Here is the calculation, done once, for one action. Repeat it for every candidate and the ranking builds itself.
The arithmetic
Target category: unplanned stops, 620 hours.
Realistic recovery in year one: 45 percent.
Hours recovered: 279.
Program cost for this action: $145,000.
Cost per hour recovered: $520.
Against an hour worth $6,000, that is roughly eleven dollars back for every dollar spent. And more importantly, it is now a figure that can sit on the same page as every other proposal competing for the same money.
The recovery percentage is where credibility lives
Everything else in that calculation is either observed or quoted. The recovery percentage is a judgment, and it is the number the first serious challenger will go after.
Three rules.
Be conservative. A defended 40 percent that delivers beats an optimistic 80 percent that stalls in month four. The second one does not just fail, it costs you the next proposal, because it teaches the committee that your numbers are aspirational.
Write down how you chose it. Prior experience on a similar asset, a documented result from another site, a vendor’s installed base, or a pilot on one failure mode. Any of these is defensible. What is not defensible is a round number with no derivation, and a round number is exactly what a challenger assumes when you cannot explain it.
Never claim a category to zero. No action eliminates a loss category. Proposing that it will is the single fastest way to lose an experienced operator or engineer in the room, because they know better and they will assume the rest of your case is equally loose.
The four mechanisms that actually return hours
Ranked here by typical cost per hour recovered, cheapest first. Your ranking depends on your own loss profile, which is why you counted.
Precision skills. Alignment, balance, fastening, and installation performed to a documented standard. This attacks unplanned stops at the root and lowers repeat failures on the same asset. In most plants these are the cheapest hours available, because the labor is already on site and the standard costs almost nothing to write.
Lubrication practice. Correct lubricant, clean delivery, controlled contamination, and confirmed application. Small spend, long payback, and it protects the precision work you just paid for. Skipping it means buying the same hours twice.
Planning and scheduling. Work planned before it is scheduled, kitted, and executed inside a protected window. This converts unplanned constraint hours into planned ones at a fraction of the cost, which is a real gain even though the hours do not disappear.
Operator ownership. The crew running the asset detects, reports, and clears the small interruptions. This is where minor stops and startup losses go to die, and it is the only mechanism on the list that scales without adding headcount.
Notice what is not on this list. Software, sensors, dashboards, and consultants. All of those can help, and none of them recovers an hour by themselves. They tell you about hours. Somebody still has to go get them.
Include the hold in the cost
Most cases price the fix and omit the mechanism that holds it: written standards, verified skills, and a monthly view with a named owner.
It is a small share of program cost and it is the difference between a result and a cycle. A case that funds only the fix is a case that funds the fix twice, and the second time is much harder to get approved.
This week
Price three actions against your top three categories. Divide cost by hours recovered. Rank them.
If you cannot estimate a cost, get a rough quote rather than leaving the row blank. An action without a price cannot be compared, and anything that cannot be compared does not get funded.