It doesn't help that the old Edison screw is terrible for cooling.
E.g. I get relatively good bulbs by requiring CRI > 90. Or just requiring no flicker, as a feature. When the producer has more than one parameter to optimize, they seem to come up with more well-rounded designs. To say nothing of the useful features I mentioned.
Yes, I likely pay 2-4 times more for these bulbs. But they perform, and they stay for years of intense use. The cheapest overall bulbs are designed for other purposes.
But it's still a perfectly good counter to "no sane option".
Unless there's some regulatorially-mandated set of minimum standards, given the thin information channel of price ("what will I have to pay out of pocket right now for Product X, with no other indicator of quality, longevity, or potential risks"), the market will pursue low costs with an iron determination, over all other factors.
Consider that a corrolary of Gresham's Law, which in its essence says that given a nominal currency value, currency of lower determinable quality (lower specie metal content) will tend to drive higher-quality currency from the market.
"But wait, Dred," you say, "You just said that there was no other indicator of value. Why is low specie currency driving out high specie?"
Because that's not the relevant value, and it actually is apparent (or at least sufficiently so to a certain number of traders in coin). It's the nominal value that's the determinant for circulation. In the case of coin, it's face value which determines market exchange. Put another way: in a world of copper-painted zinc lozenges circulating as pennies, a diamond-encrusted gold penny will still only net you $0.01 in exchange value as currency. You're better off holding on to it or trading it for its specie / gem value.
With bulbs or electronics, the logic shifts around a bit, but the result is similar. Price becomes exchange value, product quality factors (colour registration, colour range, flicker rate, dimmer compatibility and performance, heat dissipation, any electronics buzz or hum, diffusion pattern, and most of all, that one factor you cannot determine in 3-5 minutes at a store: product life (or more properly: product-quality-life, the period of time it functions adequately before driving you bananas, even if it doesn't fully fail).
You're standing in a store aisle, or staring at an utterly uninformative webpage of quite probably counterfeit and fake-reviewed Amazon LED bulbs as well as whatever LED-bulb-adjacent search results Amazon's Alexa Search Engine has decided it needs to distract you with today, and you see two things:
- LED bulb.
- Price.
And "you" is the typical shopper. Which is to say, one who's not read this particular screed.
What are you going to do?
You're going to pick the lowest-cost bulb you see and offer a prayer to His Noodliness.
This is, incidentally, a chief reason why pure market solutions tend toward abysmal product offerings in the absence of some external regulatory, ratings, or similar systems.
There's an alternative, and some retailers follow this. Marks & Spencer in the UK long made a tradition of carrying only one exemplar of a brand, generally represented as "the best there is". That is, M&S undertake the research and quality determination costs for its customers. Your research problem is reduced to "shop M&S".
In the US there are retailers who follow a similar strategy, though emphasizing the price-quality relationship. Costco is among these, where typically only a small range of products within a category is offered (Trader Joe's operates similarly, though only in food). Friends purchased a convection toaster-oven from Costco recently -- not the cheapest model out there, but one which performs quite admirably, will handle a frozen pizza and most meals for 2-3 quite well, and (whilst slightly tricked out on controls) is fairly simple and straightforward to operate. On balance, "shop Costco" has been justified in this case.
The pure-pricing model typically works best (if at all) for nondifferentiable commodities of minimum complexity. Foodstuffs, bulk construction materials such as sand and gravel, coal, oil, etc. And yet even within each of these in developed economies you'll find grading and quality control, some imposed by government mandates, some self-adopted by industry.
And outside of bulk commodities -- in services, in wages, in rents, in interest and risk markets, in public goods -- pure market-based pricing absent various corrections, breaks down entirely or leads inevitably to pathological conditions. See "Iron Law of Wages" and "Law of Rent" -- the twin prods of numerous revolutions.
The problem with your "sane option" is that at mass market scale, say, the 100 million or so households in the US, or the 300 millions or so in China, it simply does not scale. Information and expertise implies some level of clustering and trust relations. Commodity pricing doesn't provide for this.
