REFINERS HAVE OPTIONS TO DEAL WITH REFORMULATED GASOLINE

Gary Yepsen, Tony Witoshkin
Engelhard Corp.
Iselin, N.J.

Individual LP models must be utilized to evaluate reformulated gasoline's impact on any given refinery. But there are some general options available that appear useful for producing the reformulated gasoline mandated by the Clean Air Act of 1990.

However, the industry faces problems in addition to reformulated gasoline. These cannot be addressed piecemeal and must be part of an integrated solution. Among the compelling issues to solve in the next 2-4 years are:

Producing adequate supplies of oxygenates
Finding uses for or converting the components unsuitable for reformulated gasoline
Meeting growing long-term gasoline demand
Satisfying rising gasoline octane demand
Coping with the increasing use of heavy crudes.

The new law's intent is to reduce automobile emissions by establishing a formula for specific gasoline components, i.e., benzene, aromatics, oxygenates, and volatility (Table 1). The new law may also affect gasoline content indirectly by setting limits on hydrocarbons, carbon monoxide, and NOx in auto exhausts. Olefins are not specifically addressed.

The mandated changes might reduce gasoline volume at a time when refiners are operating near capacity in the face of growing demand for gasoline. August 1990, for example, saw demand reach a record 7,860,000 b/d, driving crude distillation utilization rates to 93% and conversion unit utilization to 100%.

Also, domestic gasoline consumption has grown 2%/year since 1984 (Fig. 1). It is projected to rise 1%/year for the next few years. This assumes that the Middle East crisis does not affect long-term crude pricing.

The gain will come from a steady increase in the number of drivers and the number of automobiles per driver, both of which offset the gains in automotive efficiency.

Not only has demand risen, but also the type of gasoline purchased has changed. Average clear pool octane has climbed 1.0-1.5 numbers in the past several years because of lead phaseout and growing demand for mid and premium grades (Fig. 2).

Refiners will also be processing heavier crudes in the future. These crudes increase heavy fuel oil and resid yields. Normally, these components would be blended into No. 6 fuel oil. However, because of stagnant demand for these materials, refineries must upgrade them to lighter components.

Fluid catalytic cracking units (FCCUs), cokers, and hydrocrackers have traditionally been used to perform this function. These units are now fully utilized, and additional resid conversion will have to be done by new processing units, requiring capital investments.

COMPONENT STATUS

The new law requires the addition of some gasoline constituents, principally oxygenated compounds, and the elimination of several others. Refiners must add new capacity to produce the needed components, and find ways to deal with an oversupply of those components removed from the current gasoline stream. To quantify these problems, the proposed changes must be looked at in detail.

OXYGEN

Oxygen content will be boosted to a minimum of 2.7 wt % in carbon monoxide nonattainment areas, and 2.0 wt % in ozone nonattainment areas. This translates into a methyl tertiary butyl ether (MTBE) content in gasoline of 15 vol % and 11 vol %, respectively-up from the current value of less than 1 %.

The existing network of refineries, pipelines, and terminals cannot handle the current three gasoline grades plus three new reformulated grades. Thus, most of the gasoline produced in the mid-1990s may have to be reformulated gasoline.

If this is so, future MTBE demand will require an additional 800,000 b/d of isobutylene feedstock above the 100,000 b/d now produced in refinery FCCUS. With current process and catalyst technology, this incremental increase will have to be met via isomerization and dehydrogenation, using n-butane as a feed.

Dehydrogenation is an expensive process because it is done at high temperature, low pressure, and with a high recycle rate. A 10,000-b/d unit costs between $100 million and $150 million, or about $10 billion for all the units needed in the United States.

BENZENE

Benzene must be reduced by about 1%, which will generate 8 billion lb/year of excess benzene in the petrochemical market. U.S. petrochemical benzene supply and demand are currently in balance at 13 billion lb/year.

Refiners will have to choose between two primary options to avoid benzene imbalances: either convert excess benzene or avoid its production.

In the latter case, the more desirable option, C5s and C6s can be fractionated from reformer feed and isomerized to isopentane and isohexane for blending into gasoline. While this is technically feasible, capital expenditures for fractionation and isomerization units will be needed.

