George H. Unzelman
HyOx Inc.
Fallbrook, Calif.
Maintaining transportation fuel quality in the regulatory environment during the 1990s will become a balancing act between regulating agencies and the petroleum refining industry.
Regulations modifying gasoline and diesel fuel composition will be promulgated to improve air quality and reduce human exposure to critical hydrocarbons.
Given enough time to generate capital and install processing, refiners should be able to meet fuel quality demand.
Regulatory agency targets include additional gasoline volatility control, an upper limit on benzene in gasoline, and restrictions on undesirable hydrocarbons in both gasoline and diesel fuel. To some degree the processing moves necessary to control gasoline composition also will modify diesel fuel composition.
Regulatory action may require application of alternative fuels in areas of high air pollution, such as Los Angeles. For example, Colorado requires oxygenates in gasoline in Denver and other front-range cities during winter months to control carbon monoxide emissions.
The refining industry has recognized the need to reformulate gasoline, not only to meet forthcoming federal and state regulations, but also to protect the crude oil-derived transportation fuel market.
The first stage of an extensive joint oil/auto industry research and testing program is under way to target the best fuel composition to reduce emissions, especially those contributing to ground level ozone formation. Early indications of possible specifications for reformulated gasoline are based on current emission control gasolines and EPA studies for future rulemaking.
Summer Rvp of gasoline may be 1-2 psi lower, aromatics could be limited at some point between 25 and 30 vol %, with additional control on benzene at the 0.5 to 1.0 vol % level.
Olefin content may be restricted by limiting the bromine number of gasoline. However, setting limits on total volume percent aromatics and olefins may be flawed because it does not necessarily eliminate critical reactive hydrocarbons.
The most difficult refining maneuver to meet reformulated gasoline targets, commensurate with maintenance of quality, will be aromatics reduction. Ethers are the only refinery components that have the potential to replace aromatics. However, with further Rvp restrictions the manufacture and blending of ethers will tend to alleviate the gasoline volatility problem.
Adjustments to the U.S. gasoline pool to meet reformulated fuel targets and maintain reasonable grade quality will require most of the 1990 decade.
This first installment of a two-part series on reformulated fuels examines current gasoline compositions, the regulatory issues that affect gasoline composition now and in the future, and the role various blending components will play in formulating low-emission gasolines.
The concluding installment will project future transportation fuel formulations, the processing strategies that will likely meet quality and environmental needs, and the effect alternate fuels might have on overall transportation fuel formulations during the 1990s.
CURRENT GASOLINE SITUATION
Lead phasedown had a major effect on the composition of the U.S. gasoline pool during the 1980s. Overall volatility rose, and the concentration of aromatics and isoparaffins increased. Low-octane normal paraffins were minimized.
MTBE (methyl tertiary butyl ether) was introduced as the only nonhydrocarbon, refinery-pool component to increase octane quality.
By the end of 1987 the major effects of lead removal on pool-composition changes were essentially complete. Allowable lead content in the leaded pool was reduced to 0.1 g Pb/gal, and the leaded grade was destined to decline at the rate of about 56%/year in concert with the attrition of lead-tolerant vehicles on the road.
However, the American public's appetite for sizable cars with adequate power output worked in tandem with the lead phasedown to continue the changes in gasoline-pool hydrocarbon composition. The oil industry seized the opportunity to aggressively market more profitable unleaded premium. Competition for premium volume resulted in higher-octane, premium-fuel quality at the pump, particularly at Eastern U.S. outlets. The increasing demand for high-octane premium fuel pushed gasoline-pool aromatics upward and placed more emphasis on etherification and isomerization processing in petroleum refining.
Table 1 shows the approximate composition and octane range of refinery components that made up the average 1988 gasoline pool.1 The 1988 estimates for the stream composition were somewhat overstated regarding the volume percents for isomerates and MTBE. Also, the light straightrun figure was low.
