What an hour on the printer actually costs
Updated 2026-09-04Most shops price machine time with a number they guessed once, or copied from a forum post written about a different printer in a different country. It is usually wrong in the same direction: too low, because the costs that are easy to forget are the ones that do not arrive as a monthly bill. The rate is not hard to derive. It has four parts, and one of them causes most of the error.
The four components
A machine-hour rate is the fixed cost of owning a printer, spread across the hours it will actually print, plus the variable cost of running it for one of those hours.
- Depreciation — the purchase price, plus anything you had to buy to make it usable (enclosure, upgrades, a table), divided by the hours it will print across its life.
- Maintenance — nozzles, belts, plates, a yearly service. Estimate it annually; per-hour figures invite guessing.
- Electricity — average draw in watts, divided by 1,000, times your price per kWh.
- A buffer — a deliberate uplift over the first three, covering what you did not model.
The denominator is where the error is
Depreciation and maintenance are annual figures that have to be spread over hours. Which hours you choose changes the rate more than any of the costs do.
The temptation is to use calendar hours — 24 × 365, or a working year. Do not. Use the hours the machine will realistically be printing: subtract downtime, maintenance, the gaps between jobs, the weeks work is thin. A printer you believe runs constantly might bill 3,000 hours a year, not 8,760. That is a factor of nearly three on every fixed cost, applied to every quote you send.
If you are unsure, estimate low. A rate derived from optimistic uptime undercharges on every job, and the shortfall is invisible because nothing about it looks like a loss.
Why a buffer, and not just better estimates
The buffer is not padding for its own sake. It covers the costs that are real but awkward to model per hour: a failed print that consumed six hours and a spool, the plate you replaced early, the firmware afternoon nobody billed for.
Modelling each of those precisely is possible and rarely worth it. A single multiplier over the derived subtotal — twenty percent is a common starting point — is honest about the uncertainty instead of hiding it inside a falsely precise number.
A worked example
A printer bought for 22,000 with 3,000 of upgrades, expected to last five years, realistically printing 2,500 hours a year. Maintenance runs about 1,800 a year. It draws 150 W average, and electricity costs 2.60 per kWh. The figures are in whatever currency the shop works in; only the ratios matter.
What the rate does not cover
This is the cost of the machine running, and nothing else. Material is priced per gram against the spool it came from. Your time — slicing, plate prep, removal, finishing — is labour, and belongs on the quote as labour rather than buried in a machine rate. Mixing them produces a number that cannot be checked and that nobody can explain to a customer who asks.
Keeping them separate also means each can be wrong independently, and you can see which one it was.
How Kraftius handles it
A printer stores its purchase price, additional cost, expected life, billable hours per year, maintenance, power draw and buffer, and derives the rate from them. Electricity resolves from a per-printer override, then the workshop-wide price, so changing your tariff updates every printer that has not overridden it.
You can also enter a rate directly. A manually entered rate always wins — useful when you have a figure you trust from elsewhere, and honest about the fact that the derivation is a tool rather than an authority.