MFL Pipeline Inspection Services 2026

How magnetic flux leakage actually works

The pipe wall is driven close to magnetic saturation. Metal loss reduces the cross-section available to carry that flux, so flux is forced out of the wall into the surrounding space where sensors detect it. The signal responds to the volume of metal missing, not to depth alone — which is the single most important thing to understand about interpreting the results.

What it finds, and what it will not

It is fast and effective on volumetric metal loss: corrosion, pitting and gouging, over large areas at survey speed. It is correspondingly poor at tight planar flaws such as cracks aligned with the flux, which remove almost no metal and therefore leak little. Very gradual general thinning can also under-respond, because it is the flux gradient that generates the signal. Where cracking is the credible failure mode, MFL is the wrong screening tool and ultrasonic or EMAT techniques belong in the scope.

Why amplitude is not depth

A broad shallow patch and a narrow deep pit can produce comparable signal amplitudes. This is precisely why MFL screens and ultrasonics confirms: the survey identifies where metal is missing and roughly how much, and targeted UT at those indications establishes the remaining wall that a fitness-for-service assessment actually needs. A report that converts amplitude directly to depth without that confirmation should be treated with caution.

What governs the work

API 1163 defines how an in-line inspection system is qualified — detection, identification, sizing and location validated against stated performance rather than vendor claim. ASTM E570 covers flux leakage examination of ferromagnetic tubing. API 653 governs tank floor examination, where MFL is the established screening method. NACE SP0102 addresses in-line inspection of pipelines more broadly.

What you receive

A survey record locating every indication, sized and classified; ultrasonic confirmation at the indications that matter; and the data in a form that feeds corrosion rate and remaining life rather than stopping at a defect list. The value is in what the next interval decision rests on, not in the length of the anomaly table.

Tank floors — where MFL earns its place

A storage tank floor is the classic application, and the reason is geometry rather than preference. The floor is large, flat, largely inaccessible from beneath, and corroding from the soil side where nothing can be seen. Scanning it ultrasonically point by point would take an implausible amount of time; MFL covers it at walking pace and identifies where the metal has gone. Under API 653 that screening result then drives where thickness is confirmed, whether patch plating or floor replacement is required, and how long the tank can run before the next internal.

Pipelines and in-line inspection

On pipelines the same physics travels inside a tool driven by the product. The constraints are different: tool speed affects signal, wall thickness must be within the range the magnetiser can saturate, and features such as bends, valves and wall-thickness transitions all produce signals that must be distinguished from metal loss. This is why API 1163 exists — it obliges the system, the procedures and the personnel to be qualified together and the results validated against field verification digs, rather than accepting a specification sheet as achieved performance.

How to read an MFL report critically

Three questions separate a report you can act on from one you cannot. First, was the tool qualified for the wall thickness and defect type actually present, or was a generic specification applied? Second, were the reported depths confirmed by ultrasonics at a representative sample of indications, and did the confirmations agree? Third, are the indications located precisely enough to be found again — a depth estimate with a vague position cannot support an excavation decision. A provider who cannot answer all three is offering a defect list rather than an integrity assessment.

When MFL is the wrong tool

It is worth being direct about this. If the credible failure mechanism is stress corrosion cracking, fatigue cracking at a weld, or hydrogen damage, MFL will not reliably find it and a survey that comes back clean will be actively misleading. Those mechanisms need ultrasonic, EMAT or crack-specific in-line tools. The right sequence is to establish the damage mechanism first and select the technique second — the reverse, which is more common, is how organisations end up with a clean survey and a failure.

Frequently asked questions

Does MFL work through coatings?

Yes, within limits. Coating thickness adds lift-off between sensor and steel, which reduces signal amplitude, so it must be known and accounted for during setup. Heavy or irregular coating erodes sensitivity to the point where the survey stops being reliable, and at that stage removal or an alternative technique is the honest recommendation.

Can MFL find cracks?

Not dependably. Cracks aligned with the flux path remove almost no metal and produce little leakage. MFL is a metal-loss tool. If cracking is what you are protecting against, the scope needs a technique chosen for it rather than an assumption that one survey covers everything.

Related: NDT consulting · NDT training · request this inspection.

Related: what operators and regulators actually check in a pipeline integrity audit — including the operator-qualification trap that catches NDT-certified technicians.

Related reading

• Floor scanning is one part of a wider tank inspection scope. API 653 tank floor inspection.