A UV lamp that’s visibly dim gets replaced without a second thought. The expensive failure mode is the one nobody notices — a lamp that still looks bright but has quietly dropped below the dose a bond actually needs, producing weeks of understrength parts before anyone traces a rework spike back to the light source.

Q: How much does an under-cured bond actually cost, versus just replacing the lamp sooner?

A: Replacing a mercury arc lamp or LED array before it’s strictly necessary costs a few hundred dollars in hardware and a short line stoppage. Running it past the point where delivered dose drops below the adhesive’s minimum requirement costs considerably more, just spread out and harder to attribute: warranty claims on parts that already shipped, scrapped work-in-progress once the defect is caught, and the engineering time spent tracing a quality escape back through weeks of production before anyone thinks to check the lamp. Because the intensity decline is gradual rather than sudden, the true cost of running a lamp too long rarely shows up as a single line item — it shows up as a slow creep in the rework rate that’s easy to attribute to something else first.

Q: Why doesn’t a lamp just fail outright when it’s worn out?

A: Mercury arc lamps solarize gradually — the quartz envelope darkens with use, cutting UV transmittance well before the arc itself becomes unstable, and the electrodes lose efficiency at the same time. UV LED arrays degrade through lumen depreciation, a slow drop in output driven by junction heat and current stress, with industrial arrays commonly rated for 10,000 to 20,000+ hours before reaching end of useful life — but “useful life” here means a defined intensity threshold, not a hard cutoff where the array stops emitting light entirely. In both cases, the part coming off the line looks and often feels normal well after the cure has become marginal.

Q: What’s the actual threshold for replacement, rather than a fixed time interval?

A: Time-based replacement schedules are a rough proxy at best, because two identical lamps run under different duty cycles, ambient temperatures, or duty schedules degrade at different rates. The reliable threshold is measured, not scheduled: replace or aggressively compensate for a light source once measured irradiance falls to roughly 70 to 80% of its baseline “new lamp” reading, assuming production cycle time is already optimized around that baseline. Reaching that threshold is only detectable with a calibrated radiometer taking readings at the exact working distance used in production — a visual inspection or an elapsed-hours counter alone will not catch it reliably.

Q: What does a proper measurement program actually involve?

A: Three steps, done in order and repeated on a schedule. First, establish a baseline: measure a brand-new lamp’s intensity at the exact production working distance and record it as the reference point. Second, monitor routinely: a daily or per-shift visual check of the light guide and lens for dirt or scratches catches obvious problems, but a weekly or monthly radiometer reading at that same fixed distance is what actually tracks the decay curve over time. Third, act on the trend rather than a single reading: a downward trend approaching the 70–80% threshold is the signal to schedule replacement before the next production run, not after a defect is already found downstream.

Q: Can a declining lamp be compensated for instead of replaced immediately?

A: Within limits, yes, and this buys time without accepting under-cured parts. On a conveyor system, slightly reducing belt speed increases exposure time and restores the delivered dose (J/cm²) even as peak intensity has dropped. On LED systems that support it, increasing drive current compensates for lumen depreciation, though this trades off against the array’s remaining service life and generates more heat that needs managing. Both are legitimate short-term measures; neither is a substitute for eventual replacement once the compensation itself runs out of headroom.

Q: How does dose relate to intensity, and why does the distinction matter here?

A: Intensity (irradiance, in mW/cm² or W/cm²) is the instantaneous brightness of light hitting the part; dose (J/cm²) is intensity multiplied by exposure time, and it’s dose — not intensity alone — that determines whether a bond fully cross-links. A worn lamp with lower intensity can still deliver an adequate dose if exposure time is increased to compensate, which is exactly what slowing a conveyor accomplishes. Confusing the two is a common source of unnecessary lamp replacement, or conversely, of running a genuinely under-dosing process because intensity alone looked “close enough.”

Q: What does Incure provide beyond the bulb or LED array itself?

A: Incure’s L-Series™ UV LED flood lamps pair the light source with integrated thermal management that slows the rate of lumen depreciation in the first place, and several configurations include automatic intensity feedback that increases drive current as the array ages, holding delivered dose steady across thousands of hours without operator intervention. Beyond hardware, Incure helps establish the baseline-and-trend measurement program described above, matching wavelength to a specific adhesive’s photoinitiator and defining the dose threshold that actually matters for a given bond line thickness — a related discussion of how lamp output decline plays out for light-guide-coupled spot lamps specifically is in what causes UV light guide degradation over time.

Selecting the right lamp technology in the first place — LED versus mercury arc versus electrodeless — is a separate decision covered in full in UV curing light source: an industrial guide; the measurement program above assumes that choice has already been made and instead focuses on keeping a chosen system performing at spec over its service life.

If you’re not sure whether your current replacement schedule is catching intensity decline before it affects cure quality, Email Us with your current radiometer readings, or lack thereof, and we can help you set up a baseline-and-trend program from scratch.

Moving from a fixed “replace every N months” schedule to a measured, threshold-based program is what actually protects bond quality — and it usually costs less than the rework it prevents. Contact Our Team for a technical consultation on UV system maintenance planning.

Visit www.incurelab.com for more information.