PTQ Q2 2026 Issue

forming so-called ‘hammock’ profiles. This premature shell saturation restricts access to internal pore volume, resulting in incomplete utilisation of the available nickel surface area and reduced overall arsenic capacity. Minimising ‘hammock’-type profiles is critical in severe hydroprocessing applications, particularly when process - ing feeds with elevated arsenic content or when long cycle length and maximum arsenic pick-up are required. The improved diffusion characteristics and high effective nickel surface area ensure a more uniform arsenic distribution, ena - bling superior volume-based arsenic utilisation, higher total pick-up efficiency, and more stable performance over time. The TK-49 and TK-51 arsenic guards in extrudate form, as well as TK-45 and TK-41 in ring-shaped geometries, provide flexibility in reactor loading, high structural robust - ness, and reliable arsenic capture across a wide range of hydroprocessing applications. Canister testing during the development of the latest arsenic guard revealed that arse - nic capture is directly correlated with the available nickel surface area throughout the catalyst, as previously shown in Figure 3 . Therefore, TK-51 was specifically engineered to maximise the effective nickel surface area throughout the catalyst’s pore structure, resulting in more uniform arsenic penetration and improved volumetric utilisation of the avail - able nickel surface area throughout the catalyst pellet. Industrial application A dedicated arsenic guard system was installed in a com - mercial diesel hydrotreating unit in Southeast Asia. The unit operating pressure is 70 bar (1,030 psig), and liquid space velocity (LHSV) across the bulk catalyst is 1.4 h -1 , with a product sulphur target of less than 8 wt ppm. The unit processed a feed slate with an average arsenic con - tent of roughly 1,000-1,300 wt ppb, which is among the most severe arsenic levels seen in feeds to hydroprocessing units. Prior to installing Topsoe’s arsenic guard, the diesel unit struggled to meet its targeted cycle length due to severe deactivation caused by arsenic poisoning. This resulted in

Proportional relationship between nickel surface area and arsenic uptake

Nickel surface area (m/g, arbitary scale)

Spent catalysts from the diesel hydrotreater (DHT) were analysed by scanning electron microscopy (SEM) combined with X-ray fluorescence (XRF) elemental mapping at Topsoe R&D facilities to quantify arsenic accumulation, evaluate radial arsenic distribution, and assess effective nickel utili - sation in TK-51. Figure 5 shows that with the DHT sample, arsenic is distributed uniformly across the catalyst cross-section, indicating that arsenic uptake in TK-51 is not strongly dif - fusion-limited under these conditions. The absence of a pronounced surface-biased (‘hammock’) loading profile confirms effective internal transport and penetration of arsenic throughout the particle volume. As a result, TK-51 achieves near-complete nickel utilisation, providing an arse - nic pick-up of higher than 11 wt%. X-ray diffraction (XRD) analysis, combined with Rietveld refinement, was used to identify crystalline arsenic-contain - ing phases in the spent catalyst. Rietveld refinement is a full pattern fitting approach in which the entire diffraction pro - file is modelled to extract phase composition and structural Figure 3 Optimisation of arsenic trap by enhancing nickel surface area

the refiner either skimming the top reactor beds mid-cy - cle or reducing feed rates to meet the projected turna - round date. After installing a custom - ised grading solution, includ - ing TK-51 to trap the arsenic, the unit operated with great stability and a low deactiva - tion rate of 0.5°F/month or 0.4°F/month/LHSV during the unit’s most recent cycle, as shown in Figure 4 . The increased stability translated directly into improved opera - tional flexibility, allowing the refiner to maintain target feed rates and cycle length without mid-cycle corrective actions.

Normalised HDS WABT

Deactivation rate: 0.5 ˚F/month or 0.4 ˚F/month/LHSV

100

200

300

400

500

600

700

800

900

1000

1100

Run days

Figure 4 HDS normalised weighted average bed temperature (WABT) (industrial diesel hydrotreater; feed As ~1,000 wt ppb)

PTQ Q2 2026