The phone rings. It’s Richard Hayes: I’d like to start a conversation about the way lighting designs are specified.
.
In my world engineering specifications have tolerances. If I order a piece of metal 1000mm long depending on how tightly I specify the tolerance I will pay a different amount. 1000 mm +/- 1% (+/-10mm for those who can’t do the math) is a very different beast from 1000 mm +0.5/-0mm
Bricklayers have a tolerance on out-of-vertical for a wall.
The whole construction industry is based on a set of commonly agreed tolerances.
Parts of our industry are really good at this. LED, and before them lamp, manufacturers will give us a tolerance on Lumen output and Colour. Does anyone believe all 200 mA constant current drivers actually deliver exactly 200mA ?
So we know that that downstream to our lighting design we have tolerances in the supply chain.
Yet, when it comes to a lighting design specification we do not seem to allow any tolerance.
If, as designers we are ever given reflection factors (which, let’s face it, is rare) do we seriously expect them to be exactly as specified ?
But we know that variations in assumed or actual Reflection factors have a major effect on the design.
I contacted the SLL: what is an acceptable result for an illuminance specification in lighting design. The answer was X Lux +10%/ -0. I am not in any way convinced that this is a realistic position. I’ve been in this industry for 40 years and I have never seen this written down anywhere. Why don’t specifiers provide designers limits of acceptability of their designs? This would at least allow designers to evaluate jobs properly and have a sound contractual basis for client acceptance of the design.
Every street lighting software I have seen assumes the lanterns are mounted exactly level, at exactly the correct tilt angle and exactly the correct height and have the 90/270 axis pointing exactly across the road. I once proved to the Highways agency that a +/- 2 degree out of level on Luminaires in a scheme totally ruined the U 0 and U l figures, this was why the resultant installation looked like a Zebra’s backside.
And don’t get me started on RUG (UGR in case you haven’t caught up with the latest arbitrary nomenclature change).The series of Rug specifications is a preferred series 16,19, 21…… etc. They define steps of discernable difference in disability glare. In other words you can just about tell the difference between RUG 19 and RUG 21. So why do we report Rug numbers to 1 decimal place? Why do we reject Rug specified designs requiring RUG 19 when the extrapolation from a table that bears no resemblance to the design reports a result of RUG 19.1. This is Madness.
Photometric data from tests usually the result of testing one (1 ???) sample and we base the whole design on that data. Sure labs if pushed will give an uncertainty figure for the data, does anyone ever take any notice of this ?
The industry needs to sit down, preferably with a beer and some snacks, and try to agree on a unified approach. If there are not too many of us, I may even offer to buy the first round.