A Weep at a Repair Toe During Hydrotest
A 1-1/2 in. thick SA-516 Gr. 70 knockout drum has been PWHT'd and is on hydrotest at 1.3 times MAWP. During the hold a weep appears at the toe of a 3 in. long weld repair in the longitudinal seam, and the gauge drops 15 psi in 10 minutes. The repair had been made after radiography showed a crack in the original SAW seam; it was ground flush, radiographed, and accepted. The superintendent proposes to grind the weep area, lay a cover pass with E7018 without preheat, skip PWHT because it is a small repair on a seam that was already heat treated, and re-test at 1.3 times for 10 minutes. The drum is scheduled to return to wet H2S service in 48 hours.
1. Identify what is wrong with the proposed plan, element by element.
Read the worked answer
The leak itself. A weep during hydrotest means a path through or nearly through the pressure boundary. Grinding the visible weep and covering it leaves the defect in place, and the defect almost certainly extends beyond what is visible on the surface.
No preheat. Welding 1.5 in. carbon steel without preheat, on a damp, previously tested surface, adds hydrogen to a high-restraint repair in thick section, which is how delayed cracking starts.
Skipping PWHT. There is no small-repair exemption. The original PWHT relieved the original weld; it does nothing for the new heat-affected zone, which will be hard and unrelieved. In wet H2S service that hard zone is a sulfide stress cracking site, and the vessel will be back in service in two days, which is precisely the exposure window that matters.
The retest. The 15 psi drop already constitutes a failed test. A short hold at test pressure after a patch is not the specified test, and the test must follow completion of all repairs and examination, not substitute for them.
2. Write the repair sequence you would require, from depressurization to return to service.
Read the worked answer
Depressurize, drain, clean, and confirm the area is dry; the vessel is not welded while wet. Map the leak with PT or MT and locate its extent. Excavate with a burr or grinding wheel, working beyond the visible indication by a margin, and check with PT at each stage until the defect is gone; use UT or a depth gauge to confirm the excavation does not breach the remaining wall below design plus corrosion allowance. Pre-heat the excavation and surrounding metal, then weld with a qualified WPS and qualified welders using properly stored E7018, controlling interpass temperature.
Examine the finished repair: PT or MT at minimum, plus RT or UT over the full repaired length, and a hardness survey of the weld and heat-affected zone. Postweld heat treat, locally or in a furnace, with a documented cycle. Re-test the vessel at the specified pressure and hold time with the AI present.
Before returning to service, establish the cause: the original crack and this leak may share a root cause, such as a repair made without adequate excavation or a hydrogen crack from the first repair. Fix that, or the third repair is already scheduled.
3. What administrative and qualification checks apply to this repair, and why do they matter as much as the welding itself?
Read the worked answer
The repair needs a qualified WPS and qualified welders for the actual conditions: thickness, position, process, and preheat. If the drum is in wet H2S service, the client specification may impose hardness limits and PWHT regardless of thickness, and those requirements apply to repair welds exactly as they did to the original seam. Confirm the filler metal and its storage, the welder's continuity, and that the repair is documented as a repair, not folded into the original weld record.
The added PWHT cycle also has a procedure consequence: check the aggregate time at temperature against the PQR under QW-407.2, and check whether UCS-85 test specimens still represent the finished vessel. If the vessel was impact-tested by exemption, confirm the exemption still holds.
Finally, the repair must be recorded on the appropriate report forms if the item is code-stamped, and the AI should be involved before excavation rather than after painting. The welding is a few hours of work; the documentation and heat treatment decisions are what determine whether the drum can legally and safely return to service.