Spappu, the quotation on the experience of industry on ultra low-level desulfurization is not entirely true since catalytic reformer feeds are being desulfurized to levels lower than 0.5 ppm by hydrotreating, and there are also proprietary absorption techniques that claim reducing sulfur levels further down to 0.1 ppm.
The stoichiometry mentioned by epoisses could be complemented with the need to saturate with hydrogen the bonds left opened by removing sulfur, as in
R1SR2 + H2 [→] R1H + R2H + H2S
With higher-boiling more refractory thiophenes and benzothiophenes, there is also opening and saturating of the thiophene ring.
These reactions need more moles of hydrogen than would be required by making one mole of H
2S.
As a result larger amounts of lower boiling hydrocarbons are produced which require stripping of the product to preserve flash point and boiling range.
Side reactions such as hydrocracking, dehydrogenation, and dearomatization, as well as those leading to carbon laydown, are minimized by chosing the right catalyst (and support) and operating conditions.
The selection of the latter (in particular space velocities which are about proportional to residual sulfur content) may affect also the removal of nitrogen (as ammonia) and oxygen (as water), and in the heavier fractions, such as distillation residues, the removal of metal compounds.
