LED Lights Don’t Melt Snow on Traffic Lights, Hiding Signals From Drivers
Jan. 8, 2010
A huge swath of the country is getting snow and it’s raised an unusual and potentially dangerous problem for motorists.
Click picture to see report
Communities across the country are converting to LED traffic lights, but these lights don’t emit heat, so snow doesn’t melt like it would with a regular incandescent bulb. In some cases, Drivers then can’t see the signals.
During a snow storm last year, Lisa Richter of Oswego, Ill., had a green light and was turning left. But police say a driver in the oncoming lane blew through his red light and plowed into her, killing her instantly.
This wasn’t a regular accident. Police said this traffic light, blocked by snow, contributed to the crash. The light was an LED signal, which doesn’t emit heat, so snow doesn’t melt like it would with a regular incandescent bulb.
Cities and states across the country that have converted to LEDs report an energy cost savings of up to 80 to 90 percent.
In Green Bay, Wisc., where all traffic lights are now LEDs, December’s incredible snowfall caused many to be packed with flakes.
After their intense storm last month, some drivers in Madison, Wisc., noticed their neighborhood LED signals were hiding.
“I know that the stoplights are there, but if I didn’t, it would have been very easy to fly right through them,” one driver said. “And especially with the off ramp right on the interstate, it could be a very dangerous situation.”
The state of Wisconsin switched to LEDs in 2002 to achieve the massive energy cost savings. Maintenance costs are also much lower because LEDs last a long time. Incandescent bulbs usually have to be replaced every 2 years.
“With LEDs, we have some of our heads that were installed in 2002 still operational today,” said Wisconsin state traffic signal systems engineer Joanna Bush.
Another advantage of LEDs: Bush said the old incandescent bulbs could pose safety problems of their own.
“When they fail, they go dark, like a light bulb at your house. There’s no warning and it’s dark. With the LEDs, it’s a string or two that starts to go out and the driving public might not even notice a change in the LEDs and we can get our crews out to change it.”
There’e the right way, the wrong way, and the government way.
Thanks to the last Energy bill, we get LED’s for traffic lights, CFL’s for home use, and nobody even notices the biggest waste, the Street Light.
The LED’s should be in the Street Lights and your home, and either the CFL or incandescent in the Traffic Lights.
LED’s don’t belong in icy weather Traffic Lights and CFL’s don’t belong in your home, and what the heck was Boxer thinking with when handing over our Big 3 electrical industries to China? I know who she wasn’t thinking of at the time… the public good.
LED’s sold separately.
mikelorrey (15:43:37) :
AFAIK, wrong part of the IR spectrum. You need an LED whose emission curve correlates with that of H2O’s absorption spectrum.
Say what? The lens material would be absorbing the IR emissions, converting them to heat, and that heat would melt off the ice and snow. The IR absorption spectrum of water does not matter for that.
Oh, and it could also be an optically transparent coating on the lens that is absorbing the IR, instead of the lens material itself.
Hmmm so people wanted to go green and so you end up with LED lights. But wait, they don’t generate enough heat to melt snow, solution?? big PLASTIC shields, oh the irony, to go green, they need to use an evil fossil fuel to do it, typical of the greens/left. Do as I say not as I do.
And lastly in that video, that reporter goes on to say even the old incandescent lights had a similar problem, the yellow lights weren’t on as much so didn’t heat up, but it’s sitting between two heat sources red and green, so I’m sure that hardly ever happened if at all!
Just imagine if this was the other way round and all these towns had installed regular lights resulting in dozens of accidents and a few deaths. We would hear about it non-stop from the MSM.