Yes, but what is a "pure market"? I like to believe that my local home improvement or grocery store has some kind of concept of their own brand and puts some effort into getting decent products on the shelf, whether store brand or not. So, I hope that their markets are not as "pure" as online sites where anyone can sell.
In the case of a brick-and-mortar store, part of what the retailer is paying is for a physical retail location, which you might think of as access to a specific cohort of local traffic (foot, vehicle), and customers, in the sense of patronage of a retail establishment by custom (https://www.etymonline.com/word/customer).
Since that is a finite set, and there are often (though not always) competing retailers, practices which alienate customers tend to be counterproductive.
I was actually thinking of the local B&M hardware store where the staff know their products, and customers, and I've been making a series of (to date, satisfactory) LED bulb purchases whilst writing my earlier comment. What is ultimately carried is a function of the store's specific traffic, the parent corporation, and other arrangements (there are periodic government and utility incentives for low-energy bulbs), but the local staff and management have some agency.
Existential Comics' "Freedom Monster" thought experiment is worth consideration:
I think it's an oversight to mention bulk commodities as an exception to the need for quality information. Things are bulk commodities because there are standards and regulations that enable them to be traded easily.
"Bulk commodity" in trade means that a good is comprised of units (discrete or continuous) of equally-interchangeable quality, as opposed to a good which is comprised entirely of unique entities.
If I take a brick, or a shovelful of sand, or a barrel of rainwater, a ton of iron, a bushel of wheat, a dollar or yen or pound from you, and return another to you later, it doesn't matter if you get back the same unit or a different one, so long as the quality is equivalent.
If I take a Monet or Picasso or first-edition Gutenberg Bible or the Hope Diamond, you're going to want that specific item back, not a replica, no matter how exact or precise.
If I order a package of washing-up liquid from Amazon and that goes walkabout, Amazon can simply ship another unit of soap and I'll be happy.
If I order a pair of prescription eyeglasses and those go walkabout, Amazon cannot simply redirect another customer's order to me -- the specific nature of the glasses -- prescription, frame choice, fit, lens coatings -- matter.
Bulk commodities are highly undifferentiated. Within a commodity class, they're infinitely substitutable. Once a price is determined, what you get for that price really doesn't much matter. In other words, it's the realm of market-based exchange most amenable to price-as-the-only-signal you can find.
And yet, it still doesn't rely on price-as-the-only-signal.
Even amongst commodities, there are quality measures. You want your sand to be sand, not dirt or plastic. For specific applications, river rather than beach sand. Grain is controlled for moisture content, spoilage, and insect fragments. There are quality standards imposed by mechanisms other than price.
Your (apparent) objection to my comment was actually the point I was trying to make.
But thank you.
I've put in some "filament" style LEDs, these look like they will last longer as there doesn't appear to be a power supply taking up half the bulb (although there are electronics down in the screw part).
Also, I see many people getting extremely bright bulbs and putting a large fully opaque light shade over them. That’s rather counter productive.
There apparently are 200,000 hour L70 fixtures [2], but I haven't found any 250K, 300K, 400K, or 500K hour units.
If we ignored form for pure functionality to get the full-rated life of the bulbs, I wonder what the design/specs would look like. The longest DC power supply I could quickly find a mention of is an Antec line, so I wonder what a really long-lived (50-100 year) DC power supply design would look like.
[1] https://www.accessfixtures.com/led-lifetime-l70/
[2] https://www.google.com/search?client=safari&rls=en&q=200000+...
Philips is selling its small domestic appliances business and hoping to strike a similar deal regarding the licensed use of the name Philips. They are centering solely around the Health business now.
There's a bedside lamp next to the bed and the bulb is at least 30 years old. Two generation of children have used that bedside lamp to read stuff at night or walk to the bathroom.
Meanwhile my LED - whatever the brand - are failing between 6 months and two years.
Nonetheless in the last 20 years all my normal light bulbs had to be replaced on a regular schedule which was often enough that i got slightly annnoyed by it.
