LV fault tripping HV ring main unit

[Daveash123]

Hi all, I have been on a site repairing damage caused by travellers collecting copper for the last week. We have found and replaced cabling to get the site up and running. Power has been on for a few days but today the 11KV ring main unit tripped. This feeds a transformer attached to the main switchboard. We called in an HV engineer to assist. We carried out testing and pinpointed the cause of the trip to a single phase 63A type D MCB in a remote DB. The circuit is not marked so it was disconnected until we can investigate tomorrow. We restored the power and left site for the night.
Both myself and the HV guy have never come acrosss this before and was wondering if anyone on the forum has any ideas. The fault could be a cut cable somewhere and its been disturbed or got wet today. The fault is bypassing a 63A MCB, 200A MCCB on a section board, 400A HRC fuse in the main switchboard and the 3200A ACB before tripping the ring main unit.
Obviously we will be carrying out full testing tomorrow to identify this circuit though the site is a former manufacturing plant with cables everywhere, both redundant and in use and the marking of the boards has never been updated
Any advice appreciated

 

[Julie.]

What type of protection is on the rmu?

Size of transformer, settings on Protection etc.

Could be over reach, but need setting etc to see.

 

[Daveash123]

I don't have access to that information, not allowed to access the HV side of the switchroom. The installation has been in use since 2006 and I assume the settings would not have been changed since then. The transformer supplies the switchboard via a 3200A ACB. I can investigate further in the morning
Thanks for the reply

 

[Julie.]

OK.

Let's say the transformer is 500kVA and is 5% impedance, then the maximum fault level is flc/5% or 20x

Since 500kVA is just over 26A at 11kV, this means any fault on the secondary side cannot result in a fault current more than circ 525A on the primary.

So you would normally set the instantaneous trip level greater than this (by a margin as the CTs would have a tolerance of say 5% - 5p10 CTs for example)

In this way if there is a fault on the secondary, the fault level seen on the primary would be lower than 525A and the instantaneous protection will ignore it, however if it persists the curve portion of the primary protection will operate after a time delay.

If the fault is internal to the transformer, the fault level is not limited to below 525A, so the instantaneous protection will operate immediately for internal faults.

BUT, if you set the instantaneous below 525A on the primary, then if you have a falult on the secondary (even downstream of a mccb) although the protection on the secondary will clear it, the instantaneous on the primary will operate, basically once it starts to operate, it will finish!

This is called overreaching.

I would hazard a guess that this is the problem, but need the protection settings, and transformer size & impedance to know.

 

[Daveash123]

Thanks for that, we have a HV specialist coming on Monday, I will pass on your advice
Thank you

 

[Julie.]

Thanks for that, we have a HV specialist coming on Monday, I will pass on your advice
Thank you

That's my first guess, although it could be CT unbalance - no stabilising resistance fitted, or incorrect setup on the unit protection etc.

Basically it depends on the type of protection fitted, smaller transformers though larger ones, may just have: 11kV fuses, or time delay fuses on CTs, or an IDMTL relay (with instantaneous), or IDMTL relay (with residual) or IDMTL plus differential ....... etc

Knowing the protection fitted is needed to narrow down the options.

 

[pc1966]

Thanks for that, we have a HV specialist coming on Monday, I will pass on your advice
Thank you

Please let us know what comes out of it.

 

[Rockingit]

Following this with interest, and fond memories of Tony chiming in with a diagram or four.

 

[Lister1987]

OK.

Let's say the transformer is 500kVA and is 5% impedance, then the maximum fault level is flc/5% or 20x

Since 500kVA is just over 26A at 11kV, this means any fault on the secondary side cannot result in a fault current more than circ 525A on the primary.

So you would normally set the instantaneous trip level greater than this (by a margin as the CTs would have a tolerance of say 5% - 5p10 CTs for example)

In this way if there is a fault on the secondary, the fault level seen on the primary would be lower than 525A and the instantaneous protection will ignore it, however if it persists the curve portion of the primary protection will operate after a time delay.

If the fault is internal to the transformer, the fault level is not limited to below 525A, so the instantaneous protection will operate immediately for internal faults.

BUT, if you set the instantaneous below 525A on the primary, then if you have a falult on the secondary (even downstream of a mccb) although the protection on the secondary will clear it, the instantaneous on the primary will operate, basically once it starts to operate, it will finish!

This is called overreaching.

I would hazard a guess that this is the problem, but need the protection settings, and transformer size & impedance to know.

Not completely understanding everything there but find myself slightly aroused please tell meant that only one? Something about a lady knowing thier stuff seems to do it, no idea why

 

[Moley]

Something about a lady knowing thier stuff seems to do it, no idea why

Because it's such a rare event?
I think I had better take cover now

 

[Julie.]

