Hydrogen Cracking Risk on a Thick-Section Repair at Low Ambient Temperature
A 50 mm thick low-alloy steel vessel nozzle weld requires repair of a 60 mm slag inclusion found by UT. The repair will be done by SMAW with a matching consumable. Ambient temperature in the shop is 2 degrees C, and the shop has no heated enclosure. The original WPS specifies a preheat of 100 degrees C and an interpass maximum of 250 degrees C but does not address ambient conditions. The welder proposes to start immediately with a torch preheat of the local area. The client's engineer is offsite and has not responded to the repair request. Production pressure is high.
1. Should the coordinator allow the repair to proceed with torch preheat alone, and what is the technical basis?
Read the worked answer
No. Torch preheat of the local area is not adequate for a 50 mm low-alloy steel repair at 2 degrees C ambient, and the coordinator should hold the work.
The technical basis is hydrogen-assisted cracking, which depends on four interacting factors: hydrogen content, restraint, hardness (cooling rate), and section thickness. A 50 mm section is thick enough to generate high restraint and to conduct heat away rapidly. At 2 degrees C ambient, the cooling rate after welding is much faster than the WPS assumed, which promotes hard martensite in the heat-affected zone. A local torch preheat is difficult to control and verify; it may raise the surface temperature while the through-thickness temperature remains low, and it does nothing to control cool-down rate after the pass.
The WPS preheat of 100 degrees C was written for normal shop conditions. It does not address the ambient, and the coordinator cannot assume it covers a 2 degrees C environment. The correct action is to require a heated enclosure or a controlled preheat that brings the joint to the specified preheat through the thickness, verified by contact thermocouple or temperature-indicating crayon at multiple points, plus a controlled cool-down (for example, slow cooling under insulation) after welding. The coordinator should also confirm the consumable is properly baked and handled to keep hydrogen low.
Proceeding on production pressure alone would be a judgment failure. The coordinator should document the hold, notify the client engineer, and propose the controlled repair method for approval.
2. What does the coordinator need from the client engineer before releasing the repair, and how should the hold be documented?
Read the worked answer
The coordinator needs a written, specific approval that addresses the ambient condition, not a general 'proceed with repair' message.
Required approvals and information:
- A revised or supplemental repair procedure that specifies the preheat method, the minimum through-thickness preheat, the verification method (thermocouple placement and readings), the interpass control, and the post-weld cool-down or post-weld heat treatment requirement.
- Confirmation of whether PWHT is required for this repair and, if so, the parameters. Thick low-alloy repairs often need PWHT to temper the HAZ and reduce residual stress; the coordinator should not assume it is waived.
- Consumable control requirements: baking, holding oven, and exposure limits.
- Acceptance criteria for the re-inspection (UT after a defined cooling/hold period, not immediately, to allow any delayed cracking to appear).
Documentation of the hold: the coordinator should issue a hold tag or stop-work notice referencing the repair number, record the ambient temperature and time, state the reason (preheat/ambient not covered by WPS), list the required approvals, and log the client notification with timestamps. The hold should remain until the revised procedure is approved and the preheat equipment is in place. The coordinator should also record that the welder proposed torch preheat alone so the decision trail shows why it was rejected.
3. After the repair is completed under controlled conditions, what inspection timing and follow-up should the coordinator require?
Read the worked answer
The coordinator should not treat a single post-repair UT as sufficient, because hydrogen cracking can be delayed.
Timing: require the inspection to occur after the weld has cooled to ambient and after a defined hold period. A common practice is to wait at least 24 to 48 hours after welding before final UT or surface inspection on thick low-alloy repairs, because delayed cracking may not be present immediately. The exact hold should be specified in the approved procedure. If PWHT is required, final inspection should follow PWHT, since PWHT can also open or propagate defects.
Inspection scope: UT of the repaired area plus a margin around it to catch cracking that extends beyond the repair; surface inspection (PT or MT) of the cap and, where accessible, the root; and a check of the HAZ for hardness if the procedure requires it. The coordinator should confirm the UT technique is appropriate for the repair geometry and that the technician is qualified for the thickness and configuration.
Follow-up: compare the post-repair indication map to the pre-repair map to confirm the slag inclusion was removed and no new indications appeared. Record ambient conditions and preheat readings during the repair so the inspection result can be tied to the process. If any indication is found, the coordinator should escalate before any further welding, and the repair record should be closed only when the inspection is clean and the client engineer has accepted the result.