It is 2023 and incandescent is still better than LED

Bees wax versus tallow. Candle versus oil lamps.  Whale oil versus petroleum oil.  Oil versus gas.  Gas versus arc lights.  Arc versus incandescent.  Incandescent versus halogen versus xenon versus fluorescent.  And now led.  The wheel goes on and on.  I bought my first cfl 20 years ago.  I retired the last of the oldest ones 2 years ago.  I'm awaiting the death of the rest so I can justify replacing them with the new leds. Some cfl were horrible but those from well known manufacturers performed well the crummy ones performed badly.  My first forray into leds involved wiring diodes with a resistor for 12 volt units.  Those early ones with the first flawed manufacturing processes really were horrible.  Too blue with bad quality control and cheap components. Fast forward a decade and off the shelf units from national retailers are cheap and longlived.  As to the light quality Nothing holds a candle to daylight if you will pardon the play on words.  As to complexity I would say leds allow you to lessen the complexity of your power generation so it should count as a plus for them.  One of the main sources of complexity and failure is the driver which has to regulate voltage and also fit in the tight package of the traditional Edison bulb design.  Think of the screw in leds as a stopgap measure at best.  More and more the new units are enclosed with a separate driver to control one or a string of fixtures.  These new gen ones are modular, replaceable and extremely longlived. Also since they generate very little heat their enclosures are much less complex.  No metal junction box and elaborate heat dissipating solutions incandescent lights require. Available in many spectrums based on taste and task. Stick to well known manufacturers do your research and you will love them.
Led lover David Baillie

I find to very odd, that as LEDs operate with 2-3 voltage and you use 230 voltages from the wall socket, it is 100 times more they need, so we use energy to transform energy: also need resistors, which heat up usually more than the LED itself..

If you have right voltage, you dont need resistors or/and transformers with LEDs.

I found this monster sized incandescent the other day and realised I would have to post it here!

This 1000W bulb was run on a dimmer switch (usually underpowered) in an outbuilding workshop at my parents old house. They built the workshop to help construct the house in the early 80's and it wasn't demolished until around 2004. So I reckon its has had at least 20 years of use of fairly consistent use and the filament looks as good as new. It wouldn't suprise me if my dad bought it second hand as didn't remember him having any spare giant lightbullbs of the same type.

What an eye opening post. Wow.

I searched through all of the posts that mention battery back-ups (electrical inverter power), because that is what is most relevant to me in Haiti. I buy CFL's, not because they "save the earth" or are "cheaper in the long run." I buy them because they let my inverter power run for a longer amount of time. In fact, in Haiti, the corner-store boutiques refer to CFLs as "inverter lights," because they help inverter power to last longer.

I was wondering how these calculations hold up when taking this matter into account. What about the cost of a generator being run more frequently? What about the decrease in the life-span of an inverter battery due to increased cycling? What about the strain on the actual inverter itself?

Unfortunately, I could not find any direct references to research in this vein. Perhaps one of you could point me to some...

I'd be happy to provide numbers (cost of batteries, cost of fuel, generator servicing, etc.) if someone wants to try to calculate. Or, if someone has a suggested calculation, post it here and I'll plug in for the variables.

To my mind, the discussion begins and ends with incandescents produce heat. I don't do heat. Most of my life, we had no heat, other than the occasional fire, solar in the late 70's, early 80's was terrible. Then as an adult, we couldn't afford it. I have heat now, but wouldn't use it, if I lived alone. As long as the house doesn't dip below 40 degrees, I'm good.

