MTTR vs MTBF: What Each Really Measures and When Each Matters
Table of Contents
The MTTR vs MTBF comparison is the most common source of confusion in reliability metrics. They sound similar, they are reported together, and most plants pick the wrong one to optimize. The result is teams that drive one number without moving the one that actually affects production.
This guide lays out what each metric measures, how they combine into availability, and — most importantly — how to decide which one your plant should prioritize this quarter.
The Clean Definitions
MTBF (Mean Time Between Failures) = total operating time ÷ number of failures. Answers: "how often does this equipment fail?"
MTTR (Mean Time To Repair) = total repair time ÷ number of repairs. Answers: "when it fails, how long is it down?"
They are independent metrics. A pump can have high MTBF and high MTTR, or low MTBF and low MTTR. You need both to understand reliability.
How They Combine Into Availability
The metric production actually cares about is availability: the percentage of planned operating time that equipment is ready to run. The formula:
Availability = MTBF / (MTBF + MTTR)
Example: a conveyor with MTBF of 400 hours and MTTR of 4 hours = 99.0% availability. A conveyor with MTBF of 200 hours and MTTR of 1 hour = 99.5% availability. The second one fails twice as often but is MORE available because it's repaired faster. This is why MTTR vs MTBF is not a one-or-the-other question.
When to Prioritize MTBF
Focus on MTBF when:
- Failures are safety-critical. Every failure is a potential incident; reducing failure frequency beats reducing repair time.
- Repairs are cheap and fast. If MTTR is already low, the only way to move availability further is MTBF.
- Failures cascade. One failure causes downstream effects (lost batches, damaged product). Prevent the first domino.
- Spare parts are hard to get. Catastrophic repairs with long lead times punish low MTBF disproportionately.
When to Prioritize MTTR
Focus on MTTR when:
- Failures are inevitable or frequent. For equipment with structurally short MTBF (cutting tools, seals, belts), improving repair speed is the higher-leverage move.
- Downtime cost is nonlinear. An hour of downtime on a continuous process costs far more than 4×15-min stops. Cutting MTTR collapses compound losses.
- Technician time is the binding constraint. Short MTTR means more work capacity per technician.
- Equipment is old or hard to improve. You can't fundamentally change MTBF on a 20-year-old asset. You can change how fast the team responds.
The Most Common MTTR vs MTBF Mistake
Plants obsess over MTBF reporting and ignore MTTR because MTBF feels more "preventive" and sophisticated. Meanwhile their MTTR is 4x industry median and they can't figure out why availability is poor.
Rule of thumb: if your MTTR is above the industry median for your equipment class, fix that first. It's cheaper, faster, and doesn't require changing the equipment itself. Then attack MTBF.
See also: MTTR reduction strategies and MTBF improvement.
Frequently Asked Questions
Is MTTR or MTBF more important?
Both, and it depends on your current gap. If MTTR is above median for your equipment class, fix MTTR first. If MTBF is your weak spot, fix MTBF first. Neither is universally more important.
What is a good MTTR?
Depends heavily on equipment and context. Rotating equipment in process plants: 2-4 hours. CNC machines: 30-90 minutes. Conveyor belts: 15-45 minutes. Benchmark against your industry, not a universal number.
Are MTTR and MTTF the same?
No. MTTR is Mean Time To Repair (repairable systems). MTTF is Mean Time To Failure (non-repairable items like bearings or lightbulbs). Different math, different use.
How do I improve MTBF and MTTR at the same time?
Focus resources on the top 5 failure modes by frequency × impact. Root-causing these both reduces failure frequency (MTBF) and makes repairs faster (MTTR) because the team learns the failure pattern.
Does AI change the MTTR vs MTBF decision?
AI-assisted diagnosis pushes MTTR down faster than MTBF, because diagnostic time is often 50%+ of repair time. AI-based prediction lifts MTBF but requires good condition monitoring data. Start with MTTR if you have neither.
