Choosing setting for industrial residual earth fault trip 7

[brianthegiant]

I am using a residual earth fault trip which fits onto the incomer of a 3P&N 400A MCCB; it uses a core balance ct around the 3 phases, I'm not given ct values because it comes as a package; I just have to choose the earth fault trip setting eg 0.03A-5A & trip time (0.01s-1s). The 'load' is 230kW of solar pv inverters but from the point of view of earth protection they might as well be 230kW of computer power supplies (ie they use PWM & generate harmonics from the IGBT's & earth leakage from the filter capacitors).

I know that a low setting (0.03-0.1A) will probably cause nuisance tripping due to the inverters (but not how to calculate the minimum), and that the the maximum earth fault multiplied by the maximum earth loop impedance should equal a safe value of voltage (<50v); but I've a feeling there's more to it than that?

 

[RRaghunath]

The triplen harmonics in the supply should cause concern as the same are reflected on CBCT secondary side and could cause nuisance trip.
Sensitivity point of view, choosing the max setting of 5A and 1sec should not be a problem, as 5A is presumably the primary current.
However, the right thing would be to measure the current on the CBCT secondary when the load is on and setting the relay suitably.
May be, the current setting should be double the measured current, just to take care of worst case scenario.

 

[NCTHAI]

as a general practice it is recommended to choose 10% o/c for earth fault setting. e.g. if your o/c setting is 10A, choose 1A as earth fault setting. If this relay is in series with other relays, co-ordinate with other relays.

Time delay - 1 sec or co-ordinat with other relays.

 

[burnt2x]

brianthegiant,

Please expound. Not much data (maybe reading too fast on this post)
However, if the said package comes with the usual EF relay characteristics, please check if EF relay has built-in harmonics filter, no worry on harmonics going thru EF if with filter.

Core balance CT will sum up vectorially the 3 line amps. The CBCT output will be more noise during fault-free conditions. Try calculating the 3Io(also called residual)assuming an SLG on the load. General practice, as previous posts recommend 10-30% of this residual amps. Set to 10% of this calculated residual amps. If spurious trips occur, try a bit higher setting. Definite time trip should be as short as possible not to exceed 1 sec. Or better still, try it out starting at the lower settings until no spurious trips occur.

 

[ScottyUK]

Consider that if there is no downstream E/F protection at the individual loads then you are effectively looking at coordinating whatever O/C protection is at the loads with your E/F relay. Unless you have numerous small loads and no large ones it is likely that an E/F on a single load will take out the whole installation rather than tripping only the faulted circuit.


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

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