circuit breaker interrupting capacity rating

If the branch circuit breaker fails to clear the fault it will be presented to the main the same as a bolted fault on the breaker rails in the panel and that is exactly what the main breaker rating is designed to do. The real question is what is the available fault current from the service side, not what is happening on the load side.

| If the branch circuit breaker fails to clear the fault it will be presented to | the main the same as a bolted fault on the breaker rails in the panel and that | is exactly what the main breaker rating is designed to do. | The real question is what is the available fault current from the service side, | not what is happening on the load side.

Assume the available fault current is a large portion of what the main is rated to handle. Can a lower branch breaker affect the ability of the main to interrupt the fault.

I've read some literature that says when using 22 kA mains with 10 kA branch breakers ... manufactured by a competitor ... that it can reduce the rating of the whole panel to 10 kA. Is there any truth to this or is it a bunch of marketing BS?

Though it has never happened to me, I've heard of cases where people have had short circuits that resulted in both the branch breaker and the main breaker tripping. While that can be annoying, obviously you don't want the main breaker to wait for too long before "deciding" that the branch breaker isn't able to interrupt the fault.

Not really. If the fault is upstream of the branch breakers (in the panel buss, for example) the branch breakers won't 'see' the fault. The main will clear it. If the fault is just downstream of a branch breaker, both breakers will 'see' and attempt to clear the fault. If it is close enough in, the maximum fault current could exceed the rating of the branch breaker, given the example you posted below. If this is the case, the branch breaker may fail to clear the fault. It may fail catastrophically. But this shouldn't affect the main's ability to back up the branch breaker and clear it as long as the main is rated for that maximum available current.

This is true. Since the 10kA branch breaker can only interrupt a maximum available fault of 10kA, you cannot use the above combination in any case where the maximum fault current just downstream of that branch breaker will be greater than this figure, no matter what the main is rated for.

This is a case of poor breaker coordination. It can result from the use of breakers not rated for the available fault current. For example, the branch breaker, with a lower overcurrent rating, would be expected to act sooner for any given fault current. However, if it is incapable of breaking too high a current quickly enough (or at all). Even as the branch breaker is opening, sufficient current will pass, and the main breaker will sense it, so that the main will reach its trip point. The end result is that both breakers operate.

Most molded-case circuit breakers made today are thermal-magnetic, which means that, for overcurrents below a specified value the trip time is current dependent (thermal element trips breaker) and for currents above a specified level the trip is instantaneous (magnetic element trips breaker). Any fault current which exceeds the instantaneous threshold of the upstream breaker will cause it to trip, regardless of the rating of the downstream breaker. This has nothing to do with interrupting rating.

This can be fun with electronic trip breakers. These breakers are generally shipped with all settings dialed to minimum, with the expectation that a coordination study will provide appropriate set points. Often the study is never done, and the adjustments are left at minimum. THAT is when you get these fun trouble calls; a shorted plug on a coffee maker trips the whole office building, due to the fact that an electronic trip is often faster than even a small magnetic trip.

The whole field of breaker co-ordination is fascinating. At the simplest level you want the breaker to interrupt a fault as quickly as possible, but at the same time you want to isolate the outage. This may be done by making the upstream breakers slower to respond, but now you have increased the let-through energy in case of a fault upstream of the branch breaker. There are proprietary schemes, such as zone selective interlocking, in which breakers are connected by communication cables. In the case of a fault, a downstream breaker sends a restraining signal to the upstream breaker, telling it to hold; the downstream breaker will clear. If the downstream breaker doesn't detect the fault, the fault must be upstream, no restraining command is sent, and the upstream breaker trips at full speed. The techniques go on and on.

The simple expedient of coord>

One of the major influences in available fault current is the resistance of the conductors in the circuit that failed. I always worry when I see very short service conductor runs because the conductor size itself may not be an accurate reflection of the available fault current. In panels where they have large breakers for the branches, compared to the main, it is not unusual for both to trip on a direct short since the fault current is instantanously very high. You are just in a race to see which breaker clears the fault first and it ends up in a tie. If the panel is fairly well loaded and the failing branch is open, prior to the fault you can almost count on the main tripping. I understand that is the worst scenario for the user (lots of other loads get interrupted) but it is simple physics. If the main is already operating at 60-70% of its rating and it gets another big spike it is going down. Once in a big computer installation I saw this happen with a bolted fault in a

60a circuit which actually tripped a huge breaker 2 panels away at the service entrance, bringing the whole building down. (there was a short in the Russell Stoll that fused all 3 phases together with molten copper).

