IET
Decrease font size
Increase font size
Topic Title: substation earthing
Topic Summary:
Created On: 07 January 2006 11:00 AM
Status: Read Only
Linear : Threading : Single : Branch
1 2 Next Last unread
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 07 January 2006 11:00 AM
User is offline View Users Profile Print this message



weasel

Posts: 90
Joined: 19 July 2002

Gents,
I am currently designing a new 33kv/11kV primary substation with two 7.5/15MVA transformers, a 5 panel 33kv GIS board and a 9 panel 11kV board. The site is in an urban area in close proximity to a lot of houses and i am a little concerned about rise of earth potential
The switchroom is having a new mezzanine floor installed which consits of large metal mesh type plates 2x4m which will rest on steel cross beams.

At the moment the switchboards, relay panels, battey chargers etc.. will all be bonded into the earth ring as usual, but do i need to bond in each of the metal plates & cross beams into the main earth ring as well? Or is the fact that all the plant is bonded in and they are all touching the floor adequate?

So far i have had a few different answersfrom colleagues, but none of them have sounded very convincing

cheers
steve
 09 January 2006 09:52 AM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

Hi,
I would certainly recommend that the main steel beams are bonded to the main earthing system. It is preferable for each steel mesh plate to be bonded to its support beam but if they are bolted down to the beams, rather than just sitting on them, the bolts should provide an adequate connection. Note that the bolts are unlikely to have to carry any significant fault current as this bonding is merely to provide potential grading for personnel.

Dr. M. Taylor

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 18 January 2006 05:43 AM
User is offline View Users Profile Print this message



Samuel Chan

Posts: 210
Joined: 02 August 2002

I'm not a HV engineer (is LV engineer). My electrical experience on this is (yes) to bond all to main earth and then to the main earth-pit (I'll use the sub-structural or steel-work of the station as earth-pit) so as to min. potential.
 18 January 2006 04:17 PM
User is offline View Users Profile Print this message



tonysung

Posts: 630
Joined: 14 September 2001

In HV design, normally under the UK Electricity at Work Regulations or US NESC, I personally interpret it is an "absolute requirement" to protect against any danger arising from 'step, touch (mesh) and transfer potentials'. If the HV design calculations (based on IEEE standard?) show that there is a risk, without a document in writing from a good authority, we should take steps to eliminate the risk.

The use of Faraday Cage principle to lower the touch potential etc to a safe value is, I think, acceptable. A thick, short and adequately sized bonding braid would seem make good sense to me. However, just to cover myself here, you must exercise your own engineering judgement that the local-equipotential bonding will not lead to a greater danger under any circumstances.

Regards

-------------------------
Tony Sung
 18 January 2006 04:46 PM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

quote:

Originally posted by: tonysung
The use of Faraday Cage principle to lower the touch potential etc to a safe value is, I think, acceptable. A thick, short and adequately sized bonding braid would seem make good sense to me. However, just to cover myself here, you must exercise your own engineering judgement that the local-equipotential bonding will not lead to a greater danger under any circumstances.

Regards


The Faraday Cage principle relates to equalising electric fields around a body where the cage need not necessarily touch the body, whereas what is needed here is an equipotential surface that keeps a persons hands and feets at the same potential, which isn't quite the same.

If the mesh flooring were not bonded to the equipment, the potential difference probably wouldn't be above the allowable touch voltage limits, but this would have to be proven by computer simulations. However, in this case, the cost/time taken to carry out the simulations probably wouldn't be cost effective against the cost of bonding all of the floor panels together.

Dr. Matthew Taylor

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 18 January 2006 11:02 PM
User is online View Users Profile Print this message



alancapon

Posts: 5831
Joined: 27 December 2005

There seem to be two questions here, both related to earthing. The first applies to the environment inside the substation, the second to the environment outside the substation.

Firstly, inside the substation. I'll agree with the replies already given on this one. You are trying to protect someone either operating the switchgear, or working on the switchgear. With the bonding you apply, you are trying to ensure that the metalwork the person is standing on will be at or near the same potential as the switchgear metalwork they are touching, particularly in the event of a fault scenario.

The second question relates to the area round the substation. In the initial post, the question of "Rise of Earth Potential" (ROEP) was briefly mentioned. Whilst I cannot help on the detailed "design" side, I will try and explain what ROEP is and the problems it can create.

