TECHNOLOGY Low sulfur diesel controversy continues

Roger L. Leisenring Jr.
Texaco Inc.
Beacon, N.Y.

The introduction of low-sulfur, on-highway diesel fuel into the U.S. market in October 1993 brought to light several concerns. Among these, the predominant issues were lubricity, low-temperature operability, and dye regulations.

Use of low-sulfur diesel has not been reported to create lubricity problems, as occurred with Sweden's ultralow-sulfur fuel. Industry tests have shown that new engine designs are significantly less sensitive to changes in the aromatics, cetane, and sulfur content of diesel fuels.

A review of testing to date, and a discussion of the test methods available to evaluate diesel's properties, will set the stage for a discussion of the future direction of diesel regulations.

Lubricity

The refining industry has spent considerable time studying diesel lubricity. At least in the U.S., reports of equipment failures resulting from use of low-sulfur diesel are not wide-spread.¹ A look back at the origin of lubricity concerns in the U.S., therefore, is appropriate.

They were reported during a forum at the Society of Automotive Engineers' (SAE) annual congress.2 One conclusion from the forum was that fuels suspected of causing lubricity-related problems were primarily low-sulfur No. 1-type diesel fuels. These fuels contain less than 0.05 wt % sulfur and have viscosities below 2.0 cSt at 40^o C.

After this forum, several reports indicated that the common element in all fuel-pump failures appeared to be fuels with low cloud points, formulated for winter operation, and very low sulfur contents-typically around 0.01 wt %.^{3 4} In addition, the problem was reported to be more troublesome in the spring, when daytime temperatures can increase suddenly. (There is a dual effect from low lubricity and low viscosity.)

These reports eventually led the members of SAE Committee 7 to recommend that a footnote be included in the American Society for Testing and Materials (ASTM) standard specification for diesel fuels, D975. As a result, a discussion of lubricity has been included in the appendix of D975.

How did this concern about No. 1-type fuels grow to include low-sulfur No. 2 fuels? The answer probably comes from experiences in Sweden.

In January 1991, Sweden enacted legislation that established two classes of low-sulfur diesel fuel:

• Class 1, with a maximum sulfur content of 0.001 wt % (10 ppm) and a maximum aromatics content of 5 vol %

• Class 2, with a maximum sulfur content of 0.005 wt % (50 ppm) and a maximum aromatics content of 20 vol %.

Severe hydrotreating conditions are necessary to meet these stringent specifications.

Hydrotreating has the potential to reduce the lubricity properties of diesel fuel. This is because hydrotreating reduces trace components, such as oxygen and nitrogen-containing compounds, as well as polycyclic aromatics. These naturally occurring, polar compounds adsorb onto metal surfaces to form a protective low-friction layer.

When these severely hydrotreated fuels were marketed in Sweden, about 70 light-duty vehicles equipped with rotary distributor-type fuel pumps failed; however, no wear problems were reported for heavy-duty vehicles with in-line fuel pumps.

At about the same time, the U.S. Environmental Protection Agency (EPA) and the California Air Resources Board (CARB) each issued specifications for low-sulfur diesel. The CARB regulation reduced aromatics content in California fuels to as low as 10 vol % for some refineries.

Although these requirements were discussed throughout the U.S. refining industry, the relative hydrotreating severity needed to produce these fuels, as compared to Swedish fuels, often was not covered.

Without being fully aware of the differences in hydrotreating severity, many speculated that the U.S. federal and California low-sulfur diesel fuels would be similar in composition to the Swedish fuels. Fortunately, however, Swedish fuels (essentially in the kerosine distillation range) are dissimilar to U.S. low-sulfur diesel No. 2. or CARB diesel fuels.

The U.S. has had limited experience with the use of low-sulfur diesel. It has been produced and marketed in Southern California since 1985, with no evidence of field problems.

Even with this experience, the refining industry fully participated in efforts, coordinated through SAE and the International Standards Organization (ISO), to develop lubricity measures.

Field tests

Several test methods have been evaluated in an international round robin. Two of the methods used are: a ball-on-cylinder lubricity evaluator (Bocle)-modified by the U.S. Army and designated by ASTM as scuffing-load Bocle, or Slbocle-and the high-frequency reciprocating rig (HFRR) test. These methods have demonstrated some correlation to original equipment manufacturer (OEM) fuel-pump rig tests.

