/ Home & Energy

Are your LED light bulbs burning out too soon?

LED light bulb

LED light bulbs have a tendency to make grand claims about their lifespan – it’s common to see manufacturers promise bulbs will last 25,000 or even 50,000 hours. But we’ve found many failing well before this.

One of the attractions of LED light bulbs is that they’re supposed to last a long time. And so if you’re shelling out for these bulbs – typically more expensive than other types of light bulb – you’ll want to be sure that they’ll live up to those claims.

But our tests show that not only do many LED light bulbs stop working before the end of their promised lifespan, some don’t even reach the soon-to-be-implemented EU minimum lifespan of 6,000 hours. We discovered bulbs from both Ikea and TCP that failed to reach the 6,000 hour mark for the majority of samples we tested.

Ikea bulb among failures

In the tests – which were carried out by Which? and our European partner organisations – we took five samples each of 46 different bulbs. The bulbs were switched on for two hours and 45 minutes, then switched off for 15 minutes, in a continuous cycle until they burned out.

Five different bulbs stopped working before the 6,000 hour mark for the majority of samples we tested, though the TCP and Ikea bulbs were the only ones which were sold in the UK. Both have since been discontinued.

New EU regulations which will come in from 1 March 2014 say that 90% of any batch of LED light bulbs should last at least 6,000 hours.

Another five bulbs stopped working before the 10,000 hour mark for the majority of samples we tested, despite claiming lifespans of at least 25,000 hours. None of these bulbs were sold in the UK.

In total, 66 of the 230 samples we tested failed before the 10,000 hour mark, though they all claimed they would last at least 15,000 hours.

Has your bulb burned out early?

Ikea said the bulb had passed its own tests and those in a third-party lab. It’s looking into why the bulb failed our test and has removed it from sale in countries where it was still available.

TCP said it was already aware of the problem with this bulb and withdrew it from sale when they discovered the problem. TCP added that it no longer deals with the supplier of that particular bulb and now make their LED bulbs in-house.

We’re in the process of testing the life span of many more LED bulbs, and we’ll update you if we find others that burn out prematurely. But we also want to hear from you – have you bought bulbs that haven’t lasted as long as they should?


One issue that may well be leading to premature LED failure is in the design of the electronics (drivers) used to control the LED current. Apparently poor drivers start with a current spike that may be 3x the design value on switch-on; this can destroy the LED. Good drivers will eliminate this by using a soft start circuit- but it adds to its cost.

I had seen this interesting article. I guess that the spikes produced by fridges, vacuum cleaners, washing machines and choke-ballast fluorescent lighting are more of an issue than hot switching.

I wonder if Which? and other organisations that test LED lams check for how well they cope with spikes.

fuzzy says:
31 January 2014

By not freely releasing the study with the article Which? is promoting misinformation. The headline is showing up in forums and newsletters in the US but no one can look at the test results.

Probably because you need to subscribe to Which? to access the full information? It would be useful if Which? would publish the raw information from tests on line, rather than only combining them together to give simple % ratings. Some of us may have different views on the way the lamps perform in particular ways. With LEDs for example, life might be less relevant (better ones will be available far earlier than a good LED takes to die) than good colour; light output and depreciation, for example.

Bought some early B&Q LED GU10 with 21 LEDs all failed in main bedroom use within 2 years. Bought about 8 GU10 by GE 4 watt single LED almost 2 years ago for same bedroom and OK so far. Bought about 10 Osram GU10 4 watt single LED 18 months ago and failure rate is probably about 50% (used in main bedroom and little used spare bedroom). Fortunately B&Q have changed Osram without question. Reliability of Osram lamps is very suspect.

Alan Ferne says:
31 January 2014

I have seen your report on bulbs and note that some IKEA LEDARE bulbs have been recommended as best buys. I have bought about 7 spotlight type E16 LEDARE bulbs 601.722.01 on different occasions from them. They all have failed to say 100th of original brightnes after say 10-100 hours. They have all been used in IKEA fittings, and supplied from a pertfectly normal mains supply. I have returned a number of them, but have also was in contact with their customer service last year about them, they said they were sorry but could offer no explanation. I have recently contacted them again but had have no acknowledgement or reply as yet. It raises the question about their claim (which seems to be general in the market) of 20,000hours life, ie continuous use of over two years, mine would reckon would not on average be switched on for more than 5 or 6 hours a day, ie should last over 10 years.
The reason for them is spotlights on pictures etc from a high ceiling in my living room (only accessible with a ladder), I previously had used clear incanescent 60 and 100watt spotlights (in some cases with dimmers). These LED spots have the right sort of beam but so far a complete failure and a huge waste of time.This was written summer 2013. I then purchased some 2 spotlights from Beam LED, these have been in regular use since August 2013 and another in occasional use, none of these has failed so far.

I more recently purchased

Bryn Thomas says:
1 February 2014

Bought an Osram which was supposed to last 15000hrs and 100000 clicks it fell completely short and lasted less than 5 months (Bought in September)

Before failure it dimmed and kept flicking between full luminance and dimmed. This is not a dimmed switch.