Cool thread. Any idea where to find incandescent lighting experts? I'm looking for engineers who know about manufacturing parts - and have a good breadth of knowledge about the range of incandescent lights out there (in various frequency ranges). I'm working on a project and looking for consultants, engineering students, or whomever has the most experience in this "old fashioned" technology... Easy to find LED experts, of course

I might posit that LED flashlights/worklights are greatly superior to  incandescent. Much brighter, and they give longer battery life.
The reason that they aren't as pleasing to use for interior illumination is because any given LED can only emit one wavelength of light, so to make "white" light, it is actually a composite of emitters. It kind of works, but your eye/mind seems to know better.

i use all 3 for different applications, here is why:

incandescent, for heat and light, for the chicken coop in the winter, when there are days they are completely snowed in and it might take some time to shovel a path to them to let them out. the light helps keep their water from freezing entirely.

fluorescent for indoor grow boxes, because they are fucking everywhere and last a long time, and dont make as much heat as incandescents(important for an almost fully enclosed grow box, i prefer to smoke and/or cook what im growing AFTER its been harvested), i stock piled CFLs as they are everywhere they were being given away for free, so have a bunch in storage. long flourescents are also available everywhere, folks get rid of them for free so much.

LEDs, for the general lighting in the house. these were also being given away for free in a lot of places. i don't have them as stocked up as the CFLs, but have enough for most of the rooms in my house. i like how efficient they are electicity wise. id like to learn to make LED arrays myself, so i can grow more indoors, year round microgreens, clones and seedlings to market and plant in spring.

i don't know if/how significant differences there are in the industrial processes for making them vs other bulbs, at a certain point i consider that splitting hairs, and greenwashing. i take no responsibility for the world capitalist industry has built for us. im doing what i can with what i have, and i have little to no power in these dynamics. my impact is insignificant compared to systems beyond my control, challenging those systems, like the military, and industry itself, are my priority, not peoples consumption habits, which are divisive, and unecissary, egoic and elitist, a distraction from organizing our power as people to challenge the destructive systems.

As for heating goes, we live in a newer mobile home, rather efficient design, lots of south oriented windows with central oil heat, i shut the vents to underused rooms in the winter, as well as my own bedroom. Its one of the most efficient homes ive lived in in Maine(not saying much, our housing stock is getting super old, poorly insulated, drafty). Its more efficient then electric heating, both cost wise and on the environment.

We'll be using more wood this year in a conventional fireplace. I'm trying to get permission from the landlord, and help with buying materials, to build a greenroom on the deck, to get some passive heating of the house, and a warm entry into a grow room. Plan to extend the greenhouse out from this 3 season deck, connecting the two, and use a rocket mass heater to keep it all warm. On sunny winter days i plan to just open the porch door to let in all the excess heat from the greenhouse/3 season room, into the house, which is already south oriented and toasty on sunny days.

It's 2017, and I still use LEDs as my favorite illumination.
My favorite USB powered light bars are all Piranha LEDs, and
My favorite flashlights are all KREE

paul wheaton wrote:

So I am going to say that both bulbs will last 10,000 hours and electricity is a dime per kilowatt.  100 watts for 10,000 hours is 1000 kwh.  $100.10 worth of electricity for the incandescent and $35 for the LED.  A strong win for the LED.

I have never found a incandescent to last more than 2 years or so without dropping so much light it becomes a glowworm. At present we have all LED lights and in the last three years have had to replace one bulb where the casing cracked. Now back to the price even assuming both last the same time our electricist costs 3x as much as that example 30c a KWH the largest LED bulb we have is 12W it is rated at 70W equivalent so 10000 hours on the fictional 70W bulb and the led comes in at (120KWH @.3) $36 for the LED but (700*,3) $210 for the Incandescent, now lets say that over the winter when we have heating on that bulb runs 12 hours a day and "winter" here lasts 5 months (153 days)  We have another bulb that also runs 12 hours it's not as powerful but lets say it's the same and would be replacing the same for convenience, and all the other bulbs in the house added together lets say they are another 12 hours so 36 bulb hours a day. that's 2.5kwh for (75c) incandescent and 432 (12.9c) for LED over the winter LEDs would cost $19 and $114 for incandescent. so $95 difference over 5 months

So for some of us LED's are hugely cheaper than any other choices. While we do not HAVE to heat our back room where we live at all deliberately because of two humans, two dogs and two large computers, we do still have to heat the rest of the house even if we do not mind the house being cold, a cold house gets damp and damp turns to mold which is terrible to health. And the fridge freezer stops working at 9C so it has to be higher than that! (Our total heating and hot water bill (wood) for the entire year is $350)

Now where I am not going to use LED's is on seed starting I do not understand why but the growing LED arrays take so much power we literally could not run enough in this house to start my seeds, so they will start under tubes 16 of them!

