LED replacement

[petersonra]

This is more about personal curiousity than anything else.

Are you guys installing a lot of the LED replacement lamps for fluorescent fixtures?

I had understood there is not much difference in the energy usage so I am presuming that means whatever cost savings there is comes either from subsidies or labor savings in not having to replace the lamps.

What all does it involve? Just taking out the ballast and rewiring the fixture direct?

 

[TNBaer]

No. Not at all. I keep doing the math and it NEVER pays it self. Add in the lower light levels, lower CRI, and steep lumen maintenance curve, and there's no reason at all to be doing this. In fact, the DOE has issued statements against this. See here: http://www.usenergysciences.com/content/what-does-us-dept-energy-have-say-about-leds

 

[tortuga]

No. Not at all. I keep doing the math and it NEVER pays it self. Add in the lower light levels, lower CRI, and steep lumen maintenance curve, and there's no reason at all to be doing this. In fact, the DOE has issued statements against this. See here: http://www.usenergysciences.com/content/what-does-us-dept-energy-have-say-about-leds

Same here.
We only install what LED's the local utility gives rebates for and those are never included.

 

[GerryB]

LED Dimmers

LED Dimmers

Do you need special dimmers? A friend changed bulbs in what looks like 12 volt track heads to led's. He said he put in a special dimmer but the lights blink. I haven,t gotten involved with these yet. Thanks

 

[broadgage]

As other post, the replacement of flourescent lamps with LEDs is seldom worthwhile, the claimed lifetime is not allways reached, the light output may dissapoint, and some types are very directional.
The only exception IMHO is in freezers or outdoors in artic climates, flourescent lamps dont perform well in very low temperatures, and LED retrofits are well worth considering despite their other drawbacks.

Replacing halogen lamps with LEDs is much more worthwhile and can save a lot of electricity.
Most LED lamps can not be dimmed, though a few can be.

 

[Electric-Light]

This is more about personal curiousity than anything else.

Are you guys installing a lot of the LED replacement lamps for fluorescent fixtures?

Some contractors love to push LEDs, but LEDs are neither good nor bad.

I had understood there is not much difference in the energy usage so I am presuming that means whatever cost savings there is comes either from subsidies or labor savings in not having to replace the lamps.

LEDs can produce highly saturated brilliant colors. So, LED outdoor signs can provide much more vibrant colors and far better efficacy than fluorescent lamps behind translucent plastic panels.

Efficacy and life of LEDs are unmatched by anything I know in production in one to two watt lamps. They're quickly replacing lower power flashlights and LCD backlights.

However LEDs fall behind in efficacy and output to full size fluorescent and HIDs.

What all does it involve? Just taking out the ballast and rewiring the fixture direct?

Depends on the kit. Ballast bypass is laborious as ballast replacement or even more. Philips 48" T8 LED requires a two-connection socket, the type used for rapid start. So if you're retrofitting instant start fixtures, it requires socket replacement as well, parts & labor.

In other setups, LED system is a permanent integral part of fixtures. The replacement requires entire fixture be replaced. Out-of-warranty repairs are very costly. LED fixture require model specific LED modules if they're modular. If they're non user serviceable, the entire fixture must be replaced which entails long down time and labor cost.

 

[TNBaer]

Electric Light, that's about the nicest thing you've ever said about LEDs. Be careful out there with these replacement tubes. I had a distributor come in here the other day with LED tubes that could be installed with the ballast in line. The company's spec sheet said their product had an 85 CRI. However, the Lighting Facts label included with the product listed a CRI of 67. The other thing was the manufacturer said their product was 50,000 hours but what happens when the ballast burns out? If you're installing LEDs on an old ballast you don't have a maintenance free product. Also, they stated that their product works with "most" ballasts. What happens if you experience many failures but the company says that the ballast you had is not compatible with their product? And lastly, they touted that the LED tube could be used with electronic and magnetic ballasts. I wonder how that is possible considering magnetic ballasts use twice the current of electronic ballasts?

 

[Electric-Light]

Electric Light, that's about the nicest thing you've ever said about LEDs.

