help - so far outside the norm I need a sanity check!

[opentoideas]

My current project is so far outside the norm I am hoping that the more experienced here can have a look and see if what I want to do is acceptable as I am having a hard time applying the regs to the different aspects when combined as a whole.

While this is my own property not a customer I obviously want to do it within the regs and my head is spinning and I don’t want to miss anything as selective optimism is easy to indulge when you are looking to make it work.

To set the scene – I have a couple of workshops fed from a domestic single phase supply.

I also have a 50KVA 3-phase generator supplying machinery within the workshops as required.

At present the 2 systems are separate

Up to this point all is well, however, I had the “sensible “ idea that while the generator is running I might as well supply the workshops single phase requirements…………

I have a 3 phase transfer panel and while I know I can achieve the desired outcome  it is so unorthodox that I would appreciate a sanity check.

First off the changeover is delayed and has mechanical interlocks so mains and generated supplies remain isolated from each other in all circumstances.

The plan would be to have the single phase supply come in to a split load CU and from one side of the CU feed one socket with items that I don’t want interrupted (CCTV network) and the mains side of the generator changeover.

Generator supplies the other side of the changeover.

So normal operation would be mains single phase input fed to L1+L2+L3 on the changeover input

Output from the changeover would be :

L1 – workshop 1 (other side of the split load board)

L2 – workshop 2

L3 to heaters. (roughly balanced across the phases)

When the generator is running it will disconnect from the mains and supply these single phase loads.

Am I mad to even consider such stupidity?

If not is there anything in the regs to forbid this setup?

If there is still any glimmer of hope then any specific areas to pay particular care about?

I realise that this is an overly complex solution to a non existent problem but it has proven to be an interesting thought experiment so far to work around the technical difficulties and come up with a workable solution and if it is all for nothing then it was good fun none the less!

I have all of the components required so there is no cost involved  its just one of those brain farts that’s taken on a life of its own

 

[apprentice87]

What you need is FMEA...

"failure mode effect analysis" It is ok to come up with "novel" solutions, but you had better make sure it is "fail safe" in case it kills a DNO worker....

The others will be along shortly, but i would suspect that in a commercial enterprise especially, there will be a LOAD more things you need to comply with apart from BS7671.

Having said that, there are bit sof your description that i cannot understand what you mean..

The plan would be to have the single phase supply come in to a split load CU and from one side of the CU feed one socket with items that I don’t want interrupted (CCTV network) and the mains side of the generator changeover.

Generator supplies the other side of the changeover.

Do you mean you want to have some single phase stuff supplied from the mains, and other stuff, at the same time, for the generator?? How you going to assess and manage the voltage between the two?? and what about earthing arrangements??

Not being critical though, we like interesting things on this forum!! The others will have a much better idea than me though!!

john..

 

[opentoideas]

What you need is FMEA...

"failure mode effect analysis" It is ok to come up with "novel" solutions, but you had better make sure it is "fail safe" in case it kills a DNO worker....

this is exactly why I thought I would put this out there.

just because something is possible does not mean you should do it!

difficult to describe something like this without a diagram.

think of it as single phase coming into a CU and one circuit fed with the stuff I don't want to interrupt.

feed then goes to the changeover then feeding back to a second CU that provides the rest of the circuits in the workshop from either mains / generator.

no need to manage voltage as the circuits will always be isolated and there will be a delay between disconnection of mains and connection of generated supply.

isolated that is with the exception of earth. the generated supply is earthed to a spike which ?should? also be bonded to the MET

earthing is one of the more confusing aspects and I will admit this is one part I am struggling with as logic says bond everything but I am having difficulty as the generator side would seem reasonable to bond N to E however if i cross bond the supplies then I believe this should not be done.

I should mention that the supply is TNS for once it would be simpler if TT!

 

[opentoideas]

well I have just come across another oddity.

for the moment I am wiring up the generator side as a stand alone 3 phase supply and I have come across something interesting...

the generator came from a mobile seed dresser (stand alone piece of agricultural equipment) the set and its controls were put together by what looks like a reputable company that specialise in industrial generator supply so I was assuming all would be correct however as I  am replacing the engine control system and distribution I have found that there is no N-E link installed.

the generator output as supplied went to a 63A MCB then a 30mA RCD then  presumably then directly fed the machine possibly through another distribution board.

there are bonding links between the engine housing and control panel but no continuity between the N and bonded parts.

it is possible that a N-E link was installed downstream but to my mind this would be less than ideal and defeat the point of the fitted RCD.

everything I have found points to the system having  a N-E link installed as close to the star point as possible and I had expected to find this already in place however not finding it has made me question my logic.

as it stands there is no current path for a live to ground fault and given that the generator itself and presumably the machine it was attached to would all be extraneous conductive parts this just seems wrong.

how likely would it be that this was installed with a N-E link at the machine side rather than at supply?