ROEP is a voltage measurement at the substation earth system compared with "true earth" during the time that a worst-case earth fault is present on the system. The voltage will decrease to zero ("true earth") as you get further away from the substation, like ripples in a pond when you throw a stone into it. This gradient is usually expressed in "volts per meter".

The substation building will have an earthing system attached to it. The earth resistance of this system can be calculated / measured. The resistance you get, represents the resistance to "true earth". You will also need to calculate the worst-case fault current that will have to flow through this resistance. This will depend on the design of the HV network on both the "supply" and "load" sides of the substation. Application of Ohm's law will give the ROEP voltage.

There are two measurements that are critical from a safety point of view. The first is the "step potential" in the area that experiences the voltage gradient. Some livestock, such as cattle are more sensitive to step potential than humans. Your company should have some standards that lay down what is aceptable. The second concern is telecom services provided over copper. These are effectively at "true earth" potential and the danger is the voltage difference between a telephone handset (for example) and the surface that the user is standing on. There are two limits set that determine that the site is classified as "hot". These are ROEP>650v assuming "high speed protection" (the fault will be disconnected within 200 milliseconds) or ROEP>430v for disconnection times greater than 200ms. It is conceivable that the 33kV and 11kV sides of the substation may have differing fault clearance times for an earth fault. The areas where voltages in excess of these can be found is known as a "hot zone". Any telecom equipment within the "hot zone" may need special isolation units or procedures in place to allow it to be worked on.

It should be remembered that these hazardous voltages only occur following an earth fault, until the appropriate circuit protection trips the circuit.

Regards,

Alan.

(Edited to correct fault disconnection times to "200 milliseconds" in Para 6)

Edited: 19 January 2006 at 11:13 PM by alancapon
 19 January 2006 09:21 AM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

quote:

Originally posted by: alancapon
There are two measurements that are critical from a safety point of view. The first is the "step potential" in the area that experiences the voltage gradient. Some livestock, such as cattle are more sensitive to step potential than humans. Your company should have some standards that lay down what is aceptable. The second concern is telecom services provided over copper. These are effectively at "true earth" potential and the danger is the voltage difference between a telephone handset (for example) and the surface that the user is standing on. There are two limits set that determine that the site is classified as "hot". These are ROEP>650v assuming "high speed protection" (the fault will be disconnected within 100 milliseconds) or ROEP>430v for disconnection times greater than 100ms. It is conceivable that the 33kV and 11kV sides of the substation may have differing fault clearance times for an earth fault. The areas where voltages in excess of these can be found is known as a "hot zone". Any telecom equipment within the "hot zone" may need special isolation units or procedures in place to allow it to be worked on.

Regards,

Alan.


I agree with you there Alan, and well explained. The only error I think is that the "high speed protection" fault clearance time that determines whether the 650V or 430V limits should be applied is actually 200 milliseconds not 100 ms.

We are getting into earthing of substations here really rather than bonding within a building, but it's all interesting stuff.

The voltage difference experienced in the telephone line shock scenario that you mentioned is usually referred to as the "transfer potential".

I would also add that "touch potential" within and external to the site, particularly with regard to members of the public touching the perimeter fence, is also very important.

Finally, note that as a result of recent research, the 'hot zone' voltages and those that the telecom industry are prepared to accept are due to increase significantly, which will be detailed in new releases of the earthing standards.

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 19 January 2006 08:50 PM
User is offline View Users Profile Print this message



BryanLeyland

Posts: 45
Joined: 09 November 2001

Earthing can be a bundle of worms. Especially in existing installations or in urban areas.

First, bond the reinforcing steel to the earth system. Almost without exception, this gives a better earth electrode than anything you can install. (In a recent station in NZ, a consultant had our client spend ~£150,000 on drilled earth rods because their computer program ignored the bonding to the reinforcing that I had already had had installed. When they measured the earth resistances, the reinforcing was was <0.1 ohm and the rods were about 5 ohms.)

Second. Computer programs for calcualting earth resistance in existing systems are worthless because you cannot input - and often they won't accept - data about everything that is connected to earth. The actual network is very complex and probably extends to many other installations via cable sheaths etc. So an actual test using a suppy at a frequency other than 50 Hz (from a portable generator for instance) is essential. (By dong this, we saved another client ~£100,000)

Third. Neutral earthing resistors can virtually eliminate the problem of high voltages during an earth fault. They are a much underrated means of reducing costs and improving safety. (The reduce costs because you can buy cheaper 11kV cables with minimal screens. This more than pays for the cost of the resistors.) You will find a paper on this in www.bryanleyland.co.nz.



