IET logo
 
IET
Decrease font size
Increase font size
Topic Title: 66kVA Generator
Topic Summary:
Created On: 13 July 2017 03:47 PM
Status: Read Only
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 13 July 2017 03:47 PM
User is offline View Users Profile Print this message



brightonbill

Posts: 4
Joined: 29 January 2004

I posed the following question on ask an expert.

I am involved in a project requiring a 66kVA generator as a standby supply to to a control panel. I've read quite a few posts about earthing, but came across what seemed to be a contradiction regarding the earth electrode resistance in Topic Title: life safety generator earthing.

I realise that a separate earth electrode is required, but what should be the resistance of the earth rod. In the post it reads "The earth rod resistance needs to be low enough to operate your ADS promptly for an LE fault, even if the DNO earth is absent", but in another it says "Provided the generator supplies a TN-S system (as would be usual), the rod won't be involved in passing earth fault currents, it merely references the N to earth, so it's resistance to earth isn't at all critial. From memory BS 7430 recommends < 20 Ohms".

Can anyone clarity?

The response from Chris Chew was "Generator earth is one earthing requirement. When the supply is disributed you comply to another earthing system of your choice."

So does this mean a < 20 Ohms earth rod will suffice as long as the installation is set up as a TNS?


-------------------------
electromott
 13 July 2017 04:09 PM
User is offline View Users Profile Print this message



gkenyon

Posts: 4954
Joined: 06 May 2002

You will need to comply with BS 7430, which as you say would mean 20 Ohms Max for neutral earthing of TN-S configuration.

Assume you've already sorted the requirements for the changeover switch etc., to comply with ESQCR and the other relevant requirements of BS 7671 ?

-------------------------
EUR ING Graham Kenyon CEng MIET TechIOSH
G Kenyon Technology Ltd

Web-Site: www.gkenyontech.com
 13 July 2017 04:18 PM
User is offline View Users Profile Print this message



broadgage

Posts: 2435
Joined: 07 August 2007

With an earth fault, between the CPC and a phase conductor on a TN-S system, there is a metallic path for the fault current and no significant current should pass through the earth rod, the resistance of which is therefore unimportant in this scenario.
20 ohms or less should be fine.

I would prefer a lower earth rod resistance, of say an ohm if possible.
Consider an earth rod resistance of 20 ohms and a fault between phase and the general mass of true earth, NOT THE CPC.

If this fault has a resistance of 20 ohms, then about 12 amps will flow which wont open any likely sized OCPD quickly. Moreover the CPC throughout the installation will now be live at about half line voltage, say about 120 volts. Clearly potentially dangerous, not perhaps as dangerous as it sounds since bonding SHOULD ensure that all touchable parts are at the same voltage.

If however the earth rod had a resistance of one ohm, and the 20 ohm fault between phase and the general mass of earth occurred as above, then the CPC would be raised to only about 12 volts which is tolerable.

In practice the fault I describe is unlikely, and bonding of the generator earth to services and structural metalwork reduces the risk.

Edited: 13 July 2017 at 06:51 PM by broadgage
 13 July 2017 06:01 PM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

"The earth rod resistance needs to be low enough to operate your ADS promptly for an LE fault, even if the DNO earth is absent"


Draw out the circuit and you'll realise that statement is nonsense

Your ability to clear a fault from a generator into a TN-S system is controlled via the generator subtransient reactance - which will be several order of magnitudes greater than a mains derived TN-S supply - basically your generator simply can't provide enough fault current to clear even a small rated protective device.

Basically, you need and RCD to provide earth fault protection, the electrode is simply stopping the neutral floating and allowing the RCD to operate correctly.

So, sub 20 ohms is fine (but note the unusual case Broadgage describes) - and you should connect it to the installation MET plus all the required bonding to that same MET

If the mains fails (but remains intact), the suppliers sheath remains as an earth electrode and is supplemented by your electrode

If the mains fails and doesn't remain intact (Bob the Builder and his big yellow cable finding tool0 then you are solely reliant on the generator electrode so you need to be sub 20 Ohms for you to "effectively" earth the generator system neutral

No idea if this is single or three phase, but you need to sort out the position of your N-E link (typically, it will be at the generator terminal box or at the changeover switch on a small system like this) - which I think is the point Graham is alluding to - so make sure your electrode "sees" this position under both circumstances of mains failure

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 13 July 2017 07:47 PM
User is offline View Users Profile Print this message



jcm256

Posts: 2303
Joined: 01 April 2006

For 66KVA generator An RCD device will have to be a MCCB with a bolt on RCD but BS7450 say as along as suitably chosen protective device, fuse, circuit-breaker or RCD operates within the time given in BS 7671:2008+A1.
See 7.2.5 and fig 12 in BS7450 2001 if it relates to you.
http://www.iso-iran.ir/standar...of_Practice.pdf


