Section I Shop Hydro Leak at Tube-to-Tubesheet Weld
During a shop hydrostatic test of a new 60,000 lb/hr, 350 psig saturated-steam packaged boiler built to ASME Section I, the AI observes a weep at a tube-to-tubesheet weld on the front wall after the pressure reaches 525 psig. The tube is 2 in. OD, 0.165 in. wall, SA-178 Gr C; the tubesheet is 1.25 in. SA-516 Gr 70. The manufacturer wants to grind out the leak, reweld the same joint with the same WPS and welder used for production, re-hydro at 525 psig after lunch, and stamp the boiler that afternoon. The shop traveler shows the original weld was made with a qualified WPS and welder, and the tube-to-tubesheet joint was not radiographed. The AI has not yet signed the Manufacturer's Data Report. The owner's representative asks whether the boiler can ship on schedule.
1. Can the Manufacturer make this weld repair before the boiler is stamped? What must the AI verify before accepting the repair?
Read the worked answer
Yes, a construction defect found before stamping may be repaired by the Manufacturer, but the AI must be satisfied that the repair restores Code compliance. The Manufacturer remains responsible for the work; this is not an NBIC in-service repair and no R-symbol is involved. The AI should verify: (1) the leak is actually in the weld and not in the tube wall or an adjacent ligament; (2) the defect is removed to sound metal, with the excavation shape and depth acceptable to the WPS and with MT or PT confirming no remaining crack; (3) the repair WPS is supported by a PQR and the welder is qualified for the process, position, and thickness; (4) the repair does not create an unacceptable reduction in tube wall or tubesheet ligament; and (5) the joint is re-examined by the method required for the original joint, recognizing that this joint was not radiographed. Because the repair is by welding after the original hydro, a new hydrostatic test is required. The AI should also require the repair to be recorded in the shop records supporting the Manufacturer's Data Report.
2. The shop plans to re-hydro at 525 psig with no AI present and send a photo of the gauge to the AI. Should the AI accept that plan?
Read the worked answer
No. The AI's acceptance of the hydrostatic test is a required inspection hold point, not a documentation review only. The first test was witnessed at 525 psig, which is 1.5 times the 350 psig MAWP, consistent with Section I hydrostatic testing. After a weld repair, the boiler must be re-tested at the same required pressure, and the AI must witness the test and examine all joints and connections for leaks. A photograph of a gauge does not prove that the required pressure was held, that the test was performed under controlled conditions, or that all joints were examined. The AI should refuse to sign the Manufacturer's Data Report until a witnessed re-test is satisfactorily completed. If the shop wants to avoid a second full test, it may propose a Code-acceptable alternative only if Section I expressly permits it for the specific repair; a tube-to-tubesheet weld leak repaired by welding does not fit such an exception.
3. After grinding, the shop finds that the weep came from a pinhole in the tube wall about 1 in. from the weld, not from the weld itself. How does that change the AI's decision?
Read the worked answer
The location changes the repair scope and the examination needed. A pinhole in the tube wall is a tube defect, not a weld defect; grinding the weld would not remove the cause. The AI should require the tube to be replaced or the defective length cut out and replaced in accordance with the applicable Section I tube requirements, including proper tube material, expansion or welding details, and examination of the replacement joint. The replacement is still a construction repair by the Manufacturer before stamping, so the WPS, welder qualification, and any required NDE must be acceptable. A new hydrostatic test at 525 psig is required after the tube work. The AI should also ask why a new SA-178 Gr C tube developed a pinhole before service; if the tube was damaged during installation or if other tubes are affected, additional examination may be needed. The original weld repair plan cannot be accepted as written because it did not identify the actual defect.