Not buying again £14 bulb which doesnt last as long as normal bulb os not cost saving.

Peter Scargill says:
2 February 2014

LEDS have been shown time and time again to have long life. The problem is that the figures for longevity come from the manufacturers testing the chips under ideal conditions – on the other hand the Chinese lamp manufacturers are more concerned with headline figures like wattage and lumens – ie performance – and of course, shipping weight.

They overdrive the LEDS then have to come up with innovative ways to cool them down – the very fact they are getting hot implies inefficiency. A better approach would be to under-drive them – hence then approaching the theoretical lifetime – but in a competitive world this would never happen. Using LED STRIP as widely available on EBAY is a great way for enthusiasts as you are in control of what power you feed them – there is no regulation on the strip other than simple resistors. I have a red strip running around our front door and the voltage is strictly regulated yet they are still very bright – 2 years later they are still running at 100% brilliance – and not one failure. Another lights my garden hut and has been running for something like 5 years. I expect more years out of them – on the other hand the 250v LED lamps are totally out of your hands – you have no control over power quality or voltage – and these are the ones that go wrong so very often. Consider innovative uses of LED strip if you want longevity.

I’m sure this is true – but most purchasers of light bulbs think only of the cost and efficiency. They want to buy a bulb, plug it in and forget about it. Incandescent bulbs did this for their limited lifetime and CFLs are now beginning to approach incandescents in their output and rapidity of starting. The target for LED manufacturers is to make a bulb that will plug in and work as well as an incandescent – but last much longer and use much less electricity.

And that target is one which the LED manufacturers have clearly not yet met.

I do not see that the door strip of red LEDs is comparable to any normal household lamp.

The strips have no localised heat source caused by reducing 240v to the1.6v which would be normal for a single red LED. That all occurs at some distant transformer. The LED temperature is controlled by ambient air?

Your comment does confirm, as has been said many times here, the LED itself within its design parameters, is long life item.

From what I read here, the claims of LED lamp life is a pure fantasy from the manufacturers. The only way they can convince a primarily non technical customer to part with what effectively is a very large amount money compared to a simple tungsten lamp.

Richard – I agree that people want to replace incandescent bulbs for reasons of extended life and lower running costs. The manufacturers have struggled to produce LED lamps that are compact and have a high light output. Unfortunately, lack of reliability is an issue and no replacement for the simple 100 watt bulb is readily available at a sensible price.

There is likely to be a demand for plug-in replacements for bulbs for some years to come, but new light fixtures can be designed to use LEDs to best effect, so that LED chips and electronics are not crammed together and likely to overheat. Many lampshades are large and don’t need a small LED lamp, so perhaps there is a market for larger LED lamps allowing more LED chips and better cooling.

Absolutely, wavechange. But people still have old luminaries that were designed both to accept the old-style bulbs and to fit into the design of the room. It is – or should be – easier to design a bulb that fits an existing luminaire, than to redesign a room around a different kind of illumination system altogether.

Richard, exactly what happens. But we can spend so much updating other bits of our houses – kitchens, bathrooms, decorating for example – why not replace your light fittings with ones specifically designed to best use your lamps of choice. That avoids the inefficient compromise of putting lamps into styles they are not suited to.

I did acknowledge the continued demand but wonder when we are going to move on. Plenty of people have installed fixtures that use halogen capsules but the only intelligent use of LEDs that I have seen has been in commercial fixtures. It was the same with CFLs, where the goal was to squash everything into the size of an old fashioned bulb rather than making the CFL a bit larger and a lot more reliable. In my view it was not very clever to even attempt to produce miniature high output LEDs such as the GU10 type.

Maybe I’m being pessimistic. I don’t know.

I believe Malcolm bought some LED strip lights from John Lewis. Here the LEDs are spaced out and the electronic circuitry is separate, so everything should stay cool. Assuming that there is voltage spike suppression included, I would expect these lights to have a long working life.

Nick C says:
2 February 2014

Wavechange – as I mentioned in a previous post here, it does seem that manufacturers are starting to take advantage of the different design freedoms presented by LEDs. An example of a new type fitting is one from Screwfix (who are doing quite a good job with LED) which you can find by searching ref 59016.

There are also many others hitting the market now which integrate the LED assembly into the fitting, and presumably design the whole ensemble to be reliable in a way that is much harder with 50mm GU10 or MR16 lamps. Certainly the Screwfix product is very thoroughly designed with an excellent heat-sink, a COB LED, and a separate LED driver.

Although more expensive than a GU10/MR16 lamp at £18, you have to bear in mind that you are getting the whole thing, fitting and lamp together. When you look at it that way the price is very fair for what you get.

Thanks for reminding me, Nick. I did look at the Screwfix website. What you say about new designs is very encouraging. Perhaps LEDs will have a bright and reliable future after all. 🙂

Regarding energy savings, both CFLs and LEDs (with integral electronics) seem to have poor power factors – typically 0.4. This means that although you are charged based on the wattage, the actual current these devices draw is around 2.5x what it could be. Current is what the generators are producing, so that’s bad for the energy suppliers. If you buy these lamps with separate controllers they would normally correct the power factor to nearer unity, with a significant reduction in current. Presumably one day this could be a real issue.

