SET-wash water ow
Vap for pp
Water dewpoint, velocity, H S pp check point
MIX-104
H pp check point
S
E uent plus wash water
V-100
HDS_wash water
HDS_RxtE Mix
Liquid from split
E uent plus wash
B
HDS_V3_Vap
AC-100
Velocity, H S pp check point
1
HDS_V3_Liq
E-102
HDS_R_E
2
HDS_V3_CHPS
HDS_V3_Q
HDS_V3_Wtr
HDS_C1_C6P
Q-101
E2 stripper feed
E
Figure 3 Example of potential fouling and corrosion mitigation solutions
an area not well addressed in many REAC corrosion miti- gation documents. Without good mixing, water washes or chemical injection are not very effective. With equipment process models, such as Petro-SIM, OLI, and other tools, all the parameters contributing to salt problems can be estimated. A key action will include keep- ing chlorides out of the system, as these are contaminants. Figure 2 identifies areas to look at just for REAC corrosion mitigation opportunities. Figure 3 shows some typical solu- tions that might be identified. Monitoring and process conditions Catalyst performance depends on more than steady- state operation. It requires ongoing monitoring of pro- cess, mechanical, and corrosion parameters, as well as the impact of unplanned events for equipment and catalyst. For example, the number of upsets, such as thermal cycle/shut- downs, can be linked to inspection requirements or even earlier hot bolting of flanges. This involved the integration of key performance indicator (KPIs) with integrity oper- ating windows (IOWs). API RP 584 – Integrity Operating Windows (IOWs) gives guidance on IOW development for equipment. The key is to integrate these IOWs with process KPI monitoring to create a more powerful tool. It is essential to include parameters that will identify impending short- and long-term problems, as well as target KPIs. This can include identification of changing corrosion rates, valve positions, fractionation loss, furnace operation, machinery efficiencies and exchanger fouling, as well as changes in deactivation. Predictive modelling is now a great asset; digital twin models can further improve reliability and optimisation. Some operations have cut unplanned outages by integrat- ing IOW and KPI monitoring and optimisation using digital twin models, such as those based on Petro-SIM. This is now advancing to AI-overseen applications, which should have a built-in knowledge base to manage any data anal- ysis system. Guidance correlations based solely on data analysis/ self-learning have sometimes been completely wrong, as certain knowledge information was not accounted for. For example, in some units, deactivation from cracked stock was
overestimated based on feed type data and deactivation. However, in reality, the perceived catalyst deactivation was due to fouling and maldistribution at the top of the bed, and most of the activity was fully recovered just by skimming the bed, reducing O₂ in feed. Process conditions optimisation Many parameters can be looked at to maximise catalyst uti- lisation within the operating envelopes, some of which are often overlooked. A few examples include: • Feed quality control, including fractionation control, cut point control, and routing the best feed to the best unit, such as easier feed to lower H₂S partial pressure units. • H₂ purity in recycle gas impacts performance and cycle length significantly, especially if below 95%. Even 1-2% makes a measurable difference. H₂ purity can be increased with more make/purge/use of amine scrubbers. Lower drum temperatures depending on the cost of H₂, if available. • Furnace constraints can be mitigated with a change in quench targets, reducing preheat fouling by minimising tank- age feed, drop-in recycle gas rates (if spare activity exists). • Better analysis of deactivation and catalyst conditions. • Hydraulic constraints can be overcome by running spare pumps in parallel. A more general practice now is to run all machinery to maximise capacity where pump curves permit. Summary Selecting the right high-activity catalyst is the first step in maximising its utilisation, especially now that harder feeds are processed and product quality is becoming more rig- orous. Many of the areas discussed, if not addressed, can result in the credit for a higher activity catalyst being com- pletely wiped out. Thus, for maximum catalyst utilisation, a multifaceted sustainment programme is required to main- tain performance and maximise catalyst credits. In some actual cases where every constraint has been rigorously challenged, even a ‘humble’ diesel hydrotreater resulted in millions of dollars of additional revenue. Petro-SIM is a trademark of KBC (A Yokogawa Company). Andrew Layton is a principal consultant at KBC (A Yokogawa Company). Email: Andrew.layton@kbc.global
72
PTQ Q4 2025
www.digitalrefining.com
Powered by FlippingBook