-------------------------
BryanLeyland
 19 January 2006 10:45 PM
User is online View Users Profile Print this message



alancapon

Posts: 5831
Joined: 27 December 2005

quote:

Originally posted by: taylormj4
The only error I think is that the "high speed protection" fault clearance time that determines whether the 650V or 430V limits should be applied is actually 200 milliseconds not 100 ms.



Well spotted Matthew. The "high speed protection" fault clearance time given in my post should have been 200 milliseconds.

Regards,

Alan.
 20 January 2006 04:00 PM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

I agree Bryan, bonding in the reinforcing bars is generally a good procedure. However, I would add that you do need to supplement these by a 'main earthing system' that can conduct away all of the fault current, otherwise you can end up with small reinforcing bars trying to conduct large currents and the heating effect can damage the concrete that the bars are supposed to be holding together. So the earth rods in the example that you mentioned will serve a useful purpose albeit there may be too many of them or the soil structure may not favour the use of rods.

I can't agree with the 'worthless' label on earthing computer programs. We use computer software to simulate earth faults on a wide range of complex earthing systems including railways, mobile phone transmitters, substations and power stations and find that most things can be included in the model, resulting in cost effective designs and solutions. I do agree that the simulation should be backed up by a measurement where possible. Although there is not always enough land area available to conduct one correctly, especially in urban areas.

NERs will reduce the fault current and hence the rise of earth potential, as you stated. Your investigations certainly seem to show that they are cost effective. However, NERs cannot always be used on high power circuits, where high load currents are required. In the UK, NERs are not used on the transmission network (275kV and 400kV).

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 21 January 2006 09:58 PM
User is offline View Users Profile Print this message



BryanLeyland

Posts: 45
Joined: 09 November 2001

I did not intend to suggest that connecting to the reinforcing bars eliminated the need for a comprehensive earthing network. But experience in New Zealand has shown that, almost without exception, multiple connections to reinforcing is more effective than driven earth rods. There is a standard connector available in New Zealand that is designed to be bolted to the boxing and connected to a main reinforcing bar. It is a good idea to tack weld these main reinforcing bars to the smaller bars crossing them for a few metres on each side of the connexion. We were expected to have several of these connexions in and installation.

In my opinion, the best thing is to test the resistance value of reinforcing and, if necessary, add additional earthing rods. I suspect that extra rods would be necessary in only a few cases.

Another thing that I forgot to mention is any connexion to the external low voltage neutral. One of the lines companies in New Zealand always supplied the lighting and small power in their substations from the local - usually overhead - distribution system. This has a multiple earthed neutral and they discovered ] that earth faults on the 11 kV system often burned out out the low voltage neutral connexion. We investigated this and recommended that their be preferably two good solid connections between the substation earthing system and the low voltage neutral in the street. Tests then showed that, that many substations, the low voltage neutral was the most effective earth.

In my experience, the earthing software commonly used in New Zealand is "worthless" when faced with a complex existing installation. Better programs may now exist but there is still the problem of making sure that every deliberate or inadvertent connection between the earthing system and the mass of the earth is included. (Low voltage neutral being a classic example!). I have much more confidence in site tests.

New Zealand is slowly - much to slowly in my opinion - moving to neutral earthing resistors. In recent years four children have been very badly burned through accidental contact with 11 kV lines. In all cases, neutral earthing resistors would have meant that there woujld have been less burning.

-------------------------
BryanLeyland
 04 February 2006 03:33 PM
User is offline View Users Profile Print this message


Avatar for JonathanHill.
JonathanHill

Posts: 225
Joined: 09 September 2002

2 further comments for consideration:

1 - I agree that the ROEP and gradients from both the 33kV and 11kV systems need to be assessed
2 - If cable systems are used, then a proprtion of the fault current will flow back to the sources through the cable sheaths. This can result in a significant reduction in ROEP at the new subststion, and considerable savings in buried earth electrode can result. Information on this can be found in the ENA publication on Substation Earthing (the number escapes me at present but see the Energy Networks Association website - Publications.

