The "worst" fault is known as a "bolted" three phase fault. "Bolted" refers to the phases being connected via a zero impedance connection. In reality this is not physically possible, since there will always be a resistive componant with a fault occurs.
On the other hand, it depends what you mean by "worst". As far as I'm concerned ANY fault is a BAD fault.
When sizing fuses you should consider the fault type which will result in the lowest fault current, and size your fuse to clear that fault.
The worst kind of fault I have experienced is an arcing fault that can't be seen by the installed protective relaying.
A spring tension connection on a fuse clip overheated and melted at 2400V. Once the copper path was gone, an arc bridged the gap to keep current flowing in the burned open phase. There was only fuses and overcurrent relays on the upstream feeders. This arcing fault persisted for several minutes until the feeder was manually shut down because of flickering lights downstream of the affected phase. Several pound of copper were pooled up in the bottom of the gear and the entire cubicle had to be rebuilt from the resultant damage. As a testament to the switchgear manufacturer's design, no phase-phase or phase-ground fault occured even though ionized gas filled the cubicle during the event.
These are also the worst kind of faults for residences as these are the largest cause of electrically initiated fires.
Hard faults or low impedence faults are easy for protective relaying to "see" and thus clear.
Very interesting info from you guys. I considered a fault at generator terminals with X1=0.3, x2=.2 and x0=.1 then I found SLG=5 pu and 3phase symmetrical fault=3.33 pu. If I change x0 to 0.5 then SLG = 3pu. It seems that for low x0, SLG faults are worse than 3 ph bolted faults.
Yes, there can be conditions, generally close to generators or the wye side of delta-wye (grounded) transformers, where the SLG fault will produce more current than a 3-phase fault. If the 3-phase fault is balanced, there is no current to ground and there will be no difference in the fault between a 3-phase fault and a 3-phase to ground fault.
In general, the three phase fault is the largest fault.
However, the single phase-to-ground fault may be greater than the three-phase-fault in special cases such as near generator under transient conditions or at the bushings of the Wye-Zigzag transformer.
This is because the zero sequence of the source is less than the system impedance upstream the fault.
The figure in the enclose link may be used to determine the maximum SC.
Typically, as the others are saying, the three phase 'bolted' fault will be the fault with the largest magnituude. Now "worst" fault can depend. If we are talking Arc Flash, an arcing fault of a significantly lower magnitude could be worse than one with a high magnitude. This is because the energy of an arc flash depends as much on the fault level as it does on the time the fault persists (actually the distance the subject is away from the arc has an even greater influence -- as it is a squared function -- but let's leave that out of the equation for the purpose of this discussion). In any case due to the inverse time characteristics of the OCPD, a smaller fault may persist much longer than a larger one, and hence more energy may actually be dissipated in an arc flash event. At least, that is how I understand it. Anyway...
I had the opportunity of talking to the late, great JL "Blackie" Blackburn way back in the early 70s when I was working at Westinghouse. I asked him what the worst kind of short circuit that could occur. He said that it was the three phase bolted fault. While a phase to ground fault can generate more SC amps, the three phase bolted fault injected more power (hence more enery)into the fault than any other type
Guess I'll need to challenge the great Blackburn, then. If the fault is bolted there is little resistance and little power dissipated locally. And the higher current means short clearing times and little time for the power to make energy. Maximum power transfer takes place when the power is dissipated by an impedance that is the complex conjugate of the source impedance. If fault arcs are assumed to be resistive then the maximum power occurs when the arc resistance matches the source resistance. Now throw in a good long clearing time to ensure lots of energy to damage equipment or people.
Reviving tsuro's question on the single line to ground fault -- IEEE Red Book indicates that SLG can be up to 125% of 3ph bolted faults, but that such conditions are rare in industrial installations except from generator sources due to the low zero-sequence generator Z.
However, playing with SKM on a recent project, we were seeing high SLG at the y side of a Dy transformer, as was mentioned by davidbeach and others. I don't understand why you would see such high SLG at a Dy -- the transformer zero sequence Z should be equal to the pos & neg sequence Z (right?). So what causes the high SLG fault current at at Dy?
I had a really interesting case a couple of months ago. I was fault-finding a 1200 A 400 V soft starter and there was a mistake when we changed the trigger unit so that one of the thyristors got a 120 ohms series resistor in the gate path. The resultant loss of triggering gave us a heavy DC current in motor and transformer. The 400 V breaker cleared the fault immediatley and as we were standing there, wondering what had happened - the 20 000 V just shut off. Only emergency lights there.
It turned out that the DC had saturated the transformer core and caused a very high primary current. So high that the protective relay more or less caught fire and then, after a few minutes, just died. Had to replace the protection.
That is one of the worst faults I have experienced. And it is interesting because no one had anticipated such a high primary current. A bolted short on the secondary side would have been limited by Xk of the transformer. DC saturation took a lot of that Xk away.
However when there is a bang and the lights go out, your generators stop and the utility company calls in to ask what happened as their power went of, now that's bad.
In fact its almost time to put on ones coat and go home....
This was half past two in the morning. So I went to my hotel and got some sleep while the HV guys sorted out their problem. The utility company didn't have to call - I was working with one of their fans for the boiler. On their premises.
jghrist, thanks for your response. . . but could you please clarify your statement a bit, that "The zero-sequence impedance of the system on the source side of the Dy transformer is shorted out." Are you saying something going on at the primary side is reflecting to the secondary which appears as a lower zero-sequence impedance? During normal conditions or only during fault conditions?
