US cities take second look at LED after health warning

LA has has more street lights than nearly any other city in the world and many of them are LED.

US cities are taking a second look at LED street lighting after the American Medical Association (AMA) issued a warning earlier in the year that blue light can disturb sleep rhythms and exacerbate serious medical conditions.

The AMA highlighted high intensity streetlights, such as those found in Seattle, Los Angeles and New York as being potentially dangerous.

Nearly thirteen percent of US roadway lighting is now LED and the technology is usually found in major US cities. LED's increased lifespan and their energy efficient nature are both powerful convincers to local government looking to save money, however the warnings are taking their toll.

While some cities, such as Seattle, have been very dismissive of the health reports, other cities, such as New York, have reacted by switching to lower-intensity LED bulbs, which the AMA considers to be safe.

Lake Worth in Florida is embarking on a plan to replace sodium street lights with more than 4,000 LEDs that have less of the blue light that the AMA has highlighted as being harmful.

In Gloucester in Massachusetts, the city authorities have conducted a consultation and decided to install LEDs which emit less blue light. Next month, the city is expected to finish installing its LEDs, but they will be 3000K rather than 4000K.

  • You can find out more about street lighting at this year's Lux Live. The exhibition will take place in London on Wednesday 23 November and Thursday 24 November 2016. You can find​ out more by clicking here

Comments 9

Evelyn, Nicely written. Your comments point out the absurdly speculative nature of an overall position on a nascent technology like SSL. SSL is evolving into a symbiotic hosts for increasingly sophisticated sensors that already do much more than control lumen output. The internet is the IOT and SSL networks allow it to more fully manifest in three dimensions. The genie is out of the bottle. Thank you again for your words.

This is hogwash. Next article, going to grocery stores, work, and anyplace with poor CRI and high CCT fluorescent lighting will cause you to become a vampire. A 2700-300K 'warm' LED is more pleasant to the eye, but the blue energy in the spectrum is still there. Depending on the spacing of the street lights, the illuminance one is subjected to could contain even more blue light than the 4000K street lights they are saying are bad for you.

Funny American. The standard CIE S 009 was being pusblished by them at the beginning time, and they always said that the European Standard EN 62471: 2008 was made a copy from the CIE standard. I am wondering if the US government already given up the classification for the Retinal Blue Light hazard? And the other point is that, the non-uniform spectrum of CDM-T / CFL / Fluorescent Tube, but why they still can be used in hospital? As a professional organisation I think AMA shall study into it and shall know more about the relation between CCT vs Blue Light, and should provide data to proof their point of view. And in this passage why no one talk about the new limitation of Blue Light hazard in BS EN 60598-1, BS EN 62471 and EN 62778? These standards provinding systematic classification system and limitation of use, testing considering CCT / light intensity / distance between user and the light source. Lastly, the simpliest way to lower down the photobiological hazard which bring to human, just let the city goes off at the night time, then the problem solved. #humancentriclighting

I continue to find this topic intriguing. We're all worried about waste light from road lighting (a soft target IMHO) and the risk associated with blue light that is apparently so much worse with LED technology. Why no-one has yet looked at the elephant(s) in the room - waste light from commercial premises without curtain contributes more to sky glow (UWL) than any road lighting installations, particularly those with well designed LED luminaires. Also we are exposed to much greater dosage of blue light from interior fixtures than any outdoor installation. I can accept the biomedical research into blue light risk, but lets target the true culprits not just pick the easy one.

This story goes from bad to worse. Exposure depends on intensity which has little to do with lamp lumens as light distribution and distance from the source are substantial factors; in fact, lamp elevation and/or pole spacing is based on a calculation of ground level intensity. As some point out, ambient light under daylight conditions is rich in blue light - depending on latitude and elevation it ranges from 5600K to 6300K CCY - while LED streetlighting has less than 1/200th of the blue light content of daylight. The article suggests that there is a known activation curve for melatonin suppression (" the peak sensitivity occuring between 446-477 nm") when this has not been established and it has at least been shown that intense light of any spectrum tends to suppress melatonin and that diurnal exposure profile determines LIMS threshold while exposure to reduced levels of blue rich fluorescent light - various studies 250 to 500 lux as compared to street light spec variously 6 to 15 lux - produce no measureable LIMS. One must wonder why mercury, mercury-xenon, etc which all produce more blue light and considerably more harmful UV than LEDs has been given a pass also considering that their uneven intensity profile results in a much brighter central spot. If one were truly concerned with health effects, one would consider studies that show a 15-20% increase in traffic accidents and ~30% increase in pedestrian injuries under HPS street lighting as compared to white(ish) street lights; the fact is that HPS lighting impairs color recognition and peripheral vision and, considering the LIMS hypothesis, does little to prevent driver drowsiness. As for an LIMS activation spectrum, one must discount certain claims which ignore the transmission spectra of the cornea, fluid and lens of the human eye and/or how blue light transparency declines with cumulative exposure (e.g. age). It should be noted that many LIMS studies are flawed particularly those which use fixed spectrum light in combination with tinted glasses and the majority that do not calibrate using instrument quality spectrophotometers and the far too many that use no metrology at all.