-------------------------
Jonno
 11 February 2006 01:50 PM
User is offline View Users Profile Print this message



weasel

Posts: 90
Joined: 19 July 2002

Cheers chaps,
I have waded my way through the EATS 41-24 and think i have the general earthing philosophy sorted out. I will stick in a main ring inside the barn to pick up all the switchgear and chargers, bond in the metal cross beams and then connect this ringin a few places to an external ring of buried 70mm bare copper with rods to reinforce it.
Just as a thought can anyone recommend a 'good' (i.e. easy to understand with practical examples) book that covers earthing in more detail - or is it more worthwhile to go on the EA tech earthing course?

cheers
steve
 12 February 2006 07:22 AM
User is offline View Users Profile Print this message



BryanLeyland

Posts: 45
Joined: 09 November 2001

As I said before, don't forget the reinforcng and anything else that is copnnected to ground or power cables.

And measure before spending moneyh on rods.

-------------------------
BryanLeyland
 13 February 2006 05:13 PM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

quote:

Originally posted by: weasel
Cheers chaps,
I have waded my way through the EATS 41-24 and think i have the general earthing philosophy sorted out. I will stick in a main ring inside the barn to pick up all the switchgear and chargers, bond in the metal cross beams and then connect this ringin a few places to an external ring of buried 70mm bare copper with rods to reinforce it.
Just as a thought can anyone recommend a 'good' (i.e. easy to understand with practical examples) book that covers earthing in more detail - or is it more worthwhile to go on the EA tech earthing course?

cheers
steve


Hi Steve,
Megger, or AVO, used to produce a basic introductory book / panphlet on earth testing and basic earthing principles.

......I've just found it on the internet at the following address:

http://www.megger.com/uk/story/Index.php?ID=146

I read this when I was doing my PhD on power earthing and it was a good introduction.

As well as 41-24, there is also ER S34, BS7430 (rewritten 1998) and IEEE81 and IEEE80 (2000, I think).

Alternatively, get yourself a good consultant !

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 05 March 2009 10:39 PM
User is offline View Users Profile Print this message



deansumner

Posts: 25
Joined: 05 March 2009

Gentlemen

Reading your correspondence has been very helpful.

However I am puzzled and wonder if any of you guys could help.

In the BS7430: Earthing Regulations it states for distribution earthing one of two options.

One been that a HV & LV earth nest can be combined if the earth nest resistance to earth is less than 1ohm and that the EPR (Earth Potential Rise) is less than 430V.

Two been if the above cannot be met then the HV & LV earth nests are to be insulated from each other and be segregated by a minimum of 3mtrs.

To keep things simple, I go down the route of option 2.

However as I understand the EPR is basically ohms law
I x R = V

If this is the case then say for worst case scenario situations;

11kV fault current been 13.1kA and say an earth reading of 1ohm the EPR would equate too 13,100V

Even worse for 430V, fault current been 50kA and say again an earth reading of 1ohm then the EPR would equate too 50,000V

Surely this would lead to detrimental high touch, transfer & step potentials????

Also from the regulations it states that with a separate HV & LV earth nest that the HV earth nest reading is to be 10ohms or less and that the LV earth nest is to be 20ohms or less.

Can somebody please explain???

Kind Regards
Dean
 06 March 2009 12:05 AM
User is offline View Users Profile Print this message



BryanLeyland

Posts: 45
Joined: 09 November 2001

One been that a HV & LV earth nest can be combined if the earth nest resistance to earth is less than 1ohm and that the EPR (Earth Potential Rise) is less than 430V.

Two been if the above cannot be met then the HV & LV earth nests are to be insulated from each other and be segregated by a minimum of 3mtrs.
>>>I think that this is dreamland stuff! In the real world, the "earth nests" are connected by the earth resistance between them and, in reality, both earth systems will probably be interconnected somewhere. So the only rational option is to connect everything together - and in particular, the building reinforcing and steel work. This is the best earth you will ever get and the whole shebang is at the same potential. So its safe! If you measure the earth resistance of the whole system, it will be low.


11kV fault current been 13.1kA and say an earth reading of 1ohm the EPR would equate too 13,100V >>>>But its step and touch voltage that counts.

Even worse for 430V, fault current been 50kA and say again an earth reading of 1ohm then the EPR would equate too 50,000V >> Ignoring the fact that there will be a real resistance and so the current will be lower. 50kA is for a bolted fault.

