provide for hot flash drum or stripper on the reactor outlet to minimize hydrogen and light ends carryover to the alkylation unit.

MILLER: We normally do not find detectable amounts of hydrogen in our alkylation plant feed from the BB unit. The solubility of hydrogen in the BB is very small. We do occasionally have problems with small amounts of methane and ethanes that come in with the BB feed from the butadiene unit or other sources.

These light ends accumulate in the propane stream as previously discussed. One problem that we have experienced is high vapor pressure of the propane product. Normally the light ends are vented off from the unit by venting the depropanizer overhead accumulator.

NEGIN: Excess hydrogen as well as ethane will be non condensable in the alkylation unit and must be vented from the unit. The amount of non condensables to be vented is a function of the hydrogen purity used in the diolefin removal process, with low purity hydrogen resulting in higher vent rates in the alkylation unit. In sulfuric alkylation units, the vent point is typically at the compressor discharge drum. Non-condensables show themselves by increasing the compressor discharge pressure. As a result operating costs increase and compressor throughput is restricted.

CHARLES S. McCOY (McCoy Consultants): The question was originally submitted to the Screening Committee in terms of hydrogen. Is there anything unique about hydrogen in the alkylation plant?

LIOLIOS: To the best of my knowledge, hydrogen does not interfere any more than the other light ends.

MILLER: The solubility to hydrogen is very low. What you would find is if you had slugs of either methane or ethane that this would gradually be purged from the unit due to its solubility. However, when hydrogen enters the unit, it is there to stay until you vent it off.

We have seen fluffing, or foaming, of the sulfuric acid in our reactor emulsion settlers. What causes this phenomenon and what steps have been successful in overcoming it?

FRAYNE: Fluffing, foaming, and emulsions have three primary causes. Particulate matter such as corrosion products might be a factor. Over mixing in the contactor might create stable emulsions. Chemical contamination may lead to foaming and stable emulsions.

Based on previous NPRA Q&A transcripts, process condition changes would be the first step in combating this problem. It is advisable that a knowledgable consultant investigate the problem relative to unit operating conditions. We understand that in older units there has been a benefit from minor equipment improvements.

There are chemicals that can be used to help with these problems. Surface active additives which reduce the surface tension between the acid and the hydrocarbon have been effective. These additives enhance droplet formation for emulsion resolution. The additives stay with the acid.

COMEAU: Foaming can be caused by the presence of iron sulfate particulates in the acid, sulfonic esters, or high mixer impeller speeds. Keeping corrosion products to a minimum, increasing settling time, and a reduction in mixing energy should be helpful.

LIOLIOS: Some additional process related problems that can lead to fluffing problems are:

Cold acid: At temperatures below 40 F. the settling rate of acid in hydrocarbons is diminished. Operating below 40 F. has occasionally resulted in acid fluffing.

Acid settler levels: High acid inventory in the acid settlers will lead to higher than normal acid emulsions in the acid settler. The cause of high acid inventory is normally the result of plugged level gauge glass taps. This is one of the most frequent causes of acid fluffing.

Reactor pressure control: Unusually low reactor operating pressure can cause entrainment of acid. A high propane content in the system or high reactor temperature may also lead to acid fluffing. Poor control of these parameters will cause the contents of the acid settler to begin boiling. The resulting agitation inhibits acid settling.

Improper contactor emulsion ratio: Too low of a percent acid in the Contactor ratio glass can switch the emulsion from an acid continuous emulsion to a hydrocarbon continuous emulsion. This is when the acid is being broken up into small droplets which are dispersed within the hydrocarbon phase. This type of emulsion requires additional acid settling time.

PARKINSON: In addition to the things already mentioned, i.e., low and fluctuating reactor and settler operating pressure, high propane and particulates; we have also found that unusually high acid strengths, with an effective acid strength near to or in excess of 100 wt % can also lead to foaming problems.

R. E. DAVIS (R. E. Davis Chemical Co.): In addition to what the panel has mentioned, I have made some notes of things that through the years we have found to cause these problems.