Therefore the volume percents and octane ranges are a reasonable representation for 1989. An aromatics range and average total for pool aromatics has been added to the earlier documentation which keys to industry data.1 In order to reach the clear pool octane of 88.4, aromatics have increased to slightly over 32 vol %.
In 1980 the U.S. unleaded gasoline pool was about 83 octane and the aromatic concentration was 10 numbers lower at 22 vol %.
Current aromatic content of the gasoline pool has been established from Motor Vehicle Manufacturers Association data and by extrapolation of earlier Ethyl Corp. information.2 3 The range of aromatics in stream compositions is taken from typical industry samples and fitted to the gasoline-pool total of 32.1 vol %.
Table 2 is an estimate of the U.S. gasoline grade mix for 1988.1 In the summer of 1989 the higher price at the pump for premium gasoline dampened the market and some volume shifted back to the unleaded regular grade. Table 3 is a current estimate of the U.S. gasoline-grade mix for 1989 which reflects the shift away from premium and declining volume of the leaded grade. Unleaded premium has lost about 3 percentage points but has increased to an average of 92.5 in octane quality. Leaded regular has dropped to 10 vol % and unleaded regular now has about 60% of the market. Average clear octane, at 88.5 of the grade mix, has slipped fractionally but has been relatively stable for the past 18 months. In fact, this is the first period of pool-octane stability since the lead phasedown in the mid-1970s.
To some extent the stability of hydrocarbon composition of the gasoline pool has paralleled the stationary clear pool octane.
Aromatic content has leveled and possibly even declined fractionally.
There has been a modest drain on pool octane from restrictions on summer Reid vapor pressure (essentially less n-butane) and the declining amount of lead antiknocks used in gasoline. The balancing octane factors are from more MTBE and the gradual and continuing exchange of isoparaffins for normal paraffins, primarily from isomerization.
QUALITY ISSUES
The so-called period of hydrocarbon-composition stability for the gasoline pool may continue for a limited period, but it is destined to end with the air-quality control fuel regulations that are currently projected. New rules governing fuel composition will affect hydrocarbons that are high in octane quality, specifically aromatics, light olefins, and n-butane.
As these rules restrict key octane contributors to gasoline, there is the question of maintaining product quality at the gasoline pump. The situation actually boils down to retaining an octane balance between Table 2 (grade mix) and Table 1 (pool composition).
The balance can be accomplished with the promulgation of reasonable fuel composition standards and reasonable industry timetables to install new processing.
The grade mix shown in Table 3, with an average clear octane quality of 88.5, should provide a satisfactory average octane quality for cars of the 1990 decade. Assuming leaded regular disappears from the market and midgrade is combined with premium, the grade ratio would approximate a 30:70 premium/regular fuel ratio. This is not too far afield from the traditional premium/regular ratio of 40:60 of the early 1970s. The basic industry challenge will be to shift the composition of the gasoline pool of Table 1 to satisfy both regulations and grade mix requirements.
REGULATED HYDROCARBONS
The oil industry has known the octane values for most hydrocarbons in the gasoline boiling range since early in the century. Blending values changed as gasoline composition changed, but the quality of individual hydrocarbons held within a fairly narrow range.
Except for n-butane, the normal paraffins are low in octane. Comparable isomers are much higher octane.
In unleaded gasoline, aromatics, olefins, and isoparaffins provide the basic octane structure, generally in that order of magnitude. Again, except for normal butane, the normal paraffins are limited by low-octane quality.
Today there is a good deal of background about which classes of hydrocarbons are good and not so good with respect to environmental factors. And the scientific community is continuing to learn more about individual hydrocarbons as they relate to public health.
For example, although gasoline generally contains only a small amount of benzene (1-5 vol %), the effects of this particular high-octane aromatic on human health are under study because benzene is a known carcinogen. In fact, aromatic hydrocarbons in general have attracted the attention of environmental groups because the concentration in gasoline has steadily risen to current levels.
