Ray Molony: Where are we with LED technology and how is that affecting the specification process? What confidence do you guys have that when you spec something that it will perform as you expect? How will it perform in a year or two’s time?
David Atkinson: The main thing is to go to a manufacturer that you feel confident with. One that’s a well-known brand. One feels safe with just few catalogues that you can go to. But every day you get bombarded by emails from China.
Doug James: There are two tiers to it of course. We all retreat in uncertain times to the comfort zone. We have manufacturers, the likes of iGuzzini, that you know, through their rigorous development and testing, will come up with good solutions. But you also have another layer that we’ve never had to think so much about before. Lamp manufacturers were much of a muchness, but now we’re talking about different LED suppliers. There are people like Seoul and Cree making chips, then people making emitters, like Philips and Osram. There’s suddenly this layer cake of points at which you can dive in and get the detail. A lot of people will just stick with the manufacturer which may be using Xicato or Cree, others may want to go right down to component level.
Iain Ruxton: You may or may not have the option to do that. The old model of you get a luminaire, you get a lamp and you put one in the other – that’s gone. It’s not just a new technology, it’s a new process. It changes the nature of what a light is. Now we just spec one thing. We take all the decisions in one go. Let’s face it, we’re all used to busking it a bit – we can’t do that as much.
All about adjustment
David Atkinson: Manufacturers equally are looking at having adjustability in the fixture. That’s good in museum work where you can go from 10 to 65 degrees. So you don’t have to worry about changing a complete module.
But our perception of lighting has to change. We’re in a completely new mindset now.
Gareth Jones: The problem is that some of the boundaries in the layer cake are not well defined. When you’re building a structure, you have to make sure the foundations are right before you build the rest of it, so when you from the gate of an LED module to a light engine and then to a fixture, you need to make sure your checklist is signed off for that module.
The problem is that there’s no international level playing field for how you sign that product off. There must be standards and testing so you know what to check off on your checklist.
Iain Ruxton: Even just the terminology… we haven’t really established a shared vocabulary yet.
Lumbered with longevity
Ian Stanton: What’s causing us problems is the longevity of the product. With traditional light sources the life cycle of a product could be 20 years. Now the life cycle could be two years – it could almost be extinct before we launch it.
The fitting is designed around a particular module for heat dissipation. Now there’s the prospect of active cooling, and we’ve never had moving parts in our products. We’re now talking to people about fan noise.
Paul Nulty: A bit of a bugbear I have is lumen depreciation cycles. Generally for a white LED life is about 30,000 hours to 80 per cent. We all used to design to roughly 0.8 maintenance factor, suddenly with LEDs we seem to have forgotten about that.
Iain Ruxton: It’s not just a matter of ‘is there still light coming out of it?’, not even ‘how much light is coming out of it?’, it’s ‘what’s the quality of that light?’. If it has shifted ten McAdam ellipses then it’s not the same damn thing any more. The quality of the light at the end of that 50,000 hours or whatever varies wildly.
Doug James: In a way, what we’re saying is that a lot of work has been done and made some very expensive products that are no more efficient and no better in colour rendition over life than metal halide.
Ian Ruxton: The reality is that we still don’t actually know what we’ve got. There’s no real data, it’s all based on accelerated testing.
Darren Hicks: One retailer put lots of QL fittings in over stairs. The lives were supposed to be 10 years but they actually lasted six or seven. But you have to replace the whole thing. It didn’t really save us anything at all. LEDs will be the same.
Paul Nulty: I don’t understand why we can’t standardise on LED bases and terminology. I don’t understand why the source can’t be treated as a lamp. I don’t want to be specifying something where you have to throw away the entire luminaire every time.
Iain Ruxton: Chucking away a massive great bit of aluminium because this tiny thing has worn out is absolutely unsustainable.
Doug James: And we don’t really know about the colour effect over time. Everything points to the need to be able to change up.
Ian Stanton: I think we’re getting there with interchangeability. With the modules, it’s three screws and you can change it. The difficulty with these twist-and-lock systems is thermal dissipation, and the life goes down to 20,000 hours. That’s not much different from a metal halide. So the advantages you have with maintenance are disappearing with the flexibility.
Gareth Jones: One of the things that seems to be coming out of this is that there’s a misalignment between the rated life claims of the module manufacturers and the warranties that they’re imposing. What you really want is a product you can rely on and you’re paying for the life that you want. If a manufacturer is quoting you a 30,000-hour lifetime, then the warranty should be aligned with that.
