Almost every plant that regrets its line regrets the size of it. Rarely the brand, rarely the specification, almost always the number.

The mistake is nearly always the same. Capacity gets chosen from a sales figure the plant hopes to reach, or from the rated output on a quotation, rather than from what the plant actually has to ship, shift by shift, on a normal week.

What capacity biscuit line do I need?

Set capacity from saleable output per shift rather than peak rate. Take your monthly tonnage, divide by working days and shifts, then add headroom for changeovers, cleaning and the yield you lose to startup and shutdown. A plant selling twelve tonnes a day across two shifts needs roughly 1,000 kilograms an hour, not 750.

Rated capacity is not production capacity

A line rated at 1,000 kilograms an hour will not give you 8,000 kilograms across an eight-hour shift. Nobody's does.

Between the two figures sit the things a rating does not include. Startup, while the oven comes to temperature and the first product goes to waste. Product changeovers, if you run more than one variant. Cleaning, whether scheduled or forced. Short stops, which are individually trivial and collectively the largest loss in most plants. Shutdown at the end of the run.

A realistic planning figure for a well-run plant is seventy to eighty-five percent of rated capacity across a shift. If you size the line on the rated figure, you have already sized it short.

Work backwards from the market, not forwards from the machine

The method that works is simple and uncomfortable, because it requires honesty about demand.

  • Start with monthly saleable tonnage. What do you actually ship. Not what you would like to ship.
  • Divide into shifts. Working days a month, shifts a day. Two shifts is the usual planning assumption. A third shift looks like free capacity on a spreadsheet and rarely survives contact with a maintenance schedule.
  • Add the losses. Changeovers, cleaning, startup and shutdown. If you run three variants a week, this is not a rounding error.
  • Then add growth headroom, carefully. This is where judgement enters. A line is a ten-year asset and demand does grow, but headroom bought today is capital sitting idle until demand arrives.

The case for buying slightly larger

Two arguments, both real.

The first is that width is the cheapest capacity available. A wider line delivers more output in the same footprint, drawing broadly the same overheads, using the same operators. Moving from 1,220 mm to 1,525 mm, for example, adds roughly a quarter more output without adding a building, a supervisor or a second maintenance routine. We cover this in detail in the case for a wider line.

The second is that capacity added later is always more expensive than capacity bought now. A second line means a second footprint, a second set of utilities, a second spares inventory and a second commissioning. Extending a line that was specified with room to grow means changing a section.

The case against buying much larger

Also real, and less often argued.

A line running well below its design capacity does not run efficiently. Ovens are sized for a throughput, and a half-loaded oven burns fuel it is not converting into product. Operators managing a line that never stretches lose the habits that keep it tuned. And capital tied up in idle capacity is capital not spent on the product development, distribution or packaging that would actually create the demand you are building for.

The practical answer for most plants is one step up, not two.

What the numbers look like on a floor

  • 250 kilograms an hour. A working industrial line rather than a pilot unit. Typically the right size for a regional brand moving off semi-automatic equipment for the first time, or for a larger plant adding a specialist variant it wants to keep separate.
  • 1,000 kilograms an hour. The most common band we build into. It suits a brand with established regional distribution, running two shifts, with a small portfolio of variants.
  • 2,500 kilograms an hour. A plant supplying national distribution, or contract manufacturing for someone who is. At this scale, changeover time and cleaning time stop being operational irritations and start being the numbers the plant is managed by.

What we ask for when we size a line

When a plant asks us to propose a layout, five things decide the answer.

The product, and its approximate recipe, which determines forming method and oven profile. The target output in kilograms per hour, arrived at by the method above. The fuel available on site, which is a harder constraint in some markets than any other. Floor length, width and ceiling height, which set what configurations are possible at all. And the packing equipment the line has to feed, since a line that outruns its packer has not added capacity.

With those five we return a proposed layout showing capacity, utilities load and footprint, at no charge.

The question worth asking twice

Is the number you have written down the output you need, or the output you are hoping for.

They are often the same figure, and they never carry the same risk.