US tunnel to install sodium lamps as ‘LEDs not ready yet’

The operator of a tunnel in America’s north east is set to install high sodium lamps because it claims that ‘LEDs haven’t been tested in a tunnel’.

A spokesperson for the Washington Department of Transportation said the lighting refit of the Mount Baker and Mercer Island tunnels in Seattle would use SON lamps and not LEDs as the latter had not been proven in tunnel conditions.

The spokesperson told local news station KIRO 7, that the improvements would actually make tunnel lighting 50 percent brighter by early 2017. The project is part of an ongoing upgrade to Interstate I-90.

Drivers who regularly use the tunnels have long complained that the contrast between the dark tunnel and the brightness outside often creates a traffic slowdown with people braking as their eyes adjust to the changed light conditions in the tunnels. There’s anecdotal evidence that traffic flows are better on overcast days, when the contrast between the tunnel and the open road is less pronounced.

However, the decision to use the relatively old technology of high pressure sodium over LEDs will be seen as controversial in the lighting industry. Earlier this year, the 10.7 kilometre Toven Tunnel in Nordland, Norway, became the world's longest lit entirely by LED lighting.

In what was seen as a vote of confidence in LED technology for tunnels, the operator – the Norwegian Public Roads Administration - opted for LEDs over traditional technologies to achieve low energy and high controllability.

The lighting has either meets or exceeds all lighting requirements in the relevant standards, with a luminance level in the inner zone of 0.7cd/m2, which exceeds the requirement of 0.5cd/m2. The efficient LED lamps typically use 21 per cent less energy compared to conventional alternatives. When combined with the intelligent lighting controls employed at Toven tunnel, this figure increases to 70 per cent to secure substantially lower energy costs.

The use of LEDs in tunnels is the theme of the world’s first Road Tunnel Lighting conference, which takes place in Barcelona on 8 and 9 October 2015. Entry is free for tunnel operators and consulting engineers. See the full programme and register for a free place at

Picture: Matthew Rutledge

Comments 3

This is an interesting decision. The article is too short to explain the rationale behind the decision. I presume the increased illumination within the tunnel was necessitated by the original "under-lit" conditions, or the glare from unshielded lights. (Unsielded lights in a tunnel is a REALLY BAD idea.) But what surprised me most was the editorial comment that LEDs would have used only 21% of the power! This makes me wonder what HPS lights they have chosen to use! For the same amount of light I would expect a difference of 15% or so, since presumably the opted-for HPS would have modern bulbs and ballasts. Or, that the LED option would have them dim the tunnel 3/4 of the time. You don't need to be an expert to know that that HPS have MUCH greater efficacy than incandescent bulbs!

LED torches, fantastic. LED retrofits, useful. Non-maintainable LED fittings, doubtful. It's a panel indicator taken to extremes. Indeed, what will it feel like in 4, 5, 6, 10 years time when these devices start to fail. Ker-ching says the lighting industry! Ouch says the taxpayer!

This is remarkable. The coverage of the LED tunnel contained no criticism, but this article uses 'quotes' to make clear that the argument of the engineers is shown to be doubtful. There is no weighing of the evidence, and in fact, you don't know the brief. Most tunnels have a design brief for a twenty to thirty year life (or longer) on their installs, as a full replacement/rewire involves closing major infrastructure. Conservatism is reasonable under these circumstances. Further, the SON lamps may not be specifiable in twenty years, but with the amount of installed stock around the world, there is good reason for confidence that there will be companies providing long-tail maintenance, or at worst, upgrading. On the other hand, the chances of being able to get a replacement component for a 2015 LED in 2035 is at best unproven. So if the decision was based on risk management, it's quite a sensible one. Further, without giving the relative lumens/watt and W/sqft values of two comparable compliant schemes, it's essentially impossible to read such an article and work out the strength of arguments. Better still, the capital and operating costs over the service life, and the NPV of each. When you start doing that analysis, conservatism on technology often wins out. Oh, and finally, never provide equivalences without providing the maintenance factors applied, since this easy fudge factor hides all.

Leave your comment