Trip readings too high on ground fault relay with Toroid sensor......Help?

[Hobsoni]

Hi there, Have been carrying out testing on some 3 phase sub-DB's that have outgoing supplies to 3-phase DB's for accommodation blocks. My problem is to do with readings I'm getting on a ground fault relay which has been set to 300mA. Basically, the incoming cable to the Sub DB ( 4-core 25mm SWA) enters the panel and passes through a sensor (toroid) and then terminates into an MCCB. As per the installation requirements for this Toroid device, 3-phases and neutral pass through the centre and the CPC does NOT pass through the device. From this Toroid are some control cables to a Relay which can be set (mA) and time delayed (0.06, 0.1 ms etc) and also from the relay are outgoing cables to the incoming MCCB (I presume Shunt control?) and also from the relay to the outgoing L1 terminal of the incoming MCCB. In a nut shell, whilst performing trip times on the ground fault relay I cannot get trip times less than >310.0ms whether I test at 300mA or 500mA. The device does not trip at half the rated current which is good. I am absolutely stumped, because there are 5 of these devices and each and everyone is the same!? All other readings obtained have been absolutely fine, ingoing loop to DB etc etc. I am absolutely perplexed, can someone please shed some light on what it could be please. Many thanks

 

[sparky6078]

is this with the time delay still on? or did you turn it off?

 

[Hobsoni]

The time delays were pre set at 0.06. However, as soon as I got a reading of >310 ms I turned off the time delay setting. Still the same trip times obtained. I ramp tested every unit and they trip out between 300-315 mA. Oh, for the sake of clarity..........I used two different testers (Megger and Fluke, both calibrated and still the same outcome) ??

 

[spark 68]

What make and model are the relays ?

 

[Hobsoni]

Schneider TA30 Toroid device and a Schneider RH99 Ground Fault Relay.

 

[spark 68]

Some of these protection schemas do not test properly with a normal MFT, some have their own test routines, it may be worthwhile consulting Schneider technical for more information.

I had a quick look at some technical data, but this consists of a 100 pages plus lol.

 

[Hobsoni]

That's an interesting point actually, because whilst I 'glanced' through some datasheets I came across the term 'Primary Injection Tester'.......which in my simple mind would be an RCD tester? Never straight forward though sadly, although technically you would think if you inject 300mA of fault current through a conductor.........it would replicate the same type of 'current' during a 300mA fault. So perhaps technical advice is the only answer here! Thanks for the assistance anyway guys.

 

[spark 68]

Some required primary injection, some you can use secondary injection, some types allow both, some test routines use a seperate timer, unfortunately the Schneider site is bloody awful to use.

During the quick look I had I noticed there was a self-test routine, a push button to connect to two terminals, but I don't have enough information at the moment to know if this is a valid test, Schneider spread the information all over the place.