Earthing resistance and calculations 5

[AusLee]

Hi,

I'm a bit new at substations grounding and I have given myself an exercise to do it for a 1 MVA 11/0.4kV kiosk substation but I need assistance please:

1. Is there any Free software that can help me with this?

2. A friend of mine at work uses a decent software in his department (ETA Power Station). I was surprised when he showed me the screen for the soil resistivity: you specify only the cover, upper layer (and its depth) and lower layer. Shouldn't there be a procedure to go on site and use a "Shepperd" stick (vertical rod with battery and two electrodes) and take measurements across several points on the site? And if that is done, what do I do with all this "valuable" earth resistance site survey if a software as powerful as ETAPS does not need it?

3. I understand that IEEE 80 and the rest of the stuff is more towards "zone substations"/"outdoor switch yards" with sizeable area. What about a 1 MVA kiosk substation? Historically, i think all that is required was two deep driven electrodes, one at the HV and one at the LV sides. As i'm trying to understand this better, the fault level of this substation at the LV side is around 25kA. From the equation giving the voltage at a distance r away from the electrode or the grid:

Vr = Icc * rho / (2 * pi * r)

if the short circuit current Icc alone is 25kA, the Vr will be in the order of 3-4 kV, which is way above the acceptable level of a step voltage (should be not more than 100V if the protection trips in 0.1s to my understanding, maybe IEEE 80 allows a bit more).

So if this calculation is correct, has it been "always wrong" to use only a couple of rods, and indeed a grid that covers and extends beyond the area where people might be standing near the substation?

I hope someone has a short and quick step by step guide they would like to share.

Looking forward to some instructions

 

[Marmite]

Back to basics Auslee. You have a 400V system with 230V to earth. Its physically impossible to get 3-4kV step voltage as you describe.
Regards
Marmite

 

[AusLee]

Hi Marmite,

Thanks for your reply.

I was just reading the other post where they say the same:

http://www.eng-tips.com/viewthread.cfm?qid=278625&page=2

well two questions:
1. it seems on that post they say IEEE 80 notes that only the HV side of the fault is to be considered.

I would like to question this a bit, in the ETA Power Station document i'm reviewing, you do a short circuit calculation on the network, you say which buses to include in the fault, and it will pass on to the earthing module the highest kA value, which obviously would be from the LV side of the transformer. Maybe the LV bus should not be selected in the first place to be included in the short-circuit calculation? if so, ok, lesson learnt.

2. Regarding the 3-4kV from 400V, ok, let's consider it's the HV side, if the utility has a fault level of 250MVA, then at 11kV primary, the fault level is 13kA.

Vr = 13 kA * rho / (2 * pi * 3 meters) is still in the 2kV level, much higher than an acceptable step voltage, so again, aren't / haven't every 2 rods only been sufficient for a kiosk substation?

 

[waross]

When a sub feeds a distribution network at 13 kV, 25 kV or 35 kV levels the distribution is often a four wire wye scheme with multiple protective earths.
In the case of a hard line to earth or line to neutral fault, the fault current will divide between the earth path and the neutral copnductor. The sub feeding this network is the place where very high currents and high touch and step potentials are possible.
The application to the primary of your kiosk may be slight.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[AusLee]

Very interesting waross. In Australia however it is 3 wire at the HV side, even if it splits in half each way, 1kV is still above acceptable step voltage.

Is there any free software to do these calcs?

And what about the "value" of taking a decent soil resistivity using a vertical probe vs the 4 point method and coming up with just 1 value? especially if the software will just ask for the soil resistivity as 1 value input and calculate the voltage gradients from the short circuit current and the grid layout.

 

[ScottyUK]

Lee,

It's not too hard to put the IEEE 80 calcs into Excel - it's a worthwhile little exercise and gives you a useful tool.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[FreddyNurk]

Given your location, you should also consider the relevance of AS2067 as opposed to the previously used IEEE 80. Most utilities seem to be playing catch up with the standard, though as I understand it compliance with AS2067 is referenced from AS3000 and thus all consumer sites need to comply.

IEEE 80 will, however, give a reasonable indication of the process. Marmite is correct as to why IEEE80 doesn't consider LV faults for step and touch, though you'll still need to ensure that the connections to the ground grid can handle the prospective LV fault level.

That said, the main issue that appears in terms of the difference is the method of calculation of acceptable step and touch potentials. Ground grid design doesn't appear to be different between the two standards.

Also notable is that AS2067 recommends a combined HV / LV earthing system unless appropriate segregation between the two systems can be obtained.

