U.S. REFINERS CHOOSING VARIETY OF ROUTES TO PRODUCE CLEAN FUELS

Ralph Ragsdale
Bechtel Corp.
Houston

Passage of the Clean Air Act Amendments of 1990 has prompted U.S. refiners to install new facilities to comply with stricter specifications for gasoline and diesel fuel. Refiners are choosing a number of routes to produce these clean fuels.

A roundup of the types of new facilities being built will provide a reference for those refiners who have not yet begun such projects, and an overview of the difficulties U.S. refiners are facing. Only those processing options known to be in design, construction, or operation will be presented.

CLEAN FUELS

Dubbed "clean fuels" projects, these revisions to refineries are considered essential for remaining in business, but are otherwise difficult to justify, Indeed, the cost of compliance has forced shutdowns of plants too economically marginal to support the debt required for this special kind of modernization.

it has been estimated that producing these fuels will increase refiners' costs by 2-3 cts/gal for reformulated gasoline and 12-17 cts/gal for gasoline meeting California Air Resources Board (CARB) specifications. Revisions to diesel fuel specifications will be costly as well.

Not all market regions in the U.S. will require the new gasoline formula. Those regions not classified by the U.S. Environmental Protection Agency (EPA) as noncompliant with atmospheric ozone or CO limits are exempt. Each state, however, can "opt in" to the reformulated gasoline program.

Oxygen is a new gasoline specification for the U.S. as a whole, although ethanol and methyl tertiary butyl ether (MTBE) have been blended into gasoline in some regions for several years. Uncertainty as to the extent of opt ins has made future demand for oxygenates and potential refinery modifications difficult to predict reliably and accurately (OGJ, Oct. 25, 1993, p. 66).

GASOLINE

Recent U.S. legislation requires specific reductions in tailpipe emissions, as compared to historical fuel and operating statistics for each refinery. The EPA calls this historical depiction of each refinery its 1990 baseline (OGJ, Jan. 17, p. 16).

The gasoline analysis required to determine these reductions predicts exhaust or tailpipe emissions using EPA formulas. The formulas, released and to be applied sequentially, are currently referred to as the simple and complex models.

Various terms in the models address qualities of the gasoline blend such as benzene content, Rvp, or oxygen content. This structure allows refiners to "trade off" specific gasoline qualities to achieve compliance with calculated tailpipe emissions for a given gasoline formulation.

There are, in addition, specific mini ma and maxima for some components in the gasoline blend. And CARB's requirements are even stricter than those enforced elsewhere in the U.S. (Table 1).

Planning groups within refining organizations have been charged with evaluating these requirements and defining new processing sequences and units, and revamps of existing process units, to set the future course for each refinery.

Key, to such analysis is knowledge or prediction of the geographic market for that plant's gasoline and the type of gasoline required for that market (unoxygenated, oxygenated, reformulated, or CARB). From such an analysis emerges a plan for physical modifications of the plant.

All refineries are not installing the same processing steps or sequences to comply with the new gasoline requirements. Differences in approach are created by three factors:

• Differences in refinery configurations

• Differences in product marketing situations

• Individual technology preferences.

Table 2 summarizes process technologies currently being planned, engineered, installed, or operated, for compliance with the new gasoline formulations.

It becomes apparent that a particular quality, such as benzene content, can be influenced by more than one process or operation within a complex refinery. A description of the processing options and each one's effect on the gasoline formula win help sort out these influences.

BLENDING

With such factors as the replacement of reformate octane with oxygenate octane, blends of reformulated gasolines will differ significantly from historical experience.

Little or no butanes will be added to base stocks to meet the required lower vapor pressures. Complicating the matter further, the various oxygenates have different vapor pressures.

PREFRACTIONATION

In some refineries, the 1.0 vol % maximum benzene specification can be met simply by diverting the benzene in the crude, along with benzene precursors such as methylcyclopentane and cyclohexane, around the catalytic reformer (Fig. 1).

In those plants, the benzene in gasoline from the fluid catalytic cracking (FCC) unit, plus the benzene formed in the reformer from components heavier than those normally identified as benzene precursors, will produce less than 1 % benzene in any of the gasoline blends and in the average pool.

In the scheme shown in Fig. 1, the prefractionator overhead is combined with light straight run (LSR) gasoline and fed to the isomerization unit, where the benzene is converted to other compounds.

Prefractionation also can reduce T90, if required. Preparing a heart cut to the reformer has an advantage over cutting the T90 of the reformate. In other words, prefractionator bottoms is a low-aromatic blend stock for jet and kerosine, whereas reformer splitter bottoms might need to be dearomatized in some refineries.

