Apr-1999

Performance optimisation of fixed bed processes

An extensive range of hydroprocessing technologies has been developed through a combination of long-term R&D commitment and commercial experience

C E D Ouwerkerk, E S Bratland, A P Hagan, B L J P Kikkert and M C Zonnevylle
Shell Global Solutions

Article Summary

Many oil refining processes are carried out under high pressure trickle flow conditions – hydrodesulphurisation, hydrogenation, hydrocracking, residue hydroconversion [A K Saroha and D K P Nigam,“Trickle bed reactors”, Rev Chem Eng, 1996]. Due to tightening environmental legislation on product quality specifications, performance of these units is gaining importance in the current refinery configurations. For example, the introduction of low-sulphur diesel in Europe in 1996 has led to a substantial increase of HDS capacity through new units and revamps.

Because of further tightening of specifications for automotive fuels (sulphur, aromatics, etc) in 2000+, the combination of catalyst choice and reactor design is becoming ever more critical for all of the above mentioned hydroconversion and hydrotreating processes. More stringent product specifications also decrease blending flexibility and put higher demands on overall onstream time, reliability and conversion, which must be achieved at lowest possible operating and investment costs.

The combined improvement of (i) more efficient catalyst use by optimum gas/liquid distribution and quench performance, (ii) run length extension by an increase in fouling resistance and (iii) online performance monitoring for superior operational control at increasing reactor sizes is therefore economically very important. This article deals with a number of Shell-developed reactor technologies to achieve these improvements.

Reactor internals
Maximising reactor volume utilisation with superior performance
Hydroprocessing reactors are to be designed with internals, which provide optimal conditions for catalyst performance in terms of liquid/gas distribution, inter- and intra-phase mixing, quenching performance and fouling resistance, while simultaneously being space-efficient and easy to maintain. In trickle flow units proprietary internals for homogeneous distribution of gas and liquid over the reactor cross-sectional area are installed [J Vos, “Trouble shooting in hydroprocessing operations”, Akzo-Nobel Catalyst Courier, 1996; F Emmett Bingham et al, “Performance focused reactor design to maximise benefits of high activity hydrotreating catalyst”, 1997 ERTC proceedings].

The equal distribution of liquid over the catalyst is essential to achieve maximum utilisation of the inventory [D L Yeary and J Wrisberg, “Revamp your internals”, Hydrocarbon Eng, 1997] and to avoid the potential risks formed by stagnant zones and channelling [C R Kennedy and S B Jaffe, “Analysis of tracer experiments from commercial scale trickle bed reactors”, Chem Eng Sci, 1986].

Gas/liquid distributors
In the past, liquid distribution was effected via a small number of separate liquid and vapour down-comers followed by a layer of inert distributive packing. The liquid down-comers can take a number of shapes, from plain holes in the tray to special tubes equipped with triangular or rectangular notches. The number, size and detailed design of the liquid downcomers are usually selected so that a liquid level of  about 0.05–0.10m of liquid is maintained on the tray at the lowest liquid flow.

Liquid distributor trays must be installed as level as possible since tilt will deteriorate the uniformity of liquid distribution to the bed, the target maximum deviation from the horizontal plane being of the order of 1/1000 of the reactor diameter. Figure 1a shows the measured liquid uniformity of such liquid down-comers. The disadvantage of such internals is that some catalyst inventory is lost due to the need for distributive packing above the catalyst bed, resulting in a reduced cycle length.

For some 10 years, High Dispersion (HD) nozzles have been used as standard design for Shell trickle flow hydroprocessing units, because they provide a very even liquid distribution. They have been tested and developed in R&D and proven by years of operational experience in numerous hydrotreating, hydrodesulphurisation, hydrocracking and residue hydroprocessing units worldwide.

Contrary to conventional trays, vapour and liquid pass through the nozzle together. As a consequence of the acceleration of liquid by the gas in the tube, an intimate mixing of liquid and vapour is obtained as well as an almost 100 per cent distribution uniformity (Figure 1b) over the reactor cross section. The same phenomenon also improves the tolerance of the tray towards fouling, as blockage of a nozzle is compensated by surrounding nozzles and hardly affects overall distribution uniformity.

Therefore, with the installation of an HD tray, distributive packing on top of the catalyst bed is no longer required. Also, a significant decrease in the catalyst deactivation rate is achieved. In comparison to conventional distribution trays, the use of HD nozzles will extend cycle length by reducing weight average catalyst bed temperature (WABT) and/or deactivation rate and will improve product quality by reducing the risk of dry spots in the catalyst bed.

The design is robust to installation and maintenance and has less sensitivity to tilt due to the acceleration of liquid by the gas. For heavy fouling conditions, such as in residue conversion units, special dirty duty nozzles can still provide excellent liquid distribution even if partially fouled.

Case study 1
After retrofitting of an existing hydrocracker with HD trays a decrease in WABT of 5–10°C was observed, corresponding to a 25 per cent increase in bed temperature span and a potentially corresponding increase in run length.

Case study 2
In a recent retrofit of  another hydrocracker the installation of HD trays enabled the combination of the two top beds in the reactor. Together with dense bed loading this resulted in some 20 per cent more catalyst loaded and a 30 per cent increase in run length.