Long term real life LED durability LM-79, LM-80 and TM-21 tell you nothing about

[Electric-Light]

Presently China dominates the LED industry. From components to manufacturing. The protocols around LEDs are still constantly changing. LED products are still quite fluid and standards are still developing and performance for many LED products are still unstable.

Lumen maintenance of conventional sources are backed by historical experience and data. LED lifetime is still significantly based on projection and faith.

There are three prevailing standards in use around LED luminaries.

LM-79: Brand new LED fixtures are rigged up in a lab and testing starts up. Data taken after fully warmed up. The outcome is entirely based on the performance on the evaluation day. Even the tests are often based in China and finding errors like unexplained notion like "12v 50Hz" on lab report for products sold under Philips label makes me doubt the competence of some tests.

LM-80 is a set of long term tests taken at 55C, 85C case temperature and one more temp of manufacturer's choice. Measurements taken every 1,000 hours or more for 6,000 or more hours. This is a component level test of the LED source which is not strictly defined.

TM-21 is an extrapolation (predicting forward) for up to six times the duration based on LM-80. A test batch of 10 for 5.5x extrapolation. 20 or more for 6x extrapolation. A small error can wildly skew the result just as when you're extending a line with a straight edge . The lab's room temperature is kept at 77F.

Overall I am not sure how meaningful it is. A 60W LED lamp bulb sold as general purpose 25,000 hour, for household use, suitable for use in totally enclosed fixture. Well. does this mean 25,000 to expected LED fading of 70% lumen maintenance in horizontal, open air use or 25,000 to 70% in a jelly jar? I'm pretty sure not the latter.

http://www.lightingfacts.com/downloads/lumen_maintenance_faq.pdf
So, when you see "lasts 10 years" means that they predict that it takes 10,950 hours at 3hrs/day for the LED elements inside to fade to 70%, but not take into account of other things.

You have probably seen a photographer use an umbrella to diffuse a flash. When you shoot light into an opal, some pass through but some scatter off the surface you aim the light into. This rebounce off white surfaces within the light fixture. So the conditions of white finish inside light fixtures absolutely matter. Yellowing will cause lumen loss and color shift.

This ETI SSL 546061XX LED luminary shows enough yellowing to infuse color into reflected light as well. This is an EnergyStar logo bearing, 830 lm, 11.5W 830lm LED surface mount luminaire advertised to last 35,000 to 50,000 hours depending on where you look.
The sample on left has perhaps 8,000 to 10,000 hours. Right has almost zero.

Consumer who choose to write reviews have written long before they have even reached 5% of rated life and good reviews have some purpose but they are not a good indication of long term durability.

The fixture is made of plastic and it showed obvious yellowing/browning. Some staining or scorching also visible inside the rubber goop within the LED elements. The white lacquer on circuit board/reflector, along with the fixture optics both show yellowing. LEDs generate white light by using a deep blue light to drive yellow phosphor applied directly onto the elements and the two mix together to form white. It is around 450 nm which has enough photon energy to be chemically active. This is why this spectrum of blue is used for industrial adhesive curing. The fixture claims 35,000 hour life but it has nothing close to that amount.

As this article shows, absorptions of blue photos can damage pigments over time. http://www.dailymail.co.uk/sciencet...ieces-artists-including-Van-Gogh-C-zanne.html

The red pigment in PVC wire inside this fixture sustained photon bombardment damage from exposure to intense LED light. You can see the slight difference in pigment damage along the hole where the light does not shine as strongly. The pigment with strong absorption of for blue left the pigment with strong blue absorption bleached out. The slight color difference in plastic frame on two units is due to discoloration from heat.

Thermal discoloration in general on plastic components as well as deposited capacitor fluid outgas on part of the ballast cover. Overtime, optics and LEDs face threats from contaminants such as outgas from fixture and and LED ballast components.

It is a myth that LED does not attract insect.

 

[tom baker]

OK whats the point, can you summarize?

 

[winnie]

In a nut-shell: LEDs are getting the snake oil marketing treatment (like just about everything else). Understand the testing standards so that you can make apples to apples comparisons with other tech. Sometimes the testing standards don't give enough info.

