The cost of choosing the wrong metal-loss tool

An operator commits to an in-line inspection (ILI) run months ahead. Cleaning programme, launcher and receiver work, flow management, and often a shutdown or a throughput constraint all have to be scheduled around it. If the tool returns data that cannot support the next decision, the money is spent and the risk hasn't moved. Two failure modes come up again and again. An ultrasonic tool gets starved of couplant in a gas line, or a magnetic tool's sizing uncertainty produces a long list of precautionary digs. Either way you end up with conservative repairs, deferred assessments, or a repeat run.

So choosing between magnetic flux leakage (MFL) and ultrasonic (UT) wall measurement is as much a data-quality and fitness-for-service call as it is a technology call. This article walks through how to make that call for corrosion surveys.

Standards and requirements context

The tool is only one piece of the compliance picture. These documents set out what the run has to deliver:

  • API 1163 (In-line Inspection Systems Qualification) covers qualification of ILI systems, including how vendors establish and document performance specifications and how operators verify results against field measurements.
  • ASME B31.8S (Managing System Integrity of Gas Pipelines) and API 1160 (Managing System Integrity for Hazardous Liquid Pipelines) are the integrity management frameworks for gas and liquid lines. They drive threat identification, assessment method selection, and reassessment.
  • ASME B31G (Manual for Determining Remaining Strength of Corroded Pipelines) and API 579-1/ASME FFS-1 (Fitness-for-Service) are the main routes for turning reported metal loss into remaining strength and acceptance decisions.
  • NACE SP0102 (In-Line Inspection of Pipelines) gives general guidance on planning and running ILI programmes.

Your regulator may add prescriptive assessment intervals and repair criteria on top of these. Check the jurisdiction-specific requirements separately, because they change how much sizing confidence you need.

How the two technologies measure metal loss

MFL: an indirect measurement

MFL detects the magnetic flux leakage caused by defects, whereas UT measures wall thickness directly using ultrasonic waves (iFactory comparison guide). The tool magnetises the pipe wall. Where metal is missing, flux leaks out and sensors pick it up. Depth, length and width are then inferred from signal amplitude and shape through calibrated sizing models.

What that means in practice:

  • MFL works in many pipeline environments without the liquid coupling that conventional ultrasonic tools need (American Pipeline Solutions). That is the main reason it is the default for gas lines.
  • It covers long distances well (Enduro) and is widely used for general corrosion surveys.
  • Sizing depends on the relationship between signal and geometry. Complex, interacting, or oddly shaped features are harder to characterise than smooth, isolated ones. Wall thickness, magnetisation level, and tool speed all shape the response.

UT: a direct measurement

Ultrasonic tools time the return of sound pulses, so wall thickness comes from a physical measurement, not a modelled signal. UT usually needs a liquid couplant (Enduro). In an oil, product, or water line, the product itself does the job. In a gas line you have to introduce a liquid slug or batch, and that adds operational complexity and risk.

What that means in practice:

  • Direct thickness measurement generally gives more detailed wall data, which makes it the preferred choice where detailed wall thickness information is needed (American Pipeline Solutions). The iFactory guide similarly positions UT as the stronger option in liquid pipelines.
  • Whether metal loss is internal or external is inherent in the measurement, not inferred from it.
  • Performance hinges on keeping acoustic coupling intact. Wax, scale, debris, or gas pockets can cause data dropouts, so cleaning quality and flow conditions matter far more here than they do for MFL.

Where EMAT fits

Electromagnetic acoustic transducer (EMAT) tools can detect cracks in dry lines without liquid couplants (Winterhawk). They are primarily a crack-detection option, not a corrosion-survey alternative. If your threat assessment lists both corrosion and cracking in a gas line, EMAT is a separate decision from the MFL/UT question.

Comparison for corrosion surveys

Criterion MFL UT
Measurement principle Indirect: flux leakage interpreted through sizing models Direct: wall thickness from ultrasonic signal
Product medium Gas or liquid Needs liquid couplant; gas lines need a liquid slug or batch
Sensitivity to line cleanliness Moderate High; coupling loss produces data gaps
Long-distance runs Well suited Feasible in liquid lines, but dependent on coupling and sensor condition
Wall detail and profile Derived from signal Higher resolution thickness data
Internal/external discrimination Depends on tool configuration Inherent in measurement
Operational complexity Lower Higher, especially in gas
Typical default Gas lines and general corrosion screening Liquid lines and detailed wall assessment

These are general tendencies, and individual tool generations vary. Confirm each one against the vendor's published performance specification for your pipe grade, wall thickness, and defect type.

Data quality trade-offs that drive selection

Detection, identification, and sizing are separate questions

Vendor specifications usually state probability of detection, probability of identification, and sizing tolerance at a stated confidence level. Look closely at which defect classes those numbers apply to. A tool can meet its depth tolerance for general corrosion and still carry a looser specification for pinholes, axial grooving, or narrow features. Ask for the specification broken down by defect class, and check how it was validated, consistent with the qualification approach in API 1163.

Defect morphology matters more than the headline

  • General and broad corrosion: both technologies handle it well. MFL is usually the economical choice.
  • Pitting: small, closely spaced pits challenge any tool. Establish whether the specification covers your expected pit geometry, and be wary of clusters that the tool may report as one feature or as several.
  • Narrow axial or circumferential grooving: an MFL tool's response depends on the orientation of the feature relative to the field. Confirm the magnetisation configuration suits the expected orientation.
  • Laminations and wall variation: direct thickness measurement is better placed to reveal these than a magnetic signal.

