Apr-2001

Steady state simulator to monitor treater performance

A steady state hydrotreater model was used to monitor the effects of feed quality so that changes in key process factors could be matched against catalyst age

Michael C Hu and Robert T Powell, KBC Advanced Technologies
Naoki Yomoji and Danichiro Ohshima, Kashima Oil Company

Article Summary

New low sulphur specifications (<50ppm) for gasoline and on–road diesel fuels will require refiners to increase the degree of desulphurisation in their production. This task can be accomplished by either adding new hydrotreating capacity (such as a new unit), debottlenecking existing units, employing more active hydrotreating catalysts or simply raising operating severity (eg reactor temperature).

With increased emphasis on sulphur removal, it becomes extremely critical for the process engineer to know the performance status of the hydrotreater, particularly from a catalyst stability or deactivation point of view, so that an optimum catalyst change-out schedule can be established. Unfortunately, due to variations in feed quality and fluctuations in process conditions, good catalyst stability information is not always apparent just by plotting raw process data against catalyst age.

This article presents a systematic procedure which uses a steady state hydrotreater model that accounts for feed quality effects and process condition effects to normalise raw data to a set of standard conditions so that the shift of key process factors (reactor temperature, hydrogen consumption, yield and quality) versus catalyst age, can be clearly defined.

Low and ultra low sulphur specifications require the refiner to boost overall hydrotreating capacity or, more specifically, the degree of hydrodesulphurisation, either in the area of feed treatment for cracking conversion units or hydrofinishing units for final fuel products. To meet the challenge, many technical alternatives have been developed and proposed by internal technology groups, process licensors or catalyst suppliers.

Possible solutions include a wide range of new and improved catalytic hydroprocessing technologies, such as selective hydrodesulphurisation of FCC gasoline, deep and ultra-deep diesel hydrotreating, FCC feed pretreatment and vacuum residue hydroconversion.

On the catalyst development front, the catalyst suppliers have constantly commercialised new catalysts aiming at higher activity and better stability. It has been noted that, depending on the process objective, mixed loadings of two or more catalysts in the reactor may be more effective than a single type of catalyst. The complexity of hydroprocessing technologies, catalyst selections and ever-changing fuel specification requires refineries to work closely with process licensors and catalyst vendors to select the most economical solution with maximum flexibility.

The selection of the right process and catalyst package, however, is only the first part of the task. The hydrotreater process engineer needs to closely monitor the reactor’s hydrotreating performance to ensure that the product quality meets specifications while not subjecting the catalyst to unnecessary deactivation, exceeding process constraints nor violating safe operation.

In trying to schedule unit shutdowns and comparing catalyst from different runs, process engineers will be asked the following process-related questions by operation and maintenance managers:
What is the current hydrotreater catalyst deactivation rate?
How long will the catalyst last before reaching end of run (EOR) temperatures?
What is the impact of processing a new batch of feed on the unit performance?
How is the catalyst performing compared to last week – last month – last run?

To answer these questions, engineers must have access to the historical process data. Usually, plots of reactor average temperatures, makeup hydrogen requirement, product yields and quality etc, versus catalyst age, are prepared to facilitate this process analysis. Unfortunately, due to feed quality variations and fluctuations in other process conditions, plotting of raw data may not yield a clear trend. A data normalisation exercise must be carried out to cast the raw test run results in terms of a set of standard feedstock and operating conditions.

Calculation procedures
The primary task of unit monitoring is to generate reactor performance data versus unit run time. This information is crucial for the planning of catalyst change-out, evaluating competing catalysts and conducting future project studies. To demonstrate the calculation procedures required for unit monitoring, a typical hydrotreater is used in this article as an example where sulphur conversion usually is the target and the reactor temperature is adjusted upward to compensate for the loss of catalytical activity due to deactivation.

Typically, a plot (Figure 1, on previous page) of reactor weighted average bed temperature (WABT) versus time-on-stream should be generated from the operating data. The catalyst age is defined as cumulative barrels of feed processed per pound of catalyst (bbl/lb). The WABT term is most useful if it is normalised to be free from feed quality effects and process condition effects. Under an ideal constant severity operation, a steady linear deactivation rate in terms of °C or °F/bbl/lb of catalyst would be observed.

A higher severity operation, such as lower product sulphur spec, poorer feedstock quality or higher throughput inevitably will result in a higher WABT requirement, a higher deactivation rate and shorter cycle life.

The calculation procedure required to calculate normalised WABT basically answers the question: what would the WABT be if the hydrotreater had been operated under a standard and ideal set of conditions based on the actual reactor temperature readings?

For example, the test run data may show that under the measured feed rate, reactor WABT and feed sulphur content, a certain product sulphur level is achieved. The first step is to calculate the kinetic rate constant of hydrodesulphurisation (HDS) using the following equations: