Why the first firing is the most dangerous moment in the kettle’s life
Three things are true simultaneously on day one, and never again.
The wall has no Fe-Zn alloy layer. In service, a thin adherent alloy layer forms on the inner wall and slows further attack. On first firing there is bare iron in contact with molten zinc. BG/T 1501-2015 Appendix A.2.1.1 is explicit: the risk of intergranular attack initiating cracks is at its highest here.
The kettle is full of manufacturing stress. Cutting, forming and welding all leave residual stress. If the kettle did not receive a full stress-relief anneal, that stress is still in it — and it will come out during the first heat, one way or another. Controlled, it relaxes. Uncontrolled, it distorts the vessel.
The refractory is wet. New furnace linings hold moisture. Heat it too fast and that moisture turns to steam inside the lining.
The schedule
| Stage | Rate or hold | Why |
|---|---|---|
| Ambient → 300 °C | ≤ 2 °C/h | Limits thermal gradient through the wall |
| 100 – 120 °C | Hold 6 – 8 h | Drives moisture out of new refractory |
| 300 – 330 °C | Hold 24 h | Kettle distortion and residual stress release happen here |
| 300 → 420 °C | ≤ 5 °C/h | |
| ~419.5 °C | Zinc melts — natural hold | Top up ingot to working level |
| → working temperature | Per process requirement |
Two limits apply throughout, and they are the ones to instrument for:
- Wall inside-to-outside differential < 50 °C
- Wall-to-base differential < 100 °C
From ambient to 300 °C at 2 °C/h is about six days. Plus the 24-hour hold. Plus the climb to 420 °C. A proper commissioning is a week-long operation, and every hour of it is buying years at the other end.
Why the 24-hour hold is the part not to cut
Everything else in the schedule is a rate limit — go slower and you are safe. The 300–330 °C hold is different: it is a soak, and its value is entirely in its duration.
This is the temperature band where the steel is hot enough for residual stress to relax, but the kettle is not yet carrying the load of molten zinc. Twenty-four hours here is the kettle settling into the shape it will keep.
Cut it to eight hours and the stress does not disappear. It comes out later — under the hydrostatic load of a full bath, at working temperature, as distortion.
For kettles that did not receive a full stress-relief anneal at the works, this hold is not just beneficial, it is the substitute for it. BG/T 1501-2015 §5.6.6 requires the manufacturer to supply a heat treatment curve in that case. If your supplier did not send one, ask why.
Loading the kettle
Pure zinc only, above 99.99 %. No lead in the base. With bare iron on the inner wall, purity matters more on the first charge than it ever will again.
Stack tight against the wall. The kettle is side-heated. Heat transfers into the charge through wall contact — ingots stacked in the middle of an empty kettle heat slowly and unevenly.
Leave a 100 mm gap in the middle of the stack. This is the rule that gets missed. Zinc’s coefficient of linear expansion is about three times that of iron. A tightly packed charge with no relief expands against the kettle wall as it heats, and pushes outward on a vessel that is simultaneously trying to release its own stress. Timber battens can steady the stack — they carbonise and float off once the zinc melts.
Cover the bath during melt-down. A lid, or rock wool blanket on steel supports, to limit heat loss from the top while the kettle is being heated from the sides.
Instrumentation before you start
| Thermocouples | Type E or K, in protective sheaths of a thermally conductive, zinc-resistant material |
| Bath measurement | At both ends of the kettle |
| Chamber measurement | Multiple points along the kettle length, including at the flue inlet |
| Outer wall | Additional points — the practical proxy for inner wall temperature |
| Calibration | Thermocouples and instruments verified on a schedule |
| Over-temperature alarm | Fitted and tested |
| Zinc leak alarm | Fitted and tested before firing |
| Logging | Every 30 minutes, recorded |
And before any of that: what the kettle sat on while it waited
If the kettle was delivered ahead of the furnace being ready: coat it against rust; block it clear of the ground, one end higher than the other; cover it; never allow water to collect inside; do not store with acids or alkalis.
Before installation, clean off loose rust and scale mechanically — without roughening the surface. Lifting lugs welded to the top edge for installation get cut off afterwards.
The base and the supports
The foundation is reinforced concrete with a refractory insulating layer, level to ≤ 1.5 ‰. The pit needs waterproofing or a sealed steel outer trough — a leak into a wet pit is a different order of incident than a leak into a dry one.
Furnace wall and kettle top edge both insulated; outer surface not above 55 °C.
Above 1.8 m depth (2.2 m for electric heating), one support band; above 2.5 m, two. Support points at 1.5–2.0 m. Adjustable, so the gap stays even as things move. Heat-resisting steel for anything in the hot zone. Expansion clearance along the length.
The economics
A week of controlled commissioning against a kettle that reaches its design life instead of half of it. On a kettle costing several hundred thousand dollars, plus the plant downtime of an unplanned changeover, the arithmetic is not close.
We supply the curve with every kettle, sized to that kettle’s dimensions and wall thickness, and we will go through it with your team before you light the burners.
Source: BG/T 1501-2015 — §5.6, §6.5, Appendix A.1.2–A.1.7, A.2.1, A.2.2.