I was very surprised by the conclusions of the Streetlights for Collector Roads report issued recently by the Lighting Research Center (LRC). In fact, I’m not sure it is still relevant considering how fast lighting technologies evolve nowadays. When it came out, it was already inaccurate and outdated.

The National Lighting Product Information Program (NLPIP) is administered by the LRC. Its mission is to provide lighting specifiers and other professionals with “the most complete, up-to-date, objective” information available on energy-efficient lighting products. This Specifier Report challenges the pertinence of LED lighting for collector roads. But, unfortunately, it is neither complete, nor up-to-date, or objective…

The report concludes that LED has an inferior distribution and higher cost per mile than HPS lighting, and that it does not meet RP-8 ratings for collector roads. However, careful scrutiny shows a flawed methodology. I also think the luminaires they picked may not have been the right ones, thus leading to erroneous conclusions. It almost feels like they are biased in favor of HPS lighting, which is disconcerting, considering the habitual quality of the LRC studies we have come to expect.

The LED luminaires used were, in most cases, underpowered and not appropriate. I wonder what made them compare 70.3W, 88.9W, 73.4W, 87W, 77.8W, and 74.1W LED to 182.1W, 190.2W, 191.4W, and 178.4W HPS, and 194.8W PSMH luminaires. No one can expect the LED fixtures to equal – in terms of intensity, uniformity, or pole spacing – their HPS and PSMH counterparts. These LED fixtures will also not achieve 0.9 fc average at the end of a 27 feet pole.

Strangely enough, two LED luminaires had a substantially higher wattage (138.2W and 194.8W) but the reasons are not explained.

Then, there’s the Light Loss Factors (LLF) the report has used. The NLPIP made some efforts to come up with average numbers instead of the standard and widely accepted ones. For HPS, instead of the 80% lumen at 50% life usually used, they used 0.84 Lamp Lumen Depreciation (LLD). As for LED, contrary to the logical 0.85 value it should have used (if 70% depreciation means end-of-life, then 85% is half-life), they attributed an LLD value of 0.79. Do the math and you get LLF values that seem to favor HPS technology.

On the other hand, the NLPIP admit that they do not use standard and accepted numbers, but instead calculate averages based on data provided by the manufacturers. I’m not convinced this is totally neutral and objective.

Another issue I have is with how they visually demonstrate how expensive LEDs are (Life Cycle Cost per Mile over 27 Years on page 16). One quick glance at their chart is enough to want to abandon LED luminaires altogether. In some cases, it looks like they’re more than three times costlier to own and operate than HPS! But a closer look reveals that the extreme difference is due, in part, to the fact that the graphics include the relamping of LEDs at 25k hours! They should have stopped the line at 50k, considering this is the bare minimum lifespan most manufacturers advertise for their systems.

Furthermore, because the luminaires in this test are so underpowered, they require closer pole spacing. This means more luminaires, which translates into a higher pole-per-mile value and higher relamping costs. If LED wattage better suited for a collector application had been selected, the difference in cost-per-mile between HPS and LED would be lower. The relamping cost could also be lowered even more if a system such as our LifeLED™ was used, because it requires no more effort and time than relamping an HPS fixture. You just do it less often! So with proper data and a real-life scenario, this graphic would have been much less dramatic!

The icing on the cake lies in the response the NLPIP has addressed to the Pacific Northwest National Laboratory’s correspondence regarding their report. They explain that they “tested only one sample of each luminaire, and purchased luminaires from July through October 2009.” One year is almost a century in the SSL world and we cannot afford the wait.

In the fast-paced, hi-tech industry that lighting has become, this report is unfortunately outdated. It is also a bit flawed and inaccurate. It serves to document what was the Ultimate Worst-Case LED installation back in 2009, if done with inadequate luminaires and planned in a hurry.

Despite all the arguments stated above, I must still acknowledge the efforts deployed by the LRC as it is a great challenge to benchmark technologies that change almost every day. I’m sure such a report would be very appreciated by the specifiers and the industry and I would encourage the LRC to try again, but with up-to-date numbers and current technology. And, most importantly, to make sure the results are published quickly next time. Specifiers need the latest data at hand to do their job.

Note: Shortly after this post was published, the LRC released an addenda to their study. Story continues here.