IET logo
 
IET
Decrease font size
Increase font size
Topic Title: GENERATOR EARTHING
Topic Summary: Need for electrode?
Created On: 02 October 2017 02:09 PM
Status: Read Only
Linear : Threading : Single : Branch
 GENERATOR EARTHING   - John Peckham - 02 October 2017 02:09 PM  
 GENERATOR EARTHING   - OMS - 02 October 2017 04:02 PM  
 GENERATOR EARTHING   - AJJewsbury - 02 October 2017 04:48 PM  
 GENERATOR EARTHING   - John Peckham - 02 October 2017 04:52 PM  
 GENERATOR EARTHING   - aligarjon - 02 October 2017 05:18 PM  
 GENERATOR EARTHING   - OMS - 02 October 2017 05:09 PM  
 GENERATOR EARTHING   - davezawadi - 02 October 2017 06:08 PM  
 GENERATOR EARTHING   - mapj1 - 02 October 2017 07:39 PM  
 GENERATOR EARTHING   - John Peckham - 08 October 2017 12:21 PM  
 GENERATOR EARTHING   - mapj1 - 08 October 2017 02:21 PM  
 GENERATOR EARTHING   - AJJewsbury - 08 October 2017 02:32 PM  
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 02 October 2017 02:09 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 8783
Joined: 23 April 2005

I wonder if the brains on the forum could help me with this?

BS7430 requires generators rated at 10kW or more to have an electrode with an Ra of 20 ohms or less.

If the generator star point is connected to earth and neutral and these are distributed, a sort of TNS without the "T", what is the purpose of the electrode and why is this required for more than 10kW given the electrode is not in the earth fault path? I am thinking about site temps. and temporary installations.

My thoughts are.

1. Larger than 10kW is likely to have a larger installation connected to it and there will be capacative coupling to the general mass of earth, so the could be a potential difference between the supply earth and the general mass of earth.
2. If the generator has earth fault protection then a fault to the general mass of earth MAY (not certain) cause ADS where the over current protection may not operate.
3. For IR testing we we require the installation to be connected to earth so IR faults can be detected.
4. Regulation 8 of the EAWR probably requires a connection to the general mass of earth.

Any thoughts?

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 02 October 2017 04:02 PM
User is offline View Users Profile Print this message



OMS

Posts: 22359
Joined: 23 March 2004

You don't need an electrode if you intend operating in an "isolated supply" mode - but that's unlikely on a set exceeding 10kW

Basically you need to earth the system neutral to create the TN-S output - the neutral point of 3-phase windings should be connected to a reference point to prevent things like neutral inversion and to ensure protection will operate correctly on single-phase faults. Basically the idea of the neutral being at or near earth potential

The reference point should be electrically continuous with the generator enclosure, frame and load side cpc's and should be connected to true earth when reasonably practicable to do so - this could be via an impedance if you are deliberately aiming for a IT system with appropriate first fault monitoring (example would be a set located dockside providing power to a ship or boat to keep essential systems operating)

If you choose not to do so, then you really want your system wiring in fully metallic containment so it creates a short circuit condition - the RCD is then basically acting as insulation monitoring - but you still need to be mindful of neutral inversion. Short wiring or even the generator and say a pump on a common frame would be typical of this

You don't have to provide the electrode - but you would be a brave man not to do so based on all the engineering reasons given in guidance from various places

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 02 October 2017 04:48 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 15811
Joined: 13 August 2003

I think I share some of John's curiosity - most of the regs/legislation refer to "earth" but what they (usually) really mean is the protective conductors - it generally matters little to ADS if the protective conductors are floating or only incidentally connected to true earth, as long as PE and N are solidly connected. Much of section on mobile & transportable units seems to share that thinking.