Cleanliness and operating envelope

Both tools need a properly cleaned pipeline and a run speed inside the tool's qualified window. UT is more exposed to coupling disruption, and MFL is more sensitive to speed effects on the signal. Check both against your flow regime. In gas lines, throughput constraints often settle the matter before sizing accuracy does.

Using both

A feature in Pipeline & Gas Journal on choosing between MFL, UT, or both treats dual-technology and combined approaches as a legitimate answer when neither alone meets the data requirement. A common programme design is to run MFL as a screening survey and reserve UT for segments that need detailed wall data. It also gives you a cross-check on sizing. The NDT Global case study comparing MFL and UT tools illustrates this kind of side-by-side evaluation, and the vendor white paper on UT and MFL describes the complementary strengths of the two approaches.

How ILI results feed fitness-for-service

An ILI report is an input to an assessment. It is not the assessment itself.

What the assessment needs

Remaining-strength methods based on ASME B31G, and the metal-loss procedures in API 579-1/ASME FFS-1, need defect length and depth. Depending on the level of assessment they also need width and an effective profile of the remaining wall. Simplified methods treat the defect as a bounding shape, which is conservative. Methods that use a measured river-bottom or effective-area profile recover margin, but only if the data resolution supports it.

This is where tool choice shows up in repair decisions. Higher-resolution wall data can reduce the conservatism the assessment needs. Wider sizing uncertainty has to be carried into the calculation, so more features fall below acceptance and more digs follow.

Close the loop with field verification

  1. Select a representative set of excavations that covers the range of reported depths and feature types, not only the deepest anomalies.
  2. Compare field measurements with ILI-reported dimensions, and check whether the results fall within the vendor's stated tolerance and confidence level.
  3. If measured data fall outside the specification, take it up with the vendor. You may need to adjust the sizing algorithm or apply a correction to the assessment.
  4. Carry the verified uncertainty into the remaining-strength calculation and the reassessment interval.

Corrosion growth

Growth estimation means comparing successive runs. Use the same technology, or a documented method for reconciling different technologies. Sizing differences between an MFL run and a UT run can be mistaken for real growth, or they can hide it. If you are thinking about changing technology between runs, plan a feature-matching and reconciliation step rather than relying on a raw comparison.

Illustrative scenario

This scenario is constructed for illustration and is not drawn from a source.

An operator has a gas gathering line with a history of internal corrosion from water dropout at low points. The last survey, years earlier, was MFL. The integrity team wants to know whether a few clusters near low points are growing and whether any of them need repair.

Option one is another MFL run. It suits the gas medium, needs no liquid batch, and allows direct comparison with the earlier survey, which keeps the growth assessment simple. The catch is that clustered corrosion at low points is exactly where sizing is hardest. The report may flag many features whose assessment is limited by uncertainty.

Option two is a UT run with a liquid slug. It could give better wall detail at the clusters. It also brings operational complexity, a coupling risk if the slug doesn't hold through the low points, and a technology change that complicates the growth analysis.

A reasonable programme repeats MFL for the full line and the growth comparison, then targets the low-point clusters for follow-up. That follow-up could be direct examination at excavations, or a localised UT inspection if access allows. The verification digs then calibrate how much to trust the MFL sizing across the rest of the line. The decision rests on medium, defect morphology, and the need to compare with the previous run. A general preference for one technology doesn't enter into it.

Selection checklist

Use these points when preparing the tool selection:

  • Define the threat first: general corrosion, pitting, grooving, laminations, or cracking. Cracking points to a different tool class.
  • Confirm the product medium. Gas favours MFL unless you can justify and manage a liquid slug or batch for UT.
  • Request vendor performance specifications by defect class, with the validation basis, and review them against API 1163.
  • Verify pipe wall thickness, grade, and any heavy-wall sections against the tool's qualified range.
  • Assess line cleanliness, and set cleaning-pig acceptance criteria before the inspection run.
  • Check achievable flow and speed against the tool's qualified speed window.
  • Decide what the assessment needs: bounding dimensions for a screening method, or a detailed profile for effective-area methods.
  • Plan the reconciliation method if the technology differs from the previous run.
  • Define the dig-verification programme and acceptance criteria for tool performance before the run.
  • Plan launcher and receiver operations as hazardous-energy work. That means isolating and depressurising the trap, verifying zero energy, applying lockout/tagout, controlling ignition sources in the classified area, monitoring with gas detection, and venting or purging to a safe location. Handle pyrophoric iron sulphide deposits in received debris with particular care.

Conclusion and next steps

MFL is generally the practical choice for gas lines and broad corrosion screening. UT is stronger where direct wall thickness data and a liquid couplant are available and the assessment benefits from detailed wall data. Combined programmes make sense when neither technology alone satisfies the data requirement.

As a next step, write the assessment requirement before you ask for tool proposals. Define which defect types matter, what sizing confidence you need, and which ASME B31G or API 579-1/ASME FFS-1 route you'll use. Then ask each vendor to demonstrate performance against those requirements for your pipe and medium. Finally, plan the verification digs before the run, so the data can be trusted when the repair decisions get made.