Fluorescent Lighting Issues

[ryanm]

Hi all,

I have recently started a new job (recently qualified maintenance tech), and we have had a thermographic survey done.

There were a few issues like loose connections on MCB's, which mean't the current was too high, and hence too hot too. These were easily solved, but there is one issue I'm not too sure about.

The office lighting MCB is too hot (and actually buzzes). It's a 20A MCB, with a 300mm length of 2.5mm T+E from the MCB, but then it reduces down to 1.5mm T+E (crimpped) then it's an approx 20m run to the light switch, then on to the first light fitting. From then on all the light fittings are linked with 2.5mm.

The lights consist of 16 banks, each with 4 4ft fluorescent tubes (36w) in, so a total of 2.3kW.

Now my first course of action is to replace the 1.5mm cable with 2.5mm, but what else can be done? I've checked the current at the the MCB and it is 19.5A, which is why I'm guessing the MCB is getting so hot. Will the replacement of the 1.5mm cable lower the current being pulled (due to voltage drop etc)?

I've suggested splitting the load in to two sets of lights, just incase the MCB ever tripped, then at least there would be another MCB/set of lights still on. But as this will cost the company they aren't willing to spend that much as they are re-doing the offices anyway soon (but I've been told the 'soon' has already been years)

Any advice would be great.

 

[Zeespark]

Do you have any idea of the length of cable run at all Ryan?

AndyGuinness

 

[S60TEM]

2.3 kw will only give you a running current of 10 amps but if the MCB is buzzing I would just first of all replace it for a new one aswell as replacing the length of 1.5 and remove the crimped joint. Flourescants have start up currents to think of also but if this hasn't been a problem in the past then I wouldn't worry about it at this moment in time. Unfortunately your governed by the 2.5mm cable and cannot fit a bigger MCB than the 20 amp for CCC of the cable but as you've suggested, I would advise splitting the circuit into two banks of 8 for start up current issues and to lessen user inconvenience in the event of a fault.

 

[kme]

Agree with the calcs above - your running current for those lights is 10A - possibly 18 at startup (if they were all brought on together). You deffo need to remove the 1.5 and replace the MCB; but where is the other 9.5A coming from? Something not right here.

Given the amount of lighting; I`d be surprised if you have an acceptable Zs at the extremity of the circuit.

If these lights are in a suspended ceiling; the cost for you to split the circuit into banks shouldn`t be excessive - IF the Zs is too high, you need to do something about it.

First thing is to find the reason for the excessive current.

KME

 

[ryanm]

Thank you for the replies.

Zeespark, the run from the MCB to switch is approx 20m, then from that it goes to the first light, then they are all linked together, until you get to the final one, which goes no-where. I'd estimate the total length of the linking cable is around the 30m mark, so a total cable lenth of 50m!

S60TEM, the MCB does look a lot newer than the others, but I will replace this as a matter of course.

Start up has never been an issue as far as I'm away. It is a C-Type MCB.

I'm puzzled where the current is coming from (I'm 99% sure it is only the lighting circuit in the office, as I have switched it off while everything else was on). As like you said, it should only be around 10A. Could it be to total length of the cable?

Thanks a lot

---------- Post Auto-Merged at 12:41 ---------- Previous post was made at 12:39 ----------

Thanks kme, I am going to double check the fittings as I think they are 36w tubes, but need to be sure.

 

[kme]

The cable length can`t "add" current to the load - not at the level your describing.

lets say 20M of 2.5, at 19.5A ( it drops as it travels through each light)

so 30M of cable has 19.5A to 0.5A down its length - so if we calc at 10A for approximation.

Drop is 18mV/A/M

20 x 18 x 19.5 = 7.02V

30 x 18 x 10 = 5.4V

total V drop of approx 12.4V

R1+R2 for 2.5/1.5 is 19.51mohms / metre

50 x 19.51 = 0.9755 ohms

A 20A "C" curve max Zs is 0.93 ohms

So even if your DB was at the origin, with a Ze of 0.0 ohms, you still fail to meet the disconnection time.

KME

 

[Graham Lister]

Dont think the running current of a fluorescent is just the lamp rating is it? Are these HF ballasts or choke and starter? Kme's calcs show circuit should split up as best option really.

 

[ProDave]

Apart from all the above, nobody has stated the obvious answer to your question.

Replacing the 1.5 with 2.5 WILL NOT reduce the current.

How can it? Why would you think it would?

If anything, it will marginally reduce the resistance of the cable run, and very probably slightly INCREASE the current.

As others say, you need to do more measurements as you seem to be measuring more current than there should be, and the circuit needs a re design as already detailed.

 

[Graham Lister]

If they are old choke type then there is current drawn by the ballast to consider, power factor and the fact they are a reactive load not resistive. So a simple adding up of the lamp wattage and dividing by 230V will not give the correct answer for the current actually being used to run the lights, the actual current drawn will be higher than the figure calculated by just adding up the lamp wattages.

 

[kme]

Agreed mate; but not by a factor of 2 - old florries were fairly inefficient, but I don`t think even they were 50%eff; and these are not ancient fittings, I would presume.

