PTQ Q3 2025 Issue

Overhead condenser

To use the model for troubleshooting

Vent

Model l ing, troubleshooting, case studies

Overhead vapour

Condensate

To identify deviation based on low & high value

Data historian, lab analysis, corrosion data

Petro-SIM + OLI engine

KPI (in Petro-SIM)

pH, Cl conc., solids, generalised corrosion rate * , velocity etc.

Corrosion rate, metallurgy

Stream properties

Condensate receiver

Product

OLI application

* Only when operating below ionic dew point

Sour water

Fractionation tower

Corrosion analysis

Product/ reux pump

Reux

Figure 2 Corrosion digital twin

the damage may have already occurred. As one industry source notes, corrosion acts like a ‘slow poison’. Often, signs emerge only after equipment is severely degraded or a leak has occurred.5 This reactive approach carries sub - stantial risks. In a CDU, sudden overhead line failures can trigger emergency shutdowns, posing both safety hazards and production losses. A solution to monitor corrosion continuously, anticipate risks earlier, and act before failure occurs is essential. To better understand where these mechanisms occur, Figure 1 illustrates the typical equipment arrangement of a CDU column overhead system. This includes the overhead line, condenser, reflux drum, and associated streams where corrosive phases such as sour water, HCl, and ammonium chloride are likely to form. Corrosion mechanisms The most important corrosion mechanisms that can appear in the CDU overhead can be summarised as follows: • HCl corrosion caused by the dilution of HCl vapour into liquid water. The first water droplets formed during con - densation can be extremely acidic and provoke localised corrosion at very low pH. • Ammonium chloride (NH₄Cl) corrosion formed by HCl and NH₃ vapours. These salts are dry and non-corrosive when temperatures are above the dew point but become highly corrosive when wetted. • Amine hydrochloride salt corrosion is similar to NH₄Cl but often more severe due to the presence of liquid-phase salts. • Wet H₂S damage can cause blistering and cracking, especially around welds under specifc pH and composi- tion conditions. Minimisation of overhead corrosion Ammonia (NH₃) is commonly used to neutralise HCl and control pH in the condensing phase. However, NH₃ dis - solves more slowly than HCl, which limits its ability to neu - tralise the first-condensing water. Optimal pH is generally maintained between 5.5 and 6.5 to balance corrosion and salt formation risks. Figure 1 Scope of the corrosion model – typical equipment arrangement of a CDU column overhead system

Solution: Corrosion digital twin A corrosion digital twin combines live plant data with first-principles simulation to continuously monitor and predict corrosion risk in the CDU overhead system. Built using KBC’s proprietary Petro-SIM process simulator and the OLI engine, the model calculates stream properties and simulates corrosion chemistry dynamically throughout the overhead system. By integrating engineering models with real-time plant data, Petro-SIM enables refiners to improve efficiency, reduce emissions, and gain insights far beyond what periodic field measurements can capture. To ensure meaningful and reliable corrosion predictions, the simulation setup must consider the composition of all streams near the overhead line, particularly those contain- ing H₂S, CO₂, NH₃, HCl, and water. Modelling the overhead system helps assess risks from entrained sour water and early-phase condensation. Neutralising agents like morpho- line or ammonia must be carefully evaluated to avoid creat - ing corrosive amine hydrochloride salts or triggering NH₄Cl precipitation. This rigorous approach enhances the value of digital twin analysis. The objective of this system is to be able to monitor process equipment in corrosive environments throughout the whole refinery and monitor and predict corrosion. A further objective is to take accurate measures against process equipment corrosion using this system. As shown in Figure 2 , the digital twin serves as both a monitoring and diagnostic tool. Plant data, such as lab anal - yses, corrosion measurements, and process historian values, feeds into the Petro-SIM and OLI engine, where thermody- namic modelling and corrosion prediction take place. Stream properties are then passed to the OLI application for corro- sion analysis and returned to the Petro-SIM and OLI engine. Simultaneously, Petro-SIM computes key performance indi- cators (KPIs), such as pH, chloride concentration, and salting/ condensation points. These KPIs are continuously monitored to identify deviations, support real-time troubleshooting, and guide proactive decision-making, especially under conditions below the ionic dew point where corrosion risk is highest. Case study As part of a refinery’s digitalisation and reliability strategy, a corrosion monitoring digital twin was implemented for

PTQ Q3 2025