 

[Marmite]

Attached IEEE80 sreadsheet.
Regards
Marmite

 

[UtilityGy]

I have a questions for the experts here. Suppose the primary is 115 kV Delta and Secondary is 44 kV with solidly grounded star connection.

Three O/H line leave the station with a ground underneath the three wires and no neutral. Dont you think in that case if a fault occurs 1 KM down the line on secondary side, fault current would return through the ground connection and then the grid, the neutral connection and then the transformer back to the line faulted feeding the fault.

In this situation, we will always have to design a grid with secondary fault in consideration.

My concepts are rusted for now so there is a possibility that I might be missing something in here.
Thanks

 

[AusLee]

Hi Marmite,

Now I know the answer to questions 1 and 2: no free software and there is no need for a million bore holes on site, a couple is enough as the soil resistivity will be the average (or better, the maximum) of the samples.

Thanks for the file. I have checked it, the excel equations in it are correct to IEEE 80 however some of the pictures are not, specifically the picture for the equations of Cs and Km, the existing pictures show equations that are a bit different. I'm attaching my modified version 2 if you could please have a look?

So for a substation of 1 MVA, supplied at 11kV primary with 415V secondary and a fault level of 250MVA from the utility, all i could find out from the excel sheet is that the minimum cable should be 70 sq. mm.

For the rest of it, i tried to put a grid of like 9 meters by 6 meters and a middle conductor with a rod at every intersection, and i reduced the clearance time of the fault to 0.2s, i still get excessive touch, step and mesh voltages.

Can someone please clarify just this point? now a 1 MVA kiosk substation requires all this earthing and still not enough? at a time when i think that there are plenty of 1 MVA kiosk substations who have just 1 electrode in the ground, not even the required 2 at opposite ends. What's happening? If you had to earth this substation using IEEE 80, how would you do it?

Cheers

 

[waross]

I suspect that the high step voltages that you are seeing are an issue for the line feeding your kiosk and the transformer or sub feeding that line. Ground fault current on the primary side will be supplied by the source line and sub, not by your transformer.
If your system is capable of regenerating there may be some ground fault contribution by your transformer but that calculation will be based on the capability of the machines back feeding.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[AusLee]

Thanks waross, i confirm that the 13kA at the primary side is calculated as follows: 250MVA / 11 / sqrt(3) = 13kA and is irrespective of the transformer (630kVA, 1MVA, etc..) and that any possible regenration (as you pointed out) is not possible in this case. Btw i'm taking 250MVA but there are some places where the authority wants 500MVA which will make things worse

So how do we "correctly" earth a simple kiosk substation? Can anyone please return a "corrected" excel sheet?

Thanks!

 

[jghrist]

Thanks waross, i confirm that the 13kA at the primary side is calculated as follows: 250MVA / 11 / sqrt(3) = 13kA and is irrespective of the transformer (630kVA, 1MVA, etc..) and that any possible regenration (as you pointed out) is not possible in this case. Btw i'm taking 250MVA but there are some places where the authority wants 500MVA which will make things worse

So how do we "correctly" earth a simple kiosk substation? Can anyone please return a "corrected" excel sheet?

You need to consider the split of the fault current between the earth and metallic ground wire back to the fault current source. If your 11 kV system is really only 3-wire with a solidly grounded source and no ground wire, the current will all flow through the earth, but I doubt that you would have 13 kA Ø-grd fault current with no metallic return.

 

[jghrist]

Three O/H line leave the station with a ground underneath the three wires and no neutral. Dont you think in that case if a fault occurs 1 KM down the line on secondary side, fault current would return through the ground connection and then the grid, the neutral connection and then the transformer back to the line faulted feeding the fault.[/wuote]
By "ground underneath" do you mean a ground wire? If so, this will be the same as a multigrounded neutral and will be a path for return current. Some of the fault current will flow through the earth for a line fault and cause step- and touch-voltages at the substation. Usually, a fault on the HV side of the substation is a worse condition, however.

If there is no metallic return, then there may be step- and touch-voltage problems at the substation, but there will be a worse problem at the location of the fault unless there is a large ground grid system there. You may also have problems detecting and tripping for a fault with no metallic return path.

 

[EddyWirbelstrom]

Three O/H line leave the station with a ground underneath the three wires and no neutral. Dont you think in that case if a fault occurs 1 KM down the line on secondary side, fault current would return through the ground connection and then the grid, the neutral connection and then the transformer back to the line faulted feeding the fault.

.

