Reformulated gasoline and diesel fuel will stretch existing refining processing configurations to the limit, and they will likely require some new processes. As a result, refineries could look substantially different by the end of the 1990s.
Reduced gasoline aromatics, olefins, and benzene contents and lower gasoline vapor pressure will mean that major gasoline producing processes, such as fluid catalytic cracking (FCC) and catalytic reforming, will have to change. Other processes, such as alkylation, polymerization, and isomerization will also be affected.
New diesel fuels will mean deeper hydrodesulfurization and more severe hydroprocessing operations to reduce aromatics content.
Rules that may require minimum oxygen content in gasoline will also spur the installation of processes that produce oxygenates, such as methyl tertiary butyl ether (MTBE), ethyl tertiary butyl ether (ETBE), tertiary amyl methyl ether (TAME), and maybe some others.
Reformulated fuels will also change the availability and value of blendstocks needed to make the new fuels.
This could result in the need for some new processing units, such as dehydrogenation, alkylation of higher hydrocarbons, and specialized fractionating units.
And new refining catalyst technology will be needed for many of the new and existing processes that will allow them to efficiently produce the new fuels.
This article looks at some of the process developments and refinery configurations that could be part of refineries to meet fuel demands for the 1990s and into the next century.
GASOLINE PROCESSES WILL CHANGE
Reformulated gasoline would cause big changes in catalytic reforming and FCC operations, the processes that supply more than 65% of the motor gasoline blendstocks made in the U.S.
Strict rules on aromatics and benzene content in gasoline hit the catalytic reformer hard because its reactions result primarily in aromatics and some benzene. According to a recent National Petroleum Refiners Association survey, 1,75,767 b/cd of reformate was produced in U.S. refineries from Apr. 1 through Sept. 30, 1989 (see p. 49, Table 4).
Simple reduction of reforming severity reduces these components, but it also pinches octane blending capability. Lower severity reforming does, however, result in a slight increase in gasoline yield.
Other than severity reductions, there may be little that can be done to catalytic reformers to produce blendstocks with lower total aromatics. According to George Unzelman of HyOx Inc., reduction of aromatics levels to below 30 vol % in the entire U.S. gasoline pool may not be possible given current processing capabilities (OGJ, Apr. 9, p. 43). Of course, if low-aromatics gasoline is required only in areas with the most severe air-quality problems, shifting of blendstocks from current processing capability could meet the needs.
But operational changes can substantially reduce reformate benzene content, according to UOP at a recent process seminar.1
UOP revealed two approaches to managing benzene in the reformer. One is to prefractionate the naphtha feed, sending the benzene and six carbon-benzene precursors to an isomerization unit. The other is to post-treat the reformate product to remove benzene, either by isomerization or by alkylation with a new UOP process called Alkymax (details of the process will be discussed later).
In the first scheme, benzene is still produced in the reformer even though the six carbon-benzene precursors are removed from the feed. Secondary dealkylation reactions cause some higher hydrocarbon aromatics to be converted to benzene.
The degree of dealkylation is a function of research octane clear (RONC) and the selectivity of the catalyst. Higher octane increases dealkylation, and an operation with lower reactor pressure and more-selective catalyst decreases dealkylation.
A 1960s UOP reforming unit operated at 102 RON severity produces more than 5 vol % benzene concentration in the reformate when processing full naphtha feed. Prefractionated feed, with the benzene precursors removed, will reduce the benzene by a factor of 3 for the same octane.
For a 1990s UOP reformer at 102 RON severity, benzene reduction is more than a factor of 8 when the feed is prefractionated. It must also be noted that the benzene concentration in a 1990s reformer is less than 0.5 vol %.
If a 1 vol % benzene limit is ruled, the 1990s benzene concentration from a 1990s reformer will be adequate without prefractionation. But a 1960s unit would not be able to meet reformate benzene specifications above 99 RON severity. Table 1 shows the properties of reformates from various UOP units.
MORE LIGHT OLEFINS FROM FCC
Fluid catalytic cracking will also have to change if proposed rules require gasoline reformulation. The FCC unit accounted for 2,551,392 b/cd of gasoline blendstocks produced in U.S. refineries during Apr.1-Sept. 30, 1989.
Aromatics content of the full-range FCC gasoline are estimated at 29.7 vol % for summer gasoline (see Table 6, p. 50 . Light FCC had 13.7 vol % aromatics, and heavy FCC had 50.3 vol % aromatics.