In particular: if you 'push' your LED emitters harder than you can get more initial lumen output (the lamp is brighter) but the heat makes the output decay rapidly. In some applications this isn't a problem (closet lights) in others you need long term lumen maintenance.

-Jon

 

[Electric-Light]

OK whats the point, can you summarize?

It's not the impression people get when they're told about LEDs.

LM-79: Performance Test. Measurements are taken on a brand new complete product to validate specs.
Actual product model is tested for performance, but a brand new product is used with zero hour of use. The reading is obtained the same day. It's like a product review. The correlation between tested samples and actual items on the shelf is a variable depending on the product. It's like a test drive on a brand new car.

LM-80. Endurance test. It is ran on a handful of engines and long term durability rating shared between models using the same engine.

TM-21. A handful of samples of engines were tested for 6,000 miles. Based on the findings, the performance loss up to 6x the tested time is projected forward. The projection process leaves a lot of room for individual manufacturer discretion.

Consumers are given the impression that the testing is based on the PRODUCT, not just the LED elements that makes it. They're not buying a fixture and installing their own lamp. They're buying a complete product with permanently installed LEDs. Things like fixture optic degradation or probability of failure based on LED ballast failure was not part of the test. Something troubling about this product is that functionally essential reflective surfaces are not made with materials with durability as specification grade conventional fixtures while claiming 35,000 or 50,000 hour life. An impression is generated that the only change that takes place is LEDs slowly fading over time but the rest of the fixture is like a simple hand tool with really nothing to go wrong.

 

[IMFOTP]

It is a myth that LED does not attract insect.

I can attest to that, all exterior LED fixtures, even "high quality" wallpacks have been inundated with bugs. The "no bugs" story is a flat out lie told by salesmen.

As far as components not lasting as advertised, yes, I agree the published data is under "ideal" conditions and is based on some iffy assumptions, but its still better than other options for applications I deal with. In a classroom setting we can get a $110 dollar LED troffer with 0-10V dimming that has a rated L70 of near 100000 hrs. That's almost 30years at 10hrs a day. Now...will it last that long? I'm betting not, but nobody can reasonably expect to get that kind of life out of anything costing $110. The ROI compared to a dimmed fluorescent equivalent is a little over 3yrs. Its a no-brainer, LED wins even if they can't perform completely as advertised.

 

[Electric-Light]

The ROI compared to a dimmed fluorescent equivalent is a little over 3yrs. Its a no-brainer, LED wins even if they can't perform completely as advertised.

Why compared to dimmed unless it's a end-user driven need? LEDs always struggled with making ROI without finding ways to pad up the alternative. Dimming for fluorescent was always available but they were only used when it was functionally necessary.

Field evaluation is often done by LED sales reps with methods passed down from poor estimation technique they learned in sales training course.

Fact: dimming has always been in high demand for homes.

Fact: Fluorescent dimming ballasts that can dim down to almost blackout has been around for decades. Magnetic fluorescent ballasts actually worked very well.

Fact: electronic dimming ballasts has been around since the 1990s.

Here's a 1990s electronic dimming ballast

If we're talking about the replacement of existing systems, the most dramatic error is introduced in assessing the parameter of existing power consumption. The crap gets passed down from the big brand's LED marketing department which is why there'd been non-sense like Philips Lighting who should know better than using setting point of reference for two 4' T8 as 72W. When it comes to new applications, there is no before/after but options 1 vs 2, vs. 3.

LEDs win for flashlights, household and some outdoor applications. Also, every battery powered lighting where lumen maintenance can be usually completely ignored.

As far as components not lasting as advertised, yes, I agree the published data is under "ideal" conditions and is based on some iffy assumptions, but its still better than other options for applications I deal with. In a classroom setting we can get a $110 dollar LED troffer with 0-10V dimming that has a rated L70 of near 100000 hrs. That's almost 30years at 10hrs a day. Now...will it last that long? I'm betting not, but nobody can reasonably expect to get that kind of life out of anything costing $110.

