Radiography Accepted the Weld; TOFD Did Not
A 38 mm thick, 2.25Cr-1Mo (P-No. 5A) main steam pipe girth weld was radiographed with Ir-192 using a single-wall single-image technique, 100% coverage, IQI sensitivity 2.0%, film density 2.4, and was accepted by the third-party inspector against the project's radiography acceptance criteria. Six weeks later, an automated ultrasonic examination (TOFD) run from the outside as part of a fitness-for-service assessment reports a 30 mm long, 3.5 mm high planar indication at the root, oriented roughly perpendicular to the pipe surface and showing a lack-of-fusion signature. The hydrotest at 1.5 times design pressure was passed. The client asks whether the earlier radiography acceptance stands and whether the joint must be cut out.
1. Explain how a 3.5 mm high planar root defect can survive a 100% radiograph that met sensitivity and density requirements.
Read the worked answer
Radiography is a volumetric technique; it responds to change in thickness along the beam path, not to the existence of a defect. Several factors combined here:
- Orientation. The indication lies roughly perpendicular to the pipe surface and parallel to the beam at the root region on that side of the joint. A tight lack-of-fusion plane gives almost no attenuation difference across the beam, so contrast at the defect is minimal.
- Tightness. Lack of fusion has essentially no volume. Unlike porosity or slag, it does not present a lower-density path, so there is no image to see unless the crack is open and the beam happens to align with the plane.
- Energy and geometry. Ir-192 through 38 mm of steel gives a high-energy, low-contrast image, and the source size produces geometric unsharpness that further blurs a few-tenths-of-a-millimetre feature.
- Root geometry. The root profile itself creates a density band that can mask a small planar feature sitting in it.
Meeting 2.0% IQI sensitivity and 2.4 density proves the technique met the written requirement. It does not prove the technique was capable of detecting this defect type. That distinction is the heart of the case: code compliance in the method is not the same as detection capability for the mechanism present.
2. Does the earlier radiography acceptance stand, and what does the TOFD result mean for the joint?
Read the worked answer
The two results are not in conflict; they answer different questions.
The radiography acceptance stands as a statement that the joint met the specified radiographic acceptance criteria. It does not stand as a statement that the joint is free of rejectable planar defects, because the technique was not capable of demonstrating that.
What matters is the applicable acceptance standard. If the construction code or project specification requires ultrasonic examination of thick Cr-Mo girth welds in addition to radiography, the joint has simply not yet been shown to comply; the TOFD result is then the governing one, and a 30 mm long, 3.5 mm high root lack of fusion would normally exceed planar acceptance limits and require repair. If the contract specifies radiography only, the client still has a fitness-for-service question, because the service is high-temperature creep regime in 2.25Cr-1Mo: a sharp root flaw perpendicular to the primary stress is a creep crack initiation site, and the hydrotest does not clear it. Hydrostatic proof tests load for minutes at ambient; they say nothing about creep or fatigue over 200,000 hours.
My answer to the client: the radiographic acceptance is factually correct but technically insufficient, and the joint's acceptability must be re-decided against a fracture-mechanics assessment or the specified ultrasonic criteria, not against the film.
3. Set out the decision path you would recommend before any cut-out, including the examination and documentation you would require.
Read the worked answer
Cutting out a 38 mm Cr-Mo girth weld is expensive and carries its own risk, so the decision should be evidence-based and documented.
- Confirm the indication. Re-examine with phased array from both sides and both directions, and with TOFD from the opposite side if access permits, so the flaw is characterised in length, height, through-wall position and orientation. Single-sided, single-technique planar sizing is not enough to condemn a joint.
- Characterise the root. PT or MT of the root where accessible, plus a hardness survey across the HAZ and weld, to establish whether the joint also has a metallurgical problem, not just a geometric one.
- Establish the governing criteria. Retrieve the contract, the construction code edition at time of contract, and any supplementary UT requirements. Determine whether UT was required and omitted - that changes the discussion from fitness for service to a quality escape.
- Engineering critical assessment. If the criteria are fitness-for-service, run an assessment with the measured flaw size, real toughness, and the actual creep and fatigue duty, including start-up and shutdown cycles.
- Decide and record. Repair, monitor with a scheduled inspection interval, or accept with documented justification - and get the owner's written agreement, because the owner carries the consequence.
- Close the systemic gap. If the inspection plan did not require UT on thick Cr-Mo girth welds, revise the plan. This defect was predictable, and a review of the plan should have caught it before welding started.