Here in Nashville and the surrounding area, service electricity runs hot. There's 126-127 volts at an outlet in any given house, and that's known to shorten the expected or labeled life of lightbulbs. In my current (no pun intended) home, and this has happened in other houses I've lived in, I have incandescent eave flood lights on one corner of the house that I've never changed and still work, and they've been there at least the 8+ years I've been here. I also have another corner of the house with eave lights that burned out regularly until I screwed in an LED flood light, and it's been working for four years now. I used to be able to get 130v rated incandescent bulbs at a mom & pop hardware store but not anymore, and those bulbs would give me longevity with Nashvilles hot power. I have a good friend who is an electrical engineer and I asked him about such happenings one day and he didn't have an answer for me. He told me about a closet in his house that regularly blows light bulbs, while other sockets don't. One would think that service voltage is a finite amount, and it is, and I have yet to figure out why one socket location will blow bulbs regularly and others don't.

On another bulb lifespan note, it's my understanding that the amount of hours a bulbs life is rated for, is based on controlled laboratory environments, with the bulb being turned on once and left on, and it's the switching on & off of bulbs is what really shortens their life.

For me, the argument breaks down over two things, both mentioned over a year ago:

1) the fact that unless you're heating year-round, you're adding heat to your home during the time of year you're also trying to cool it

I live in Toronto, Canada. We have to heat around six to eight months of the year for comfort, and by that I mean so we aren't covered in water from the mains when our pipes freeze solid and crack. The rest of the year, we are more concerned with cooling, as we experience the hot, humid summers that a continental climate bring us.

One aspect this argument always glosses over is conversion loss. We heat preferentially with natural gas, because it is more cost-effective than trying to heat with electricity. One of the main downsides to my current apartment is that it has electric baseboard heat. It is too expensive to use except to keep the pipes from freezing, which necessitates extra blankets and sweaters for us, oftentimes. In this specific case, were I willing to swap out my incandescents for LEDs in the summer, it might make more sense than the baseboard heaters, but I doubt it. I like to think about the whole system involved: something is being combusted in most cases to power a heat engine to make enough electricity to make up for conversion and transmission losses, just so it can be made into heat again.

You know what gives better radiant heat than an incandescent bulb? One of those radiant electric heaters with the parabolic dish behind it to direct the heat. And heating with electricity is pretty wasteful. So apart from spectral benefits of incandescence, which the LED can beat in terms of versatility and energy efficiency, and amount of pollution incurred, in which, of course, the LED leads, but probably not for long as demand for cleaner efficient light increases, why is the incandescent better? Because it doubles as an inefficient heater, which is good sometimes half the year round?


2) the fact that the target that is being aimed at isn't stationary

LEDs are a developing technology. They are constantly being improved, and I wouldn't be surprised if demands of the market weren't a driving factor there. I would love to see more R & D being done on incandescents, but I have read only one article on the subject, about MIT developing an incandescent bulb with an insulated filament that was something like 40% efficient, which beat the LED numbers handily at the time, but that was like a decade ago. It would probably last longer, as the insulated filament would be less prone to thermal shock, but that efficiency was found by eliminating waste heat. The transparent insulative sheath around the filament would heat up the air inside the bulb, and some would transfer to the outside bulb, but that would remain cool.

But as there is no money in designing a bulb that lasts forever, the only way to get one is by accident, when they're actually trying for something else, like increased efficiency due to decreased waste heat generation. And as the efficiency to compete with LEDs comes from the heat that is the benefit of the incandescent-as-inefficient-heater, well, you'd still have the radiant aspect, as in, skin warmed by bouncing photons, but unless the more involved manufacturing process is cleaner than the one producing LEDs, a lot of the argument has shifted.

So it is almost 2018. I think the superiority of the incandescent overstated except in certain specific scenarios that don't include the reality that most places have a summer period that makes the generation of additional heat by light fixtures an inconvenience. I also think that the fact that LEDs can be tuned to different parts of the light spectrum can't be overstated as a game-changer with regards to the efficient augmentation of light levels in indoor or assisted outdoor situations.

