Revamps 2026 Issue

the train was delivering crude to the heater close to its design preheat temperature. Six months in, the outlet tem- perature had dropped notice- ably. A year in, the shortfall was large enough that the heater was visibly working harder. Fuel gas consumption had risen, stack temperatures were up, and the unit was approaching its firebox duty limit earlier than planned. Simply knowing the train is fouled is not enough. Without understanding which exchangers are responsible,

Figure 3 Certain tubes such as those circled were significantly more restricted, causing flow to redistribute away from blocked tubes and reducing effective heat transfer area

how badly each one is affected, and what is driving the dep- osition, any cleaning strategy is essentially guesswork. The problem returns on the same timeline regardless of what was done at the turnaround. Troubleshooting measures thermal performance mapping The first step was a rigorous thermal performance assess - ment of every exchanger in the train. This meant collecting reliable operating data, inlet and outlet temperatures on both sides, flow rates, pressure drops, and fluid properties, and running a current thermal rating against each exchang- er’s clean design basis. The result was a fouling resistance profile across the entire train, showing clearly where the thermal performance gap was largest. The mapping exercise revealed something the plant team had not fully appreciated: fouling was not evenly distrib- uted. Exchangers at the cold end of the train, handling crude below 150°C, were performing reasonably close to their clean ratings. The real deterioration was concentrated in three exchangers at the hot end, where crude tempera- tures were in the 220-280°C range and shell-side streams included heavy vacuum gas oil (HVGO) and atmospheric residue. High-temperature crude containing reactive asphaltene components is significantly more prone to forming tenacious carbonaceous deposits, which explains the distribution clearly. Pressure drop analysis Thermal performance data alone does not give the full picture. Tracking pressure drop trends alongside thermal performance distinguishes between a uniform fouling layer that reduces conductivity and partial tube plugging that dis- torts internal flow distribution. In the hot-end exchangers, pressure drop data showed a clear upward trend over the run on the crude side. The pressure drop increase was dis- proportionately large relative to the thermal performance loss, indicating that while most tube surfaces were fouled, a subset of tubes was significantly more restricted, causing flow to redistribute away from blocked tubes and reducing effective heat transfer area beyond what average fouling resistance calculations were capturing (see Figure 3 ).

Velocity and wall temperature assessment Fouling in crude service is strongly influenced by local fluid velocity and TWT simultaneously. Low crude-side velocity allows the boundary layer to thicken, increases residence time near the tube wall, and reduces the shear forces that limit deposit adhesion. High wall temperatures accelerate the cracking and polymerisation reactions that convert dissolved asphaltene species into insoluble, adherent coke precursors. The combination of low velocity and high wall temperature is the worst possible scenario. A velocity assessment using operating flow data and known tube geometry found several hot-end exchangers running at crude-side velocities well below design. In some cases, under 0.6 m/s against a design basis of over 1.0 m/s. At these velocities, the fouling rate is substantially higher than the design fouling factor assumed in the original ther- mal rating. The exchangers were caught in a self-reinforcing cycle: low velocity accelerated fouling, fouling restricted flow, and lower flow further reduced velocity. Estimated TWTs in the most heavily fouled exchangers were approaching 300°C locally, well above the threshold at which coke precursor for- mation proceeds at meaningful rates in this crude blend. Cleaning strategy and pass arrangement modification Rather than cleaning all exchangers uniformly at the turn- around, the historical practice, a prioritised approach was developed based on fouling severity and the thermal value of each exchanger. The three hot-end exchangers were scheduled for mechanical hydroblasting. For two of these, the tube-side pass arrangement was converted from two- pass to four-pass, increasing crude-side velocity at the cur- rent throughput without changing the shell or bundle. This modification alone raised average crude-side velocity in those exchangers by approximately 60%, moving it above the threshold at which fouling rate decreases significantly. For cold-end exchangers where fouling was manageable, online chemical dispersant injection into the crude stream was evaluated to extend the interval before mechanical cleaning would be needed. Conclusion The three cases discussed share a common thread, even

Revamps 2026