Regarding 100,000hrs Lumen maintenance claims for LED lighting

December 4, 2018

Regarding 100,000hrs lifespan claims for high bay/high power luminaries-    

 

The majority of Industrial LED lighting products have a lifespan rating of 50,000hrs for what is known as L70 specification.

 

The L70/L90 figure, refers to the percentage of original light output –after the stated lifespan period has passed.

 

This is known as the products lumen maintenance, a 50,000hrs L70 product will lose 30% of its original brightness after 50,000hrs.

 

For example:

 

A site area requires a minimum level of lighting to meet safety requirements as all lighting products progressively have a reduced light output as time goes by all lamps gradually depreciate. The lighting scheme planner must take into account how much light will be delivered after 50,000hrs has passed.

 

As a result, if a room/area requires a lamp which outputs 10,000Lm to meet lighting safety levels for 50,000hrs the planner will select a luminaire with an output of 14,285Lm with a 50,000hr L70 rating – so after 50,000hrs, the onsite light levels are still legally compliant and safe.   

  

Note: If lighting products are not selected this way, the following problem can arise:

 

The site lighting levels were compliant to safety requirements at the time of installation but after 30,000/50,000hrs, the light levels have dropped, and the site end up with insufficient lighting levels this makes the site legally venerable [anyone who has an accident at that time, could make a legal claim against the site].  

 

Claims of 100,000hrs product life span, have little relevance if the lighting product, only outputs 10% of its original brightness after 30,000/50,000hrs and causes a breach of lighting safety laws onsite.

This also results, in the site having to replace the worn out/dim lighting units for new ones at additional cost of labor/new lamps site closure costs. 

 

Can a LED high bay product work for 100,000hrs at L70 specification?   

    

There are two factors which governs this:

 

1. The lifespan of the LED driver [the units power supply] this is governed by the lifespan of the drivers weakest component most commonly the driver’s electrolytic capacitors.

 

The driver’s lifespan will vary – depending on the following aspects:

 

A. The temperature on site- if temperatures are higher than expected, the drivers lifespan can be reduced.

 

B. Power quality – if the mains supply has issues such as dips/interruptions/surges it will place additional electrical stress on the driver’s protection circuitary which may wear out/overstress the driver – causing it to have a reduced lifespan –or fail prematurely.

 

C. The number of times the driver is switched on/off – all drivers have a limit to this [ known as switch on /switch off cycles] if working for 12 hours a day for 365days – 100,000hrs equates to 22 years of on/off switch cycles! 

 

2. LED lifespan – as mentioned, all light sources depreciate over time. An LEDs lumen maintenance depends on the LEDs operating temperature when working in the product.

 

All lighting products have a working temperature range, known as Ta – temperature ambient for example, a rating of -10C to 45C, means the unit is not designed to deliver long term performance -if ran at a constant 50C ambient.

 

If a product has a 50,000hr L70 rating and a Ta of -10C to 45C, it will not deliver the correct light output after 50,000hrs if ran above 45C.   

 

It is important to note, the site’s ambient temperature has to be taken into account – bearing in mind that temperatures at ceiling height [where the lights are] can be greater than found below [sometimes by 10C to 15C or more – depending on the site].  

 

Note: It is possible – that a lighting product could have a greatly extended lifespan- providing the product runs from a clean mains supply [with no issues] and runs constantly at low ambient temperatures.

 

Predicting this, all depends on the design of the individual product and its LED specification- but there are many factors to consider.   

  

Our industrial range has been designed/ tested for worst case scenarios – E.G. if the units rating is 50,000hrs/L70 -10 to 45C, the unit can run constantly at 45C for 50,000hrs and still achieve 70% of its original light level [so the site’s light levels will still be compliant after 50,000hrs] 

 

Be warned-some products on the market, cannot run constantly at their max, Ta rating and deliver 50,000hrs L70 performance – the Max, Ta rating may refer, to the highest temperature the fitting can withstand before failing – not the max, working temperature that allows 50,000hrs /L70 performance. 

 

Q- If a 50,000hrs /L70 [with a max Ta of 45C] lighting product is running well below its’s max, Ta rating – can we predict the lumen depreciation [light loss] after 100,000hrs?

