In short
MTBF measures random failures during the “useful life” phase — not early defects or wear-out. L70 measures LED module lumen maintenance — not system lifespan. In many field installations — particularly in humid or coastal environments — moisture, capacitors, or solder joints are the practical lifespan-limiting factors rather than the LED reaching L70.
The luminaire is a series system
A street luminaire consists of LED module, driver, optics, gaskets, cable entries, solder joints, and housing. In a series system the weakest component determines system reliability — exactly as a chain breaks at its weakest link.
The LED module’s L70 value is a module claim. It states that the module, under standardised test conditions, retains 70 % of its original luminous flux at the rated hours. It says nothing about the gaskets’ remaining seal at that point, the capacitor’s remaining capacitance, solder joint fatigue, or cable entry integrity. Reporting the LED module’s L70 as “the luminaire’s lifespan” is technically misleading.
The bathtub curve — three phases
The reliability engineering model for failure rate over time is called the bathtub curve — a name derived from the shape of the graph. It describes three distinct phases:
- Early failures (infant mortality) — Decreasing failure rate during the first months to years. Caused by manufacturing and installation defects: poor soldering, misaligned gasket, pinched cable, unsealed entry, incorrect torque. Burn-in and thorough quality assurance (QA) catch many of these before delivery. Remaining defects surface early in weak installations.
- Random failures (useful life) — Approximately constant failure rate. Isolated component failures unrelated to ageing, external voltage transients (surge), lightning strikes, mechanical damage. This is the phase MTBF describes — and only this phase.
- Wear-out — Rising failure rate. Electrolyte drying, solder joint fatigue from thermal cycling, gasket ageing, LED lumen degradation. This is where L70 and L90 fall — but often later than other system components.
Why MTBF misleads
MTBF (Mean Time Between Failures) is defined as the average time interval between two random failures in a population of units during the constant failure rate phase. The mathematical definition is correct. The problem is how the figure is presented. (Technical note: MTBF strictly applies to repairable systems — time between failures. Street luminaires are typically non-repairable in the field and are replaced at end-of-life; the technically precise measure for non-repairable units is MTTF, Mean Time To Failure. Industry convention applies MTBF to both situations.)
The most common industry misconception
An MTBF of 100,000 hours does not mean the product is expected to last 100,000 hours. It means the average failure frequency during the useful life phase is 1 failure per 100,000 operating hours — meaning roughly 1 % of a large population can be expected to fail during any given 1,000-hour period in that phase.
MTBF captures neither early production defects nor wear-out mechanisms. An electrolytic capacitor whose actual lifespan at operating temperature falls short of the installation’s intended service life — which can apply to undersized components — does not show up in the MTBF figure. Wear-out belongs to a different phase of the bathtub curve. A high MTBF therefore does not rule out early wear-out of a specific component.
A more informative way to report expected lifespan is via B10 lifespan (the time at which 10 % of a population has failed) or via Weibull analysis of the system’s weakest components, accounting for environment and operating conditions.
What fails first
Based on published field studies and engineering analysis for Nordic operating conditions, a representative ordering is shown below. This is an engineering judgement based on available field studies and materials science — no published study provides quantitative support for this exact ranking specifically for Nordic street luminaires. The order varies considerably with design, installation environment, and maintenance regime.
| # | Failure mechanism | Type | Driven by |
|---|---|---|---|
| 1 | Moisture and water ingress | Mixed¹ | Gasket ageing and pressure/temperature cycling (wear-out); incorrect installation or manufacturing defect (early phase) |
| 2 | Electrolytic capacitor (driver) | Wear-out | Temperature (Arrhenius), high operating temperature |
| 3 | Solder joint fatigue | Wear-out | Thermal cycling (ΔT per day × years) |
| 4 | Gasket ageing (IP class) | Wear-out | UV, ozone, heat, compression set |
| 5 | LED lumen degradation (L70) | Wear-out | Junction temperature × time |
| — | Surge / transient | Random | Lightning, grid disturbances — sudden death regardless of age |
¹ Moisture ingress frequently occurs early in the field — caused by installation errors or manufacturing defects — and can also be a late-life gasket wear-out failure. Both mechanisms are documented; the “mixed” classification is more accurate than a pure wear-out label.
Note that LED lumen degradation (L70) sits low in the list. In many installations the luminaire is never replaced for reaching L70 — it has been replaced for something else before then. That is why system lifespan and module lifespan are different things.
The IP class ages
IP66 at delivery means the luminaire met that standard at the time of delivery. Gaskets age regardless of material, but at different rates and through different mechanisms. EPDM is more susceptible to UV and ozone; silicone gaskets maintain their elasticity better at high temperatures and under UV, but share the wear-out mechanism called compression set — a permanent reduction in the gasket’s ability to spring back under sustained compression. In both cases, sealing performance declines over the years. The IP class at year 15 is therefore not the same as at delivery.
