Inside An LED Light

In the renewable energy world they use a technique to calculate total energy cost called LCOE -- Levelized Cost Of Energy. This is basically a calculation of energy footprint starting from digging into the ground to get at the oil or iron ore or silica to make the devices, all the way to the cost of power delivery including the trucks and cars that serve the utility companies -- and everything in between. LCOE calculations show that renewable energy is significantly less wasteful and the real cost of the energy is much smaller per kWhour than fossil-fuel energy. I like to bring this up when any of my friends who are not renewables fans tell me it's "unfair" that utilities subsidize residential solar energy, since it "unfairly" competes with fossil fuel-based energy sources that are not subsidized. Well, there are studies I've read (sorry no links but maybe I'll find them again) that take into account LCOE and I tell them the findings (as much as I can remember) and they usually shut up :) Then I hit them with the front-end subsidies that the oil+gas (and automotive) industries have always enjoyed, to counter the argument about how unfair the back-end (utility) subsidies are for renewables. That usually ends the argument. But my friends are reasonable people, and so it doesn't work on everybody :)

Peter Knazko

Hey Fran, Great video, certainly got me thinking. I see others have already provided some sources on the LED manufacturing impacts, wanted to provide one more that synthesizes and references a few other studies. Might be handy/interesting. https://scholar.google.com/scholar?hl=en&as_sdt=0%2C48&q=Aalto+University+led+carbon+dioxide&oq=Aalto+University+led+carbon+dio#d=gs_qabs&u=%23p%3DbfkBRlaXQBcJ If you do download the PDF the section with the interesting bits is on page 299.

TheSpaceBear

My 2 cents as an electrician - When I started installing LED lamps for customers, I was using Philips, which at the time were going for around $35-$40 per lamp (PAR30 style). They were a lot more affordable than the $100+ lamps that were going into museums and galleries, and the competition at that time was, oddly, from brands that were more known as office electronics manufacturers than as lighting brands. At that time the biggest issue was color rendering at the low end of the price range, but most lamps were made to commercial specs and had crazy long warranties like 25,000 hours or more. Those warranties stopped when more manufacturers got involved and prices dropped. Nowadays it's hard to find a consumer grade LED lamp with any kind of lifespan warranty on it, but you can get them for a few bucks each and the color is decent. Commercial grade still cost a lot and have warranties that last years, but are mostly limited to outdoor lighting for poles, wall packs, and big flood lights. The thing I'm glad to see is that more and more manufacturers are moving away from trying to make LEDs fixtures that use a traditional bulb design and more often use integrated LEDs in a tape or custom module. That allows for better thermal design than trying to fit 6 gallons of poo into a 5 gallon bucket, which is what you end up with in lamps like the GU10 MR16 that you tore down. So really, the main problem is quality. A well made LED using slightly better ingredients can last years and years, negating the carbon footprint it takes to make it. Cheap LEDs are probably worse than bad at all levels.

Peter Argyropoulos

I wonder if LED bulb manufacturers have agreed on a maximum life expectancy with planned obsolescence like they did with incandescent bulbs back in the day. https://spectrum.ieee.org/tech-history/dawn-of-electronics/the-great-lightbulb-conspiracy

I have bought so many of these only to have many fail after only 1 year. Fran's illustration of that funky slip contact to the LED board makes it obvious why, for some of them. But the trend dictates what Home Depot stocks, and now it's all LED, for better or worse. One day someone will show us the real cost. And not going to say anything about the RF noise for us poor Hams from the SMPS in these things {head slap}. Anyway, it seems like 400% of resources to get 90% energy efficiency may become an issue when global warming and the coming Democrat landslide converge. Not sure these great electronics solutions are really better than Edison's filaments. Maybe that's still the answer if the filament can be made less lossy.

Peter Knazko

That is a great question,wow!