 

As mentioned before, this depends on the individual design of each product – in this case we look at our 150W UFO high bay light as an example:

 

This 150W UFO employs the Phillips Lumiled LED.

 

When the product is running in 25C ambient temperatures – the LEDs maximum temperature is 67.9C.

 

The LED temperature [in the product] is compared with the LED manufacturers lifespan prediction for the LED, this prediction is based on what is known as  an LM80 report –this shows the comparison between the manufacturers lifespan prediction vs actual test data.  Please see Philips Lumiled LM80 data below:

 

LED temp,         L70 time

 

Ts= 105C         >54,000Hrs

Ts= 85C           >54,000Hrs

Ts=55C            >54,000Hrs

 

For Luxeon 30302D  part no, L130xxxx003000W21   CCT=White . If = 165ma. Limited by TM21 rule.

 

As the LED runs at 67.9C at 25C ambient in the 150W UFO- and the max, Ta rating is 45C, we can state that when the 150W UFO runs at 45C ambient – the LED’s temperature is 87.9C. 

  

The LM80 report shows, that the LED will achieve 54,000Hrs /L70 even when running at 105C/165ma!

 

This does not mean the product can run at 62.1C ambient [where the LED would run at 105C] as the LED driver’s temperature limit would be exceeded. 

 

Important note: There are two ways of looking at the LEDs lifespan –LED manufacturers makes a lifespan prediction for the LED – but unless the LED has already been tested for 50,000/100,000hrs there is no absolute proof this will be the case.

 

The lighting industry uses a different system- called TM21 calculus, which works as follows:

 

The LED is measured/tested at 3 different temperatures and is monitored for light depreciation over time.

 

The results are compared with the manufacturer’s prediction- if the data matches the prediction so far, the test house will state the LED is on track to perform as predicted for 6X the elapsed test period – for example, an LED which has been tested for 1,000hrs [and matches the prediction] can have a TM21 prediction of 6,000hrs.  

 

Some lighting manufacturers, state the LED manufacturers predicted lumen maintenance- not the TM21 calculated lumen maintenance [often quoted from a LED data sheet – not based on thermal testing of the actual product].

 

The LM80 report for the Phillips Lumiled is on-going, only 9,000hrs of test time has elapsed – as the test progresses we expect an increase of the L70 hours at 85C. After 30,000hrs of test time we will know more.

 

Please see example of Lumiled lifespan prediction curves at differing LED temperatures on page below.

 

The dotted lines represent the manufacturers LED prediction – the solid lines represent the calculated TM21 prediction so far.

 

We see that LEDs running at 105C do not achieve L70 maintenance after 55,000hrs – but LEDs running at 55C can achieve L60 [60% of original light level] past 100,000hrs.

 

However- if the LED runs 30C warmer, the LED will not achieve L60/100Hrs.This demonstrates, the site ambient temperature can affect the predicted lifespan.

 

 

As the LED runs at 67.9C at 25C ambient in the 150W UFO- and the max, Ta rating is 45C, we can state that when the 150W UFO runs at 45C ambient – the LED’s temperature is 87.9C.

   
The LM80 report shows, that the LED will achieve 54,000Hrs /L70 even when running at 105C/165ma!
This does not mean the product can run at 62.1C ambient [where the LED would run at 105C] as the LED driver’s temperature limit would be exceeded.  

 

Important note: There are two ways of looking at the LEDs lifespan –LED manufacturers makes a lifespan prediction for the LED – but unless the LED has already been tested for 50,000/100,000hrs there is no absolute proof this will be the case.


The lighting industry uses a different system- called TM21 calculus, which works as follows:
The LED is measured/tested at 3 different temperatures and is monitored for light depreciation over time.


The results are compared with the manufacturer’s prediction- if the data matches the prediction so far, the test house will state the LED is on track to perform as predicted for 6X the elapsed test period – for example, an LED which has been tested for 1,000hrs [and matches the prediction] can have a TM21 prediction of 6,000hrs.   


Some lighting manufacturers, state the LED manufacturers predicted lumen maintenance- not the TM21 calculated lumen maintenance [often quoted from a LED data sheet – not based on thermal testing of the actual product].


The LM80 report for the Phillips  Lumiled is on-going, only 9,000hrs of test time has elapsed – as the test progresses we expect an increase of the L70 hours at 85C. After 30,000hrs of test time we will know more.


