Hydrostatic Test Certificate Under Schedule Pressure
During a hydrostatic test of a newly fabricated ASME Section VIII Division 1 vessel, the pump-discharge gauge reads 1.32 × MAWP, but the vessel-mounted test gauge at the top head reads 1.22 × MAWP. The vessel holds pressure, but a weep is visible at a threaded drain plug on the bottom head; the plug was sealed with PTFE tape and tightened by hand. The independent inspector has left site. The shop superintendent says the weep is normal for a threaded joint and asks you, as the shop's mechanical engineer, to sign the test certificate so the vessel can ship Friday. The client's specification requires a hydrotest with two calibrated gauges, both within current calibration, and no leakage. Your gauge calibration stickers expired three weeks ago. The vessel is for chlorine service; the client's QA representative will not return until Monday, and a shipping delay costs the shop a $40,000 penalty.
1. Identify the code, specification, and ethical issues in this situation. What can and cannot you certify?
Read the worked answer
The technical issues are straightforward and all point the same way:
- The two gauges disagree by about 8% of MAWP (1.32 versus 1.22). At least one is wrong, and the vessel-mounted gauge is the one that matters for the test pressure. The applicable code test pressure is normally 1.3 × MAWP (adjusted for stress ratios), so a reading of 1.22 × MAWP on the vessel gauge is below the required test pressure even if the pump gauge is accurate.
- Both gauges must be within current calibration under the client specification. Your gauge is three weeks out of calibration, so the test record is not traceable and the test is not acceptable as witnessed.
- A weep at a threaded drain plug is leakage. The client specification explicitly requires no leakage, and the code does not permit accepting a leaking joint as a successful hydrotest. PTFE tape and hand tightening are not a controlled sealing method for a pressure joint.
- The independent inspector is absent, so the required witness chain is broken.
Ethically, signing would be false certification: it would assert a test pressure, gauge calibration, leak-free condition, and witness status that are not true. For chlorine service, the consequence is not a paperwork problem; a leak in service could injure people and damage property.
What I can do is record the test as attempted, document the readings and the leak, and certify only what I personally observed. I cannot sign the test certificate as a passed hydrotest.
2. Set out the actions you would take and the communications you would make, in order.
Read the worked answer
Immediate actions:
- Stop the test and depressurise the vessel in a controlled way. Do not continue to hold pressure while the leak is unresolved.
- Mark the vessel as 'test incomplete - do not ship' and remove it from the shipping queue. Record the time and the reason.
- Photograph the weep, record both gauge readings, gauge serial numbers, calibration status, water temperature, and the time of each reading.
- Repair the drain plug properly: inspect the threads, use the specified sealant or a suitable plug, and tighten to a controlled torque. If the thread is damaged, rework it under the approved procedure rather than relying on tape.
- Replace or recalibrate both gauges and obtain current calibration certificates.
- Arrange a witnessed retest with the independent inspector or an acceptable alternate, and notify the client QA representative of the nonconformance and the proposed retest date.
Communications: tell the superintendent in plain terms that the test has not passed and cannot be certified, and why. Offer a recovery plan - expedited calibration, plug repair, weekend retest, remote or alternate witness - and put the plan in writing. Inform the client's QA function proactively and issue a nonconformance report. Do not accept a verbal instruction to sign; if pressed, escalate to the quality manager or engineering manager and, if necessary, to the authorized inspector.
A useful distinction to state clearly: I can write a factual record that says 'test attempted, leak observed, gauges out of calibration, retest required.' I cannot write 'test passed.'
3. Discuss the pressure to sign, including the cost argument, and explain how you would handle a supervisor who insists.
Read the worked answer
The $40,000 penalty is real, but it is not comparable to the consequences of a false certificate on a chlorine-service pressure vessel. A chlorine release can cause acute injury, evacuation, and regulatory investigation; the personal and corporate liability from a falsified test record would dwarf the penalty. The cost argument also assumes the only options are 'sign' or 'delay until Monday', which is false: calibration can often be expedited, a plug repair is quick, and a retest can be witnessed by an acceptable alternate. The schedule problem is solvable without falsifying the record.
Handle the supervisor by separating the technical facts from the commercial pressure. State: the vessel gauge read 1.22 × MAWP, the specification requires 1.3 × MAWP; the plug leaked; the gauges were out of calibration; the inspector was absent. Then state the options and their consequences. Offer to own the recovery plan and to communicate with the client. If the supervisor continues to insist, refuse clearly and in writing, and escalate through the quality organisation and the authorized inspector. Keep the refusal professional and factual rather than accusatory.
Professional codes and the law support this position: engineers are expected to hold paramount public safety, to sign only for work they are competent to and did verify, and to avoid deceptive statements. Many jurisdictions also protect employees who refuse to engage in falsification. The right answer is not to negotiate the test result but to fix the joint, recalibrate, and retest - and to make sure the record shows exactly what happened.