Hmm. Reminds me of an incident I heard about years ago of a train crash in the U.S. Two trains were heading for a junction point. One of them didn’t get the signal that they were supposed to stop to let the other pass, and they crashed. There was a legal case about this, and what was ultimately discovered was that one of the drivers missed the stop signal, which was placed long before the junction, because a reflective material had built up in it (perhaps because of a leak in the signal allowing moisture to enter), which at certain times of day appeared to show a green light, even though the bulb for that light was not powered on. Tragic case.
latitude (14:12:11) :
I just saw on The Weather Channel that there was a few snow flakes in the air at a beach just south of West Palm Beach.
photon without a Higgs (16:48:32) :
latitude (14:12:11) :
they just said a snow and sleet near Miami too
“photon without a Higgs (15:33:45) :
latitude (14:12:11) :
photon, the high was 73.3F at Palm Beach Gardens yesterday
[…]
For people to find out that there was record heat at an airport in Florida would make the UHI case crystal clear to laymen.”
Did they place the thermometer behind the place where they test jet engines after maintenance work?
“Someone needs to do an engineering study and work this out properly, taking into consideration all costs, including installation, maintenance, electricity usage and unexpected occurences like snow and accident damage.”
I worked in the traffic industry, specifically designing traffic light controllers. Of course an engineering study was done, and LEDs came out on top. The decision had nothing to do with power or going green, but pure economics.
LED lights cost more upfront, but due to their long lifetime the maintenance costs are much lower than incandescent. Imagine the cost of sending a crew out to replace a bulb every few years, having to hire traffic controllers, block off a lane, etc. Minimising having to work on the road is a significant factor in choosing to adopt many new traffic technologies.
The other main benefits of LED lights are consistent colour and high intensity. Power savings are low down on the list… As one poster pointed out none of the street lights are LEDs.
The idea that this is some green conspiracy is idiotic. It would be trivial to include a heating element in the lights and I’m sure some enterprising company is doing that right now.
That is only true for people with normal color vision. Among Caucasian males (north european) Red Green color blindness (actually atypical perception not blindness), is almost 10% . To add complications there are several types and no tow have the same degree of red green colorblindness. I am one of those 10%. In good light with large color samples I can easily tell most colors apart, and identify them by name. In dim light, or with very small color samples (like thin lines on a chart) it is usually hopeless. I cannot tell dark purple from black or dark red or dark brown from dark green, or yellow from light greens.
When I was in search and rescue, they would flag trails with red survey ribbon, by tying it on evergreen trees. I would walk right past the flagged tree and never notice the several inches long red ribbon. When I was told the tree was flagged I would have to visually search the tree to find the red flagging even when I knew it was there.
For me, dark green and dark brown are indistinguishable in poor lighting, Dark red might as well be black or dark gray under the same conditions. Light green and white (in the form of lights) are indistinguishable unless they have significant gray tone differences or distinct color casts of another color.
I was 14 or 15 before I realized that the “green light” on a traffic light was actually green, I thought it was simply a figure of speech, I always thought it was just a dirty lens on a white light. The new LED traffic lights are actually easier for me to tell the difference between the amber and the green lights. The older incandescent lights had a warm green color which was hard to tell from pale amber lights. The newer LED green lights tend to be more of a blue-green color and the amber has much higher color saturation, making it much easier to tell the difference between the two.
On numerous occasions I have nearly blown through a traffic light that I did not even know was there, because the green or amber light blended in with city lights behind it and I did not even realize I was entering an intersection until the light changed to red. This is especially a problem when the background has lots of different types of lights like sodium vapor, mercury vapor, plain incandescent, and colored signs.
In locations where cities and towns do not follow conventional placement of lights (in Colorado it is Green top, amber middle, red bottom) I cannot use physical location of the light to help tell if it is an amber or green light.
Orange traffic cones for me, if dirty, are almost invisible if placed where their visual background is a dirt embankment or a green of similar gray tone.
Do not assume your visual impact of colors is typical of others. It is most certainly not.
In the case of folks who like to make color coded charts on the web, about 50% of the time, I cannot tell what the chart is saying without the use of colorblind tools to help me distinguish the colors of the chart traces. In some cases I literally cannot see some of the traces even if I know they are there because they totally merge with poorly chosen background colors. Light cyan and an off white background simply merge I will not even know the cyan trace exists unless you point it out to me.
