Maintenance factors: Why everything you know is wrong

Maintenance factor has worked very well for many years but eventually, like so many things, the MF principle became ossified to little more than a tick-box exercise.

The background

Fifty years ago the world was a different, dirtier, place. Factories produced air-borne pollutants the size of small rodents, every office was a pall of cigarette smoke and there were still enough open coal fires in the home to warrant the Clean Air Act of 1968. It’s a wonder any of us survived.

Out of that gloom came the idea that lighting schemes should be designed to take such filth into account. After all, there was little point in designing a scheme to deliver 500 lux if a layer of dirt settled all over the fixture within months of it being switched on.

And there was more to this than just the effect of light loss due to dirt. Over time, the light output of the light source, typically a fluorescent tube, would degrade; the efficiency of the control gear and the translucence of diffusers would also deteriorate.

All of this needed to be brought together in a single figure, and thus Maintenance Factor (MF) was created. Formulae were created, factor values were agreed and tables were produced. It worked very well for many years but eventually, like so many things, the MF principle became ossified to little more than a tick-box exercise. And that’s where we find ourselves today.

Maintenance Factor today

The details for calculating Maintenance Factor are set out in document BS EN 12464-1 ‘Light and Lighting – Lighting of work places’. This is the standard that specified lighting requirements for people in indoor work places, meeting the needs for visual comfort and performance. As regards MF, it states that the figure should be calculated for ‘the selected lighting equipment, environment and schedule’.

This means that, in the event where there is more than one luminaire within a space, individual calculations are required for each luminaire, with the single MF figure being determined pro rata from the collation of those calculations. A ‘guess’ that MF = 0.8 does not comply with the requirements of the standard.

Here’s why maintenance factor has never worked

There isn’t a lighting engineer working today who doesn’t automatically think ‘0.8’ whenever Maintenance Factor is mentioned. It is bred into every one of us. But does that figure stack up? Let’s look at how the MF formula described above sees it:

What kind of building is this and what kind of maintenance regime provides these numbers?

This is a ‘clean’ environment. The space is lit using (T5) fluorescent tubes that are replaced annually and the luminaire is kept clean and dust free. The room is deep-cleaned/re-decorated annually.

In other words, this is not a space that exists in the real world. It would even be difficult to find this room in a hospital.

This is a ‘normal’ space that  has been cleaned and decorated properly every three years. The fluorescent lamps are replaced and the luminaire cleaned every three years. This is a recognizable maintenance regime, but the Maintenance Factor does a lot of damage to the cost of the lighting installation.

And herein lies the problem. A MF of 0.8 sounds reasonable. We can all imagine a lighting installation losing 20 percent of its illuminance over time. But 40 percent? This simply unacceptable. So we decided, as an industry, not talk about it and trust no one will notice. And no one has noticed, until the LED came along and opened the lead-lined box that is Maintenance Factor.

These factor values are taken from CIE097.-2005: Maintenance of indoor lighting systems

What happens next

The familiar saw-tooth lumen degradation curve demonstrated how a regular programme of lamp changing could maintain light output over time, together with a luminaire cleaning regime. This worked well because of the ‘limited’ life term of the light source, but the LED has changed all of that.

There is current evidence of a ‘race to the bottom’ whereby competing LED manufacturers claiming better performance from their luminaires by ignoring MF altogether, using the argument that none of the factors making up Maintenance Factor can apply to a light source that has such an extended life-term.

After all, there are no lamps to change, so what’s the issue? It helps that technical decisions are being taken by finance departments, rather than Facilities Managers.

Environmental conditions now are so very different to the times when MF was introduced that it’s easy to see why some LED manufacturers imagine that no degradation factors need to be included into their illuminance calculations.

These manufacturers have convinced themselves that their products are 100 percent efficient and that’s how they present their calculations to clients. It’s not true, of course, and it works against those manufacturers who understand that Maintenance Factor is still very present in lighting scheme design.

LED manufacturers would have us believe that:

  • There is no lumen degradation with an LED
  • There is no degradation in driver performance
  • There is no degradation in room reflectances
  • There is no degradation in light output caused by dirt build-up on the luminaire

Again, none of this can be true.

LED lumen degradation is already enumerated in the L70/80/90 figures displayed in luminaire data sheets, though the figure rarely relates to the entire luminaire. Its usually just the LED source.

Although driver efficiency is usually glossed over, there is a fall-off in circuit performance over time. Rooms still get dirty, despite the cleaner air that we enjoy. LED luminaires, because of their heat sinks, will attract dirt particles. These particles will eventually create a ‘dust duvet’ around the heat sink. This will impact on the heat dissipation from the LED chip, impacting on light output (and LED life).

The problem that we have is that there is no standard method for evaluating these figures. As far as we know, manufacturers are not researching the impact of dust build-up on heat sinks and we know very little about the deterioration over time of LED drivers. All we know is that they will fail and we must assume that the ultimate fail comes after a period of deterioration.

What needs to happen next

We have to acknowledge that an MF of 1.0 cannot be justified. It’s not good enough to hide behind the relatively long life period of the LED. At some point L70/80/90 will be reached in even the most efficient of luminaires and there is a performance implication leading to that point.

Regarding driver performance, even good quality constant lumen output drivers will come up against the maximum available lumen output at some point in the luminaire’s service life. Degradation is delayed, not removed.

We also need to acknowledge that, if a scheme is calculated to deliver maintained illuminance during its lifetime, then it would need to take L70/80/90 into account. Taking an installation all the way to L70 means a 42 percent uplift during the initial design calculations, even if no other factors are taken into account.

And there is an ethical issue at play here as well. Back to the saw-tooth curve, where  it was acknowledged that maintenance involved replacement of light sources. LED manufacturers attempt to finesse this argument by ignoring the end-of-life scenario whereby entire luminaires often have to be replaced, well within the overall life term of a ‘conventional’ lighting installation. Its simply not good enough to stop the LED maintenance curve at 50,000 hours, or whenever is convenient, when the client may be looking for 100,000+ operating hours from the installation.

The final word….for now

It’s good to hear that SLL is actively looking into the issue of Maintenance Factor evaluation. It’s been a long time since CIE gave us CIE 097:2005; that was a whole lighting generation ago and the landscape is completely different, both in terms of the technology being employed and in the nature of the companies who are delivering it.

Credit: thanks are due to Richard Hayes of 42 Partners for the figures quoted above.