Aromatics are not only high in octane quality as shown in Table 4, but they also have low vapor pressures and, consequently, the capacity to modify the volatility effects of light hydrocarbons in gasoline blending operations.4 N-butane, the only normal paraffin with an (R + M)/2 blending octane above 90, has the disadvantage of having a blending Rvp of about 60 psi.
This is a distinct disadvantage with the current summer EPA limits on gasoline Rvp specifications to help control ground-level ozone formation. Also, the schedule for the early 1990s calls for further restrictions on gasoline Rvp.
ATMOSPHERIC REACTIVITY
While benzene is undesirable from the standpoint of human health, and n-butane has limitations because of high-blending-vapor pressure, both compounds have low reactivity values based on the gas-phase reaction with the hydroxyl radical at atmospheric conditions. The so-called OH reaction rate gives an indication of photochemical reactivity with respect to ground-level ozone formations
Table 5 lists these values along with blending Rvp for representatives of three hydrocarbon classes: paraffins, aromatics, and olefins. Values also are shown for five oxygenates.
Both n-paraffins and isoparaffins have low atmospheric reactivity. The exchange of n-paraffins for higher-octane isoparaffins in the gasoline pool, which will continue as more isomerization units are installed, will have little effect on reactivity.
On the other hand, isomers are more volatile and their continued influx to the pool will require removal of additional n-butane.
In general, the heavy aromatics exhibit greater reactivity than benzene and toluene (see Table 5, xylenes and heavier). Because of high boiling points, there is less concern about reactivity with respect to evaporative emissions.
However, as butanes and other light hydrocarbons escape as vapor, some of the heavier hydrocarbons are carried along. At the same time, the heavy aromatics can enter the atmosphere from spillage and from under the hood of vehicles during soak periods, especially during warm weather.
Also, heavy aromatics are comparatively more difficult to combust and have greater tendency to appear as part of tailpipe hydrocarbon emissions. They also contribute to tailpipe benzene from decomposition during engine combustion.
The EPA, as well as state agencies, has studied benzene restrictions in gasoline and may place some limit on overall aromatics. The most favorable gasoline aromatic is toluene because it has low atmospheric reactivity, high octane quality, less toxicity than benzene, and tends to combust more completely in the engine than the heavier aromatics.
As a group, the olefins are high in photochemical reactivity. Most of the olefins in today's gasoline come from FCC gasoline and represent an important factor in front-end octane. In other words, the light olefins, along with n-butane, contribute octane quality in the low-boiling segment of gasoline as defined by the ASTM distillation curve, Several of the light olefins have blending octanes that exceed 100 at low concentrations.
Generally, C4 olefins are routed to alkylation, petrochemical operations, and MTBE processing. Little is blended directly to gasoline.
Listed in Table 5 are two very photochemically reactive C5 olefins, 2-methyl, 2-butene and 2-methyl, 1-butene, that appear in the front end of FCC gasoline. Both have relatively high vapor pressures (15 and 19 psi, respectively) and are subject to being significant contributors to evaporative emissions. While the heavier olefins in gasoline have equal or greater photochemical reactivity, there is less tendency to escape to the atmosphere.
Suggestions to control total olefins in gasoline via a limit on bromine number may be flawed because it would not necessarily restrict these most critical hydrocarbons from gasoline. The problem was summarized succinctly by Edgar R. Stephens, professor of environmental sciences and chemistry at the University of California at Riverside. In a letter to the Advisory Council of the South Coast Air Quality Management District in 1987, Dr. Stephens is quoted as follows:
"The most reactive component of gasoline currently marketed in Southern California is the olefin fraction. This has been recognized for many years. Thirty years ago, it was suggested that elimination of olefins from gasoline might eliminate smog, without the necessity of control devices for every car. When research showed that this quick fix would not work, the idea was abandoned, although a limit of 15% was placed on the olefin content of gasoline. Now that so many strategies of marginal effect are being seriously considered perhaps it is time to reconsider the olefin limitation as a contributor to smog control."
Later in the same correspondence Stevens made the suggestion that critical olefins could be removed by etherification.
