I'm performing an Arc Flash Analysis for a client and have come across a situation where I have numerous breakers not rated for the available fault current. Previously, the transformer ahead this line-up had failed and it had an impedance of 5.75%. The calculated fault current at that time was around 64.5K. The breakers are rated for 65K. I've heard that breakers shouldn't be used unless the available fault current is less than 90% of the AIC rating of the breaker. What is the rule on this?
Short Circuit Ratings of breakers
- Thread starter awmickens
- Start date
- Status
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
awmickens said:
Breakers need to be applied based on the UL fault current rating at a specific X/R ratio. If the system X/R ratio is different than the one used by UL there is an upward adjustment that needs to be performed on the available fault current instead of derating the published AIC rating of the breaker.
This process is kind of like saying breakers should only be applied based on 80% of their rating, instead of properly saying they should be sized to 125% of the load.
I do not believe a 10% factor provides enough room for the typical required X/R adjustment.
ron
Senior Member
- Location
- New York, 40.7514,-73.9925
If there is no derating of the AIC rating due to x/r (which I think begins at >6.6), then I would be okay with a 65kAIC breaker being installed where there is 65kA available.
kingpb
Senior Member
- Location
- SE USA as far as you can go
- Occupation
- Engineer, Registered
Did that 64.5kA also include the reduction in fault current due to any cable between the transformer and breaker/panel? As an engineer, you can perform a calculation based on through fault current instead of withstand current. The basic implication is that except for the main breaker/bus, the feeder breakers would never see that much current go through them. It is a fallacy that the breakers need to have the same rating as the main and bus. People will say, but what if a fault occurs on the bus, the breakers won't be rated properly. However, what they forget is that if a fault was to occur on the bus, the panel blows up, and it doesn't really matter what the breakers were rated for.
mayanees
Senior Member
- Location
- Westminster, MD
- Occupation
- Electrical Engineer and Master Electrician
ron said:
I agree with Ron.
Of course make sure you include anything and everything that effects the available fault current, including: the actual test values for X/R & Z for the particular transformer; accurate Utility X/R and contribution data; spinning load contributions; cable/busbar impedance; etc.
If my calcs came to 65.1kA, I'm doing anything I can do to find additional impedance to get that number below 65. I'd even look at breaker contact resistance if the numbers were that close.
My rationalization is to always provide a minimum-essential system design, thereby providing the client with the best value for his money, and thus rationalizing my existence (my fee).
John M
Johnmayanees said:
That an interesting comment. I assume that you are making these calculation from a construction dwg or something similar. Cable distance accuracy is at best suspect. IF you had the situation you mentioned, would you go back an make actual measurements?
- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
So NEC 110.9 is a fallacy?kingpb said:
mayanees
Senior Member
- Location
- Westminster, MD
- Occupation
- Electrical Engineer and Master Electrician
.. good point
.. good point
And of course that needs to be balanced with reality. On a recent Data Center design there were several panels that were marginally above a calculated 65kA fault current. The first thought was to run extra cable to assure a calc under 65kA, but one of the engineers brought up the point - was anyone willing to argue in a court of law that the panel that exploded saw 64,900 amps instead of 65,100....
In that particular case the cost to increase the panel ratings from 65 to 100k was minimal, so the decision was made to buy the 100k panels. We were looking at additional cable runs of 50 to 100 feet.
But in some cases the cost to upgrade switchgear and Static Transfer Switches is so high that it makes sense to take a conduit lap around the room!
In the case of the OP, his is an installed system, so there are fewer options. He's left with development of the most accurate system model possible, and then the calculated fault current numbers will fall where they fall.
John M
.. good point
Absolutely Bob, if the numbers were that close, I would accurately quantify the distance. And certainly that goes both ways...if the numbers were slightly above or slightly below. I sure don't want to engineer a system that's potentially dangerous.bob said:
And of course that needs to be balanced with reality. On a recent Data Center design there were several panels that were marginally above a calculated 65kA fault current. The first thought was to run extra cable to assure a calc under 65kA, but one of the engineers brought up the point - was anyone willing to argue in a court of law that the panel that exploded saw 64,900 amps instead of 65,100....
In that particular case the cost to increase the panel ratings from 65 to 100k was minimal, so the decision was made to buy the 100k panels. We were looking at additional cable runs of 50 to 100 feet.
But in some cases the cost to upgrade switchgear and Static Transfer Switches is so high that it makes sense to take a conduit lap around the room!
In the case of the OP, his is an installed system, so there are fewer options. He's left with development of the most accurate system model possible, and then the calculated fault current numbers will fall where they fall.
John M
mayanees
Senior Member
- Location
- Westminster, MD
- Occupation
- Electrical Engineer and Master Electrician
What Program are you using???