Smart meters offer the potential to take power factor into account and I suspect this is a reason why energy suppliers are keen that we should have these meters in homes. It is another reason for consumers to object to having them.

Old choke-ballast fluorescent fixtures frequently contained a power factor correction capacitor, but it is not practical to include this in small lamps.

My impression is that homes tend to be better lit, possibly because with low energy lighting we can benefit from more light for less money – like having your cake and eating it. Add in low power factor and moving to low energy lighting is not as environmentally beneficial as we might have hoped.

Malcolm – It is possible to improve power factor for low energy lamps. Here is an article that refers to LED lighting: http://www.ecnmag.com/articles/2012/11/optimized-power-factor-corrected-driver-led-bulbs

The photo shows a fairly large capacitor and inductor, so this approach would suit a lamp where the LEDs are separate fro the control circuitry. The designer has though to include a spike suppressor too.

wavechange, lighting in the commercial and public sector is routinely power factor corrected, usually to 0.9 or better, driven either as contractual requirements or those of the energy supplier. It is standard practice and easily achieved. However, it takes up space and adds a bit of initial cost.

Unfortunately many suppliers of domestic lamps only think of first cost (also hampered by the obsession with cramming LEDs plus driver circuits into unsuitable shapes and spaces and CFLs only with cheap integral electronics). Presumably as CFL and LED domestic lighting grows further things will have to change.

Our future problem is one of generator capacity, and this is solely driven by maximum demand; reducing current by improving power factor could make a contribution to alleviating this problem. Assuming domestic lighting consumes 4% of total electrical generation, improving the power factor could reduce the required generation capacity to 2%. Every little helps.

Thanks Malcolm. It is reassuring to know that this power factor correction is done routinely. It is going to be more difficult to tackle domestic lighting, but something well worth doing.

At least there is no power factor problem with use of electricity for heating.

Seeing that domestic lighting might only account for 4% of electrical energy demand, and that work in progress might be able to reduce that by a significant percentage, I think it’s high time we moved on to tackle the bigger issues like waste, thermal inefficiency, poor insulation, and inadequate control mechanisms across the entire energy sector and gave commercial organisations a bit of the bashing and heavy lecturing [including from Which? on occasions] that the poor domestic consumer has had to put up with over the last few years as their favourite light bulbs have been banned and the recommended replacements have turned out to be costly duds that interfere with their radio reception.

Apologies for being one of those doing the lecturing, John. 🙂

As an early adopter of CFLs, switching from old fashioned light bulbs seemed such an easy way of saving power and decreasing the impact on the environment. Despite the resistance to change, I suspect that many would not be keen to go back to climbing on stepladders every thousand hours or so to replace lightbulbs, and the nuisance/cost of having lamp holders replaced when they succumb to the heat generated by these bulbs.

I am surprised how many people have chosen LED lamps at such an early stage in their development.

You are right, of course, about the bigger issues.

I also was an early adopter of CFL’s and entirely agree with you on changing from incandescents to CFL’s. This has made practical and economic sense, which should have been enough to persuade people to adopt them without the compulsion brought about through prohibition. But some people don’t like CFL’s and they now have no choice. My view on LED’s is that the big gains in terms of energy saving have come from switching to CFL’s and that going down the LED route at this stage in their development is premature and in many respects unnecessary. Apart from anything else there are serious environmental implications in terms of manufacture and disposal to take into account. I have ranted elsewhere about excessive energy consumption in the commercial sector so I shall switch off now.

I agree that compulsion was the wrong strategy, John. It would have been far better to advertise the savings that can be made by switching to CFLs. It would have also been good to encourage manufacturers to design fixtures containing the control electronics, so that all that needs to be replaced is the lamp. That has the additional benefit that the electronic circuitry is protected from the heat generated by the lamp.

Nick has pointed out that new lamp holders and fixtures are becoming available to make best use of LED lighting, which is very encouraging.

I mentioned the phasing out of halogen lighting, which starts this year. Again I think the carrot would be a better incentive than the stick.

Ian L says:
2 February 2014

Please include halogen bulbs in any future study. Their claim of 2 years is laughable – if we get a quarter of that we rejoice. We have lost enough bulbs now to know that Diall (B&Q) are really bad – and those sold by J Lewis and Waitrose aren’t brilliant. If you know of any reliable sources we would like to hear of them

Ian L – There is not much point in looking at halogen bulbs when they are to be phased out. Old fashioned light bulbs use a lot more energy and halogen bulbs are not a great deal better. Richard has given us another reason to get rid of them.

I think the point is that a lot of people have installed fittings that are suited to halogen bulbs [or bought houses with such fittings already in place] and cannot now afford to substitute LED lamps and will not even contemplate it unless longevity improves dramatically. Those people need to know which of the many halogens on the market offer best value for money overall and will be suitable replacements for any of their existing lamps that fail. I suspect that mixing halogens and LED’s in a multi-light arrangement is neither aesthetically pleasing nor a good idea technically.

