A care home client asks for a strategy document that explains how the long-term issues of an LED installation should be handled.
Its still the case that most lighting design and light planning does not look much further than the initial handover date to the client. LEDs are a ‘long-lasting’ light source and what happens when they start to fail can be left to someone else to worry about. It’s not that simple, though. These things never are. This is what I presented to my client:.
.
LED life and what happens when LEDs wear out:
The most onerous aspect of the shift to LED lighting was the issue of what happens when a fixture reaches the end of its ‘operational life’. When it was only necessary to replace a failed light bulb, it wasn’t an issue; but we now how face a different situation and it’s important that everyone understands the implications of a failed fixture.
NOTE: ‘operational life’ doesn’t necessarily mean the end of actual life. LEDs tend to fade away rather than ‘blow’. But the fading away means that the light from the LED diminishes both in lumen output and, usually, in colour quality.
.
What should we expect from an LED luminaire?
Recent legislation in the EU – and fully supported by the UK lighting industry requires that an LED luminaire be capable of disassembly so that only failed components need to be changed. Luminaire typically consist of four major parts:
- The body
- The optical system
- The LED array
- The LED control gear (the driver)
The body: We don’t expect the body of a fixture to fail unless its in a hazardous or extreme environment where material integrity can be compromised.
The optical system: The optics of a luminaire (reflector / diffuser) can be affected over time and may need to be replaced at some point
The LED array: LEDs certainly deteriorate over time – and this is one of the main foci of this document.
The LED control gear: drivers fail, usually with a similar regularity as the LED arrays that they are controlling. This is also discussed here.
Changing an LED array and its associated control gear should be as straightforward as possible, though some manufacturers prefer to retain control over that action. The more accessible the product, the easier it is for the end user to carry out the replacement work themselves.
.
How often should LEDs need to be replaced?
And this is the nub of the issue. LED manufacturers produce data that describe the length of time that their products should be operational. There are two metrics that we need to look at:
Over a stated length of time, we expect a percentage loss of illumination (the ‘L’ figure) and a percentage of LED failure (the ‘B’ figure)
We have a ‘standard’ operational life figure that is based on 50,000 hours usage. Most manufacturers report on the performance of their luminaires, using L70 (a 70% output) or L90, a 90% output figure (relating to a 30% loss or a 10% loss in outputs).
Together with the L figure, we also see figures like B10 (10% fixture failure) or B50 (50% fixture failure).
What we should always be looking for is something like 50,000hours, at L90: B10. This means that we can expect the best in terms of light delivery over time and the most reliable in fixture longevity.
.
Do fixtures need to be replaced?
This has been the ongoing focus of discussion around LED product development. Accepting that components will need to be replaced at some point, the most efficient way to achieve this is to have fixtures that can be opened without damage and exhausted components easily removed and replaced with new ones. Ideally, the work can be done by a qualified electrician and does not require the original manufacturers to be involved in the process beyond providing the replacement components.
.
Are these life figures affected by LEDs being dimmed?
Yes – the effect of dimming is to reduce the electrical and mechanical stress on the LED components. In a care home environment this is especially important because we will have luminaires working 24 hours a day, at various levels.
NOTE; with 8760 hours in a year, it means that we achieve 50,000 operating hours in just over five and a half years.
An installation that has been designed to deliver 300Lux will then be delivering 270 Lux and we can anticipate a failure of 10 fixtures per hundred installed. These are not insignificant figures, despite being the best that we can look for.
Getting performance figures on dimmed installations is very difficult because of the random nature of most lighting control, but anecdotal evidence (an oxymoron there, perhaps) suggests that we can look forward to a doubling of performance lives. And that suggests that we may be looking at a ten-year term before LEDs need to e replaced (though we mist keep an eye on the components within the control gear (drivers) because they don’t get any real relief by dint of reducing light output.
.
Can LED outputs be adjusted during manufacture?
This is one of the unusual aspects of LED lighting. Because the LED is an electronic component, the circuitry can be set so as to reduce or increase the LED light output. As you’d expect, one has the effect of increasing life, the other of reducing life. As an example, figures from Tridonic provide a good example of this:
Assuming typical performance criteria this is what happens when the drive current is altered:
For an L90/B10 array (gear running at 55C within the fixture):
- at 100mA: 72,000 hours
- at 180mA: 50000 hours
- at 210mA: 50,000 hours
- at 370mA: 44000 hours
We have the choice as to which drive current to use. The starting point would be 180mA (if there’s no difference between 180mA and 210mA, why bother to drive that 16% harder?)
From that position we can make a judgement: are we under- or over-delivering on light output, which is the crucial test for circadian entrainment. If we’re over-delivering, then we can consider reducing the drive current to 100mA and gain substantially on expected life; if we’re under-delivering, then we have the facility to raise the drive current to 370mA and mitigate the reduced operational life by looking harder at the dimming levels.
.
Summary:
- Basic criteria for LED performance should be an operational life of 50,000 hours, that offers only 10% reduction in light output (L90) over that time with only 10% likelihood of failure (B10).
- Dimming will improve on these figures.
- The cost of component replacement depends on how easy it is to remove exhausted/failed components, which should preferably be done on site by local contractors, without the need for removing fixtures.
.
.