Lumens, Lux and Efficacy: Reading an LED Datasheet Properly

A lighting datasheet can run to four pages, but most specification decisions come down to three numbers: how much light the fitting produces, how much of it lands where it is needed, and how much electricity it takes to do so. Those are lumens, lux and efficacy respectively. They are related, but they are not interchangeable, and a surprising amount of disappointment on site traces back to one being read as if it were another.

Lumens — how much light leaves the fitting

The lumen (lm) is the unit of luminous flux: the total quantity of visible light a source emits in all directions, weighted for how sensitive the human eye is to different wavelengths. It says nothing about direction and nothing about the space the fitting is going into. A 4,000 lm luminaire is a 4,000 lm luminaire whether it is mounted at 3 m in an office or 12 m in a warehouse — but the light on the floor in those two cases is very different.

The figure to look for is luminaire lumens — sometimes labelled delivered, system or output lumens — rather than LED or source lumens. LED manufacturers rate their chips at a laboratory junction temperature with no optics in front of them. Once the chips are mounted in a housing, run at a realistic temperature, driven by a real driver and covered with a diffuser or lens, some of that light is lost. A well-designed luminaire keeps those losses small; a poor one can lose a quarter or more. If a datasheet quotes only a source figure, ask for the luminaire figure measured to LM-79 or BS EN 13032, because that is the number a lighting design will actually be calculated with.

Lux — how much light arrives on the surface

Lux (lx) is illuminance: lumens per square metre falling on a surface. It is the unit that standards are written in. BS EN 12464-1 asks for 500 lux maintained on an office desk, 100 lux in a corridor and a range of values across a warehouse depending on whether an aisle is occupied and what is being read there. Lux is a property of the installation, not the product: it depends on the lumens available, the mounting height, the beam distribution, the reflectance of the walls and ceiling, and how many fittings there are.

Two things about lux catch people out. The first is the inverse square law: double the distance from the source and the illuminance falls to a quarter. A fitting that gives 300 lux at 4 m gives roughly 75 lux at 8 m. The second is the word maintained. Standards specify the illuminance that must still be there at the end of the maintenance cycle, after the LEDs have depreciated and the diffusers have picked up dust. Designers apply a maintenance factor — often 0.8, sometimes lower in dirty environments — so a scheme that needs 500 lux maintained is designed to deliver around 625 lux on day one.

Efficacy — how much light per watt

Efficacy is lumens per watt (lm/W), and it is the number that decides the running cost. As with lumens, the useful figure is luminaire efficacy: luminaire lumens divided by the total circuit watts including the driver, not chip lumens divided by chip watts. The gap can be significant. A chip rated at 200 lm/W can end up in a luminaire that delivers 120 lm/W once optical and driver losses are counted, and both figures may legitimately appear somewhere in the literature.

Efficacy also trades against other qualities. Warmer colour temperatures and higher colour rendering both cost a few percent of efficacy. Low-glare optics, which redirect and diffuse light, cost some more. A fitting with the highest lm/W on the comparison table is not automatically the right one; it may simply be the one with the least going on in front of the LEDs.

Reading the datasheet in practice

• Check which lumens are quoted. Luminaire (delivered) lumens are the design figure; source or LED lumens are a component rating and will overstate what reaches the space.
• Check the test conditions. Lumen output and efficacy are normally stated at 25°C ambient. A high bay running at 40°C under a roof in July will deliver a little less — a credible datasheet will show a derating curve or state a maximum ambient temperature.
• Do not compare a lux figure on one datasheet with a lumen figure on another. Lux depends on the installation; the only fair product-to-product comparison is luminaire lumens and luminaire efficacy under the same test standard.
• Ask for the photometric file. An IES or LDT file lets a designer calculate real lux levels for your actual room, mounting height and layout, which is the only way to know whether a scheme will meet the standard.
• Look at colour rendering and colour temperature alongside efficacy. The cheapest lumens are usually cool, low-CRI ones, and there are spaces where that is the wrong trade.

Lumens describe the product. Lux describes the installation. Efficacy describes the bill. A specification needs all three, and it needs to know which is which.

None of this is difficult once the three quantities are kept apart. The habit worth building is simple: when a number appears on a datasheet or a quotation, ask what it is measuring and under what conditions. A manufacturer who can answer that clearly — luminaire lumens, at stated temperature, to a named test standard, with a photometric file available — is giving you something a lighting design can be built on. One who cannot is giving you a marketing figure, and the site inspection will not be interested in those.