VOLATILITY

Reid vapor pressure (Rvp) constraints will be imposed to reduce ozone levels. Gasoline volatility must be lowered to 9 psi, primarily by removing butane. Total butane removal could reduce pool volume by 1-2% and generate 70,000-100,000 b/d of excess butane.

Butane can be converted to isobutane and isobutylene for alkylation or MTBE production. Even so, the excess butane falls short of producing the 800,000 b/d of isobutylene potentially needed to meet MTBE demands.

AROMATICS

Aromatics content, currently about 32% of the finished unleaded pool, must be reduced to 25% or less. This will have a substantial impact on the FCCU and catalytic reforming operations.

FCC gasoline averages 29% aromatics, while reformate averages 63% (Table 2). The need to reduce aromatics and introduce oxygenates will force operational changes in both processes.

The aromatics problem is magnified by the fact that the crude slate has been increasingly heavier. Pilot plant studies show that, as increased resid is charged to the fluid catalytic crackers, FCC gasoline aromaticity increases (Table 3).

REFORMER OPTIONS

Reformer operation can be altered to reduce benzene and aromatics content in the reformate. One possibility is to prefractionate the reformer feed, which minimizes benzene precursors (C5s and C6s). A less desirable route involves benzene extraction from the reformate.

In addition, reformer severity can be lowered to reduce aromatics content.

Steps that control reformer benzene and aromatics production also decrease reformer hydrogen production, which will require investment in additional hydrogen plants. Alternately, reformer operating rates can be reduced in order to reduce benzene and aromatics in the gasoline pool.

FCCU OPTIONS

The fluid catalytic cracking unit contributes approximately 50% to current gasoline pool levels: 38% directly from FCC gasoline, and 12% indirectly, from alkylate.

Although FCC gasoline meets current Clean Air Act criteria for Rvp and benzene content, its 29 vol % aromatics level exceeds the maximum 25 vol % pool limit. This creates disadvantages when it is used as a blendstock, as it is in reformulated gasoline. Aromatics must be lowered in FCC naphtha or reformate to meet the aromatics pool limit.

Changes in FCC gasoline aromatics content can be attained by either:

Processing more naphthenic or paraffinic gas oils
Altering operating conditions so as to lower unit conversion
Undercutting full distillation range gasoline.

The first option is a difficult one for the average refinery because crude sources are limited and, even if more naphthenic or paraffinic crudes were available, they are substantially more expensive. In general, this is not a viable option.

The second option, changing operating conditions or philosophies to lower aromatics through lower conversion, also has drawbacks. Pilot unit data with numerous commercial catalysts imply that operating variables (reactor temperature, catalyst type, catalyst-to-oil ratio, weight hourly space velocity, etc.) affect the aromatics level only to the extent they affect conversion. In other words, FCC naphtha aromatics content is directly proportional to unit conversion (Fig. 3).

To reduce aromatics to acceptable pool levels, FCCU conversion must be lowered by approximately 6 vol %.

This is undesirable because lower conversion reduces gasoline yield by 4.5 vol % and increases production of less desirable cycle oils by 6 vol %. In addition, lower conversion may reduce octane by 0.5 RON (0.4 MON).

An alternative route is to undercut, or distill out, the heaviest 10% fraction of FCC gasoline. This fraction contains 80-100% aromatics (Fig. 4). This lowers undesirable hydrocarbons to the required level and could increase gasoline octane by 0.5-1.0 RON (0.4-0.8 MON).

Undercutting also has some disadvantages. It lowers gasoline pool volume by 4 vol % and raises the issue of how to use the rejected heavy aromatics stream of approximately 300,000 b/d nationwide.

These aromatics cannot be used in diesel fuel because of cetane specifications, or in jet fuel because of smoke point problems. Hydrocracking is a feasible solution to this problem, but it requires new hydrocracking capacity and additional hydrogen sources.

Even though each of these alternatives is viable, gasoline undercutting is probably most desirable because it permits operators to run the FCCU to maximize light olefins (C3s, C4s, and C5s) while controlling aromatics.