Incidentally, I concurred in Ian L’s opinion on B&Q’s Diall lamps as a generality but I have to say that a recent purchase of a 35W Spiral CFL from B&Q has outshone (!) my expectations. I have found it impossible to say that any one manufacturer or brand is consistently good or bad across the board, which makes buying such expensive items quite a lottery. I hold Philips in very high esteem for the combination of quick-start, brightness, endurance, availability and economy, and I choose them whenever possible depending on the application, but we have a couple of their small spiral CFL’s in table lamps that take an age to come on after moving the switch, which is a bit annoying [although, when they do spring into life, they give a high light output immediately].

Current halogen lamps are Class C. These will be phased out at the end of 2016 (not sure whether this excludes directional lamps – spots for example). aAfter that lamps must meet Class B – essentially more efficient. Currently infrared coated lamps look like being suitable, but only at low voltage. Whether mains voltage class B halogens will be available in 2 years I don’t know. Does anyone? However the technology necessary is available.

I have been wondering how manufacturers will raise the efficiency of halogen lamps to comply with EU requirements, and the obvious solution is to run the lamps at a higher temperature. This will decrease the operating life, of course, but that could be offset to some extent by improved quality control. The coefficient of variation is high at present.

Unless the EU specifies the mean operating life of halogen lamps we could see more efficient lamps on the market with an expected life of 1000 hours, similar to old fashioned bulbs.

A technology used is to coat the bulb inside with an infrared reflective layer. This puts heat radiation, that would otherwise be lost through the bulb, back onto the filament to maintain its temperature and thus reduce the electrical power needed. The function of the halogen gas filling is to redeposit evaporated tungsten back onto the filament, which is what gives the lamp a longer life at a higher filament temperature; this gives a significantly better light output efficiency than an ordinary bulb. This combination can give Class B (EU energy class) incandescent lamps. Lives of 3000 hours (up to 3 years domestic use) are being bandied about. This works best on the shorter thicker filaments, as used in low voltage lamps; whether it will be as successful in mains voltage lamps, which have much longer thinner filaments, I have not discovered. Does someone know?

One way or another I expect that halogen lamps that meet the efficiency requirements set out by the EU will be available as direct replacements for existing lamps. Hopefully this will buy time to allow further development of reliable LED lighting.

When quartz halogen lamps were introduced, I remember reading that the filament temperature must be kept high to or the process of redepositing tungsten on the filament would not work, leading to premature lamp failure. I understand the theory but have also seen many halogen lamps used at low brightness for an extended period, with no sign of blackening or failure. I presume that the problem has been overcome by using very small quartz capsules which become extremely hot.

Filaments lose tungsten by evaporation which not only thins the filament (eventual breakage) but the tungsten was deposited on the cooler bulb and blackened it – reducing light output. The halogen cycle combines the halogen with tungsten before it is deposited on the bulb wall but releases the tungsten at the high temperature filament where it is deposited (but not in the place it came from, so gradually the filament will become locally thin and fail). So higher temperature filament plus no bulb blackening gives more efficient light output, and the cycle prolongs life. Precise manufacture is essential – so the most reliable lamps will come from the reputable manufacturers.

I understand the halogen cycle, Malcolm. What I don’t understand is the claim that a high filament temperature is essential when halogen lamps show no sign of blackening when operated at low brightness. I assume that it is simply because there is little evaporation of tungsten when dimmed.

Manufacturers could publish the coefficient of variation with their lifetime claims and that would show that some products are better than others.

When halogen lamps are dimmed there is less evaporation but the halogen cycle either ceases, or does not work as well, so a little tungsten may be deposited on the bulb wall. This will be removed when the lamp is run at or nearer its maximum output.

I assume that this is the case but I have never seen any examples of blackened halogen bulbs, so conclude that we don’t need to worry about operating temperature.

I tried halogen bulbs briefly. A complete waste of time and money. Well over twice the price of incandescents and with an advertised service life of only 2000 hours (which they never achieved) they were hopeless in theory and even more hopeless in practice.

Scott X says:
3 February 2014

Obviously most of suppliers exaggerated their products’ lifespan. When you proved them failed, probably all models have been phasing out the market. The verification time is too long – far more than 250 days for 6,000 hr mark. When did Which start the test and when will they complete the test of 15,000 hrs or even 25,000 hrs? Which also did not report if the suppliers were able to show their test verification results. I am wondering if the suppliers had verified the product lifespan before shipped out the products. In order to help this situation, the standard agencies must introduce a quicker way of spec verification.

I have assumed that longevity is not tested in real time but deduced by extrapolation.

I cannot see how the manufacturers can make a valid extrapolation, considering the very long estimates for operating life. I would like to see these estimates to be more conservative until there is real evidence, preferably based on practical use in the domestic environment rather than ideal conditions in test labs.

We are accustomed to goods being not quite as good as they are claimed to be, but this is more serious. I had thought that those posting about their problems might be unrepresentative, but friends have had similar problems with LED lighting.