Please see example of Lumiled lifespan prediction curves at differing LED temperatures on page below.


The dotted lines represent the manufacturers LED prediction – the solid lines represent the calculated TM21 prediction so far.


We see that LEDs running at 105C do not achieve L70 maintenance after 55,000hrs – but LEDs running at 55C can achieve L60 [60% of original light level] past 100,000hrs. 


However- if the LED runs 30C warmer, the LED will not achieve L60/100Hrs.This demonstrates, the site ambient temperature can affect the predicted lifespan.

 

 

LM80 Prediction Chart

 

TM21 is the solid line- manufacturers prediction is the dotted line.

 

Note: Unlike the TM21 calculus -this is only a prediction- this prediction may change/improve as time goes by.

 

Bear in mind, that over 100,000hrs –the lumen maintenance can be affected by the following:

 

1. Miss-judgment of onsite ambient temperatures – or unexpected high ambient, through climate change [this could have a dramatic affect - if the error is compiled over 22 years]

 

2. Sunlight hitting the fittings – creating additional unexpected heat.

 

3. Alterations in the use of the factory – causing additional heat and dust.

 

5. Badly estimated cleaning schedule – leading to build up of dust on luminaires – causing overheating/reduced lumen maintenance.

 

6. Power quality issues – such as dips/interruptions/surges /overvoltage and onsite supply faults.

 

7. If the luminaire uses heatsink paste – it may not have a 100,000hr lifespan.

 

8. The LED driver may not last for 100,000hrs.     

 

TM21 calculus is the industry accepted way of making accurate reasonable lifespan predictions – which also take into account LED tolerance variations [the B value] and other limitations such as driver lifespan.

 

If purchasing a product that promises 100,000hr/L70 performance –we would ask for TM21 proof + thermal test results, and we would also require a warranty that guarantees /covers L70 lumen maintenance when ran at the highest Ta rating.

 

Otherwise it may be just wishful thinking on the supplier’s behalf –based on LED manufacturer’s predictions [but with no hard evidence – such as thermal tests and TM21 calculus]

 

We recently examined a 200W high bay light -which is sold as having a 100,000hrs lifespan. We found it used the same Phillips Lumiled LEDs as our high bay UFO’s. This means the competitor unit had exactly the same lumen calculated depreciation as ours [54,000Hrs /L70].

 

If our UFO ran in a constant ambient temperature of 45C, the LED would run at 87.9C. This would give an LED predicted L60 lifespan of 90,000hrs and a L70 lifespan of 60,000hrs [but this assumes the driver will last as long].

 

The competitor unit may use a different heatsink [running the LED hotter or colder] this would affect its lifespan predictions- this can be assessed, only with product thermal testing and LM80 data. 

  

Note, that LED efficiency is improving all the time – a new product does not use LED from 5 year ago- so the new product use LEDs which have not been tested for 50,000hrs.

 

This means, that most of the new products do not have LM80 report for the LEDs allowing long term TM21 predictions to confirm the manufacturers LED lifespan expectations.

 

So – when we see a product claiming 100,000hrs [with a Ta up to 45C] we know, that the LED has not been tested for 100,000hrs – and the TM21 prediction may only have reached a prediction of around 50,000hrs.

 

This suggests the manufacturer is using wishful thinking, instead of industry proved methods of proof.as well as ignoring to mention- at which point in time the lamp becomes un-usable [ with regards to light level safety] 

 

For example – a 100,000hrs lifespan luminaire, may output 1% of its original light after 100,000hrs [creating a safety hazard] but is deemed to be OK by the terms of the warranty. 

 

Q- Will the LED driver last for 100,000hrs?

 

Again, as designs differ with manufacturers, we can only speak for our own high bay UFO light –

 

On the MGLites UFO, the LED driver’s T case temperature at 25C ambient operation is 52C.

 

On Paper, this gives a driver life span prediction of 100,000hrs – but if the site temperature is a constant 40C [at ceiling height] the drivers T case would be 67C – giving a prediction of around 60,000hrs – not 100,000.

 

So – a 100,00hrs lifespan driver could mean the products max, Ta rating could be around 35C – not 45C.

