| From | To per hour | Example |
|---|---|---|
| FIT (failures per 10⁹ hours) | × 10⁻⁹ | 250 FIT = 2.5 × 10⁻⁷ per hour |
| Failures per year | ÷ 8,760 | 0.01 per year = 1.14 × 10⁻⁶ per hour |
| Failures per million hours | × 10⁻⁶ | 0.3 per 10⁶ h = 3 × 10⁻⁷ per hour |
| MTTF (years), constant rate | 1 / (MTTF × 8,760) | MTTF 100 years = 1.14 × 10⁻⁶ per hour |
Traps
- MTBF is not a dangerous failure rate. It usually covers all failures, safe and dangerous, detected and undetected. Dividing 1 by MTBF gives λ total, not λDU.
- MTBF is not a lifetime. An MTBF of 100 years does not mean the device lasts 100 years; it describes the constant-rate period only.
- Per demand versus per hour. Some data (for example human error) is a probability per demand or per task, not a rate.
Quick check
λDU of a certified transmitter around 10⁻⁷ per hour means about one dangerous undetected failure per 1,140 device-years. If your plant has 500 such transmitters, expect a dangerous failure found at proof tests roughly every two years across the fleet.
Written for First Out AI as engineering guidance. It supports engineering work and does not replace it; check every value against your standards, certificates and company rules with a competent functional safety engineer.