The HFRR was recommended by the ISO round-robin task force late in 1994 for consideration as a standard bench test, with a specification within the test method for determining the lubricity of diesel fuel.

A bench test to measure the lubricity characteristics of a base diesel fuel (without additives) may be acceptable. Caution is essential, however, when assessing fuels containing additives.

As a case in point, Chevron Corp. and Stanadyne Automotive Corp. conducted a vehicle test using a 1982 Chevrolet C-10 truck with a 6.2-l. diesel engine. The engine was equipped with a standard Stanadyne fuel-lubricated, rotary distributor-injection pump.5 The vehicle was operated on a chassis dynamometer using a 0.7 vol % aromatics test fuel without additives.

The lubricity rating for this fuel, as measured by ASTM D5001 Bocle, was about 0.85 mm wear scar diameter (WSD). The average mileage for catastrophic injection pump failure was 375 miles.

The final test was conducted with the original 0.7 vol % aromatics fuel, plus 25 ppm of carboxylic-acid-based, lubricity-enhancing additive (corrosion in hibitor). This mixture produced about a 0.60-mm WSD, as measured by ASTM D5001.

The vehicle operated for 20,000 miles on this fuel without experiencing any problems. If this additive-containing fuel had been evaluated in the U.S. Army modified Bocle, it would not have shown any improvement in lubricity; however, the 20,000-mile field evaluation demonstrated a definite improvement.

Shell Canada Ltd. and the U.S. Army performed similar tests, with comparable results.^6-12 Shell's results are shown in Table 1 [19704 bytes].

California's experience with lubricity-enhancing additives may support the theory that the Slbocle is acceptable for measuring the lubricity characteristics of a base diesel fuel, but not for assessing fuels with additives.

In February 1994, CARB recommended that petroleum manufacturers add lubricity-enhancing additives to diesel fuels with scuffing-load capacity of less than 3,000 g, as measured by the Slbocle test.

Many in the industry believe that high dosages of carboxylic acid-based, lubricity-enhancing additives can have detrimental effects. One could assume, therefore, that California fuels containing lubricity-enhancing additives are likely to have additive concentrations considered customary, or at least not extreme, compared to the high concentrations indicated by the Slbocle to provide an improvement over base fuel.

If this is the case, the Chevron and Shell studies support CARB's conclusion, made in March of this year, that after more than a year of experience, there is no indication that pump damage, such as occurred in Sweden, has occurred with California fuels.¹³

Still interested in developing a test that measures lubricity and correlates with field data, ASTM formed a Lubricity Task Force. The objectives of the task force are to propose test methods for evaluating the lubricity characteristics of diesel fuel and, if needed, a specification limit.

At the time of this writing, two test methods, Slbocle and the HFRR, have passed three levels of approval in ASTM. Results from the final approval level are expected in 3-6 months.

Cold-flow properties

The low-temperature operability of diesel fuel continues to be a concern of the Engine Manufacturers Association (EMA). Consumers' interest in this area was dampened as a result of the mild winter of 1994-95.¹⁴ The concerns now being expressed deal more with whether current ASTM requirements adequately protect the consumer.

To address this issue, ASTM formed the Diesel Fuel Low Temperature Operability Task Force under the responsibility of Section E2 (diesel fuel). The objective of the task force is to evaluate alternative test methods to the cloud point test for predicting low-temperature operability of distillate fuels in North American diesel-powered vehicles, and if necessary, to revise the guidelines.

The task force will attempt to resolve four main issues:

• The maps of the tenth-percentile minimum ambient temperature, which appear in the Appendix of ASTM D975, are vague, inaccurate, and should be rewritten so that the average person can understand their meaning. It has been suggested that ASTM continue to make the maps available, but include a table of the temperature values.

• D975, Table 1 [19704 bytes], Footnote I, states: "...satisfactory operation should be achieved in most cases if the cloud point is specified at 6^o C. (10.8^o F.) above the tenth percentile minimum ambient temperature." This statement continues to be a concern. Several industry sources believe that this 6^o C. differential should be removed entirely, which means they believe a vehicle would fail at precisely the minimum ambient temperature.