This is one of the problems with long-life lamps, particularly LEDs. and much depends upon how they are used – temperature predominately. You might find the attached document interesting – it looks at how Philips evaluate LEDs.
There are two life aspects to LEDS: one is when they fail to light (often given as hours when 10% and 50% are likely to have failed) which is extrapolated; second is when the light output has dropped below a certain value – often 70%. These parameters are of value to professional users for maintenance and function. Domestic consumers may be dealing with manufacturers who lack the necessary quality. The most likely causes of early failure with many domestic LEDs are, I suspect, death of the electronics and failure of the LED chip contacts, due to both poor quality and overheating.

wavechange, I have posted a link that deals with this extrapolation, using an established method – it’s waiting for approval.

Malcolm – With LED lamps, Philips would be my first choice. I’m familiar with the document you have linked to but that was produced nearly seven years ago. What concerns me is the statement: “LED light sources, on the other hand, do not tend to fail catastrophically. Instead, the light output
degrades gradually over time.” Hopefully this applies to lamps produced by Philips, but it is certainly not representative of LED lamps in general, where sudden premature failure is not uncommon.

What I would like to see is a systematic analysis of failure of LED lighting taking into account a range of factors including catastrophic failure and significant decline in light output, type (e.g. GU10), lumen rating and brand. In the case of catastrophic failure, it would be interesting to learn whether or not voltage spikes are a significant factor. I was impressed by the technical input we had when we were discussing radio interference by LED lamps.

Nick C says:
3 February 2014

I don’t think it helps at all, but I am certain that all the failures that people have experienced are with the driver circuitry rather than any of the LEDs themselves – this is definitely the case for the (relatively few) early life failures that I have seen. I am also fairly certain that the statistics that are given for lifetimes are based on predicted life for the LEDs and do not take account of the possibility of the driver circuitry failing.

The characteristic of light output degrading over time will apply when the whole lamp keeps working for the many hours of use it will take to get to this end-of-life characteristic of LEDs.

Scott X says:
4 February 2014

It may give a good excuse to the suppliers that it is impractical to verify it before shipment.

Scott X says:
4 February 2014

Malcolm, where is the link?

Scott, if it was the one you wanted it was as a reply to your request, which is on P1 of these comments. Here it is again.
Scott, if it is the EC report you mean then this is the link. http://ec.europa.eu/enterprise/sectors/electrical/files/emc/ms-campaign-fourth_en.pdf

Scott X says:
4 February 2014

Hi Malcolm, I referred to your below message about extrapolation, not the 2011 report.

“wavechange, I have posted a link that deals with this extrapolation, using an established method – it’s waiting for approval.”

Scott – sorry. That one is 8 posts up.

Scott X says:
5 February 2014

Thanks, Malcolm! 6k hrs are tested in light source and projects the rest of lifespan according to IES standards.

Scott, there are two ways of defining “life” that are in use for lamps whose light output degrades over time (discharge lamps like CFLs for example, LEDs). One is mortality – when a certain % of lamps will have died – because in “professional” lighting this may determine how and when you decide to replace lamps (so too many are not out at a time). The other is when their light output will drop below that required to produce the design lighting level for the space – often 70% of the initial light output. The Philips LED.paper above deals with the first – mortality prediction. The IES method (Illuminating Engineering Society of America) I believe you refer to deals with the second -fall off in light output. In domestic lighting we generally replaced failed lamps. However some lamps can give a much reduced light output well before failure and could well be replaced then.

The Philips paper mentions failure, but focuses on the decline in light output as a function of current and junction temperature, factors which designers need to take into consideration.

Failure rather than declining brightness seems to be the main concern of those who have installed LEDs.

Lamp manufacturers can predict or guess the lifetime on their products used under ideal conditions in test labs but they need to give us guidance if domestic use could significantly affect longevity. Use in an enclosed or semi-enclosed fixture adversely affects the life of CFLs, and I have a Tesco lamp that has the following advice on the package: “Do not use in enclosed fixtures. Usage in recessed fixtures could result in reduced product life”. This is very sensible advice because overheating is detrimental to the electronic circuitry. I do not know if LED lamps come with advice concerning use in enclosed fixtures. Another likely risk of overheating of LEDs is down lighters installed under a thick layer of loft insulation.

I relit my house a year ago. LEDs under the kitchen cabinets, CFLs for table lamps and some ceiling lights, and tungsten halogen in locations where I wanted to use dimmers – some ceiling pendants and kitchen spotlights. So far I have had good results – one early failure of a spotlight replaced by the retailer. I like the colour of halogen, including when it is dimmed. I’ll consider LED substitutes when I see better colour and reliable life (and price). It will come. My main lighting load (CFLs) in the living room can be varied from 14W to 90W depending on the use. I am quite content with this.

Scott X says:
4 February 2014

Can Which provide the detail test results such as the failure distribution during the test. It can reveal if the products failed in infant mortality, normal life or end of life stages for analysis.