Not completely understanding everything there but find myself slightly aroused please tell meant that only one? Something about a lady knowing thier stuff seems to do it, no idea why

With power transformers, one test done is to short out the secondary, and apply voltage on the primary, increasing it until you get to 100% full load current.

If this occurs at 10% of the rated voltage, then the transformer is said to have a "voltage impedance " of 10%

This actually means that if you put 100% voltage on a shorted transformer, you would get 100% Volts/10% impedance = 10x full load current.

10% = 10x
8% = 12.5x
5% = 20x
4.75% = 21x
And so on.

This means the maximum fault level through a transformer is limited predominantly by its own impedance - look at the rating of the transformer and you know the maximum fault level that could pass through the transformer.

In practice the level will be slightly lower than this due to cables, upstream transformers etc - but is a really good estimate from just the kVA and impedance.

So a 1MVA 11kv tx with 8% would be a maximum of:
52.5A (flc @ 11kV) x 12.5x = 656A @ 11kV
1445A (flc @ 400V) x 12.5 = 18kA @ 400V

In practice, much less than this, especially further from the transformer.

If you measure more current than this, the fault cannot be downstream of the transformer, so won't be cleared by any secondary side protection. It must be internal to the transformer, and you can operate the primary side protection instantly.

Generally distribution transformers have very simple protection - just 11kV fuses which like LV fuses have a single characteristic curve, but also have a small explosive charge, this operates a pin which is fired out of the fuse which trips all phases - I doubt it is this in this case as you don't generally overreach.

Time limit fuses are an older type of protection, which again I doubt it is this in this case as you don't generally overreach.

The most likely protection in this case would be an IDMT or IDMTL (InDependent Minimum Time [Lag]) relay with an additional instantaneous setting.

This type of protection has a characteristic curve (similar to a MCB) but which can be "moved around" in time (faster or slower), and in current level (higher or lower current) - like some MCCBs with electronic protection, but also you can change the actual shape of the curve itself (~4-5 selectable curve types).
The instantaneous protection is basically what it sounds like, another setting, independent of the IDMTL part where you can set a current value (above the maximum through fault level described above) where if it sees this it operates instantly. (You can also get time delay functions on this, but you wouldn't use them in this application)

For larger transformers or more important ones, there are other protection options available, I could bore anyone to death with the possibilities, but it's too long a post already, and you are probably already asleep!

 

[Rockingit]

I could bore anyone to death with the possibilities, but it's too long a post already, and you are probably already asleep!

On my second coffee and wondering how you do the process of shorting out and reaching 100% load - I assume by calculation of the correct reduced primary voltage versus the measured impedance of the dead short?? Because otherwise you just get meltdown, surely?

 

[pc1966]

On my second coffee and wondering how you do the process of shorting out and reaching 100% load - I assume by calculation of the correct reduced primary voltage versus the measured impedance of the dead short?? Because otherwise you just get meltdown, surely?

At 100% load in to a short you are only putting in 5-10% voltage to the primary, and also a significant portion of the transformer's impedance is inductive so you are only dissipating a couple of percent of nominal power (mostly copper losses in the windings I think, from distant memory of when I studied such things).

 

[pc1966]

We called in an HV engineer to assist. We carried out testing and pinpointed the cause of the trip to a single phase 63A type D MCB in a remote DB. The circuit is not marked so it was disconnected until we can investigate tomorrow.

The fault is bypassing a 63A MCB, 200A MCCB on a section board, 400A HRC fuse in the main switchboard and the 3200A ACB before tripping the ring main unit.

It might seem like a basic question, but is it possible the RMU has some external monitoring/control and that 63A MCB fed some sort of panel for that?

 

[Daveash123]

As far as I know it does not have any monitoring. There are no control cables between the main switchboard and the RMU

 

[pc1966]

Personally I find it hard to see how the RMU could be tripped by a MCB fault. At, say, 10kA MCB breaking limit (as I assume the related panel has a lower PFC...) that is just 3x the ACB's rating so it ought to be well under the HV side protection threshold.

So either the HV side is broken or very badly configured (which you would think would have been noticed when previously in use), or that MCB's circuit is somehow upsetting it via a control signal or somehow tricking it in to thinking there is an earth-fault or similar.

 

[pc1966]

Just to add this is way outside of my area so believe anything @Julie. tells you before what I have to say!

 

[Daveash123]

Update on this, we have found out that there is a small panel within the switchroom that is connected to the RMU, no one on site knew about this and the guy that attended the initial fault and reset the RMU didn`t know either. It looks like a power supply with batteries but it is not labelled and is at the opposite end of the switchroom from the RMU. We have assumed that the feed to this has been cut and when the batteries discharged the trip occurred. However, it has been off for a while now and not tripped since. We have installed a new supply to the panel but leaving it off until the specialist can confirm what it does. He was due yesterday but didn`t turn up. The 63A MCB was causing the trip but only because it supplied the DB that supplied the panel in the switchroom.
Thanks for all comments and advice