  At the Dec. 5, 1994, meeting of ASTM's Section E2, the author presented historical information indicating that the 6^o C. offset does have technical merit.¹⁵ The question now is whether this has changed, because the supporting data were taken in tests of 1960s and 1970s-era vehicles.

• The practice of preparing "winterized" diesel fuel by blending ASTM Grades No. 1 and No. 2 produces a final blended product that may not fully meet the specifications of either Grade No. 1 or Grade No. 2. It will require considerable effort for the task force to develop a recommendation to cover existing and future specifications for such a fuel.

• Currently, the only statement within D975 regarding a test method to determine low-temperature operability refers to the cloud point test. An alternative test method is needed to evaluate fuels containing low-temperature operability additives.

As stated in the "Winter Diesel Question and Answers" pamphlet published by the American Petroleum Institute (API), two test methods have been used to predict vehicle and fuel low-temperature performance: the cold filter plugging point (CFPP) test and the low-temperature flow test (LTFT, or ASTM D4539).

The LTFT is the only test method for which significant vehicle operability test data have been gathered to support its use for predicting North American vehicle and fuel operability. The limiting aspect of this test method is the amount of time needed to run it (8 hr or more).

Dye regulations

Federal regulations require the dyeing of diesel fuel. The EPA low-sulfur diesel regulation, which took effect on Oct. 1, 1993, and was later revised, requires that noncomplying high-sulfur (0.05 wt %) fuels be dyed red.¹⁶

U.S. Internal Revenue Service (IRS) regulations require tax-exempt low and high-sulfur diesel fuels to contain the dye Solvent Red 164 at a concentration spectrally equivalent to 3.9 lb/1,000 bbl (ptb) of solid dye standard Solvent Red 26 (at least 5.6 ptb active Solvent Red 164). Each dye supplier has its own dye formulation and dosage recommendation to meet the "spectral equivalent" of 3.9 ptb of Solvent Red 26.

The industry uses two test methods to measure red dye concentration in the field.¹⁷ One test uses spectrophotometric absorbance, with the supplier's dye formula as a standard. The second method was developed by the IRS, working with the U.S. Air Force Aerospace Laboratory. It is a computer-facilitated spectrophometric method, in the visible range. This technique uses a second-derivative method for background compensation.

The second derivative of the diesel absorption spectrum is used to eliminate any background broad-wavelength absorption derived from the base color of the fuel. This background interference otherwise would add a positive bias to the absorption values at the analytical frequencies used.

The second-derivative analysis produces a plot with two definitive maxima, plus a strong minimum at maximum absorption. The linear difference between the maximum and minimum can be used to quantify the peak against standard solutions of dye.

The amplitude difference between the first maximum at 538.0 20 nanometers (nm) wavelength and the subsequent minimum at 561.0 20 nm determines the concentration of the red dye. Standards of known dye concentration in xylene are used to generate a concentration curve before dyed diesel samples are tested.

The IRS has determined that Solvent Red 26 standard must be 99.8% pure. Fuel tested for dye must meet the standard of at least 3.9 ptb (11.1 mg/l.) of 99.8% Solvent Red 26 dye.

The IRS first reported the repeatability of this test method to be 0.83 mg/l. at 95% confidence level. Because the test was developed by a single laboratory, however, the industry was able to convince the IRS to conduct a round robin to establish the repeatability and reproducibility of the method.

The round robin involved 20 different instruments analyzing 20 samples of No. 1 and No. 2 diesel fuel of varying colors and dye concentrations.

In response to the need for a portable analyzer, Boston Advanced Technologies Inc. developed a small portable device to measure the concentration of red dye in diesel fuel. The instrument (PetroSpec DT100) has shown good results when tested on the fuel samples used in the round robin tests. In addition, the IRS has purchased this equipment for its field compliance officers.

In March 1995, IRS and Treasury Department officials met with industry representatives to discuss a proposal that the dyeing regulations eliminate splash blending and mandate dye injection equipment at terminals for dyeing low-sulfur diesel for nontaxable use. The IRS established four parameters for the dye injection equipment.

