Mean Time to Failure: Definition

Definition: Mean Time to Failure (MTTF) is the average operating time expected before a non-repairable component fails and must be replaced. It applies to items such as bearings, circuit boards, sensors, and filters that cannot be meaningfully repaired after failure, and is calculated by dividing total operating hours by the number of failures observed in a component population.

What Is Mean Time to Failure?

Mean Time to Failure is a reliability metric that quantifies how long a non-repairable component operates on average before it fails. Unlike metrics applied to repairable equipment, MTTF describes a one-way journey: the component is installed, operates for some period, and eventually fails irreversibly. At that point it is discarded and replaced rather than repaired and returned to service.

The distinction between repairable and non-repairable assets is fundamental to selecting the correct metric. A pump is a repairable system: when it fails, it is diagnosed, repaired, and returned to service. The average time between successive failures on that pump is measured by MTBF. The bearing inside the pump is a non-repairable component: when it fails beyond a wear threshold, it is replaced. The average operating life of that bearing before replacement is its MTTF.

MTTF is particularly valuable for spare parts planning and proactive replacement scheduling. Knowing the statistical average life expectancy of a component type allows maintenance teams to stock replacement parts appropriately and schedule replacements before failures occur, rather than reacting to unplanned breakdowns.

MTTF Formula and Worked Example

The formula is:

MTTF = Total operating hours / Number of failures

Worked example: A facility has 10 identical circuit boards installed in control panels. Over an 18-month observation period, 4 of the 10 boards fail. The total operating hours accumulated across all 10 boards during this period is 123,300 hours.

MTTF = 123,300 / 4 = 30,825 hours

This means a board of this type can be expected to operate for approximately 30,825 hours on average before failure. If the boards operate 24 hours per day, 30,825 hours translates to roughly 3.5 years of expected service life per board.

A practical implication: if the facility wants to replace boards proactively before failure, scheduling replacement at 70-80% of MTTF gives a 70-80% probability that the board has not yet failed, providing a buffer against premature failure while avoiding running components to destruction.

The Three Reliability Phases: The Bathtub Curve

MTTF calculations are only meaningful during one of the three phases of the component reliability lifecycle, often visualized as the bathtub curve due to its characteristic shape when failure rate is plotted against time.

Phase Failure Rate Cause MTTF Applicability
Infant mortality High, decreasing Manufacturing defects, improper installation, poor commissioning Not applicable: failure rate is not constant
Useful life Low, approximately constant Random failures; no dominant degradation mechanism Valid: stable failure rate makes MTTF meaningful
Wear-out Increasing rapidly Accumulated fatigue, material degradation, end of design life Not applicable: failure rate is increasing

Applying MTTF data to components that are already in their wear-out phase significantly underestimates actual failure risk. Components should be replaced before entering the wear-out phase, using MTTF data to identify when this transition typically occurs for each component type.

MTTF vs. MTBF vs. MTTR

Selecting the correct reliability metric requires understanding what each measures and when to apply it.

Metric Applies To What It Measures Example Asset
MTTF Non-repairable components Average life from installation to failure Bearing, circuit board, sensor, filter
MTBF Repairable systems and equipment Average uptime between successive failures Motor, pump, compressor, conveyor
MTTR Any repairable asset or system Average time from failure to full restoration Any asset that is repaired and returned to service

In practice, MTTF and MTBF are often confused or used interchangeably, but the distinction matters for reliability calculations. A pump (repairable) has an MTBF. The pump bearings (non-repairable) have an MTTF. Applying MTBF logic to a component that is replaced rather than repaired will misrepresent the failure economics of that component type.

Practical Applications of MTTF in Maintenance

Proactive Replacement Scheduling

The most direct maintenance use of MTTF is proactive component replacement before failure. Rather than running components until they fail (with the associated unplanned downtime, secondary damage, and emergency labor costs), maintenance teams use MTTF data to identify replacement windows. Scheduling replacement at 70-80% of MTTF provides a statistical buffer: the majority of components of that type will not have failed at that point, but the team acts before entering the high-risk zone. This converts unplanned breakdowns into planned work orders.

Spare Parts Inventory Planning

MTTF enables quantitative spare parts stocking decisions. If a facility has 50 identical sensors with an MTTF of 20,000 hours and each sensor operates 8,000 hours per year, the expected failure rate is approximately 2.5 failures per year per sensor. For 50 sensors, that is roughly 20 failures per year, informing minimum stocking levels and reorder points. This is a direct alternative to reactive stocking, where parts are ordered only after failure and often expedited at premium cost.

Budget Forecasting

By combining MTTF data with component cost and installation labor rates, maintenance managers can project annual component replacement costs with reasonable confidence. This converts reactive maintenance spending from an unpredictable variable into a plannable budget line, supporting more accurate maintenance budget preparation.