The last category in Table 5 lists atmospheric or photochemical reactivity figures for methanol, ethanol, MTBE, ETBE (ethyl tertiary butyl ether), and TAME (tertiary amyl methyl ether). All of the values are low and are comparable to paraffin-hydrocarbons.
ALCOHOLS
Table 6 lists blending octane, blending Rvp, and the boiling point for the five oxygenates that have potential as blending agents in gasoline. Methanol must be used with cosolvents if blended to gasoline. Ethanol is currently blended downstream of the refinery to the extent of about 60,000 b/d and constitutes 0.8 vol % of U.S. gasoline.
Downstream blending accounts for the octane differential (average) of 0.1 number between Table 3 (88.5 octane) and Table 1 (88.4 octane). Ethanol as 10% of gasohol contributes significantly to the difference.
The experience of the last decade indicates that the alcohols have little attraction for refinery blending. They are not water tolerant, and they exhibit high blending vapor pressures with hydrocarbons.
For example, methanol has about the same blending vapor pressure in gasoline as n-butane, while ethanol blends in the Rvp range of n-pentane and isopentane. Their low photochemical reactivity values tend to be negated by azeotropic characteristics which greatly increase blending vapor pressure in combination with hydrocarbons at low concentrations.
This represents a serious economic penalty for refiners who must blend to ASTM specifications. On the other hand, pure methanol has an Rvp of only 4.6 psi.
Also, mixtures of methanol and gasoline such as M-85 (85% methanol) are relatively low in vapor pressure. The President's Clean Air Act proposal, as well as many agencies representing critical air pollution areas, strongly support methanol fuels.
METHYL TERTIARY BUTYL ETHER
MTBE has a blending octane ranging between 106 and 110, Rvp between 8 and 10 psi, and photochemical reactivity of 2.6; close to that of n-butane. It is manufactured at the refinery, blended, and transported in the same wet systems that handle hydrocarbons.
Growth for this methyl ether has been rapid, from essentially zero in 1980 to 100,000 b/d in today's gasoline pool.
This growth has been logical and natural for the petroleum refining industry. A high percentage of early construction took place in the Southwest U.S. where both isobutylene from steam cracking and methanol feedstocks were economically available.
Because of the convenient isobutylene, the capital cost for MTBE processing facilities was moderate. With the pressure of the lead phasedown, an MTBE plant could not only generate high octane blend stock, but also unload alkylation capacity to process alternative-olefin feedstock. It allowed a timely addition to pool-octane quality from two refinery units.
More recently, refiners have based ether producing facilities on the isobutylene from FCC operations. Because less isobutylene is available at lower stream concentration, these units are smaller and have higher per-barrel cost.
As the environmental demand for refinery-compatible oxygenates grows, it will be necessary to manufacture more isobutylene from n-butane by isomerization and dehydrogenation; a more capital intensive approach. For example, Phillips 66 Co. recently announced engineering for such a plant (7,500 b/d) at Borger, Tex. An alternative for small refiners will be to purchase merchant MTBE, which should be available from U.S. producers as well as from world-scale methanol plants producing coproduct ether.
While MTBE will continue to increase in volume in the gasoline pool, it is not a one-on-one replacement for aromatics. In general, the blending-octane quality is higher than most of the pure aromatics listed in Table 4 but the blending vapor pressure of the aromatics is much lower.
Capacity to absorb n-butane is important to gasoline-blending economics and will become increasingly so with more stringent environmental restrictions on Rvp. For example, the blending Rvp for toluene is about 0.5 psi vs. MTBE in the 8-10 psi range.6
Boiling point difference is another factor because it affects the ASTM distillation curve. Pure MTBE boils at 131 F. vs. a 250 F. + average for aromatics.
LOWER SEVERITY REFORMING
While it is interesting to compare the substitution of MTBE for aromatics on a direct basis, the processing adjustment at the refinery would be made by modifying naphtha reformer operations as well as other units. However, the reformer can be used as the example.
