Inspection and testing — and what the standard numbers actually mean
Chinese standard numbers are not a discount. On the tests that matter for a kettle, the Chinese standard and the ISO standard are the same document.
Standards, translated
| What | We test to | Which is | US equivalent |
|---|---|---|---|
| Weld ultrasonic testing | GB/T 11345-2023 | Identical (IDT) to ISO 17640:2018 | ASME BPVC V Art. 5 |
| Weld UT acceptance | GB/T 29712 | Modified from ISO 11666 | — |
| Plate ultrasonic testing | GB/T 2970 | Comparable to EN 10160 | ASTM A578 |
| Penetrant testing | JB/T 4730.5 | Comparable to ISO 3452 | ASME BPVC V Art. 6 |
| Weld quality classes | BG/T 1501-2015 Table 7 | — | — |
| Radiographic testing (optional) | GB/T 3323 | Comparable to ISO 17636 | ASME BPVC V Art. 2 |
For reference: ISO 11666 acceptance levels 2 and 3 correspond to quality levels B and C of ISO 5817. Reports are issued in English.
This mapping exists because “we test to GB/T 11345” means nothing to an inspector in Rotterdam, while “identical to ISO 17640:2018” means exactly what it says. Same test, same document, different number on the cover.
Inspection and test plan
| Stage | Method | Standard |
|---|---|---|
| Incoming plate | Ultrasonic, plate by plate | GB/T 2970 Class III |
| Incoming plate | Chemical re-verification | GB/T 222 |
| Incoming plate | Mechanical re-verification | GB/T 228.1 |
| Cutting | Material tracking — contract, heat number, grade and size carried onto every piece | BG/T 1501-2015 §5.2.1 |
| Forming | Warm forming temperature by infrared pyrometer; thinning ≤ 5 % | §5.3 |
| Fit-up | Platform flatness ≤ 1 mm/m²; local gap ≤ 5 mm over less than 20 % of seam length | §5.4 |
| Weld preparation | Back-gouge, grind, penetrant test where required | §5.5.3 |
| Structural welds | Ultrasonic testing | GB/T 11345-2023 (= ISO 17640:2018) |
| Weld metal | Silicon content | Si ≤ 0.03 % — Table 10, Grade 1 |
| Dimensional | Tolerance class B | Table 8 |
| Coating | Visual and dry film thickness | §6.3 |
Coverage is agreed with you before we start, and written into the inspection and test plan. Full-length testing of every structural weld and partial coverage are both legitimate specifications — what is not legitimate is leaving it vague and letting each side assume. Tell us what your inspector expects, or ask us what we propose, and it goes in the ITP where it can be checked.
Product grades
BG/T 1501-2015 defines two finished-product grades. Which one applies to your kettle is written into the order — the difference is not cosmetic, and it is worth knowing which one you are buying.
| Grade 1 | Grade 2 | |
|---|---|---|
| Plate | Si ≤ 0.02 %, P ≤ 0.015 %, S ≤ 0.010 % | Si ≤ 0.03 %, P ≤ 0.025 %, S ≤ 0.020 % |
| Plate UT | GB/T 2970 Class III | GB/T 2970 Class IV |
| Weld class | I, II | III |
| Weld metal Si | ≤ 0.03 % | ≤ 0.12 % |
| Dimensional tolerance | Class B | Class C |
| Geometric tolerance | Class E | Class F |
Witness it yourself
Certificates are easy to produce. An inspection you watched is not.
The plant is open for pre-shipment inspection. Send your own inspector, or appoint a third-party agency you already use, and we will schedule the ultrasonic testing so it can be witnessed — and send you the inspection and test plan in advance, so your inspector arrives knowing exactly what to look at.
That is a stronger check than anything we could send you after the fact, and it is the one we would want if we were buying.
What comes out of all this testing goes to you as paper: see the documentation package. For why the weld silicon line in the table above matters more than it looks, see the weld is the weakest link.
Source: BG/T 1501-2015 — §5, §6, Tables 7, 8, 10. GB/T 11345-2023 is an identical adoption (IDT) of ISO 17640:2018.