In-line inspection tools: MFL, ultrasonic, EMAT and caliper

An in-line inspection tool is a machine driven through the line by the product itself, recording as it goes. It is the only way to inspect the whole length of a buried pipeline without digging it up, and the whole industry depends on it. The four families below measure genuinely different physical quantities, and they are not substitutes for one another.

Magnetic flux leakage — the metal loss workhorse

Powerful magnets saturate the pipe wall with magnetic flux. Where the wall is full thickness the flux stays inside the steel. Where metal is missing there is not enough cross-section to carry it, so flux leaks out and sensors between the poles pick up the leakage.

What it is good at: corrosion. External corrosion on a buried line, internal pitting, under-deposit attack, general wall loss. It runs in gas or liquid, tolerates a dirty line better than the alternatives, and it is the tool most operators run first.

What it cannot do: it never measures thickness. It measures how much flux escaped and infers a depth from that, which is why MFL sizing carries a real tolerance band — commonly stated as something like ±10% of wall thickness at 80% confidence. That band is not a formality. A feature reported at 40% depth may genuinely be 30% or 50%, and your remaining-life calculation inherits the whole spread.

It is also directional. Axial MFL magnetises along the pipe and sees circumferential features well; a long narrow axial groove — a gouge, or a seam-weld corrosion channel — barely disturbs the flux and can be badly under-called. Circumferential MFL exists precisely to cover that gap, and running both is common on lines where axial features are credible.

Ultrasonic wall measurement — the accurate one

Each sensor fires an ultrasonic pulse at the wall and times two echoes: the inside surface and the outside. The difference is the wall thickness, measured directly.

What it is good at: accuracy. Tolerances are typically a few tenths of a millimetre rather than a percentage of wall, so it is the tool when a fitness-for-service assessment has to be done on the results. It also sees mid-wall features that MFL is blind to — laminations, hydrogen blistering, HIC — because it images through the thickness rather than around it.

What it cannot do: run in gas. Ultrasound needs a liquid couplant between the sensor and the wall, so a UT tool in a dry gas line reads nothing. Gas operators who want UT data have to batch a liquid slug through with the tool, which is a significant operation in itself.

It is also intolerant of a dirty line. Wax, debris and heavy scale block the signal and produce lost readings, so cleaning runs come first — often several.

EMAT — the crack tool

An electromagnetic acoustic transducer generates the ultrasonic wave in the pipe wall itself using a magnetic field and a coil, rather than transmitting it in through a liquid. No couplant is needed, which is the whole point.

What it is good at: cracking in gas lines. Stress corrosion cracking colonies, fatigue at the long seam, cracking in girth welds, and hook cracks in older ERW seams. Configured for it, EMAT can also detect disbonded coating — which is where external SCC starts, so finding the disbondment finds the problem earlier than finding the cracks.

What it cannot do: size shallow cracking reliably. Dense colonies interact and read as fewer, larger features than they are. A confirmatory dig with surface inspection is still how an EMAT call becomes a number you would put in an assessment.

Caliper and geometry — the shape tool

Mechanical arms or sensors mapping the internal bore, recording every change of shape and its position.

What it is good at: dents, ovality, wrinkles, buckles and bore restrictions. It is usually the first tool run in any new inspection programme, for a plainly practical reason: it proves the line will actually pass a more expensive tool without jamming it.

What it cannot do: see metal. A dent with a gouge in the bottom of it looks identical to a plain dent, and that distinction is the difference between a feature you monitor and a feature you cut out. Dents with metal loss are assessed under API 579 Part 12; plain dents are much less severe. Caliper alone cannot tell you which you have.

Matching tool to threat

The mistake that costs the most

Running one tool and treating the report as an inspection of the pipeline.

An MFL run is an inspection for metal loss. It is not an inspection for cracking, and it is a poor inspection for axial features unless the circumferential variant was used. A line with a clean MFL history and an SCC problem is not a rare situation; it is the normal way SCC is discovered, usually by a leak.