Emerdency supply?

[SUNRAY]

New one on me.

I locked of a 3p 300A supply to a dis board at main switch room. Checked the dis board for dead and started strippng old circuits and breakers out. Cut a T&E and tripped the 6A MCB.

Turns out one of the single phase outgoing circuits feeds switch gear which can choose between 2 feeds from different dis boards, due to a fault it was trying to choose both feeds and hence back feeding to the board I was working on.

No mention of the system on the board or by the client in the site induction.

How far does one go with checking for dead?

 

[holmslaw]

..

 

[ban-all-sheds]

Checked the dis board for dead

But not at the outgoing terminals of each MCB....

Cut a T&E and tripped the 6A MCB.

Presumably in the other board?

Could have been worse...

No mention of the system on the board or by the client in the site induction.

Arguably some kind of EAWR offence going on there?

How far does one go with checking for dead?

Can you ever go too far?

 

[ericmark]

The same sort of thing happened to me. Stripping out redundant cables from old batching plant having been told they were all dead. I of course did not rely on this and tested cable and it was dead. Cut it and dropped to floor then it started sparking.

Seems the cable had been re-used and powered a seldom used motor and until that augur was used it was dead.

After that it was agreed cables would only be stripped while plant was not being used. Lucky no one injured that time.

The worry was we had not omitted any tests. Even with the borrowed neutral you can use a clamp on ammeter but there is no real safe way to work on complex systems without detailed up to date plans. Which is of course why the 17th Edition requires them.

I have questioned a few times the lack of mechanical interlock. Years ago all change over contactors (Star delta starters) had mechanical interlocks as well as electrical but today it seems the designers rely on fuses to open circuit in case of any contact sticking. However using the same contactors for other uses does present a problem in that a sticking contact may not be detected.

Sorry to say in maintenance I have seen many times items being replaced for similar units without the care that should be exercised. The typical one is replacing standard contactors for semi-conductor types without considering the emergency stop control.

And however careful one is these type of mistakes will be made and accidents will happen. Sorry to say but fact and I have never worked since mine. Where a lock was removed and machine switched back on.

 

[SUNRAY]

Hi Guys

I checked for volts between all phases, Neutral and earth, ie 10 tests, with the boards local isolator on and all MCB's on then switched off and locked the local isolator and rechecked the board for volts, all tests showed dead.
I believe the backfeed did not occur until after I performed the tests.


I agree the change over switchgear should not have been able to back feed but these things happen, I sincerely hope it was due to a fault in the switchgear which has not yet been accessed to inspect, hope to do that tomorrow.

I didn't mention in the initial post that the backfeed occurred on a 32A circuit, the 6A that tripped was in the board I was working on. At that point I checked for volts again with the local isolator still off and found 230ish on the board. flicking off the MCB's soon identified the live circuit.

No you can never go too far testing for dead, historically I would have always shorted all 3 pases & N & E together when working on something like this but I confess I have not done that for a while now.

 

[SUNRAY]

Got in yesterday to investigate back feed.

The machine was started more than an hour after I checked for dead. It automatically selected the good supply then. The fault was simple, one pole of the contactor on the feed from the panel I was working on was stuck, a thump stopped it and its now been changed.

A note for ericmark, I still use mechanical interlocks and in fact they are usually supplied in a star delta kit, all you have to do is throw away the pneumatic timer which I find often don't survive the first year. I find that many panel builders work on a 'painting by numbers' process and either don't know about them, cant be bothered to fit them, or worse, theres not enough space for them.

 

[Spark123]

Even still the contactors should have been wired so that electrically one cannot make unless the other one is physically out (i.e. not stuck in) by using an interlock via an opposing contactors N/C contact.

A note for ericmark, I still use mechanical interlocks and in fact they are usually supplied in a star delta kit,

I also see them a lot in forward/reverse DOL gear.

 

[bernardgreen]

I hope with that circuit arrangement the type of contactor was also changed to one where stuck contact would not be possible. Forced break instead of spring return to break where the actuator if jammed by a welded contact cannot move far enough to operate an auxillary contact in the coil for the contactor connecting load to the other supply.

beaten to it

 

[SUNRAY]

Even still the contactors should have been wired so that electrically one cannot make unless the other one is physically out (i.e. not stuck in) by using an interlock via an opposing contactors N/C contact.

A note for ericmark, I still use mechanical interlocks and in fact they are usually supplied in a star delta kit,

I also see them a lot in forward/reverse DOL gear.

It does have, in the form of an auxiliary contact but not mechanical as they are in different enclosures.

The first forward/reverse I had to work on was star/delta, it needed replacing as it had caught fire and was badly damaged that took a bit of working out.

 

[Spark123]

I thought the physical makeup of the contactor should be break before make, hence if the contactor was stuck in the N/C contact should remain open interlocking the other contactor out.
If this was true then there is no way the power should have been able to feed back through to the other supply.

 

[bernardgreen]

I thought the physical makeup of the contactor should be break before make,.

They normally are. But most have individual return to normal springs. one spring for each contact. A single insulated bar moves all contacts from NC to NO when the coil is energised. When the coil is de-energised the bar moves back by its own spring but does not pull the contacts from NO to NC. Each contact depends on it own return spring. If one or more are welded in the NO position the bar can still move. With forced break there are two bars one pushing the contacts to the NC position, either this break the welded contact apart or if the weld is not broken then the bar cannot move and operate the auxilary contacts.

 

[SUNRAY]

I thought the physical makeup of the contactor should be break before make, hence if the contactor was stuck in the N/C contact should remain open interlocking the other contactor out.
If this was true then there is no way the power should have been able to feed back through to the other supply.

Totally agree. The contactor had developed a fault, I didnt bother opening it to investigate further, just chucked it in the bin.

and replaced it with an identical contactor as issued by the on site storeman.

 

[Spark123]

When we have had contactors stuck slightly in before on say West motion on an EOTC, it stops the east motion electrically. Maybe we just use a different brand of contactor.
The only thing with just scopping it in the bin and replacing it with an identical type is that there may be nothing stopping it happening again. I take it your company carried out a full investigation into the dangerous occurance?

 

[viewer]

The problem of a 'second' supply is quite complex. There is no easy answer. Yes, proper drawings etc - but never trust them! Mechanical elements e.g. contactors and RCCBs etc can fail to unsafe. Fatalities have happened when all tests are fine and there is a feed through something/circuit which was not working during testing - I'm thinking of a caravan site where the second (unknown to the electrician) supply was through a photocell. I've had the same situation occur, but was lucky as not touching anything when the cloud covered the sun. I can only suggest linking all with croc clips, but this is not always feasible. There are lots of comments on part P and 17th ed etc, but the truth is that a mistake made now can be invisible to most routine tests and only show up 20 years later.

 

[Spark123]

Photocells, PIRs etc are a bit of a nuisance, as is someones 16 year old darling turning the power back on so his Play station will work

For the photocells/PIRs it is down to provision of information, similar with the likes of motor control panels etc.