Normal color vision has highest sensitivity in the yellow green peak of the suns spectrum (high visibility chartreuse yellow used by emergency signs and fire engines in some places), not in red.
If you place a dark blood red object on a dark background I will not even know it is there until I trip on it if the light level is low.
Since some industries require normal color vision, people like me with red green color blindness tend to concentrate in other industries. I used to work in a computer data center on a crew of 4, ( 3 men one woman). All three of the men on the crew were red/green color blind. When we needed to know if a warning light was green or yellow we had to ask the woman to verify it. Or (Thank you Sun Microsystems ) use command line instructions to determine the warning light status.
If you use color coding in warning lights or charts, include physical cues like location or line patterns on charts to give the 10% of your audience who cannot determine what you are trying to communicate by color alone a clue. Also make damn sure your color key has exactly the same color triplet as the line on the chart.
Nothing irritates me more than someone who has a chart with two traces that are almost identical to me (say a yellow and light green) and when you use a color picker tool to compare them, you find out that neither trace has the same color triplet as the color key blocks on the chart.
/end rant 😉
Larry
photon without a Higgs (16:52:54) :
they just said a snow and sleet near Miami too
Gore Brand Global Warming:
Directions – Wind up model, mix with one part hot air and broadcast over affected area. Comes with a Money Down the Tubes guarantee.
Warning: Avoid contact with populated areas. Keep out of reach of Politicians and Children.
{… DirkH (17:01:35) :
“photon without a Higgs (15:33:45) :
latitude (14:12:11) :
photon, the high was 73.3F at Palm Beach Gardens yesterday
[…]
For people to find out that there was record heat at an airport in Florida would make the UHI case crystal clear to laymen.”
Did they place the thermometer behind the place where they test jet engines after maintenance work? …}
No, space is precious, they mounted it on top of the Honey Bucket Holding Tank Digester. That’s a “green” location…. bio-degradation you know…
I agree with the idea to install a temperature triggered heater element.
It wouldn’t be very difficult, all that’s required are some power resistors, a transistor and the right type of thermistor. Pick the thermistor type so that the resistors are switched one as the outside temperature falls below freezing and it will regulate power through the resistors to keep the assembly above freezing temperature. During summer it will consume virtually no power.
The problem is that they didn’t think of this until it became a major issue.
Heating element can be laminated into the “lens” and operated via thermostat as well as back-reflective sensing (you can use an LED for that as well). The back-reflective sensing measures the reflection off the inside of the lens, and when it’s ALSO sufficiently cold, one could operate the de-icing element on the lens. The element need only be a line; it doesn’t have to de-ice all the surface.
One can’t rely on wind or gravity to get rid of any stuff that’s stuck on the lens.
BTW: If it’s not cold enough for ice/snow (typically above 4 degrees C), then excessive back-reflection may indicate a fouled lens; which, if the signals are network, could be used to schedule a cleaning crew visit. Keep in mind though that the reflective sensor probably also responds to low sun angles where Murphy’s Law operates. If the sun is not intense, then one can discern the signal’s LED pulses from the “constant” sunlight.
The LEDs are doubtless already more expensive to install, adding electronics and a heating element would increase the cost further.
There are ways to do it that require NO electronics. You use material properties (positive temperature coefficient of resistance) to do the job. Then you only have the cost of the material and extra wiring to deal with. Reliability is very high.
And as some one pointed out above, labor (and the attendant vehicle etc.) is the main cost in traffic light maintenance. So LEDs are a good thing.
mikelorrey (15:01:32) :
Me too. The capacitance of liquid water is going to be different from snow.
But then your sensor has to be out in the elements. Not good for sensor life.
Unless you have designed sensor based eqpt. you have absolutely no idea how difficult it can be. Murphy is very unforgiving. KISS.