Neil Painter: An LED warranty is generally to LM70, for instance. How do you measure LM70 in a retail environment? It’s still emitting a light output. Warranties on other sources are generally based on failure.
Max Guzzini: When we talk about lifestrong, the big challenge is how to dissipate heat. All the information you get from the lamp manufacturer doesn’t consider how you design the product. Much more than before, the luminaire manufacturer has to understand how to dissipate heat.
Few companies understand the roadmap of LED evolution. If they do, they can ensure that, when they design a fitting, they don’t use a technology that is already obsolete.
Rory Marples: Let me give you a simple example. If our standard product uses an Osram chip, and you as a lighting designer want us to produce a light distribution that varies slightly from the standard product – say a 40-degree beam angle rather than a 20-degree angle. On the surface, that’s a straightforward exercise, but we need to look at how, from a component perspective, we create that distribution. There is a multitude of add-on lenses for specific products. If you want a 40-degree beam angle and there’s not compatibility between the lens and the chip, we then have to look at another chip. Then you have to redesign the thermal performance of the product.
We live in exciting times
Tim Downey: I think this is a really exciting time, what we’re doing is getting back to where we were 20 years ago, before dichroics and before metal halides when, as a lighting designer, you had to design a light fitting using different components. There were all these fantastic projects were you designed every light fitting on the job.
It’s gone full circle, the best way to specify LEDs now is to write a performance brief. And you judge it on technical criteria, optical criteria and visual criteria.
We’ve got some educated clients who are aware that they don’t have to have what they’ve always been told they need before. They know they can have something they want.
For manufacturers, there are areas they are in now that they might loose. But there are areas they are not in that they will gain. We all realise that coming down the track are the Samsungs and the Toshibas.
I was giving a speech at the SSL, and I was asked to put forward a case for abolishing the codes. My position was that we should be moving away from lux-based standards, because where we’re heading more and more surfaces will emit light. There are lots of ways of thinking about a problem and it’s not about light hitting surfaces any more, it’s about light coming from surfaces. Luminance not illuminance.
The only way to deliver these things is with LEDs. Hard on the heels of this are OLEDs. Then you can forget about optical quality. It’s all about how bright a space is and wether it’s good enough to do some work.
We have to evolve or die. The challenge for luminaire manufacturers is to start offering a range of products that enable us to realise our design ideas. Soon the market will be shared with some really large names that make Philips look like minnows.
The SLL and the dinosaurs in that room were all, to a man, happy to see evolution of the code. Interestingly, the younger people agreed with me and asked: ‘When can we get some kind of standard that deals with surfaces?’
Ray Molony: How do you specify LEDs currently?
Tim Downey: To do a performance spec we have to talk about not only the optical performance but also the environment that the light is used in. Our performance specs are part educating documents where we say ‘technically, it needs to do this, but holistically it needs to contribute this to the overall project, and aesthetically it needs to fit in with the entire design remit’. We put down a minimum of three and a maximum of five manufacturers. We ask if they’ve done something similar before.
The client doesn’t judge the spec on a product any more, but on what it’s got to do.
Ray Molony: So you’re back to quality brands by a different route?
Tim Downey: We have to be inventive to prevent people judging things solely on cost.
Iain Ruxton: The proven record stands for so much. The end-user wants to know it works.
Darren Hicks: The very short history is ruining LEDs.
Tim Downey: Current products are a little bit better from an energy point of view, but a lot better from a long-term point of view. You can’t look at circuit watts now, but 30,000 hours down the track. The big savings that you’ve got to point clients at is maintenance.
The light quality is getting really good, the control gear is getting better – there are solid state drivers coming down the track.
Neil Painter: Reflectors continue to gather dirt and dust. At Nando’s we’re reactive at the moment and do spot relamping. Our restaurants generally have low ceilings so it’s quite easy.
We have started to go down the LED route in our new restaurants. Maintenance is something that’s interesting. We’re doing a lot of uplamping and we’ve found some very interesting challenges, particularly around dimming. A lamp may be dimmable but it’s not plug and play.
Iain Ruxton: LEDs as a technology are inherently awesomely controllable… but only if you’re starting from scratch. If you have an existing system that’s fundamentally completely incompatible with LED, it’s a hideous, nasty hack.
Tim Downey: You can’t just relamp forever.
Gareth Jones: I wouldn’t underestimate the longevity of relamping. The LED retrofit lamps that are coming on the market now, apart from cost, are very good. Recent studies have shown that the growth in retrofit lamp products will outstrip retrofit luminaires for quite some time.