I've never been not nice about LEDs. I'm only critical about misapplications they market for that I find to be improper and claims that are more than a bit puffed up.

A little engine that works great in a Honda Civic isn't necessarily efficient if you gang up a hundred of them to power a big vessel. Bunker fuel burning marine diesels are among the most efficient engines known to man kind, but they can't be scaled down to something small enough for a car and give the same performance (50% fuel to mechanical efficiency).

So, learn to accept running a Titanic on VTEC engines is inefficient and not cost effective.

They're both great technology. The applications of LEDs I see today aren't always proper.

careful out there with these replacement tubes. I had a distributor come in here the other day with LED tubes that could be installed with the ballast in line. The company's spec sheet said their product had an 85 CRI.

Those are another product of marketing gimmick. Some can function with ballast in place, but of course, it reduces the overall system efficacy.

If you're installing LEDs on an old ballast you don't have a maintenance free product. Also, they stated that their product works with "most" ballasts. What happens if you experience many failures but the company says that the ballast you had is not compatible with their product?

LEDs can not run straight across the line like incandescent lamps. LED "lamps" have integrated passive or electronic driver/power supply/ballast. A fluorescent ballast is a driver but LED marketeers see ballast as a dirty word. It's a matter of semantics, like 4WD vs AWD.

For the ones that work with existing ballasts in place, leaving them in place is a installation cost saving at the expense of overall efficacy. It's also a ground for insurance denial on ballast fire originating from existing ballast as it constitutes usage for unintended application.

So contrary to what LED sales like you to believe, LEDs are no strangers to outage due to ballast/driver/power supply failure.

And lastly, they touted that the LED tube could be used with electronic and magnetic ballasts. I wonder how that is possible considering magnetic ballasts use twice the current of electronic ballasts?

A magnetic ballast simply behaves as a limited current 60Hz power source. An electronic ballast's 60KHz power may not allow LED module's driver to function properly. Ballast current and power input varies depending on load.

Check out new display cases at Target grocery area. These LEDs are not nearly as efficacious as T8 fluorescent, therefore to use them for general lighting is a poor application. They're on a motion detector switch though.

It is not an acceptable option with fluorescent, because the ramp up time would be far too excessive, so until recently they're started at store opening and left on. With LEDs, they can be short cycled and reach full output immediately in frigid conditions.

Leaving fluorescent light on all the time is very detrimental in display cases, because the heat needed to stay warm is a load on refrigeration system, yet they take too long to warm up for frequent shut down.

 

[cowboyjwc]

The driver on LED's is still your weak point. You could have 50,000 hours of bulb life, but most drivers won't last that long. a lot of the cost savings with LED is not with retrofits but with new installs. According to what I've heard the maintenance level is way down.

 

[weressl]

No. Not at all. I keep doing the math and it NEVER pays it self. Add in the lower light levels, lower CRI, and steep lumen maintenance curve, and there's no reason at all to be doing this. In fact, the DOE has issued statements against this. See here: http://www.usenergysciences.com/content/what-does-us-dept-energy-have-say-about-leds

That article is 2 years old. The LED technology had went through several generations of changes since. Output, life and CRI had improved multifold. SMD and SMT technology had also emerged and makes this drastically different. It is time to start looking at again and jump in at ground floor.

 

[1793]

The driver on LED's is still your weak point. You could have 50,000 hours of bulb life, but most drivers won't last that long. a lot of the cost savings with LED is not with retrofits but with new installs. According to what I've heard the maintenance level is way down.

I have installed some LED fluorescent tubes for a customer of mine. They provided the tubes and following the instructions I removed the ballast, wired sockets direct, pop in the LED's and away we go. Light output seemed to be equivalent to what I removed. Price of each tube was, I think, ~$50.00 each.

 

[Besoeker]

A slightly different take on the LED issue.
We have been using LED panel indicator lamps for over two decades now.
Efficiency isn't so much of a concern. Half a dozen 20mA LED cluster lamps doesn't have a huge impact on say, a 200kW VFD panel.
What started us using them was a customer's request as part of the project spec. At that time they were relatively expensive compared to the traditional units with BA9 filament bulbs but, in the context of the equipment, a very minor cost. Costs have fallen to the point that we would fit the LED indicator lamps on that basis alone.
But the major plus point is their longevity. In 20 years we have not had a single failure to my knowledge.
On older kit, regular failure of the BA9 bulbs was routine. Even when we operated 130V rated units at 110V. Not very good if you expect indicator lamps to indicate what they are supposed to indicate.