Imagine being able to tune out all the colors of light that reflect off of the leaves, and all the parts of the spectrum that are harmful to plant growth, and then imagine filling in the missing parts with more of the helpful parts of the spectrum. That's what LEDs offer us.

I would like to see more efficiency in incandescents, but I want to see more work done on LEDs. The presupposition that unfiltered sunlight is probably more healthy for plant growth isn't a reason to reject LEDs, given the potential benefits. Incandescents are way better than CFLs, that is certain, but I think that LEDs are almost another animal entirely.

I can agree, though, that incandescent bulbs are much better space heaters than LEDs.

-CK

LEDs cuts CO2 emissions by 1/2 a billion tons in 2017
See:
https://om.blog/2018/01/leds-cuts-co2-emissions-by-1-2-a-billion-tons-in-2017/

A heat-pump and LED lighting is a much more efficient way to produce heat and light than an Incandescent, plus the HP can also produce "Cooleth"

I.e. during the winter a HP and LED lighting can produce 2-3 times as much heat for the same energy input as an incandescent.  And during the summer, when the incandescent is STILL cranking out unwanted heat, the HP can crank out cooleth instead.


Incandescent is no longer "better" at anything.

I think it's more accurate to say that the circumstances wherein incandescent bulbs are better than LEDs are highly specific and rarefied.

I think it was clear-cut when we were talking about CFLs and the really bad, early days blue LEDs, and even most of the cheaply made ones still fitting the edison bulb fixtures.

I don't think the title of the thread is accurate, though, especially with the updated year edit.

Should those new incandescents that MIT developed with the insulated filament that hits a crazy efficiency by NOT generating excess heat come to market, this might change, but they will no longer be any use as space heaters.

-CK

I contemplated this thread awhile back and came to a similar conclusion as Peter.

Paul advocates RMHs and Incandescents, but RMH are superior to most things when it comes to the efficiency of giving off heat. LEDs providing efficient lighting and little heat, while RMH giving off lots of heat seem to be a better pair than dealing with incandescents.

As I muddle along through life, finding alternative ways to heat a house or building is much easier than finding alternative ways to provide light to one - which is the purpose of a lightbulb. And electricity has gone up more than 10% in the last 2 years alone. I'd rather use the extra electric saved from LEDs to power a gadget that provides convenience. (Ex. electric shaver)

With all that said, the world still needs to embrace natural sunlight more into the design of things - the original incandescent in the sky :)

---

The only things I don't like about LEDs in consideration to permaculture are their complexity, and I'm not sure if they'll get any simpler to manufacture/assemble in the future.

Jarret Hynd wrote:
The only things I don't like about LEDs in consideration to permaculture are their complexity, and I'm not sure if they'll get any simpler to manufacture/assemble in the future.

In my opinion, the biggest problem with LEDs currently is that they keep trying to come up with ways to plug this into an outlet designed for this

Much of the waste in LEDs is caused by including the LED drivers in every bulb.  It would be more cost effective and efficient to use a single driver powering a string of LEDs, plus by separating the driver from the bulb (and the LEDs) it means you could replace just the part that fails and not the whole assembly.

As we go on I suspect we will start to see internal achitecture/design change to take advantage of LEDs used this way.  Perhaps ceiling panels with LEDs built in so the whole ceiling lights up.  


One idea that I've been toying with is installing something like picture rails with hidden LEDs hidden setup to reflect off the ceiling.  I would use low voltage wiring to the LEDs with the drivers hidden behind the light switches.

With an off-grid setup you could even run DC from the battery pack to the switches and avoid the need for an inverter.   And you wouldn't need to use heavier gauge wiring like you would if you used incandescent lights.  Even at 12V DC, 14 ga wire can deliver more than enough power to light up a large room with LEDs.

Context is important, too. As you point out, Peter, you will get much further on a 12v system using LEDs than incandescent.

I think the observation about many LED issues stemming from incompatibility with the Edison bulb model is bang-on.

I want to see what happens when that particular design hurdle is jumped. And I would love to see pictures of your described LED picture rail system.

