Jul-2000

Increasing hydrocracker capacity at low cost

Detailed case studies are presented, illustrating the planning and working methods that were followed in revamping the hydrocrackers at two refineries, in Germany and Canada

Gurunath G Bharne and Christine A Groenendaal
Shell Global Solutions International BV

Article Summary

In a climate of depressed refining margins and higher investment thresholds, carrying out debottlenecking or revamps becomes difficult to justify, and refineries can only hope to get project approvals by limiting the costs. The hydrocracker is the central upgrading unit and is, justifiably, often the candidate for revamping.

When conceiving, developing and executing revamps, an integrated all-round technical capability is a big help. Detailed knowledge of catalyst technology, process engineering and operational feedback will be essential in producing safe, sound and cost effective solutions. Minimising modifications not only saves direct costs but also helps speed implementation.

R&D, design, project handling and operating experience has proven instrumental in finding the right answers in several revamp projects undertaken by Shell. Two recent examples of the company’s hydrocracker revamps are given in detail here.

The first case relates to Shell’s Godorf refinery in Germany, where the revamped unit came on stream in mid-1999. The second case addresses a detailed debottlenecking design completed for the Scotford refinery in Alberta, Canada, as a key link in Shell’s Athabasca oil sands project.

Case study 1
Godorf hydrocracker revamp
The Godorf hydrocracker, near Cologne, Germany, is a partial conversion hydrocracker originally commissioned in early 1983. The original design capacity was 3000t/d of vacuum flashed distillates (vacuum gasoils) and the cycle length 18 months with fluoriding of catalyst for an online activity boost.

Employing Shell hydrocracking technology, the unit was designed as a single stage hydrocracker with two reactors in series and a dedicated fractionation section. The main product is the hydrowax exported as feedstock to an ethylene cracker in close proximity.

The capacity evolution
Before the revamp implemented in mid-1999, the unit operated at a maximum sustainable capacity of about 5600t/d, a plant conversion of about 60 per cent and a cycle length of 30 months. The capacity evolution is summarised in Figure 1.

The capacity increase can be linked to two specific thrusts with capacity creep in between. The first thrust was in the form of a revamp project implemented in 1989 to achieve a new increased design capacity of 4500t/d.

The associated major changes included modifications to the hydrocracker feed preparation (high vacuum distillation) units, the addition of a second parallel string of two reactors, modifications to the heat exchangers and furnace in the reaction section and replacement of the main fractionator in the fractionation section.

The second impetus came from a switchover from the single catalyst configuration (Criterion C-424 catalyst) to a stacked bed configuration (Criterion C-424 and Zeolyst Z-603 catalysts in series) with little else modified. C-424 filled the first reactors but also the top bed of the second reactors. The objective was to simultaneously achieve an increased throughput (5500t/d), a longer cycle (30 months), improved naphtha/kero yield (1.6 per cent wof increase) and improved hydrowax quality (2.5 MBCI points reduction – MBCI is symbolic of ethylene yield potential and is defined in terms similar to BMCI, the US Bureau of Mines Correlation Index). Both projects were successful and delivered the design premises.

Higher capacity without excessive capital requirements
With the hydrocracker capacity saturating at around 5600t/d on an annual basis, simultaneously limited by several constraints, the next challenge was flexible and, paradoxically, tough. How much more capacity could the unit achieve without major capital injection? And exactly what would that entail?

It would be important to avoid the need for new expensive equipment items such as high pressure equipment in the reaction section, compressors, furnaces and large columns. The project had to be implemented on a fast track to coincide with a scheduled inspection and catalyst changeout stop in mid 1999.

A feasibility study, based on a 1997 test run, identified the then existing limitations. The most fundamental of these, the light naphtha make limitation imposed by flooding in the top section of the main fractionator, had to be removed before further progress could be made. The newly developed Shell Con-Sep trays (discussed in more detail later), if used to replace the existing internals in the flooded section, would provide an elegant answer, boosting the potential light naphtha capacity by up to 50 per cent.

On the reaction section side, the new catalyst combination, designed to operate without the fluoride-based activity booster and targeting a higher naphtha yield, would now need to aim at a throughput of 6400t/d to “fill up” the revamped fractionator. The feasibility study further indicated that this increase from 5600t/d to 6400t/d  should be possible with limited modifications.

This led to the formulation of clear design targets for the reaction section and the fractionation section. These targets and the solutions to achieve them are discussed below in more detail.

The reaction section
The target was to aim for an increase in capacity from 5600t/d to 6400t/d while maintaining the cycle length at 30 months and the plant conversion at 60 per cent. Contributing to the additional feed supply would be own thermally cracked vacuum flashed distillate, making the feed somewhat more refractory (higher nitrogen-containing). In addition, Godorf had a clear wish to abandon the use of online fluoriding. The hydrowax MBCI would need to improve and product selectivity to naphtha would need to be better.