Apr-2003

ULSD production: improved feed stream quality

Much of hydrotreater revamp investment costs in future will be in dealing with the most difficult-to-remove sulphur species, because of their low reaction rates

Daryl Hanson and Steve White, Process Consulting Services

Article Summary

Meeting future specifications for ultra-low sulphur diesel (ULSD) of 10–15ppmw sulphur will require significant hydrotreating investment. Much of this investment is directly related to the hard-to-remove sulphur compounds found in the true boiling point (TBP) range of hydrocarbons above 650°F. For many refiners, optimising upstream fractionation can materially improve recovery of easy-to-treat hydrocarbons of below 650°F currently feeding the FCC or hydrocracker, thus significantly reducing overall investment costs. Conversely, ULSD product yield can be materially increased for a given level of capital investment.

For many refiners, deficient atmospheric crude, crude vacuum, FCC, and delayed coker unit main fractionator performance will dramatically increase ULSD hydrotreater revamp capital cost, reduce hydrotreater throughput, and increase operating costs. Refiners processing very high sulphur heavy crude oils have even more incentive to improve crude unit performance because recovery of <650°F hydrocarbons is generally low and the straight-run (SR) diesel contains higher concentrations of the hard-to-remove sulphur compounds.

Furthermore, the SR diesel products yielded from heavy crudes have longer distillation tails because fractionation is inherently more difficult. As ULSD hydrotreater feed distillation 95 vol%-endpoint (EP) temperatures increase, the amount of hard-to-remove sulphur compounds increases. These compounds largely control ULSD product sulphur content, thus, when cost-effective, they should be fractionated from the hydrotreater feeds.

Most refiners will need to invest significant capital to revamp their existing diesel hydrotreaters now producing 350–500ppmw sulphur product. Supplying road diesel at the pump having 15ppmw sulphur will require ULSD hydrotreater product sulphur of 8–10ppmw or less due to transportation system contamination and sulphur analytical repeatability. ULSD hydrotreaters will need higher activity catalyst, increased hydrogen partial pressure, lower reactor liquid hour space velocity, improved reactor feed distributors and other design changes [Hu M C , et  al, Rigorous hydrotreater simulation, Petroleum Technology Quarterly, Spring 2002]

 Many options for revamping current diesel hydrotreaters for ULSD production are being considered. Much of the focus is on new catalyst technologies. Yet there are many unanswered questions including the consequence of normal upstream unit product TBP distillation variability on ULSD product sulphur. For example, the penalty currently for producing off-specification 550ppmw sulphur diesel for a short period is low because hydrotreater reactor temperature can be increased to make a lower sulphur blend stream.

In the future, blending off-specification material will be extremely costly when the target product sulphur content is 8–10ppmw. Meeting ULSD product sulphur content will require all feed stream distillations to be tightly controlled so that the amount of difficult-to-remove sulphur entering the hydrotreater does not exceed ULSD product limits.

FCC and delayed coker middle distillate streams contain higher percentages of the hard-to-remove sulphur compounds while crude unit straight-run diesel streams have lower concentrations that depend on crude source and distillation endpoint. The majority of the hard-to-remove mono-beta and di-beta alkyl substituted dibenzothiophene compounds are contained in the 650–700°F boiling range.

Improved feed stream fractionation may offer significant opportunity to increase recovery of hydrocarbons containing the easy-to-treat sulphur compounds while dramatically reducing the long distillation tails containing the most refractory sulphur compounds. Production of ULSD diesel should include optimising upstream fractionation to ensure that improved feed quality is part of the overall investment strategy. 

Undercutting
Undercutting feeds to the ULSD hydrotreater has been discussed in numerous articles. Undercutting implies reducing the feed stream yields to eliminate the 650–700°F hydrocarbons containing the hard-to-remove sulphur compounds. While this is the lowest capital cost option, undercutting also decreases diesel recovery. In one instance, undercutting to eliminate the hard-to-remove sulphur compounds reduced the delayed coker light coker gasoil (LCGO) yield by 35%. Furthermore, undercutting will increase cat feed hydrotreater (CFHT) and/or FCC unit feed rate and have a significant impact on these units feed rate, operating costs and product yield structure.

An alternative to simply undercutting is to consider low-capital fractionation improvements that increase recovery of the easy-to-treat hydrocarbons from the FCC or hydrocracker feeds while eliminating the hard-to-remove sulphur compounds from the ULSD hydrotreater feeds.  

Sulphur species
The Figure 1 block diagram (next page) shows the three primary feed streams to the ULSD hydrotreater. Atmospheric and vacuum crude columns, and FCC and delayed coker main fractionators produce SR diesel, light cycle oil (LCO), and LCGO streams, respectively. Different sulphur species are distributed throughout these streams. Table 1 lists the major sulphur compounds and their approximate hydrocarbon boiling range.

For the purpose of this discussion, easy-to-treat sulphur compounds are contained in the <650°F hydrocarbons, whereas the 650-700°F boiling range have the majority of the hard-to-remove sulphur species. In reality, the transition between easy-to-treat and hard-to-remove sulphur compounds is not a sharp line. The mono-beta alkyl substituted dibenzothiophenes begin to distil in 620-650°F hydrocarbons. Nevertheless the presence of stericically-hindered sulphur species changes significantly with the >650°F TBP temperature hydrocarbons.

Easy-to-treat recovery
Refiners need to identify the easy-to-treat hydrocarbon losses from each unit producing feed for the ULSD hydrotreater. Incremental <650°F hydrocarbons will come from crude unit FCC feed, delayed coker heavy coker gasoil (HCGO), and FCC HCO/slurry product streams. Improving recovery will depend on existing crude, FCC, and delayed coker primary fractionator performance.