1991 (HSE)may be still current, never heard if it was out of date.
http://www.hse.gov.uk/foi/inte...499/oc482_2.htm


Why 20 ohms why not (10 ohms like combined lighting earthing electrodes), well you don't want to be lower than the mains source earth electrode which may be at a far flung transformer, otherwise a mains earth fault will prefer your electrode . (But there again DPC 18 th says will recommend local installation earth with PME supply). If it is a standalone generator no mains supply earth spike needed for consumers RCDs.
Regards
jcm

Edited: 13 July 2017 at 08:51 PM by jcm256
 14 July 2017 12:11 AM
User is offline View Users Profile Print this message



mapj1

Posts: 9568
Joined: 22 July 2004

I think there is confusion between different types of L-E fault.

If it is the cut cable or a fault in class 1 equipment, then the "E" that matters is the CPC, and as such has a modest impedance back to the genset neutral/star point.
Depending how long and thin your wiring is, then the fault currnet may be dominated by the genset's impedance to short duration events (but not the same as you may deduce from a static test of volt drop on load, where the controls have time to throw more fuel in the engine)

In such a case, an MCB or fuse will hapilly act as ADS on all but the longest thinnest run.

The second case is when the fault 'E' includes a current path via terra-firma. The common stinger of an example with temporary stuff is if a cable gets draggd over something sharp like a barbed wire fence and the live core gets peeled, but the rest of the insulation is undamaged, and then it starts to rain.
Because the phase of the genset is earthed to true ground via the fault, but the chassis /CPC is as well, there is a potential division between the two resistances and if the genset electrode is inadequate or absent, all 'earthed' metalwork becomes quite live.
For this situation an RCD is needed, as for all but the most extensive earth electrodes, or circuits protected by very small MCBs, nothing is guaranteeed to trip in that case without one.
For emergancy supplies, cleary the DNO supply may have been severed, so to rely on them for a terra-firma to CPC connection, while maybe the cable is being jointed, would be unwise in the extreme, and a local electrode must be furnished.
The DNO are allowed electrodes with resistance to true ground of 20 ohms or less for a single "pole pig" transformer feeding one farm on TT type situation, but a larger substation would be a lot better grounded than that.
Aiming for something similar is the intention here.

-------------------------
regards Mike
 14 July 2017 10:11 AM
User is offline View Users Profile Print this message



brightonbill

Posts: 4
Joined: 29 January 2004

Thanks for all the info everyone.

The gen set is a three phase unit in a newly created room (former store room) immediately next to the transformer room. The new generator supply cable is to be routed through existing ducts to the building opposite where the 4 pole ATS is located in the plant room, and from there to the control panel. Total length 130m.

I have allowed for bonding the star point to earth and the MET of the LV switchboard. The generator comes with an MCCB but not any earth fault protection.

The substation does have lightning protection. Can this plus the independent earth electrode count as one (thing) to bring the impedance down or do we have to rely solely on the independent electrode?

As you say, if JCB man goes through the main cable and takes the the earth with it, the lightning earthing will still exist.

-------------------------
electromott
 14 July 2017 12:20 PM
User is offline View Users Profile Print this message



OMS

Posts: 22393
Joined: 23 March 2004

You can combine the LPS electrode and the generator electrode into a global earthing system - if the overall LPS is less than 10 Ohms as it should be then just use that - you don't need any further electrodes

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.
 14 July 2017 01:32 PM
User is offline View Users Profile Print this message



AJJewsbury

Posts: 16007
Joined: 13 August 2003

Consider an earth rod resistance of 20 ohms and a fault between phase and the general mass of true earth, NOT THE CPC.

If this fault has a resistance of 20 ohms, then about 12 amps will flow which wont open any likely sized OCPD quickly. Moreover the CPC throughout the installation will now be live at about half line voltage

Although it's highly unlikely you'll ever have a fault of negligible impedance to true earth (even when we try very hard using lots of buried metal it's hard to get below several Ohms), which could change the numbers significantly. Also it should take a double fault for a live conductor to come into contact with the general mass of the earth (same as coming into contact with a person) - typically there should either be earthed (as in connected to the c.p.c.) metal or double/reinforced insulation interposed - a situation generally so rare that BS 7671 doesn't require us to take account of (or where there is recognised risk (e.g. hand held tools outdoors), BS 7671 already requires 30mA RCD protection for the circuit that might originate such a fault anyway - so disconnection should be assured).

Generators on mobile units only need to be "earthed" to the chassis - there's no requirement in that situation to have a deliberate connection to true earth - the rules might be different but the laws of physics are just the same.

