What a Lighting Record Has to Show at Warranty Time

Why this matters

Four years after a job, the customer measures the room, gets a number under the specification, and asks you to file a warranty claim. The manufacturer will not argue with your meter. They will ask what the flux of the luminaire is, because that is what their document is written on, and it is not what you measured. Illuminance in a room and flux from a luminaire are two different quantities, and every term that sits between them - dirt, outages, ambient temperature, control setting, whether anything in the space moved - belongs to somebody. A record earns its keep at exactly this moment or it never does, and the fields that do the work are the ones nobody thinks to write down on the day.

The two quantities, and why a room reading cannot reach the warranty

What you measure is footcandles on a plane in a room. What the warranty covers is the luminaire's light output under stated operating conditions. Between them sit every loss the light loss factor product accounts for, and that card owns how the product is built and how a maintained figure is re-based to an initial one.

So a room reading on its own can never substantiate a lumen-maintenance claim, however carefully it was taken. It establishes that the room is short. Attributing that shortfall to the source means accounting for every other term first, and each term you measured or logged on the day is one the manufacturer cannot hand back to you.

That is the design principle of the record. It is not a diary. It is the set of fields that strip the room's losses off a measurement until what is left is the quantity the warranty is written on.

What a warranty conditions itself on

Read the document, because these are the clauses that decide whether you have a claim, and they are conditions rather than promises.

• An ambient temperature limit at the luminaire, which is why the temperature in your space is a warranty field and not a comfort note.
• A stated operating-hours basis, which is why a lighting schedule and any change to it belongs in the record.
• The ordered configuration: lumen package, drive current, driver. Change any of them in the field and you are outside the tested article.
• A control and supply range the product must have been operated within.
• No unauthorized modification, which is where a retrofit lamp put into somebody else's housing lands; the retrofit card owns why that assembly is a different fixture.
• A remedy clause. A product warranty commonly remedies the product and not the lift time, the labor or the light you were short in the meantime. Whether yours does is a question for the document and, where it matters, your own attorney.

Two of those are things you can only prove with data logged while the building was running. Nobody reconstructs an ambient history in arrears.

The record, field by field

Six fields, per luminaire type. Anything else is nice to have.

1. Identity. Full ordering code as installed, quantity, install date, and the location of the units. The code as installed, not as specified, because substitutions happen at the counter.

2. Photometric basis. Which report the design used, whether it is an absolute test of the finished luminaire or a relative one, and the initial output with its test condition. The photometric file card owns what that report contains and what it does not.

3. Design basis. The maintained target with its plane, its height above finished floor and its grid, plus the light loss factor and each term that built it. A design that reaches you as a single number instead of a product cannot be decomposed later, and decomposition is the entire job at claim time.

4. Operating conditions. Ambient temperature at the luminaire, logged rather than estimated; operating hours or the schedule they derive from; the control setting; supply voltage at the far end of the run. These are the warranty's own gate fields, restated onto your side of the fence.

5. Maintenance events. Cleaning dates, replacements, outage counts, and any change to the space: racking raised, a wall repainted, a partition added. None of that is depreciation, and if it is not written down it will be assumed to be.

6. Measurements over time. Every survey with plane, height, basis, grid, extent, meter, daylight state, control setting and date. The repeatable-reading card owns how to take one; this field is about keeping it comparable to the next one.

The filled-in record: a four-year claim on a warehouse

Fields 1 to 3, from the closeout package. 96 units of one ordering code, installed four years ago, absolute photometric report of the finished luminaire with its shipped driver. Design: 30 fc maintained, horizontal at 30 in above finished floor, nine-point grid per bay, light loss factor 0.78, built as lumen maintenance 0.90 at 60,000 hours x luminaire dirt depreciation 0.90 on a 24-month cleaning interval x ambient temperature 0.97 x room surface dirt depreciation 0.99. Initial basis 30 / 0.78 = 38.5 fc.

Field 4, from the logs. Schedule 14 hours a day, 6 days a week, 50 weeks a year = 4,200 hours a year, so 16,800 hours at four years, which is 16,800 / 60,000 = 28 percent of the warranty's hours basis. A temperature logger left at the luminaire plane under the roof deck for one summer recorded a peak of 41 C and 6 percent of operating hours above 35 C. Getting that logger up and down was a lift rated for the height with the fall protection the equipment requires under 29 CFR 1910 Subpart D in general industry or 29 CFR 1926 Subpart M in construction, and it was clamped to the strut rather than to the housing, because nothing gets fixed to an energized luminaire.

Field 5, from the maintenance log. Both scheduled cleanings performed and dated. Two units out of 96 dark at the time of the survey. No racking change, no repaint, no new partition.

Field 6, the survey. 22.0 fc maintained, horizontal at 30 in above finished floor, same grid and extent as the design, taken at 22:00 with the skylights contributing nothing, controls logged at 100 percent.