The equipment used to dye diesel fuel must:

• Dispense dye at a specified rate

• Be calibratable

• Have secure control devices

• Incorporate automatic malfunction shut-off.

The IRS suggested the additive injection equipment be secured with a government-issued seal, and that operators be required to notify the IRS whenever the seal is broken. A final decision on the IRS/Treasury proposal has not been reached.

The future

Both EPA and the EMA have stated that new diesel fuel specifications will play a role in helping engines meet future exhaust emission standards. In addition, engine designs may play a role in the development of future diesel fuel specifications because of the improvements fuels provided in the 1994 engine design changes.

By lowering the fuel sulfur content to 0.05 wt % from the typical pre-1993 level of 0.25 wt %, the sulfate portion of diesel exhaust was reduced significantly. Engine manufacturers would not have been able to meet the particulate standard for 1994 engines without this sulfur reduction; therefore, this regulation was rational and technically justified.^17L

In August 1992, however, EMA approached API with a long list of diesel property changes. The list included aromatics reduction, sulfur reduction, distillation curve control, API gravity control, and cetane enhancement.

In January 1993, the EMA recommended that API raise the minimum cetane number of highway diesel fuel from the current ASTM D975 level of 40 to a new level of 55. The EMA felt that a higher cetane number would assist them in complying with the 1998 NO_x emissions standards, because some studies have found cetane number to have an effect on NO_x emissions from diesel engines.

API responded with an analysis of the cost and cost-effectiveness of raising cetane to lower NO_x emissions. The analysis concluded that raising cetane number was not as cost-effective as modifying engine hardware to lower NO_x emissions. In fact, raising cetane to 55 to reduce NO_x is 10 times more costly than changing engine hardware. Increasing cetane number from 40 to 55 reduces NO_x emissions by about 0.¹⁵ g/bhp-hr, which is only one-sixth the reduction required for certifying 1998 engines.

The most recent emissions studies by the Coordinating Research Council (CRC) evaluated the effects of aromatics and cetane on exhaust emission using prototype 1994 and 1998 Detroit Diesel (DD) Series-60 engines.^{18 19} Emission results were measured using the EPA transient federal test procedure on the 1994 DD Series-60 engine tuned for 5 and then 4 g/bhp-hr NO_x calibration.

The purpose of calibrating the 1994 engine to a 4 g/bhp-hr NO_x level was to examine fuel effects on emissions for engines restricted to lower NO_x levels. In other words, the engine was tuned to represent a 1998 engine. (At the time of the study, the 1998 DD Series-60 engine was a true prototype because of hardware changes and a new computer control module for fuel metering and timing.)

CRC tested a number of fuels and developed predictive exhaust-emission models from the results. Using the models from the CRC's VE-1 and VE-10 programs, Fig. 1 [30433 bytes], Fig. 2 [30955 bytes], and Table 2 [20971 bytes] were developed.

In 1994, the National Institute for Petroleum & Energy Research (Niper) conducted a survey of low-sulfur, on-highway diesel fuel. The survey revealed that an aromatics content of 30 vol % and a cetane number of about 45 are typical for low-sulfur diesel No. 2.

Results from the prototype 1991 DD Series-60 engine showed that, with a decrease in fuel aromatics to 20 vol % from 30, (maintaining a cetane number of 45), NO_x emissions decreased 2.9% and particulate emissions, 3.6%.20 In the same engine, an increase in cetane number from 45 to 50 (maintaining 30 vol % aromatics) caused NO_x emissions to decrease 1.7% and particulate emissions, 6.8%.

The prototype 1994 DD Series-60 engine, calibrated to 5 g/bhp-hr NO_x, showed that with a decrease in fuel aromatics from 30 vol % to 20 vol % (45 cetane number), NO_x emissions decreased 2.0%, with no change in particulate emissions. In the same engine, an increase in cetane number from 45 to 50 (at 30% aromatics) caused NO_x emissions to decrease only 1.2%, with no change in particulate emissions.

The prototype 1994 DD Series-60 engine calibrated to a 4 g/bhp-hr NO_x showed that with a decrease in fuel aromatics from 30 to 20% (45 cetane number), NO_x emissions decreased by 1.7%, again with no change in particulate emissions. In the same engine, an increase in cetane number to 50 from 45 (at 30% aromatics), caused NO_x emissions to decrease by 2.2%, with no change in particulate emissions.