LEDs become even less affordable as a retrofit especially when it comes through a retrofit sales company. The price has to cover the material, labor and sales company profit which makes ballast retrofit one of the worst proposals even though it might make sense for homeowner DIY because, he's more concerned with the cost of having to buy a replacement for a failed ballast, and labor is no cost.

LED selling is based on providing good performance upfront, collecting full payment with promises of long term energy savings, so there's a demand for lowest initial cost item with good initial performance. It's no accident that initial lumen is always emphasized for LEDs. The best selling lowest priced ones have durability problems. Instant light at full output and great calendar life in stop and go use while giving fair lm/W is essential for residential and low occupancy area market where lumen maintenance into five figure hours isn't important.

However, LED retrofit sales are usually trying to convert small and medium businesses with traditional fluorescent applications with long annual burning hours. You can see just how badly builder grade LEDs deteriorate. For 24/7 use like convenience stores, it degrades more in first year than premium fluorescent does in lifetime. If you need 700 lumen output at the lamp, the best way is to trim it down just over 700 with minimal drop rather than use something that starts at 1,000 and decay to 700.

Compare the cost of 6 lamp T5HO system w/ the highest grade premium lamps vs system to 30000 lumen class LED

When you look at actually matching fluorescent lamps in long burning applications, LEDs become extremely expensive. The incremental cost difference to squeeze 8% saving of 7 watts per fixture comes at the cost of nearly $80/fixture which results in no ROI in A vs B comparison and it is only possible to generate ROI against a very inefficient existing system.

 

[IMFOTP]

@Electric-Light

You are aware that the LED industry extends beyond LED T8 Lamp Replacements, correct?

 

[Electric-Light]

@Electric-Light

You are aware that the LED industry extends beyond LED T8 Lamp Replacements, correct?

The first post was about an LED luminaire which has a permanently installed ballast and LED module pair. Any problem means the entire fixture becomes a plastic-clad e-waste with nearly zero recyclable content and ends up in our landfill or atmosphere. It also features inferior quality optics that yellows which does not affect the test criteria in place but causes lumen loss and color shift. "Affordable" price compared to traditional fixture can be attributed to the use of shoddy materials. The LED board is assembled in China on Chinese sourced phenolic board with LED elements built with Chinese sourced chemicals. Made in China content is approaching 100%. All these Made in China LED and solar stuff is more than undoing buy American movement preceding the LED while externalizing pollution generation.


The phased out 75W A19 filament lamps were 1250 lumens. An LED type lamp not suitable for enclosed fixtures may put out 1100 lumens brand new and says 15000 hour life and touts 75W equivalent. There's an added problem that such things fail to provide a true match. It infers it matches 75W lamp performance for 15000 hour life, but the life rating means the time they estimate for it to fade by 30% to 770 lm (70%) under the logic people are unlikely to notice it. If this logic was applied, a modern point of reference is a 1,000 hour 43W, 800 lm halogen A19 which has a negligible output degradation. 75W A19 is phased out, so it makes more sense to compare against currently available consumables.

Use of absurd life rating like L70 is responsible for inappropriately casting LEDs under a favorable light despite they produce unfavorable light compared to filament lamps (pun intended). L90 rated LED luminaires that do not rely on automatic energy consumption boosting LED decay compensation ballast are very uncommon.

 

[jsmalone1]

To keep it simple, I'm 5 years in on many LED installs, 3 years in with my homes and many other customers & all my friends now that the price point is reasonable. I had 1 (one) LED light fixture fail at install and that's it. I believe everyone is entitled to their opinion and so am I. In my opinion 60-80% energy savings is not snake oil, that's what my meter consistently shows. Or better, for the negative comments I have attracted here, it's math--60 watts or 9 watts? 85%? I no longer repair or replace any fluorescent, HID, incandescent fixtures or bulbs, etc. I have 60 watt LED bulbs approaching 15,000 hours of use and not a single failure. Plus there is so much more, no apparent cycling effects, huge temperature operating range with no change in light output, and of course, very long life span. BUT, I have had many people that insist on fluorescent and I just tell them that I'm sure they are on sale, good luck. Or even better, ask my wife and all my friend's wives. They all love LED and have all asked why we didn't do this sooner---price.