Why should somebody strip out their luminaires and change?
Tim Downey: But you can’t just change the lamp to get the full benefits of LED.
Doug James: You’re dead right, but a lot of people, frankly, don’t care. The manufacturers that are making retrofit lamps in a separate division. Luminaire development and production is a another division, and they each have their own goals and missions to conquer the world.
Tim Downey: The retrofit market will die away quickly as the benefits of a full LED install start to make themselves clear.
Darren Hicks: That’s going to be a long way off.
Iain Ruxton: It can’t be. If I’m prepared to pay a premium for the benefits I get from LED technology, I can go the LED route. Or I can go the way I went before and not pay the premium. If the old way doesn’t exist any more, LED is the only choice, LED has to come in at the same price.
Tim Downey: People are beginning to realise that capital expenditure and operating expenditure have to be linked. Now the changes in opex are potentially more significant than they used to be.
Neil Painter: The amount of disruption during a wholesale change of lighting systems, even over a 10-year period, would be high. It’s not just about the cost of the product. I don’t think you can be looking at very tight timescales.
Paul Nulty: I have a sneaking suspicion it’s going to go another way. The really exiting thing at the moment are the big players coming to the market – the Samsungs and Toshibas. I think they’re really going to change the state of the market. What they’re really interested in is the domestic market, where they want retrofit. Ultimately, increased competition will drive down prices and there will be increased competition in terms of quality, so the quality of the light output will go up and efficacies will increase.
I suspect we’re going to end up at a point where we have some very good, very useful retrofit solutions. And I suspect we may get to a point where we do treat LED and luminaire separately again because I think the domestic market will be the key driver.
Market forces outside professional lighting circles will drive this.
Doug James: I slightly disagree. I think that the domestic market for lighting doesn’t drive the professional market. It never has and I don’t think it will because we’re moving from one type of technology to another. I think they will develop separately.
Iain Ruxton: The two channels have different agendas, and they will go in parallel.
Darren Hicks: One retailer has gone for a weird idea of taking fluorescent tubes out of their 600 x 600mm fittings and putting LED tubes in. As an engineer, you say that’s deeply wrong, but the finance director would have looked at it and said: ‘It’s got a one-year payback.’ The figures stack up really well, but it looks horrible and the wooden floors and products look dead.
Doug James: That’s the kicker, because if you want to get good colour rendering and get to a level close to watts per metre of decent T5, you can’t do it any more efficiently than you can with a T5.
Darren Hicks: T5 is still the correct answer. Retailers kind of get that, but there’s somebody at the top saying ‘we want LED’.
Dave Tilley: It fills my heart with joy that this is the first time that I’m fully on board with what the lighting designers are saying. It seems to have shifted full circle, which is very nice indeed. I think one of the key points that Max made earlier, is the critical one, is about application.
It’s about what you are trying to achieve and I am also pleased to hear that the table is able to recognise that the world is going to change and there are going to be other names to reckon with. I think the important thing to stress as well, someone at the very beginning said about trusting a brand, there are other brands out there that we have never heard of that are actually very good, and we as lighting people need to open up and do exactly what Tim said, and understand specifications and understand how things are put together more because otherwise, one, there is the risk we will miss an opportunity, and two, there’s a risk we’ll sideline brands that are very strong and have good quality products to offer.
On the retro-fit lamps side, it’s a means to an end. The point that Darren and Neil have made is a valid one. At the level that I operate, on the multi-chain retailers, we don’t have the luxury that Tim and his team have of dealing with flagships and people that do understand the difference between maintenance and build. Although that said, they understand it, they are not prepared to put the two disciplines together, so what we have to do is to compromise. But compromise within a reasonable application. The other thing to recognise is that because we are dealing with difference technologies, and we are moving away from slavishly following lighting design and lux levels, what we are actually looking at is ‘look and feel’.
You can change an MR16 and put in an LED equivalent and get a very good look and feel. Neil operates a restaurant where there are three 28-arm standard pendants where we used two manufacturers’ candles. The reason we used two, was that we wanted to try two — one had a relatively high output compared to the other. And that worked extraordinarily well by balancing those two outputs. It’s quality of product, it’s about look and feel, it’s about application and let us not forget, there still will be failures and there still is the issue that at the end of a certain point in time, you will be throwing away a lot of metal.
Onus on the supplier
Tim Downey: That’s true, and the onus then is on the supplier to replace it. So the contracts they are writing now, are not their own departments that buy and sell the light fittings, they are entering into framework agreements with suppliers who supply the entire product, after-sales service, and give us a new system in 10-15 years.