 

[Electric-Light]

A slightly different take on the LED issue.
We have been using LED panel indicator lamps for over two decades now.
Efficiency isn't so much of a concern. Half a dozen 20mA LED cluster lamps doesn't have a huge impact on say, a 200kW VFD panel.
What started us using them was a customer's request as part of the project spec. At that time they were relatively expensive compared to the traditional units with BA9 filament bulbs but, in the context of the equipment, a very minor cost. Costs have fallen to the point that we would fit the LED indicator lamps on that basis alone.
But the major plus point is their longevity. In 20 years we have not had a single failure to my knowledge.
On older kit, regular failure of the BA9 bulbs was routine. Even when we operated 130V rated units at 110V. Not very good if you expect indicator lamps to indicate what they are supposed to indicate.

LEDs last a long time and complete failure is rare if you give them some "personal space". When they're crowded together and pushed to the maximum permissible current, the heat will cause some of them to fail. For panel indicator using the typical amber and yellow-green low luminance LEDs, the only criteria is that they're visible. It's not a biggie if the output decay by 75%.

Not that its a surprise, but driving them at the upper limit reduce their life and increase the chance of sudden failure. With semiconductor devices, limit means limit unlike electrical devices where 15A means you can actually push 15A through it and key to reliability is to design with generous margin of safety.

For modern outdoor applications on equipment requiring visibility in sunlight, they may utilize the higher brightness nitride LEDs.

 

[weressl]

LEDs last a long time and complete failure is rare if you give them some "personal space". When they're crowded together and pushed to the maximum permissible current, the heat will cause some of them to fail. For panel indicator using the typical amber and yellow-green low luminance LEDs, the only criteria is that they're visible. It's not a biggie if the output decay by 75%.

Not that its a surprise, but driving them at the upper limit reduce their life and increase the chance of sudden failure. With semiconductor devices, limit means limit unlike electrical devices where 15A means you can actually push 15A through it and key to reliability is to design with generous margin of safety.

For modern outdoor applications on equipment requiring visibility in sunlight, they may utilize the higher brightness nitride LEDs.

I think that for the sake of this discussion it would be good to discuss only the LED and power supply engines for illumination purposes and not for indicators. The difference of requirements may not be as great as between a poer transformer and a current transfomer, but not that far from it either.

 

[Jacob S]

Aren't the LED drivers the weak point? It is common place to have DC PSU's fail, and aren't LED drivers basically just that? I can see these failing all the time in these tubes.

 

[Electric-Light]

Aren't the LED drivers the weak point? It is common place to have DC PSU's fail, and aren't LED drivers basically just that? I can see these failing all the time in these tubes.

LEDs are like fluorescent lamps in that they're not self regulating so additional circuitry either on board or external current limited power supply is needed. The said device is a ballast even though they don't like to use that word.

Yes, it's a link that can fail, just like in fluorescent system.

Also, another weak point is that LEDs life span is rated to 70% of initial output. A 30% decay in output is unthinkable with fluorescent even in the days of halophosphate F40/CW. Modern T8s lose no greater than 6% over their useful life which could be as much as 52,000 hrs.

LEDs need to be paired with a dimming circuit with a feedback control System needs to be oversized compared to fluorescent. The system needs to be overspeced by about 1.5x. When the chip is near the end of useful life and has decayed to 70%, the system needs to be operated at 1.42x the original power to compensate for chip decay.

This can not be done with pushing them beyond specs, so the system needs to be spec'd so that initial output at maximum setting is 142% the needed output. Initially the system is dimmed down and operated at 70% power. as the chips degrade through natural decay, it would have to be gradually ramped up to overcome it.

 

[brantmacga]

Our primary customer is spec'ing nothing but LED's now. We've installed probably close to 3000 interior LED downlights for them in the last couple of years (new and retrofit), maybe a couple hundred LED lay-in's in the past few months. They're also using 70%+ exterior LED lighting.

