Ze readings >2000Mohms

[ELECELEC]

Fluke MultiFunction Tester 1662

TNS Earthing system.

Switchboard - 415v - 3 Phase + Neutral + Earth.
MCB 1000A
3 individual Earths (185mm) connected to the Switchboards Earth Busbar via a main earth bar which is then connected to the sites Earthing Grid that is bonded with exothermic welding.

Performed Ze reading on main incoming supply to Switchboard with Earth disconnected, without any issues.

Ze - L1 - 0.13 PEFC - 1.9 kA
Ze - L2- 0.13 PEFC - 1.9 kA
Ze - L3 - 0.13 PEFC - 1.9 kA

The issue is only apparent on some of the outgoing supplies.

Lighting board A with earth disconnected on incoming supply - Ze readings -

Ze L1 - 0.27 PEFC - 891A
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

Lighting Board B with earth disconnected on incoming supply - Ze Readings -

Ze L1 - >2000Mohm
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

Lighting Board C with earth disconnected on incoming supply - Ze readings -

Ze L1 - 0.25 PEFC 960A
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

HVAC Switchboard with earth disconnected on incoming supply Ze Readings -

Ze L1 - 0.13 PEFC 1.9kA
Ze L2 - 0.13 PEFC 1.9kA
Ze L3 - 0.13 PEFC 1.9kA

For the life of me, I can't understand why I can get Ze readings on some outgoing supplies, and not the others.
And in some cases only getting a Ze reading on L1, but not on L2 + L3.

I've tried various other Multifunction Test equipment and issue remains.
I've had an outage on the board, inspected the incoming supply, internal busbars and torqued everything.
Performed a DUCTOR test on the board, without any issues.

Any help would be appreciated

Kind Regards

 

[davesparks]

That's some interesting readings!

A Ze of 0.13 is way too high for a 1000A supply and a PFC of 1.9kA is way too low. How far away (in terms of conductor length) is the supply transformer? The inductance of the transformer may be screwing with your test results.

It would help if we you could clarify exactly what you are testing and what is being disconnected for these tests?
Ze is only measured at the incoming supply position with the installation isolated and the means of earthing disconnected.
All other earth fault loop impedance tests are Zs and, as they require the installation to be live must be conducted with all earthing and bonding connected. Plus they must only be carried out after dead testing has proved the continuity of the earthing up to that point.

 

[ELECELEC]

That's some interesting readings!

A Ze of 0.13 is way too high for a 1000A supply and a PFC of 1.9kA is way too low. How far away (in terms of conductor length) is the supply transformer? The inductance of the transformer may be screwing with your test results.

It would help if we you could clarify exactly what you are testing and what is being disconnected for these tests?
Ze is only measured at the incoming supply position with the installation isolated and the means of earthing disconnected.
All other earth fault loop impedance tests are Zs and, as they require the installation to be live must be conducted with all earthing and bonding connected. Plus they must only be carried out after dead testing has proved the continuity of the earthing up to that point.

My mistake the MCB for the board is 500A.

The Distribution board itself is downstream from another distribution board (800A MCB), which ties into a transformer via busduct.

2 single core 300mm cable for each phase + neutral at a length of 32M.
1 single core 300mm cable for Earth at a length of 32M.

We are required to provide Ze readings for each LV Distribution board on site.
Incoming supply Earths are always disconnected, and Main Circuit Breakers / rotary isolators are always switched off + isolated before taking a Ze reading.

This specific is task is to capture Ze reading on the distribution (500A MCB) and then capture Ze readings for all downstream LV distribution boards - 3 Lighting DBs and 2 HVAC DB's.

Yesterday, We revisited a lighting distribution board (which is an outgoing supply of the board in question) that we previously had acceptable Ze readings for. The circumstance of the Distribution board upstream of the Lighting Board had 0 load, acceptable Ze readings were obtained. Tested again yesterday with load on the upstream board of 200Amps and the Ze readings were all >2000Mohms.

 

[James]

@Julie. will be able to give a much better description than I can
however, in brief when you are that close to the transformer the readings from a multi tester are useless because they can only measure the resistance element of the circuit and not the inductance.