Addressing the Stratco type reactor, the impeller clearances are too great; feed nozzles falling off or holing through; acid nozzle falling off or holing through between the shell; a hole in the shroud causing short circuiting of the acid recycle directly to the impeller from the outer shell; too low pressure on the settler; excessive feed rates; too little settling time; very low acid strengths; as the panel mentioned, with C3 in the system with too low pressure to control the boiling; too low acid recycle; too high acid recycle; temperature too high; fresh acid reconstituted rather than regenerated which is one of the major problems in the industry right now; the spent acid getting into the fresh acid tank; too low percent acid in the emulsion; settler acid level too low due to a lying gauge glass; C5 olefins or C6 olefins in the feed; or in particular be aware that C2 olefins will do the same thing.

Quite a few laboratories do not have a column that will separate the C2 olefins in the presence of the other hydrocarbons and give a good analysis. Other causes are: settler distributor broken; vortex breaker missing; caustic or amine from feed pre treaters; the iso-to-olefin ratio too low; and speaking directly of the Kellogg unit, the zone differential pressure as being too low. There is not enough differential pressure between the zones to move the acid forward.

Are there any alternatives to NaOH for effluent treating? Are any refiners using an acid wash followed by bauxite treatment of raw alkylate for removal of neutral acid esters? If so, was any improvement in research octane observed? Was there reduction in corrosion?

LIOLIOS: As mentioned before, our current effluent treating design is a fresh acid wash followed by a warm alkaline water wash. The results of our recent startups have shown good results with this design. However, we do believe that a fresh acid wash followed by a bauxite treater would be a very good system.

To the best of our knowledge, there are no systems of this type in operation. The advantages of both systems are reduced fractionation fouling and corrosion, reduced acid consumption, and in some plants, specifically overloaded plants, some incremental octane increases.

EUBANKS: Some Chevron plants use fresh acid wash, followed by either caustic or alkaline water wash to treat reactor effluents.

We believe the acid wash step provides about 0.5 research octane number clear benefit by keeping polymer forming esters out of the fractionation section. We also see a reduction in corrosion caused by SO2 released in reboilers by these same esters. Keeping the water wash hot (125-140 F.), helps de compose esters before they get into the columns.

TOM VARADI (Merichem Co.): Merichem also recommends an acid wash followed by an alkaline water wash instead of the more conventional caustic wash approach to maximize the reduction of acid and neutral esters in the alkylation reactor effluent.

What is the range of sulfuric acid consumption for amylene, 1-butene, trans-2-butene, cis-2-butene, i-butylene? What are the corresponding octanes of alkylates produced? What is the accepted range of sulfuric acid consumption for butadiene?

LIOLIOS: The expected acid consumption and octane values for butylene isomers are listed below. The values for amylene assume an average refinery amylene distribution and a 50:50 mix of amylenes and butylenes.

What progress is being made in refinery processes to alkylate, dimerize or polymerize ethylene into something usable in gasoline?

COMEAU: There are three processes I know of that are under development or in use. Two processes licensed by IFP use ethylene as a feedstock.

The Dimersol E process charges an ethylene and propylene feed from a cryogenic fractionation section and dimerizes olefins at 400 to 500 psig at 100 F. in the presence of an alkyl alumina catalyst into a full boiling range olefinic gasoline. Fifteen percent of the dimate is butylene with the remainder being C5+ olefin. There are no commercial installations as yet.

The Alpha Butyl process dimerizes ethylene to 1-butene. The 1-butene can be used in alkylation, but a more valuable use is as a feedstock to a lower linear density polyethylene plant. Ethylene must be polymer grade, 99.9% pure, and would require cryogenics along with excellent fractionation. An Alpha Butyl unit is being successfully operated in Saudi Arabia.

The Catskill process, marketed by Brown & Root and developed by Chemico Research & Licensing company, alkylates ethylene or propylene with the light aromatics found in the high octane reformate. A commercial test has been completed at a mid-western refinery. Currently CR&L is continuing work to develop a catalyst with a longer life and to study optimal reactor configuration and operating conditions.

EUBANKS: Chevron has not pursued these technologies because we believe them to be economically unattractive. It is true that they need a high purity ethylene. If you are going to take it out of the absorber gas on the FCCU, then you must have some kind of concentration upgrading equipment. We calculate that if pure ethylene is used for a C10 product that the cost of ethylene alone would be $1.50/gal. When you add manufacturing expenses to this, it becomes very unattractive.

