German refiner produces ultralow sulfur diesel

June 2, 2003
Bayernoil Raffineriegesellschaft MBH recently revamped a diesel hydrotreater in its Neustadt refinery to produce ultralow sulfur diesel (ULSD)—one that contains

Bayernoil Raffineriegesellschaft MBH recently revamped a diesel hydrotreater in its Neustadt refinery to produce ultralow sulfur diesel (ULSD)—one that contains <10 ppm (wt) of sulfur. The unit started up in May 2002. Bayernoil conducted numerous test runs in June-September 2002 to ensure that unit performance satisfied design specifications.

The unit started producing ULSD in October 2002 in preparation for the official introduction of the European diesel specification in January 2003.

Combined with a successful revamp in 1999-2000, the current project gives Bayernoil two diesel hydrotreaters that produce ULSD and one distillate hydrotreater that makes light heating oil (<0.2 wt % sulfur).

Bayernoil's experience demonstrates that existing units can be revamped to produce ULSD at substantially lower costs than building a grassroots unit. As much as 18% more unit capacity can result from moderate equipment modifications.

This article also covers different approaches to maintain and enhance the reliability of producing ULSD.

Because every refiner processes a different crude slate and has different sets of processing equipment, the choice of technology options to produce ULSD is site specific. A successful revamp requires a thorough evaluation of all options and an understanding of the limits of each option. In many cases, a combination of options is required to meet a refiner's needs at the lowest capital expenditure.

In preparation for producing ULSD, Bayernoil evaluated various options and adopted this strategy to minimize costs:

  • Revamp existing diesel hydrotreating units.
  • Use a high-activity catalyst.
  • Use online sulfur analyzers.
  • Optimize operating conditions and feed composition.

Diesel regulations

Oil refiners in Europe, similar to those in the US, have faced the problem of overcapacity, poor gross margins, and tightening of environmental regulations.

The European Commission mandated a reduction in diesel sulfur levels from 500 ppm (wt) to 350 ppm (wt) in 2000 with a further reduction to 50 ppm (wt) in 2005. Refiners started marketing diesel with 50-ppm (wt) sulfur in 2000, which several European countries encouraged through tax incentives. In Germany, early introduction of 10-ppm (wt) diesel, starting Jan. 1, 2003, was encouraged via tax incentives.

A similar trend is occurring in the US and the rest of the world. US refiners must produce a highway diesel with a 15-ppm (wt) sulfur cap starting in mid-2006.

Revamp options

To minimize capital investment, refiners should consider low or moderate-cost options to revamp existing diesel hydrotreaters before investing in new grassroots units. Options that increase the capabilities of existing units to produce ULSD are (in ascending order of capital cost):1-7

  • Use improved catalysts. Advanced hydrotreating catalysts, with desulfurization activities 25-75% higher than older catalysts, can significantly increase the desulfurization capability of existing hydrotreaters. Additional changes, however, may be required to produce significant amounts of ULSD.
  • Adjust feed end point and tailor the feed diet. One way refiners can produce ULSD in existing units is to use "easier" feeds, such as straight-run kerosine and light gas oil. Many existing hydrotreaters can produce some ULSD by cutting out most of the refractory sulfur compounds contained in high-end-point materials and by reducing aromatics in the feed, combined with the use of high-activity catalysts. Refiners, however, must find a home for these higher-end-point refractory cuts such as heating oil.
  • Raise reactor temperature. This obvious option has limited effectiveness and will incur additional costs due to a shorter catalyst life.
  • Improve reactor efficiency. Better vapor-liquid distribution in hydrotreating reactors can increase catalyst utilization significantly. The extent of improvement depends on the type of existing vapor-liquid distributors and feedstock quality.
  • Add catalyst volume. More catalyst in an added reactor is another simple and moderate-cost method to increase desulfurization capability of an existing unit. The amount of additional catalyst depends on the feed diet and process conditions.
  • Increase H2-to-oil ratio. Increasing the treat gas rate (H2-to-oil ratio) can enhance the catalyst's desulfurization by reducing the inhibition effect of H2S and NH3. The effect is rather limited, however.
  • Remove H2S from treat gas. H2S inhibits catalyst activity; therefore, removing the H2S in the treat gas will improve desulfurization activity. Although this option is more effective than increasing the H2-to-oil ratio, the refiner should evaluate the cost compared to other options.
  • Increase H2 partial pressure. This is another option that improves sulfur removal and prolongs catalyst life. Again, the refiner should evaluate the cost (e.g., adding a hydrogen purification system or using imported hydrogen) against other options.