Kevin Cairns says:
5 February 2014

I have purchased 2 OSRAM CFL Square bulbs 16 W neither of which has lasted more than a few months. Not cheap either

Scott X says:
5 February 2014

Obviously this Which test report reveals incorrect spec claim is a common practice and most of users prove the actual life is much shorter than the claimed one. What is the next step to remedy it? The users would accept the LED lifespan of 6,000 to 10,000 hrs that is much better than incandescent or halogen lamps but unhappy to see the false claim. The enforcement bodies should help this industry to grow in a healthy fashion.

LEDs are widely used in professional lighting and have good reliability. In domestic use there are so many suppliers of unknown quality producing lamps down to a price and probably not complying with standards that it is an indictment of these than the technology. Short lives tend to be the ones reported, used in unknown ways – particularly temperature in use – so I think it difficult to draw a conclusion. In principle, LEDs from reputable manufacturers with good electronic drivers used correctly are good. How to weed out the rubbish – not a clue other than avoid suppliers you can’t approach when you have a problem. I’d stick with well-known Eurpoean manufacturers who have reputations to maintain.

I can’t help feeling this is creative packaging blurb. The LEDs may “last” 6 or 10 or 20,000 hours, but the control gear is not given any guarantee. They are not quoting the possible life of the “package”, just the LED. A bit like a car with a 100 year life; well the seats and steering wheel or even the wheels. Pity about the rest of it.
Am I correct it assuming this LED life is a projection from a light emmision loss curve? What shape is this curve? Anyone know. Is LED life quoted to extinction of visible light or a percentage or out?
The gizmology.net website has a reasonable explanation of how LEDs work. See part way down column on RHS.

Scott X says:
5 February 2014

I believe the lifespan refers to the whole unit – LED lamp, not just the part of LED component, isn’t it?

Yes, it the electronic driver is integral with the LED chip board. For LEDs where the chip and the driver are separate components then each should be given a lifetime.

Brian, Usually for professional users it is light output that determines “useful” life. This is determined from the data measured up to 6000h, then extrapolated; the extrapolation is an accepted technique and, if anything, is pessimistic. The curve used is an exponential fit, so it is a drooping curve. A “quality” LED would typically be (this, in fact, is based on 10 000 hour data) : 99.5% light output at 1000 hours, 96% at 6000h, 94% at 10000h, 85% at 28000 (one light deprecitation life used) and 70% at 63000h (the other common light lifetime).
Degradation is greatly dependent upon quality – the silicone encapsulant can degrade rapidly on cheap LEDs, the plastic backing darkens and reflects less light (high power chips use ceramic instead), high temperature (often caused by high current to extract maximum initial light) speeds this up.
This all applies to the chip only of course.

Not having any technical or scientific know-how and to get some sort of handle on the issues involved I just did a quick-&-dirty, ordinary consumer’s, assessment of some of the Philips lamps sold by John Lewis, comparimng them with the now-withdrawn standard 60 Watt incandescent bulb which retailed, in today’s money, for about 60p each with multipack discounts commonly available. They were reckoned to have a life of around 1,000 hours so the outlay cost is roughly 60p per 1,000 hrs.

The choice for a “classic” shaped bulb is now as follows –

a. 42W Halogen [810 Lumens, est life 2,000 hrs, price £2.50 [£1.25 per 1000 hrs]
b. 16W CFL [810 Lumens, 8,000 hrs, £5.50 [70p per 1,000 hrs]
c. 9.5W LED [806 Lumens, 15,000 hrs, £14.95 [£1 per 1,000 hrs]

The Halogen is over twice the price to buy and only saves 30% energy compared with the old lamps. The CFL is only a little more expensive to buy but saves 57% energy so has a very quick pay-back; if it goes dim some time in the next three years or dies prematurely I wouldn’t grieve over it. The LED costs a lot more but saves 84% energy and for this type it’s all about the lifespan; the energy saving is huge and if it only lasts half its projected life then it is worthwhile. But so many people are saying they pack up just after they’ve thrown away the packet – it’s not easy to tell whether this relates to the “classic” lamp that typically goes in a table lamp or centre light fitting where there is usually adequate airflow, or whether it concerns down-lighters or spotlight installations. To my mind the 15,000 hrs projected life of the LED stated by John Lewis seems a responsible estimate. At this stage in the development of LED lamps, where there are still many uncertainties, I go along with what Malcolm has said about buying top manufacturers’ products from reputable retailers. You can hunt around on the web and find cheaper LED’s with longer lives, and you might find the same ones as John Lewis sells at lower prices on the internet. If they suit your fittings and give the quality of light you like its probably worth giving them a try. If you can afford the upfront cost, it’s worth replacing any remaining tungsten incandescents and halogens now and not waiting for them to pop. Although the electricity bought for lighting the home doesn’t set the meter wheels spinning like your other appliances do it’s one of the few areas where you can make an easy saving.

I like the practical approach, John. Even though a few of us are very interested in technology and how reliability is assessed, most readers are interested in value for money. 🙂

Perhaps we should wait until LED lamps fall in price and improve in quality. I do hope they don’t become widely hated, as CFLs did.

