Deep-dive guide
Factory OEE and Downtime Domino Simulator: What Every Lost Hour Really Costs
Executive industry analysis
Overall equipment effectiveness is the simplest honest score of how much of a factory's promise actually reaches the dock. A line is scheduled for a number of hours and designed for a speed, yet only part of that time runs, only part of the running time is at full speed and only part of the output is good. Most plants sit well below the figure they imagine, and the gap is rarely one dramatic failure. It is dozens of small losses that nobody owns, each too small to trigger a meeting, together large enough to decide whether a month makes money.
The reason the problem is hard is that the three losses multiply instead of adding. A line that runs 87 percent of the time, at 88 percent of ideal speed, with 97 percent good output, has an OEE of about 75 percent, not 87. Every extra point lost in one factor reduces the value of the other two. This is why a plant can look busy, with machines humming and people moving, and still return only three quarters of the output it was built for while paying for one hundred percent of the hours.
The domino model follows the loss through the money. Downtime lowers availability, which leaves fewer hours to run; slow cycles cut the units made in those hours; scrap removes some of them from sale; the remaining good units must then cover the plant cost of the whole schedule, because wages, energy and depreciation are paid per scheduled hour. Because plant cost is largely fixed, a small drop at the start can flip the last box from profit to loss.
The iceberg separates the bills you see from the margin you never earned. Above the waterline sit materials and plant cost, which appear in the accounts. Below it sit the margin lost in stopped hours, the margin lost to slow running, the sales value of scrapped units and the restart cost after each stop. None of these is an invoice, so none is reviewed, yet together they are often larger than a typical improvement budget. Use the simulator to price your own lost hour.
Formula and methodology blueprint
The model works on one month of 480 planned production hours. Running time is planned time minus downtime. Units started are running hours times the ideal rate times one minus speed loss, and good units are started units times one minus scrap. Revenue is good units times price. Material is charged on every unit started, plant cost on every planned hour, and each stopped hour adds a restart cost equal to a share of the hourly plant cost.
- Availability
- A = (480 − downtime) ÷ 480
- The share of planned time during which the line actually runs.
- OEE
- OEE = A × P × Q
- Availability times performance times quality; the three losses multiply rather than add.
- Units started
- S = (480 − downtime) × ideal rate × (1 − speed loss)
- What the line really produces in the running hours before inspection.
- Good units
- G = S × (1 − scrap)
- Units that can be sold at full price; scrapped units still consumed material and time.
- Net profit
- Net = G × price − S × price × 45% − 480 × cost per hour − restart
- The monthly result after material, plant cost and restart losses.
- Value of an hour
- H = ideal rate × (1 − speed loss) × (1 − scrap) × price × 55%
- The margin that one extra running hour would add, used to price downtime.
Cascading elasticity
Downtime and speed loss act on the same volume, so each point is worth about the same, but downtime also adds restart cost and is usually easier to find. Scrap is the most expensive percentage point because a scrapped unit has already consumed material and time. Hourly plant cost and price are the steepest levers because they are multiplied by the whole schedule and the whole output. Ideal rate scales everything, so a wrong nameplate rate quietly misstates every loss.
Three real-world case studies
Bootstrapped starter
A small shop with one hand-fed line
A small workshop runs a single line at forty units an hour for premium goods. Stops are frequent but the price is high and plant cost is modest, so the month stays positive with little room. Loading it shows how a handful of lost hours matter when there is only one line to carry the whole schedule.
Scaling mid-market
A high-speed line making cheap parts
A large plant runs a fast line at three hundred units an hour on a low-priced part. Downtime and scrap are low by industry habit, yet the volume is so high that each lost hour is worth thousands. Loading it shows how a tiny OEE gain on a fast line pays more than a big gain on a slow one.
Black-swan crisis
A line starved by stops and rework
A plant loses a third of its planned hours to breakdowns, runs well below ideal speed and scraps one unit in nine while energy costs have risen. Material and plant cost keep rolling regardless. Loading it shows the month turning deeply negative and which recovery lever brings it back first.
