Downtime cost & PM ROI
What unplanned downtime really costs, and what a preventive-maintenance program pays back — the numbers that justify a PM contract instead of running to failure.
Unplanned-downtime cost & preventive-maintenance payback
net = failures × hours × $/hr × reduction − PM cost
What this gives you
This is the tool that turns “we should do more preventive maintenance” into a number a plant manager can sign. Enter how often a machine fails in a year, how many hours each failure keeps it down, and what an hour of that downtime really costs — then the reduction a PM program is expected to deliver and what the program costs to run. It returns the money unplanned failures drain every year, the share a PM program would avoid, the net annual saving after the program’s cost, and the return on that spend. The relationship is simple: net saving = failures × hours × cost/hour × reduction − PM cost.
Why “cost per hour” is the number that matters
Two of these inputs are easy to look up and one is not. Failure count and hours down come straight off the maintenance log. The cost of a downtime hour is where estimates go wrong, almost always on the low side, because people reach for the repair cost and stop there. A true downtime hour includes production you can never make back, scrapped or off-spec product at restart, operators still on the clock with nothing to run, expedited freight to catch up, and sometimes a penalty for a late shipment. On a constrained line the honest figure is the contribution margin of everything that line would have made in that hour — frequently several times the maintenance-only number. Get this one right and the rest of the calculation follows.
Simple or detailed cost model
If you already carry a blended downtime rate, leave the model on Simple ($/hr) and enter it directly. If you would rather build the number from its parts, switch to Detailed and enter three figures instead: lost production per hour, repair-labor per hour, and parts per failure. The cost of one failure is then hours × (production + labor) + parts, and the annual downtime cost is that per-failure cost times the number of failures a year. Both models feed the same avoided-cost, net-saving and ROI math — the detailed model just makes the hourly rate defensible line by line instead of asking the room to trust one blended figure. Alongside ROI the tool also returns a simple payback in months (PM cost divided by monthly net saving) and the 5-year net saving, the two numbers a budget owner usually asks for next.
Field note — keep the reduction estimate conservative
A PM program does not stop every failure, and in its first months it will not stop the ones already in motion. A 70 percent reduction is an aggressive, mature-program figure; 40 to 60 percent is a more defensible first-year claim. Model the number you can defend, not the best case — if the ROI still clears with a conservative reduction and an honest hourly cost, the argument holds up in the room. If it only works at 90 percent and a padded hourly rate, it will not survive the first budget review.
Worked example
A line goes down 6 times a year, 8 hours each time, and an honest downtime hour costs $5,000 once lost throughput and restart scrap are counted. That is 6 × 8 × $5,000 = $240,000 a year lost to unplanned failures. A PM program expected to cut that by 70 percent avoids $168,000. Subtract the $40,000 a year the program costs and the net saving is $128,000 — a 320 percent return on the PM spend, a 3.75-month simple payback, and $640,000 over five years. Even at a conservative 50 percent reduction the program still nets $80,000. That is the whole point: when the hourly cost is real, running to failure is almost never the cheaper option.
Building the same hour from parts in the detailed model — $4,000 of lost production, $150 of repair labor, and $2,000 of parts per failure — gives a per-failure cost of 8 × ($4,000 + $150) + $2,000 = $35,200, or $211,200 a year across six failures. The 70 percent reduction avoids $147,840, and after the same $40,000 program the net is $107,840 — a 270 percent return built from numbers each department can vouch for.