John, a good way of looking at it. It then depends upon what you want to achieve. If it is just energy, then the LED will save you 78p a year or more cw CFL (often 1000h a year is quoted for use). If it is for nice lighting, then you might use halogens for their colour and dimmability in particular situations, CFLs for general use, and choose LEDs carefully – apart from reliability. some deteriorate in output relatively quickly and some have poor colour.
Hopefully we will be left with a selection of lamps to use in different lighting applications. LEDs will continue to improve in light output, electrical efficiency and price and will be a dominant light source for both domestic and commercial use. The phase-out of of the less efficient halogens may well be postponed until 2018 or later, and hopefully mains voltage versions will be available with more efficient technology, as low voltage are. Fluorescent and LEDs with separate drivers will, I hope, become adopted in domestic luminaires.

In public buildings, commercial premises and street lighting it is usual to replace lamps when a proportion have failed. This is more cost effective even if it is not necessarily environmentally sound. Long lasting lamps that do not fail very prematurely are attractive.

In the home, many of us have continued to use lamps with a short life. If not unsafe, it is definitely inconvenient to have the single lamp at the head of the stairs, in the toilet or bathroom fail. With incandescent bulbs one approach is to fit long life versions, which offer approximately double the life. An alternative is to replace them periodically and put the used lamps in rooms where there are multiple lamps in use or in locations where it is very easy to replace a bulb.

With LED lamps it may be worth producing more durable versions, where the LEDs and electronics are not operated close to their limits and we can be reasonably sure of a long operating life.

The purpose of lighting generally, particularly in public areas, is to provide a designed level of light for safety and to carry out a task. So replacing a lamp (or them all) when their light output falls to a predermined level (predicted by burning hours) ensures the minimum number of lamps are installed balanced against energy cost. That is environmentally friendly.
LED lamps are reliable and long-lived if they are sourced from reliable manufacturers, run on the correct drivers, and in suitable luminaires. They can, and will, last 50 000hours and give the declared light output and colour. The problem in this debate is firstly, there are some (many?) rubbish manufacturers and suppliers in the domestic market, and secondly, LED lamps in unsuitable configurations are being used in luminaires for which they have not been assessed – primarily being run too hot. Were we to spend our money on luminaires designed for LEDs, and use reputable makes, we would be a lot better off. It is not the technology that is wrong, it is the misuse of it.

Malcolm and I have been advocating purchase of LED lamps from a reputable manufacturer. We hope that the manufacturer will want to protect their reputation and not try to sell inferior products. Another factor is electrical safety. We don’t want lamps that could set the house on fire or break in use, exposing live parts.

Intuitively, I would feel safest choosing lamps branded GE, Osram, Philips and several other well known manufacturers.

What about B&Q, Ikea, Tesco and other large retailers? I doubt that any of these are manufacturers of lamps and have no idea which companies supply them. If Which? gives a B&Q LED lamp Best Buy status, should I be setting aside my innate preference for well known brands? One of my reasons for asking is because I bought a couple of cheap and cheerful Tesco spiral CFLs a couple of years ago to find out how they compared with bigger brands. After heavy use, they are still working well.

I have no intention of buying unbranded lamps but perhaps it is not necessary to pay top prices either.

Large retailers will buy lamps (as most other stuff) from manufacturers under their own (retailers) brand name, so you won’t know the manufacturer (they don’t make them themselves). The manufacturer may well change over time. You hope, as in all own brands, that the retailer has an interest in supplying good quality lamps so they don’t get innundated with complaints. So you will get good products. But I suspect you will also get poor products – price not quality driven. One thing all these products should be is safe – there are strict regulations regarding product safety. Are they enforced? We don’t seem to have the resourcs. Your best bet is to check they are CE marked and use them in accordance with instructions.

I have more confidence when a retailer indicates the manufacturer, but this is less common than it used to be – presumably because supermarkets are able to negotiate large discounts and push down the profits of the manufacturers.

CE marking is of limited value because it is effectively a manufacturer’s declaration that their product complies with the appropriate regulations, and also CE marks are put on dangerous counterfeit electrical goods. I believe that CE marking is unfit for the purpose, even though it seemed like a good idea when it was introduced.

The point I was making about CE Marking is that a major retailer is responsible for the authenticity of the product they have sourced, so one hopes they have sourced them reliably. I agree that using a less reputable seller may leave you open to counteits. CE Marking is not abused by the major manufacturers – it is not worth their while to prejudice their business and they generally operate, in my experience, ethically. So the combination of reputable supplier and CE Mark – and read the instructions – is as good as we can get until there is proper policing. Too many suppliers and not enough police for that to happen.
Incidentally, I looked at a warranty today for an LED bulb from a UK supplier (Chinese made). Life was given as 15000 hours, warranty 2 years at a maximum of 4000h/year. However, the normal domestic lamp usage is quoted as 1000h/year so the warranty expires for many after 2000h. Not great.

For “counteits” read “counterfeits”. Oh for editing!

