A test package “coming back” is not the same as a hydrotest “failing”

On site, two separate events are described in one sentence: "the package came back". One happens before water goes in (missing document, old isometric revision, open punch), the other under pressure (leak). Different causes, different costs, different denominators.

I looked: I could not find a combined rate for failed hydrotests or returned packages on B31.3 plant projects; other tools gave the same result. The verified example I found is a single case: two leaks in a newly installed 12-inch 304 stainless line at hydrotest, caused by microbiologically influenced corrosion (MIC), not a weld defect. The total number of joints tested is not stated, so no rate follows.

Another case: in Kuwait, in an underground composite (GRE) gas line, a joint near the valve pit and a tee leaked; the paper lists non-compliance with manufacturer recommendations, insufficient support in the pit, and a tee manufacturing defect, and the supply of the large-diameter tee delayed the project by 6 months (not metallic pipe).

At an NGL plant, with re-planning, 75.6% of packages were completed and approved by the time panels arrived; this is a completion rate, not a rejection rate.

Lesson: classifying "leak = bad weld" may be wrong; investigation is needed. The cost of a repeat test cannot be given as one number either; it has components: draining, repair, re-examination, refilling, re-mobilization, waiting on the critical path. You can price them from your own records.

Four numbers are enough to measure: packages submitted for readiness, returned at first submission, actually pressurized, leaked at first test. For each return and leak, the cause and the duration. Without these numbers, giving a percentage would mislead.

Sources: Mahajanam et al. (2022) failure analysis; Gunasekaran, NDE 2019 (CP-312); Ibrahim, SPE 2017 (OnePetro); West test package specification; Aramco/OGN.