I assume the above could read 'Three O/H line leave the station with an underslung earth wire'.
Assuming uniform wooden pole spacing and earth resistivity at each pole electrode, could the Garrett method - described in IEEE 80 Section 15.9 - be used to calculate the division of current between the earth and earth wire ?

AS3835.2 - 2006 section 5.9.2 shows a simplified method of determining the impedance of the O/H earthwire in parallel with pole earth electrode resistance. This impedance in parallel with the earth grid resistance would then be used to calculate total earhtfault current 3Io. The current division between O/H earth wire current and current in the earth could then be calculated.
I have attached a sketch similar to the O/H line 'equivalent ladder network' shown in AS3835.2 Section 5.9.2.

 

[AusLee]

Thanks all.

I've attached Rev 3 of the excel sheet where the fault from the source is 6500A which is 13kA split in two, still high step and touch voltages.

Let's keep it simple please

I have a kiosk substation, 11/0.415kV, fault level from the supplier: 250MVA. The size of the substation is 5 x 3 meters (15 x 10 ft). It will be built on top of a soil with resistivity 100 ohm.m (sry i used the word ground before). Using the IEEE excel sheet, how do you design the earth gird?

Much appreciated is someone can correct the excel sheet.

Many thanks and looking forward to your reply.

 

[UtilityGy]

Auslee,

Sorry for jumping in your post.

Thanks for your response Jghrist and EddyWirbelstrom.

1. In Stations those have a Delta secondary or a floating star connected secondary leaving the station and then a fault occurs down the line, and if there is no ground return, Line breaks and hands to a wooden pole or falls on sandy ground. How current will return back to Source ?
Will the flow back simulate its self as a zero sequence flow with in Delta winding?


2. My Earlier post, In case of a ground return available through a overhead ground wire, which is bonded to the mesh at the station and when the current will return back during a single line to ground fault in case of a solidly grounded star connected secondary winding leaving the station, Will it still be considered as Ig=(I (fault)- I (earth)) ? I(Earth) should be used for GPR calculation that measn very little potential.

Jghrist, It looks like my grounding concepts are messed up again, I need to go back and read all your posts again.

Anyways I would appreciate your word on this.

Thanks

 

[EddyWirbelstrom]

1. If the 11kV is not earthed ( transformer secondary delta ) then in the event of an earthfalt on the 11kV system, only a very low capacitive current will flow in the earth.
2. Most 11kV systems will be earthed ( transformer secondary star point earthed – or zig-zag transformer ). If the O/H line includes an earth wire ( OHEW ) then in the event of an earth fault on the 11kV system, the total earthfault current ( 3Io ) will split between the earthgrid and the OHEW. An approximate method to determine this split is shown in AS 3835.2:2005 Section 5.9.2. More accurate methods are shown in IEEE 80 Annex C.
I have used the SKM Ground Mat Ver 2004 to calculate the step and touch voltages for two earthing configurations of the 11kV substation. I have assumed that the split factor is 50%.
The total earth fault current supplied from the source is split equally between the earthgrid and the OH earthwire.

Configuration 1.
See attached file Eng-Tips_Thread238-279051-01.xlsx
Two 4.5m long, 8mm DIA electrodes spaced 5.83 meters apart in soil with uniform resistivity of 100 Ohm-m. The total earthfault current ( 3Io ) from the source of 443A splits equally between the earthgrid and the overhead earthwire ( OHEW ).
The SKM Ground Mat study results show that touch voltages exceed the ‘tolerable touch voltage’of 351V. Further earthgrid enhancement is required.

Configuration 2.
See attached file Eng-Tips_Thread238-279051-02.xlsx
The earthing grid comprises of a rectangle 5m x 3m with horizontal earth conductors 1.0m apart. At each corner of the grid is a 4.5m long, 8mm DIA electrode.
The total earthfault current ( 3Io ) from the source of 443A splits equally between the earthgrid and the overhead earthwire ( OHEW ).
The SKM Ground Mat study results shows the area where the touch voltages are below the ‘tolerable touch voltage’of 351V w.r.t remote earth. The boundary for thesafe area is defined by the yellow plane and includes all the area inside the earthgrid and a small area outside the grid.
All step voltages are below the ‘tolerable step voltage’of 351V.

 

[EddyWirbelstrom]

See attached file Eng-Tips_Thread238-279051-02.xlsx for configuration 2.

 

[EddyWirbelstrom]

Please delete rows 54 to 91 of the Input Data tab of the En-Tips_Thread238-279051-01-01.xlsx