The DOE has issued a report that tries to lay to rest fears that LED-based light is any more harmful to humans or materials than other light sources.
The US Department of Energy last put an oar in these waters a year and a half ago, when Spanish researchers claimed that light from LEDs is capable of damaging human retinal tissue. The DOE issued a report, Optical Safety of LEDs, that added to the chorus of voices rebutting the claims of a blue-light hazard in LED-generated light (we covered the news here).
Now the DOE has published a more in-depth report, True Colors, that aims to lay to rest once and for all three related concerns about LEDs. The subtitle lists them: "LEDs and the relationship between CCT, CRI, optical safety, material degradation, and photobiological stimulation."
"Optical safety" refers to the blue light hazard, and the report quantifies the situation using the spectral weighting function, KB,v. "Material degradation" refers to damage such as fading that might be experienced by, e.g., fabrics or paints; it is characterized by the CIE Damage Function, Sdf. "Photobiological safety" is about the ability of light, specifically blue light, to alter human circadian rhythms. No spectral weighting function has been agreed upon for this relatively recent concern; the DOE's paper uses the functions Mγ ("one of several proposed efficiency functions for melanopsin") and CS (from a 2005 paper by Rea, Figueiro, Bullough, & Bierman).
All blue-pumped, phosphor-converted LEDs show a pronounced peak in the blue region in their spectral power distributions (SPDs). The question is whether there is sufficient energy in this peak to trigger these various concerns about damage to humans and materials,
The DOE first cautions that a common source of confusion related to LEDs' SPDs is that they are often presented as relative SPDs, with the peak scaled to one. Since quantity of light, as well as spectral distribution, is involved in any calculation of light's effects, SPDs must be compared on an absolute basis. The absolute SPDs in the figure (representing 20 CALiPER-tested products with CCTs between 2700K and 6500K, and CRIs between 62 and 98) have been normalized to equivalent luminous flux.
The core of the report's findings are presented in three graphs on page 5. They show, for five light sources, blue light hazard efficiency (KB,v) vs. CCT, CIE spectral damage potential (Sdf) versus CCT, and melanopic flux versus CCT. For all three potential types of damage, blue-pumped LEDs are low in the pack; violet-pumped LEDs are toward the high end. In all cases blue-pumped LEDs lie below the data points for blackbody radiation and for models of daylight (CIE D Series) and fluorescent light (CIE F Series).
Here is the summary for photobiological effects:
The photobiological effects of light are related to the spectrum and intensity of light, but are not specific to any type of light source. Especially when nighttime exposure is a concern, choosing lower-CCT sources will generally reduce the photobiological risk potential. In critical applications, evaluations beyond CCT are warranted.
The report also looked at different modes by which LEDs produce white light: blue-pumped, hybrid (phosphor-converted plus red), violet-pumped, and color-mixed. The report notes, "While the difference between the LED products [of different light-producing modes] can be substantial (up to 26%), none of the products exceeds blackbody radiation by more than 8% for any of the risks considered."
I have not seen anything in writing taking issue with the DOE's methodology or conclusions. (If you have seen such, please let us know in the comments.) Perhaps we can now dial down the panic level about the blue peak characteristic of LED-produced light.
— Keith Dawson , Editor-in-Chief, All LED Lighting