Vendor and Component Evaluation

MTTF provides an objective basis for comparing component performance against manufacturer claims and against alternative suppliers. If a manufacturer specifies an MTTF of 50,000 hours and the observed MTTF in operation is 30,000 hours, that gap is quantifiable evidence for a specification discussion or supplier change. Tracking actual MTTF by manufacturer and model number builds a performance database that informs future procurement decisions.

How to Improve MTTF Performance

Proper Installation and Commissioning

Many components fail well before their theoretical MTTF due to poor installation rather than inherent component weakness. Misaligned bearings, overtightened fasteners, contaminated lubricant during installation, and electrical connections made outside specification all accelerate failure. Standardized installation procedures and commissioning checks directly extend realized component life toward its design MTTF.

Operating Within Design Parameters

Components operated beyond their design ratings degrade faster than their MTTF data predicts. Overloaded bearings, motors running hot due to blocked cooling, and sensors exposed to vibration beyond their specified range all experience accelerated wear. Monitoring operating conditions through condition monitoring and predictive maintenance platforms helps ensure components operate within the envelope their MTTF values assume.

Lubrication and Contamination Control

For mechanical components, lubrication quality is one of the largest controllable variables in MTTF. Under-lubricated bearings fail substantially earlier than their rated MTTF. Contaminated lubricant is similarly destructive. Establishing and following correct lubrication specifications, intervals, and quantity controls is one of the highest-return maintenance practices for extending component life toward rated MTTF.

Tracking Remaining Useful Life

Advanced reliability programs use condition monitoring data to estimate the remaining useful life of individual components in service, rather than applying a population-average MTTF to every unit. A bearing that has been monitored for vibration signature change may have a much more precise remaining life estimate than the statistical MTTF figure alone provides. This individual-asset approach reduces unnecessary early replacements while catching components that are degrading faster than average before they fail in service.

The Bottom Line

MTTF is the reliability metric that converts component failure history into actionable maintenance planning. For every non-repairable component in a facility, MTTF answers the question that drives two of the most consequential maintenance decisions: when to replace it, and how many to keep in stock.

The most direct application is proactive replacement scheduling. Replacing components at 70-80% of their MTTF converts unplanned breakdowns into planned work orders, eliminating the emergency labor costs, expedited parts costs, and extended downtime that reactive replacement incurs. For high-consequence failures on critical assets, even earlier replacement thresholds are justified by the asymmetry between planned and unplanned repair costs.

MTTF data improves over time. The more failure events and operating hours a facility accumulates for a given component type, the more accurate its MTTF estimate becomes. Building this database systematically through a CMMS that captures component installation dates, runtime, and failure records is the foundation of a mature reliability program.

Replace Components Before They Fail

Tractian tracks component runtime automatically and alerts your team when assets approach their replacement threshold, eliminating both premature replacements and unexpected in-service failures.

See How It Works

Frequently Asked Questions

What is Mean Time to Failure?

Mean Time to Failure (MTTF) is the average operating time expected before a non-repairable component fails and must be replaced. It applies to components such as bearings, circuit boards, sensors, and filters, and is calculated by dividing total operating hours by the number of failures observed in a population of identical components during their useful life phase.

How is Mean Time to Failure calculated?

MTTF equals total operating hours divided by the number of failures. For example, 10 identical circuit boards accumulate 123,300 total hours with 4 failures: MTTF = 123,300 / 4 = 30,825 hours. The calculation applies to the useful life phase where failure rates are approximately constant. Including data from infant mortality or wear-out phases will distort the result.

What is the difference between MTTF and MTBF?

MTTF applies to non-repairable components that are replaced after failure: bearings, sensors, capacitors, and filters. MTBF applies to repairable systems or equipment that are restored to service after failure. MTBF measures the average operating time between two successive failures on the same asset; MTTF measures the average life from installation to the single failure event that ends the component's service life.

When should replacement be scheduled relative to MTTF?

Scheduling replacement at 70-80% of MTTF is a commonly used approach. At this point, the statistical probability that the component has already failed is relatively low, providing a reasonable buffer against premature failure. The exact percentage depends on the consequences of an in-service failure: higher-consequence failures justify earlier proactive replacement, while lower-consequence components can be run closer to MTTF.

What are the three phases of the bathtub curve?

The three phases are: infant mortality (high decreasing failure rate shortly after commissioning, caused by manufacturing defects or poor installation), useful life (low constant failure rate, the phase where MTTF applies), and wear-out (increasing failure rate as the component approaches end of design life). MTTF calculations are only valid during the useful life phase when failure rates are stable.

How does MTTF differ from remaining useful life?

MTTF is a population statistic: the average life of a component type based on historical failure data across many identical components. Remaining useful life (RUL) is an individual asset estimate: the predicted time before a specific component in service will fail, derived from its current condition monitoring data and degradation trend. MTTF provides the baseline expectation; RUL provides a more precise estimate for components that are being individually monitored.

Related terms