A typical early prefractionator was a tower that produced, as reformer feed, a heart cut from a chimney tray called the "center well." Any unit built after 1950 has no center well or heart cut, and is either a simple depentanizer or a rather intense deisohexanizer.

The deisohexanizer has significantly more trays and reflux and cuts through the middle of the hexanes. The relevance of this is that, in some cases where a deisohexanizer is in operation, the 1% benzene maximum can be achieved simply by shifting the cut point of the operation.

Some refineries, however, are planning to take further precautions against benzene in blended gasoline, for either of two reasons:

• Compliance with the 1 % benzene maximum is calculated to be unachievable or borderline in their plant if only prefractionation and isomerization are employed.

• Achieving markedly less than 1 % benzene will, in these plants, allow for significant trade off for a more important product quality, such as total aromatics or T90.

FRACTIONATION

The additional precautions that may be required involve reformate fractionation. Historically, such fractionation has been employed to enable the extraction of aromatics (benzene, toluene, and xylenes, known as BTX) in a variety of units.

Some refineries producing reformulated gasoline will extract benzene as a means of reducing gasoline benzene content. Others will feed reformate light front-end (LF) to a benzene saturation unit to reconvert the benzene, as opposed to extracting it or simply prefractioning "deeply" to avoid producing benzene (Fig. 2).

At least one refiner is designing both deep prefractionation and benzene saturation of reformate LF. Another refiner is planning to collect T90 + material from reformate and other streams and hydrocrack it to the T90 - boiling range.

Some refiners will want to increase reformer feed to avoid installing a hydrogen plant. This approach requires reformate fractionation.

For most refiners, the need to install hydrotreating units for catfeed (FCC feed), diesel, and FCC gasoline, coupled with reduced reformer RON, will necessitate the expansion of hydrogen facilities, regardless of whether reformer heart-cut prefractionation is employed.

ISOMERIZATION

The production of reformulated gasoline moves the justification of C5/C6 isomerization from borderline to virtually essential (Fig. 2).

Because reduced reformer RON will offset high oxygenate octane, the octane boost will not be essential. The isomerization unit will, however, convert the benzene in the prefractionator overhead.

BENZENE SATURATION

An isomerization unit can tolerate a limited amount of benzene in the feed. Therefore, when the formation of benzene in the reformer is allowed, when reformate is fractionated, and when benzene is not extracted, LF will be processed in a dedicated benzene saturation unit (Fig. 2).

EXISTING CAT. REFORMER

Catalytic reformers will operate at lower severities - around 93 RON - signaling the end of the lead-phaseout era, during which replacing low-severity units with high-severity ones was the best option. Interestingly Europe is planning to skip that evolutionary phase by replacing lead with oxygenates.

Those plants in the U.S. where BTX has been or will be extracted will continue to run at high severity if aromatics prices are favorable. Of course, the need to install a new hydrogen plant, or expand the existing one, is affected by these options.

HIGH-OLEFIN CATALYST

New FCC catalysts are capable of producing more olefins without increasing dry gas production. The obvious advantage of these catalysts is increased production of alkylate, MTBE, or tertiary amyl methyl ether (TAME).

Increasing the output of plant oxygenates reduces purchases of oxygenates. Increasing alkylate production replaces and dilutes aromatics in the blends.

The shift in FCC yield distribution can be at the expense of FCC gasoline yield-a shift which is normally detrimental to overall economics. Consequently, refiners are carefully evaluating high-olefin FCC catalysts rather than automatically embracing them.

MTBE, TAME

Many refineries are installing MTBE synthesis units to process FCC isobutylene product (Fig, 3). These units alone reduce the required purchase of oxygenates by 40-50% in most plants. The downside is reduced production of alkylate, which is needed to decrease gasoline aromatics levels.

The economics of installing an MTBE unit are usually favorable. One U.S. refiner has installed a butane dehydrogenation unit to increase available isobutylene for MTBE production. This route can result in net export of MTBE from the refinery. The economics depend primarily on the market value of MTBE and, to a lesser extent, the value of methanol.

TAME units also are being installed. TAME is attractive for the same reasons as MTBE, in addition to the fact that amylene alkylate is relatively poor quality and consumes more acid than butylene alkylate. In addition, TAME's low vapor pressure provides an advantage over blending amylenes directly into gasoline.

SKELETAL ISOMERIZATION

Skeletal isomerization of n-butylene is emerging as a lower-cost route to producing isobutylene or isoamylene for MTBE and TAME production (compared with dehydrogenation of isobutane or isopentane). With five available licensors of the process, the marketing focus is two-fold:

• Refinery application, based on FCC products and operating jointly with MTBE or TAME synthesis.

