A Customer Specification Stricter Than the Construction Code

The owner's inspector is right, and the fabricator has confused a code minimum with a contract requirement.

The construction code sets a floor. UCS-66 provides an exemption route, and if the contract were silent, the fabricator's reading would be defensible. But the customer specification is part of the contract, and it removes the exemption explicitly and consciously: 'irrespective of code exemption'. Once that sentence is in the contract, impact testing is required by the construction documents, and the code's exemption analysis becomes irrelevant to whether tests are performed. It still matters for the material purchase specification, but not for the welding qualification decision.

The correct framing is that this is not a code interpretation dispute at all. It is a contract compliance question, and the contract is unambiguous. If the fabricator believes the requirement is unnecessary, the route is a documented deviation request with technical justification, accepted in writing by the owner before welding - not a unilateral decision on the shop floor after the PQR was written.

Impact testing changes the qualification basis, not just the test list.

When impact testing is required, the supplementary essential variables become applicable - the tables covering base metal thickness and P-number, filler metal F-number and A-number, position, preheat and interpass, PWHT, electrical characteristics and technique. A PQR qualified without impact testing is a PQR qualified against essential variables only; it has never been tested against the supplementary set, so the fabricator cannot assert that its variables fall inside the qualified range.

Practical route:

1. Compare the existing PQR against the supplementary essential variables and confirm that the variables proposed for production match it exactly. Any difference - preheat range, interpass, PWHT condition, filler A-number, thickness - invalidates the extension.
2. Qualify a new test coupon or supplement the PQR with impact specimens. The coupon must be welded with the same essential and supplementary variables as production, using production consumables from the same classification.
3. Specimen extraction and testing per the applicable QW-462 figures, with notches in the weld metal and in the HAZ, tested at −45 °C, three specimens minimum, 27 J average with no individual value below the code minimum.
4. Document the results on the PQR and reissue the WPS with impact testing noted as a qualification basis.

Any WPS issued on the old PQR while impacts are required is unsupported. That is a QA finding, not a technicality.

First establish the facts: which joints were welded, under which WPS, before or after the contract requirement was known, and whether the WPS used was supported by a PQR qualified against the supplementary variables.

If the WPS was not so supported, the welds are not qualified welds under the contract. That does not automatically mean they are defective - it means their acceptability is unproven. Options, in order of preference:

• Retrospective qualification with a coupon welded under the same conditions, if the production variables can be shown to be inside the newly qualified range and traceability to the actual heats, consumables and parameters is intact. This is legitimate if the records are real.
• Production impact testing of run-off coupons or trepanned samples where the code and specification permit, as evidence for the specific joints.
• Cut-out where neither route is available.

The resolution structure: a written nonconformance identifying the contractual basis; a technical justification; the chosen route with its evidence; owner acceptance in writing; and a corrective action that puts contract review before WPS selection on the next job. The root cause is that someone read the code and never read the specification - a contract-review failure, not a welding failure.