NYC's 250,000 Street Lights To Be Replaced With LEDs By 2017

An anonymous reader writes "New York City Mayor Michael Bloomberg announced that the city's 250,000 street light fixtures, which currently use incandescent bulbs, will be replaced with LEDs by 2017. It's part of a plan to reduce the city government's emissions by 30%. The LEDs have a lifespan of 20 years, more than three times that of the current incandescent bulbs, and Bloomberg says it will save $6 million in energy and $8 million in maintenance every year. It will be the largest LED retrofit in the country. 'The first of three phases to replace the standard "cobra-head" high-pressure sodium street lights, which will upgrade 80,000 at a time across the five boroughs, is expected to be completed in December 2015 with the final phase expected to be completed by 2017. Following the replacement of roadway lighting, decorative fixtures in the city's business and commercial districts will be addressed.'"

Comments

High-pressure sodium isn't "incandescent"

[Anonymous Coward]

It's a luminous vapor, rather than light from a heated solid filament. It's already much more efficient than incandescent bulbs to produce the same amount of light.

Re:I wish they'd do it here.

[AmiMoJo]

We have had them in my crappy city in the UK for a couple of years now. They put out better light than the old orange bulbs and seem brighter. The orange colour of the old bulbs is actually known to inhibit night vision, so white LEDs are safer.

20 year lifespan

[Bayoudegradeable]

Let me first say that I live in New Orleans, so go ahead make all your inept government remarks now. That said, we did begin making changes in our traffic signals to LED lights and the big claim of "20 year lifespan" was made. Less than 5 years later I see many of the LED bulbs (really, clusters of bulbs, like a Lite Brite set) are now replaced with the traditional traffic signal bulbs. Not only did the LEDs not last very long, they aren't being replaced with LEDs but with the old style bulbs. Hope NYC gets LEDs from a better vendor than we did.

Re:20 year lifespan

[mspohr]

There is no ballast in an LED light.
Fluorescent and sodium vapor lights have ballasts, not LEDs.
All of the traffic lights in my area were switched to LEDs many years ago. I have never seen a single light that wasn't working properly.

Re:20 year lifespan

[necro81]

Because traffic lights go on and off and on and off all day long. They're a terrible idea for LED because the ballasts wear out doing that

LED lighting systems don't have ballasts. True, LEDs require power conditioning (for these applications, it's some sort of switched mode AC/DC converter with constant current output), but those kinds of circuits are highly efficient and robust. LEDs experience essentially zero degradation from being turned on and off repeatedly. All those blinky lights on the front panels of computers, all the flashing indicators on routers and switches, those are all LEDs.

You are probably thinking of fluorescent lamps (tubes and CFLs), for which frequent on/off cycling is indeed a good way to make them die soon. No one makes fluorescent traffic lights precisely for this reason.

Re:20 year lifespan

[geoskd]

LED bulbs require current limiting in the power supply as well or they will burn up- so functionally the have the same requirement as a HPS or fluorescent bulb.

LEDs function effectively as a diode in a system. When you apply forward voltage, they turn on. For any given voltage, the resistance (and the light output) is constant. Florescent and arc lamps by contrast have a wildly variable resistance at any given voltage, and as such, voltage control cannot be used to effectively protect these lamps from over current. Setting a fixed voltage is easy to do in an efficient manner, setting a fixed current is less simple to do efficiently. In any event, they are worlds and gone different behaviors, so thinking of both types of power supply as a "ballast" is inherently incorrect and potentially harmful.

It should also be noted that the power supply for an arc lamp, or a florescent, have to be capable of producing some pretty high voltages, making them marginally more dangerous. LED drivers by contrast produce the same very low voltage at all times.

Re:Costs

[Anonymous Coward]

Nope, its actually cheaper to replace them in large blocks than to replace them one at a time. Old tech told me once that when they maintained the long tubes at the factory high up. Once a few went it was only a matter of about a year or two before the rest of the them did and it was more disruptive, time consuming, and costly to replace them one at a time than to do it all at once. So, I've followed this process for most of my larger lighting projects. If you are going to replace one brake light replace them all. If you are going to replace one headlight replace them both.

Also, the nice thing about LED lighting is that the way it fails is it just doesn't produce as much light as it once did. So in 20 years if they want to put off the costs for another couple of years, it's entirely possible to do so.

Re:I wish they'd do it here.

