Ground fault

[timm33333]

What would be the difference (setting, working principal, etc) between a ground fault relay for a motor circuit, and the ground fault relay of a non-motor circuit. Thanks.

 

[dpc]

No difference in principle. But if the relay is protecting a single motor, it can be set to operate with little delay for very low levels of ground current, since there is nothing it needs to coordinate with. On a typical feeder, you have to consider coordination with downstream ground fault and phase overcurrent devices, if there are any.

 

[krisys]

DPC, You have clarified it in few simple sentences!!! It re-affirmed my understanding as well.

Now I need your advice for the following:

For a 415 V, 45 KW motor supplied from solidly earthed 415V system, what should be a typical Earth Fault setting. Our Consultants have recommended 10 Amps with 0.1 sec time delay. The Earth fault protection unit gets the input from the CBCT.

In my view the sensitivity can be in milli Amperes (mA), say 500 mA with a time delay say 1 sec. Does it make sense?

 

[timm33333]

Actually the ground fault sensitivity which is in mA range (2-30 mA) is for personnel protection. And the ground fault sensitivity which is in ampere range (5-50 A) is for motor protection. So a sensitivity of 10A is perfect for motor.

My question is that why do manufacturers say that the "motor ground fault relay" cannot be used for the ground fault protection of non-motor circuits? My understanding is that if the sensitivity of a motor ground fault relay is lowered to mA range (for example 30 mA), then it should be Ok to be used for the ground fault protection of non-motor circuits.

 

[dpc]

If you set the ground fault pickup too low, it can mis-operate during motor inrush and starting. There's no real advantage to a setting of 500 mA over 10 A for protection of the motor. Also, if you use a delay of 1 second, you are creating the risk of much greater motor damage should a fault occur. These are typically definite time devices, so with a delay of one second, the tripping is delayed for one second regardless of the fault magnitude. Also, any ground fault devices upstream must be slower than this device to properly coordinate. A delay of 0.1 sec is more appropriate, IMO.

 

[rcw retired EE]

What dpc said.

Motor starting inrush current can saturate a GF CT or at least cause errors and false trips. A feeder circuit may not have the large inrush currents seen on motor circuit. Alsao, a feeder circuit probably has downstream overcurrent protection devices that should coordinate with the feeder GF protection. A feeder GF relay typically will be set with a longer time delay than a motor circuit GF relay. The motor circuit GF doesn't have to wait for any downstream device to trip.

 

[ppedUK]

Don't forget that any ground fault relay setting in a motor circuit using a contactor would have to be correctly co-ordinated with the contactor breaking capacity and the motor short circuit protective device. Contactors are not generally designed to break fault currents (earth faults or phase faults) and a solidly grounded 415V system (krisys example above) will have a high earth fault current.

 

[krisys]

ppedUK, thank you.

Now I recall why the motor feeders should have relatively higher time delay with high sensitivity (like 500mA). If the fault is with higher earth fault current, which is exceeding the contactor SC Current breaking capacity, the E/F element should allow the fuse or MCCB to clear the fault. If the earth leakage current is sustained at a low level, then it is a confirmed earth leakage case. Such a small leakage current cannot be detected by other protective devices in the motor starter. Also it is safer for the contactor to open the for the earth leakage current.

It may be noted that in a solidly earthed 415V system, the earth fault current could be as high as (exceed) the phase fault current.

The question of why do we need such a sensitive setting for the motors? Yes the sensitive setting would easily detect the weak winding and trips the motor before the motor fault makes the faulty motor beyond repair. In other words, it minimizes the extent of damage to the motor.

dpc,
Please tell now whether my explanation makes sense.

 

[timm33333]

Motor has three live wires (no neutral) so why would it matter if the neutral is solidly earthed or not?

 

[waross]

Well, the codes in North America require four conductors to connect a three phase motor.
The fourth conductor may, in some cases, be the metallic raceway from the source to the motor. More often it is a fourth wire.
This is called the equipment grounding conductor.
In the event of an internal earth fault in the motor, some of the current will return to the source via the equipment grounding conductor.
If the ground fault is an arcing ground fault, it is important to interrupt the current before the stator laminations are damaged.
It won''t take much arc damage to the stator laminations before the stator is scrap metal.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[timm33333]

There are two kinds of ground fault relays: First type are used on solidly grounded systems, and Second type are used in resistance grounded systems.

If a system is resistance grounded, then for 3-phase 5-wire feeders (L1+L2+L3+N+Gnd) we will need the Second type ground fault relay.

Question: If a system is resistance grounded, then what type ground fault relay (First type or Second type) should be used for 3-phase 4-wire (L1+L2+L3+Gnd) feeders?

 

[dpc]

No, it doesn't really make sense to me. A delay of 0.1 seconds is more than enough time for the MCCB to trip for any fault the exceeds the contactor's interrupting rating. Not all faults start at very low levels, and even if they do, they often escalate rapidly. If you always delay 1 second, you will be generally be causing more motor damage.

Regarding resistance grounding - at 415 V, this must be high-resistance grounding, so there is really no purpose to the ground fault relay of either type. The whole point of high-resistance grounding is that tripping is not required for a single line-to-ground fault. You might want ground fault protection on the mains and feeders to help in the event of a double line to ground fault, but I've never seen motor ground fault relays applied on a high-resistance grounded low voltage system unless they were already there and the system was converted from solid grounding to resistance grounding.

 

[timm33333]

High resistance grounding only means that the ground fault current is limited to 5A-25A, it does not mean that ground fault protection is not required. Ground fault protection should still be used for high resistance grounded systems for sensitivity of 2mA-25A; however ground fault protection cannot be used for high resistance grounded systems for sensitivity of more than 25A.

 

[dpc]

timm,

25 A is not high resistance grounding at 415 V. It will almost always be less than 5 A. I think you have it backwards. Ground fault CURRENT detection is generally used for low-resistance grounded systems. That is one of their advantages - selective coordination is possible. With high resistance grounding, we generally do not trip. At least in my corner of the planet.

 

[timm33333]

How about we set the ground fault protection as this: no time delay for non motor loads, 10 m-sec delay for motor loads, and 20 m-sec delay for main incoming feeder? (The sensitivity would be the same, say 4A, in all three cases.)

 

[timm33333]

Sorry disregard last post because as discussed above, breakers are not tripped in high resistance grounding system.

 

[3I0]

Breakers are tripped in high resistance grounding systems if the resistor is not rated for continuous current (i.e. 10 second resistor rating is seen on most medium voltage systems here).

There may also be code requirements that outline under what conditions you can leave the system energized with a single ground fault on a high resistance grounded system, i.e. must annunciate the presence of a ground fault on the system with a visual/audible alarm, system voltage 5 kV or less, 10A or less resistor rating for Canadian code (10-1102). If this is a mining application, there may also be additional requirements.

A 10 millisecond coordinating time interval between your feeder and mains would not be long enough in any case, for a tripping scheme on a high resistance grounded system.