 

[winnie]

One of the key points which electric-light brings up is that LEDs are often designed in ways where they will get dimmer over time (still using the same power) without outright failing.

There is a 'specmanship' issue here, designing to give the best initial specs even if the degradation over time is worse, or designing for best specs over narrow test conditions, rather than real world conditions.

Yes, LEDs are quite good and getting better/cheaper. But they are not necessarily the best technology for every situation, and like every other tech the marketing folk will coat even good tech in snake oil.

-Jon

 

[Electric-Light]

To keep it simple, I'm 5 years in on many LED installs, 3 years in with my homes and many other customers & all my friends now that the price point is reasonable. I had 1 (one) LED light fixture fail at install and that's it. I believe everyone is entitled to their opinion and so am I. In my opinion 60-80% energy savings is not snake oil, that's what my meter consistently shows. Or better, for the negative comments I have attracted here, it's math--60 watts or 9 watts? 85%? I no longer repair or replace any fluorescent, HID, incandescent fixtures or bulbs, etc. I have 60 watt LED bulbs approaching 15,000 hours of use and not a single failure.

The power going in is only one side of the story and equivalency is not applied consistently. 400W MH is around 42,000 lumen when it is new. A GE 165W retrofit LED for 400W MH is listed as 20,000 initial lumens. This is probably a reasonable estimate for MH moments before they fail.

The theory for L70 is that 30% is around when we start to notice a drop. Today's LED lamps are rated at 10,000 to 15,000 hours to 30% decay, but inappropriate use will cause an accumulation of light level loss with each generation of retrofit.

A 100W legacy light bulb is 1,600 lumens. Making comparison to this is a moot point since incandescent lamps are only rated 750 to 1,000 hours and traditional wattage are no longer produced. LED lamps with 15-1600 initial lumen are sold as 100W equivalent. After 30% degradation, it is closer to 1,100 lm which is equivalent to presently legal 53w halogen. Filament lamps require replacement often but the light output loss is negligible so 1,100 lumen is fairly well maintained until failure. The 53W halogen is not really equivalent to 100W 1,600 lm but it is equivalent to 1,600 lm LED and LLD 0.70 is the current IESNA LED design recommendation and for a 10,000 hr lamp that burns thousands of hours a year, I strongly agree and personally I would split it down the middle for a lamp with L70 of 50,000.. but the IES recommendation is to still use 0.70.

53W halogen yields almost half saving in power, and the 30% reduction in output will only be perceived like a 17% drop. Retrofit equivalency subjective evaluation on the day of installation and matching the head of LEDs to tail of HIDs in a bid to sell watt cutts will result in accumulation of lumen loss.

My justification to match terminal lumen of LEDs to initial lumen of halogen is the fact halogen has negligible light loss as well as ambient related variation. I'm in no way arguing these halogens are anywhere near as efficacious as LEDs but you get what's going on here. They're often pushed way too aggressively to cut down initial cost and wattage to crank out ROI on paper without producing comparable replacement.

Plus there is so much more, no apparent cycling effects, huge temperature operating range with no change in light output, and of course, very long life span. BUT, I have had many people that insist on fluorescent and I just tell them that I'm sure they are on sale, good luck. Or even better, ask my wife and all my friend's wives. They all love LED and have all asked why we didn't do this sooner---price.

They are quite tolerant on lower end of temperature, although HIDs and incandescent are also ambient independent. HID's warm up time is either show killer or irrelevant. Integral ballast LEDs and CFLs both suffer loss of useful life at high operating temperatures.

One of the key points which electric-light brings up is that LEDs are often designed in ways where they will get dimmer over time (still using the same power) without outright failing.

Which is a trait also held by mercury vapor lamps. They were phased out in favor of metal halide that do burn out. The general practice of using initial lumens for matching and presumption of lasting at same performance until "L70" value needs to stop, because this is not appropriate except when comparing to MH using values more reasonable than end to initial.