I've only replaced one LED lay-in that i know of for them, and maybe a couple of the 6" downlights that failed within the first day or so. I know two years isn't a long time, but i would've replaced several fluorescents by now in the same facilities.


Its here people; if its not to you yet, its coming. its getting to where you don't have to sell anyone on LED; the customer knows about it when you show up. better get ready to start pricing accordingly for the lack of maintenance calls you're going to get. that's what the manufacturer's seem to be doing.

 

[Electric-Light]

Our primary customer is spec'ing nothing but LED's now. We've installed probably close to 3000 interior LED downlights for them in the last couple of years (new and retrofit), maybe a couple hundred LED lay-in's in the past few months. They're also using 70%+ exterior LED lighting.

I've only replaced one LED lay-in that i know of for them, and maybe a couple of the 6" downlights that failed within the first day or so. I know two years isn't a long time, but i would've replaced several fluorescents by now in the same facilities.


Its here people; if its not to you yet, its coming. its getting to where you don't have to sell anyone on LED; the customer knows about it when you show up. better get ready to start pricing accordingly for the lack of maintenance calls you're going to get. that's what the manufacturer's seem to be doing.

I'm just saying LEDs typically do not make economic sense. It may make sense in the dream world of where there is no such thing as interest and money is available for no cost so long as you pay back eventually.

But in reality, contractors like them, because its lucrative and there is value in lump sum money.

For a cab company, a Toyota Prius might not make sense over Scion xB which they're using these days in operating cost.
But, if their goal is to appease tree hugging liberal yip yaps and want to pull environmentalism and sustainability PR stunt through social media marketing, it might make sense. So I would wager that the cost difference is more of a advertising expense more so than capital equipment. So, I feel that demonstrable technical advantage isn't always needed to sell things.

There's a lot of misconceptions about LEDs. They're always more efficient than "fluorescent" is one of them. "makes almost no heat" is another. Seldom mentioned is the fact that they lose six times as much relative light output compared to modern T8 lamps over the useful life span. (5% vs 30%).

Another common issue is that the reduction in output often exceeds the reduction in power.

One of the cylinders is misfiring in my beater car and gets horrible mileage. But because of its brand image, it is seen as environmentally responsible than a full size SUV which gets about the same mileage on the newer engine that burns much cleaner. Liberal yip yaps would see the SUV as earth wrecking conspicuous consuming bastard while my crap beater is seen as frugal despite the fact it pollutes many times more.

 

[Besoeker]

I'm just saying LEDs typically do not make economic sense. It may make sense in the dream world of where there is no such thing as interest and money is available for no cost so long as you pay back eventually.

With respect, I think you might be quite mistaken about that.
The cost of running the LED equivalent downlighters in place of the halogens in my office is about ?50 ($80) less per unit over projected lifetime. Which is about ten times the cost of the LED unit. So even if the halogen cost nothing, which it doesn't, and didn't need replacing, which it would, I'd still save my 50 squid minus the cost of the lamp.
It's a no brainer.

 

[Electric-Light]

With respect, I think you might be quite mistaken about that.
The cost of running the LED equivalent downlighters in place of the halogens in my office is about ?50 ($80) less per unit over projected lifetime. Which is about ten times the cost of the LED unit. So even if the halogen cost nothing, which it doesn't, and didn't need replacing, which it would, I'd still save my 50 squid minus the cost of the lamp.
It's a no brainer.

Perhaps I should have been more explicit. I meant compared to existing high efficacy lighting options, so compact ceramic metal halide or fluorescent. Also, depending on burn time per year and energy cost, the expected payback period assuming common interest rate might not meet expected project ROI.

Say you have an old V8 car that gets 12mpg. The payback period of Prius comparing it against that car is favorable when you don't introduce a normal Honda Civic into the equation and get into differential cost payback period.

Given the same usage per year, how long does it take to recoup the cost difference of a hybrid car and high mpg car through the difference in the mpg between the two replacement options? It is necessary to compute opportunity cost or interest expense.