Dave - no it won`t; but the 1.5 really shouldn`t be there anyway.

 

[Sidewinder]

One other thing in the OP that puzzles me is how the loose connections at the mcb's were causing excess current too?

 

[Evans Electric]

Couple of things, first shove a clamp meter on to get shut of all this guessing .

The ambient temp in the office will make a difference to the MCB , (beware of them in Boiler Rooms) along with the fact that the MCBs each side will be warm also.

If its a suspended ceiling you could divide that lot into two cicuits in half a day , claim 'elf 'N Safety. I posted a similar problem on here with fluorescents tripping the 16A breaker in the summer , ambient temp was affecting the thermal part of the tripping mechanism . Ended up with two Hager HRC fuse units instead of breakers . The ones that replace an MCB.

 

[SLIPSHOD & SLAPDASH]

I am pretty sure fluorescent light fittings ( or any lights) should not be backed up by a 20A breaker. Its nothing to do with the loading. It's more than likely mentioned in the manufacturers instructions, and also in BS7671 but don't ask me for a reg no.

 

[Graham Lister]

I am pretty sure fluorescent light fittings ( or any lights) should not be backed up by a 20A breaker. Its nothing to do with the loading. It's more than likely mentioned in the manufacturers instructions, and also in BS7671 but don't ask me for a reg no.

Well that is a lot of commercial installations needing a new breakers then! Many fittings have an inline fuse fitted, I do agree some dont and should probably be on a lower rated breaker.

A lot of the HF ballasts that I see failing, tridonics mainly, dont half go with a hell of a bang, they dont trip the breaker but blow up a fused track on the PCB. If an inline fuse is fitted in the light fitting of 3A then this blows instead of the fused track when the ballast reaches the end of its life and if replaced without changing the ballast goes with a hell of a flash when power is reconnected back to the fitting. Still not sure what goes in the ballast to make them fail but at 10 years they all seem to start failing. Sorry for waffling on!

 

[kme]

It also raises the question - is the switch rated at 20A??

 

[Evans Electric]

Tridonic ballasts ......yes get a lot blowing themselves to bits .

 

[ryanm]

One other thing in the OP that puzzles me is how the loose connections at the mcb's were causing excess current too?

Sorry, I mean't excess/greater resistance.....so the current would actually reduce wouldn't it.

---------- Post Auto-Merged at 11:15 ---------- Previous post was made at 11:10 ----------

Apart from all the above, nobody has stated the obvious answer to your question.Replacing the 1.5 with 2.5 WILL NOT reduce the current.

How can it? Why would you think it would?

If anything, it will marginally reduce the resistance of the cable run, and very probably slightly INCREASE the current.

As others say, you need to do more measurements as you seem to be measuring more current than there should be, and the circuit needs a re design as already detailed.

Hi,

I've realised I was wrong to say that. I was half thinking that the voltage drop wouldn't be as great in 2.5mm cable, so the current wouldn't be as great? I understood as the voltage dropped in a cable run, the current increased. But obviously it can't thinking about it more V=IR proves this.

---------- Post Auto-Merged at 11:29 ---------- Previous post was made at 11:15 ----------

Thanks a lot to everyone for the replies too! And not ripping me apart too much for not having the knowledge I know I should, but I am still learning

The current at the 20A MCB is 19.5A. I replaced the MCB yesterday too, and nothing has changed, there is still a worrying buzz from it.

I'm not sure what the switch is rated too if I'm honest.....it's old, with no markings on. It's metal clad though.

There are definitely 64 x 36W T8 tubes. Just a thought, would replacing the T8 tubes with T5 (with adaptor) reduce the current enough to be safe?

I noticed yesterday there are two emergency lights connected to the system too. A twin flood-light type, and a small fluorescent one. I will disconnect these to see if they are faulty (very old too), as they could be pulling too much current?!

Other than that, is the only REAL solution to split the load?

Thanks again guys

 

[kme]

YES.

The switch sounds like a 10AX type - so if that is being asked to switch this load, it could be actually switching a load 300%+ higher than its rating (19.5 x 1.8)

Apart from all of the above, a look at the calcs on the last page (which ignore the 1.5) have the circuit max Zs as too high - it would be interesting to find out, by measurement, what "ZsDB" and cct. Zs are; we could see how far out the calcs are in reality.

Further, (its too late now, but...) I personally wouldn`t have de-energised the circuit unless they would let me change it, for the simple reason that YOU have now energised a non-compliant circuit. What makes it slightly worse is that you KNOW it doesn`t comply, but you re-energised it anyway. Had the circuit been turned off, for ANY reason, I`d have refused to switch it back on, due to the concerns raised previously in this thread.

So, to clarify your last question - You need to do SOMETHING. you`ll struggle to get 4.0mm or 6.0mm terminated at the fittings, so the only sensible move is circuit segregation. As previously said, if this is a suspended ceiling, it shouldn`t be too hard.

n.b. replace the m/c switch with a 20AX; or fire the lighting through a contactor.

KME