Crack in a batch digester cone weld after a failed repair
Batch digester No. 4 is a 40-year-old, 22-mm-thick SA-285-C shell with a 3-mm stainless overlay on the top cone. Design is 1,250 kPa(g) at 175 C; it cycles six times per day between 25 C and 170 C. A wet fluorescent magnetic-particle inspection of the cone-to-shell knuckle weld found a 140-mm-long, 2.5-mm-deep surface crack in the heat-affected zone. The shell has lost 3 mm to general corrosion, leaving 19 mm. The mill has a 10-day outage window and wants to grind out the crack and weld repair in place without post-weld heat treatment because the local PWHT equipment is committed elsewhere. The crack is oriented parallel to the weld. The mill's last similar repair leaked after 14 months. Production is 2,000 adt/d and the digester is the only one that can cook the specialty grade.
1. What damage mechanisms are plausible, and what does the crack orientation tell you?
Read the worked answer
The crack is in the HAZ, parallel to the weld, in a cyclic batch digester. That orientation is consistent with caustic stress-corrosion cracking along the HAZ or with fatigue driven by the knuckle discontinuity and thermal cycles. The previous repair leaked after 14 months, so the root cause was not removed: residual stress, caustic concentration, or a crack left at the tip. A parallel-to-weld crack is not a simple hoop-stress axial crack; it points to HAZ microstructure and residual stress. Before welding, do a fitness-for-service assessment. A 2.5-mm crack in 19 mm is 13% of wall, but it is crack-like and must be treated as a growing defect, not as uniform metal loss.
2. Would you accept the proposed in-place weld repair without post-weld heat treatment? Justify your answer.
Read the worked answer
Not as proposed. ASME Section VIII may not require PWHT for 22-mm SA-285-C, so the code alone is not the issue; the service is. Caustic SCC and fatigue are controlled by residual stress and microstructure. Welding without PWHT leaves high HAZ stress and can re-crack, and the failed prior repair is evidence. Use local electrical PWHT at 595-650 C, hold about one hour per 25 mm, with a heated band at least three times the wall thickness, or use a qualified temper-bead procedure. Otherwise do not return to full service. Grind out the crack and confirm removal by PT and MT before welding.
3. If repair proceeds, outline the procedure and acceptance tests.
Read the worked answer
Remove the crack to sound metal with a 1:3 taper, and verify by PT and MT. Preheat 100-150 C, weld with a qualified WPS and low-hydrogen electrodes under controlled heat input, then perform local PWHT and cool slowly. After PWHT, MT and PT the weld, check hardness below 200 HB, and hydrotest at 1.3 times MAWP. Check the stainless overlay for disbonding. If PWHT is refused, derate the vessel or change the service; do not rely on a short inspection interval alone.