The decomposition. Actual in-service factor: 22.0 / 38.5 = 0.571, against a designed 0.78. Now strip the terms the record can pin, each measured or logged rather than assumed:

• Dirt, 0.91. One optic cleaned and its station re-read the same night, 26.4 fc before and 29.0 fc after, so 26.4 / 29.0 = 0.910. Essentially at the designed 0.90, which is what a kept cleaning schedule buys.
• Burnout, 0.979. 94 of 96 working, 94 / 96 = 0.9792.
• Room surface dirt, 0.99. Unchanged at design; no surface in the space was altered, which field 5 records.
• Ambient, 0.94. From the manufacturer's derating curve for that ordering code, read at the logged operating temperature rather than at the 25 C bench condition the published output carries. The design had assumed 0.97.

Implied source lumen maintenance = 0.571 / (0.91 x 0.979 x 0.99 x 0.94) = 0.571 / 0.829 = 0.689.

What that number means. The source is at roughly 69 percent of its initial output at 28 percent of the warranted hours, against a warranty of 80 percent at 100 percent of those hours. That is not a maintenance finding and it is not a design finding. It is a claim about the product, expressed in the quantity the product's warranty is written in, and no survey without fields 1 through 5 could have produced it.

What the record also proves against you. Six percent of hours ran above the 35 C gate, peaking at 41 C. That is in your own log and you disclose it, because the alternative is having it found. Whether a bounded excursion voids the claim, pro-rates it, or is treated as immaterial is a question for the warranty document and your own attorney; the operational fact you can bank is that you have a dated, decomposed number and an honest condition history. Those are two different kinds of thing and the record only ever produces the second.

The same shortfall at the site next door

Same chain, same product, same year, different contractor. The file holds one line: "measured 22 fc, under spec."

The manufacturer's first three questions end it. What plane and height, because 22 fc at grade and 22 fc at 30 in above finished floor are different numbers in the same building. What was the control setting, because a scene at 80 percent produces this reading with nothing wrong. When was it last cleaned, because at a 24-month interval that term alone can carry a tenth of the light.

None of the three can be answered in arrears. The shop cannot establish that a shortfall against design exists, let alone whose it is, and the conversation stops before the word warranty is used a second time. The two sites are equally short and only one of them has a claim, which is a difference made entirely by six fields and a temperature logger.

What flips this

Where the space is daylit and the controls respond to it, the operating-hours field stops being a schedule and becomes a duty log, because a luminaire that spends half its life dimmed has not accumulated the hours its runtime suggests, and the warranty's hours basis is a full-output basis unless the document says otherwise. Where the product came through a distributor on a project quote rather than from the manufacturer directly, the party you are claiming against may be the distributor under the sale terms, which is a contract question and belongs with your attorney.

Checking your own figures

• The design product recomputes. 0.90 x 0.90 x 0.97 x 0.99 = 0.7778, reported as 0.78, and 30 / 0.78 = 38.46, reported as 38.5 fc initial.
• The factor is applied once, in one direction. 0.78 divides once to reach 38.5 fc initial. The 22.0 fc survey figure is a maintained field reading and is never multiplied by a loss factor.
• The in-service factor and the stripped product. 22.0 / 38.5 = 0.5714, reported as 0.571. Stripped terms: 0.91 x 0.979 x 0.99 x 0.94 = 0.8291, and 0.571 / 0.8291 = 0.6887, reported as 0.689.
• Every stripped term is measured or logged, and each is printed against its design value. Dirt 0.910 against 0.90; burnout 0.9792 against 1.00; room surface 0.99 against 0.99; ambient 0.94 against 0.97. The ambient term is the one that moved against the design, and it moved in the direction the logged 41 C predicts.
• The hours arithmetic. 14 x 6 x 50 = 4,200 hours a year; 4,200 x 4 = 16,800; 16,800 / 60,000 = 0.28, reported as 28 percent.
• The dirt probe is a same-visit pair. 26.4 and 29.0 fc at one station on one night with one meter, so any fixed multiplicative error in the instrument divides out of 26.4 / 29.0 = 0.910.
• Six fields promised, six fields given, and each of the six is used at least once in the filled-in record: identity, photometric basis, design basis, operating conditions, maintenance events, measurements.
• Every illuminance carries plane, height and basis. 30 fc maintained, 38.5 fc initial, 22.0 fc maintained, 26.4 and 29.0 fc maintained at the probe station, all horizontal at 30 in above finished floor.

References

• The manufacturer's written warranty for the exact ordering code, which owns the ambient limit, the hours basis, the configuration conditions and the remedy, and the derating curve that produced the ambient factor
• IES recommended practice covering light loss factors and maintained illuminance, in the edition your specification or employer standard names, which binds only through that document; an IES recommended level is a design target, not a code minimum
• 29 CFR 1910 Subpart D and 29 CFR 1926 Subpart M for fall protection wherever a logger, a probe or a hand reaches the luminaire plane
• See related: What a Light Loss Factor Is Accounting For; How to Take a Light Level Reading Somebody Else Can Repeat; What a Photometric File Actually Describes; Why a Retrofit Lamp in an Old Fixture Is a Different Fixture; What Lumen Depreciation and Lifetime Ratings Mean