Total downtime: the honest sum of every hour you lost

At month end the operations manager pulls the production report, and the totals are lower than anyone expected. The equipment seemed to be available most of the time. Then someone actually adds up the hours the asset was stopped: the planned maintenance window, the three breakdowns, the emergency repair after the weekend shift. The "most of the time" feeling does not survive the arithmetic.

Total downtime is that arithmetic. It is the single, honest number for how much time was lost, and it has the great virtue of hiding nothing.

What it actually measures

Total downtime is the sum of all the hours an asset could not run, whatever the reason. It has just two parts: scheduled downtime, the work that was on the finalised weekly maintenance plan, and unscheduled downtime, the repairs that were not.

That split is the whole point. The total tells you how much capacity you lost; the two halves tell you what to do about it.

How to work it out

Total downtime = Scheduled downtime + Unscheduled downtime

In a month, an asset loses 50 hours to planned work and 25 to surprises.

Total downtime = 50 + 25 = 75 hours

Across a 30-day month of 720 hours, that is 10.4% of all available time gone, two-thirds of it planned and one-third a surprise. A glance at that balance already tells you where to look: a heavy scheduled share points to maintenance strategy and outage planning, while a heavy unscheduled share points to reliability problems and the need for more proactive work.

What good looks like

Be a little careful with benchmarks here, because the honest picture has two layers. The part that should be small is the unscheduled half: best-in-class unplanned maintenance downtime sits around one per cent of available time, against five per cent or more in a typical plant. Planned downtime then adds a few points on top, so a genuinely well-run operation often carries total maintenance downtime in the region of five per cent, not near zero. Top-quartile maintenance-related downtime, in other words, is low single digits, with the unscheduled slice the one you really drive toward the floor. All of it shifts with industry and with whether the process runs continuously or in batches, so the only fully fair comparison is against your own history.

The split is where the value lives

The reason to separate the two halves is that they cost completely different amounts and call for completely different fixes. A repair forced on you by a breakdown typically runs about three times the cost of the very same job done on a plan, once you count the overtime, the expedited parts and the production lost while you scramble. So the mission hiding inside this metric is conversion: moving hours out of the unscheduled column and into the scheduled one by catching failures early, in the window between the first detectable sign and the actual breakdown. A plant whose total downtime holds steady while its unscheduled share shrinks is winning, even though the headline number has not moved.

Where it can mislead you

• This figure is for maintenance-related stoppages. Hours lost to thin demand or a feedstock shortage are real, but they do not belong against this benchmark.
• Continuous processes usually show less maintenance downtime than batch ones, so cross-industry comparisons flatter some plants and punish others unfairly.
• Read it next to a rate-based measure such as overall equipment effectiveness, which catches the slow-running losses that pure downtime never sees.
• Beware the comforting belief that downtime "costs us nothing because we have spare capacity". That redundancy hides the true cost rather than removing it: the extra lines and spares you carry, the overtime, the scrap at every switchover, and the catastrophic failure that eventually arrives anyway. And do not assume a tidy schedule means a tidy reality, because in many plants well under half of the planned work actually gets completed as scheduled.

Lost time is the easiest thing in a plant to feel and the hardest to admit. Total downtime puts an end to the argument by laying every lost hour on the table at once, then quietly splitting them into the ones you chose and the ones that chose you. The first kind is an investment; the second is the bill, and it is the one worth attacking first.