Table 7 exhibits yield inspection data for two operating conditions. Volume per cent aromatics of the C5+ (gasoline) fraction and (R+M)/2 octane has been estimated from industry data.
As reformer severity (reformate RON) is increased from 90 to 95, aromatics of the gasoline fraction (C5+) increase about 8 vol %, (R + M)/2 octane increases 4 numbers, and gasoline yield falls 3.7 vol %.
Conversely, the addition of 5 vol % MTBE to pool gasoline would allow reformer severity to back off about one (R + M)/2 octane, thus reducing reformate aromatics by 2 vol %. Gasoline yield would increase about 0.9 vol %.
Further, the Rvp of the gasoline yield would drop slightly, about 0.1 psi. Overall, the gasoline pool would directionally increase in API gravity as a result of the introduction of MTBE and the elimination of some aromatics. The trade off would tend to move the entire gasoline pool in the direction of a cleaner-burning composition.
ETHYL TERTIARY BUTYL ETHER
Obviously, MTBE is not the only ether that has the potential to replace aromatics from gasoline. ETBE has a slightly higher octane quality than the methyl ether and lower blending vapor pressure; 3-5 psi depending on the hydrocarbon composition of the gasoline. Because mixed aromatics have a blending vapor pressure of less than 1 psi, the ethyl ether would be more attractive for Rvp control. Much depends on feedstock economics and whether ethanol feedstock will be subsidized. At this writing the U.S. Internal Revenue Service is receiving comments following a Jan. 4, 1990, hearing on the proposal to make ETBE eligible for the 6/gal tax credit for the percent of ethanol contained in it. A favorable decision would shift some current MTBE units to ethanol or mixed methanol/ethanol feed.
Future units could be designed for ETBE rather than MTBE, particularly in the farm states that might follow the federal move with state subsidies. ETBE has strong support from the Executive Branch. President Bush in remarks Jan. 8, 1990, to the American Farm Bureau in Orlando, Fla., said, "Just a few months ago, we proposed the expansion of the producer tax credit for alternative fuels to include ETBE. This will mean more markets for growers and cleaner air for all Americans."
TERTIARY AMYL METHYL ETHER
The most reactive hydrocarbons shown in Table 5 are the C5 olefins, 2-methyl, 2-butene and 2-methyl, 1-butene. Both are components of FCC gasoline and have blending octane quality in the 100 range, and Rvp's of 15 and 20 psi. They can be converted to tertiary amyl methyl ether (TAME) via the reactions with methanol. The conversion results in a methyl ether with octane quality above 100 and blending Rvp in the 1-2 psi range.
Current plants in England and France manufacture TAME gasoline rather than a relatively pure ether stream. The front end of FCC gasoline is used as the olefin feed stream, and the critical C5 olefins are converted to ether. There is little doubt that the C5 olefins, some of the most reactive with respect to ozone formation, will be under severe pressure in the 1990s as the study of fuel reformulation progresses. In 1989 ARCO Chemical Co. authors summarized the advantages of TAME to the environment:8
"Of all the ethers, TAME can probably provide the most environmental benefit. It reduces tailpipe emissions and converts some very volatile and highly reactive C5 olefins in the gasoline pool into a very low vapor pressure and clean burning ether. This Rvp reduction will also provide the refiner with added flexibility in meeting the future Rvp control regulations. "Unlike the Rvp controls which will remove butanes from the gasoline during the summer months at a significant economic penalty to the refiners, the TAME operation has the potential to remove 70,000-90,000 b/d of the highly reactive 2-methyl butenes from the gasoline pool all year round with no economic penalty."
ANTIKNOCK ADDITIVE
MMT (methylcyclopentadienyl manganese tricarbonyl) has been used commercially in gasoline as an octane improver since 1957, primarily in leaded gasoline.9 10 It has been successfully blended in Canadian unleaded gasoline for many years at the 1/16 g Mn/gal level.