 

Please see Meanwell driver lifespan V case temp, chart below:

 

Also bear in mind – the aging effects of unpredictable onsite power quality issues – the effects of dust build up [increasing luminaire heat]- and if organic dust is eventually burnt onto the paintwork through time – as well as moisture ingress [ if the gasket seals don’t last for 100,000hrs].  

 

 

 

 

 

LED Driver lifespan prediction

 

 

Are there other factors involved?

 

Where using parallel wired LEDs – a percentage of the LEDs will fail to achieve L70 target – as they may run slightly under the average light output level [this is known as the B value – if 10% are predicted to run below L70 the LED array is said to have a B value of B10].

 

Not all luminaire sold takes this component tolerance variation into account [we see cheap E bay products that do not use additional LEDs to allow for this] - as a result the predicted L70 figure may be incorrect.

 

If a luminaire is used for 12 hours a day /365 days a year – for 100,000hrs, it would have been working onsite for 22.83 years! 

 

A 100,000hr warranty would have to cover the materials of the unit for this time period! – Also bare in mind drivers have a limited number of times they can be switched on/off during their life.

 

Can the supplier who states they have 100,000hrs reliability answer the following and provide proof?

 

1. Q- Would the products paintwork  have peeled off by then- if the paint rusts, could it thermally insulate the fitting and overheat?

 

2. Q-Will the diffuser plastic still be clear – or will it eventually turn yellow after 22 Years?

 

3. Q-Can the LED driver withstand that many switch on/switch off cycles over 22 years!

 

4. Q- Even when not in use- products have a limited shelf lifespan [such as the capacitors in the drivers/ heatsink paste] does the driver [designed to work for 50,000hrs= 5.7years 24/7 constant use] have a 22 year shelf life?

 

5. Q- if wired to the mains for 22 years – will AC supply conditions, such as dips/interruptions/surges/over voltage, cause the drivers fuses to be worn out and fail?

 

6. Q- LED products are increasing in efficiency and design every year- After 22 years – we expect graphene enhanced LEDs and graphene filament lamp technology to be available. Bearing in mind the rate of progress – could the user find that after 5 years, the leap forward in LED efficiency, could make it economically viable to replace the 100,000hr product much sooner?

 

The energy manager would have to consider 22 years of 140 lumen per watt performance verses 5 years of 140 lumen per watt with 17 years of the next generation of LED products.

 

How would this work out, if the next generation runs at 200lumen per watt – or more?

 

7. Organic solvents/gasses and chemical such as Sulphur dioxide, degrade the LEDs phosphor [decreasing lumen output and altering color temp] – over 22 years, could exposure to these chemicals build up over time [we heard of a LED street light, that was affected by traffic fumes- and LED lamps suffering the same fate- when working at a rubbish tip – due to gases emitted from decaying organic matter and disposed solvent products.

 

If installed into a swimming pool facility – would the gaskets eventually fail before 100,000hrs, and let chlorine gas into the fitting [casing the LED phosphors to fail]    

  

Q- Is it really cost effective?

 

The fitting would have to comply with safety lighting levels after 100,000hrs – so the lamp would have to produce at least 40% more light from the start – as opposed to being just 30% over, this means the fittings would use 10% more energy – for a prolonged period of 22 years!

 

Bare in mind, the price of oil/gas/electricity is more likely to go up than down in the future.

 

As the fittings age – they will use slightly more power – but just a few watts [this is factored in on our light plans] the 100,000hrs fitting will not reduce its power consumption, as it depreciates in lumen output – the result is lower long term efficiency, than using a standard 50,000hr/L70 fitting running at 30% above the final brightness level.

 

This must factor into the cost of ownership – we note, it is never mentioned to the customer by the sellers.

 

When we consider this 10% additional energy consumption, with the fact that LED products get more efficient, and cheaper every year- and then consider the effect of the last 22 years in terms of cost/efficiency of lighting products [and the current rate of progress] – we feel it the total cost of ownership, may not be appealing as it first seems- the gamble may pay off, if LED products increased in price.   

   

Bear in mind – the prices of LED products tend to fall – not rise over time.

 

Sometimes customers are un-aware, that employing sensors/control gear with luminaires, can create the savings they are looking for. Very long-term fixed energy savings – in a fast moving/developing market, could be a red herring- and a potential false economy!  