- Andy.
 14 July 2017 02:27 PM
User is offline View Users Profile Print this message



brightonbill

Posts: 4
Joined: 29 January 2004

Thanks again

I'll request the lightning protection test results to see what the resistance is. I assume that this will be bonded to the MET therefore can I just connect to the MET if this is the case?

Also, the building that houses the control panel (the one 130m away) has it's own lightning protection; hence I assume that this will effectively double up the earthing.

Do you agree?

-------------------------
electromott
 14 July 2017 11:51 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9568
Joined: 22 July 2004

A guarded yes, from an earthing point of view it will be most likely be fine. Be jolly careful that you know where your NE link is, or if you have more than one of them in circuit at once, that the CPC cable between is not in effect going to carry a large part of the neutral load current when running from the DNO supply, by dint of some lower impedance parallel path.

It is an aside but
Generators on mobile units only need to be "earthed" to the chassis - there's no requirement in that situation to have a deliberate connection to true earth - the rules might be different but the laws of physics are just the same.
is only really a sensible approach if all the loads are contained within the same vehicle as the genset. If however the situation is more like a fairground or circus, where the genset is one vehicle, but then LNE cables run to other class one loads that may be on the ground, or even in or on other vehicles, then it does become very sensible to have a terra-firma earth to the CPCs, as it is no longer a single self contained earth island, and a load chassis to live fault, or a cable to terra-firma fault, may cause either one vehicle or all to become partly live.

-------------------------
regards Mike
 15 July 2017 09:46 AM
User is offline View Users Profile Print this message



lyledunn

Posts: 1108
Joined: 13 August 2003

Is the 20ohm electrode impedance not an outcome of the desire to mitigate stress voltage in the event of a phase to earth fault on the primary side of the standard DNO tx? It then becomes a useful tool to reference neutral and to permit the automatic detection of earth faults on the secondary side of the system?
Notwithstanding the situation Broadage described the need to hold the electrode resistance to 20 ohms with a genset using TNS is less critical. I suppose by keeping it around 20 the installation could also opt for a TT arrangement.
This is a question and not a learned response!

-------------------------
Regards,

Lyle Dunn
 15 July 2017 09:59 PM
User is offline View Users Profile Print this message



mapj1

Posts: 9568
Joined: 22 July 2004

A DNU supply with a 20 ohm LV earth is only really typical of a small transformer, perhaps the pole pig in the country, where LV earth (the one on the star point and neutral) and the HV earth (transformer core and containment, and related metal bits that support any HV isolators) are two seperate electrodes, far enough apart that they are essentially not going to interact, so an HV fault should not bounce the neutral on the LV side. TheHV earth is the path that fires the expulsion fuse on the primary, and or operates an earth fault detector (think a giant scale RCD with sensing coils and interrupting breakers in seperate places).

The transformer construction is normally such that a primary to secondary fault is (almost) impossible without also livening up the metalwork of the core and containment, so the HV side ADS does not really need the LV earth.

If the impedance is low enough that an HV fault will not cause too much ground bounce, then electrodes from HV and LV may be combined. However that is only appropriate at sites with electrodes that are of a sub-ohm performance. Even then current advice is to take care and not to use the one ohm rule of thumb, but to actually calculate worst case earth rise during fault before deciding if combined ('cold' is OK, or separete ('hot') is best.
OF course the PSSC at the end of a long run of 11kV may seem surprisingly low to those used to 230/400V operation of a kiven no of kW.

One could TT a genset, but it is never done in my experiance, given the lengths between genset and load, the saving on 2 core cable versus 3 (or the 3 phase equivalent ) is not worth it. TNS it is.

-------------------------
regards Mike
 17 July 2017 06:56 PM
User is offline View Users Profile Print this message



jcm256

Posts: 2303
Joined: 01 April 2006

Of course there is some variants' to this earthing of generators, you know the one where you have to throw a ladder against to read the nameplate and check the neutral earth link. None fitted that could be because low voltage to high voltage then out the 11Kv lines. The step-up transformer winding configuration is normally delta on the low and star on the high voltage winding (earthed through a neutral resistor). In such cases advice to anyone who is checking as part of EICR, slink away say nothing until you study facts. Also checking generators at airport while the main generator for the terminal is neutral earthed , the smaller standby ones down the runway for runway lighting has no earth/neutral link, these could be run in parallel , what would your comment be (multiple diesel generating sets, it is necessary to ensure that one (and only one, to avoid the possibility of circulating currents) of the neutrals of one of the running generating sets is earthed at all times). Let the hare sit, say nothing, though so.
You see it is not 100% all plain sailing.
jcm
Statistics

New here?


See Also:



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

..