The problem of which CCT has to be used exist, most likely, because municipality trying to find universal solution to simplify inventory. But “one size (and one CCT) fir all” does not always well applicable in urban environment. Mesopic factor makes sense for Highways, where hardly any pedestrian presence and speed is higher than 25 m/h. But if City streets have a speed restriction at 25m/h, than multiplier hardly applicable, and high CCT only cause complains in residential areas about circadian sleep rhythms disturbance.

Dear Big Matt and others- The AMA is echoing statements given by the LRC (Marianna Figiuero & Mark Rea), as well as the Chairman of the IES Human Health and leading pioneer in this field George Brainard. Your statement of the 4000K being naturally present is true in daylight, but that's not nighttime. Clearly, all available research shows that darkness is the most optimal environment for preserving the healing response, and darkness by its very definition is no light at all. Blue light is clearly shown to suppress nocturnal melatonin production, the peak sensitivity occuring between 446-477 nm. You can't say that there is still 50% blue in the spectral array at 3000K without discussing CRI and the spectral flux graph is the most honest way beyond the antiquated CCT/CRI ratings. Certainly, Chile is the best example - they mandate 15% or less blue in the spectral array, which can be achieved in some 3000K lumenaries LED or otherwise. Mark and Marianna's cautions simply relate to a more stringently define dosimetry - accounting for duration of exposure, angle of incidence, etc... That was clearly stated in the recent Architectural Record article which quoted them. Any statements to efficacy of LED luminaries is assumptive at best without the CRI factoring in, and still fails to take into account the message I've been spreading for nearly 10 years now, which is: how can you consider a lighting design efficacious if it creates ill health? It drives up health care costs, decreases quality of life and speaks to creating a host of diseases - obesity and cancer as standout examples here. My comments past this is that municipalities can save so much money by simply reverting to the RP-8/GL-6 guideline, which states that the minimum illumination recommendations can be met by vehicular headlights when they are traveling 25mph or more. To evenly illuminate a freeway of cars normally going 65mph, is counterproductive and a waste of taxpayer funds. LA might have a point due to traffic jams, but I'll argue that fewer vehicular collisions occur when they're all sitting on the freeway. The grid is less stressed when these luminaries require to be on, at night when everyone is sleeping - its not a peak demand time. Use the blue rich LED during the day, that's when its needed. It creates less stress on the grid at peak demand times, plus bolsters health. To Barry's point: I don' think the cost of tunable white can be justified. All light is alerting, so warm white is still the best option. Warm white (properly specified without all the blue) can be alerting but avoids suppressing nocturnal melatonin.

Artificial exterior lighting above 4000K has been in existence for close to 150 years. Natural exterior lighting above 4000K has been around for 4.5 billion years. The AMA simply hiccupped misinformation spread by the IDA due to their concerns about viewing the stars from the city. Unfortunately the IDA goals stand in direct competition with global warming goals as 3000K LED fixtures require roughly 40% more energy than 5000K fixtures due to their lower efficacy and their lower mesopic multiplier outlined in the IES design procedures. If you want an independent point of view, look at the Lighting Research Center's response to the AMA statement. Ironically, the 3000K fixtures still produce 50% as much short wavelength light (when adjusted for the additional wattage required) as the 5000K fixtures (so both are harmful if short wavelength light at night is truly bad for you).

Wouldn't Variable CCT LEDs and correct Color Index -Gamut Index etc. #HumanCentricLighting measures alleviate most of these issues? After all, we can't have Phase Shifts from 6000K Fixtures creating Insomniac Drivers--#TheRollingDead.

Leave your comment