The aromatics level of the full-range FCC cut is higher than the proposed 20-25 vol % aromatics limit. This means that some change may be necessary to reduce the FCC cut aromatics content slightly.
A more important circumstance, however, is olefins content of the FCC gasoline. High octane gasoline requirements since the lead phaseout in 1985 have resulted in higher olefinicity of the FCC gasoline.
Therefore, a strict limit of about 5 vol % olefins in gasoline to limit reactive hydrocarbon emissions could dramatically affect FCC operations during the 1990s.
To reduce olefinicity of the FCC gasoline, FCC units can be run to be more selective to light olefins. The reduction in gasoline yield would be replaced by higher yield of light olefins (propane-propylenes, butane-butylenes) that would be further processed in alkylation and MTBE or ETBE units.
These units produce gasoline components that are more attractive for blending into reformulated gasolines. Alkylate is low in aromatics and olefins, and it also has a low vapor pressure and high octane number blending values.
The butylene stream would also be valuable as a feed to oxygenate-producing units, such as MTBE units. MTBE also has low vapor pressure and high octane number blending values, and it also provides a source of oxygen, should that be mandated.
Table 2 compares the operation of a UOP FCCU in the gasoline and light-olefins modes .2 Table 3 shows combined FCC gasoline and alkylate yields for both cases.
Changes in FCC selectivity can be enhanced by FCC catalyst formulations. These formulations can be made from some of the newer octane-boosting FCC catalysts developed to boost octane of FCC gasoline fractions.
Dealuminized zeolites and special matrix formulations are features of these catalysts.
ISOMERIZATION
Isomerization Of C4, C5, and C6 streams has been part of refinery process configurations for many years. C5/C6 isomerization capacity jumped during the late 1980s in response to the lead phase out.
C5/C6 isomerization produces highly paraffinic gasoline blendstocks with high octane quality. The products will make ideal gasoline blendstocks in a reformulated product slate.
Reformulation rules will likely require increases in C5/C6 isomerization capacity. With reduced reforming severity, and possibly reduced reforming capacity, more light-straight-run naphtha (C5) will be available for isomerization.
C4 isomerization has become more important because of volatility limits that require butane to be backed out of gasoline. Isomerizing the butane to isobutane will provide more alkylation feed. Dehydrogenation to isobutylene will provide more feed to oxygenate producing units.
PRESENT ALKYLATION UNITS
Present sulfuric acid and hydrofluoric acid alkylation units produce a gasoline blendstock that fits ideally with reformulated gasoline specifications. Alkylate has minimal aromatics and olefins contents (less than 0.7 vol % in each), but the volume of material blended into gasoline at U.S. refineries amounted to only 831,348 b/cd blended into summer gasoline in 1989 (see Table 4, p. 49) comprising about 12 vol % of summer gasoline made in U.S. refineries.
Obviously, the attractive properties of alkylate could spur capacity additions in many U.S. refineries during the 1990s. But alkylation capacity cannot be added easily without modifications to FCC operations, unless other feed supplies from other sources are made available.
One of those other sources could be gas processing facilities in the U.S. and abroad. If demand for isobutane and isobutylene rise to meet reformulated gasoline requirements, economics could be favorable for gas processors to install C4 ISOMerization, isobutane dehydrogenation, and even MTBE or ETBE units (see separate story in this article).
Isomerization units at gas processing facilities could then supply the additional feedstock to support increased alkylation capacity in refineries.
DIESEL PROCESSES
Process changes will also be required to produce low sulfur and aromatics diesel fuels. To meet the 0.05 wt % sulfur limit proposed by EPA for 1993, hydrotreating operations will have to increase capacity and operating severity (OGJ, May 29, 1989, p. 47).
For lower aromatics diesel fuels, hydrotreating operations will have to substantially increase processing severity, perhaps bordering on hydrocracking (at least mild hydrocracking) operating conditions.
But pilot-plant testing of some new hydrotreating catalysts have yielded some promising results that may reduce the projected capital expenditures needed to meet a low-aromatics specification (OGJ, May 7, p. 109). If refiners consider building aromatics-saturation units, it might be better to choose hydrocracking, which has about the same capital investment requirements and would provide greater processing flexibility. Some aromatics reduction in the diesel pool could be accomplished by selecting only those distillate blendstocks with high cetane value (see second article). This could be a favorable route for some refiners, but many would lose substantial feedstock flexibility.
