A 250Kva three phase 415vgenerator is installed in a small hotel complex and offers an alternative supply to the power company during peak tariff periods.
The generator neutral point is grounded via the power company earth, which is contrary to the requirement that the gen-set should have its own independent electrode with an earth electrode resistance not exceeding 10ohms. The Power Company earth is TN-C-S However, the installation is protected with rcds. These operate arbitrarily on change-over from mains to genny. The change over is automatic and on load but there is a small delay. I thought that perhaps this would lead to voltage transients that may cause nuisance tripping of the rcds. The site electrician grounded the frame of the gen-set to a temporary earth electrode. This seems to have solved the problem. I would appreciate your comments on the possible causes that tripped the rcds.

Lyle,
This is an interesting subject actually and I say thanks for the information that you submitted with your post.
I must say that I'm a little rusty when it comes to Generator Installations.
I'm making a wild assumption in asking if the Generator is a Star (Wye) connected type?.
What I can gather, Lyle, this sounds like a problem with current beng introduced by the Star-point of the Generator, back into the Installation Earthing system.
I could be wildly off the mark here, but I would never fail to drive a seperate Electrode for a Generator, regardless of size.

Hmmm... Interesting problem. I'm trying to sketch out the arrangement and figure what might have caused it.

I'm assuming that the RCD(s) are located after the transfer switch (i.e. that you don't have separate RCDs on each supply before it).

How was the gen-set frame grounded before? Was it just bonded to the neutral/PME earth? If so, then now that the temporary rod has been added, is that frame-to-neutral bond still in place?

Just a slightly off-topic question Lyle.
How is the Generator connected into the Installation?.

Was the genny connected in through a double pole contactor? If so then the neutral would have been disconnected too. I think the cause would be static building up on the genny frame, which then runs to earth through the first pole to connect when the contactor goes over. If there are any computers in the system, their power supply filters will provide the earth route.

Ughh! I meant quad pole. Too late at night, must go to bed...

Hello strangers.. been movin house and all that..

With regard to the genny problem, as far as i was aware you should have no connection of the earth of the building to the genny earth,, once the genny is running shouldnt you be rely on the genny earth??? Maybe you should convert the building to a TT system??

With regard to the trips triping, maybe you are catching them mid cycle when the genny switches over..??

always a difficult problem gennys, not sure anyone is sure what to do

Uhuh LD; come across this little number before. Finished up with splinters from scratching my head too much!

The gen-set must be earthed, and so too an independent earth provided for the installation. Any earth between the genny and the installation can be removed if you get nuisance-tripping provided that the main earth terminal is securely grounded. I have even seen the PME link removed in standby mode - but this shouldn't be required.

The RCD will be downstream of the change-over switch, and nuisance-tripping will also occur from on-load switching transients causing an imbalance in the active-neutral, combined with parallel paths through earthing arrangements. The intermittent nature may be due to this only occuring when certain loads are 'online' at the time the change-over occurs.

If the genset and installation are electrically seperate ( from the earthing point of view ) you should have no problem as the ECL exists only 'within' the installation 'circle'.

Sorry guys, been away on a course. It was good of you all to respond.
Ok, between myself and the site spark we have established that take the genny spike out, rcds trip - not always but often. Put the spike back (and its actually just pushed into soft earth adjacent the generator housing)problem resolved. RCDs ARE ELECTRONIC and are of course,placed downstream of the 4-pole change-over switch. Trumpy, Paul; genny is star connected. Neutral connected to frame and thence to installation earth. Paul Cornwall; I cant see how the two earths would be separated lest a link was fitted. I think this arrangement is only used where harmonics are of a concern on parallel sets. UK Sparky, I need to think about what you are saying before further comment.
I do appreciate all your comments.

Uk sparky, just as a foot note. I am having a little difficulty thinking about removing the link from the met to the genny. This would require a new electrode at the installation met thereby converting system to TT. The system currently is TN-C-S on poco supply and TN-S on genny. The genny earth is to reference neutral with earth to stop the genny neutral floating. Fortuitous contact with earth via the met is certain as this is a steel frame building. The genny earth could really be described as superfluous in this case. However, I do recognise the point that when the gen-set is independently spiked there seems to be a cure to rcd tripping so obviously something is afoot! I have drawn the circuit arrangements out and I can see no advantage in separating the gen-set earth from the met.
The reason that I highlighted the fact that these rcds are electronic with loss of neutral protection is that it has been well reported that upstream faults can upset them.

I can see no advantage to separating the gen-set earth from the building earth either, in fact I can see several problems with such an arrangement.

I'm thinking along the lines of switching transients as that 4-pole transfer switch is operated. I've run across RCD-tripping problems on just a single-phase system where there was a little earth leakage (from heating elements) and a double-pole isolator had the line side closing a fraction of second before the neutral.

The varying loads on the system at different times could account for why it trips on some changeovers and not others. It could also be dependent upon the point in the A.C. cycle at which the contacts close.

Let this Yank take a wild guess or two...

Assuming that your rcd's trip at a low level, say 30ma, it mould resemble a problem I had with a UPS system once; in that case, there was a connection between the utility neutral and the back-up system neutral that shouldn't have been there.

Paul, you may also be on to something...if the generator is diesel, and has a heater in the fuel tank, and that continues to be powered by the utility while the generator is operating, this sort of problem might come up.

Failing these ideas, are we sure that the transfer switch is operating properly- or is the system getting a little crossover power for an instant during the transfer?

Just some thoughts.

Hmm this one could be a dinger -

you say that you sell power back to the grid so you will have some form of syncronisation working in there, this could set up some form of imbalance on the lines to trip the RCD.

How ever Im thinking on the lines of the genset neutral earthing if it has this fitted - this will cause some problems with the rcd's worthy of some investigations....

Rugged

Ruggedscotty,
Now there is a username that just screams of hard work!.
Welcome to ECN, mate.
Great to have you along.

Yet another Scotty for the wallchart!
Alan

thanks for the welcome - Im looking forward to contributing more.....

If you want to ask me anything feel free - Im based in scotland work for a large company here and Im in the building management side of things.....

Rugged

Welcome aboard Rugged. ECN has quite a few members from north of the border now. We sassenachs had better watch out!

Alan,
You've got a wall chart?.
That's actually quite funny, in a creepy stalker sort of a way!.