Ian Stanton: What is a good warranty? What does everyone see to be a good warranty on an LED product? If someone is claiming a 50,000-hour life, what do want as a specifier and end-user as a warranty. What does that warranty encompass?
Iain Ruxton: It is pretty meaningless at the moment. Warrantying an LED product right now is a largely pointless exercise.
If you sell me something now and you warranty it for five years and in three years time I need it replaced, whether it’s because it’s failed or someone’s smashed it with a ladder, it’s not just about failures. Things get damaged as well which may cause you to replace something.
And even if you wanted to, your diode suppliers won’t be making the same diodes. So you can’t actually do it, so warranties just become pointless.
A lighting company recently told me that they’d give my client any warranty if they are prepared to pay for it. They will warranty anything for as long as you like, if you are prepared to pay enough because warranties are completely meaningless.
Dave Tilley: If you go back to 2007 when I wrote my first article for Ray, it was about what’s the value of a warranty. The point that you are making is perfectly correct. In the world of LEDs, even more poignant now than it was when realistically, we were talking about work at the time we were doing with Neil when he was at New Look. Where we were warrantying the product with the manufacturer obviously, and we were running the warranty as a database so that if a product did fail, it did get replaced under the warranty.
With LED, the point that Darren made is true. We talk to manufacturers all the time about the length of their products at certain parameters and they’ll say, for example, 30,000 hours, which in our world is six years. But then you ask what the warranty is and they say a year. There’s a real disconnect between what manufacturers are saying the longevity of the product is and what they are prepared to warranty it. And even if they were prepared to say five years, the point that you have made that if the product will still exist in five years, and indeed, will the person that signed-off the warranty still be at the company to enable that warranty claim to be enacted? It’s about the process of managing the warranty and it becomes more complicated now we’re in this wonderful world that is LED.
Light as a service
Iain Ruxton: One of the places that could take it, that meets what Tim is saying about the AA situation where at end of life, at throw away the metal time, it’s the manufacturer’s responsibility to resolve, refurbish or whatever to deliver it, then you start moving to something that some people have started to talk about quite seriously, which is the idea of light as a service.
Light as a utility, where you don’t buy light fittings anymore, you get your light supplied by a supplier of light in the same way that you get your electricity, your water. So it does become exactly what Tim was talking about, where Ian doesn’t sell me some lights and then give me some meaningless warranty and then some point in the future we maybe buy some more lights. He (instead) sells me lighting for 20 years and I pay for it quarterly or whatever. People are talking about this kind model.
Doug James: Philips have been doing it with floodlighting for years.
Ray Molony: Is that called energy performance contracts?
Max Guzzini: If you let me know how much you want to pay, I can give you a warranty or whatever you want. But this is not the answer. It is the guarantee you can give according to the market price. It’s how you manage the warranty. All the companies are going to start fighting to give a longer warranty with the same prices. I think it’s going to be suicide for some companies.
Ian Stanton: If you specified, for example, a linear wall washer for a façade, and the worry is that if a couple of chips were to fail five years after you specified it had been supplied, the difficulty is that in theory, the way that technology is going, the lumen output of that product will maybe double or even triple. So how are you going to keep the continuity of this light effect?
Paul Nulty: What we desperately need is an industry standardisation on lumen output or intensity because that’s the driver. Yes, if there is development to be made in LED technology, then fine, but do that in reducing the energy consumption of the LED, not necessarily increasing the output.
Time for a design freeze?
Gareth Jones: We need a ‘design freeze’.
That’s exactly what the retro fit lamp market does already. It gives you the standard blocks of luminous output and angular distribution. It doesn’t matter if the LED changes inside, the manufacturer maintains and tweaks the electronics year on year, to make sure that either the current is going less or they have got less chips in, to give you the same specification of lamp every time. There’s no reason why that can’t be done in a luminaire.
Rory Marples: One of the fundamental issues is the LED and the performance of the driver. If you have a more intelligent driver, then you can compensate for performance of LEDs. Even now, the drivers with relatively straight-forward DIP switches that allow you to alter the drive current to the chip and therefore the output.
Doug James: What I would like to see is luminaire bodies with easily replaceable LED light engine components and easily replaceable drivers. So you have a plug and maybe a couple of screws for a driver and you have a plug and a couple of screws for the emitter. And then the body remains. The body is then good for another 30, 40 years.
Ray Molony: An interchangeable modular they can replace?