 

[pc1966]

The description of what you are doing does not make much sense to me, Are you trying to measure the Ra of the sites Earthing Grid? Having that is the region of 0.13 ohms is sort of sensible. But that is not Ze which is relative to the TN-S supply's earth point and for a 1000A system it ought to be far less and your PFC far bigger. Probably above what a MFT can accurately resolve!

As for the odd differences when testing in different places, have you tried new test leads?

 

[davesparks]

My mistake the MCB for the board is 500A.

0.13 still aounds too high for the Zs

The Distribution board itself is downstream from another distribution board (800A MCB), which ties into a transformer via busduct.

2 single core 300mm cable for each phase + neutral at a length of 32M.
1 single core 300mm cable for Earth at a length of 32M.

We are required to provide Ze readings for each LV Distribution board on site.

Again, that is Zs and not Ze, Ze is the external earth fault loop impedance.

Incoming supply Earths are always disconnected, and Main Circuit Breakers / rotary isolators are always switched off + isolated before taking a Ze reading.

How are you actually disconnecting the incoming earth to these DBs? This isn't normally possible.

Where exactly are you connecting the test leads when taking these measurements?

 

[ELECELEC]

0.13 still aounds too high for the Zs


Again, that is Zs and not Ze, Ze is the external earth fault loop impedance.



How are you actually disconnecting the incoming earth to these DBs? This isn't normally possible.

Where exactly are you connecting the test leads when taking these measurements?

It's not a Zs reading. Why would we be disconnecting the main earth of the supply cable to the distribution board to record a Zs?

We've captured Ze readings on over 100+ DB's on this PowerStation thus far.

This isn't an issue of where the test leads are being connected.

I think I'm starting to agree with the distance of the Transformer / length of the conductors.

 

[ELECELEC]

Next thing I think I'm going to do is have no load on the Main DB, as there's roughly around 200A being pulled at present. then try and record Ze readings for each of the downstream DB's, see if that changes anything. Maybe the running of one of the HVAC boards is interfering with the readings.

 

[davesparks]

It's not a Zs reading. Why would we be disconnecting the main earth of the supply cable to the distribution board to record a Zs?

Ze is the earth fault loop impedance external to the installation, you are measuring earth fault loop impedance within the installation so normally that is Zs, hence why I have bee saying Zs. But now you've revealed this information:

We've captured Ze readings on over 100+ DB's on this PowerStation thus far.

It's a PowerStation, that's a whole different kettle of fish and things can be very different.

Is there not an engineer there who could shed some light on the issue?

This isn't an issue of where the test leads are being connected.

OK, but it is always worth checking the simple things first as we all make that kind of mistake sometimes. I've spent ages in the past looking for a similar issue until I realised I'd got the tester set to a 2 wire low current test instead of a normal high current test.

I think I'm starting to agree with the distance of the Transformer / length of the conductors.

The distances you've mentioned above are normally plenty to get far enough from your average substation transformer to get sensible readings.

 

[ELECELEC]

is always worth checking the simple things first as we all make that kind of mistake sometimes

Totally agree, first thing I did was change the leads / meter and check the settings. Same outcome.

It's a PowerStation, that's a whole different kettle of fish and things can be very different

Yes a huge difference, and is quite difficult to explain the exact installation with words / without uploading a one liner. Multiple 415v switchboards feeding multiple downstream 415v switchboards, which then feed multiple downstream distribution boards e.g lighting DB’s.

It’s the engineer that has requested Ze readings on all switchboards / distribution boards, excluding the switchboards connected to the step down transformers via bus duct. The previous commissioning team failed to take any results on first energisation.

 

[Julie.]

Fluke MultiFunction Tester 1662

TNS Earthing system.

Switchboard - 415v - 3 Phase + Neutral + Earth.
MCB 1000A
3 individual Earths (185mm) connected to the Switchboards Earth Busbar via a main earth bar which is then connected to the sites Earthing Grid that is bonded with exothermic welding.

Performed Ze reading on main incoming supply to Switchboard with Earth disconnected, without any issues.

