API SIFE study materials
API SIFE study resources

Hydrostatic Test Performed With No Inspector Present

API SIFE · Put your knowledge to work in context.

All case studies

Hydrostatic Test Performed With No Inspector Present

You are the owner's source inspector for a shop building 14 identical ASME Section VIII Division 1 vessels (SA-516 Gr. 70, 1.0 in. shell, design 300 psi at 350 F). The approved inspection and test plan (ITP) lists the hydrostatic test as a Hold point requiring 48 hours' notice and your physical presence. Your surveillance budget covers six visits. The shop's QC manager has already signed off fit-up, PWHT chart review, and NDE. On Tuesday you learn that the shop hydro-tested vessels 5, 6, and 7 on Saturday while you were off site, and QC signed the hold as 'released - records reviewed.' The records show no leaks, gauges within calibration, and a 30-minute hold. Vessels 8 through 14 are still open and none have shipped. The purchase order makes the ITP a contractual document.

1. What is the status of the tests already performed on vessels 5, 6, and 7, and what is your first action?

Read the worked answer

Status. The hold point was not released by the party entitled to release it. A manufacturer cannot self-release a contractual hold; only the inspector (or the client in writing) can. The code requirement to test the vessel may have been met, but the contractual verification requirement was not, so the tests cannot be accepted as witnessed verification.

First actions, in order:

  1. Issue a written nonconformance report the same day, with the ITP clause cited, and notify the client's contract administrator and the shop QC manager simultaneously.
  2. Place a hold tag or equivalent administrative stop on vessels 5-7 so they are not shipped or stamped for release.
  3. Determine whether re-test is still feasible. Since the vessels are open, not lined, and not shipped, re-hydrotest under witness is normally the cleanest disposition.
  4. Establish root cause with QC: who scheduled Saturday work, who was notified, and why the release signature was treated as routine. Fix the notification workflow before vessels 8-14 are tested.
  5. Record in your surveillance log that the tests were not witnessed so no downstream reviewer assumes otherwise.

2. How would you restructure the quality surveillance plan for the seven remaining vessels on a six-visit budget?

Read the worked answer

Re-rank by risk, then buy coverage where it matters.

  • Group the remaining hydrotests into two campaigns (four vessels, then three) so one witnessed visit covers several units. This converts seven hold points into two.
  • Amend the ITP so hydro cannot be scheduled on weekends or outside agreed windows without written agreement, and extend notice to five working days.
  • Convert pure document reviews (MTR review, WPS/PQR checks, PWHT chart review) to desk review, since those are already partly done and do not need a site visit.
  • Keep first-of-kind fit-up and final dimensional as witness points for one vessel only, then sample.
  • Spend one visit on NDE and one on final documentation and nameplate before release.
  • If the arithmetic still does not cover the critical holds, request three additional visits or part-time resident coverage from the client with a short written justification; do not silently downgrade a hold to fit the budget.

Confirm the revised plan with the client in writing, since the ITP is contractual.

3. The shop argues that the records prove the test was satisfactory, so re-testing wastes money. How do you respond and what do you document?

Read the worked answer

The records are shop-generated and describe conditions the shop controlled. The purpose of the hold is independent observation of the variables that records capture poorly: gauge identity and range, actual metal temperature, venting, leakage at every joint, and the behavior of the vessel at hold pressure. A record set cannot demonstrate that the inspector saw those conditions.

Document:

  • The nonconformance, with dates and the specific ITP hold.
  • That hydrostatic testing for vessels 5-7 is unverified, and whether re-test or client concession is proposed.
  • If the client grants a concession, the compensating measures you recommend (verification of gauge calibration certificates, confirmation of water and metal temperature, post-test dimensional checks, review of the pressure chart, dye penetrant of weld toes at nozzles) and the residual risk accepted.
  • That you have not signed or initialed any record representing the tests as witnessed.

Re-testing three open vessels is inexpensive compared with the cost of a disputed release. Present it as a schedule fix, not a punishment.