The consequence: the design must ensure maintained sealing over time, not only at type-test. Pressure-equalising membranes (PTFE pressure-equalising) equalise air pressure differences that otherwise pump humid air in during temperature drops. This reduces condensation risk without compromising the IP class.
What to ask instead
When a supplier quotes MTBF or L70 as a lifespan claim, the following questions are more informative:
- What is the capacitor B10 lifespan at 70 °C? Temperature is the single most important factor for electrolyte lifespan (Arrhenius rule: every 10 °C increase halves lifespan).
- How many thermal cycles do the solder joints tolerate? A street luminaire typically undergoes 1 thermal cycle per day — switched on at dusk, off at dawn. Over 20 years that amounts to ~7,300 cycles, which is the parameter used to dimension solder joint fatigue lifespan (IPC-9701).
- What gasket material, and how is the IP class re-tested after 10 years? An IP66 type-test at delivery says nothing about year 12.
- Is a pressure-equalising PTFE membrane fitted? If not: how is the breathing and condensation issue managed?
- What system lifespan is guaranteed — not module lifespan? It is the system that is installed, not the module alone.
VALDUR and system lifespan
We report driver capacitor lifespan based on Arrhenius at actual operating temperature, not just a module L-value. VALDUR is equipped with a PTFE pressure-equalising membrane for pressure equalisation and EPDM gaskets dimensioned for Nordic climate. See driver capacitors and lifespan for the details.
Next level of understanding
The capacitor is the single component that most determines when the driver dies. The Arrhenius rule explains why temperature is everything.
Every 10 °C increase in operating temperature halves an electrolytic capacitor’s expected lifespan. It is the single strongest lever for actual system lifespan.
Lifespan & reliability
Driver capacitors and lifespan
The Arrhenius rule, optimal operating temperature, and how capacitors are dimensioned for 25 years.
Frequently asked questions
Can two luminaires be compared using their MTBF values?
Only with great care. MTBF values are calculated differently by different suppliers — methodology differences, assumptions about temperature and load vary. Comparison requires that both use the same standard (e.g. MIL-HDBK-217 or IEC 61709) and the same operating conditions. Furthermore, the MTBF comparison only captures the random failure phase — not the wear-out mechanisms that dominate in the field.
What is B10 lifespan and when should it be stated?
B10 (or L10 for optics) states the time at which 10 % of a population has failed. It is a more honest measure than MTBF when you want to understand when mass failures start occurring in an installation. B10 should be stated for the weakest system components — above all the driver’s electrolytic capacitors at stated operating temperature.
Does a high MTBF always mean high quality?
Not necessarily. A high MTBF means few random failures occur during the product’s middle life. But if components with early wear-out are present in the design — such as an undersized electrolytic capacitor — that does not affect the MTBF figure. It is possible to have both high MTBF and short actual lifespan.
Does L70 apply to the whole luminaire or just the LED module?
L70 (and L80, L90) are defined in IES LM-80 and TM-21 as the module’s lumen maintenance — meaning how long the LED chip retains 70 % of its original luminous flux. It says nothing about the lifespan of drivers, gaskets, or solder joints. Reporting the LED module’s L70 as “the luminaire’s lifespan” is a common error — not a lie, but a missing system perspective.
Summary
MTBF is a well-defined measure of random failure frequency during the product’s middle life. It says nothing about wear-out. L70 is a well-defined measure of LED module lumen maintenance. It says nothing about the rest of the system.
In many field installations — particularly in humid and coastal environments — actual system lifespan is limited by moisture, capacitors, solder joints, or gaskets rather than the LED reaching L70. In dry, well-installed environments, LED degradation can be the primary limiting mechanism. System reliability determines actual lifespan, and system reliability is governed by the weakest link in the chain.
The questions that actually determine lifespan: capacitor B10 at stated operating temperature, number of thermal cycles tolerated by solder joints, gasket compression set value, and the presence of a pressure-equalising membrane.
Sources
- IES. (2015). LM-80-15: Measuring Luminous Flux and Color Maintenance of LED Packages, Arrays and Modules. Illuminating Engineering Society.
- IES. (2011). TM-21-11: Projecting Long-Term Lumen Maintenance of LED Light Sources. Illuminating Engineering Society.
- IEC. (2017). IEC 61709:2017: Electric Components — Reliability — Reference Conditions for Failure Rates and Stress Models for Conversion. International Electrotechnical Commission.
- ReliaSoft. Life Data Analysis Reference: The Bathtub Curve and Product Failure Behavior. HBM Prenscia.
- Valetti, L., Piccablotto, G., Taraglio, R., & Pellegrino, A. (2023). Long-term monitoring campaign of LED street lighting systems: focus on photometric performances, maintenance and energy savings. Sustainability, 15(24), 16910.
- van Driel, W. D. & Fan, X. (Eds.) (2012). Solid State Lighting Reliability. Springer. (LED system failure mechanisms and lifetime prediction.)