DaveR says:
7 February 2014

I have a pub and have replaced about 60 ordinary bulbs (BC and ES) and GU10s with LEDs from LED Hut website. They cost on average around £6 to £7 plus VAT. I have had them for about 8 months and are on for a minimum of 12 hours a day up to 17 hours weekends – up to 4,000 hours in total. This could be equivalent to well over 2 years in a typical house. None have failed in that time. They are guaranteed for 5 years anyway.
They have the advantage of running very cool which is particularly useful when lighting shelves of spirit bottles! Due to the daily time on, I calculate that my money will be recovered in between 4 and 6 months. An investment of £400 will save well over £1,000 a year.
It is surprising how effective the light output is despite many comments. Three 100 watt halogen bulbs were replaced with three 7watt LEDs over the pool table with completely satisfactory light output. 300w down to 21w. Eight 40 watt bulbs in two windows were replaced with 3 watt candle LEDs. 320w down to 24w
Changing a bulb, often in an awkward light fitting, every few days is a thing of the past. LEDS for me are the only sensible choice.

I don’t know how you have had the pub, Dave, but perhaps CFLs could have saved you a lot of electricity for 20 years or more. Some pubs have been using CFLs for years and others are still using wasting electricity using old fashioned bulbs and halogen lamps.

Best of luck with the LEDs. I have heard very mixed reviews about LED Hut lamps but a five year guarantee is good.

This is very interesting but I assume that Voltimum is a trade body whose members include some of the well known manufacturers of LED lighting. I don’t trust trade bodies, though their publications provide a very useful insight into many issues. The author of this article in Voltimum hands out plenty of criticism but does not provide any information to support the claim that well known brands are in fact superior.

It is interesting that the best performing lamp mentioned in the recent Which? tests is consigned to the end of the list when the tests are mentioned, no doubt because the company is not a partner in Voltimum. 🙂

Voltimum is a portal for major manufacturers – see its website. Do not be too cynical about major manufacturers when they present information. It is best interpreted if you have some knowledge of the industry to weed out technical facts from marketing. I would trust major manufacturers more than LED distributors to give me reliable information.

“Voltimum is the leading internet portal for the electrical industry. It went live in 6 countries in Europe in 2002 and now, in 2013, operates in 13 countries across the globe in 7 languages.

Visit http://www.voltimum.com for details of all Voltimum’s country operations.

The portal is backed by 7 leading global manufacturers of electrical installation products: ABB, Legrand, Nexans, OSRAM, Philips, Prysmian and Schneider Electric, who invested over Eu20 million in the start-up of the project and continue to fund it today. Other leading manufacturers have joined the initiative using the portal as a showcase for their product information, news, training, events and expertise. The portal is also supported by leading trade associations and industry bodies.”

I have very little trust in trade bodies and similar organisations, Malcolm. They are there to support the interests of their members and not necessarily to provide the public with unbiased information. I doubt that claims made in their public output come under scrutiny by the Advertising Standards Authority, though I would be very happy to learn that their public output is independently monitored.

Irrespective of whether manufacturers are large or small, and irrespective of whether or not they are members of a trade body, the consumer deserves safe, durable products that are fit for their purpose.

One of the main reasons that Which? exists is because companies repeatedly let consumers down. I’m going to continue to be cynical. 🙂

Having dealt with manufacturers for many years I am happy to say that many – especially the major ones – do impart genuine information. They have the expertise and in-depth knowledge that other people – including Which? – do not have. Their resource needs to be used, along with others, to ensure that facts are properly acquired and presented. Ignore them at your peril. I have posted quite a few links to documents that certain manufacturers provide that give very detailed technical information.

I am well aware that larger manufacturers provide some very useful and reliable information, Malcolm. However, I have little trust in trade bodies and similar organisations funded by companies and run for their benefit.

While I feel more confident buying products backed by a large organisation, sometimes their consumer products are not up to the standard that we would expect. The Which? test reports provide plenty of poor quality products made by some of the larger companies.

I think we are getting off the topic of dodgy LEDs.

The basic point I am making is that this conversation – and many others – depend upon reliable information and facts for objective judgements to be made. That is not to exclude emotive responses – not by a long chalk – but it needs to be a balanced argument. A question is who do you trust? Not simply to provide the facts, but to interpret situations impartially?

I trust independent assessment of products and information. Regarding consumer products, I trust Which? and other independent organisations to take a balanced approach and demonstrate impartiality. Manufacturers, large and small, are there primarily to make money for their shareholders.

I hope that we can demonstrate that LED lamps from reputable manufacturers do offer significantly better performance and reliability and also with respect to freedom from problems such as radio interference. We certainly need some impartial information and not the thoughts of an organisation representing certain manufacturers.

This is an example of the kind of information a manufacturer will produce to support its products. This sort of fact is necessary to help make objective decisions. Trade associations represent all their members and play a vital role in ensuring that information is presented to, for example, regulatory bodies. They have played a major role in developing energy-efficient products and ensuring that EU policy is kept sensible. Which? is not always impartial in presenting its views to members. It is up to interested members to look at available information to make up their own minds.

With or without help, the EU sometimes comes up with some daft ideas. Many will remember its deliberations on the curvature of bananas. At present I am looking at the durability of marine inboard diesel engines in the new Recreational Craft Directive, where the expected lifetime is only 480 hours.