My expensive philips hue, no issues so far.
18 years now, maybe longer, but it still turns on and has yet to burn out. Of course, we don't use it much, and it's only something like 30 watts (clear bulb).
Someday, maybe, it'll get replaced.
They were a dollar Canadian after government incentive. You get what you pay for.
Prior technologies have expected failure built in. An incandescent bulb's filament will fail eventually (making the filament more robust reduces light output so now your bulb lasts longer but is less useful or costs a lot more to run) and either type of fluorescent very slowly loses the gas that makes it work.
They're own-brand ones, sold by a major supermarket chain here in the UK (Sainsburys), so I'd expect them to be 'ok' at least. Certainly not 'cheap'.
Compared to ye olde incandescents, the complexity and material make up is significantly less recyclable and looks to my eye to require a lot more invested energy to manufacture.
-
I've wondered this about so-called 'smart meters' too - how much energy does it take to make one, and how long to 'pay off' that investment?
Alternatively, plastic takes about 100 kJ/g to produce. So if the bulb has 1 g of plastic, that's about 0.03 kWh of energy. Nothing compared to the savings in electricity using it.
Not always, they could be subsidized to get people off of incandescents.
I've shaved off around .15 kwh of energy that was used 24/7 ($25/month in California) based upon usage information from our energy company.
Downside is that we're a decade into net-connected thermostat and HVAC systems and none of them support any of this.
“efficiency” is meaningless when heat is desired. if you pump 100W into a device, and it “wastes” all that power, you have successfully generated 100W of heat. only heat pumps can be more efficient.
Exactly. And if you have a whole house/apartment electric heating system it is very likely that it is in fact heat pump based. So you are better off using that for heat than the lights.
If your whole house/apartment heating system is gas or oil based, that will probably be even cheaper than the heat pump electric version.
Hmm, where? I have never knowingly been in a house with a heat pump in the UK, and lots of cheaper flats have electric radiators and immersion heaters.
Anyway, my point is that the inefficiency lost to "waste heat" is not 100% waste.
During the summer, it's also extra inefficient to pump 100W into a device producing heat and then pump even more power somewhere else to cool down the room.
All the heat that comes out of a lightbulb is being generated at 100% efficiency :)
(Although it's likely more expensive to heat your home electrically instead of via natural gas)
For CFL, it's a bit more complicated, but keep in mind that they do have quite a bit of glass (takes energy to make and form), need some mercury, and at least as much electronics (if not more) as LEDs. Furthermore CFLs don't necessary fail completely, but degrade over time. Thus they're a bit less efficient than LEDs to begin with, but a lot less after a few years.
So overall, unless the LED is so shitty as to consistently fail within weeks, there's no doubt their total energy bill is competitive, and it's even more clear if you limit the comparison to products of similar quality.
I think it's more about how well balanced all of the individual LED's in the bulb are. Unbalanced circuits have more current flowing through some paths than others and leads to premature failure of individual elements and it's usually rapidly downhill from there for the rest of the bulb.
Philips bulbs seem to do much better in the ceiling. If only they'd make 3000k bulbs...
Hopefully you bought a problematic model, and they’re not uniformly going downhill.
How old are the ones that are failing for you?
They weren't cheap, but they weren't mega expensive either.
The cheaper the design the higher the failure rate. The heat seems to kill these things over time. For example I just ordered another warranty replacement from one brand called Hyperikon that uses similar types of designs for floods in my Kitchen. These have a much higher failure rate. Others seem to last forever. Between using electrolytic Capacitors and other tings that are prone to heat and small enclosed areas it is no wonder that the newer, cheaper bulbs fail quicker.
Since one big reason for going with LED is to be more energy efficient and greener, it would be interesting to compare the carbon footprint of these cheap LED bulbs that fail quickly with an incandescent and CFL over the entire lifespan. If the bulb lasts longer the numbers are pretty easy, but given short lifespan of some and their use or rare earth elements it would be interesting to see the full analysis of the environmental impact of each one.