It would be a worry if the protective conductors attained a voltage significantly different to true earth, with an significant available current (over 10mA say), and PE and true earth was simultaneously accessible - like with capacitive coupling from a HV/LV transformer. But I'm not sure where an equivalent danger would come from with a completely islanded generator system - bar a live conductor shorting to true earth while completely missing the protective conductors. To be honest, conventional (overcurrent based) ADS doesn't work particularly well in that situation anyway and any circuits liable to such faults (e.g. supplying garden cutting equipment) would normally be 30mA RCD protected, so should disconnect should any serious currents flow between PE and true earth in either direction.

The only situation when I can see that true earthing of a generator is a must is where some of the downstream installation has been (or might be) TT'd - which in these days of everyone TT'ing every little outbuilding and hot tub, is perhaps a thought not to be dismissed even for a small domestic stand-by generator.

20 Ohms does seem to be a rather arbitrary - too high to be terribly useful if ADS is by overcurrent protective device and faults to true earth are a worry and too low to be easily achievable in many situations. I think the best I can say for it is that it matches the DNO's limits for their transformer electrodes, so should there be any downstream TT installations, you can go ahead an connect those without any further worry (whereas if it got into the hundreds or thousands of Ohms RCDs might stop working, especially higher rated ones).

- Andy.
 02 October 2017 04:52 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 8783
Joined: 23 April 2005

OMS

I am think about a generator say supplying multiple sites at a pop concert or county show.

Agree the neutral and earth connected to the star point with both distributed but no earth electrode so no "T" so the earthing system would be "NS".

There would be some fortuitous earthing from the generator skid touching earth and any other class 1 parts touching earth.

You say the neutral needs to be at earth potential is that with respect to the distributed earth or the general mass of earth?

I am not thinking of earthing impedances more a BFO generator supplying the stands at the Borchester county show for 2 days via HO7RNF cables installation method "laid on the wet grass".

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 02 October 2017 05:18 PM
User is offline View Users Profile Print this message



aligarjon

Posts: 3850
Joined: 09 September 2005

Does it matter where the spikes are on an installation like this. If they can be fitted mid installation, where ever is convenient so to speak it,should be quite easy to achieve 20 0hms or lower in a wet field with a couple of spikes. The only stipulation is the earth cable size which is 1.5mm.

Gary

-------------------------
Specialised Subject. The Bleedin Obvious. John Cleese
 02 October 2017 05:09 PM
User is offline View Users Profile Print this message



OMS

Posts: 22359
Joined: 23 March 2004

OK - start the set up and wander around with a screwdriver - stick in in the grass at random locations and measure from it to the system earth at the distribution boxes (or neutral) - what voltage do you expect to see ?

What would this voltage be under a fault condition

Why do the DNO earth the system neutral at all on the LV side of transformers - what's different to a generator ?

As I said, the important thing is the reference between N and E (the system CPC's) - but you also need to be mindful of the differences between the system "earth" and "true earth" - it's reasonably easy to put the reference at true earth by means of an electrode

That electrode could be the earthing point of another system incidentally

Given that you may well be reliant on a certain amount of RCD protection for ADS on the supplied system, and you cannot easily guarantee the integrity of insulation on that system, then it seems to me that you really need that reference to be at true earth potential (akin to a DNO transformer)

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 02 October 2017 06:08 PM
User is offline View Users Profile Print this message



davezawadi

Posts: 3843
Joined: 26 June 2002

The 20 Ohms question is almost certainly that it was copied from the DNO handbook. Depending on the site then 20 Ohms might be OK, but in some places it is not achievable without major earthworks, which is not really an option. An electrode suitable for the installation as TT , say 100 Ohms seems a more reasonable value, and along with all the bonded kit can usually be got without too much trouble. Clearly there needs then to be treatment of the installation as fully TT (presumably you are referring to temporary, ie BS7909 land).

If it is a fixed install then the 20 Ohms is probably not too onerous and you can operate as TN-S in the conventional manner, but with the way BS7671 is going there is little difference between this an TT! (Due to RCD protection everywhere).

Faults between an installation and real Earth have never really worked with ADS, but keeping the potential of the N somewhere near the ground cannot be a bad thing.