Ze - L1 - 0.13 PEFC - 1.9 kA
Ze - L2- 0.13 PEFC - 1.9 kA
Ze - L3 - 0.13 PEFC - 1.9 kA

The issue is only apparent on some of the outgoing supplies.

Lighting board A with earth disconnected on incoming supply - Ze readings -

Ze L1 - 0.27 PEFC - 891A
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

Lighting Board B with earth disconnected on incoming supply - Ze Readings -

Ze L1 - >2000Mohm
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

Lighting Board C with earth disconnected on incoming supply - Ze readings -

Ze L1 - 0.25 PEFC 960A
Ze L2 - >2000Mohm
Ze L3 - >2000Mohm

HVAC Switchboard with earth disconnected on incoming supply Ze Readings -

Ze L1 - 0.13 PEFC 1.9kA
Ze L2 - 0.13 PEFC 1.9kA
Ze L3 - 0.13 PEFC 1.9kA

For the life of me, I can't understand why I can get Ze readings on some outgoing supplies, and not the others.
And in some cases only getting a Ze reading on L1, but not on L2 + L3.

I've tried various other Multifunction Test equipment and issue remains.
I've had an outage on the board, inspected the incoming supply, internal busbars and torqued everything.
Performed a DUCTOR test on the board, without any issues.

Any help would be appreciated

Kind Regards

Ok, there is much to unwrap here – and I am struggling to do so!



Let’s try talking it through – why do we want the Zs - there are two reasons, and let’s apply them to a domestic situation then industrial, then generation.



Firstly

We need to find out the MINIMUM fault level so we can confirm the protection operates (this is the same at transmission, distribution and domestic levels – and everywhere in-between).

To do this we normally remove any paths which we cannot guarantee will always be there – such as in domestic where the Bonding connects the MET to the water pipes, which connects to the boiler, which is connected to the FCU and hence the board via the CPC of that circuit, or the supplementary/EP bonding in the bathroom connecting the radiator to the lighting circuit and the boiler, through the water to the MET etc etc.



The easiest way is to disconnect the CPC from the board for the circuit under test, and measure the R1 + R2, noting we must ensure in doing this that there is NO other connection to earth, then add it to the source impedance (usually measured as Re) to get Zs – providing this is low enough to operate the associated fuse, MCB, MCCB…. or whatever) we are golden.

In a domestic this is generally all we need to do.



The second reason.

…Is to determine the MAXIMUM fault level to ensure the equipment is able to interrupt the circuit.

In this case we actually need everything in place which may contribute to this level – usually this is just ensuring all parallel paths (such as the circuit CPC and bonding with all circuits connected etc), however when we deal with industrial and generation sites, this may not be sufficient.

We could just carry out R1+R2 tests (added to Re), but with everything remaining connected to ensure all paths are in place, or more simply Use a MFT to measure Zs directly again with all paths in place.

In domestic this isn’t usually necessary as firstly we have one source Re through one main fuse into a CU which is type tested to coordinate with the standard supply fuses at all likely fault levels, and secondly, we only need to measure at that one point (the CU).



In industrial however, this is often required as we cannot guarantee such coordination, so it does have to be measured, (and why industrial kit may be 3kA, 6kA, 10kA, 16kA…..), but also remote switchboards may actually have lower fault levels - for example the MCCBs in the main switchboards need to be 50kA capacity (including the feed to the remote switchboard), but the remote switchboard itself may only need 16kA MCBs, and another switchboard downstream only 10kA etc etc.

So, if we are looking for Zs and it’s for minimum fault level we ought to use Re + R1+R2, we shouldn’t really measure Zs directly.

But, if we are looking for Zs and it’s for maximum fault level we ought to measure Zs directly, and we shouldn’t really use Re + R1+R2.



One issue with industrial is that there is often a huge bank collectively of motors, both large and small, which when running actually act like generators when a fault is applied - usually this doesn’t impact the steady state fault level, but it does significantly impact the initial fault level.


