The lumen’s limits: why old measures of light output don’t always work for LEDs, and what to do about it

Whether you’re a facilities manager, an office manager, or a homeowner, you’re going to have to change a lightbulb at some point in your life and, these days, more likely than not, you’ll be swapping an incandescent or fluorescent bulb for an LED. You’ll want to buy the right bulb; one that won’t burn your retinas or make your kitchen feel like a high school chemistry lab. To do that, you’ll want to know watt you’re looking at (don’t worry, we limit ourselves to one cheap pun per article).

The history of measuring lighting volumetrically dates back to the “candela” or the light emitted by one candle. Since the candle was replaced by the incandescent bulb, the lumen has long served as the benchmark lighting industry measurement of light volume. A 40-watt filament light bulb used in a bedside lamp emits 450 lumens. A 32-watt, 4-foot long fluorescent tube emits 2,600 lumens. An industrial high-bay light emits upwards of 50,000 lumens.

This unit of measurement has been a necessary constant as we’ve replaced old technology with new. Lumens are designed to be the apples-to-apples of the lighting world. When replacing one light source with another, you need a medium of exchange, and lumens exist to serve as that medium. However, LEDs may be the end for the old faithful lumen. In practice we’ve found the lumen to be a significantly inaccurate indicator of brightness where LEDs are concerned.

In short, the spectral distribution of LED lighting (ie. the way LEDs are able to approximate natural light) forces a rethinking of the way light is measured to account for the sensitivities of the human eye.

First, let’s return to the lumen. In order to convert a light source to lumens, the wavelength reading of the light emitting from the source is multiplied by the human eye’s sensitivity to each of those wavelengths. This, the theory goes, allows us to answer the question “how bright is this light?” no matter its color composition.

Our experience replacing countless incandescent lights, more and more fluorescent lights, and, even, metal halide and other high-intensity-discharge lighting, has revealed a shortfall in the value of the lumen.

“The more fluorescent work we do, the more we notice that equal brightness is delivered with fewer lumens from an LED.”

Where you would theoretically just match the LED lumen output with the lumen output of the conventional light source, we’ve found that you need to apply a discount to the lumens emitted by old lighting before applying that lumen number to LEDs. With halogen and incandescent lighting, for example, you need to reduce the lumen count by at least 10-20% in order to avoid excessive brightness. A 90-watt halogen PAR30 that emits 1,300 lumens should be replaced by 1,100 lumens of good LED light.

This slight lumen discount, however, has proven to increase as the quality of the original light source worsens.

Fluorescent lighting has a bad reputation—and for good reason—it hurts your eyes and is filled with mercury. Though many tolerate it in return for its low cost and longevity benefits, the longer a fluorescent burns, the more likely it is that chemical changes will occur and its color will change, making it less white and less bright (or worse, purple). The more fluorescent work we do, the more we notice that equal brightness is delivered with fewer lumens from an LED.

Meanwhile, as we venture into the replacement of high-intensity-discharge (aka metal halide or high pressure sulfur) lighting, which is considered to be the worst artificial light and intolerable for most interiors, we find that the lumen discount is even greater. Why?

The parts of the eye that sense colored light, cones, are grouped according to the wavelength, or color, to which they are sensitive. But our cones get tired. When a specific wavelength is emitted in high intensity, the relevant cones become fatigued and less sensitive to that wavelength, making the light seem dimmer as a result.

The lighting industry has grown up around the lumen and related measures like the footcandle and candela in an effort to standardize light output. However, none of these measures, to our knowledge, account for cone fatigue. LEDs are capable of emitting light evenly across the visible spectrum and are, therefore, brighter than existing industry measures suggest.

Our work involves anticipating these changes. In practice this means adjusting the fixture brightness and/or fixture count within a space to account for the difference in the brightness of the LEDs.