Complex motor start cuasing RCBO to trip problem

[pmgd]

HI,

I have a difficult to diagnose problem. I've had two qualified electricians try and sort it, but no solution yet.
It is tricky (interesting?) and would love to hear comments/suggestions.

I have a long barn conversion.

Supply enters one end into a CU. There is a 50A 30mA RCBO feeding a 16mm T&E which goes about 50yds to an annex where it terminates in a standard split load CU. (Two 30mA RCDs with assorted MCBs each side).

Because the earth conductor in 16mmT&E is of insufficient size, an earth spike is additionally connected at the annex end.

One MCB supplies a 'pump room' which supplies a bore hole pump and - recently - a variable speed water pressure pump.

Every time the variable speed pump starts it blows the RCBO. (Never the RCD).

But, it is not faulty. If plugged into a regular socket on my main house ring main it works fine.

I have a socket near the main CU also connected to the RCBO - for test purposes.
If the pump is plugged in there it trips the RCBO ONLY IF the annex DB main switch is on, and at least some MCBs are enabled.

The situation is fully reproducible - although putting MCBs on in turn gives inconsistent results.

My personal view is that this is related to the earth spike having a substantially lower inductance to earth for the startup motor current and 30mA is current flowing through some domestic appliance in the annex to the spike. However, this is difficult to argue as the problem occurs when the pump is pugged in at either end of the 50yds 16mm cable.

Supply appears to be TN-C-S (although there are 3 cables coming from the transformer outside).

Any clever ideas?

Peter

PS All Zs etc within spec. 1V shown on test equip between N and E in whole installation!

 

[pmgd]

Def not the pumps fault as if plugged into a totolly seperate building - or even the main CU of my own - it all works fine. I cannot get over thinking it is somthing to do with the earthing of the annex. However - I will read your essay - slowly and soberly tomorrow darkwood. Thank you VERY much for taking the time ....

 

[Knobhead]

Can't you get it in you’re head:
Without the VSD the pump runs OK. Using the VSD the supply trips!
VSD’s have an inherent leakage, if you look through the literature for it you will find it!

 

[Guest55]

i like how tony answered the query perfectly well in post #4 yet has been completely ignored by everyone else since lol

Here T , have a facepalm on me........

 

[Knobhead]

 

[pmgd]

Tony, biff55: I'm afraid that I disagree.
Yes VSD do present a challenge. I have researched them on the net quite a lot. I talked to the manufacturer yesterday.
The challenge they present appears to be due to the particular waveform they generate and perhaps common mode noise. NOT simple leakage.

Also - the pump works fine in a different socket. Doesn't that rather suggest that it is the socket at fault not the pump.

Also - I've had nuisance tripping in the annex for a year - apparently related to nothing we could identify. This was when the VSD did not exist! -- but a bore hole pump (regular capacitor start) did. Clearly something is wrong with the supply, we are just using the VSD to locate it.

So - NO I am not just going to swap out the VSD. I want to fix the supply.

Yes - I suspect giving the pump room its own (non RCD) supply will fix it. However - as I already said that is extremely difficult. Although I may be forced to spend the money eventually - I'd prefer to try without.

 

[snowhead]

What is in the pump room that must be on an RCD?

 

[pmgd]

17th edition regs unfortunately. It is fed with T&E which is behind plasterboard and within 3cms (?) of the surface and is not otherwise mechanically protected. Unfortunately!!!

 

[Guest55]

17th edition regs unfortunately. It is fed with T&E which is behind plasterboard and within 3cms (?) of the surface and is not otherwise mechanically protected. Unfortunately!!!

aah , bad design then.
you should design out the need for rcd's for equipment that have high leakage current at the installation stage.
your problems have then been further compounded by the selection of inappropriate protection devices.
back to the drawing board me thinks.

 

[Darkwood]

Tony, biff55: I'm afraid that I disagree.
Yes VSD do present a challenge. I have researched them on the net quite a lot. I talked to the manufacturer yesterday.
The challenge they present appears to be due to the particular waveform they generate and perhaps common mode noise. NOT simple leakage.

Also - the pump works fine in a different socket. Doesn't that rather suggest that it is the socket at fault not the pump.