And how different is the capacitance of wet snow from liquid water? What factors would confound that differential? What happens when the sensor is covered with mud, or dirt. Or in a harsh industrial environment (say near a steel mill) metal particles.
Murphy is a bastich.
Suggestion: Incorporate radioactive elements to melt the snow and ice.
“Better Living Through Thermonuclear Biochemistry”
Did I miss any toes? 😉
Somehow the conservation alarmists never seem to look for unintended consequencies of their wacky ideas. A good example are energy saving florescence light bulbs. My experience is that at given power rating, the light is several lumens lower than an incandesent light bulb with the same power rating. However the next jump is not small, ie 60 watts to 75 watts. If you want the same brightness, you have to buy a bulb with a higher watts rating. In addition, we will now be contaminating our land fills with toxic particles that contain mercury. The LED is yet another example of not asking what could go wrong? Of course, every example in top ten wacky ideas to save the planet fail to mention any limitations to the climate saving fiascos.
M. Simon (20:08:27) : And how different is the capacitance of wet snow from liquid water? What factors would confound that differential?
Out here in Kaleeforneeya (as Ahnold says it 😉 we have the occasional pigeon that likes to make a nest in the lights. It is said they like the high safe place AND the warmth. It would be interesting to find out if there are fewer “nesting” issues with the LED lights. But if you put in “warming” equipment, especially if it senses birds, you might find a whole new set of problems coming home to roost 8-}
In my informal observations, they seem to prefer red to green (but not by a lot) and I’ve never seen one in the yellow.
I’m quite pleased to see so many responses in this thread focused on improving the technology to address the problem rather than throwing the baby out with the ice-water so to speak. The most common suggestion by far seems to be incorporating additional heating elements that can be controlled when the temperature drops with many people wondering how that would effect energy consumption. So for fun some back of the envelope calculations yields:
It takes about 1W to melt a block of ice weighing 0.01kg . 1 mm of water over a surface of 1 square meter weighs approximately 1kg. Ice weighs just slightly less then that and snow would be reduced by a factor of 10. The radius of a standard traffic signal lens is 0.152 meters giving an area of 0.0725 sq meters. So 1 mm of ice covering the lens would be just less than 0.0725 kg. We would then need about 7.25W to be able to melt that. If the buildup is primarily due to snow the energy required to melt that is an order of magnitude less due to the air content of snow and we could melt 10mm of snow for the same amount of power. 10 mm works out to just under half an inch. So 1″ of snow takes around 15W. For a single traffic light that is 45W for all three indicators. At each 4 way intersection you would need at least 4 signals or 180W. Add in turn signals that only require an amber/green indicator with the shared red lamp and you’ve just added an extra 120W which brings us up to 300W for a typical intersection on a 2 lane road.
So doing a quick google search for applicable power rates for public utility lighting in a state that receives lots of snow/ice (Grand Rapids, Minnesota) gives us a rate of $0.0272 per kWh. 3 intersections would be able to run heaters continuously for an extra 2.72 cents an hour or 65.28 cents per day.
The gridded map of Grand Rapids streets works out to about 17×17 intersections so for simplicities sake 300 intersections would add 65$ per day to the budget. If we bound it to the months where the average low does not drop below freezing we have around 200 days @ur momisugly 65$ or an annual budget adder of roughly $13,000 significantly less than any lawsuit I would surmise.
For any accumulation greater than 1″ heater requirements and costs would have to be adjusted accordingly. Improved sensing of moisture/temperature could counterbalance the additional heater requirements.
When I first started this thought experiment I had anticipated much lower power requirements to remove snow/ice accumulation. I also anticipated much lower operating costs but when you think about the number of traffic lights in a given area and actually start summing them all up even a little additional power per light can quickly drive up utility budgets.