Surely this would lead to detrimental high touch, transfer & step potentials????
>>Not necessarily!

Also from the regulations it states that with a separate HV & LV earth nest that the HV earth nest reading is to be 10ohms or less and that the LV earth nest is to be 20ohms or less.
>> Its time they were updated then!

I have just been for a walk in Massachusetts and I noted that HV, LV and everything else was connected to a common HV LV neutral/earth wire. Sensible!

-------------------------
BryanLeyland
 13 March 2009 12:02 PM
User is offline View Users Profile Print this message



taylormj4

Posts: 25
Joined: 30 December 2002

Hi Dean,
The HV and LV earths can be segregated and this is done on 11kV sites by a lot of the electricity companies in the UK at say 11kV distribution substations or pole mounted transformers. I think the earth grid around the substation is usually bonded to the HV earth and the LV earth connection is then insulated for a set distance away from the substation and say earthed at the first pole out or other appropriate location. It is important that the earths be segregated if the HV substation is 'hot' for an 11kV earth fault as you should not allow the rise of earth potential (for an 11kV fault) to transfer onto customers LV earths in their premises (which is what would happen if the LV and HV earths were combined).

For your rise of earth potential (RoEP) calculations, note that the rise of earth potential can never be higher than the supply voltage. So you cannot have a RoEP of 13,100V on an 11kV system nor 50,000V on an LV system. When calculating the fault level, you need to take into account the fault resistance. As an example, I am presently designing a windfarm earthing system (11kV connected), which has a fault level of 1254A. The site earth impedance is approx 10 ohms. This would give a RoEP of 12,540V, which is impossible on an 11kV system. When the fault levels were recalculated to take into account fault impedance, the fault level is as follows: 720amps with a 4 ohm fault impedance, 463amps with a 9 ohm earth impedance and 323A with a 15 ohm earth impedance. So you can see the importance of getting the fault impedance taken into account.

Just to confuse matters a little, the actual current that you should use in the 'ohms law calculation' is the "ground-returning fault current". That is the part that actually goes into the ground, as this is the only element that causes RoEP. Any part of the fault current that returns to source through other paths (overhead line earthwire, cable sheaths etc due to mutual coupling with the faulted phase) will not cause RoEP.

Hope that all helps.
Matthew Taylor

-------------------------
Matthew Taylor B.Eng C.Eng MIET PhD
 14 March 2009 11:28 AM
User is offline View Users Profile Print this message



deansumner

Posts: 25
Joined: 05 March 2009

Thanks Matthew

That has helped allot.

Also Bryan you are correct, apart from as Matthew has stated HV & LV earth electrodes should be separated if a fault on the HV system(11kV) gives a rise of potential to higher than 430V on the LV system(430V).

However the current BS7430:1998 (Earthing code of practice) is dated and I believe now that the 1ohm rule will now be scrapped, as a 1ohm resistance to earth reading does not nessacarilly give you an Earth Potential Rise (EPR or as you refer RoEP) less than 430V, although I suppose this is backed up by the BS stating it should be 1ohm or less AND an EPR less than 430V.

The other aspect the BS states if this cannot be achieved the HV & LV nests should be separated by at least 3mtrs, this is not good guidance as the LV earth nest should be outside the 430V contour of the HV nest.

Also Bryan with respect to 10ohm or less for HV and 20ohm or less for LV, yes the current BS is vague and by this I think it means that just the earth nests themselves should be below this value not taking into account structures, fences, PME earths parallel paths etc.

I have been told a draft copy for the new BS7430 was issued around 18months ago, but as of yet am still paticantially waiting for the official revision.

Matthew

Is it possible for you to explain why it is deemed safe if an EPR is less than 430V??

Kind Regards
Dean
 19 March 2009 12:59 PM
User is offline View Users Profile Print this message



wilkinson

Posts: 1
Joined: 12 February 2009

Have a look at ENATS 41/24 section 11.12 earthing of ancilliary metalwork. This should help.
As a rule of thumb all exposed metalwork within a substation should be bonded to earth.
IET » Energy » substation earthing

1 2 Next Last unread
Topic Tools Topic Tools
Statistics

See Also:



FuseTalk Standard Edition v3.2 - © 1999-2014 FuseTalk Inc. All rights reserved.