Mar-2023

Improved diesel hydrotreating catalyst loading scheme

A European refiner used an independent catalyst testing approach to confirm their existing hydrotreating unit’s ability to cope with different LCO blending targets.

Tiago Vilela and Nattapong Pongboot
Avantium

Article Summary

Refineries process blends of straight-run gasoil (SRGO) and light cycle oil (LCO). LCO is conventionally processed in hydrotreaters along with straight-run middle distillates to upgrade its economic value. From a product quality perspective, LCO has a relatively lower cetane number (poorer ignition performance in diesel engine) compared to straight-run middle distillates derived from the crude distillation unit. The aromatics content (low cetane components, mainly 2-ring aromatics) of LCO from FCC units can be as high as 85 wt% in a high severity FCC operation (such as high-octane gasoline or propylene mode).

Generally, the cetane number of LCOs ranges from 15-25 compared to 40-60 for straight-run diesel. It is also important to note that the cetane number is directly proportional to the total aromatics content.1 As such, the amount of LCO permitted in the diesel blending pool is often limited by this combustion property, forcing refiners to dispose the remaining LCO to the low-value fuel oil blending pool as the viscosity adjuster. To make matters worse, disposing of LCO as fuel oil is becoming more and more constrained by declining demand for heavy fuel oil as the world is moving towards zero-carbon emissions.

In addition to high aromatics content, a significant portion of organic sulphur (normally 0.2-2.5 wt%) is in the form of alkyl dibenzothiophenes (DBT), while organic nitrogen (typically 100-750 ppmw) is mostly constituted of non-basic organic nitrogen compounds (such as 5-ring membered carbazoles). These organic sulphur and nitrogen components are known to be refractive, posing challenges to ultra-low sulphur diesel (ULSD) operation. In most cases, LCO processing requires more severe hydrotreating (higher temperature) at the start-of-run (SOR) to meet the same product sulphur target (<10 ppmw for ULSD), thus limiting the cycle length. It should be noted that cycle length can also be limited by diesel ASTM colour specifications. It is common for the product colour to deteriorate over time from declined hydrogenation activity.

In general, feeding LCO along with straight-run middle distillates requires higher hydrogen consumption due to hydrogenation of unsaturated hydrocarbon compounds. From a ULSD perspective, a preferred reaction pathway is saturating the first aromatic ring of alkyl DBT (for better sulphur accessibility of metal active sites) prior to the sulphur removal by hydrogenolysis, thus adding to the total hydrogen consumption (see Figure 1). Removal of organic nitrogen compounds, an essential step before converting alkyl DBT, also contributes to the additional hydrogen consumption by increasing the total nitrogen content.

Special grading requirements
In addition to aromatics, the fact that LCO also contains a certain level of olefinic compounds (typically indicated by Bromine number) is a challenge in terms of grading bed design. Catalyst activity must initially be low enough and gradually increase over the reactor length to prevent rapid heat release, local hydrogen starvation, bed fouling from polymerisation, and coke deactivation. This additional special grading requirement can limit the volume of higher-activity hydrotreating catalysts (which is particularly important when processing LCO) when the reactor volume is fixed, such as existing reactors.

To accommodate a high portion of LCO in hydrotreaters, the hydrogen intake capacity must be large enough with adequate reactor volume, hydrogen partial pressure, and hydrogen circulation rate to ensure an acceptable catalyst deactivation rate. With higher temperature rises in the catalytic bed, a higher quenching rate will also be required to maintain catalyst bed thermal stability, thus adding to the total hydrogen circulation rate. Finally, a higher hydrotreating temperature requirement accelerates catalyst deactivation due to a faster coking rate.

These requirements are particularly important for existing hydrotreaters seeking the opportunity to process more LCO, as these old hydrotreating units are often limited in reactor volume, pressure rating, and hydrogen compression capacity. For this, it is important to evaluate the implications of introducing more LCO in diesel hydrotreating units and effectively evaluate the impact on hydrodesulphurisation (HDS) conversion and hydrogen consumption.

Catalyst loading schemes
In demonstrating how independent catalyst testing helped a European refiner confirm their existing hydrotreating unit’s ability to cope with different LCO blending targets,  it was crucial to focus on hydrogen consumption, cycle length, and aromatics content. Moreover, different catalyst loading schemes were evaluated to determine which one was the best fit for the existing equipment.

The catalyst evaluation was performed at Avantium laboratories in Amsterdam using a dedicated Refinery Catalyst Testing (RCT) high-throughput unit with 16 parallel reactors employing Avantium’s proprietary technology, which will be discussed and described further in Figure 3.

The test program consisted of a run of about 20 days (excluding activation) where four ULSD catalyst configurations (CoMo, NiMo, and stacked beds of NiMo + CoMo) loaded in duplicate reactors were exposed to three different feed blends (SRGO + LCO) with three levels of LCO: 15%, 30%, and 45%. The operating temperature was adjusted to reach a product sulphur of around 8 ppm.

The complete set of results obtained from the test was very consistent, showing the expected correlation among different measurements, such as hydrogen consumption, gas make, liquid product density, and product aromatics content. These results, combined with an exceptional reactor-to-reactor repeatability, confirmed the experimental test’s validity, relevance, and accuracy.

Experimental
The test program aimed to determine the effect on catalyst performance when processing blended feeds of SRGO + LCO in diesel hydrotreating units at start-of-run conditions (SOR). Eight catalytic systems of CoMo, NiMo, and stacked beds of NiMo + CoMo were exposed to the different blends of SRGO + LCO during a period of 20 days (excluding activation). Different feeds were introduced into the catalysts by changing the amount of LCO blended with the SRGO, while the operating temperature was adjusted at each condition to reach a product sulphur of 8 ppmw – initial temperature estimates were provided by the customer.

The minimum time of stabilisation used after each temperature adjustment was 24 hours. At least one liquid sample was collected at the end of this period, while the composition of the gas effluent from each reactor was analysed sequentially using a GC analyser (Agilent 7890B). The liquid sampling time was eight hours, so the amount of liquid collected was around 5 ml. The total concentration of sulphur and nitrogen was measured on the liquid product samples using an Xplorer TN/TS analyser according to the standard method ASTM D2622.