-CK

I've always liked that Lazarus Long quote.   I've actually done all of those things except setting a bone (although I know the basics), writing a sonnet, and dying gallantly.  I could probably write a sonnet (badly), but I don't plan on dying gallantly.

I want to die when I'm 130, in bed with a beautiful woman, shot by her jealous husband.

It seems like emissions of mercury might be lower now as a result of using less coal and less fluorescents.  This seems to coincide with the electricity mix being more renewable and the lighting mix being more LEDs.  

I think Paul makes a good argument for the effective use of "waste heat".  It is an excellent way to improve energy efficiency, whether that is motivated by improvements to the environment, economy, innovation, or all of the above.  I live in a cold northern state too, so I might understand what Paul means by that.  In the winter I use the oven more.  I stop-up the shower drain so that the heat goes into the house first before I drain the water.  And I even use this to rationalize my taking longer showers, haha.  Sometimes I even use that gray water to flush a toilet (although I think it needs an automated system to be practical).  

Does hot water use more energy than lighting?  It can.  However the bad news is that in the summer the tables are turned.  All of a sudden I'll need another system to make solar hot water.  I need to cook less or cook outside more.  And unfortunately that argument for justifying incandescents is diminished.  Maybe with longer days, less artificial lighting is needed.  That supports Paul's argument.  

I think the e-waste in the US is handled a little better than in other nations, not that it's totally benign.  An average of around 100 pounds of e-waste per American per year is quite a lot.  As a society we're getting more electronic, but in many ways we're using less toxic materials.  TVs and computer monitors don't contain 5 pounds of lead each anymore.  Nickel-cadmium batteries are much less common than they were.  It's probably only a matter of time before lead-acid batteries are replaced with lead-free devices.  Lead-free plumbing and "RoHS" electronics are very well established at this point.  Lead-solder is mostly a specialty product now, used in applications like medical, military and by hobbyists, that sort of thing, not in mainstream electronics products which I figure represents the majority of e-waste.  The amount of lead used must have gone down sharply.  We're producing more e-waste but it's a lot less toxic that it used to be.  I could be wrong but I'd guess the rate we're producing pollution isn't rising nearly as fast as the rate we're creating new technology, at least where e-waste is concerned.  

Using the property of incandescence for lighting is a nice idea, and in some applications it is sufficient, but in others perhaps it isn't quite sufficient.  There is also the problem that materials are limited in their operating temperature, and this tends to make it a bit difficult to improve their efficiency.  Tungsten is quite an unusual and phenomenal material, especially for a *metal* of all things, to have a higher melting point than most ceramics at 3400 °C!  It's only a matter of time until someone insists on using it in a rocket mass heater core, lol.  The refractory ceramics that we usually use for that include:

alumina  2100 °C
zirconia  2700 °C
silicon carbide 2800 °C

To increase a lamp filament temperature we'd have to reach for materials like:

tantalum carbide 3800 °C
hafnium carbide 3900 °C
tantalum hafnium carbide  4000-4200 °C  (costs < $10/g)

Tantalum is considered a "conflict mineral" in Africa, although political issues aside, the occurrence of tantalum and hafnium on earth is no less than that of tungsten.  (Although the battery material lithium is 100 times more abundant than tungsten and evidently there can still be difficulty sourcing it, I guess because it is used in greater quantity.)

One interesting possibility is to use a design that doesn't need these materials, for example a plasma lamp. This is basically what arc lamps are anyway, whether it's a sodium lamp or a mercury lamp.  Don't forget we still have the very efficient, long-lasting neon lamp, argon lamp, and xenon lamp and similar lamps that don't need mercury.  A xenon lamp produces an exceptionally high-quality white light.  Although these gases are relatively rare, they are well-distributed in the atmosphere, hopefully making their price stable in the low dollars per liter range.  Small quantities cost somewhere around: $20/L for xenon, $2/L for krypton, and $0.50/L for neon, and $0.01/L for argon.