The antiknock, even at 1/32 g Mn/gal, could offer close to 1.0 (R + M)/2 octane depending on gasoline composition and octane level. As an additive in U.S. unleaded gasoline, the octane improver could help lower the level of aromatics as well as contribute to reduced gasoline volatility. Table 7 can be used to estimate the positive effect of MMT on naphtha reforming operations which contribute approximately 34 vol % of all U.S. gasoline. MMT was banned by the Clean Air Act amendments of Aug. 3, 1977, for use in unleaded gasoline unless EPA granted a waiver. Subsequently, two waiver requests were denied based on studies showing statistical increases in hydrocarbon tailpipe emissions from fuels treated with the manganese antiknock. Ethyl Corp. plans to submit a third waiver to EPA during 1990, based on new studies involving a 48-car fleet, representing high-volume production vehicles from U.S. manufacturers. Based on historical data, the use of MMT at low levels in gasoline would probably shift the balance among tailpipe carbon monoxide, nitrogen oxides, and hydrocarbons slightly due to chemical effects on catalyst performance. From an environmental standpoint, the risk would seem minor compared to the advantage of reducing overall gasoline volatility and aromatics without capital investment at the refinery.
Refiners will need all the quality help they can get as new rules on fuel composition are promulgated by EPA. While the auto makers have traditionally opposed MMT, octane quality from whatever source allows flexibility to increase compression ratio as an aid to meet CAFE (Corporate Average Fuel Economy) requirements. The advantage to the refiner is obvious when one considers that in a 100,000-bbl gasoline blend, about two 55-gal drums of MMT can produce the octane equivalent of 5,000 bbl of MTBE. (Calculations are based on raising unleaded regular from 87 to 88 ON with 1/32 g Mn as MMT, octane blending value for MTBE of 108 and no Rvp adjustment.) In the final analysis it would seem that an antiknock for unleaded gasoline has positive features for the automakers and regulators as well as petroleum refiners.
REFERENCES
- Unzelman, George H., "Future Role of Ethers in U.S. Gasoline," paper No. AM-06, NPRA annual meeting, San Francisco, Mar. 19-21, 1989.
- "U.S. Gasoline Outlook, 1989-1994," Information Resources Inc., Washington, D.C., October 1989.
- Hall, C.A., "Effect of Government Antiknock Regulations on Demand for Aromatics in U.S. Gasolines," Annual Fall Conference, U.S. Chemical Marketing Research Association, West Germany, Oct. 16, 1979.
- Morris, W.E., "Octane Blending Effects of Aromatics," paper No. AM 80-43, NPRA annual meeting, New Orleans, Mar. 22-24, 1980.
- Piel, W.J., and Thomas, R.X., "The Role of Oxygenates in Reformulated Gasoline," Mobile Source Clean Air Technology Conference, Detroit, Feb. 21-22, 1990.
- Unzelman, George H., "Gasoline Volatility-An Increasing Industry Problem," paper No. AM-86-34, NPRA annual meeting, Los Angeles, Mar. 23-25, 1986.
- Feldman, M.B., and Rangnow, D.G., "Modern Gasoline Economics," Hydrocarbon Processing, December 1982.
- Miller, D.J., and Piel, W.J., "Ether Options: MTBE/TAME and ETBE," paper No. AM-8958, NPRA annual meeting, San Francisco, Mar. 19-21, 1989.
- Unzelman, George H., "Learning More About an Antiknock Called MMT," Octane Week, Oct. 5, 1987.
- McChesney, J.M., and Burns, A.M., "MMT An Antiknock for the Past, Present and Future," National Conference on Octane and Oxygenated Fuels, San Francisco, Mar. 21-23, 1989.
- Cohu, L.K., et al., "EC-1 Emission Control Gasoline," ARCO Products Co., Anaheim, Calif., September 1989.
- Unzelman, George H., "Diesel Fuel Quality: Refining Constrictions and the Environment," paper No. AM-87-33, NPRA annual meeting, San Antonio, Mar. 29-31, 1987.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.