On a separate note one of the earlier commenters mentioned LED failure due to the repeated switching of the signal indicators. A huge benefit of LEDs is that they don’t age like a traditional bulb due to high initial surge currents and can be switched on and off very rapidly. Dimming of an LED bulb is accomplished by switching it on and off faster than the eye can see and reducing the on and off time to achieve the correct intensity called pulse width modulation. The slow switching time between signals would not cause premature failures to LEDs.
I had actually often wondered what the power savings might be if traffic lights employed a controller that reduced the intensity at night. During the day the LEDs should be as bright as possible, but at night they could easily reduce intensity by half and still be clearly visible (provided they aren’t iced over). After working through some of these calculations it might be worth pursuing as additional energy savings if not already done.
JM
So, why are people so fixated on heating solutions?
The Snow Scoop Tunnel Visor seems like a clever and cheap solution. (From Joel (13:57:12))
————
James Mayo (21:44:14) :
>Dimming of an LED bulb is accomplished by switching it on and off faster than the eye can see…
I really despise the PCM brakelights on cars these days. How fast do they have to
flash so they are distracting when I turn my head quickly?
I’d be happy if they just had a bigger resistor (another way you can dim a LED – reduce the current through it) or use a simple power supply to resude the current
out.
Salt solution.
Glue a cereal bowl to the top of the light with rock salt in the bottom.
The pigeons will bathe in this salt bath and “pollinate” the entire housing with salt spray. The salt will melt the snow, and eventually poison the avians.
That’s called killing two birds with one stone… err.. rock.
Ric,
The only reason I can think that heating solutions are the dominant recommended solution is the description of the problem leads you down the path of looking at a way to achieve the higher thermal output of incandescents.
I agree the scoop seems like the best way to solve this problem. Its passive, has a one time fixed cost and only needs to be fitted to units that suffer the problem which doesn’t burden the cost of the additional heating elements for those living in warmer climes.
As far as LED dimming goes PWM can be done with the same number of components. Since brightness is a function of duty cycle varying levels of brightness are easy to achieve versus fixed resistors that gives you two discrete states. Most LCD screens have an LED backlight now that uses a power supply to control the current. It is definitely a lot better as the power supply can reduce or increase current within microseconds as opposed to milliseconds, making it impossible to tell it is switching at all, but the added cost isn’t really practical if you have lots of LEDs as it pretty much forces them to be wired in series to maintain the same current through all of them. If you had parallel groups of LEDs from the power supply they would not have the same current going through each and for a brake light that has several LEDs wiring all of them in series would require a very high voltage to drive that many at once.
In the automotive market I’ve heard of meetings with suppliers that last all day arguing over 1 pennies worth of savings in a single component. Given how sensitive everything in a car is to costs if there is a microprocessor already controlling the LEDs then I’m sure most manufacturers would rather control brightness by PWM then any other method.
In all the LED lighting applications I’ve worked with we try to have a minimum of 120Hz switching so as not to be noticeable. Your monitor is updating at a rate of 60-75 Hz and most people don’t have a problem, although a very small percentage of people seem to be sensitive to those sorts of things. The early DLP projectors had a color artifact known as ‘rainbows’ that only about 1% of people could see as a result of the color wheel inside the projector not spinning fast enough.
JM
Steve Goddard (10:19:55) :
Many serious accidents happen at traffic lights, because people speed up to get through at the end of a cycle. Roundabouts are much safer because you have to slow down entering a roundabout, and because there is no reason to rush. You never have to wait for a light to cycle.
Accidents aren’t caused by the traffic lights at an intersection, but by the drivers.
Every traffic circle (roundabout/rotary) I’ve ever seen in the US had at least one set of tire tracks running straight across the sod from one entry to the exit on the opposite side, except for one on Ft. Dix, NJ — the tracks ended at the chipped paint on the M-48A5 tank sitting in the center.
Give the traffic engineers a break. For over a decade, they’ve been told that snow and cold are about to become rare events, so how can we blame them if they finally started believing the spin and installed lights that won’t function in snow? Blame those that preached that we’re all going to hell in a warming hand basket.