I seem to remember the "Optimisation Unit"
claimed to save energy, by reducing the voltage by 7.5%. There was also a version built into a consumer unit. In the old days of florescent lamps with wire wound ballasts a slight over voltage resulted in a massive over current, and a slight under voltage stopped the lights striking, so the idea was to stop the over voltage so save energy. However it only really helped with the florescent lamp, and once the electronic HF ballast came out, it was no longer required.
When the voltage was changed from 240 volt +/- 5% to 230 volt + 5% - 10% in real terms the voltage did not change, called harmonisation we dropped 10 volt and rest of EU went up 10 volt, so now all at 230 volt nominal voltage. But the increased tolerance resulted in no need to change anything. However a 220 volt fluorescent uses well over rated watts when used with 240 volt, and a 240 volt fluorescent is going to have problems striking at 220 volt, hardly any other item was affected.
With the arrival of solar panels and EV charging points the voltage became critical, to ensure they fail safe if either whole supply lost or PEN lost they have a voltage window of 207 to 253 volt, if the voltage goes out of that window they close down, so the DNO was forced to go around actually dropping the voltage tapping to ensure voltage stayed within the range.
I had an old 65 watt fat fluorescent fitting, and the 65 watt tube supply had dried up, but the 58 watt worked OK, slightly shorter life, but within acceptable limits, however when my volts dropped from 245 which it had been for years to 230 volt, that was the end of being able to use 58 watt tubes, I was forced to move to LED tubes.
So our supply has changed, and also we now have items likely to warp the wave form, and cause spikes. Since I don't have an oscilloscope I can't check for spikes and warped wave form, however in my old house with just two RCD's I noted they would have bouts of tripping, would hold for 2 years then trip 6 times in a fortnight and then another 2 years without tripping.
I could find no fault, and assume spikes on the supply, the same spikes could take out LED lights, and since not been around that long, if fitted in the 2 year spike free time they would seem OK, but fit in the fortnight of spikes and they could fail, and it is unlikely any tests would show the reason.
In 10 years time we may have located the problems and done some thing to stop it, but LED lights are relativity new, we have not had them long enough to work out problems.
In the old days most electronic equipment started with a step down transformer, then came the front loading video machine, and the switched mode power supply designed to emit less heat so the unit could go under the TV without over heating, since then the switched mode power supply has become standard in all sorts from freezers and washing machines to LED bulbs. Which means electronic chips with no transformer to remove spikes.
So we have the SPD, which it seems no one quite knows what to do with them, one minute we are not fitting a MCB to supply them as the resistance of the MCB will reduce their effectiveness, next we have a MCB. Some are twin module units, and some single, and it is all very well talking about risk assessments, but electricians like me at 70 year old are baffled as to how to do that risk assessment, in real terms we fit them as we simply don't know if required or not, but think better to fit and be on the safe side.
Then we wonder why the 500 volt insulation tester is showing fail? Be it RCD testing or insulation testing the whole procedure has been turned on its head. We were always told test a RCD with nothing connected, then we are told DC can stop them working so need testing as well under load so if DC is freezing them we will know. And in spite of having type AC, A, F, and C on the RCD tester we are told to comply only test as type AC?
In real terms we all need to return to school and re-train, but not yet as the collage and University lecturers need to work it all out first.
Yes not directly related to LED life, but the whole system is interacting with each other, and even as electricians we are surmising (guessing) what is going on.
So back to the SPD, are these the reason why some people do not have issues with LED life and others do? But if no spikes in the first place, then SPD will not help, so a report "I have not got a SPD and my bulbs last and last." means nothing.
Personally I have a SPD fitted, but was only in this house for 6 months without one, so can't really say reason for few LED failures is the SPD. It could also be due to having electronic light switches, or still having one or two tungsten around.
For a detailed explanation see here
