The number that matters: corrosion rate
BG/T 1501-2015, Appendix A.2.3.4, states it plainly: follow the operating rules and the kettle corrosion rate can be held to 1.0–2.5 mm per year.
On a 50 mm wall, that is a straightforward calculation. What the standard is really saying is that kettle life is a controlled variable, not a property of the kettle — provided the kettle was built right in the first place.
So there are two questions, not one. Was the kettle built so that a good corrosion rate is achievable? And is the furnace and the operation keeping it there?
Most disappointing kettle lives are failures of the second question blamed on the first.
Question 1 — what the kettle has to get right
Four things are fixed before the kettle ships and cannot be corrected afterwards.
Plate chemistry. Si ≤ 0.03 %, and Si + P + Al ≤ 0.045 % combined. Silicon, phosphorus and aluminium make the Fe-Zn alloy layer non-adherent, so it sloughs off and attack continues instead of slowing. Carbon, counter-intuitively, should not be minimised — the standard advises against plate below 0.05 % C, because the wall needs strength to resist the static pressure of zinc at temperature.
Weld metal chemistry. Weld silicon can be thirty times plate silicon depending on the consumable. A kettle with low-silicon plate and high-silicon seams fails at the seams. Covered separately here.
Geometry. Above 2 m depth the standard calls for a full rounded corner, because a square wall-to-base weld sits exactly where hydrostatic pressure is highest, and the internal corner traps dross that then insulates and drives local overheating.
Residual stress. All forming done warm at 850–900 °C rather than cold. No cruciform plate joints. Parallel welds at least 0.5 m apart. Double-bevel joints welded alternately from both sides.
Question 2 — what the furnace and the operator have to get right
This is where the three-to-ten-year spread actually comes from.
Inner wall temperature is the master variable.
| Temperature | What happens |
|---|---|
| 445 – 460 °C | Normal bath operating range |
| Below 480 °C | Inner wall safe zone |
| Above 480 °C | Corrosion accelerates sharply |
| 490 – 510 °C | Peak corrosion band |
Inner wall temperature is hard to measure directly. The standard’s practical proxy: the difference between bath temperature and outer wall temperature normally runs 30–50 °C, tightening to 15–20 °C during steady balanced production.
Heat input density is what pushes the wall temperature up.
| Bath temperature | Permitted heat input | Critical hourly throughput incl. jigs |
|---|---|---|
| 445 °C | 31 kW/(m²·h) | 1/30 of zinc volume |
| 460 °C | 24 kW/(m²·h) | 1/40 of zinc volume |
Note the direction of that relationship. A lower bath temperature permits a higher heat input. If your bath is well sized, running cooler lets you run harder.
Furnace chamber temperature. Below 700 °C on oil or producer gas. Below 650 °C on natural gas or electric heating.
Throughput discipline. Hourly throughput at 1/30 to 1/40 of zinc volume, and — the standard is explicit about this — run it level. Chasing output by raising heat input inevitably shortens kettle life. Overloading is called out by name.
Logging. Temperatures recorded every 30 minutes. Not for bureaucracy: without a record you cannot see the trend that is quietly killing the kettle.
The first 48 hours are worth more than the first year
On first firing, the wall has no Fe-Zn alloy layer at all. This is the moment of highest risk of intergranular attack and cracking in the kettle’s entire life. At the same time the vessel is releasing manufacturing stress and the refractory is giving up moisture.
| Stage | Requirement |
|---|---|
| Ambient → 300 °C | ≤ 2 °C/h |
| 100 – 120 °C | Hold 6 – 8 h — drives moisture out of new refractory |
| 300 – 330 °C | Hold 24 h — distortion and residual stress release |
| 300 → 420 °C | ≤ 5 °C/h |
| Throughout | Wall inside-to-outside differential < 50 °C; wall-to-base < 100 °C |
And the loading rules that go with it: charge with pure zinc above 99.99 %, no lead in the base, ingots stacked tight against the wall (the kettle is side-heated), and a 100 mm gap left in the middle of the stack — zinc expands about three times as much as iron, and a tight charge will push on the wall as it heats.
A kettle rushed through this schedule will not reach its design life regardless of how it was built. This is the single largest avoidable loss in the industry, and it costs nothing to avoid. Full commissioning guide →
Design decisions that buy you years
Go deeper. Depth is normally 1.3–2 × width. Deeper means more thermal mass and less temperature swing when you charge; more wall area, so lower heat input per m²; and dross settling well below the work.
Size the zinc volume to throughput, not to the workpiece. Zinc volume should be 30–40 × maximum hourly throughput including jigs (20 × for wire). Under-size it and the only way to hold temperature is more heat input — straight into the danger band.
Protect the kettle from the burners. Baffle or heat shield at the nozzles; insulation from 100 mm above the dross line downwards, so the wall does not take direct radiant heat where it is most vulnerable.
Support it properly. Above 1.8 m depth, one adjustable support band; above 2.5 m, two. Support points at 1.5–2.0 m intervals. Adjustable, because the gap changes with temperature. Heat-resisting steel for anything inside the hot zone. Expansion clearance along the length.
Fit a zinc leak alarm. Escaping zinc shorts the detector, triggers the alarm and shuts down heating. Where site conditions allow, an emergency zinc pit at 25–30 % of total volume, conical for recovery.
So how long does a kettle last?
If the kettle was built to the chemistry and geometry above, and the furnace holds the inner wall below 480 °C at a heat input inside Table A.1 with level throughput and a proper commissioning — the corrosion rate stays at 1.0–2.5 mm/year, and you can work out your own answer from your wall thickness.
If any of those are missing, the number the supplier quoted you was never going to be the number you got.
What to ask before you buy
- What is the plate silicon limit, and the weld metal silicon limit?
- Which finished-product grade, Grade 1 or Grade 2?
- What NDT, to which standard, at what coverage?
- Will you supply the commissioning heat-up curve?
- What heat input density does your sizing assume?
That last one is the tell. A supplier who has sized your kettle against your throughput has done the engineering. A supplier who has quoted you a box has not.
Source: BG/T 1501-2015 — §3, §4.2, §5, Appendix A.1.2, A.1.5, A.1.6, A.2.1–A.2.4, Table A.1.