-------------------------
David
BSc CEng MIET
david@ZawadiSoundAndLighting.co.uk
 02 October 2017 07:39 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9346
Joined: 22 July 2004

OK, an example of a real fault at one of those "snake in the wet grass" type events. A genset was powering a distribution cable to a large tent, and the cable was damaged during dragging through a hedge, and unknowingly the jacket and insulation to a LIVE core got torn, exposing live copper to view over a few inches. Not nice but not noticed until a pool of muddy water formed around the cable, as it rained.
The impedance from live to terra-firma was presumably lower or comparable to the impedance form CPC 'earth' to terra-firma 'earth' and so all metalwork at CPC potential then rose to a significant fraction of 230V. Result, reports of tingles, and a nasty belt for someone fixing something under a portakbin toilet block.
The fix is a rod of almost any impedance, and an RCD of a few hundred mA or less at the very origin, as then such imbalances are automagically disconnected.
A similar fault modes arises with gensets on trailer or trucks - no rods needed, until you take power off to a remote place by extension lead. Otherwise you can get a truck that shocks everyone who tries to climb in.

Equally electrodes are not without risks. I have been in a case where rods where prohibited as the threat from buried ordnance was considered more serious than the electrical one. We parked one of the vehicles on a large bit of metal in damp patch and used that as a compromise...

-------------------------
regards Mike
 08 October 2017 12:21 PM
User is offline View Users Profile Print this message



John Peckham

Posts: 8783
Joined: 23 April 2005

Sorry for my late reply password problems getting on to the site.

I understand the point of referencing the generator to earth using an electrode and, as OMS says, the RCD acting as an insulation monitoring device.

OMS also intimates that an electrode is not needed if the generator set and installation are operated in an island mode.

Mike

Your example of the snake in the grass event where a damaged sheath causes a live conductor to come in to contact with the ground and people receiving tingles and one person receiving a nasty belt, was this an actual event and did the generator have no electrode and/or RCD protection?

Dave Z

Yes I think the 20 ohms in BS 7430 was possible a carry over from transformer earthing.

In respect of the comments on the value of earth resistance could be any resistance would it be better to have an Ra not exceeding the value stated in Table 41.5 for the highest rated RCD on the installation, subject to a maximum of 200 ohms as in Note 2 of the Table for all final circuits? At the moment BS7671 551.4.4.2 requires every circuit (distribution circuits) to be 30mA RCD protected. That is not practical for a large installation but it only applies to TN, TT or IT systems but if the generator did not have an electrode (an INS system) that regulation would not apply? How would you provide for earth fault protection on your distribution circuits?

How risky would it be to operate an outdoor installation from a generator >10kVA without an earth rod?

-------------------------
John Peckham

http://www.astutetechnicalservices.co.uk/
 08 October 2017 02:21 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9346
Joined: 22 July 2004

Sadly a real toilet block, a real genset, a real puddle, a real shock. A real embarrassment too and quite a bit of nervous clucking about how lucky it had not been more serious and choice of cable types and protection for it before lots of things changed by the next such event.

Took place in 2004 summer time, so before quite so many RCDs as you would expect to see today, and as these were hired in Gensets that had the provision to be operated in parallel, the Earth fault relays were set either to off or at least really wide, so the did not see each others NE links as a fault, so in effect no RCD action at origin. Tow in genset, home brew cables of unknown vintage in many cases, and not properly staked, big puddle along fence by toilet block, fault was better grounded than the generator. As I said, quite a bit was changed and the process better formalised ready for the next time.

-------------------------
regards Mike
 08 October 2017 02:32 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 15811
Joined: 13 August 2003

, fault was better grounded than the generator

So in a way might it have been better if the system hadn't been deliberately earthed at all? - in effect a separated system - first fault would then merely "degrade" the system to something a bit like TN-S.
- Andy.
Statistics

New here?


See Also:



FuseTalk Standard Edition v3.2 - © 1999-2017 FuseTalk Inc. All rights reserved.

..