Hypothetical - exceeding switchboard ratings

[FreddyNurk]

Within my employer there has been some recent discussions regarding ratings of some of our equipment, and instances whereby that rating may be exceeded. To this end, a couple of considerations occurred to me, and I'm curious to hear others' views on it.

Situation: Board previously certified as adequate for service (by engineer of record), despite the fact that under certain conditions, should a fault occur, board rating would be exceeded.

A project is raised to replace some equipment connected to the board with other new equipment of slightly higher rating. EOR states during detailed design that board is acceptable with new equipment. Commissioning team raise an issue that during commissioning, number of connected units must be limited in order to not exceed switchboard rating whilst team is in room containing switchboard. EOR agrees to limitation of connected devices and administrative management of system to ensure that rating is not exceeded whilst personnel are present.

Connected equipment consists of LV generators; as background.

So, the hypothetical is thus: EOR approves a piece of equipment as suitable for service with full knowledge of possible exceeding of ratings under certain (admittedly rare) conditions. EOR agrees to and proceeds with mitigation strategies for use whilst personnel are present around equipment. Is the EOR's consideration appropriate, or should the EOR either have:
a) Presented consideration that the exceeding of the ratings not pose any risk to personnel working in the area
b) Retracted the certification that the equipment is adequate for service, given the implication that further measures are required to ensure personnel safety for the equipment.
There is no consideration of replacement of switchboard in the immediate future.

Before anyone asks, this is not an ethics assignment. Any thoughts on whether the EOR is acting in an appropriate manner?

 

[desertfox]

Have a read of this link

http://www.trane.com/commercial/library/vol273/withstand.asp

It clearly states conditions to be met, it doesn't say that it's alright to exceed the rating if it did then what's the point of designing and testing to a given fault rating, here is an extract:-


Overcurrent protective devices (such as fuses and circuit breakers) should be selected to ensure that the short-circuit withstand rating of the system components will NOT BE EXCEEDED should a short circuit or high-level ground fault occur.
System components include wire, bus structures, switching, protection and disconnect devices, distribution equipment, etc., all of which have limited short-circuit ratings and would be DAMAGED or DESTROYED if these SHORT-CIRCUIT RATINGS are EXCEEDED. Merely providing overcurrent protective devices with sufficient interrupting ratings will not ensure adequate short-circuit protection for the system components. When the available short-circuit current exceeds the withstand rating of an electrical component, the overcurrent protective device must limit the let-through energy to within the rating of that electrical component.

 

[FreddyNurk]

racookpe has a point that I really agree with, there is no risk to personnel if they're not in the same room as the equipment, and where possible, designs should allow for this separation to happen.

djs, the reasons for the lack of clarity presented around the prospective fault current is that this location is powered by diesel generators, there is no incoming feeder or single breaker for a utility feed.

Thus (and this is what ScottyUK would have considered when he pointed out breaker ratings) the prospective fault depends on number of machines connected and so on. In our discussion case, when all units are online, the peak fault exceeds the board rating as stated in the test documentation (type test certificate), for a period in the range of around a 1/4 of a second. After this point its expected (as per manufacturer's documentation and typical tests) that the fault decays to far less than the board rating. This is entirely due to the size of the unit, much larger machines behave in the same manner, though the numbers differ.

This is all to do with the technical details of whether or not the rating has been exceeded though, rather than the original question posed.

 

[MintJulep]

The rating is almost always proven by actual tests, because it is difficult to predict the complex interaction of the bars, supports, braces, droppers, and so on.

How many samples are tested?

Enough to cover every possible permutation of complex interactions? Of course not.

Is manufacturing variation considered correctly? Are all possible material defects considered correctly? Even probably defects?

A "worst case" sample? Probably not, because has you correctly state, it's difficult to determine what is really "worst case".

The truth is that most tests are woefully inadequate to do anything remotely useful other than make people that don't understand this feel good.

 

[ScottyUK]

I don't disagree with you, but the modelling capability has only been available to the mass market for 20 - 25 years which isn't very long in terms of equipment life. Like I said, power T&D is a tremendously conservative industry.