This is why there is a difference between switchgear used in domestic and industrial – a 6kA domestic MCB would have to withstand something like 1.5 x its braking capacity (6kA = 9kA) over the first few milliseconds – which has generally fallen to 6kA by the time it actually opens. However, an industrial one due to the motor contribution may have to withstand 2.5 x its braking capacity over the first few milliseconds – which will not have actually fallen to its rated braking capacity by the time it actually opens - this is fine because the standards for industrial MCBs take this into account, and they are tested to confirm this.



This is actually worse on a generation site – you measure Re when it’s connected via a weak link to the national grid (and hence the generation), but when it’s running the local generation provides well more fault level than the grid would – the design must take this into account.



So all the drivel done with, in your case

• What are you measuring the Zs for – why does the Engineer want this – it’s likely to be for minimum, but it matters as to how you configure for your tests.
• You appear to be measuring Zs at a switchboard rather than Re – but with earths disconnected – if the earth is disconnected properly I would expect Zs = infinite, if it isn’t then the earth is not disconnected
• If this is an active site remote activities can impact your readings (you disconnected the earth, but remotely someone placed a spanner or something between metal clad switchgear and the metallic framework of the building – actually linking your “disconnected” earth to earth – many many reasons)
• When a MFT measures the Re (or Zs) directly, it does so by applying a number of direct short circuits – but of very brief durations, noting the spike in current and estimating the fault level - if this is done on a site with major contribution via motors or generators this could be totally un-representative of the steady state value – the reading could be wildly out (too high or too low) – A MFT effectively measures a very peak value, and back-estimates this to what normal steady state fault level would produce this sort of peak. If the X/R ratio of the location is significantly different than a “normal” X/R ratio such as directly on a transformer, or having local generation (true generators or motors) then the results can be spurious.

 

[pc1966]

I can't see any reasonable explanation for Ze over 2000 ohms other than something is open.

Sure the X/R aspect close to a transformer, etc, might alter the MFT accuracy but that is going be be relatively small, say 0.15 ohm actually being 0.1 or 0.2, that sort of magnitude.

 

[Julie.]

Just thought I might add some other boring information.

So if you short a generator directly, theoretically the current in each phase would be like this:





Thus, the peak current could be over 2.8 (2 x sqrt(2)) the steady state value, in practice slightly less, and the decay rate changes throughout the network as the X/R ratio changes.

So equipment near generation needs to break whilst the current is still falling from it's initial peak, towards the steady state. Remote kit (such as domestic) falls quickly, so the equipment only needs to interrupt close to steady state values as by the time it's opened, the initial peak has died away.

To see this in practice, here are some results from testing a 12kV 50kA circuit breaker (it was my design btw)





In this case, which is a through-fault (that is some other protection/breaker is expected to clear it) the fault current is 127kA peak, decaying to a steady state level of 53.3kA for a duration of 3.08 seconds. (The standard is 125kA/50kA/3.0secs) interesting to note by the way is that the high eddy currents caused arcing around the earthed enclosure (which is pretty normal btw).




In this case, the breaker closed onto a fault at 51ms, and opened 36ms later.

In this case the voltage was 15.2kV, the currents 127kA, 131kA, and 90kA, and the arc times (from opening) 2ms, 7ms, and 7ms per phase.

This was a design to retrofit old 12kV switchgear in both power stations and large industrial sites - this test was for industrial level equipment, but we also tested it for higher X/R associated with power stations.

I am just boring people now!

 

[ELECELEC]

Thanks for your replies guys.

Did some more digging.

Looks like the downstream HVAC board is interfering with our readings, but still don't understand why.

High load - HVAC MCC in service = 198A at upstream MCC - No Ze being recorded on any of the outgoing supplies (3 lighting DB's). >2000Mohms across 3 phases.

Low Load - HVAC MCC isolated = 38A - at upstream MCC - Ze readings recorded on outgoing supplies / all acceptable ( 3 lighting DB's)

Is this pointing towards the MFT not being able to record External earth loop paths at high load currents?.

High resistance possibly? Will try Thermographic reading on the incoming supply to rule this option out.

 

[bennie]

Sorry, I've read all this but I still don't understand why your disconnecting the earth at each DB to carry out the test. What cable type of cable is supply them? And are you turning the DBs off before you carry the test out? Also are you reconnecting the earths before you move on to the next test?