  

To recap-

 

LEDs lumen maintenance is TM21 calculated- to avoid using long term LED manufacturer’s lifespan claims – as great errors can occur over such a long time period.  

 

TM21 lumen maintenance predictions take into account additional /overruling product limitations – such as driver lifespan /LED tolerance variations - and is based on the actual products thermal testing compared with the LEDs LM80 testing results – which is a controlled industry approved method, of providing usable LED lumen maintenance data from the manufacturers LED predictions - as opposed to directly using the manufacturers [unproven] lumen maintenance predictions in the blind hope ,that no other variables [such as those mentioned here] will cause their 100,000hr usefulness claim to fail. 

    

Claims of TM21 calculated predictions of 100,000hrs /L70 would require the LED to have 16,666Hrs of LM80 of elapsed test time – LED efficiency improves every year- the newer more –efficient LEDs are new to the market and usually have not undergone more than 9,000hrs of test time so far -so, the max TM21 prediction you would expect to see would be limited to 54,000hrs [unless the product is using LED from 2 years ago – but the efficiency/energy savings would be lower than new products].

 

It may be provable, that the LEDs in a product -will run onsite for 100,000hrs – but if the LED driver only lasts for 70,000hrs, it would be pointless.

 

If the driver last for 100,000hrs – we would expect the products max Ta to be low -We note that many UFO style products use the very same LED driver – made by Meanwell. The HGB240 is often used on 200W models and if ran in an ambient of  40C – would have a lifespan of 60,000hrs.

 

If a product is said ,to last for 100,000hrs and also have a max, Ta of 45C- they would have to use a driver with a higher Ta rating than the Meanwell HGB240! – we would exepect the product to cost much more, and be accompanied by definitive proof such as, TM21 calculus predictions- and a warranty that guarantees the lumen maintenance after 100,000hrs of operating time.

 

Onsite temperature conditions affect the products lumen maintenance -  a 10C error in guessing the highest site temperature can reduce the light output after 100,000hrs by 20,000hrs – or may lead to driver failure before then.

 

The gasket seals may it last for 100,000hrs or 22 years.water ingress may occur.

The paintwork may peel off after 100,000hrs – it may breach food safe standards- or look unsightly.

 

Important note:

 

It should be noted that the lifespan line of the LED is not linear – if the output decreases by 30% over 50,000hrs ,it does not mean that it will decrease to 60% after 100,000hrs. Looking at a manufacturers LED prediction [not TM21calulus] we see that this LED drops to 60% after 60,000hrs – even though, it achieved a 30% drop over 50,000hrs.

 

This is why using TM21 calculus is more reliable, than using raw prediction data with wishful thinking.LED lumen maintenance decay curves are not linear!

 

It should be noted, that all light sources depreciate over time – once the onsite light levels have dropped below the site’s required lighting safety standard level, the lamps must be changed by law.A 100,000hrs service life LED fitting may fail to provide legal/safe lighting levels after some time [depending on varying site conditions].

 

We would not advise anyone to create a budget/cost of ownership around this – as it is hard to predict at what point in time the onsite light levels, will become non-complaint- and need replacing.

Q- Does the green line  representing the LED lumen depreciation line], look like it could reach the 100,000hr point at 70% or even 50% brightness- or does the depreciation rate increase dramatically after 50,000hrs?

 

Looking at the first angle of the line [before 50,000hrs]– it could be falsely assumed we have 100,000hsr at L70, this could be the basis on which the 100,000hr useable lifespan claim is being made – the same logic could be applied to the yellow line, at 22,000hrs the line looks as if it will go diagonally to 50,000hrs at 60% - but in reality it does not.

 

We would be sceptical -unless there is genuine proof provided from 3rd party test houses ,showing a comparison with thermal test results and LM80 data –with 3rd party TM21 calculus involved.

 

Please see lumen maintenance comparison ,of differing light technologies below:

 

Lumen depreciation of differing lamp technologies.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Please reload

Featured Posts

Regarding 100,000hrs Lumen maintenance claims for LED lighting

December 4, 2018

1/5
Please reload

Recent Posts

February 6, 2017

Please reload

Archive
Please reload

Search By Tags
Follow Us