An ECA result is only as credible as the flaw characterization it is based on. If the size, shape, orientation, or location of the flaw is not well established by inspection, the ECA may assess a flaw that is smaller, simpler, or less critically located than the actual defect. The connection between NDE capability and flaw model is one of the most important — and most frequently underestimated — aspects of defensible ECA.
NDE sizing uncertainty must be explicitly incorporated into ECA flaw models. Ignoring it can produce ECA results that appear to accept a flaw as safe when the actual flaw could be significantly larger than the inspection data suggests.
The Flaw Model in ECA
In ECA, a crack-like flaw is characterized by its height (depth through wall), length (extent along the surface), orientation (axial, circumferential, transverse, oblique), location (surface-breaking, embedded, weld metal, HAZ, base metal), and aspect ratio (height to half-length ratio). These parameters directly control the crack driving force in the fracture mechanics assessment. The flaw model is derived from inspection data — but that data is subject to detection probability and sizing uncertainty.
NDE Sizing Uncertainty — Why It Cannot Be Ignored
All NDE methods have detection probability limitations and sizing uncertainty. PAUT and TOFD provide better sizing capability than conventional UT for planar flaws, but they still carry uncertainty that depends on flaw orientation, access, surface condition, procedure, qualification level, and analyst competence. For small flaws near the detection threshold, the actual flaw may be somewhat larger than the reported size due to the combination of detection and sizing uncertainty.
How to Handle Sizing Uncertainty in ECA
Recognized ECA standards require that sizing uncertainty be incorporated into the flaw model. This may be done by adding a sizing tolerance to the measured flaw dimensions before performing the fracture mechanics assessment. The magnitude of the tolerance depends on the NDE method used, the procedure, qualification evidence, and specific inspection conditions.
- BS 7910 Annex K addresses measurement uncertainty and inspection qualification requirements
- Flaw dimensions used in ECA should represent the 80th or 90th percentile of the likely actual flaw size, not simply the best-estimate measured value
- Using the reported measured size without adding sizing uncertainty may produce a non-conservative result
- The NDE method and procedure should be appropriate for the expected flaw type and orientation
Flaw Orientation and Its Impact on ECA
The orientation of a crack-like flaw significantly affects the crack driving force. Axial flaws in pressurized vessels and pipes experience hoop stress as the primary loading. Circumferential flaws experience both axial pressure stress and bending. Transverse flaws in welds may be oriented perpendicular to the weld residual stress profile. If the flaw orientation assumed in the ECA does not reflect what the NDE actually detected — or what the detection method is capable of reliably detecting — the assessment may not be appropriate.
Interaction Between NDE Qualification and ECA Acceptance
In some applications, the ECA is performed as part of acceptance of new construction welds (Alternative Acceptance Criteria in API 1104, ECA-based acceptance in offshore pipeline standards). In these cases, the NDE procedure and qualification must be compatible with the flaw acceptance sizes determined by ECA. If the NDE cannot reliably detect flaws of the critical acceptance size, the ECA-based acceptance is not valid regardless of how accurately the fracture mechanics was performed.
TES Canada ensures that ECA flaw models are matched to the actual capabilities of the NDE methods used. We review inspection procedures and sizing performance before applying ECA to ensure that flaw dimensions include appropriate sizing uncertainty and that the NDE method is capable of detecting and characterizing the flaw types of concern.
Standards & References
- BS 7910 — Guide to Methods for Assessing the Acceptability of Flaws in Metallic Structures — Sizing uncertainty and measurement uncertainty (Annex K)
- API 579-1 / ASME FFS-1 — Fitness-for-Service — Crack-like flaw assessment
- API 1104 Appendix A — Alternative Acceptance Standards for Girth Welds
- DNV-RP-F108 — Assessment of Flaws in Pipeline and Riser Girth Welds
Need support with this type of technical challenge?
TES Canada can help you assess the issue, select the right inspection or engineering approach, and develop a practical integrity management solution.
Contact TES Canada →