System Availability Calculator
The availability of a system of machines or components in series, in parallel or k-out-of-n, from each unit's availability or its MTBF and MTTR. Updates as you type.
This is reliability availability (time not lost to failure and repair), not OEE availability (run time ÷ planned production time): for OEE use the availability calculator, and for one asset's numbers the MTBF and MTTR calculator.
Your system
Pick how the units are arranged, then enter each unit. The defaults are the worked example below.
Block diagram
What each unit does to the system
Want a second look at this result?
Leave your work email if you would like the OEE Lab team to get in touch about your numbers. To keep the result, use Copy link or Print above.
No newsletter, no spam.
See every stop as minutes of OEE lost.
A system figure is only as good as the MTBF and MTTR behind each block. Fabrico reads every stop from the PLC or IoT sensors and shows it as minutes of OEE lost per line and shift, so MTBF and MTTR build up per machine from real stops, and where cameras are fitted the video of any stop can be streamed on request next to its machine data.
Request a demo Contact usWhat the system availability calculator does
It combines the availability of several units into one figure for the system they form. A unit can be a machine, a pump, a drive or any component that fails and gets repaired.
You choose the arrangement. In series every unit must run; in parallel any one unit can carry the whole load; in k-out-of-n at least k of n identical units must run.
For each unit you enter its availability in percent, or its MTBF and MTTR in hours. The tool returns the system availability, the expected downtime hours a year, the weakest link in series and what each extra unit adds in parallel.
The formulas, with the working
These are the standard reliability block diagram formulas, applied to steady-state availability. Reliability textbooks such as C. E. Ebeling, An Introduction to Reliability and Maintainability Engineering (McGraw-Hill, 1997), use the same series, parallel and k-out-of-n rules.
One unit: A = MTBF ÷ (MTBF + MTTR)
- One unit: MTBF 99 h and MTTR 1 h give 99 ÷ (99 + 1) = 99.00%.
- Series: A = A1 × A2 × ... × An. Two 95% units give 0.95 × 0.95 = 90.25%.
- Parallel: A = 1 - (1 - A1) × (1 - A2) × ... × (1 - An). Two 95% units give 1 - 0.05 × 0.05 = 99.75%.
- k-out-of-n, identical units: A = sum for i = k to n of C(n, i) × Ai × (1 - A)n - i. For 2 of 3 at 90%: 3 × 0.9² × 0.1 + 0.9³ = 0.243 + 0.729 = 97.20%.
- Downtime a year = (1 - system A) × hours per year.
The formulas rest on three assumptions. Failures are independent, each unit has its own repair so repairs can run at the same time, and every figure is a long-run steady-state average, not a forecast for next week.
Worked example: the default numbers
Example: a bottling line runs a filler, a capper and a labeller in series, 8,760 hours a year. The filler is at 98% and the labeller at 95%; for the capper you know MTBF 99 h and MTTR 1 h.
| Unit | Availability | Downtime alone | System if it never failed |
|---|---|---|---|
| Filler | 98.00% | 175.2 h | 94.05% |
| Capper (99 ÷ 100) | 99.00% | 87.6 h | 93.10% |
| Labeller, weakest | 95.00% | 438.0 h | 97.02% |
| System (0.98 × 0.99 × 0.95) | 92.17% | 686.0 h |
The system is available 92.17% of the time, so it expects (1 - 0.92169) × 8,760 = 686.0 hours of downtime a year. The labeller is the weakest link.
Raise the labeller to 99%, as good as the capper, and the system goes to 0.98 × 0.99 × 0.99 = 96.05%, about 346.0 hours of downtime. That is 340 hours a year back from one machine.
The three units' own downtime adds up to 700.8 hours, 14.8 more than the system figure. The formula lets failures overlap in time, so read 686.0 h as an estimate, not a promise.
How to read the result
- Series: the system is always below its weakest unit. Start with the red block; the last column of the table shows the most each fix could win.
- Parallel: the table adds one unit at a time, in the order you listed them. The first spare usually gives the big jump, so check what the next one really adds before you pay for it.
- k-out-of-n: the table shows what one or two more identical units would add for the same k.
- Downtime hours are easier to discuss than 99.9%. Put a price on them with the downtime cost calculator.
Common mistakes
- Mixing OEE availability with reliability availability. OEE availability counts every stop against planned production time, including changeovers and waiting for material. This tool counts failures and repairs only, and OEE vs technical availability explains the gap.
- Assuming redundancy helps when failures share a cause. Two pumps on one power supply, one control panel or one batch of bad oil can fail together. The parallel formula assumes they cannot, so it overstates the gain; an FMEA is a good place to find shared causes.
- Calling capacity "parallel". If two half-speed machines must both run to meet the rate, the system is 2-out-of-2, which is the same as series. MTBF for multiple machines separates true redundancy from capacity in parallel.
- Mixing MTBF and MTTR definitions. Figures from different windows, or MTTR with waiting time for one unit and without it for another, give a system number nobody can trust. How to calculate MTBF sets out one method.
- Treating the weakest link as the bottleneck. The least available unit is not always the slowest one. Throughput limits are a separate question; see bottleneck analysis.
What is system availability?
It is the share of time a system of units is able to run, given each unit's failures and repairs. It comes from the units' own availability and how they are arranged: in series, in parallel or k-out-of-n.
How do you calculate availability for machines in series?
Multiply the availabilities. Three machines at 98%, 99% and 95% give 0.98 × 0.99 × 0.95 = 92.17%, which is below the worst machine.
How do you calculate availability for redundant units in parallel?
Multiply the chances that each unit is down, then subtract the result from 1. Two units at 95% give 1 - 0.05 × 0.05 = 99.75%, if their failures are independent.
What is k-out-of-n availability?
It is the availability of a system that needs at least k of its n identical units running. Three units at 90%, where any two are enough, give 97.20%.
Is this the same as OEE availability?
No. OEE availability is run time divided by planned production time and counts every stop, while this calculator counts failures and repairs only. For OEE, use the availability calculator.
Can I turn MTBF and MTTR into availability?
Yes. Availability = MTBF ÷ (MTBF + MTTR), so an MTBF of 99 hours and an MTTR of 1 hour give 99%. Switch a unit row to MTBF, MTTR and the tool does it for you.