Worst case for short circuit calculations

[REEngineer]

Consider a standard double bus bar system. One transformer feeds each half and they're separated by a tie breaker. The tie breaker is normally open. For short circuit calculations, it seems we should assume the tie breaker is closed. This is possibly the worst case. As we know this case would happen rarely, do we really need to base our calculations on this arrangement? Or can we assume safely that the tie-breaker is always open and reduce short circuit current level.

 

[rbalex]

I generally endorse your analysis ? with a few caveats. First, I would not permit a manual only scheme. Second, even an automatically controlled transfer scheme would need to have a very short parallel duration.

Having said that, you need to recognize it is still not consistent with the last clause in the first sentence of Section 110.9. Since the full rule is neither clearly permissive nor prohibitive regarding duration it generally must be interpreted in the most restrictive manner. This is especially true when you realize ??the current that is available at the line terminals of the equipment? is the combined current through both sources when they are paralleled. (Underline by rbalex)

Even though none of the principal circuit breakers (main-tie-main) will be called to interrupt ??the current that is available at the line terminals of the equipment,? all downstream breakers can still theoretically ?see? it under closed transition condition. Of course, a transfer is still not likely to precipitate a downstream fault than any OPC would need to clear; nevertheless the Code says what it says.

The American Petroleum Institute (API) has tried several times (over 12 years) to get CMP 1 to clarify this since double-ended, automatic secondary selective schemes are very common in petrochemical facilities. CMP 1 has essentially said ?get special permission? under Section 90.4.

 

[eric stromberg]

I'm with Bob.

I can say this. The very large industrial facility, for whom i've done much work, does not consider the case when all three breakers are closed (main-tie-main) because the 'time duration' is very short and, therefore, is an 'acceptable risk.' (Good thing there are remote control panels to close the tie)

On the other hand, i've worked with other industrial facilities that keep all three breakers closed. In this case, it would be crazy not to consider both sources while doing short circuit calcs.

 

[jim dungar]

Consider a standard double bus bar system. One transformer feeds each half and they're separated by a tie breaker. The tie breaker is normally open. For short circuit calculations, it seems we should assume the tie breaker is closed. This is possibly the worst case. As we know this case would happen rarely, do we really need to base our calculations on this arrangement? Or can we assume safely that the tie-breaker is always open and reduce short circuit current level.

Not all Main-tie-Main systems are designed or intended operate with all three breakers closed. It seems most of the ones in this area include a key interlock scheme that prevent the supplies from being paralleled.

 

[ron]

As mentioned by rbalex in the earlier post, there has been several attempted to get the NEC modified to allow for an exception if the overlap of sources would be short, and each time they have been rejected indicating in comments that it is an AHJ issue for exception or that both sources should be included.
I know of no AHJ that would give a variance or non-conservative interpretation for something like this.

 

[rbalex]

...
I know of no AHJ that would give a variance or non-conservative interpretation for something like this.

I don't either - but it is usually because they don't understand that a downstream failure concurrent with a short duration closed transition is astronomically remote. Literally. It is on the order of 10 to the negative 18 when the transition is 2 seconds or less.

If we used the same level of conservative thought throughout, we wouldn?t use electricity at all

Edit: corrected ONE to the negative 18 (Still a certainty ) to TEN to the negative 18.

 

[Mike01]

Another worst case.

Another worst case.

Just to throw something else out there about worst case serino. What about when you have a co-generation system (peak shaving) where the generator runs in parallel with the utility during the peak hours of operation? I believe you would then have to include the generator / motor feeding into the available fault current from the utility line?

 

[kingpb]

Yes, generators operating in parallel will provide a large contribution to a fault, large motors as well.

As far as a main/tie/main arrangement, I think it is totally dependent as to the industry, client, operating intent, safeguards in place, etc. It is also necessary to decide whether the design is based on through fault current, or total bus fault current.

 

[mayanees]

... my opinion

... my opinion

I believe that if the 2 busses can ever be paralleled, even if there's a closed transition between the breakers, and the other one then drops out, the downstream system needs to be rated to withstand both sources.
Even though the duration is short, that's the time there would be a higher probability of a fault, during the transfer to another system.

The only way I would consider only one of the sources (that being the higher rated system) would be if there was a Kirk-key interlock that absolutely prevented them from ever being closed together.

This is a big issue with critical power delivery because there are always multiple sources to provide backup or bypass operation. And my experience is if the two can come together, the electrical system must be able to handle the combined fault curent.

Because, the consequences of a combined fault contribution onto an underrated bus can be catastrophic - even life-threatening.

Perhaps if a real-world scenario made it essential that two systems meet during some kind of emergencty procedure, and combine on an underrated bus, there could be a procedure that isolates human contact from the equipment during the act. That's the ony way I would consider it. And I would have that procedure laminated and posted to assure compliance - or at least to limit my liability.

JM

 

[kingpb]

i would agree, that is certainly a conservative approach, and would probably be viewed as the safest. Unfortunately, in some cases equipment ratings would not be purchasable, and would also not be viewed as totally practical in some industries, particularly the power/utility industry.

 

[REEngineer]

Thanks all for showing interest.

1) I am talking about petrochemical industry where such arrangements are common. Interlocks exist and there is normally no co-generation. Only in case of an outage, plant's generator will feed downstream busbars and when utility power is restored there'd be a short period in which both sources(Plant's generator and utility transformer) would be connected to same busbar.

2) Where I work IEC standards have higher priority. Does anyone know if IEC has the same issue or not?

 

[weressl]

another NEC fluke

another NEC fluke

I don't either - but it is usually because they don't understand that a downstream failure concurrent with a short duration closed transition is astronomically remote. Literally. It is on the order of 10 to the negative 18 when the transition is 2 seconds or less.

If we used the same level of conservative thought throughout, we wouldn’t use electricity at all

Edit: corrected ONE to the negative 18 (Still a certainty ) to TEN to the negative 18.

Absolutely. Yet the whole idea of Class I, Division 2 rests on the presumption that ignitable mixture will NOT be present when the Div. 2 rated electrical device fails and becomes a source of ignition.

Shees, gimme a break! How many substations are in a chemical plant? 1 or 2? How many electrical devices in the Cl.I, Div. 2 area in the same plant? 1000+?
Blow up a substation and kill nobody, becasue it is an automatic transfer in an unattended loaction. Blow up a chemical plant and you certainly will kill somebody and injure numerous others.

I don't care what the right answer is but at least be consistent with the assumptions.

 

[rbalex]

2) Where I work IEC standards have higher priority. Does anyone know if IEC has the same issue or not?

It has been a few years since I have worked under IEC rules but they do not change quite as often as the NEC.

As I recall, the rule was more general and similar to the second statement in 110.9. Something like," Equipment intended to interrupt current shall have an interrupting rating sufficient for the current that must be interrupted." (These are not the exact words.) It was left as a design exercise to determine the conditions of operation for "...current that must be interrupted."

 

[kingpb]

There is no IEC standard that correlates to the NEC, and the NEC is not applicable to IEC equipment. IEC standards are more similar to IEEE standards. Whereas the NEC is installation safety practices. There is a British standard for wiring regulations which is applicable to the IEC world.

As far as short circuit ratings and switching schemes, you need to look at how the system will be operated. If the two busses can be connected together at any time for simultaneous operation, then you must consider ALL fault contributions. If this makes the available fault current to high for the equipment, then other switching arrangements must be considered, or alternative equipment design, such as current limiting reactors must be utilized.