When lights keep failing in a plant, the fixture is usually the symptom, not the cause. A fixture that dies in eighteen months on a line running three shifts is telling you something about the air around it, the power feeding it, or the classification of the room it hangs in. Ordering the same model again buys the same failure on the same schedule. This checklist runs through the conditions that actually shorten fixture life on an industrial floor, in the order a distributor works through them on a quote. Some answers sit in the industrial lighting range and some in the hazardous location range, and half the job is knowing which side of that line the room falls on.
When lights keep failing, describe the failure precisely
Three questions separate most causes on their own.
- What stopped working? A dead driver, a strip of dark LEDs, a cracked lens, and a fixture that still lights but has gone dim are four different failures.
- How many, and where? Failures clustered near an oven, a press, or a dock door point at the environment. Failures scattered evenly across the building point at power.
- How long did they last? Units dying inside a year point to heat or electrical stress. A bay quietly going dim over three years is lumen depreciation, not a fault.
Have those answers before you look at a catalog.
Cause 1: The air is hotter than the fixture is rated for
This is the most common cause on a plant floor and the easiest to prove. Every LED fixture carries a maximum ambient operating temperature, the Ta value, and it is measured at the fixture, not at the floor where people stand. Air near a furnace, an oven line, or a curing tunnel sits far above the general room reading.
The rating also moves with wattage on the same fixture. The UFO-I high bay is a clear example: the I4 model is rated to a Ta of 80°C at 100 W and 120 W, 75°C at 150 W, and 70°C at 180 W. Same housing, three different ceilings depending on how hard it is driven. Ordering the highest wattage in a family and assuming the top ambient rating comes with it is a common and expensive mistake.
If the mounting location runs hot, the fixture has to come from a line with a printed ambient rating, which is what the high-temperature lights category exists for. Our high-temperature buyer guide covers how to verify one.
Cause 2: Dust is insulating the heat sink
An LED high bay is cooled passively. The fins on top are the only path the heat has out, and a layer of dust across them works like a blanket. A fixture correctly spec'd for a 55°C ceiling can run past its design temperature after two years in a grain, cement, wood, or foundry environment with nothing visibly wrong with it.
Two things fix this: a cleaning interval matched to the dust load instead of a generic annual walk, and a housing designed against accumulation. The UFO-I uses a sloped back that sheds dust outward from the center rather than collecting it between fins. If maintenance cannot safely reach the fixtures, that belongs in the spec conversation.
Cause 3: Water and dust are getting inside
Ingress failures look like corrosion inside the lens, condensation that never clears, or a driver that quits after a wash cycle. The usual root cause is an IP rating chosen for the wrong threat.
IP66 means dust-tight and protected against powerful water jets, which covers rain, hose-down splash, and heavy airborne dust. It does not cover a high-pressure, high-temperature sanitation cycle, which is a different test entirely. If a hose, foam cannon, or steam wand reaches the fixture on a schedule, that is a washdown zone and it belongs to the food-processing line. Our IP69K vs IP67 guide sets out what each rating proves.
The other half is the cable entry. A correctly rated fixture with an unsealed or wrong-size gland fails at the gland, and the fixture takes the blame.
Cause 4: Vibration and impact are cracking things
Presses, compressors, overhead cranes, and rail lines put continuous vibration into a structure, and vibration loosens hardware, fatigues solder joints, and works gaskets loose. Forklift masts and hoisted loads add direct impact.
The spec to look for is the IK code, which rates impact resistance. The EX-B high bay is rated IK08 with its glass cover and IK10 with the wire guard fitted, and it uses stainless steel hardware, the kind of detail that decides whether a fixture survives a decade of vibration. The EX-O2 jam jar is rated IP66 and IK08 and ships with a metal guard as standard. Where impact is a known risk, order the guard with the fixture.
Cause 5: The power is the problem, not the light
If failures are scattered across the building rather than clustered in one hot or dirty area, look at the electrical side. Drivers fail on surge, on sustained voltage outside their input band, and on switching transients from large motors and welders sharing the service.
Two specs matter. The first is surge protection: the UFO-I carries 6 kV built in, and on a site with heavy inductive loads that is a floor, not a bonus. The second is the input voltage band. That fixture accepts 120 to 277 V or 277 to 480 V depending on the model ordered, so the ordering code has to match the measured service voltage at the panel, not the voltage someone assumed. Where a plant has a history of driver failures, add surge protection at the panel as well.
