Boiler returned to service after a tube replacement: hydro test, safety valves and three unexplained low-water trips
A 10 t/h saturated-steam boiler (design pressure 1.25 MPa, operating pressure 1.0 MPa) has been shut down for major repair after 14 water-wall tubes and two economizer tubes were replaced following a bulging failure. For three weeks before the failure the unit logged three low-water trips; each was reset by the shift operator within minutes without any investigation. The repair contractor ran a hydrostatic test at 1.25 MPa, held it briefly and reported the boiler 'qualified'. A weep appeared at one expanded tube joint and stopped by itself within a few minutes. Test water temperature was 12 °C and ambient temperature 8 °C. The two drum safety valves were not removed, re-set or re-sealed after the tube work; one was gagged during the test, and the gag has since been taken off. The owner wants steam by tonight and asks you to sign the inspection certificate.
1. Assess the hydrostatic test. Was it validly performed, and what is the significance of the weep that sealed itself?
Read the worked answer
Test pressure is wrong. The test pressure for a boiler is a multiple of design pressure, not the operating pressure: 1.25 × 1.25 MPa = 1.5625 MPa, i.e. about 1.56 MPa. Testing at 1.25 MPa only reproduces normal operating stress and demonstrates nothing about joint integrity at the required level. The test was therefore not a valid hydrostatic test and must be repeated at 1.56 MPa after the joint is corrected.
The weep is a failure, not a pass. At test pressure every expanded joint must be leaktight. A joint that weeps and then self-seals has been stressed beyond its seating and is not proven; the correct action is to re-expand or re-tube the joint, then repeat the full test. A leak that stops by itself is exactly the defect a hydro test exists to find, because it will reopen under thermal cycling.
Do not reject the test for the wrong reason. Water at 12 °C with ambient at 8 °C is acceptable: both are above the 5 °C floor used to avoid freezing and brittle behaviour. Making that the objection would miss the real defects and weaken the report.
The test record must also state the pressure gauge used (calibrated, correct range), the holding time, the inspection of every joint at test pressure and at design pressure, and the fact that the gag was fitted and removed.
2. What must be done with the safety valves before the boiler can be sealed and returned to service?
Read the worked answer
After pressure-part repair the safety valves are no longer in a proven state and must be treated as part of the repair:
- Remove both drum valves; overhaul and set them on a test bench by a qualified valve-testing agency, not by the repair crew on site.
- Establish the set pressures from the code table for the rated steam pressure, with at least one valve on the lower setting so that it opens first; the highest set pressure must not exceed the design pressure beyond the permitted tolerance. The set pressures must be justified in writing, not chosen to match what the valves happened to be at.
- Fit lead seals and record the set pressure, the test agency, and the certificate number in the inspection file.
- A gag may be used only to isolate a valve during the hydro test, it must be recorded, and it must be removed and the valve verified afterwards. The fact that a gag was fitted and removed without a record is itself a finding, because a gagged valve is a valve that cannot protect the boiler.
- After steam-up, confirm each valve lifts and reseats at operating pressure and that the drain and vent lines are clear.
Only when the valves are re-set, sealed and the steam test is satisfactory can the boiler be sealed. Signing the certificate before that would certify a boiler whose last line of overpressure protection has not been demonstrated.
3. The owner says the low-water trips are unrelated to the tube repair. How do you handle the return to service and what is your certification decision?
Read the worked answer
The three trips are not unrelated administration; they point at the feedwater and level-control system, and repeated manual reset without investigation is a serious safety-culture failure. The low-water cutoff is a safety interlock, not an alarm to be cleared.
Before any certificate:
- Verify the repair itself: the repair unit must hold the appropriate licence, the repair must be treated as a major repair, tube material and expansion records must be checked, and the repaired welds/joints examined as required.
- Functionally test the low-water cutoff by actually lowering the water level (blow-down of the level column or simulated signal), confirm the burner and fuel supply trip and lock out, and confirm manual reset is required.
- Inspect the water-level gauges and electrodes for fouling, check the feedwater control valve and pump capacity, and look for the root cause of the three trips (foaming, faulty level electrode, control valve sticking).
- Confirm the operators are instructed not to reset a trip without recording and reporting it; note the practice in the report.
Decision: withhold the certificate and issue a written defect notice. The boiler returns to service only after a valid hydro test, safety-valve re-setting, satisfactory interlock tests and closure of the low-water investigation. If the owner insists, record the refusal and report the situation through the supervisory channel; a boiler that trips on low water three times in three weeks is a boiler that will eventually be run dry.