Forty-eight circuits on one rack, every reading in its place.
The same sample unit as the workbench, built as a clay model: three tiers of insulated pipe on a rack, the supports, dead legs, penetrations and low points where corrosion under insulation starts, and a reading point at each of them. Drag to orbit, scroll to zoom, click a pipe for its record.
Run the sweep and a probe works through the circuits in workbench order. Each reading point turns from unread to a wall-thickness result as the probe passes, so you can watch where the ranked order finds the damage and how early.
What sits between the probe and the steel.
A pulsed eddy current probe reads wall thickness through the jacketing and insulation without stripping anything. The layers below are the ones it has to see through, and the last one is the only one that matters.
Water gets in at the seams. Corrosion starts at the supports, low points and dead legs, where it stays.
Jacketing keeps rain out until a seam opens or a walkway boot lands on it. Absorbent insulation such as calcium silicate then holds water against the coating, and the coating, if there is one, is what buys time.
The temperature band does the rest. Between roughly 50 °C and 120 °C a carbon-steel wall stays wet and warm, which is the fastest corroding condition API 583 describes. Below that, condensation cycles; above about 175 °C, the wall dries out and the risk drops.
The probe reports one number per point: remaining wall. The workbench decides which points to read first.
The probe switches a field on through the insulation and then off. Eddy currents in the steel fade at a rate set by how much steel is there.
The time it takes the signal to fall away is converted to an average wall thickness under the probe footprint. Nominal 9.5 mm reads as the slow green curve; a wall at 7.1 mm reads as the fast red one.
A crew can read a few hundred points in a campaign. Ranking the circuits first means those points land where the wet, warm, uncoated steel is, instead of wherever the walkdown started.
P-320-018
Send one export.
Get a written scorecard back.
Reply to the thread or write to the address below. No meetings needed unless you want one.
jamil@clearanvil.com