Equipment total cost of ownership: what a production machine costs after the invoice is paid

Everything that has to happen before the first good part

Between delivery and production sit costs nobody quotes: removing the machine being replaced, foundations or floor reinforcement, power supply upgrades, compressed air and extraction connections, guarding and interlocks to bring the installation into conformity, lifting and rigging, commissioning support, trial material consumed while dialling in, and training for operators and maintainers. On a substantial machine these frequently rival a meaningful fraction of the purchase price. They also land in the same financial year, which is why underestimating them is the classic way a project overruns before it has produced anything saleable.

The running cost per hour and what belongs in it

Build an hourly cost containing the things the machine consumes whether or not anybody thinks about them: electricity and compressed air at metered draw rather than nameplate, consumable tooling and its regrind or replacement cycle, coolant and its disposal, planned maintenance labour and parts, and the operator attendance the machine actually requires. Attendance is the item most often assumed away — a machine sold as capable of unattended running usually needs someone within reach for load, unload and intervention, and the difference between one operator per machine and one across several changes the economics more than the price did.

Support risk is a cost even when nothing breaks

Ask before purchase how long the builder will supply parts, whether the control system uses components that will still exist in a decade, whether diagnostic software is licensed or locked, what happens if the local agent changes, and whether anybody other than the builder can service it. A machine whose controller is discontinued becomes unmaintainable long before it wears out mechanically, and a retrofit is expensive and disruptive. Where a machine is unique on site, add the cost of the critical spares that must be held because the lead time would otherwise stop production for weeks.

Downtime cost depends entirely on where the machine sits

An hour lost on a constrained machine costs the contribution the plant could not earn; the same hour on a machine with spare capacity costs almost nothing. So the same reliability figure carries wildly different value depending on position in the route, and a cheaper, less reliable machine can be perfectly sensible off the constraint and indefensible on it. Estimate expected unavailability from the builder's data and from experience with similar equipment, then price it according to where this machine will actually sit — including the possibility that a future product mix moves the constraint onto it.

Owning the number and using it in the decision

The estimate should be built by engineering with maintenance and finance contributing, and the assumptions written down rather than embedded in a spreadsheet nobody can interrogate. Then it must survive contact with the decision: capital cases are frequently rewritten to fit an available budget, and the first casualties are spares provision, training and installation contingency, all of which reappear later as unbudgeted spend. Revisiting the estimate against reality a year after commissioning is the discipline that makes the next one credible, and almost nobody does it. Publishing the comparison in the same format each time also lets a committee weigh two very different machines against each other instead of debating whose spreadsheet is more generous.

Frequently asked questions

How far ahead should we cost a machine's life?

Over the period the plant realistically expects to use it for the product it was bought for, not over the accounting depreciation life, which is a tax and reporting convention rather than an engineering estimate. Many machines outlive the product that justified them, so consider what else the asset could do afterwards. Where the product's future is genuinely uncertain, a shorter horizon with an explicit residual assumption is more honest than a long projection everyone privately doubts.

Should a used machine be evaluated the same way?

The framework holds, with two additions. Condition assessment before purchase is essential and worth paying an independent engineer for, because the seller's account of remaining life is not evidence. And support risk is usually much higher: parts availability, control obsolescence and the absence of documentation are common. A used machine can be an excellent decision when the technology is mature, the same model exists elsewhere on site, and the price reflects the support position rather than ignoring it.

Is the cost of the space a machine occupies worth including?

Include it where floor space is genuinely constrained, because then the space has an opportunity cost — something else could have stood there. On a site with room to spare, allocating a notional rent adds arithmetic without changing any decision. Where it does matter, compare footprint honestly by including the access, maintenance clearance and material staging each option needs, rather than the machine outline on the drawing.