These light olefins can be used to produce excellent blending components for reformulated gasoline, such as polygasoline, dimate, alkylate, MTBE, tertiary butyl alcohol (TBA), ethyl tertiary butyl ether (ETBE), tertiary amyl methyl ether (TAME), and tertiary amyl ethyl ether (TAEE).

Isobutylenes and isoamylenes are the primary olefins in greatest demand because they are used to produce MTBE and TAME, the preferred oxygenate blending components for reformulated gasoline.

Light olefin output can be increased moderately by using the proper mix of catalysts, additives, and operating conditions as discussed below.

PROPYLENE

Yields of C3 olefins may be increased 28% by changing from a 4% rare earth oxide (REY) low/inactive matrix surface area (MSA) catalyst to a non-rare earth exchanged, ultrastable zeolite high/active matrix surface area catalyst that includes 3% of Z-100, a ZSM-5-based octane enhancement catalyst, in the blend.

Tests show that propylene yield increases even further when reactor temperature is increased from 960 F. to 1,030 F. (Fig. 5). It should be noted that as unit conversion increases, compression and fractionation capacities will also have to be increased.

Propylene is a raw material for the production of the gasoline blending components polygasoline, dimate, and alkylate. Although the first two are high in C6-C9 olefins, these olefins do not present a problem because they have reasonable vapor pressures and good octanes.

BUTYLENES

Yields of butylenes, as with propylene, increase as reactor temperature rises, and with the use of a low hydrogen transfer catalyst and Z-100 additive (Table 4). These changes increase butylene output by 30-40%. Again, additional compression and fractionation capacities are required downstream.

Isobutylene is the most versatile of the butylenes. It yields excellent blendstocks for reformulated gasoline when converted to ethers. Its production can be increased with a low rare earth, high/active matrix octane catalyst, Z-100 additive, and increased reactor temperature. These increase the ratio of isobutylene to butylene from the 0.27 obtained with high rare earth gasoline catalysts to 0.38.

Butylenes can be converted to five gasoline blending components: polygasoline, via catalytic polymerization; alkylate, with the addition of isobutane; MTBE, using methanol as a raw material; TBA, in the presence of water; and ETBE, by reaction with ethanol.

ISOAMYLENES

The changes in catalyst types and operating conditions discussed in the preceding also increase isoamylenes production. The ratio of isoamylenes to amylenes in the C5 olefins produced increases from 0.44 to 0.52 (Table 4). Isoamylenes can be converted to excellent blendstock ethers, such as TAME and TAEE, that have low Rvp and high octane values (more than 103, R+M/2).

Alkylate produced from C5 olefins also possesses low Rvp (2.0 psi) and good blending octane values (91.3, R + M/2). Aromatics content in any of the above blending components is less than 0.5 Vol %.

Current technology allows for an increase in light olefins in the LPG stream and at the front end of the gasoline stream (e.g., isoamylenes). This will provide some of the feeds needed for oxygenates production.

This discussion placed no mechanical or other limits on the FCCU, even though most units in the real world operate at some limit.

To take advantage of the catalysts, additives, and operating conditions that maximize light olefins production, investment will be required in order to upgrade the FCCU and associated fractionation and compression sections.

One reason that the FCCU may require no significant capital is that FCC catalysts are under development to foster light olefins and iso-olefins, and to raise gasoline selectivity. Given the timetable in the law, these catalysts are targeted to be in refiners' hands in 1992 or 1993.

In the long term, refineries can only obtain reformulated gasoline blending components from new or upgraded units, for example:

Upgrading FCCU and downstream fractionation
Adding prefractionation and isomerization of C5s and C6s from reformer feed to reduce benzene production
Adding dehydrogenation and isomerization to butane for MTBE or TBA production
Increasing resid upgrading capability (coking, hydrocracking, etc.)
Adding capacity to hydrocrack excess light aromatics (300,000 b/d) produced by undercutting FCC gasoline
Increasing refinery hydrogen supply to compensate for reduced reformer severity.
These additions will be capital intensive and have an estimated cost well above $20 billion.