Generator - fault current for mccb 1

[mrbj]

Hi,
please help resolve the following discussion.
What value of rms sym fault current do you use when attempting to determine the settings of a downstream protective device (ie an mccb)when the supply is from a generator?.
Do you use either the sub-transient, the transient or the steady state conditions.
ie.
600kVA generator at 415/240Volt approx 835Amp
X"d at 0.10 giving Irms = 8350
X'd at 0.15 giving Irms = 5567
Steady state sus short circuit current at approx 2600A.
I err towards the transient, my collegue towards the steady state.

 

[dpc]

You would generally use the sub-transient but it depends on the situation.

To determine if the MCCB is going to trip on its instantaneous element, you would use the subtransient reactance.

For determining operation of short-time delay elements (3-10 cycles), the transient reactance is more accurate.

The synchronous reactance could be used to determine sustained generator fault contribution, but in the real world, this isn't too helpful because the generator SC current at the synchronous reactance is generally below the generator full load output at normal voltage. In addition, many generators have some type of field forcing that will keep the generator fault current well above the synchronous reactance current for a long time, often in the 300% current range.

The synchronous reactance would corresponds to the sustained short circuit output of the machine say from 30 cycles out indefinitely, assuming no change in field current. Normally, some other type of protection will trip the generator (undervoltage, underfrequency) before the generator would be able to trip a MCCB up on the breaker's slower thermal curve.

 

[mrbj]

Thanks dpc.
just to take this a step further.
Would you say that if the sub-trans fault current were greater than the "instantaneous" trip current of the mccb then it would be sufficient in magnitude and remain for enough time to trip the mccb, or would it decay so fast that the mccb would be unable to react within the time that the fault current is available. -- ditto for the transient fault current, which is basically my collegues point.
ps we are talking about an mccb with adjustable micro-processor based electronic trip settings.
regards,
mrbj

 

[dpc]

If the generator fault contribution using xd" is greater than the MCCB instantaneous pickup setting, it should trip. It takes very little time for the MCCB trip unit to respond to the current, much less than a cycle, although the breaker clearing time will be longer, the "decision" to trip happens very quickly. Also, depending on the trip unit, many instantaneous units also respond to the asymmetrical current which can be quite a bit more than the symmetrical.

If you know the Td" time constant of the generator, that will give you an idea of how quickly the subtransient current will decay.

If you want to be absolutely sure that the MCCB will trip on a generator fault and don't have the time constant, I might use the transient reactance to be conservative. On the other hand, if you are trying to set the MCCB to NOT trip, you would have to use the subtransient reactance and probably take the worst-case asymmetry into account.

Typical subtransient time constants for synchronous machines range from 0.02 to 0.05 sec. Transient time constants range from 0.4 to 1.8 sec.

Modern molded-case breakers can CLEAR a fault on instantaneous one to 1.5 cycles.

Hope that helps.

 

[mrbj]

Thanks again dpc,
This is the first time I have used the site and it is great to get feedback like this.
The T"d is 0.012s
The T'd is 0.08s
I assume from the above that the time period for the T"d is 0 to 0.012s and the time period for the T'd is 0.012 to 0.092s.

My colleague, still won't shift on his method ( golden rule written in stone )of 3 x generator output current, for max Ifault to operate the protective device and further states that at the point of intersection of the transient and sustained is three times the gen o/p current and that is what should be used. It also appears that some manufacturers actually produce mccb's with max Isc of 3x to 4x on the device.
Could this be an old rule of thumb that he is using and true for HRC fuses with slower fusing times.

Further, the device is not protecting the upstream generator, that is done by a dedicated control unit which has an electronic version of an IDMT relay set at 3x for 10secs.

The mccb protects downstream cables etc and all that is required is to ensure that enough fault current is available to trip the mccb in case of a fault.
Depending on the settings of the device,and using the 3x method will in my opinion hardly give enough fault current to clear the device, from what you say, applying the transient method, the generator will provide and sustain the fault current for sufficient time so that the device will trip.
Any further comments much appreciated.
Regards,
mrbj

 

[mpparent]

It's also a good thing if you can plot the generator capability curve, which is something most EE programs like SKM or ETAP will do for you. That way, you can know exactly where the curve is in relation to your downstream devices and ensure they trip before your gen. breaker does.

Mike

 

[dpc]

Your time constants are fast, probably due to small size of generator. You might want to verify that these are the short circuit and not the open circuit time constants.

Actually the transient time constant starts from the beginning of the fault. So initially, the total fault current is the sum of the subtransient current, the transient current and the synchronous current. Gradually, each contribution dies out according to its time constant.

I have the equation for the generator decrement curve based on the data you have, but it would tough to do it in plain ASCII text. A good reference book should have it.

After one time constant, the corresponding current has decayed to about 37% of their max value. We approximate this by the use of X"d and X'd. These should be the saturated reactances, in case you have both.

So after 5 cycles, the transient component has decayed to about 37% of your calculated value.

You're right, that if the generator has a PMG and some field boosting, the steady state current will be about 3 times FLA.

I'd feel pretty comfortable using the transient reactance with an instantaneous trip MCCB in this case, even with the short time constants.

I might try to find a way to post the equation tomorrow, if you're still in need.

 

[mrbj]

Thanks to all,
mpparent, unfortunately I do not have access to such programs, many thanks anyway.
dpc, I realise now that I should have incuded more info on my first thread, my apologies.
Reactances are saturated,
generator has PMG,
Xd is 2.63,
T'do O.C field time constat 2.5s,
Ta armature is 0.019s,
short circuit ratio is 1/Xd
I have obtained a time / current three-phase short circuit decrement curve with No load excitation at rated speed, for the generator.
Reading Current off the curve at
0s gives approx 8500A
0.012s gives approx 6000A
0.08s gives approx 2500A
0.4s onwards gives approx 2600A
The curve dips to 1700A at about 0.15s then recovers.
Many thanks for your help over this issue.
My colleague (boss) still insists on using the 3x method.
Any last thoughts ??
regards,
mrbj

 

[dpc]

The decrement curve gives you the information you need to make a decision. You didn't mention the size or type of the MCCB, so I can't offer an opinion on whether or not it will trip.

But you should probably also consider the generator's undervoltage protection. It may shutdown the unit pretty quickly at close to zero volts.

 

[RRaghunath]

Using 3x method for coordination of over current relays seems appropriate.

But, for the short circuit protection, in the particular case the transient current of the generator should work well.