[ebno-10db]

Part of my town (the main streets) have LED street lights, and we have no problem with snow or ice. BTW, LED's are in the neighborhood of 20-30% efficient, so they don't run ice cold (pun intended). It may seem so though if you've only touched indicator LED's (flashing lights on equipment).

Re:I wish they'd do it here.

[fuzzyfuzzyfungus]

I wonder how many smaller cities have already done this?

I think that it's not uncommon (though traffic signals usually go first, since LEDs have been cheap and good at red, green, and amber for longer than they've been either cheap or good for white, and bulbs-behind-filters have always had even more miserable efficiency than bulbs in general).

LEDs are still pretty expensive, and white ones (because they are usually blue ones pumping a phosphor layer) are still less efficient than one might like; but one big advantage is lifespan.

A replacement lightbulb doesn't cost much; but sending out guys in bucket trucks to deal with dead ones adds up.

Re:I wish they'd do it here.

[K. S. Kyosuke]

Well, white LEDs are most likely lousy for low levels of illumination because our color perception shifts with the illumination: our visual cortex expects redder colors in darkness, so physically white faint light looks unnaturally blue, and incandescent light bulbs correspondingly look too red when you attempt to use them for daylight levels of illumination. I'm patriotically proud to point out that this is called the Purkinje effect. ;-) While the fact that LED light appears brighter may lead to energy savings beyond the simple increase in energy conversion efficiency, I wonder how it will change the perception of traffic signs. The red ones will probably appear even darker. What about traffic safety?

Re:Stick with sodium

[Joce640k]

What about amateur astronomers?

Amateur astronomers actually make a *lot* of the discoveries and do a lot of the photography.

Re:incandescent != sodium

[SJHillman]

I did a quick Google to satisfy my curiosity and found a few things:
- While high pressure sodium gives off more lumens per watt, LED has better effective illumination (in part due to how our eyes can detect different wavelengths)
- LEDs are more directional, eliminating up to 40% of light loss due to reflectors
- In the end, an LED might only need to give off 20 or 30% as much light to still illuminate the same area effectively

Source: http://www.al-e.com/led-vs-sodium-lamps

Re:Many of those bulbs due for replacement anyway

[djlemma]

They are replacing High Pressure Sodium lamps, which are not incandescent. The funny thing is, by the standard measure of efficiency used in the industry, the new street lamps probably will be LESS efficient than the old ones. HPS lamps can get above 100 lumens per watt pretty easily, and low pressure sodiums can even get up to 200 lumens per watt. They've been able to get efficiency like that in labs for LED's, but for production fixtures it's not very common.

Of course, LED's often win out in real-world comparisons, because all the lumens are more efficiently directed where they need to go. Still, to get that much brightness, it's going to cost quite a lot of money.

Re:I wish they'd do it here.

[Samantha Wright]

I'm not sure where you're getting the "expects redder colors" part from. The Purkinje effect simply describes the fact that we're more sensitive to blue light at lower intensities—we see it better. This is purely physical, and due to the assymmetry in the response curve of all of our photoreceptors. While most direct light sources activate the cone receptors, this bias is sufficient to make us think of our monochromatic rod cell night vision as slightly bluish, which is why nighttime scenes are depicted as being blue in art, even though you're literally only seeing something grey. Rod cells have such a wide response range in the blue portion of the spectrum (not shown on graph) that some people can see very violetish frequencies with them, causing eyestrain as we get indecisive about how to dilate the pupil.

Sodium lamps are extremely monochromatic; they only put out a very small range around 600 nm because of the chemical reaction that they operate on. Any white bulb either incandescent or LED, even ones with a bluish tint, will illuminate red signs much better than a traditional sodium-vapour lamp.

Re:I wish they'd do it here.

[smooth wombat]

Streetlamps for pedestrian safety reduce safety

Bullshit. Street lighting has been found to reduce pedestrian crashes by approximately 50%.

peed bumps increase traffic crashes and reduce safety

Double bullshit. Overall, the treated streets experienced a 39 percent decrease in crashes per year after speed bumps are installed. The 39 percent decrease on speed bump streets is a statistically significant difference (t = 2.8) from 1.39 to 0.85 crashes/year, meaning crashes most likely do decrease on speed bump streets due to bump installation. As well as this gem which asks a different question but which provides the same evidence against your "common sense".