Leakage Current of 630 sq. mm XLPE Cable

<...the permissible leakage current limited in the cable...> is determined by local standards, which may even be facility-based.

<...what would be the reason for failure of NGR....> The most likely possibility is overcurrent, particularly as <...Recently a cable was ground on LV side...>.

In a case like this, the best source of information is the cable manufacturer and the test results when the cable was first commissioned.

The 3 amps is most likely AC cable capacitance current - not leakage, which is usually a DC test.

If protection does not operate, since NGR are often only rated for enough seconds to exceed maximum protection operate time, NGR failure is expected - if your continuous rating is correct it should have survived fault. What is specification for NGT & NGR ?

But if NGR coils were corroded, causing "hot spots" or dirty or wet there may have been a fracture or flashover which cascaded into total failure. You have not described failure - open or short circuit, signs of arcing - how could we know, is it direct air cooled, oil immersed?

It would seem prudent to energise the replacement NGR at 400V continuously and check its temperature rise.

NGR is made grid resistance with tappings with 25%,50 &100%.50% of grid resistance have burnt The leakage current in each run is more tan 400 miliamps.Do I require to replace all cables.

.

It depends upon what has caused the fault. It is not possible for anyone outside the facility to see the problem and identify the cause.

Data for 18/30kV single core XLPE insulated cable to DIN VDE 0276-620 indicates capacitance of 0.14 microfarad per kilometre. 0.14 μF has reactance 22740 ohm @ 50Hz - @ 18kV that draws 18000/22740 = 0.79 amp. No surprise data sheet lists loading of 0.8 amp/km!

You have not stated cable length or if that 0.4 amp is just cable disconnected from any transfo or surge arrestors, so cannot judge.

As a rough measure, dissipation factor of polyethylene is 0.0002 @ 60Hz sine, suggesting a "resistive" leakage current of 0.8 amp x 0.0002 = 0.16 milliamps/km.

I do not think what you are measuring is leakage current. In any case, insulation would be a DC test & less than AC dissipation suggests - the cable manufacturer will have data for real cables of their manufacture.

In a perfectly symmetrical sinusoidal 3 phase system, the current In the NGT/NGR would be zero. But real systems have harmonics & they do not balance out. So what you see in the NGR is probably harmonics.

The individual elements in NGR may be OK for 600 amp on their own in free space, but overheat eventually in a stack - as has been posted NGR would usually only be short time rated at full voltage since 999 out of 1000 times protection will dead the system in seconds.

A 3 ampere current on a 600 ampere rated resistor seems to be OK.

The reason your earlier resistor failed is because the protective relay was not set properly to protect the resistor. There is normally a current transformer perhaps 600/5A ratio on the ground lead to the resistor. The inverse current relay is set to protect the resistor, perhaps rated 600A for 10 seconds might be typical. The output of the relay trips the upstream breaker for the transformer, removing all source of fault power. The timing for this relay is set to allow all downstream ground fault sensors to clear the fault. If the fault is not cleared in time, the we must assume the downstream protection failed or the fault is within the transformer zone, so you have no choice but to clear the entire system.