The Kiln Stopped Mid-Cycle: What Happens to the Batch Inside

The Kiln Stopped Mid-Cycle: What Happens to the Batch Inside

Summary

Power and gas are interrupted on the utility side, not the factory side, and a kiln cannot stop politely between cycles. What the control room decides in the first hour, what the ware carries afterwards, why a second fire is not a reset, and the clauses that keep the loss from landing entirely on the buyer.

The Kiln Stopped Mid-Cycle: What Happens to the Batch Inside

An order does not stop at a convenient point. When the power drops, or gas pressure falls below what the burners need, the kiln stops where it is: part-way through the cycle, where the body has changed but the glaze has not finished.

Products in this guide: Pastel Bunny Ceramic Mug 420ml · Cute Cartoon Design Porcelain Bowl

For the buyer the news arrives later, as a date that has moved. For the plant, the first hour is a set of decisions taken with incomplete information, and the cost falls on one setting of somebody's order. It may be yours.

See the full ceramic sourcing process map.

The interruption is rarely inside the factory

Ceramic firing is energy-intensive, and the vulnerability sits upstream of the gate. A grid under peak load curtails large users first. An interruptible gas contract is priced lower because the plant accepted that supply can be cut. Pressure falls at the end of a line shared with other plants. A substation fails, or a main is damaged by work in the street.

The kiln is an awkward load. A tunnel kiln loses a section of its length; a shuttle kiln is a batch vessel, so the whole load is hit at once and the kiln may be down for a day while it cools. Both are easier to stop than to restart, which is why short interruptions are absorbed quietly.

The risk is upstream, in where the plant buys heat and how the contract is written.

What the kiln does in the first hour

A kiln is not a light switch. Refractory, setters and ware hold heat, so temperature falls slowly: a short interruption can be survived, a long one cannot. What matters is not whether the burners relight but whether the ware spends enough time at maturing temperature. Holding a load at a lower temperature is not firing it.

Instrumentation decides much of the outcome. A plant that logs temperature against time knows which point the cycle reached and can argue with evidence. A plant reading a dial once a shift is guessing, usually in its own favour.

Three decisions in the control room

None of the options is free; each trades one batch against the plant's next two weeks.

  • Hold and resume. Viable only if supply returns inside a defined window and the cycle can be extended. The soak is lengthened, the batch stays a batch, and the failure mode is a slow drift nobody corrects.
  • Controlled cool. Burners off, dampers managed, cooling deliberately slow. The load is lost as production, but the lining, setters and furniture survive and thermal shock cracking is avoided.
  • Abandon and unload. Fastest route back to production for the next order, worst for the batch inside. Ware pulled hot can crack in the hand, and shelves may be damaged.

The choice depends on what the plant is measured on. A plant paid by the piece will try to hold; a plant with a large order behind yours may cut its losses and clear the kiln.

What a stopped cycle leaves in the ware

The batch is not simply underfired. The body may not have completed its transformation: higher porosity, lower strength, a tone off the approved body colour. The glaze may be immature, showing as patches of matte, pinholes, or an incomplete fusion that reads as dirt. Decoration fired at the wrong point shifts colour, rarely uniformly across a load.

Then the part buyers underestimate. Ware can leave an interrupted setting looking acceptable and fail later: crazing after a few wash cycles, a handle that lets go in service, a rim that chips at the first knock. The failure is latent, so the customer finds it, not the inspector.

Why a second fire is not a reset

Putting the load back in and finishing it is a reasonable instinct and a poor assumption. A second fire adds a second thermal history: gloss changes, most pigments move, and crazing risk rises because glaze and body no longer expand together.

The comparison problem is worse than the physical one. Your approved sample was fired once; refired production is judged against a once-fired reference, and the difference reads as a defect even when the piece is sound. The reference has to be explicit before the argument starts. The master sample guide explains how sealed references keep batches identical.

The delivery arithmetic nobody quotes

Lost days are not days lost. The constraint moves from the kiln to the queue behind it, allocated by contract, by relationship and by whoever shouts loudest.

Re-setting a load takes time of its own, committed orders keep their slots, packing and booking compress against the new date, and freight space in a peak month is not elastic. The date moves by the interruption plus the recovery, not by the interruption alone.

The useful question is not how long the outage lasted but where the order sits in the recovery sequence, who decides priority in the plant, and whether that answer exists in writing. The replenishment calendar maps the ordering windows for the whole year.

What to write before it happens

Each clause is cheap to agree in a calm week and expensive to negotiate during one.

  • A notification window. Any interruption affecting a setting is reported within a stated period, with kiln and setting identified. A short written notice, not a press release.
  • Lot separation. Interrupted ware is kept as its own lot, marked, and not blended into conforming production. Blending is the default in a busy plant because it is faster.
  • A concession ladder agreed in advance. Colour difference against the sealed reference, functional criteria, and the price or re-make consequence, settled while nobody is under pressure.
  • Re-fire disclosure. Any re-fired ware is declared, with the number of cycles. A buyer who does not ask is unlikely to be told.
  • Partial shipment. The right to take the conforming portion instead of waiting for the whole, so one setting does not hold a season.
  • Force majeure that excludes the foreseeable. An interruptible utility contract is a risk the plant accepted when it signed; planned maintenance is a choice. Neither is an act of God.

On the next visit, ask what the backup generator actually runs: the kiln blowers and controllers, or only the lights. Ask about a second fuel, and the last outage.

A plant that answers with specifics has lived through it. A plant that says it never happens is telling you the plan is improvisation, and that the first batch to find out will be someone's order.

An interruption is an ordinary hazard of an energy-intensive process. What separates suppliers is whether the lot is identified, whether the buyer is told while the decision still matters, and whether the standard was agreed before the kiln stopped.