By the way, as Paul said, since heating costs can be more significant than lighting costs, using some of these gases to fill windows on a house or greenhouse can make them just as insulating as a wall panel, since krypton has only one-third the conductivity of air, and xenon has only one-fifth.  Krypton might be close to affordable for this application, where even at the low-quantity price, I'm figuring that the cost to fill a 1/8" gap is roughly $0.50 of Krypton gas per sq.ft. of window area.  Considering the lengths we go to to heat buildings and greenhouses, this might be a relatively straight-forward solution to have in the tool kit.  I also think vacuum panels are quite interesting, although apparently it requires a fairly high vacuum, so that seems a bit more challenging to do.  

By the way, to respond to a comment from last year, if you take an air conditioner's rate of cooling, for example 5000 BTU/hr , and divide this by the SEER rating, say 10, then you get the average power draw in watts, in this case 500 watts.  This formula is pretty simple to remember and allows you to know what you're signing up for when dealing with these appliances.  

Mike Phillipps wrote:...
I think Paul makes a good argument for the effective use of "waste heat". ... And unfortunately that argument for justifying incandescents is diminished.

I'm not entirely sure that argument has much value, depending on how you heat your house.
I primarily use solar space heating with a heat pump backup.
Both of these provide far more heat per watt of electricity than a lightbulb does.  So (for me) it's far more efficient to use LEDs for lighting and use sunshine, or the heat pump, for heat.  
Bonus, during the summer time, using LEDs for lighting means I don't need as much energy to cool the house either.

If your primary source of heat was wood, then the amount of wood you save by the additional heat put off by incandescent light-bulbs is insignificant.  Running a 60 watt bulb for 8 hours produces about as much heat as burning 4 ounces of wood.

There have been numerous studies that show that when you consider the complete life cycle cost to the environment, LEDs come out FAR ahead of incandescent, and that even takes into consideration all pf the extra waste materials in current LED "lights bulbs" We are starting to see a paradigm shift in how LED lighting is used that is going away from the screw in bulbs that has far less waste material.  
In a lot of current applications they are over-driving the LEDs because you can get much more light for the same cost of the LED, however this is less efficient and causes the LED to burn out sooner.  If the LED is driven at it's designed power level, most will outlive everyone on this board, and use 1/2 as much power for the same light output as many current LED bulbs.
Cree currently makes LEDs that put out over 200 lumens per watt...as long as you don't push more than 1 watt through them.  
If you run the same LED on 3 watts, it will produce more light than a 40 watt incandescent, but will only last a few years.  Or you could use two of the LEDs at 1 watt each, and get about the same amount of light as a 400 watt bulb, only use 2/3 the energy and they will last 40-50 years.

As I understand it the cost of insulated windows is primarily driven by the cost of building hermetically sealed windows rather than the cost of the gas used to fill them.  Well, that and marketing.

Peter VanDerWal wrote:Both of these provide far more heat per watt of electricity than a lightbulb does.  

If you choose to heat your home with light bulbs in a completely different room than you are in, then you are correct.

If, on the other hand, you choose to to place light bulbs strategically where there are people, and the lights are close to people, then people will begin to feel too warm and they will turn the thermostat down.  

If you add in collections of micro-heaters in strategic locations, then people can feel very comfortable and the thermostat might be set to 60, or maybe lower.  This is dramatic savings in energy.  

And this is a strategy that an LED cannot do.

the heat pump I installed this past summer keeps 1/2 my house around 65 and the rest (bedrooms) around 62-63.  It draws as little as 250 watts (4 light bulbs) and even then doesn't run continuously until the outside into drops into the 30s.  Most nights the heat pump doesn't even kick in until after we go to bed.
A geothermal heat-pump would be even better.

And if you heat with wood, well as I said, the heat output of a incandescent running for 8 hours = 4 ounces of wood, or about 1/2 ounce per hour.  I don't know anyone that is that meticulous about how much wood they burn.

And there is always the issue of having to get rid of the extra heat during the rest of the year.  Now you're wasting energy twice, once creating heat you don't want, and then again to get rid of the eccess heat.

I live in an off-grid house, with electric power supplied by photovoltaic panels and batteries.  With lighting provided by 5-10 Watt LED lightbulbs, I have just enough power to ensure a fair level of comfort, day or night.