Some of these options were successful with less-stringent diesel sulfur specifications in the US and Europe. A 15-ppm cap or lower, however, is significantly more difficult to meet than a 500 or 50-ppm specification.

Bayernoil's revamp experience

Bayernoil has three diesel hydrotreaters; two units produce ULSD and one produces light heating oil (0.2 wt % sulfur).

In 1999-2000, Bayernoil revamped two units to improve desulfurization capabilities (OGJ, Sept. 10, 2001, p. 68). The refiner revamped one unit to produce <50 ppm (wt) sulfur diesel vs. 350 ppm (wt) before the revamp, and the second unit to handle 20% more feed, produce <0.2 wt % sulfur, and achieve more than a two-fold increase in reactor efficiency.

This article covers the most recent revamp, that of the third unit, the largest of the three, to produce ULSD.

Click here to enlarge image

Before the revamp, this unit had a nominal capacity of 45,000 b/sd, a design pressure of 900 psig, and a relatively small reactor, indicated by a liquid hourly space velocity greater than 2 hr-1. The unit previously produced <350 ppm (wt) sulfur from a feed blend of mostly virgin stocks with up to 15% FCC light cycle oil and a small amount of visbreaker gas oil.

Feed sulfur was 0.2-0.6 wt % but sometimes spiked up to 1 wt % when the refinery processed higher-sulfur crudes.

Revamp scope

Bayernoil's objective was to improve desulfurization capability of the third hydrotreater and increase unit capacity to 53,000 b/sd.

Major revamp scope considerations were to minimize capital investment and to avoid major additional modifications and investments in the future in case the European Commission mandates new diesel sulfur specifications.

Bayernoil installed a new large reactor in series with the existing reactor, which increased catalyst volume by 300%.

Bayernoil added six new exchangers to the feed preheat train and rearranged the existing exchangers to recover more heat from the stripper bottoms and reactor effluent, which increased energy utilization efficiency. The feed preheat furnace needed no modifications, even with 18% more throughput.

Bayernoil installed a new high pressure, high-temperature separator to replace the existing one, which is now a low-pressure separator. The new high-pressure separator is also located closer to the reactor effluent exchanger train.

The new and rearranged feed preheat exchangers decreased compressor head by about 30 psi. Estimated pressure drop in the new reactor was 23 psi.

The exchanger modifications, therefore, improved unit hydraulics and energy efficiency even with higher gas and liquid rates, and a new reactor.

Bayernoil performed several additional modifications to debottleneck the unit's hydraulic constraints:

  • Feed pumps. Impellers replaced to increase charge rate.
  • Makeup gas compressors. Moved tie-in from the suction to the discharge of the recycle gas compressor to increase total treat gas rate.
  • Reactor effluent air coolers. Bundles replaced to handle greater flow rates.
  • Low-pressure separator. This was replaced with the former hot HP separator to handle the greater flow rates.
  • Product stripper. Trays modified for greater loads.
Click here to enlarge image

Fig. 1 shows a flow diagram that highlights the new and modified equipment.

Cost-effective reactor design

Bayernoil considered several reactor and process vendors to design the new reactor. The refiner selected H2Advance, Cherry Hill, NJ.

The reactor has an inside diameter of 18 ft, which is a world-scale size for trickle bed reactors. It has three catalytic beds separated by quench zones and a total reactor tangent-to-tangent height of 70 ft. The reactor shell weighs more than 450 tonnes.

Large reactors require multiple-bed designs to provide proper temperature control and fluid redistribution.2 Many conventional quench zone designs, however, are ineffective or require excessive reactor height to accommodate the reactor internals.

H2Advance's quench zone design reduces the reactor capital requirement by 5-10%. It uses a novel quench nozzle and a mixing chamber that reduces the quench-zone height while improving the thermal equilibrium and phase mixing.

The mixing chamber design provides thorough mixing of the fluid phases to attain a uniform fluid temperature from the different bed quadrants above the quench zone.

