Refining India June 2026 Issue

Improved DHDT efficiency with AI-based inferential prediction An AI-based inferential prediction model can significantly improve diesel hydrotreater catalyst life, performance, and product quality

Susanta Sadhukhan, Saurabh Gupta, Kubade Rk, Nandanwar P T, and Ajith D BPCL Kochi Refinery

T he diesel hydrotreater (DHDT) is a critical unit in a refinery that is responsible for removing sulphur, nitrogen, and aromatics from diesel-range feedstocks to meet stringent environmental regulations and product quality standards. Effective monitoring of the catalyst life, reactor performance, and product quality is essential to ensure optimal operation, economic efficiency, and regulatory compliance. The catalyst in a DHDT, typically composed of CoMo or NiMo on alumina, gradually deactivates due to fouling, sintering, poisoning (such as metals, nitrogen compounds), and coke formation. Another main parameter that affects the catalyst life is the need to increase bed temperature due to feed quality. In diesel hydrotreating processes, the quality and composition of the feedstock significantly influence catalyst performance and longevity. Heavier boiling range feedstocks, typically characterised by higher final boiling points and greater content of complex hydrocarbons, such as aromatics, resins, and asphaltenes, pose several challenges that can negatively impact catalyst life. Increased concentration of sterically hindered sulphur species, such as dibenzothiophenes (DBTs), in the feed aggravates the challenge of ultra low-sulphur diesel (ULSD) production, resulting in increased catalyst bed temperature and lower catalyst cycle length. The theory and information shown in Figure 1 are well known to all technologists involved in hydrotreating. It is an important factor to measure and control the feed quality of the DHDT unit to maintain good catalyst life and performance. Maintaining feed distillation has several impacts, such as:

u If the final boiling point (FBP) exceeds the catalyst life requirement, the catalyst bed temperatures need to be increased above the desired level. This directly shortens the catalyst cycle length. v If the catalyst bed temperature is not increased to maintain catalyst life, the product specification will fall outside the desired level, leading to the generation of off-grade product. w If a leaner feed quality is maintained, such as a lower FBP of atmospheric gas oil (AGO) from the crude distillation unit (CDU) than what the design specifies, any peaks in the AGO distillation are avoided. These peaks could result in a product diesel sulphur spike or catalyst deactivation in the downstream DHDT unit, which will lead to a production loss in the CDU. Challenges in monitoring the feed quality of the BPCL KR DHDT unit The DHDT at BPCL Kochi Refinery, licensed by Topsoe, was commissioned in January 2017 and

Mol. structure

Boiling point (˚C)

Typical HDS rate

Compound

HC S

Diethylsulphide

600

Thiophene

Benzothiophene

221

Dibenzothiophene

333

4, 6- Dimethyldibenzothiophene

360

<0.1

Figure 1 Theory of desulphurisation reactivity: a sulphur species versus reactivity

Refining India