Finally, testing in the prototype 1998 DD Series-60 engine showed that with an increase in cetane number to 50 from 45 (at 30% aromatics), NO_x emissions decreased 2.1%, and particulate emissions, 3.3%.

Change in aromatics was not investigated with this engine. Because the change in cetane number resulted in the same NO_x emissions as in the 1994 DD Series-60 engine calibrated to 4 g/bhp-hr NO_x, similar changes in NO_x emissions as a result of changes in aromatics concentration could be expected.

Another way to look at the data is to maintain the same change in fuel property (in this case, aromatics) and vary the model year of the engine. Table 3 [11756 bytes] shows that, as engine technology advances to meet more stringent exhaust emissions, fuel changes have a decreasing impact on emissions.

Fig. 3 [31050 bytes] presents a more global look at emission standards and the effect that fuel changes have in meeting these standards. Included in the figure are: work Texaco Inc. conducted in the early 1990s, a wider range of fuel properties than were investigated in the CRC studies, and variability in engine emissions due to production tolerances (called "engine production emissions").

Fig. 3 [31050 bytes] clearly shows the small effect on emissions from large changes in fuel property. Boxes A and B represent, respectively, NO_x and particulate emissions as aromatics and cetane were varied.

The circles representing 1988 and 1991 production-emission variability were presented by an engine manufacturer. The 1994 circle was an estimate based on the 1988 and 1991 circles.²¹

From this figure, it appears that the variability in engine production emissions has just as much of an impact on emissions as the fuel.

The effect of diesel sulfur content on exhaust emissions can be readily calculated.²² The sulfur component of diesel fuel is converted to sulfate during and after combustion, and is measured as part of the particulate matter. Changes in sulfur content affect only total particulates and not NO_x emissions.

The 1998 particulate emissions standard is the same as the 1994 particulate standard (0.1 g/bhp-hr). Based on this, trying to comply with the 1998 particulate standard by reducing sulfate emissions through further reduction of fuel sulfur would achieve no emissions benefit.

Current fuel sulfur specifications allow engine manufacturers to meet known future standards for particulate emissions. Even if the standard were reduced to 0.05 g/bhp-hr, the sulfate portion of the particulates collected would only be 0.002 g/bhp-hr, based on a fuel sulfur content of 0.01 wt %.

From these bodies of data, it is obvious that new engine designs are significantly less sensitive to changes in the aromatics, cetane, and sulfur content of fuels. In addition, fuel formulations have some effect on diesel exhaust emissions. These effects are relatively small, however, and major reduction in emissions can be expected only with hardware design.

If fuel changes have only a small impact on engine emissions, further engine modifications and exhaust aftertreatment will be required to meet future regulations.

Engine modifications helped achieve compliance with U.S. emission regulations. For 1998, however, a 20% reduction in NO_x is required (NO_x is legislated to be reduced from 5.0 to 4.0 g/bhp-hr). After this date, engine modifications may not be adequate to reach future regulated NO_x levels.

The use of some form of exhaust aftertreatment may be necessary. There are three forms of exhaust aftertreatment: lean NO_x catalysts, oxidizing catalysts (for hydrocarbons and particulates), and soot traps. The probability of using any of these technologies is presented in Table 4 [16204 bytes].²³ Within the next 5-10 years, it appears that some use of exhaust aftertreatment is probable.

At the May 1995 SAE meeting on diesel exhaust aftertreatment, future diesel fuel needs were presented and discussed. Future diesel catalysts may have some ability to reduce NO_x while maintaining the ability to oxidize particulates, CO, and hydrocarbons. The use of catalysts containing significant amounts of platinum may require further reductions in the sulfur content of diesel fuel.

The definition of an ultralow-sulfur diesel fuel appears to be 0.01 wt % or less. It has been stated that this level of sulfur in fuel would improve the efficiency of the catalysts by permitting improvements in design and application at higher exhaust temperatures, without an increase in sulfates to contribute to particulate emissions.