Ian Stanton: It’s not such much they are looking at the interchangeable modulars, but the remote phosphors… they are specifying it because they feel the remote phosphor looks better. It’s what people want to see rather than all the multi-chip. It’s just a visual perception, rather than performance.
Dave Tilley: One of the things that’s always been lacking in the lighting industry is good quality education. The problem is that the distribution market doesn’t drive it down far enough into its branch network. The industry’s got two stark choices going forward: it either tries to miss out distribution and effectively change the way the process has operated for years, or it desperately tries to educate.
I phone round distributors from time to time, just to see what kind of response you get when you ask for an MR16 LED. The response is woeful, absolutely woeful – even to some degree at branch manager level.
Darren Hicks: We will move towards specialist maintenance.
Doug James: The problem we’re discussing is valid, but it’s not unique to this technology.
Darren Hicks: It might solve itself because it’s a specialism.
Tim Downey: There are plenty of good examples of where we’re going with this. Look at the car industry. You can’t service a car now. Lighting, at a certain level, is heading in the same direction. The question for people like iGuzzini is wether or not they want to get into solving problems further down the line, so a warranty means solving a problem. After-sales service will be as important as supplying the product.
Iain Ruxton: What they’ll sign up to provide is light, not lights.
Rory Marples: I’ve been involved in a project where we as an organisation have taken it upon ourselves to over-order a product on the premise that if there is a failure, we have the replacements available immediately.
Iain Ruxton: So the business to get into now then is building warehouses?
Doug James: This all works perfectly well as long as the mandarins don’t keep pushing the boundaries of the energy codes. As an industry, we’ve been driven down a route towards certain types of products because, effectively, other products have been outlawed. What they’ve done is look at the figures the LED industry is putting out and said: ‘We can achieve that now, 55 lumens per watt or whatever is realistic.’ The next time the LED industry says it can achieve 100 lumens per watt, they’ll jump on that and say ‘now that’s our standard’.
Someone needs to take those guys in hand. I believe the government’s issue is energy generation. I don’t think they give a fig about true sustainability but they do have a problem generating enough energy to keep people happy in an expanding society and economy.
Gareth Jones: If you’re in Europe, theres’ nothing forcing you to use an accredited lamp, and there’s nothing forcing you to adhere to a performance standard. Apart from the enhanced capital allowance [ECA] for luminaires, which is a very loosely monitored programme.
Darren Hicks: Are all manufacturers trying to make sure their fittings comply to ECA and get them on the lists?
Gareth Jones: I’ve asked a number of luminaire manufacturers what their approach is and investigated the way in which the ECA scheme portrays itself. A number of manufacturers are writing letters to certify that their products comply, some write letters that certify they comply in part. There’s no monitoring of that.
If you look on the ECA website, there’s a set of rules there. You must do lumen maintenance tests on the luminaire and test that it passes certain criteria. And then you phone up the ECA helpline and they aren’t interested in the tests as long as the manufacturer is satisfied with their suppliers and the product reaches the energy performance level required.
Adopting best practice
Gareth Jones: Things are going to get better because they’re going to be probably adopting what I think is real best practice in the industry, which is what the US guys do with Energy Star. They said: ‘How do we pragmatically, cost-effectively enable manufacturers to understand their product performance?’ They should be looking at this much more quickly in Europe.
If I use an LED and I’ve got established LM80 data, the chip-level data, on lumen maintenance and the temperatures and currents that are used for the LM80 tests are the same or more onerous than those on the LED in the luminaire under normal operating conditions, then I can be assured that the LEDs are going to live in a similar manner to how they will in the LM80 data.
The LED manufacturers are doing all those tests anyway. The luminaire makers don’t have to do that, but what they do have to do is ensure that temperatures and current are consistent with the LED lifetime data.
That is what the ECA scheme will probably end up doing.
Doug James: What’s fascinating for me is that the economy that’s way behind the rest of the world in terms of light technology – the US economy – has the best standard for LED. And my experience of the US market is that all the designers will rigorously use it.
The participants in the Lux magazine roundtable forum on LED specification and standardisation were:
David Atkinson, DALD
Iain Ruxton, Speirs + Major
Paul Nulty, Paul Nulty Lighting Design
Neil Painter, Nando’s Restaurants
Dave Tilley, DJ Consultancy
Tim Downey, studioFractal
Gareth Jones, Lux TSI
Doug James, Mindseye
Darren Hicks, Design Brook
Rory Marples, iGuzzini
Max Guzzini, iGuzzini
Ray Molony, Lux magazine
Ian Stanton, iGuzzini