And that would also change the feedstock balance in some refineries, requiring other processes to fully utilize the stocks not used for diesel production.
SOME NEW APPROACHES
Adjustments and modifications to the major gasoline producers, reforming, FCC, and isomerization, may not be able to supply enough gasoline at demanded quality.
Therefore, some new processes and processing configurations will be required.
Some of these configurations simply change the operating conditions and selectivity of some traditional processes. Others are new processes or processes not traditionally used in refineries. Obviously, most of these new processes have yet to gain substantial commercial experience.
FCC, HDT, AND EXTRACTIVE DISTILLATION
One configuration change combines FCC operations with hydrotreating and extractive distillation to optimize the quality of gasoline, diesel, and jet fuel.' The configuration limits the gasoline 90%off distillation temperature to 350 F. instead of 430 F.
Lowering the gasoline fraction endpoint to 350 F. will significantly reduce gasoline aromatics, but the leftover higher cut will need a home.
This heavy naphtha cut is unsuitable for diesel blending because of the high aromatics content and moderate olefinic composition. Aromatics cause low cetane index, and olefins will lead to color stability problems because of their tendency to degrade by oxidation.
This cut, however, could be processed in an extractive distillation unit to recover valuable petrochemical intermediates, such as benzene, toluene, and xylenes (BTX). Several extractive distillation solvents have been used, such as Sulfolane, that can separate the aromatics from the 350-430 F. cut.
If the leftover aromatics in the cut are then blended back into the C5-350 F. gasoline material, the octane performance of the gasoline cut would be about the same as if an equivalent amount of toluene were blended into the cut.
However, if a hydrofinishing step is used to improve color and color stability by removing olefins from the leftover material, the aromatics-free 350-430 F. cut becomes an excellent blending naphtha to lower aromatics in diesel.
Studies on commercial gasoline cuts also showed that if the aromatics in the leftover cut are hydrotreated from 50 vol % down to 10-20 vol %, using single-stage, high-pressure hydrotreating, the hydrotreated product could be considered a good candidate for use as a premium jet fuel.
ETHERIFICATION OF C4/C5 CUTS
Etherification, processing to produce ethers such as MTBE, will likely grow in popularity as rules governing gasoline reformulation come into effect. MTBE production has grown substantially since the lead phaseout in the 1980s as an octane booster.
MTBE should experience increased use during the 1990s not only for its octane increasing capability, but also because of its low vapor pressure blending value and because it is a source of oxygen that may be mandated in gasoline.
But both methanol and isobutylene feedstocks could be tight as a result of increased MTBE use. Other ethers may have to be produced to meet octane, vapor pressure, and oxygen content specifications.
Ethers made from C5 materials may likely fill that requirement.4 Those ethers are TAME and tertiary amyl ethyl other (TAEE).
There are two major sources of the C5 cut, FCCU's and steam crackers. In both cases, the reactive iso-olefin concentration is in the range of 15-30 wt %.
The cut from steam cracking is usually available down-stream of a first-stage pyrolysis gasoline unit. The FCC C5 cut is usually sent to alkylation, but limited alky capacity could free some of it up for etherification.
C5's sent to a TAME or TAEE unit provide a material that increases RON 2-3 numbers, increases MON 1.5-2.5 numbers, and reduces vapor pressure by 1.5-3 psi Rvp.
The economics of the C5 etherification process, along with some important product properties, are shown in Table 4.
OLEFINS TO FUELS
Processes to directly convert light olefins to gasoline and distillate have been proven recently on both refinery demonstration runs and pilot-plant tests.5 The processes use Mobil Research & Development Corp.'s ZSM-5 shape-selective zeolite catalyst.
The olefins-to-gasoline-and-distillate (MOGD) process converts propylene and butylene streams to high yields of olefinic distillates. Hydrotreated distillate has a cetane number of 52, less than -50 F. pour point, less than 4 vol %.aromatics, and contains negligible sulfur or nitrogen compounds.
The olefins-to-gasoline (MOG) process converts dilute olefin streams, such as FCC fuel gas and unsaturated gas-plant C3 and C4 streams, to high-octane gasoline. Octane level of the gasoline is 95 RON and 82 MON.