Click here to enlarge image

During process test runs in June 2002 at near-design feed rates, an on-spec diesel product indicated minimal flow channeling in the reactor. At 99.8% feed desulfurization, any measurable channeling or bypassing will increase the product sulfur to more than 10 ppm (wt).

Click here to enlarge image

Radial temperature measurements at the bed inlet showed that the maximum temperature differences were less than 2° C. during the test run (Figs. 2-4), which indicated an effective distributor and quench-zone design.

Click here to enlarge image

Gay Engineering & Sales Co. Inc. FLEX-R thermocouples, installed during the turnaround, provided bed temperature measurements. Bayernoil installed three sets of circular thermal couple bundles, each with six temperature readings, in each of the three beds. Thermocouple locations in the catalyst bed are: one near the top with a 2.75-m diameter and two at the bottom of bed, an inner circle with a 2.75-m diameter and an outer circle with a 4.90-m diameter.

ULSD test runs

After the revamp, Bayernoil successfully started up the unit in mid-May 2002. The refiner conducted test runs to evaluate reactor performance. Table 1 shows the results of some of the test runs.

The unit successfully achieved the design targets of 53,000 b/sd and <10 ppm (wt) product sulfur. Bayernoil can produce ULSD from a wide range of feeds, with sulfur ranging 0.3-0.64 wt % and cracked stock contents of 5-20%.

From mid-May to early October 2002, the unit produced 50-ppm (wt) sulfur diesel for local markets. Since early October 2002, however, the unit has produced ULSD in preparation for its official introduction in Germany on Jan. 1, 2003.

Monitoring data show that the current catalyst fill will achieve its expected 2.5-3 year cycle length. The catalyst fill contains 64% fresh Akzo Nobel NV high-activity KF757 catalyst, loaded in the last two beds of the new three-bed reactor. The remaining 36% is regenerated catalyst loaded in the old two-bed reactor and the top bed of the new reactor.

The new reactor was intentionally oversized to allow Bayernoil to potentially process higher-sulfur feeds in the future. Adding catalyst volume to a new reactor is relatively inexpensive.

Using some regenerated catalyst for this load also reduced the capital investment for this revamp project. In the future, the refiner can optimize the catalyst system to achieve longer cycles.

ULSD production reliability

A major concern of refiners producing ULSD is for the unit to operate reliably so that it produces on-spec diesel all the time. Many factors can cause the unit to produce off-spec product, such as a spike of higher-than-expected feed sulfur, cracked stock in the feed, lower-than-required reactor temperature, or other unit upsets.

Several approaches help minimize the occurrence of off-spec diesel. Refiners can:

  • Use on-line analyzers to continuously monitor product qualities, such as sulfur, density, or cetane, in real time.
  • Send the product stream to an intermediate product tank and frequently monitor its quality before sending it to the blending facility. The refiner can sell any off-spec diesel as a lower-grade product such as heating oil or route it to the feed tank for reprocessing.
  • Use a feed holding tank to reduce and smooth out fluctuations in feed quality. The holding tank should be blanketed with nitrogen or fuel gas; oxygen may induce fouling in heat exchangers and reactors.
  • Develop an accurate process model or correlations that operators can use to quickly correct any unexpected incidents. This will help avoid or at least minimize off-spec materials.

Refiners have used online sulfur analyzers to monitor product sulfur levels since the introduction of 500-ppm diesel in 1993. Their use increased significantly in 2000 when 50-ppm diesel was marketed in some European countries.

In the ULSD era, online sulfur analyzers are a necessity for ULSD production units or blending facilities. Online sulfur analyzers, however, must be calibrated and maintained regularly to ensure accuracy.

Bayernoil installed Antek Instrument LP C6000 Series online sulfur analyzers on the product stripper rundown line and at the product holding tank to monitor product sulfur levels in real time.

Additionally, Bayernoil takes periodic diesel samples to analyze in the laboratory using the officially approved test method, ASTM D5453, for official certification. Online sulfur analyzer results are well within the acceptable range of ±0.5 ppm.

