Mean Time Between Failures (MTBF)

Mean Time Between Failures (MTBF) is the average operating time between two consecutive failures of a repairable asset. Computation: MTBF = (total operating time) / (number of failures), with units typically of hours or cycles. MTBF is one of the two main metrics of system reliability; the other is MTTR (mean time to repair). Together they determine equipment availability: Availability = MTBF / (MTBF + MTTR). This equation underlies the Availability component of OEE (Nakajima 1988). For non-repairable equipment (e.g., a single-use electronic board) the counterpart is MTTF (Mean Time to Failure). The statistical background matters: MTBF is not just an arithmetic mean — it is interpreted under an assumed failure-time probability distribution. The most common assumption is the exponential distribution (constant failure rate λ = 1/MTBF); a more realistic assumption is the Weibull distribution (Weibull 1951) — its shape parameter β separates wear-in, random, and wear-out regions (the 'bathtub curve'). Therefore it is wrong to say a pump with MTBF = 1,500 hours will fail 'exactly at 1,500 hours'; it is a mean-only statement. MTBF is the fundamental input to preventive-maintenance planning — preventive-maintenance intervals are typically set at 30–50% of MTBF. It feeds the equipment-replacement problem, job-shop scheduling, and the Availability component of OEE. References: Weibull 'A Statistical Distribution Function of Wide Applicability' (1951); Ebeling 'An Introduction to Reliability and Maintainability Engineering' (1997); Birolini 'Reliability Engineering: Theory and Practice' (2007); IEC 60050-191; the military standard MIL-HDBK-217 (electronic-component MTBF). SMB measurement practice: annual roll-up from the maintenance log; MTBF becomes meaningful after at least 5–10 failure events.