Discontinuity Indications After Re-Radiography of a Restrained Butt Weld

The coordinator should first separate the two indications by type, location, and orientation, because each drives a different investigation.

1. Transverse 12 mm linear indication in the cap. Transverse-to-weld orientation is the key discriminator. It is not the classic lack-of-fusion or slag line that runs along the weld axis. A transverse crack-like indication in the cap most often points to one of three mechanisms: (a) a transverse crack from high restraint plus low ambient temperature and fast cooling, (b) a crater or stop/start crack from a restart point, or (c) a repair-induced crack from the arc-strike repair itself, especially if the repair was not preheated or was ground too aggressively. At 4 degrees C with a fully restrained 25 mm section, the shrinkage stress is high and the hydrogen/cooling risk is real. The coordinator should treat this as planar and potentially crack-like until proven otherwise, which is more severe than a volumetric slag line of the same length.

2. 3 mm rounded root indication. Rounded indications are usually volumetric (porosity, a small slag pocket, or incomplete root penetration with a rounded profile). At 3 mm in a 25 mm wall it is likely acceptable under most acceptance criteria, but it must be sized and compared to the applicable code limit rather than dismissed.

Priority: the transverse cap indication governs. The correct immediate step is not automatic cut-out; it is to characterize the indication (UT or phased array to confirm whether it is planar and to measure through-wall height), review the repair record for preheat and interpass temperature, and check whether the indication aligns with a weld restart. Cutting out a weld on suspicion alone is expensive and may not be required if the indication is a shallow, acceptable volumetric feature. The coordinator should escalate to the welding engineer and the client with a documented rationale, not make a unilateral cut-out decision.

The coordinator has films, the WPS, and the repair record, but is missing several facts that change the disposition:

• Travel direction and restart locations. If the transverse indication sits at a stop/start, the cause is likely a crater crack, which is a workmanship issue with a defined repair path. If it is mid-pass away from any restart, restraint cracking is more likely and the whole weld is suspect.
• Preheat and interpass records for the repair. The repair record should show whether preheat was applied. At 4 degrees C, a 25 mm carbon steel repair without preheat raises hydrogen-assisted cracking risk sharply. A missing preheat entry is itself a finding.
• Whether the 40 mm arc strike was ground out and PT/MT checked before repair. An arc strike is a hardened, crack-prone zone; if it was welded over without removal and inspection, the repair may have buried a defect.
• Actual chemistry and carbon equivalent. Higher CE widens the preheat requirement and the cracking window.

The 4 degrees C condition matters in three ways: it lowers the preheat margin (the WPS preheat may be inadequate for the ambient), it accelerates cooling and promotes hard martensite in the cap, and it increases the chance that the transverse indication is a hydrogen crack rather than a benign feature. The coordinator should require a documented preheat verification and consider whether the repair should have been held at a minimum preheat above ambient plus a controlled cool-down. Before any further welding, the coordinator should also confirm the pipe is not restrained in a way that concentrates stress at the repair location.

The coordinator should present a decision tree tied to evidence, not to the indication's existence alone.

Acceptance with no repair is justified only if supplemental UT/PAUT shows the transverse indication is volumetric or shallow (below the code height limit), it is not crack-like, and the root indication is within the rounded-indication acceptance limit. In that case the coordinator documents the sizing, the code clause, and the client concurrence.

Localized repair is justified if the transverse indication is confirmed planar but confined to the cap region and the surrounding weld is sound. The repair plan must include: removal by grinding to sound metal verified by PT, re-inspection of the excavation, preheat appropriate to the section and ambient (and a minimum preheat above the 4 degrees C ambient), controlled interpass, and re-radiography of the repaired area plus a margin. The coordinator must also verify the welder's repair qualification.

Cut-out of the joint is justified if: the transverse indication is a through-thickness or multi-pass crack, it extends into the root, it is associated with a buried arc-strike defect, or the repair record shows no preheat at low ambient with high restraint. Cut-out is also the right call if UT cannot resolve whether the feature is planar, because an unresolved transverse crack in a restrained 25 mm weld is a structural risk that monitoring cannot mitigate.

Monitoring is generally not appropriate for a transverse crack-like indication in a pressure boundary; it is defensible only for a confirmed volumetric feature that is within limits and where the client accepts a documented fitness-for-service basis. The coordinator's job is to make the evidence thresholds explicit and route the decision to the welding engineer and client with the supporting data.