Technical guide • Eddy current testing

Eddy Current Data Analysis: Interpretation, Automation & Level III Oversight

Eddy current testing does not end when the probe is withdrawn. The reliability of an inspection depends on how the acquired signals are evaluated, compared with appropriate references, reviewed, resolved, and documented.

Core principle: An eddy current indication is an electromagnetic response—not a direct photograph of a defect. Analysis requires knowledge of the inspection technique, component geometry, material, calibration, acquisition quality, and known limitations.

What does eddy current data analysis involve?

Signal interpretation

Review amplitude, phase, signal shape, frequency response, spatial position, and behavior relative to known reference indications. Separate potential degradation from geometry, support features, lift-off, deposits, and other influences.

Evaluation and disposition

Apply the inspection procedure and applicable reporting or acceptance criteria. Distinguish detection from characterization and sizing; document uncertainty when the available evidence does not support a definitive conclusion.

Technical oversight

Confirm data quality, analyst qualifications, procedural compliance, independent review where required, resolution of disagreements, and traceability of the final reportable result.

Primary, secondary, and Level III review

In programs that require independent analysis, primary and secondary analysts evaluate inspection data under defined procedures. Their results may be compared to identify differences in detection, characterization, sizing, or reporting. The required independence and resolution process depend on the governing program.

Primary analysis

Conduct a systematic initial evaluation, identify relevant indications, apply established analysis rules, and record reportable results.

Secondary analysis

Provide a separate evaluation where required, helping identify missed indications, inconsistent calls, and interpretation differences.

Resolution and oversight

Investigate discrepancies using the original signals, calibration data, previous history, technique limitations, and additional technical review. The authorized Level III or designated resolution authority makes and documents decisions within the applicable program.

Independent Qualified Data Analysis (iQDA) is a defined practice in certain inspection programs, notably steam generator applications. It should not be assumed to be mandatory for every ECT application or interchangeable with ordinary peer review.

Why automated analysis needs oversight

Automated analysis and screening can help manage large inspection datasets, prioritize indications, and improve consistency when the system is properly configured and validated. However, automated calls depend on algorithms, training or rule sets, signal quality, thresholds, and the inspection conditions under which performance was demonstrated.

Potential advantages

  • Rapid review of repetitive, high-volume data
  • Consistent application of specified screening rules
  • Identification of indications for analyst attention
  • Useful comparison and trending tools when validated

Important limitations

  • False positives and false negatives remain possible
  • Unusual signals may fall outside validated conditions
  • Noise, geometry, deposits, and mixed mechanisms complicate interpretation
  • Software output does not replace required qualified review or technical accountability
Validation matters: Before relying on automation, evaluate performance using representative datasets, known indications, acquisition conditions, and defined acceptance criteria. Confirm how software changes, thresholds, and exceptions are controlled.

Array data analysis is not simply a better image

Eddy current array (ECA) systems acquire signals from multiple sensing elements and may display spatially referenced C-scans or other maps. Interpretation still depends on coil arrangement, operating mode, element spacing, multiplexing, encoding, scanning direction, and the underlying impedance-plane or channel responses.

Surface arrays and tubing arrays have different coverage and interpretation challenges. A colorful map can help locate and compare indications, but it does not directly establish flaw depth, morphology, or structural significance. Qualified analysis should preserve access to relevant underlying signal data and reference responses.

Historical data comparisons and signal trending

Previous inspection data can be extremely valuable when evaluating whether an indication is new, has changed, or is associated with a recurring geometric feature. Meaningful comparisons require attention to probe design, frequency, calibration, gain, phase rotation, scan speed, data registration, and analysis settings.

A difference between inspection campaigns does not automatically establish flaw growth. Conversely, a stable-looking signal does not prove the absence of degradation. Historical comparisons support decisions when acquisition and analysis conditions are sufficiently comparable and the uncertainty is understood.

What happens when analysts disagree?

Review stepQuestions to addressPurpose
Check acquisition qualityWere calibration, coverage, probe condition, scan speed, and data quality acceptable?Determine whether the original dataset supports reliable evaluation.
Review all relevant channelsDo other frequencies, modes, mixes, or array elements support or contradict the call?Test competing interpretations against the full available response.
Compare inspection historyWas a similar signal present previously, and are the datasets comparable?Look for meaningful change while avoiding misleading comparisons.
Evaluate alternate techniquesWould a different probe, scan direction, speed, or NDT method provide useful evidence?Reduce uncertainty when practical and authorized.
Document resolutionWho has authority, what evidence was considered, and what remains uncertain?Produce a traceable, technically defensible decision.

When an indication remains ambiguous, the appropriate response may be additional analysis, retesting, an alternate NDT method such as ultrasonic testing, a conservative disposition, or explicit documentation of limitations. The decision must follow the applicable inspection and engineering requirements; no single action is universally correct.

Forensic review: what ECT can and cannot establish

Detailed review of stored eddy current data may support investigations into missed indications, unusual signal responses, inspection performance, or changes between examination campaigns. This is sometimes described as forensic data analysis.

However, ECT alone generally cannot prove the root cause of a material failure. Metallography, fracture analysis, operating history, engineering assessment, and other evidence may be necessary. Analysis findings should distinguish observations, interpretations, and hypotheses.

Questions to ask when selecting a data-analysis provider

Technical competence

  • Which techniques and component types are supported?
  • What are the analyst and Level III qualification requirements?
  • How are primary/secondary analysis and disagreements handled?
  • What experience exists with the expected degradation mechanisms?

Quality and traceability

  • Can the team review raw data, calibration, and inspection history?
  • What software and automation validation are used?
  • How are changes, exceptions, and final decisions documented?
  • What confidentiality, record retention, and reporting controls apply?

Frequently asked questions

Does a clear eddy current signal mean the component is defect-free?

No. Detection depends on the technique, setup, component, discontinuity characteristics, and demonstrated inspection capability. An absence of reportable indications is not proof that every possible flaw is absent.

Can automated analysis replace qualified analysts?

Automation may support screening and interpretation, but whether it can be used for specific decisions depends on its validation and the governing procedure. Qualified oversight remains essential wherever required by the program.

Can array C-scans determine flaw depth directly?

Not inherently. Maps show processed electromagnetic responses. Depth or sizing estimates require suitable calibration, technique-specific relationships, and demonstrated performance.

Is Level III review required for every inspection result?

Not necessarily. Level III involvement, independent review, and resolution requirements depend on the applicable code, procedure, written practice, and inspection program.

Independent Level III consulting & data-analysis oversight

EddyCurrent.com accepts inquiries concerning independent eddy current Level III technical consulting, data-analysis oversight, procedure and technique review, complex indication resolution, and inspection-program support.

Each inquiry is evaluated individually for technical scope, qualifications, availability, confidentiality, and project requirements. Depending on the circumstances, services may be provided directly, coordinated with another qualified specialist with the client's agreement, or declined. No inspection, certification, or regulatory approval is implied by an initial inquiry.

Discuss a Project ↗Explore Service Providers ↗

Technical disclaimer: This material is educational and does not replace an approved inspection procedure, governing code, qualified engineering assessment, or site-specific Level III decision. Inspection methods and analysis practices must be selected and validated for the actual application.