LED-based street lights remain a controversial topic throughout the globe with about the only universally-accepted belief on the solid-state lighting (SSL) topic being that the LED retrofits can save energy. One of the complaints about outdoor lighting has been that the products increase the sky glow, especially with short-wavelength blue light, that is visually unpleasant to humans and problematic for scientists such as astronomers. But a new US Department of Energy (DOE) research project, “An investigation of LED street lighting’s impact on sky glow,” largely refutes the sky glow concerns that are specifically attributable to street lights, although the overall increase in urban lighting in general may be another story.

Before digging into this story, you may want some background on the sky glow topic. We had an excellent feature two years back written by scientist and entrepreneur Ian Ashdown. Ashdown revealed computer models and theories that predict sky glow and the general science behind the issue. Moreover, the new DOE report credits Ashdown among others for contributing to the new research.
 
It’s important to note that the DOE research, while broad in terms of the different conditions it studied, was narrow in scope. It focused on LED and legacy street lights that are typically installed across the US. For example, it didn’t contemplate what might be achievable with a new high-pressure sodium (HPS) luminaire with improved optical design intended to remove uplight. Likewise, the study focused on the typical LED luminaires being installed in place of HPS or other HID fixtures at present, and not on some new SSL luminaire architecture. Also remember that the study was really focused just on street lighting with accepted accounting for reflectance off street-level surfaces. But it did not consider buildings that might cause additional uplight via reflections or indeed other sources of uplight, such as indoor spill, that is often prevalent in urban areas.

The DOE project relied largely on the SkyGlow Simulator model that has been widely used in the astronomical community, and that was developed by Miroslav Kocifaj from the Slovak Academy of Sciences. The work relied on approximately 200,000 computer runs that generated the results. The methodology centered around sets of tests evaluating one variable while other characteristics were kept constant. The modeling assumed ground reflectivity of 15% — a value commonly used in atmospheric modeling. Four levels of uplight were considered and two levels of lumen output. Perhaps the most significant variable was spectral power distribution (SPD) where ten SPDs from typical street-light products were evaluated.

The model returns sky glow impact on the horizontal plane of an observer in terms of diffuse irradiance or scotopic lux (lx). And the DOE has reported the results of the unweighted radiant power and the scotopically-weighted sky glow. Conditions for a near observer located at the edge of a typical city were reported along with those from a far observer located 40 km from the center of the city.

The results fill pages in the report as you may expect, but here is the summary. In the unweighted case, LED street lights reduce sky glow at the near and far observers relative to all legacy products. In the unweighted tests, the results depend on the specific sources tested. The LEDs produce as little as 20% of the legacy baseline and as much as 160%. When measured at 40 km, the uplight characteristics of the LED products, with uplight curtailed by optics that direct the beam downward, reduce the contribution of street lights to sky glow by 95%.

The DOE did note that the industry can do better moving forward. For example, specifiers can choose LED street lights with SPDs that have less short-wavelength energy. And remember that SPD and CCT are not equivalent metrics. As we have reported, there has been a swell of public opinion, and even a recommendation by the American Medical Association (AMA), to limit CCT to 3000K or warmer. The DOE report states that a lower CCT doesn’t always deliver a preferred SPD, and good lighting design practices that minimize the number of fixtures installed along with autonomous dimming can be far more valuable than CCT limits in reducing sky glow and assuring healthy lighting for humans.