William Nimmo

I've read through all the comments here and on YouTube, and one thing I'm not seeing is a discussion of the cost benefits to LED lighting in terms of who's paying. Even if we suppose that the embodied energy of an LED bulb is the same as a halogen over the course of its lifetime (I suspect strongly that it's not true, in general), non-LED lighting has costs that tend fall more heavily on the individual or small municipality (in the case of streetlights), whereas the cost of LED lighting shifts more toward environmental costs. That's not great, obviously, except for the fact that they ARE so much more efficient and, in general, much longer-lived. I know there have been articles in my local newspaper about how small towns around me are switching their streetlights to LED despite the high initial cost because they'll pay for themselves in just a couple of years. And the savings in air conditioning in the summer (when electrical demand is highest) by not running incandescent bulbs can mean a lot to individuals and businesses. This is an economic equity issue. Everyone who uses LED bulbs, especially cheap ones, has seen cases of crib death or premature death. But as the saying goes, the plural of "anecdote" isn't "data," so the curve of failure-over-time needs to be looked at for any given LED bulb to know much about whether it can offset the embodied energy of its manufacture. There's likely a bathtub curve there somewhere. This is probably also something that could be addressed with regulation, if there were any political will to do so. If LED bulbs were manufactured to mandated standards of longevity, they'd cost more, but then we'd be getting a clear environmental benefit for that added cost. Incidentally, ship and train transportation is absurdly efficient in terms of weight x distance / fuel. As for color rendering, better phosphor formulations help mitigate that to a large degree. This is one situation where the longevity of LED bulbs is a disadvantage, since bulbs with poor color characteristics can stay in service for longer. But I've looked at my LED bulbs with a cheap little spectrometer I have and I've definitely seen an improvement over the last 5-10 years. I lean hard left politically, and I'm not in any way a climate change denier, but the environmental movement has a particular strain of puritanism and hyperfocus on climate change that seems to permeate the movement and cause undue amounts of guilt in really well-meaning people. We humans will never have NO negative impact on the environment; we just have to do the best we can with the tools we have. I think LED lighting is clearly a part of that, and I think regulation could be a bigger part of it. And, of course, I'd love to see someone do a deep-dive on the embodied energy of LED bulbs, because hard data is the only way to make good decisions. Thank you for coming to my TED talk.

Circuitmike

Might be interesting to know why it failed. See if you can power up the LED chips from a bench supply. I bet they are working. It's probably the driver that failed, might be just a capacitor for example. On really built down to a price stuff they can end up using components that can't handle a high enough temperature. Then this can be compounded by the light fixture not allowing enough airflow. For example that kind of bulb is often installed in a ceiling recess. It really needs some laws to enforce the longevity so they can't save a cent using unsuitable components.

Intetesting video, and thanks for sharing. I recall reading an article that for every laptop made, the equivalent of a car's size of waste is generated. And it went into embedded energy of products to good detail. Phones are a shocking example. I still have and use (second phone) my Note 3, it's on it's 3rd easily REPLACABLE battery now, and physically held up well being 100% plastic (bar the screen) with dents and scuffs, but nothing actually unservicable. Yhe screen is also of a very good quality glass and has barely shown a mark in years. I've got a 512GB microSD in it and it's a very credible music and video player. Another phone (of the Fruit flavour) was a total opposite. Aluminium sides scratched up badly, the main screen is so prone to scratches, it may as well be plastic, and first drop (cased) broke the glass back. Can't get into the battery, can't upgrade the memory etc etc. It seems manufaturers want these products to go obsolete as quickly as possible is what I'm getting at, and they are doing it more and more. I have my own thoughts on LED lamps, and have pulled many apart and tried to decipher how they work. One thing that stands out is they all pretty much push the LED's to their absolute limit hence why there may be so many failures. I actually modified a particular LED bulb as a test by removing the plastic ball off top, and drilling holes in the casing. Even with all that extra passive cooling offered, it still ran frighteningly hot, but turned out on the LED sub-board, there was a temperature sensor, basically measuring the temperature and regulating everything based off that. Best LED's I have is some "linear bulb" replacement which is incredibly simple to boot, using just a single large rectangular aluminium PCB and a driver IC inside a plastic tube. It runs cool and has lasted 7 years on 24/7 with no ill effects at all (and I bought a spare) A heap of "filament" bulbs, I bought a few different types, and then heaps of the ones that ran cool and all (12) are still working perfectly. Many of the high CFI / High PF CFL bulbs I installed (NZ) 10+ years ago are still running with good brightness. Obviously a mature product by that time. (They were Ecobulb /Energy Mad branded for those interested) But heaps in the other house were various cheap brands and didn't fare so well! I think we got 4 years from the best one.

Anton

I do scale modeling. I use a CFL bulb as one of my workbench lights. We occasionally get mice, here in the sticks. One of our cats is a skilled mouser. But she will never hang out in my cave. Maybe it's the smell of some of my paints and thinners etc. Even though it's well ventilated. In any event, I like to leave a light on so any mouse will think there's activity in here. I leave the CFL light on. And I just changed one out, after being on 24/7, for almost 3 years. And no mice! But a dreadful demise awaits those who appear in any other room.

MrBelmosan

I've been using "smart bulbs" with embedded wi-fi for a couple years now. Some used in a closed ceiling mounted light fixture have no way to vent the heat and have never lasted a year. I've opened up one of the failed ones and it's interesting to see what goes into them. They even potted some of the board but left the wi-fi antenna exposed. Fun stuff.

