Jul-2014

Upgrading hydroprocessing

Refinery profitability may be improved by operating with a cheaper crude oil diet though, in practice, this is harder than it sounds. Cheaper crudes are cheaper for a reason: they are generally heavier, sourer and more contaminated than refiners would normally consider ideal. Consequently, converting them into premium grade products at high yields and without undermining plant reliability presents a considerable challenge.

Kevin Carlson
Criterion Catalysts & Technologies

Article Summary

It is clearly possible, however. One North American refiner, for instance, has successfully processed large volumes of difficult, discounted crude oil to return an impressive product slate of clean fuels that meet the highest international specifications. This refiner’s success is partly due to the close attention it pays to its hydroprocessing assets, the way that it operates them and its ability to respond positively to challenges as they arise. An example of this is the way that the refiner, with help from Criterion, tackled the acute drop in performance of one of its two kerosene hydrotreaters, ultimately turning the problem around and generating business benefits in terms of increased cycle length and enhanced operational flexibility.

The site hydrotreats straight-run kerosene at relatively low pressure (<500 psi) in two units. The first of these tends to receive higher grade kerosene to make jet fuel, with its tight colour and colour retention specifications. It is mainly hydrodenitrification (HDN) and saturation taking place in this reactor. The second unit is assigned the tougher kerosene and it generally makes an ultra low sulphur diesel (ULSD) specification product for blending purposes.

Hydrodesulphurisation (HDS) is the prime objective here. Colour is not so important for this product; its value lies in its ability to improve the ULSD pool in terms of volatility, gravity and viscosity.

Unfortunately, the practice of using the second unit to treat low-grade kerosene ran into difficulties. This was soon after the refinery had begun to import more opportunity and synthetic crudes. The kerosene distilled from these synthetic crudes had higher sulphur and nitrogen levels and challenging API gravities. All this kerosene was directed to the second unit, which had just been loaded with fresh catalyst, Criterion’s Ascent DC-2533. The impact of the feed change was felt almost immediately. The refinery had to increase the operating temperature of the unit by 8–11°C to meet the ULSD specification. And the catalyst deactivation rate tripled. The refinery technologists tried to manage this by establishing a weighted average bed temperature (WABT) budget and reducing the throughput of the unit.

However, despite the refiner’s best efforts, the cycle length was curtailed 12 months after its start. Given the likelihood that synthetic crudes would continue to feature strongly in the crude diet through the course of the next cycle, the refinery approached Criterion for help in devising the best way of maximising the cycle length. A team of hydroprocessing specialists from the refinery and Criterion was assembled to examine the composition and properties of the feed and to formulate the best operating strategy for the hydrotreating unit and the crude distillation unit used to process the synthetic crude. As part of this exercise, Criterion team members proposed switching to a new-generation CoMo–alumina catalyst, DC‑2618, based on Criterion’s Centera technology platform. The new catalyst had already demonstrated outstanding HDS and HDN activity and stability in other low to medium pressure ULSD units running difficult feeds.

Given the uncertainty surrounding the quality of the feed to the unit, the team was unable to predict accurately the effect of the new operating strategy and catalyst on cycle length. It was hoped, however, that the changes would add at least four months to the 12 months achieved previously.

The unit was loaded with Centera DC‑2618 and restarted with 100% kerosene from synthetic crude, which had been held in storage during the shutdown. The start of run WABT was as predicted. After this poor quality kerosene had been processed, the WABT fell by 6°C, and it became clear from the subsequent rate of temperature rise that the catalyst was retaining its activity better than its predecessor, even though the unit continued to be fed kerosene of variable quality.

Data obtained during the first months of operation led the team to revise its estimate of the cycle length. Technologists at the refinery began to believe it would be possible to reach the 24-month mark, which is double the length of the last run.

The enhanced performance of the unit has been put down to the catalyst’s outstanding HDN activity. While levels of 60–70% were typical during the previous run, analysis of data from the unit shows that the DC-2618 catalyst is achieving an HDN level of 90–95%. This means there is less nitrogen inhibition of the catalyst; consequently, the required HDS level can be achieved at reduced temperatures. This is highly significant: it has improved the colour and colour stability of the product from the second unit to the extent that it may be used as a jet fuel component.

Detailed performance studies by the team, taking into account feed quality and various operational scenarios, indicate that it should be possible to run for 24 months with high volume fractions of difficult kerosene feeds. At the same time, the refinery has gained the flexibility to swing between ULSD and jet fuel production, depending on the prevailing product market economics.

The refinery has a strong record of using its hydroprocessing assets to process less favoured crudes and achieve good margins. Therefore, the company is working with Criterion to increase the cycle length beyond the projected 24 months. It is also looking at other feed and product possibilities that have opened up as a result of the improved performance of the second unit with Centera catalyst.

For more information: Kevin.Carlson@criterion.com