A recent revamp of this hydrotreater in Bayernoil's Neustadt refinery allows it to produce a <10 ppm sulfur diesel. Photo courtesy of Bayernoil Raffineriegessellshcaft MBH.
Click here to enlarge image

Two test runs show the importance of online sulfur analysis. In the first test run, Bayernoil raised the reactor temperature of a diesel hydrotreater and kept feed quality constant. Sulfur in the product rundown from the product stripper, measured in real time with an online sulfur analyzer, indicates reactor performance.

Click here to enlarge image

Fig. 5 shows that the product sulfur level responds kinetically to the change in weighted average bed temperature (WABT) with a few hours of lag time—2-3 hr depending on charge rate and holdup time in the reactor, product stripper, and lines.

In the second test run, Bayernoil changed feed composition by varying the amount of FCC light cycle oil and held reactor temperature constant. Because light cycle oil usually contains more sulfur in the form of refractory sulfur compounds, it is more difficult to desulfurize than virgin light gas oil.

Click here to enlarge image

Fig. 6 shows that the product sulfur level changes in direct response to the amount of light cycle oil in the feed.

References

  1. Lee, C.K., and McGovern, S.J., "Clean diesel: overview and comparison of clean diesel production technologies," Petroleum Technology Quarterly, Winter 2001/2002.
  2. Derr, W.R., McGovern, S.J., and Lee, C.K., "Role of trickle-bed reactor design in meeting future clean-fuels regulations," World Refining, October 2002.
  3. Yeary, D.L., Wrisberg, J., and Moyse, B.M., "Revamp for Low Sulfur Diesel: A Case Study," presented at the 1997 NPRA Annual Meeting, Mar. 16-18, 1997, San Antonio, paper AM-97-14.
  4. Ouwerkerk, C.E., Brandland, E.S., Hagan, A.P., Kirkert, J.P., and Zonnevylle, M.C., "Performance Optimization of Fixed Bed Processes," presented at European Refining Technology Conference, Nov. 16-18, 1998, Berlin, Germany.
  5. Tippett, T., Knudsen, K.G., and Cooper, B.H., "Ultra Low Sulfur Diesel: Catalyst and Process Options," presented at the 1999 NPRA Annual Meeting, Mar. 21-23, 1999, San Antonio, paper AM-99-06.
  6. Desai, P.H., Gerritsen, L.A.,and Inoue, Y., "Low Cost Production of Clean Fuels with STARS Catalyst Technology," presented at the 1999 NPRA Annual Meeting, Mar, 21-23, 1999, San Antonio, paper AM-99-40.
  7. "Centinel Hydroprocessing Catalysts: A New Generation of Catalysts for High-Quality Fuels," Criterion Catalyst Co., Houston, October 2000.

Based on a presentation to the AIChE Spring National Meeting, Mar. 30-Apr. 3, 2003, New Orleans.

The authors
Klaus-Dieter Rost ([email protected]) is a senior process engineer responsible for project development at Bayernoil Raffineriegessellshcaft MBH, Ingolstadt, Germany, a position he has held since 1999. Previous to that, he served in a similar position at Lurgi Oel Gas Chemie GMBH for 7 years. Rost holds a MSc from University of Erlangen, Germany.

Rupert Herold (Rupert.Herold @Bayernoil.de) is a process engineer at Bayernoil Raffineriegessellshcaft MBH, Ingolstadt, Germany, a position he has held since 1999. He is responsible for the operations of hydrotreaters, crude and vacuum units, catalytic reformers, and several other units. Herold holds a MSc in chemical engineering from University of Erlanger, Germany.

C.K. Lee (CKLee123@aol. com) is a principal of Petro- Tech Consultants, Voorhees, NJ. His areas of expertise are hydroprocessing, catalytic reforming, and clean fuel technologies. Previously, he worked for Mobil Technology Co. for more than 20 years. Lee holds a BS in chemical engineering from Cheng Kung University, Taiwan, and a PhD in chemical engineering from the University of Houston.

Tai-Sheng Chou (TaishengC @aol.com) is a hydroprocessing specialist for H2Advance, Cherry Hill, NJ. He has 26 years' experience in technical service, process design, and research of hydrotreaters, hydrocrackers, and hydrogen plants. Chou previously worked for Mobil Technology Co. and General Motors Corp. He has a BS (1970) in chemical engineering from National Cheng Kung University, Taiwan, and a PhD (1976) in chemical engineering from the University of Houston.