NEGIN: Significant processing has existed for some time to convert ethylene into practically any final hydrocarbon including gasoline. Conversion into gasoline has not advanced very far, mainly because there are many more economically rewarding conversions that can be done with ethylene. We have a list of five processes. They are: Dimersol E (olefin reforming) by IFP, Catalytic Pyroform (olefin reforming) by KTI, DIP Alkylation by Phillips, MOGD (oligomerization) by Mobil, and finally DHCD (dehydrocyclodimerization) by UOP.

AL EDELMAN (Exxon Research & Engineering Co.): We have been active in sulfuric acid alkylation. Recently we have added two new commercial units to our list, a licensed unit and one of our own. We have a third unit currently in design. They use the autorefrigeration system which we believe is very efficient. With previous capacity, we will have about 80,000 b/d of installed capacity on line. Our technology incorporates quite a bit of feedback from commercial units especially since we operate many of the units ourselves.

HYDROFLUORIC ACID (HF) ALKYLATION

Does anyone alkylate amylenes in an HF unit? What is the effect on the alkylate RONC, Rvp, and acid consumption?

BIGGS: At one plant I know, they ran 2 to 3 vol % of C5 olefins in their mixed C3-C4 feed to the alkylation unit. They did not see any change other than a little higher end point of the alkylate. On a pure component basis, they believe that the C5 olefin feed will run 4 or 5 octane numbers lower than the C4 olefin feed. They feel that the Rvp will be slightly lower, and the acid consumptions will be slightly higher.

COMEAU: We have one refinery which has been alkylating amylenes as a way to lower FCC gasoline vapor pressure and to fill up the alkylation unit. Amylene alkylate per literature has an (R + M)/2 of 91, and a vapor pressure of 0.4-0.5 psi. Our plant processes 5-10% C5 olefin in the alkylation feed with no octane change over their normal C3-C4 olefin feedstock. Polymer formation is up 9 b/d. Acid consumption is up from 0.22 lb/bbl to 0.3 lb/bbl due to the increased contaminants like diolefins, C6+ materials, and sulfur compounds.

Sulfur compounds have increased in the olefin feed, but extraction efficiency of the sulfur compounds in our Merox unit has been good allowing only a 10 ppm sulfur increase in the olefin feed. Therefore, higher acid consumptions can not all be attributed to the increase in sulfur.

Alkylation of the amylenes has only been economic since the fall of isobutane prices and the implementation of the lower Rvp specification on gasolines. An article entitled "Possibilities for Amylene Alky" in the September 1977 issue of Hydrocarbon Processing details the operation of HF and Sulfuric Alkylation units on amylene feeds. Our operating data is as follows:

FCC gasoline Rvp - 5.0

Mixed olefin production C4 = and lighter - 27.96

C5's and C5 = - 4.05

Total alky feed (%) - 32.01

C3 = - 25.62

C4 = - 30.65

C5 = - 7.31

Olefin % of FF - 63.58

iC4 olefin ratio overall/Rx - 7/14

FISCHER: We tried alkylating amylenes earlier this year when reduced Rvp's were imposed. Because of some special unit constraints, we saw some large changes in the unit's performance. Since we presently do not have a Merox unit on our alkylation feed, the sulfur in the feed increased from an historical average of 100 ppm to a peak level of 160 ppm. On a theoretical basis this would have increased our acid consumption by over 25% or about 1,100 lb of HF/day.

Also during this time we had a feed upset which resulted in fouling of the trays in the isostripper which had the effect of reducing iso/olefin ratio from 14 to 11 causing additional acid consumption and lower octane. We ran maximum internal and external regeneration, and our acid strength decreased from 88 to 86%. The overall effects of the lower acid purity, reduced recycle ratio, and higher feed sulfur during this period of amylene alkylation resulted in a loss of 1 RON and 0.6 MON. We saw no effect on alkylate Rvp's since the amount of amylene in the feed only typically ran at 24%.

We still believe that amylene alkylation is a viable means to help meet lower Rvp requirements. We are currently in the design stage of adding an LPG Merox extraction unit to deal with the feed sulfur and to help eliminate the problems it causes. The fouled trays in the isostripper will be cleaned or replaced in the current turnaround to help restore the isobutane recycle purity.

