Oct-2004

ULSD problems and solutions

A review of the common pitfalls in scoping, designing, implementing and operating ULSD units. The authors offer various solutions, ranging from optimising hydrogen partial pressures to avoidance of recombination reactions

Scott Sayles, Jim Bailor and Robert Ohmes, KBC Advanced Technologies

Article Summary

Production of ultra low sulphur diesel (ULSD) is scheduled to begin in June 2006 in the USA. Refiners have many plans to produce ULSD, which vary in complexity from plant to plant. Using their tailored methods, each refinery will provide a portion of the total ULSD demand within the USA. However, some refiners are still concerned with their ability to deliver the product. These concerns range from the method of production to the pipeline transportation system. Other refiners are struggling with routinely optimising these new assets once the project is complete and the first shipment of ULSD is produced. Other countries have initiated ULSD production and have experienced startup, operational and production problems.

Kinetic models such as KBC’s HDT-SIM can provide accurate ULSD predictions, which are useful for ULSD design, operation and monitoring. In addition, the KBC HTR-SIM model predictions are suited for developing accurate ULSD LP vectors, which can help the refinery anticipate ULSD production variations due to changes in feedstock rates and qualities for improved ULSD operational control.

Clean fuels production is a worldwide initiative with some regions ahead of the USA in implementation, and others behind. The worldwide refining community recognises the key role clean fuels plays in the improvement of the environment. The refining industry has successfully faced many such challenges by delivering continually cleaner, low cost transportation fuels to the consuming public.

The “clean fuels” title is used to cover a wide range of fuel characterisations for different initiatives: reduce sulphur levels, add oxygenates, reduce aromatics, increase octane or cetane and meet additive package requirements. Selected from this wide-ranging criterion are the specific issues and factors surrounding the refinery requirements to produce 100% highway ULSD according to USA regulations. Several options exist for the refiner to produce ULSD. Selecting the best option involves carefully weighing many important issues:
Can an existing hydrotreater be revamped to make ULSD or should a grassroots unit be built?
- Which option gives the best flexibility for meeting current and future ULSD quality specifications other than sulphur?
- Which option gives the most flexibility for upgrading heavier and/or cracked feedstocks to ULSD?
- How to ensure the ULSD product still meets specifications when it is delivered to the customer?

Many refiners have existing diesel hydrotreating units. However, most of these units require major modifications to make ULSD and some of these units are totally not suitable for upgrading diesel to ULSD. The difficulties in producing 15wppm sulphur have been well documented as a function of conversion of hindered sulphur species in various feedstocks. An additional concern is the ability to commercially measure sulphur levels of 15wppm or less.

Desulphurisation mechanism
The hydrodesulphurisation (HDS) mechanisms most widely accepted are shown in Figure 1.

The first route is direct hydrogenolysis, which converts almost all of the mercaptans, sulphides, disulphides and thiophenes. Most of the unsubstituted dibenzothiophenes and benzothiophenes are also converted. The second route requires partial hydrogenation of the aromatic ring in the dibenzothiophene structure prior to sulphur removal. This step seems to be more effective when a Ni/Mo catalyst is used, although some of the second-generation Co/Mo catalysts may have overcome this requirement.

The results of a single stage pilot unit study, compared to predictions by the HTR-SIM hydrotreating kinetic model, provide an example of how total product sulphur varies with reactor temperature. In Figure 2, as reactor severity is increased, the product sulphur level is reduced until a floor or plateau is reached at about 5 to 7wppm. Reducing sulphur below this point typically takes a noble metal catalyst and a second reaction stage. Producing less than 10wppm sulphur diesel has been demonstrated commercially at the pressure levels shown below:
Unit pressure level         Number of units
<500psig                                      2
600 to 700psig                             3
>750psig                                    12
                                       Total    17

Catalyst deactivation is a concern at the higher severities required for ULSD production. The reported commercial catalyst deactivation rates are between 1.8 and 3.3ºF per month for the higher-pressure units [Patel et al, How are refiners meeting the ultra low sulfer diesel challenge?; NPRA annual meeting, March 2003]. These deactivation rates will produce acceptable run lengths.

The ULSD sulphur test method cited in the regulations is ASTM 6248. A recent round robin test using this method gave the precision of the method in terms of the ASTM reproducibility at the 15wppm sulphur level to be 13.15 ±5.56wppm [Brtko et al, Sulfur content in ultra low  sulfur diesel fuels: regulatory and measurement issues; NPRA annual meeting, March 2003]. This ASTM test method is under continuous development and is believed to be able to meet the testing needs in the future.

However, for the refiner to have a 95% confidence level that the sulphur level in the product leaving the refinery will be less than 15wppm, the sulphur level tested in the field using this ASTM method must be 10wppm or less. Many refiners are targeting 10wppm or less to provide testing, blending and handling flexibility. As will be discussed in further detail, this leaves little room for cross-contamination.

Product quality
The refiner’s goal is to produce ULSD that not only meets the regulatory requirements of 15wppm sulphur, but also the additional commercial requirements of a high quality product.