It is premature, however, to specify a sulfur limit until prototype engine modifications have reached a level of diminishing returns for engine emissions. In this way, the necessary catalyst efficiency will be determined, after which a realistic diesel-fuel sulfur level can be explored.

References

1. General Motors Service Bulletin No. 376303, February 1994.

2. SAE TC-7 Update Report to ASTM, Ron D. Tharby, June 19, 1991.

3. Letter from Thomas Sheahan, Chairman of SAE F&L Division, to Ed White, Chairman of ASTM D-2 on Petroleum Products and Lubricants, June 7, 1991.

4. Report submitted to ASTM D-2 on Petroleum Products and Lubricants, Subcommittee E, Section 2, San Francisco, California, June 18, 1990.

5. Nikanjam, M., and Henderson, P. T., "Lubricity of Low Aromatics Diesel Fuel," SAE Paper 920825, February 1992.

6. Minutes of the ISO TC22/SC7/WG6-Diesel Fuel Lubricity, Feb. 27, 1995.

7. U.S. Army Fuels and Lubricant Quarterly Bulletin, Dr. Paul Lacey, pp. 2-4, Vol. 16, No. 1, December 1993.

8. U.S. Army Fuels and Lubricant Quarterly Bulletin, Fact or Friction, p. 10, Vol. 17, No. 2, March 1995.

9. U.S. Army Fuels and Lubricant Quarterly Bulletin, pp. 1-3, Vol. 12, No. 4, September 1991.

10. Ward's Engine and Vehicle Technology Update, p. 7, July 1, 1992.

11. U.S. Army Fuels and Lubricant Quarterly Bulletin, p. 3, Vol. 14, No. 3, June 1992.

12. "Fuel Lubricity Requirements of Stanadyne Fuel Injection Pump," Letter from S. J. Lestz, Director, Belvoir Fuels and Lubricants Research facility (SwRI), to M.E. LePera, Commander, U.S. Army Belvoir Research, Development and Engineering Center, June 8, 1991, File:02-1955-150, ISO TC22/SC7/WG6/N86.

13. Final Report for 1993-1994 on Lubricity of Diesel Fuels from California Refineries, California Air Resource Board Project No. C93-095A, Mar. 16, 1995.

14. McCarthy, Chris, "Update on the Effect of Government Regulations on Diesel Fuels," Paper FL-94-115, 1994 National Petroleum Refiners Association Fuels & Lubricants Meeting, Nov. 3-4, 1994.

15. Leisenring, Roger L., Jr., "Table 1 [19704 bytes], Footnote I, Justification", ASTM Section E2 Meeting, Dec. 5, 1994.

16. Regulation of Fuels and Fuel Additives: Fuel Quality Regulations for Highway Diesel Fuels Sold in 1993 and later Calendar Years, EPA, Federal Register/Vol. 55, No. 162/8-21-90/Rules and Regulations, p. 34,120.

17. Nalco/Exxon Energy Chemical, L.P., Diesel Dye Update, Cheryl George, No. CLM 95-3, May 1995.

18. Ullman, Terry L., Spreen, Kent B., and Mason, Robert L., "Effects of Cetane Number on Emissions From a Prototype 1998 Heavy-Duty Diesel Engine," SAE Paper 950251, February 1995.

19. Ullman, Terry L., Spreen, Kent B., and Mason, Robert L., "Effects of Cetane Number, Cetane Improver, Aromatics, and Oxygenates on 1994 Heavy-Duty Diesel Engine Emissions," SAE Paper 941020, February 1994.

20. Ullman, Terry L., Mason, Robert L., and Montalvo, Daniel A., "Effects of Fuel Aromatics, Cetane Number, and Cetane Improver on Emissions from a 1991 Prototype Heavy-Duty Diesel Engine," SAE Paper 902171, October 1990.

21. Panel Discussion, James E. Sibley, SAE Fuel & Lubricant Meeting, October 1994.

22. Southwest Research Institute Proposal 03-3639(012), Research Consortium: "The Role of Fuel Sulfur in Achieving Diesel Exhaust Particulate Standards," Letter to Texaco, Nov. 5, 1985.

23. Farrauto, Robert J., Engelhard Corp., SAE TOPTEC, May 24-25, 1995.

The Author