It’s Time For a Sodium-Free Illumination Diet!
By Steven Oster
 

Traditional lighting technologies such as high and low-pressure Sodium for street and security lighting have accumulated a large body of illumination data and standards over the last fifty years. Architects and lighting Engineers consult published data to determine which sources and luminaires will provide the required illumination for their projects. They go by the book, and it is a well established book. But as high-brightness white LEDs begin to compete with legacy technologies for outdoor lighting, it’s time to add a new chapter to the book…or perhaps to write a new book. Because comparing rated Lumen output for current technologies such as Sodium to LEDs is like comparing apples to oranges; it does not correspond.

As a designer of LED light engines for outdoor use I first noticed this more than a year ago. LED luminaires outputting a fraction of the total Lumens emitted by Sodium sources were actually as good or better. How could this be? I kept re-checking light meter readings and wondering what I was observing; it didn’t make sense. Yet the results were consistent. A recent example is a 120W high-pressure Sodium unit replaced by a 58W neutral white LED luminaire. The white LED effective illumination is better, even though the Lumen output of the Sodium bulb was much greater. And beyond my personal observations, studies comparing LED street lighting to high-pressure Sodium are revealing the same phenomenon. Part of the explanation, as others have pointed out, is the directional nature of LEDs. The light engine is positioned within the luminaire to radiate in the desired direction, using no reflector and thus eliminating the typical 30% to 40% reflector loss. The light generated is output more efficiently; so far so good, but we’re not there yet.

How can LEDs be so effective with not just 40%, but as much as four or five times less luminous output? A November, 2008 study1 in Oakland, California for street lighting compared LEDs to existing high-pressure Sodium and reveals the LED illumination as good or better even though measured Lumens per square meter (Lux) on the ground was less than for the HP Sodium source. The rest of the answer is found in the nature of the light itself and the sensitivity of the eye to the wavelengths emitted. There are two light-sensitive components in the retina of the eye, rods and cones. Rods are sensitive to shorter wavelengths with the peak is at 507 nm, while the cones are more sensitive to longer wavelengths with the peak is at 555 nm. These wavelengths and the eye’s relative sensitivity are shown in Figure 1. Rods give us our night vision capability (Scotopic) vision, while cones give us our daylight (Photopic) vision. The areas under the blue Scotopic and green Photopic lines in Figure 1 define the wavelengths useful for illumination and their relative effectiveness as a function of retinal sensitivity. A chart of the visible light spectrum shows the colors corresponding to the wavelengths in the graph.

             Figure 1 Scotopic and Photopic Sensitivity of the Eye and the Visible Light Spectrum

The ideal light source would output its energy at 507 nm and 555 nm, with enough output at ~600 nm (red) to yield a white color of maximum efficiency with very little energy wasted in wavelengths for which the eye is less sensitive.

Figure 2 shows the spectral output for a typical low-pressure Sodium lamp, shown in orange, overlaid with the Scotopic and Photopic sensitivity regions.

                        Figure 2 Low-Pressure Sodium Spectral Output (typical)

Nearly all the light energy output is at ~589 nm, down from the sensitivity peak of the Photopic region and almost missing the more sensitive Scotopic region. It’s obvious why, despite its rated efficiency, LP Sodium is such a poor overall lighting source. High-pressure Sodium was an improvement, shifting output to other wavelengths and yielding a better color. Figure 3 shows the output for a typical high-pressure Sodium lamp4, indicated by the area under the orange line.

                       Figure 3 High-Pressure Sodium Spectral Output (typical)

Note the distribution of the output wavelengths; nearly all are in the 570 to 600 nm range with a few low-power emissions elsewhere. The output is predominantly well below and to the right of the peak for Photopic sensitivity, and the output barely overlaps with the crucial Scotopic region. This is a better distribution than for LP Sodium, but it still leaves a lot to be desired. The green and blue emissions below 500 nm plus the red above 600 nm explain why the combined output is whiter in color than LP Sodium. But the marginally better light color comes with a tradeoff; it is achieved by adding Mercury to the carrier gas contained in the bulb. This renders the combination toxic. 

Now let’s look at the output of a high-brightness Neutral White LED5, Figure 4, shown by the area under the black line:

              Figure 4 Neutral White LED Spectral Output (Philips Luxeon Rebel)

 
The difference with Sodium is readily apparent. Most of the LED’s output is in wavelengths that correspond more closely with both Photopic and Scotopic sensitivity, Scotopic vision dominates in low light conditions, which of course is when these light sources are used. So it follows that white LEDs have a substantial advantage in effectiveness. Since LED output is more aligned with the sensitivity of the eye, fewer Lumens offer better effective illumination. The distribution of energy in the red wavelengths also explains the better light color of the LEDs compared to Sodium. The eye integrates these wavelengths, which combine to produce the light color. The LED output color is much closer to daylight and is perceived as better, with better color rendering. Consider the very poor color rendering index of Sodium light6 at 20 to 25 compared to ~75 for Neutral White LEDs and the picture is now complete. This explains why even four or five times greater total energy output by Sodium sources does not provide better illumination than LEDs.

The high-brightness white LED technology available today is not an evolution of lighting technology, it is a revolution! Comparison of legacy outdoor commercial and street lighting to LED outdoor lighting based on rated Lumen output alone is invalid. It’s time to rewrite the book!

References:
1. U.S. Department of Energy, Solid-State Lighting Technology Demonstration Gateway Program, http://www1.eere.energy.gov/buildings/ssl/gatewaydemos_results.html
2. Dept. of Hyperphysics, Georgia State University, http://hyperphysics.phy-astr.gsu.edu/HBASE/vision/bright.html
3. University of California at Berkeley, Transportation Center, Integration by the Human Eye: Implications for Warning Signal Design http://www.uctc.net/papers/207.pdf
4. PLUS lighting (manufacturer) 400W HP Sodium lamp specifications
5. LumiLEDs, data sheet http://www.philipslumileds.com/products/luxeon/luxeonrebel
6. Cooper Lighting Online, specifications 350 400 500 600 700 750 Wavelength (nm)