To test the MOGD process, a large-scale test run was made in a Mobil refinery. It used a commercially produced ZSM-5 catalyst in refinery-scale equipment in a run that lasted 70 days.
The test run used an available wax hydrofinisher, modified to three MOGD reactors, each with about 3.5 m of catalyst depth. Charge stock was a mixture of propane propylene/butanes/butylenes (62% olefins) pumped directly from an FCC unsaturated gas plant. A steam stripper was used to meet distillate flash point.
The test run had four objectives: demonstrate the process in commercial-scale equipment, demonstrate controllability in a large multi-reactor adiabatic unit, demonstrate catalyst regenerability, and provide sufficient distillate product for fleet testing. Table 5 shows distillate properties.
The MOG process was demonstrated in a 4 b/d fluid-bed pilot plant over 8 months. The run was carried out with simulated fuel gas and fuel gas plus propane/propylene feeds prepared by blending.
The ability of the MOG process to handle typical levels of potential fuel-gas contaminants was also demonstrated. Hydrogen sulfide did not affect olefin conversion, gasoline yield, coke yield, or light gas distribution.
Large amounts of basic nitrogen compounds can also be tolerated by the process. Large amounts of ammonia in the feed decrease ethylene conversion, but the conversion can be recovered by increasing catalyst circulation rate and reactor temperature. Table 6 shows typical C5 gasoline properties.
BENZENE-REDUCTION PROCESS
A new process for reducing benzene in gasoline has been developed by UOP.1
The process converts benzene in gasoline streams by reaction with light olefins over a fixed bed of catalyst.
Typically, the light-olefinic feed to the alkylation reactor comes from the gas-condensation section of an FCC, although it may also come from the gas plant of thermal processing units, such as cokers. These units all produce streams containing ethylene, propylene, and butylene, all of which are potential feeds to the unit.
The gasoline feed to the unit is split into light and heavy fractions, with the benzene-rich, light fraction fed to the alkylation reactor. Fig. 1 shows the integration of the Alkymax process into a reformer flow scheme.
The benzene in the feed is converted selectively to isopropylbenzene (cumene) and di-isopropylbenzene. Of the benzene that is reacted, roughly 80% is converted to cumene and the rest to di-isopropylbenzene.
The propylene in the feed is converted to cumene, di-isopropylbenzene, and oligomers. The principle oligomer product is nonene.
Some hexene, heptene, and octene are also produced, along with a small amount of C10+ olefins. Table 7 shows a comparison of products with and without the process. The process does not yet have commercial experience, although the catalyst used has been commercially proven.
Although there appears to be a wide variety of processes and process configurations available to help refiners produce the reformulated fuels of the 1990s, many of them could not be used without major capital investment. Large quantities of the fuels of tomorrow won't be made without substantial costs.
But it should be noted that refining process and catalyst technology were key elements that helped refiners meet previous challenges. As the industry changes its operations to meet the fuels challenges of the 1990s, process and catalyst technology will, again, be critical to the success of refiners' efforts.
REFERENCES
- Peer, R.L., Bennett, R.W., Felch, D.E., and Kabza, R.G., "Platforming, Leading Octane Technology into the 1990s," UOP Technology Conference, Apr. 17-19, Houston.
- Lomas, D.A., Cabrera, C.A., Cepla, D.M., Hemler, C.L., and Upson, L.L., "Controlled Catalytic Cracking," UOP Technology Conference, Apr. 17-19, Houston.
- Keyworth, Donald A., Asim, Mehmet Y., Reid, Terry A., and Wilcox, Jack, "Combining fluid cat cracking with hydrotreating and extractive distillation to optimize the quality of gasoline, diesel, and jet fuel products," National Pet(oleum Refiners Association annual meeting, Mar. 26-27, San Antonio.
- Des Couriers, J., Leonard, J., Nocca, J., Bonnifay, P., and Andrews, John W., "Etherification of C4/C5 Cuts: Which Route to Choose?" National Petroleum Refiners Association annual meeting, Mar. 25-27, San Antonio.
- Yurchak, S., Child, J.E., and Beech, J.H., "Mobil Olefins Con, version to Transportation Fuels," National Petroleum Refiners Association annual meeting, Mar. 25-27, San Antonio.
- Wood, B.M., Reno, M.E., and Thompson, G.J., "Alkylate Aromatics in Gasoline via UOP Alkymax Process," UOP Technology Conference, Apr. 17-19, Houston.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.