What Program are you using???
Mr. Mickens,
I can say from being an SKM user that the program automatically derates breakers based on the X/R environment they're in versus the X/R that the breakers were tested at. A quick look at some random breakers in the SKM library shows X/R testing at 4.899 for several different manufacturers, which is to say that if the electrical system you're working in has an X/R rating above 4.899 the breakers need to be derated. And truth-be-told, that's not a very high X/R ratio!
I would bet that the software you're using automatically compensates for the X/R values.
And if you're not using software, and you're doing it by hand, Good Luck with that!
John M
What Program are you using???
awmickens said:
Mr. Mickens,
I can say from being an SKM user that the program automatically derates breakers based on the X/R environment they're in versus the X/R that the breakers were tested at. A quick look at some random breakers in the SKM library shows X/R testing at 4.899 for several different manufacturers, which is to say that if the electrical system you're working in has an X/R rating above 4.899 the breakers need to be derated. And truth-be-told, that's not a very high X/R ratio!
I would bet that the software you're using automatically compensates for the X/R values.
And if you're not using software, and you're doing it by hand, Good Luck with that!
John M
winnie
Senior Member
- Location
- Springfield, MA, USA
- Occupation
- Electric motor research
jim dungar said:
I think that kingpb is just saying that the bus itself has an impedance that will reduce the fault current available at the breakers.
As a practical matter, fault current ratings seem to be pretty well 'quantized'. If the available fault current at the main breaker is X, the available fault current at downstream breaker in the same panel will be only slightly lower, and thus all of the breakers in the panel will effectively be required to have the same rating (unless, of course, a series rating is involved).
But consider the situation where the fault current available at the main is 10,100A, and once you include the impedance of the main and the bus, the fault current at the first branch breaker is 9,900A. You could have a 22KA main and 10KA branch breakers, and it would be a fully rated system. (Okay, that example was a stretch in more ways than one
)- Location
- Wisconsin
- Occupation
- PE (Retired) - Power Systems
winnie said:
I thought he was describing the condition of a branch breaker not seeing the available fault current at the panel but rather seeing only the thru fault current that is limited by the branch circuit wiring. It can be argued that the only time the branch breaker sees the maximum line side current is when the fault occurs directly on the load side of the breaker.
My 2 cents:
Transformer impedence can be off by 10% of the nameplate. So I wouldn't want the breakers AIC rating any less than 10% above the calculated value. That may be where the 90% number you heard came from.
Unless you are assuming infinate current available on the primary. That is one typical way of doing fault studies. If that's the case, chances are, the actual available fault current may be a lot less. If you have some data available on the primary wiring, sometimes you can use that to your advantage.
Steve
Transformer impedence can be off by 10% of the nameplate. So I wouldn't want the breakers AIC rating any less than 10% above the calculated value. That may be where the 90% number you heard came from.
Unless you are assuming infinate current available on the primary. That is one typical way of doing fault studies. If that's the case, chances are, the actual available fault current may be a lot less. If you have some data available on the primary wiring, sometimes you can use that to your advantage.
Steve
mayanees
Senior Member
- Location
- Westminster, MD
- Occupation
- Electrical Engineer and Master Electrician
Exact calc..
Exact calc..
When I write about 64.9kA vs 65.1 kA I'm referring to measured values of impedance and X/R ratio.
I'm basing my calcs on a call to the factory where the transformer was manufactured to get the actual test results for the S/N of the xfmr.
There is no 10% variance for that number.
And there is no infinite Utility contribution assumption. There is a letter in the project file from the Utility listing the utility contribution, complete with contribution and X/R value.
There are no assumptions when the levels are close to the ratings of the equipment.
There is always the possibility of the Utility changing their numbers, so it's worthwhile to look at an infinite contribution.
But in the case of the OP, it's an existing system. There's a need to qualify the existing equipment. Use all of the availabe real data to do the assessment.
John M
Exact calc..
When I write about 64.9kA vs 65.1 kA I'm referring to measured values of impedance and X/R ratio.
I'm basing my calcs on a call to the factory where the transformer was manufactured to get the actual test results for the S/N of the xfmr.
There is no 10% variance for that number.
And there is no infinite Utility contribution assumption. There is a letter in the project file from the Utility listing the utility contribution, complete with contribution and X/R value.
There are no assumptions when the levels are close to the ratings of the equipment.
There is always the possibility of the Utility changing their numbers, so it's worthwhile to look at an infinite contribution.
But in the case of the OP, it's an existing system. There's a need to qualify the existing equipment. Use all of the availabe real data to do the assessment.
John M
Last edited:
- Status