Industry benchmark matrix
| Metric | Danger zone | Industry average | Top 10% elite |
|---|---|---|---|
| OEE overall | Below 60% | 60–75% | Above 85% |
| Availability | Below 80% | 85–90% | Above 92% |
| Performance (speed) | Below 80% | 85–92% | Above 95% |
| Quality (first-pass yield) | Below 94% | 96–98% | Above 99% |
| Unplanned downtime share of planned time | Above 20% | 8–15% | Below 5% |
Variables and levers
Unplanned downtime per month
Breakdowns, changeovers, material waits and other stops during the 480 planned production hours.
Lever: Log every stop by cause for two weeks, then attack the top three with quick changeover and planned maintenance.
Speed loss
How far the line runs below its ideal rate because of slow cycles, micro-stops and cautious operators.
Lever: Run the line at its ideal rate on a trial shift and fix the micro-stops and settings that force the operators to slow down.
Scrap and rework rate
Share of started units that fail inspection and cannot be sold as first quality.
Lever: Trace rejects back to the first station where the defect appears and fix the cause there, not at final inspection.
Ideal rate (units per hour)
The fastest rate the line can sustain on good material with no stops, taken from the design cycle time.
Lever: Check the nameplate rate against real cycle time and remove the slowest station, since the whole line runs at its pace.
Selling price per unit
What a good unit earns before material and plant costs; material is modelled at 45% of the price.
Lever: Review price and mix, because every hour you recover is worth more when each good unit earns a healthy margin.
Plant cost per scheduled hour
Labour, energy, depreciation and overhead that are paid for every planned hour whether the line runs or not.
Lever: Cut the energy and overhead that run while the line is stopped, and trim scheduled hours that have no orders behind them.
How to use this simulator
- 1Enter last month's downtime hours, speed loss and scrap rate from your production records.
- 2Set the ideal rate from the design cycle time and the price and plant cost per hour from your accounts.
- 3Read the five domino boxes to see whether availability, speed or quality turns amber first.
- 4Press Find my breaking point to learn how many lost hours or how much scrap erases the month.
- 5Compare the current plan with the quick-changeover plan over 24 months in the Parallel universe tab.
Mistakes to avoid
- Measuring OEE against a flattering ideal rate. If the nameplate speed is optimistic, performance looks fine while the real loss hides in the rate itself.
- Counting only breakdowns as downtime. Changeovers, waiting for material and short stops are usually larger in total and are rarely logged.
- Pricing a lost hour at full margin when the line has no orders to fill it. Recovered time only pays when there is demand for the extra output.
- Chasing one factor and ignoring the others. Raising speed can increase scrap, and fewer stops can mask a quality problem that appears later.
Frequently asked questions
What is a good OEE?
Many manufacturers quote 85 percent as world class, but a typical plant sits between 60 and 75 percent. Compare yourself with your own line over time, because product mix and process type change what is realistic.
Why do availability, performance and quality multiply?
Because each one applies to what remains after the previous loss. Time lost to stops reduces the hours that can be run slowly, and slow output is then the pool from which scrap is taken.
Is a lost hour always worth the full margin?
Only if you could sell the extra output. If the line is demand constrained, the true value of a recovered hour is lower. Check your order book before you price downtime at full margin.
Why is plant cost charged on stopped hours?
Because wages, depreciation and base energy are paid for scheduled time. The simulator therefore treats them as fixed, which is why stoppages hurt so much.
How do I find my breaking downtime?
Press Find my breaking point. The tool pushes each input towards its worse end and reports the value at which the monthly result reaches zero, ranked by how close each lever is.
How accurate is this?
It is an educational planning model with the simplifications stated above, not engineering or accounting advice. Replace the inputs with your own measured figures and compare with your shop-floor data.
This is an educational planning model with stated assumptions. It is not financial, tax or legal advice. Replace the inputs with your own figures and verify important decisions with a qualified professional.