Cause 6: The air is corrosive
Chemical processing, wastewater headworks, pulp and paper, coastal sites, and battery rooms put compounds into the air that attack fasteners, gaskets, and coatings long before they touch the LEDs. The failure shows up as seized hardware, then a corroded gasket surface, then water ingress through the joint it used to seal.
The specs that answer this are the housing material and the salt-spray hours. Copper-free aluminum is the housing spec on hazardous location fixtures like the EX-B and the EX-O2 for that reason, and salt-spray testing gives a comparable number: the UFO-I is published at 500 hours, the EX-B at 1,000 hours. Neither is a lifespan. They let you compare candidates on the same test instead of on marketing language.
Cause 7: The area is classified and the fixture is not rated for it
This is not a reliability question, it is a compliance question, and it outranks everything above. If the room is classified on your area-classification drawing, the fixture has to carry the matching listing regardless of how well built it is.
Two details catch people out. First, the class and division have to match. Class I, Division 2 covers flammable gas or vapor present under abnormal conditions, combustible dust falls under Class II, and fibers under Class III. The EX-B is listed under UL 844 for Class I, Division 2, Groups A, B, C and D, Class II, Division 1, Groups E, F and G, Class II, Division 2, Groups F and G, and Class III, which is why it suits plants where gas and dust zones share a building.
Second, the T-code moves with ambient temperature. The EX-O2 is rated T6 up to 25°C ambient, T5 up to 40°C, and T4 up to 60°C, so a hot classified room needs both numbers checked together. Our classifications guide explains what each label covers.
Cause 8: Nothing failed, the light level dropped
Sometimes the complaint is not that fixtures are dead, it is that the floor is too dark, and two causes look identical from below. One is lumen depreciation, which is why L70 figures are quoted at a stated ambient: the EX-B publishes L70 above 150,000 hours at 55°C, and running hotter shortens it. The other is a layout that was never right, or that stopped being right when racking went up or the line was rearranged.
Put a light meter on the work plane before ordering anything. If the reading is well below the design level and no fixtures are out, the fix is spacing and beam angle, not replacement units. Send your dimensions through the photometric layout request and you get a calculated fixture count and spacing plan back rather than an estimate.
What to have ready before you reorder
- The failure description: what failed, how many, where in the building, and after how long.
- Measured ambient temperature at the fixture location, not the room average.
- The washdown or dust exposure, including cleaning method and frequency.
- Service voltage at the panel and any history of driver failures elsewhere on site.
- The area-classification drawing for any classified zone, with class, division and group.
Send those five and we will match a fixture against the failure, or say plainly when the fixture was fine and the layout was not.
Common questions
Why do LED lights keep failing in my plant when the same model works fine in the office? Because the two rooms are different environments, not because the fixture is defective. Ambient heat, dust on the heat sink, vibration from process equipment, corrosive air, and switching transients from large motors are all present on a production floor and absent in an office. Measure the ambient temperature and the service voltage at the failing location, then compare both against the fixture's published Ta and input voltage band.
How do I tell whether heat or power is killing my fixtures? Map the failures. Units failing in a cluster near an oven, a furnace, a press, or a dryer point at ambient heat. Units failing evenly across the whole building, including cool areas, point at surge or voltage. Failures right after every wash cycle point at ingress. The distribution usually identifies the cause before anyone opens a fixture.
My fixtures still light up but the floor is darker than it used to be. Is that a failure? Usually one of two things. Genuine lumen depreciation, which is what L70 ratings describe and which accelerates when a fixture runs above the ambient its rating was quoted at, or a layout that no longer matches the room after racking, a mezzanine, or a line change. Take a meter reading at the work plane before ordering replacements, since new fixtures in a bad layout will look dim too.
Can I use a standard industrial fixture in a classified area if it is well sealed? No. A hazardous location listing is a certification against a defined test standard, not a measure of build quality, so a well-sealed industrial fixture is still not code-legal in a classified zone. Check the class, division and group on your area-classification drawing and match the fixture's listing to it. The hazardous location range is organized by classification for that reason.
How often should plant light fixtures be cleaned? That depends on the dust load, not the calendar. A clean assembly area may need only an annual inspection, while a grain, cement, or foundry environment can insulate a heat sink within months. Set the interval against how fast accumulation shows on the fins.
Sources: NemaLED product spec sheets for the fixtures referenced above (UFO-I, EX-B, EX-O2), IEC 60529 ingress protection definitions, and UL 844 hazardous location listing categories as printed on the referenced spec sheets.
Need a fixture spec'd to your environment? Send the room dimensions, classification, and ambient, we'll come back same day.
Request Quote