Using less energy for lighting = having enough energy for other household appliances

That would be unimaginable with incandescent lightbulbs.

They are not an option for me.

End of story.

I don't see anything wrong with LED lighting when you are off-grid and producing just enough power to run them, without requiring the commercial bulbs with all the fancy electronics inside.  The waste is the overbuilt electrical system in on-grid houses, in that case, not LEDs.  It wouldn't want to use more power to run an incandescent bulb that also throws heat when it's 30C in the summer.

As the last two posters have noted, incandescent bulbs are not conceivable in an off grid scenario. LEDS are far superior. It's 2018....ditch the incandescent and ditch the grid.

A couple quick comments about questions raised in the initial post ... from an electronic techno-geek with many years experience working in the semiconductor wafer fabs. The Intels, the Motorolas, the AMD's, & the IBM's of the world. LEDS are extremely toxic. I don't have scientific or medical data in front of me to back that up but they're made of gallium, arsenic, boron, & phosphor, etc. You know, the same things bombs & rat poison are made from. Anything made using semiconductor techniques is highly toxic just by the very nature of what makes it a semiconductor in the first place. Chlorine is another common semiconductor ingredient. Plus a long list of strong acids. Semiconductors are reasonably safe after completion & encapsulation though. Until they burn or get tossed into a landfill. I'd venture to guess that if just one LED were crushed & eaten it would be fatal. You don't want to live downwind of a production facility no matter how well they attempt to scrub the exhaust of impurities. Look up the Bhopal, India Boron Trichloride plant explosion. Not fun.

That being said ... LED's are awesome in so many ways. Very very very low power consumption. Very very very long life. That's why I use them in my small scale off grid world. Good window placement, a couple of solar panels, a charge controller, a battery or two, a bit of wire, & a few well placed LED's do the job reliably. Not exactly the traditional approach but most anyone can figure out the technical details of custom LED creations with just a little research. The manufactured AC operated LED units work just as well. For a lot more initial expense if you already have a basic solar setup. Buying individual LED's in bulk is ridiculously cheap compared to store bought arrays. LED's work only from DC sources. Therefore, any "AC" operated LED lighting is actually converting the AC into DC first. Significant losses involved with that. Not a big problem if you're paying the power company (except the paying the power company part) but very significant to anyone generating their own electricity. Tesla lost. Edison won. E=IR. Deal with it.

There was a question concerning whether or not DC wiring is simpler & therefore a good choice for LED installations. The simple partial answer is yes it's easier. That's why I use it for my situation. Plus I just want a dual system. AC from the power company hydroelectric dam & my tiny completely independent DC system. There are four power grids in the USA. East of the Rockies, west of the Rockies, Texas, & my handful of panels. There are serious potential drawbacks to DC though. Good understanding & planning is crucial. 120VDC will zap you just as hard as 120VAC. Most home solar systems are usually much less voltage though. So there is a certain amount safety in that. The main drawback to DC is I2R losses. Pronounced eye squared are. Current squared x resistance. The short & sweet solution without a long explanation involving too much math ... really the only solution unless you can do some supercooling of your wires ... is USE THICK WIRES & keep them as SHORT AS POSSIBLE. Easy enough to find wire gauge vs ampacity data online. Use double the diameter if you can. Make the connections as PERFECT as possible. Long wire runs usually are better suited for using an inverter then distributing as AC. Unless you enjoy having extra panels to make up for the losses involved with long DC wiring.

Eating LED's? Was originally looking for pastured rabbit info. Apparently got side tracked.

More info about how & when DC wiring is better. For a typical AC setup it is assumed that the user would need to be able to plug any normal electrical device into any outlet. Building codes require specific wire sizes & receptacle sizes for that reason. Costs are more or less determined by that. DC can be much more flexible & cheaper.

Simplest method doesn't even need wire except enough to connect a few broken pieces of solar cells together with a enough extra to connect to a AA or AAA battery. <<always use a diode to protect your solar cells because humans are LOUSY charge controllers>> Charge a few AAA batteries to operate a flashlight. Or one of those stick on under the pantry shelf lights, etc. A bit simplistic but it's effective light. Wireless wiring.

