Voltage optimisers and lighting

Sintra · Oct 27, 2013

Done an install a couple of years back in a large building with three large sub switchrooms fed from a common main switchroom.
All lighting throughout the installation is either hf fluorescent or pl. As I say this was installed around 2 years ago with only the odd lamp failure as would be expected.
Recently lamps have been failing more or less daily and the client got us to relamp all fittings around 4 weeks ago which has been done.
On a daily basis there are numerous lamps failing so I called in today to investigate.
First thing I noticed was most of the lamps that are still lit have blackened ends so I tested the voltage at random fittings and the highest voltage I measured was 218V. Checked a couple of the local DB's and again voltage was around 218 - 219V L-N.
Got into the sub switchrooms and noticed that voltage optimisers have been installed that have been set to 380V L-L.
I'm reckoning that these voltage optimisers are causing the lamp failures. Any opinions welcomed.

Sintra · Oct 29, 2013

I have a few pictures that I will attach later but basically the output of the optimiser is at 383V L-L and the fitting ballast states 240V +/- 6% lol.

Marvo · Oct 29, 2013

Hmmm, maybe it is a voltage issue then. Where I am the standard voltage is 380/220, it's not uncommon to see only 210v at the load but most European items run fine even though it's at the lower end of their voltage tolerance.

With problems like this I'd be tempted to bounce it back to the customer and tell them to sort it out with the designers of the optimisation installation.

Sintra · Oct 29, 2013

Well here are the pics of this stupid box and ballast rating. Looking at the data plate it seems to be set to 5% 380V so turning it up a bit is not an option.

Silly Sausage · Oct 29, 2013

Immediate observation, 240 - 6% > 383 / sqrt 3.

Paul.M · Oct 29, 2013

Anything with a CE mark on it made from 1990 can be run on 220V. If the voltage is below this, it is automatically is going against manufacturers instructions!

Basically the Optimisers are a big dimmer switch, and they always work without any problems lol.

As said above, get the lighting circuits off the Optimisers. End of re-lamping every week. Good luck Sintra.

Engineer54 · Oct 29, 2013

Maybe i'm missing something here, but all that unit seems to be, is a straight transformer rather than a regulated voltage optimisation unit!! I'd want to see far better output regulation than what's being shown here. If the whole idea is to optimise voltage to 380/220 V then that's what i'd want to be seeing on the output.... The optimisation units that i've had a look at on the Web have a wide'ish tolerance of the input voltage. that can still output a regulated 380/220 V with a very small +/- tolerance.

As for the lower than normal/expected voltage (optimised voltage) at the lighting fittings, this could well be down to the cable size used and the length of circuits involved. Which is always going to be a factor when retrofitting a voltage optimisation unit into an existing installation.

Knobhead · Oct 30, 2013

I notice on the optimiser it states 3Ph 3Wire not four wire. Where’s the neutral coming from for the output?

If the neutral has been simply linked across from the incoming there will be all kinds of problems with voltage imbalance.

Engineer54 · Oct 30, 2013

Tony said:
I notice on the optimiser it states 3Ph 3Wire not four wire. Where’s the neutral coming from for the output?

If the neutral has been simply linked across from the incoming there will be all kinds of problems with voltage imbalance.

Yep i was going to mention that myself... but the post kept posting up. before i was finished writing!! As was also going to advise getting back to the manufacturers. Really need a schematic drawing and a working data sheet (cant think of the correct name?? lol!!)

Knobhead · Oct 30, 2013

An interesting read, I found the technical data sheet.

An unreferenced star output.
Tertiary delta winding for harmonic filtration.
From what I can make out Fixed input taps.

It’s just a Y∆Y transformer with no active voltage regulation I can see. Had it been a Y autotransformer then the star point would be referenced.

No wonder it’s causing problems. I would only use it for balanced 3Ph loads. I wouldn’t be keen on it then due to problems of L→N faults.

The problem with YY transformers is they are virtually transparent to harmonics hence the tertiary winding.

Knobhead · Oct 30, 2013

Just found the same company make a plus range with dynamic switching for on load tap changing.

Doesn’t look like this baby is the plus model.

Knobhead · Oct 30, 2013

From the schematics I’ve found I don’t think these are intended for Ph&N use.

To my surprise it is a Y autotransformer with a high impedance part of the winding reversed to form an artificial star point.

With the tertiary, nine windings in total.

Sintra · Oct 30, 2013

Paul.M said:
Anything with a CE mark on it made from 1990 can be run on 220V. If the voltage is below this, it is automatically is going against manufacturers instructions!

Basically the Optimisers are a big dimmer switch, and they always work without any problems lol.

As said above, get the lighting circuits off the Optimisers. End of re-lamping every week. Good luck Sintra.

Not easy to get the lighting off the optimisers lol. Optimisers are in switch rooms feeding directly into a 400A panel. Then from this panel the sub main circuits are taken to local dist boards.

Sintra · Oct 30, 2013

Engineer54 said:
As for the lower than normal/expected voltage (optimised voltage) at the lighting fittings, this could well be down to the cable size used and the length of circuits involved. Which is always going to be a factor when retrofitting a voltage optimisation unit into an existing installation.

I have to agree that I don't think any thought has gone into the installation of these "optimisers". All final circuits have been properly designed and installed and as I say no issues up until these where fitted.

Sintra · Oct 30, 2013

Tony said:
From the schematics I’ve found I don’t think these are intended for Ph&N use.

To my surprise it is a Y autotransformer with a high impedance part of the winding reversed to form an artificial star point.

With the tertiary, nine windings in total.

Thanks Tony for your input. At least I can go back to the client with something a little more concrete than my original thoughts.