John N Nelson

I've switched every lamp in my house and car to LED, with the single exception being the bulb in the oven. Before that was CFL bulbs, which I switched to from tungsten, but I hated the light spectrum and the mercury vapor problem, so they were retired and recycled the moment LED lighting matured. Both CFL and LED bulbs contain electronics, and like any electronic product, the more of them you own, the more likely you are to observe the bell-shaped curve of product lifespans, where a few die young, most live to their rated lifetime (on average), and a few seem immortal. In my home I bought the same screw-in bulb for use everywhere, and I intentionally chose the least expensive bulb that met my minimum brightness, CRI and efficiency needs. I wrote the date of installation on every bulb with a sharpie, so I could keep track. About 1/3 of them were dead within 18 months, with one expiring just 1 week after being installed. However, the rest have been going steady for nearly 4 years now. My main lesson-learned so far is simply to purchase only from established manufacturers, ones that have had time to improve their processes to reduce the rate of early failure. There are other technical factors to consider as well. CFL bulbs, like any fluorescent light source, flicker at 120 Hz, which can annoy some folks directly, and can also make indoor photography a nightmare. While some specialty LED lights use separate red, green and blue LEDs, both CFL and commodity LED use phosphors that are closely related, as both are excited by UV light. The flicker of CFLs was due to the phosphor decay being much faster than 1/120 second (= 8 ms). LED bulbs are driven by a pulsed current source (PWM) that typically switches at rates far higher than 1/120 second, generally 10x to 100x faster, eliminating the flicker. However, the earliest commodity LED bulbs were based on the old CFL circuitry and also had lots of flicker. My best LED investment, by far, has been my car headlight bulbs. The spectrum and brightness improvements immediately made night driving far safer. In particular, I can read road signs from much further away, and even faded lane stripes show up very well. When researching headlight bulbs, the most powerful ones include fans, due to the limited airflow at the back of most headlights. Electromechanical parts like fans seldom last as long as electronics, so I selected the brightest fanless bulb that was from an established brand, had good ratings, and was affordable. The bulb I chose had 8x the brightness of the tungsten (non-HID) bulb it replaced, which is safer for others sharing the road than those ultra-bright bulbs with fans. Edit: 8x the physical brightness is barely over 2x the perceived brightness, due to the non-linear response of the human eye. No blinding beams from the front of my car!

BobC

A high-quality LED light ought to last many years. Problem is that the good ones come with a higher price tag. Enter Chinese manufacturing, which produces cheap low-quality versions that are sold in mass quantities. The Public is sold on the idea of efficient lighting but wind up buying an inferior product that, long term, doesn’t deliver on the “green” promise.

I had to make some small edits and reupload this one - so sorry if you made a comment on the precious version on the YT side....

Fran Blanche

I think 5 years is a reasonable assumption. The diagram is for that component I am pointing at that looks like a heat sinked transistor...

Fran Blanche

Like I said, I'd like to see the real data, taking everything into account.

Fran Blanche

That lens is indeed a thing of beauty.

Dr Andy Hill

As I say, you have to take into consideration all the energy used, including all transportation of materials and the power source for the manufacturing processes, which in this case is all coal.

Fran Blanche

I wouldn't use price as a gauge of the waste created and energy used to produce the item - as I said, it is made in China for a reason!

Fran Blanche

This reminds me of a Big Clive video, not that this is a bad thing, lots of people do tear down videos, all with their own style. I agree to overcome the manufacturing footprint, I think it would have to survive at least 5 years. I think it would have to operate for 10 years to really make sense. Like you, this is just my speculation. I have had some LED bulbs last over 5 years. On another matter curious about the circuit diagram that popped up in the video and wasn't mentioned. Or did I have a micro narcoleptic incident and miss the explanation.

Dr Andy Hill

Of course those materials would already be imported to be used for things other than the bulb so it is shared over multiple products. Also the quantity of products produced also offsets the carbon footprint per item of importing those resources.

William Alsing

I’ve been asked this question before and also had people tell me [X] technology is terrible for this manufacturing energy consumption reason. Here’s a of a report that goes into detail about this by the US Department of Energy, there are several other good articles listed at the end of it. https://www1.eere.energy.gov/buildings/publications/pdfs/ssl/2012_LED_Lifecycle_Report.pdf tl;dr, figure 1 pretty much spells it out, the power consumed by an incandescent bulb over it's lifetime to produce light vastly outweighs that of the energy consumed manufacturing an LED light. If you think about it if it took more energy to manufacture an LED light than it takes to run an incandescent light over it's lifetime then the energy savings of an LED light wouldn't add up to the price difference and no one would buy them.

WizardTim

I've heard that solid state halogen supplies are particularly hard on non-resistive loads. That doesn't excuse an LED lamp lasting only a year though. These replacements are often marketed as lasting an entire decade, but of course with no warranty to back it up. On another note, this might be skewed, but I'd say one measure for the amount of energy required to make bulbs is the price tag. Energy costs money, that cost is folded into the unit cost. Again, it might be skewed but if you're not saving money by buying LED just on a money-per-longevity standpoint, it's probably not saving the planet either.

Great Joe