Also - I've had nuisance tripping in the annex for a year - apparently related to nothing we could identify. This was when the VSD did not exist! -- but a bore hole pump (regular capacitor start) did. Clearly something is wrong with the supply, we are just using the VSD to locate it.

So - NO I am not just going to swap out the VSD. I want to fix the supply.

Yes - I suspect giving the pump room its own (non RCD) supply will fix it. However - as I already said that is extremely difficult. Although I may be forced to spend the money eventually - I'd prefer to try without.

As i pointed out with a solution that keeps rcd cover ...

 

[pmgd]

Darkwood: Many thanks for the long explanation yesterday. I've now read it in detail. A couple of points:

1. FYI an adjacent property had a problem whereby use of a large flymo in house A cause RCD to trip in house B. (Mine is house C). A little like the scenario you mention.

2. You say (imply?) that there are clever RCBOs available which 'recognise' the inverter maveform and discount it. If so - please can you point me at a spec/partnumber etc?

3. Electrician coming next week - will do all the tests. Meanwhile I have just received a clamp meter and will do some more accurate measurements myself.

4. "Vetted". Well Chartered Engineer & elecSa approved is all I know.

 

[Darkwood]

Darkwood: Many thanks for the long explanation yesterday. I've now read it in detail. A couple of points:

1. FYI an adjacent property had a problem whereby use of a large flymo in house A cause RCD to trip in house B. (Mine is house C). A little like the scenario you mention.

2. You say (imply?) that there are clever RCBOs available which 'recognise' the inverter maveform and discount it. If so - please can you point me at a spec/partnumber etc?

3. Electrician coming next week - will do all the tests. Meanwhile I have just received a clamp meter and will do some more accurate measurements myself.

4. "Vetted". Well Chartered Engineer & elecSa approved is all I know.

Regarding your point 1 that's what I briefly suggested earlier regarding proximity of transformer but with your added info regarding neighbours then this is more than likely the cause... ive heard of this problem before with neighbouring properties creating nuisance tripping in a different building, its possibly caused by issues between the properties using a common transformer that's in close proximity the exact cause was never identified but because it may be a neighbouring installation or the DNO's issue it gets very hard to resolve.

Point 2- Schneider do an RCCB in their SI range that's designed to recognise the generated earth leakage from power drives, computers, electronic lighting etc ---product code 25524 - this is a 40amp double pole RCCB unit so it would be fitted in a separate enclosure and doesn't provide overload protection so you feed it from a standard mcb then connect load to outgoing side.

Point 4.... let him peruse this thread, its not common knowledge this issue your having so may head him in the right direction. Note this hopefully will eliminate the inverter causing tripping but may not stop the other nuisance tripping issues you had although it does filter stray frequencies and may stop the nuisance tripping caused due to neighbouring installations and proximity of transformer but I cant say if it will or not.

 

[lewlec]

Hi
Just wondering did you get any IR tests done?

 

[pmgd]

Hi All,

well I did have an IR test scheduled but I have just cancelled it.

What I didn't mention was that the annex in question is rented out as a holiday home - and I really would prefer to not disturb them. I'll wait until november when we may have a spare week.

Also, I followed advice here and purchased some "proper" test equipment. Specifically a min/max clamp meter.

I discovered that the Variable speed pump has a max of 35mA at startup, and 20 thereafter. Not sure why my ammeter in the earth line didn't notice this, but there you go.

I have two solutions at the moment - both rubbish, but they work:

1. Connect the whole of the pump room to the other property - and show the problem doesnt recur .. thus blaming the pumps.

2. Replace the RCBO with an RCD and MCB. Logically equivalent, but I have experimental observations which says one blows and the other doesn't. [ I do notice the RCBO has an earth lead which according to literature enables it to 'blow faster'. ]

I still have a very funny earthing system.

There is one more thing I'd like advice with however:

With the new clamp meter I measured the current in the earth spike lead. Small 3mA. OK. But then I realised that my property and the property next door share a gas supply. The supply comes in copper into each property. In accordance with best practice installers have bonded these incoming copper pipes to the local earth.

To my mind this then creates more opportunity for ground loops etc etc. I measured the current in the gas bonding line. It was non-trivial. Varied but perhaps 10-20mA.

Any comments?