Digital Twin for Heat Exchangers

Shell-and-tube heat exchanger digital twin — ECT and IRIS tube inspection mapping, tubesheet ligament UT, shell-side corrosion and plug history in one model. Track tube wall loss, plugging maps and bundle condition across cleanings to predict failures and optimise retubing decisions. See the digital twin platform and NDT consulting.

What makes heat exchanger integrity different

Heat exchanger integrity is dominated by the tube bundle, and tube inspection produces a volume and shape of data that general asset systems handle badly. A single bundle can hold thousands of tubes, each examined along its length, each producing a trace that must be associated with a tube position in the tubesheet layout. Without that positional binding the data is a list of numbers rather than a picture of the bundle.

Technique selection follows material and mechanism. Non-ferromagnetic tubing is screened with bobbin eddy current for volumetric loss, with array or rotating probes following up where circumferential cracking is suspected, because the bobbin coil averages around the circumference and can hide a short circumferential flaw. Ferromagnetic tubing needs a different approach entirely — remote field, magnetic flux leakage or partial saturation — since permeability variation swamps conventional eddy current.

The decisions the data feeds are plugging and retubing. A plugging decision is per-tube; a retube decision is about the population and its trend. Both need history: a bundle where the same quadrant degrades each cycle is telling you something about flow distribution or maldistribution that a single inspection cannot, and that pattern is only visible when successive inspections are held in the same positional model.

Damage mechanisms worth planning around include under-deposit corrosion where flow is poor, erosion at tube inlets where turbulence is highest, vibration-induced fretting at baffles, and stress corrosion cracking where chlorides concentrate. Each has a characteristic location, which is the argument for planning examination by mechanism rather than sampling uniformly.