Ask a plant manager how long a changeover takes and you will usually get the clean answer: the time between the last biscuit of one product and the first biscuit of the next.

Ask the finance team what a changeover costs and you will get a longer silence, because almost nobody counts it properly.

How long does a changeover between biscuit types take?

A product changeover within one dough family, a die change and a recipe recall, takes under an hour on a well-specified line. A changeover between hard dough and soft dough production is a different operation, because the forming section has to be rebuilt, and on a conventional plant it can occupy much of a shift.

The four costs, in order of how often they are missed

  • The obvious one: production hours. The line is not making product. At 1,000 kilograms an hour, four hours of changeover is four tonnes that were never made. This is the number everyone counts.
  • The one people forget: the run-up and run-down. Product either side of the changeover is out of specification and goes to waste or to rework. On a long tunnel oven this is more material than people expect, because the oven does not empty instantly.
  • The one nobody counts: what you did not schedule. This is the largest cost and it is invisible, because it never appears as a loss. It appears as a decision not taken. A plant that loses most of a shift per switch stops switching. Short runs of a second product stop being worth scheduling. A promising variant never gets the trial run. A seasonal line is judged too expensive to attempt. The line has quietly narrowed what the business is allowed to sell.
  • The one that arrives later: labour. Rebuilding a forming section is skilled work, done under time pressure, and skilled operators are harder to find and harder to keep than at any point in the last three decades. A plant whose changeover depends on two people who know the machine has a risk it has not written down anywhere.

Do the arithmetic for your own plant

Take your changeover frequency across a year. Multiply by the hours lost each time, including run-up and run-down. Multiply by your saleable output per hour and your margin per kilogram.

Then, separately, write down the products you chose not to run because of what the switch would cost. That second list is usually worth more than the first number.

Why the conventional answer is two lines

Faced with those costs, most plants that want both crackers and cookies buy two lines. It works, and for a large plant with sustained volume in both families it is often the right answer.

It is also expensive in ways beyond the second quotation. Two footprints. Two sets of utilities. Two spares inventories. Two maintenance routines. Two lines to keep loaded, which means two demand forecasts that both have to be right.

For a mid-sized plant, the second line is frequently under-loaded for its first three years. That is capital working at half speed.

The third option

A flexi line carries both forming arrangements on the line together, in registration, so moving between hard dough and soft dough production is a short planned operation rather than a rebuild. On our lines it takes roughly fifteen minutes.

Sunbeam commissioned its first flexi-line forming section in 1997, alongside India's first three-drive moulder. It was not a market that was asking for it at the time. It came from watching plants organise their entire production plan around avoiding a switch they could not afford to make.

What changes when a changeover is fifteen minutes

The arithmetic above inverts.

Production hours lost per switch become negligible against the value of the run that follows. Run-up and run-down still cost something, but they stop being the dominant term. And the invisible cost, the runs you never scheduled, disappears entirely, because a two-hour run of a second product becomes a perfectly reasonable thing to plan.

That is the real argument, and it is a commercial one rather than a technical one. A plant that can switch cheaply can chase demand. A plant that cannot must forecast it, months ahead, and live with being wrong.

What to ask a supplier

Three questions, and they are worth asking of any supplier including us.

  • What exactly happens during the changeover, step by step, and who does it. A number without a procedure behind it is an estimate.
  • Has it been timed on a plant floor, with product either side. Workshop timings and production timings are different numbers.
  • What happens to registration when you switch back. A changeover that is fast but needs an hour of dialling in afterwards is not a fifteen-minute changeover.

We would rather a customer asked those questions and watched a changeover filmed end to end with a clock running than took the figure on trust. It is a claim that should be provable, and it is.