LIOLIOS: An additional reason for alkylating amylenes is to reduce the bromine number in the final gasoline blend.

J.A. GEARHART (J.A. Gearhart Enterprise Inc.): Kerr McGee alkylated the entire C3, C4, C5 fraction from the FCC unit very successfully. The overall gasoline blending program was not penalized. It was a good way to control volatility of the gasoline.

COMEAU: Mr. Gearhart, which refinery was that in?

J.A. GEARHART (J.A. Gearhart Enterprises Inc.): Wynnewood, Okla., refinery (Kerr-McGee).

OSCAR A. ALDAMA (Coastal Refining & Marketing Inc.): Can you expect the same benefit in HF units of alkylating 1-butene vs. cis 2-butene or trans 2-butene as in sulfuric acid units?

LIOLIOS: In sulfuric acid alkylation, we do not believe that there is any octane difference between 1-butene and 2 butene.

In HF alkylation I believe that 1-butene and 2-butene have slightly different octane values.

OSCAR A. ALDAMA (Coastal Refining & Marketing Inc.): I guess that what I am getting at is if you isomerize your BB's into mostly 1-butene, can you expect the same kind of spread in the octanes as you mentioned for the sulfuric acid units?

LIOLIOS: Are you talking about sensitivity between RON and MON?

OSCAR A. ALDAMA (Coastal Refining & Marketing Inc.): Not just sensitivity, but an increase in the research octane number. I believe you mentioned that for sulfuric acid units, you can expect about 3-4 numbers higher with 1-butene vs. cis 2- and trans 2-butenes in the research numbers. I am asking if you can expect the same kind of benefit with HF units.

LIOLIOS: In sulfuric acid units, the two isomers react almost the same. So isomerization is not that beneficial. In HF alkylation there is a slight difference in the octane of the two isomers.

H.C. WARD (UOP): There is a benefit to processing 2 butylenes in an HF alkylation unit: you produce a higher octane alkylate from 2-butene than from 1-butene. One of the advantages of having a selective hydrogenation unit in front of an HF unit is that you isomerize some of the 1 butene to 2 butene. There are ways of operating the unit to do some isomerization internally.

LARRY LEW (Phillips Petroleum Co.): Phillips markets a Hydrisom process which converts butene-1 to butene 2 in our HF alkylation unit feed. The Hydrisom process was first installed at our Sweeny, Tex., refinery in 1988. In addition to dramatically increasing the butene-2 content of the HF unit feed, the process also virtually eliminates butadiene, which helps to minimize acid soluble oil (ASO) production. The following table gives the relative research octane values of the various butylene isomers in the HF alkylation process:

Earlier this year we replaced our 40 year old HF alkylate unit at our Sweeny refinery with a new HF unit. The C4 alkylate octane produced from the new unit has been confirmed in the range indicated for butene-2 alkylate.

What is the effect of high levels (125 ppm or more) of mercaptans in the feed to an HF alkylation unit? Are any changes required in the acid regenerator?

WELCH: High levels of mercaptans will produce a light ASO, usually yellow-brown in color. The light ASO will not be removed at normal rerun tower operating conditions. The rerun column temperature must be lowered to remove the light ASO along with a corresponding increase and loss of HF acid. High mercaptan levels may cause the acid strength to drop rapidly, and an emulsion may become evident in the settler. In addition, propane may not meet specification, and the alkylate may go doctor positive and remain that way for a few days, even after the acid has been cleaned up.

BIGGS: We feel the effect will be a higher acid usage and offcolor alkylate. And yes, the acid regenerator should be run at a cooler temperature to allow the light sulfur compound a chance to drop out the bottom of the regenerator.

EUBANKS: I agree with the statements that have been made. With respect to ASO make, it is light; and also there will be higher HF losses in the unit as a result of the high mercaptan levels.

FISCHER: Corrosion in the unit could also increase due to reduced acid purity. With our high feed sulfur, we feed the regenerator on the fifth tray.

LIOLIOS: I have basically the same comments. However, regenerators appear to be limiting in many plants, since the units are operating above design capacities. Modifications to the regenerator will likely be required to extend that capacity.