Mar-2012

Decrease catalyst costs by regeneration, analysis and sorting

Use of regenerated catalyst can be maximised by careful sampling and analysis of spent catalyst to construct a reactor contamination profile

PIERRE DUFRESNE, Eurecat
FRANCOIS LOCATELLI, Eurecat France

Article Summary

Oil refining and petrochemical plants are currently undergoing serious challenges from a technical and economical point of view, as well as from an environmental point of view. The volume of installed capacity of hydroprocessing catalyst is increasing in order to cope with the development of nearly sulphur-free clean fuels and the need for deeper heavy ends conversion. At the same time, low refining margins are driving industry to cost optimisation, while environmental regulations are getting tougher.

In addition to the growing fresh hydroprocessing catalyst market, regenerated catalyst usage is increasing, with an estimated worldwide consumption of roughly 40 000 t/y. This additional tonnage of regenerated catalysts is extremely useful for refineries hoping to reduce their procurement expenses as well as limit their generation of chemical waste. Unfortunately, the ratio between the regenerated and the fresh catalyst market in distillates hydroprocessing is below 50%, meaning that not all spent catalysts are regenerated at least once. This low reuse rate could certainly be increased by careful evaluation of the spent catalyst.

Regeneration of hydroprocessing catalysts restores activity
The end-of-cycle for hydroprocessing units is usually determined by: a scheduled unit shutdown; a unit upset, such as too high a pressure drop or compressor failure; or a catalyst activity that is too low for meeting product specifications. In this last case, activity decrease is mainly due to partial blockage of the active sites by coke and sometimes contaminants. Carbon content on spent HDS catalysts largely varies from 5 to 25 wt%, with an average for diesel units of 8-15 wt%. It is common knowledge now that the regeneration of spent hydroprocessing catalysts does restore activity. By using an oxidising atmosphere at a temperature of 450-550°C, carbonaceous species can be removed from the catalyst. Regeneration has the benefit of eliminating the first cause of deactivation: coke deposit. But it does more by converting the sulphide phase, which may have become partially sintered, back to an oxide phase similar to the fresh catalyst. This works well for conventional catalysts, while the latest generation of catalysts may need an additional treatment to achieve better recovery of activity.

How are hydroprocessing catalysts contaminated?
Various contaminants are found on spent hydroprocessing catalysts at the end of cycle. They originate either from the feed (vanadium, nickel, arsenic and sodium), from additives used during transport or refining operations (silicon, lead and phosphorous) or from corrosion (iron).

Ni + V
The nickel- and vanadium-containing molecules present in the heavier fraction of vacuum gas oil and resid feeds are readily decomposed to nickel and vanadium sulphides on the catalyst surface. After regeneration, the sulphides are converted back to oxides, also promoting some alumina sulphatation. Depending on the hydrotreating application, it is usually not advised to reuse catalysts containing more than 1 to 3 wt% of vanadium (often expressed as Ni+V, as these two metals generally exist together, even if nickel is not itself considered a poison).

As
Organoarsenic compounds contained in some crudes are extremely reactive under hydroprocessing conditions and are thus readily decomposed and adsorbed on the catalyst surface. Therefore, a very steep arsenic gradient is always observed between the top and the bottom of the reactor. Arsenic is considered a severe poison for HDS activity, as a quantity of 0.2 to 0.4 wt% can prevent catalyst reuse.

Na
Sodium is usually not present in the various hydroprocessing feeds in normal operation. However, it could be present on the catalyst due to desalter malfunction, because of various upsets leading to upstream introduction of caustic soda (for instance, upsets of Merox units), or when seawater is introduced to the unit via heat exchanger leakage. Sodium then adsorbs on the catalyst surface and has a significant deactivation effect, which is usually more noticeable after regeneration. The maximum tolerated poison level for sodium is typically between 0.1 and 0.2 wt%.
.
Si
Silicon originates essentially from lighter fractions of coker or visbreaker operations, where polydiméthylsiloxanes are used as anti-foaming agents. Silicon could also originate from silicon-based chemicals injected into pipelines for a reduction in pressure drop. Silicon is not a very strong poison during the cycle, and deactivation seems higher after regeneration.

Pb, P
Lead is no longer seen in spent catalysts in western countries, due to the general phase-down of tetraethyl lead as an octane booster. Poisoning by phosphorous may be seen sometimes, but not often, as phosphorous-containing additives are not so common yet.

Fe
In distillate hydrotreaters, iron contamination comes from the corrosion of upstream equipment. As such, iron scale or fine particles usually do not penetrate deeply into the catalyst pores and do not have any strong poisoning effect. Its contribution to pressure drop build-up is more problematic, as iron particles may accumulate at the top of the bed or in the interstices between catalyst granules.


How to maximise the quantity of reusable regenerated catalyst. Sorting is a solution.
Regeneration, however, will not cure anything as regards deactivation by contaminants, which remain adsorbed after thermal treatment. This does not mean that a contaminated batch cannot be regenerated. What is necessary is to profit from the concentration gradient that is always observed in fixed-bed reactors in case of contamination.

SAS (sampling, analysis, segregation) service can save a great deal of good catalyst
When catalyst is contaminated, the analysis of a global representative sample does not tell us anything regarding the concentration gradients throughout the reactor bed and is thus not sufficient.