LED's draw very little current so wire size can be smaller that standard AC wiring. Remember the I2R thing. Another way of saying heat. Less current = less heat ... squared. In terms of simple wiring less heat is a good thing. Not so in Siberan winters. Krupnik was invented to cure that kind of cold. This comment concerns light & small motors not warmth. With a little creativity small batteries & small wires can be used in very unique ways. A word of caution here ... fuses .... always include a fuse appropriate for the wire size. Electronic gizmos have a negative temperature coefficient. Things (heat) gets ugly fast. It bears repeating that nasty smell it emits is TOXIC. FUSE FUSE FUSE. A relatively thin wire will operate many LED's.

Let's say you had a remote display case that you wanted to add a fairly large light to. Or maybe an aquarium pump. Install a motorcycle battery. Custom make LED light arrays. Use a DC pump. Once a week rotate a fresh battery in or just charge it. Strategically placed LED's help prevent running into things at night but not enough light to see details. Think airplane floor lighting strips. Very effective & very low power. Tiny batteries & tiny wires for tiny places & tiny loads.

If one was so inclined ... strictly theoretical of course ... they could modify a flaming Barbie launcher into a mobile flaming Barbie launcher. Just for grins have it pull a Barbie sized trailer of her tiny solar panels so it could supply DC for the drive & steering motors. Would function every sunny day without fail. Launching Barbies moments after a short trip near the giant fresnel lens. It could happen.

For larger appliances & motors a common wire feed from a battery bank is generally easiest & cheapest. Or even converting to AC. The typical loss spec for that conversion was about 13% a few years ago. Might be slightly less now. Improper DC wiring will rapidly lose much more than an inverter uses.

Wires too long &/or too skinny for the load is the enemy. Use fuses or circuit breakers. Protect your panels with a diode or an actual charge controller that has one designed in. Within those limits are endless possibilities.

Mike Barkley wrote: Simplest method doesn't even need wire except enough to connect a few broken pieces of solar cells together with a enough extra to connect to a AA or AAA battery. <<always use a diode to protect your solar cells because humans are LOUSY charge controllers>> Charge a few AAA batteries to operate a flashlight. Or one of those stick on under the pantry shelf lights, etc. A bit simplistic but it's effective light. Wireless wiring.

Please make sure the battery you want to use is tolerant to over-discharge and overcharge before doing this. (The diode protects the battery from discharging via the solar cells when there is no light.)

Just because LEDs use less power doesn't mean you can use smaller wires.

You don't select wire size based on the power the wiring has to handle, but on the current (Amps) it needs to handle.  
Since Power = Volts * Amps.  For any given power level if you reduce the Volts you increase the amps.

If you change from 120V AC circuit to a 12V DC you end up having 10x as much current for the same power.  If you use the circuit to power an LED that uses 1/5 as much power, as an incandescent, it actually has to handle 2x as much current (amps)

Next, the idea here is to make things run more efficiently.  Using the bare minimum size wiring for long runs is not efficient (electrically speaking) since it will have more losses due to resistance, this is actually worse when using low voltage because if the wiring drops 1V at 120V that is less than 1% loss, but if it drops 1V at 12V that is over 8% loss, you typically want to keep wiring losses below 5% .  
Using a larger wire will reduce the resistance, which will reduce the voltage drop and increase efficiency.

Finally, as has been pointed out above, if you are going to wire a house to meet electrical code, you have to use wire that meets electrical code.  The minimum size wire that meets code(in most places) is 14 AWG, so there is no point in trying to use anything smaller.  Besides, since 14awg is so common it is probably CHEAPER to use than smaller wire that doesn't meet code.

Just a quick calculation.  Lets assume you are running 3 LED lights that combined draw 24 watts and that you are powering them from a 12V battery bank, and that the wires from the bank to the lights are 50 feet long.  If you use 14AWG wiring, you will have a voltage drop of approx 0.5V and lose a bit over 4% of your energy heating the wire.  So realistically you wouldn't want to use less than 14 AWG anyway.