• Merchant plant scenario, based on butylene or amylene products from ethylene plants.

As of this writing, skeletal isomerization in refineries is being evaluated, but no plans to install the process have been announced.

GASOLINE FRACTIONATION

Refiners are designing and building FCC gasoline fractionators for a variety of reasons.

Fractionation is required to prepare feed for MTBE and TAME units and to comply with the reduced T90 specification (Fig. 3). In at least one plant, pentanes will be hydrotreated for sulfur and olefin reduction before being blended into gasoline.

The most flexible approach for compliance with blend and pool maxima for olefins, sulfur, nd T90 specifications appears to be hydrotreating a carefully selected portion of a heart cut from full-range FCC gasoline. With this scheme, all three specifications can be met by controlling the fractionator, assuming there is also a catfeed hydrotreater to meet the 40 ppm maximum sulfur specification for reformulated gasoline.

The heart cut is fashioned to match the process units employed. For example, the front end could be cut to send amylenes to alkylation or TAME, and the back end to meet the T90 requirement.

The bottoms from the FCC gasoline fractionator (the T90+ cut) will be blended into jet/kerosine or diesel with the bottoms from reformate fractionation, these bottoms might require dearomatization.

CATFEED HYDROTREATING

Until now, catfeed hydrotreaters have been justified most often on the basis of improved FCC yields, or as an alternative to regenerator stack scrubbing. Only, rarely have they been chosen for the purpose of meeting gasoline sulfur specifications.

Tests show that decreasing sulfur significantly reduces tailpipe emissions, and new regulations require gasoline sulfur levels as low as 40 ppm. For all but sweet crude operations, catfeed hydrotreating is being chosen as a key step toward producing ultralow-sulfur gasoline.

As mentioned, some plants will have to hydrotreat a portion of FCC gasoline as well.

SULFUR RECOVERY

New hydrotreaters for gasoline and diesel components require expansion of the refinery's sulfur-recovery plant. Selecting lower-sulfur crudes is not an option for most refiners, as long as processing low-quality crudes generates higher margins.

HYDROGEN PLANT

As expected, most plants requiring new hydrotreaters will have a hydrogen deficit unless a hydrogen-generation unit is added. Some refiners, however, are able to debottleneck existing hydrogen plants to restore hydrogen balance.

Two of the operations described previously will decrease hydrogen production: Reducing catalytic reformer feed rate (heart-cut operation) and reducing severity to meet the maximum aromatics specification.

Of the various new units being built, catfeed hydrotreating requires the most hydrogen. For those plants not intending to meet reformulated gasoline specifications but planning to hydrotreat diesel to achieve 0.05% sulfur, it is possible to maintain hydrogen balance by upgrading offgas streams containing low-purity hydrogen using cryogenic, pressure swing adsorption, or membrane technology.

DIESEL

Current federal diesel specifications include a maximum sulfur content of 0.05 wt % and a minimum cetane index of 40. California diesel specs have the added restriction of a maximum 10 vol % aromatics (20 vol % with cetane improver, if a qualification test is passed).

Process technologies currently being selected to meet those specifications include:

• Hydrocracking to produce diesel and reduce sulfur

• Hydrotreating to reduce sulfur 0 Aromatics saturation to reduce aromatics.

HYDROCRACKING

Refiners with hydrocrackers are in the best position to produce 0.05% sulfur diesel without significant capital investment (Fig. 4). It is difficult, however, to justify shifting a hydrocracker from maximum-gasoline mode to distillate mode, because of the negative effect on refining margins.

The production of diesel from hydrocrackers that process light cycle oil (LCO) will require an added aromatics-reduction step to meet the 10% maximum aromatics spec (Fig. 4). Aromatics-saturation units are being designed to meet the most stringent diesel specifications.

Details of this option are outlined in Fig. 5. To date, this aromatics specification is required for California only.

HYDROTREATING

Many U.S. refiners have low-sulfur diesel hydrotreaters in operation, design, or construction. These units process straight-run diesel, or cuts from other conversion units, to produce 0.05%-sulfur product.

Some small-to-medium-sized plants are planning to cease marketing road diesel to avoid capital expenditure.

INVESTMENT DECISIONS

Clean fuels legislation in the U.S. has resulted in the planning, design, and construction of new refining facilities along with the debottlenecking of existing process units. As described, a variety of processing steps have been selected.

In the case of low-sulfur diesel, the factors influencing investment decisions include the existing configuration of the refinery, the financial ability of the plant to incur significant debt, and whether the plant can choose to market on-road diesel.

Copyright 1994 Oil & Gas Journal. All Rights Reserved.