Breaker functionality for Short Circuit Conditions

[minesh21]

Hi,

I have some theory questions regarding breaker functionality. We do a lot of breaker coordination studies which attempt to isolate faults. Most of the time we are only considering phase to phase faults, because breakers downstream of the main substations are generally thermal magnetic and do not have ground fault options. Here are my questions:

1. If ground faults are the most common occurrence, why do we focus more on phase to phase fault coordination when designing TCC curves downstream of substations and boards less than 1000A?
2. If a phase to ground fault were to occur at a lighting and appliance panelboard downstream, no breakers would trip except the main that has ground fault protection. Essentially, your system is NOT coordinated. How is this acceptable?
3. Hypothetically, lets say you didn't have ground fault in your upstream main breaker or anywhere in the system. Lets say you have a ground fault downstream somewhere. The fault flows up the equipment ground conductor all the way to the source transformer, then comes back down the phase conductor to complete the circuit. Wouldn't the breakers still trip even without ground fault settings? Because you still have a large amount of current flowing on one phase, which should cause the breaker to trip. My thinking is that the current wouldn't be large enough to trip the breakers on instantaneous, which is why you have ground fault detection in the first place to detect the lower current levels. Is this correct?
4. Building on item 3 above, when you have a phase to phase fault, the breaker directly upstream of the fault senses a large current and therefore trips immediately on instantaneous. This is a coordinate system because it trips faster than the breaker upstream. Is the difference between phase to phase and phase to ground just the amount of current that flows? Therefore, you need ground fault detection separately versus just having regular phase detection?

 

[LarryFine]

I asked your questions to an engineer friend. Here is the gist of his answers,

1. Both should be taken into consideration. Ground faults can exceed phase faults under certain conditions.
2. It isn't. Coordination should prevent a single circuit from taking out an entire panel or service.
3. Ground-fault protection reduces stresses and damage. It's based more on limiting current than limiting time.
4. Ground-fault detection senses currents that shouldn't be there, so can activate at lower current levels.

 

[ron]

1. Because often there is only one level of GF protection, so there is nothing downstream to coordinate with at the GF trip level (except hospitals and elevators, and other specific code required situations). A L-G fault will often spread to a L-L and L-L-L fault with time if not interrupted which is what will be worst case to trip upstream OCPDs.

2. A L-G fault may trip any breaker that has a trip characteristic sensitive enough to sense the fault level. Just because it is a L-G fault, doesn't mean a thermal magnetic trip element doesn't react. It is still current.

3. Fault current of any type wants to go back to its source neutral by taking all paths in relation to their impedance. Anyone that says it wants to go to ground or only takes that path of least resistance is not accurate. Any OCPD in that path of any type set sensitive enough with a characteristic of that type, will trip when appropriate.

4. Both are fault current. Generally of different magnitudes, but sometimes L-G fault current can be higher than L-L depending on the type of system and where you are in that system, so again, any protection will trip that is set sensitive enough with the most sensitive tripping first if the OCPDs are set selectively (not overlapping, which is sometimes hard with breakers if you are not selecting carefully)

 

[minesh21]

I think my misunderstanding is I thought breakers cannot trip and isolate a line to ground fault if it doesn't have ground fault protection. But I think it still can.

 

[minesh21]

1. Because often there is only one level of GF protection, so there is nothing downstream to coordinate with at the GF trip level (except hospitals and elevators, and other specific code required situations). A L-G fault will often spread to a L-L and L-L-L fault with time if not interrupted which is what will be worst case to trip upstream OCPDs.

2. A L-G fault may trip any breaker that has a trip characteristic sensitive enough to sense the fault level. Just because it is a L-G fault, doesn't mean a thermal magnetic trip element doesn't react. It is still current.

3. Fault current of any type wants to go back to its source neutral by taking all paths in relation to their impedance. Anyone that says it wants to go to ground or only takes that path of least resistance is not accurate. Any OCPD in that path of any type set sensitive enough with a characteristic of that type, will trip when appropriate.

4. Both are fault current. Generally of different magnitudes, but sometimes L-G fault current can be higher than L-L depending on the type of system and where you are in that system, so again, any protection will trip that is set sensitive enough with the most sensitive tripping first if the OCPDs are set selectively (not overlapping, which is sometimes hard with breakers if you are not selecting carefully)

1. Yes but my question is why don't we have ground fault protection at all breakers. We typically only put this at the upstream boards. If we put ground fault everywhere then you can isolate a ground fault rather than taking out the main. I've seen coordinated data center studies that have tripped out the main because of a line to ground fault way downstream. This is because there is no ground fault protection, only thermal may breakers that don't have ground fault protection.

2. So a thermal may breaker can trip on a ground fault even if doesn't have grouund fault protection? How? Is it because current through one phase is large enough to trip just like a phase fault? If so, why even have ground fault protection? Is this for lower currents?

3. So going back to the example, the light and appliance panelboard way downstream could trip on instantaneous just as it would a phase fault? Because current is still flowing through one phase at a large level?

4. Yes we coordinate the breaker (or do our best). But I always thought this was for phase faults, not ground fault.

 

[ron]

1. Yes but my question is why don't we have ground fault protection at all breakers. We typically only put this at the upstream boards. If we put ground fault everywhere then you can isolate a ground fault rather than taking out the main. I've seen coordinated data center studies that have tripped out the main because of a line to ground fault way downstream. This is because there is no ground fault protection, only thermal may breakers that don't have ground fault protection.

2. So a thermal may breaker can trip on a ground fault even if doesn't have grouund fault protection? How? Is it because current through one phase is large enough to trip just like a phase fault? If so, why even have ground fault protection? Is this for lower currents?

3. So going back to the example, the light and appliance panelboard way downstream could trip on instantaneous just as it would a phase fault? Because current is still flowing through one phase at a large level?

4. Yes we coordinate the breaker (or do our best). But I always thought this was for phase faults, not ground fault.

1. Especially for a data center, false tripping of a GF could be dangerous too. Data centers have static switches that sometimes transfer quickly, but one phase at a time and may trick a low setting on a GF breaker. Be careful. It is a philosophy issue. You can put them anywhere you like for an extra expense.

2. GF protection is just a lower set point. Current is current. The sensing element is the same.

3. instantaneous is just the fast acting portion of a TCC. There is an instantaneous portion of a phase characteristic and a GF characteristic for a TCC.

4. You can coordinate Phase and Phase TCCs AND Phase and GF if you like. Most don't because the GF is generally set at much smaller values