This discussion is locked.
You cannot post a reply to this discussion. If you have a question start a new discussion

Earth Fault Loop Impedance with Safety Generator

Former Community Member
Former Community Member
Do I take this into consideration? If so how do I size wire in a building with a safety generator? What is my "Ze" when the supply switches from DNO to the safety generator? How do diesel generators behave during faults?


  • There are many posts on this elsewhere on the forum, search generator, you may like to read them. The Generator will almost certainly require an Earth fault relay (large RCD) and your carefully calculated disconnection times will probably increase. There is little you can do to change this, but power is probably better than no power.
  • Another point to consider is that an earth fault on a system supplied by a generator is often much less dangerous than a similar fault on a system supplied by public mains.

    Consider the following illustrative examples.


    1) An installation connected to 240 volt public mains. Resistance of earth conductor is 0.12 ohms and resistance of phase conductor is also 0.12 ohms. Resistance of neutral conductor is 0.12 ohms. Full load current is 50 amps, with a full load voltage drop of 12 volts.

    An earth fault occurs and with a total earth loop impedance of 0.24 ohms the fault current will be 1000 amps. Whilst the fault persists there will be a voltage drop of 120 volts in the phase conductor and a voltage rise of 120 volts on the earth conductor. Any person in contact with a class one appliance and with true earth will get a shock a 120 volts. Dangerous, hence the need for prompt disconnection of such faults.


    2) Now consider the same installation but supplied from a standby generator instead of public mains. The generator has a full load current of 100 amps and a maximum short circuit current of 250 amps.

    An earth to phase fault occurs as previously.

    With the fault current limited by the design of the generator, the voltage rise of the earth conductor and of class one appliances connected thereto is now only 30 volts rather than 120 volts. Very low risk, even if the fault persists for longer than would be acceptable under normal conditions with the mains supply.


    RCD protection of general use sub circuits will still operate as before and gives extra protection.

    Think twice about sensitive earth fault protection on life safety circuits. If the building is on fire, do you really want the fire pump to trip out due to 100ma of earth leakage ?

  • Former Community Member
    0 Former Community Member
    broadgage:

    Another point to consider is that an earth fault on a system supplied by a generator is often much less dangerous than a similar fault on a system supplied by public mains.

    Consider the following illustrative examples.


    1) An installation connected to 240 volt public mains. Resistance of earth conductor is 0.12 ohms and resistance of phase conductor is also 0.12 ohms. Resistance of neutral conductor is 0.12 ohms. Full load current is 50 amps, with a full load voltage drop of 12 volts.

    An earth fault occurs and with a total earth loop impedance of 0.24 ohms the fault current will be 1000 amps. Whilst the fault persists there will be a voltage drop of 120 volts in the phase conductor and a voltage rise of 120 volts on the earth conductor. Any person in contact with a class one appliance and with true earth will get a shock a 120 volts. Dangerous, hence the need for prompt disconnection of such faults.


    2) Now consider the same installation but supplied from a standby generator instead of public mains. The generator has a full load current of 100 amps and a maximum short circuit current of 250 amps.

    An earth to phase fault occurs as previously.

    With the fault current limited by the design of the generator, the voltage rise of the earth conductor and of class one appliances connected thereto is now only 30 volts rather than 120 volts. Very low risk, even if the fault persists for longer than would be acceptable under normal conditions with the mains supply.


    RCD protection of general use sub circuits will still operate as before and gives extra protection.

    Think twice about sensitive earth fault protection on life safety circuits. If the building is on fire, do you really want the fire pump to trip out due to 100ma of earth leakage ?

     





    Exactly what I was thinking, the generator will output a lower circuit voltage. However, because the MCB will take longer to clear won't the field or excitation collapse before the MCB opens? I want to avoid this possibility, particularly in a safety supply.


  • If you really think that there are 2 options: one is a much larger generator than you need, say 1MVA for a 100 kVA supply. This is very expensive. The second option is to understand that the likely risk of supply failure is very small, much less than a fault at the same time. This means that having any supply is better than nothing, so you accept that a fault could occur but is very unlikely. Simple risk assessment to use the modern term. If an MCB fails to open what is the risk? Very small as the power available is also small. This is very important to understand.
  • Former Community Member
    0 Former Community Member
    davezawadi (David Stone):

    If you really think that there are 2 options: one is a much larger generator than you need, say 1MVA for a 100 kVA supply. This is very expensive. The second option is to understand that the likely risk of supply failure is very small, much less than a fault at the same time. This means that having any supply is better than nothing, so you accept that a fault could occur but is very unlikely. Simple risk assessment to use the modern term. If an MCB fails to open what is the risk? Very small as the power available is also small. This is very important to understand.




    Do you really want to black out a building or hospital because of a fault in single 16 amp circuit due to a light fixture?


    I'm thinking that with knowing the behavior of the alternator and engine some basic equations can lead to sizing the circuits such that a fault will not clear the generator. 


  • That will not occur because 32A (instant trip) is available. Faults that will not clear will be on main or submain circuits where the power available is not sufficient. Earth faults on final circuits are the same, the present protection will work.
  • Indeed, subcircuits that are small in relation to the generator capacity will be protected against short circuits/gross overloads by fuses or circuit breakers in the usual way.


    Mains and submains, may not be properly protected, but that is is many circumstances a price worth paying, as has already been said "better than no electricity"


    Large final subcircuits may not be properly protected and a bad fault may trip or stop the generator, again a price probably worth paying.


    A rule of thumb is that standard types of fuses or circuit breakers should be rated at no more than about 10% of the generator capacity.


    Example, a medium sized diesel generator with a full load capacity of 540*/670* amps per phase used for 100% backup to an office building.

    Most final circuits are 32 amps or less and will be protected in the usual way without any special considerations.

    Submains of 100 amps are NOT protected in generator mode and a bad fault might trip or stop the generator. Unlikely in practice since all such circuits had the ORIGINAL HRC fuses fitted from decades before.

    Large circuits to lifts and chillers also not protected, but such machines have various protective devices and the fuses only in practice protect the cable. Faults on correctly specified large SWA cables with buildings are almost unknown.


    *540 amps "prime power rating" for long hour use such as base load generation in locations without utility service. 670 amps "standby power rating" for short term use in case of utility failure or for peak hour use only.
  • Former Community Member
    0 Former Community Member
    I needed that rule of thumb! 10% sounds reasonable, or if you can get an instant pickup value around the gen's 5 cycle L-E value.


    Correct me if I'm wrong, but for an L-E fault, a generator will output 1000% of the FLC for a few cycles before dropping. Is there any truth to this?
  • No.
  • Former Community Member
    0 Former Community Member
    What are the L-E fault for the first few cycles?