On Sun, 09 May 2004 07:59:08 -0700 Paul Hovnanian P.E. wrote: | snipped-for-privacy@ipal.net wrote: |> |> On 09 May 2004 02:16:42 GMT Greg wrote: |> |> | If the branch circuit breaker fails to clear the fault it will be presented to |> | the main the same as a bolted fault on the breaker rails in the panel and that |> | is exactly what the main breaker rating is designed to do. |> | The real question is what is the available fault current from the service side, |> | not what is happening on the load side. |> |> Assume the available fault current is a large portion of what the main |> is rated to handle. Can a lower branch breaker affect the ability of |> the main to interrupt the fault. | | Not really. If the fault is upstream of the branch breakers (in the | panel buss, for example) the branch breakers won't 'see' the fault. The | main will clear it. If the fault is just downstream of a branch breaker, | both breakers will 'see' and attempt to clear the fault. If it is close | enough in, the maximum fault current could exceed the rating of the | branch breaker, given the example you posted below. If this is the case, | the branch breaker may fail to clear the fault. It may fail | catastrophically. But this shouldn't affect the main's ability to back | up the branch breaker and clear it as long as the main is rated for that | maximum available current.

Of course, avoiding any damage is best, but at the very least I want to make sure something, somewhere, can break the fault.

I saw in one of the TV show episodes where firefighters were working a fire in a house. The service wires going into the house were red hot. While that is probably due to a mis-installed and/or damaged panel in the house, it shows the level of fault that can occur without utility fuses interrupting. A friend of mine described a scenario where he was at the place he worked when a someone doing some cable TV installation accidentally drilled too far through a wall and hit a 480/277V panel or feeder. The fault didn't clear on the secondary side. Eventually, the transformer damage faulted the primary side and the primary fuse stopped the show (my interpretation of his description).

|> I've read some literature that says when using 22 kA mains with 10 kA |> branch breakers ... manufactured by a competitor ... that it can reduce |> the rating of the whole panel to 10 kA. Is there any truth to this or |> is it a bunch of marketing BS? | | This is true. Since the 10kA branch breaker can only interrupt a maximum | available fault of 10kA, you cannot use the above combination in any | case where the maximum fault current just downstream of that branch | breaker will be greater than this figure, no matter what the main is | rated for.

So how would you wire up a panel if the utility says that your available fault current is, for example, 15kA, when individual branch breakers are only rated to 10 kA? Two major product lines (Eaton/Cutler-Hammer CH and Square-D HOMELINE) have no branch breakers available above 10 kA. Even in other product lines, higher rated branch breakers are available only lesser choices.

Will people have to be demanding lower fault currents from the utilities, which otherwise seem to be increasing them as they deal with the increasing demand? Energy-efficiency is also contributing, by pushing down temperature rise ratings of transformers, which have some higher fault currents due to larger conductor sizes.

Should one install a transformer to reduce the available fault current (with a high interrupting main breaker on the primary side)?

|> Though it has never happened to me, I've heard of cases where people have |> had short circuits that resulted in both the branch breaker and the main |> breaker tripping. While that can be annoying, obviously you don't want |> the main breaker to wait for too long before "deciding" that the branch |> breaker isn't able to interrupt the fault. | | This is a case of poor breaker coordination. It can result from the use | of breakers not rated for the available fault current. For example, the | branch breaker, with a lower overcurrent rating, would be expected to | act sooner for any given fault current. However, if it is incapable of | breaking too high a current quickly enough (or at all). Even as the | branch breaker is opening, sufficient current will pass, and the main | breaker will sense it, so that the main will reach its trip point. The | end result is that both breakers operate.

Could it be an overly fast main breaker?

I would think that a really serious fault approaching the available fault current would put things down in the flat line of the interruption curve, and both breakers should be trying within a fraction of a cycle, and at least the main should succeed by at least the next zero crossing.

If given an available fault current in excess of available branch breakers, wouldn't you at least want your main breaker to be one that interrupts as rapidly as possible to minimize the downstream damage from such a fault? Or would it, because it trips too often in non-damaging cases, lead people to think that in a scenario where there is damage, to just reset it without investigating causes and damages?

Actually, when you really study it, even 10kA is hard to deliver to a residential service. I don't have the data at my fingertips, but, believe it or not, the meter is a major consideration in calculating residential fault current. The other two are the transformer and the length of the service lateral.

The only place where you routinely get into trouble in a residential setting is in high-rise apartments, where there might be a 3000 amp bus duct up through the middle of the building with each apartment having a tap. In that case, you just can't use Home-Line or similar breakers. Don't rule out the use of current-limiting main fuses in these cases

I lived in a suburb, built in 1955. We had 8 houses on a 10 kVA pole transformer. Everyone had a 60 or 100 amp service. About 1/3 had central A/C. Line would dip to about 110 on a hot summer day. Nobody noticed except me, and that's because I measured it!

Full load secondary current = 42 amperes (hard to believe, but true!). Assuming Z of 5% that would give a short circuit current, at the transformer, in the neighborhood of 840 amperes! A 50kVA transformer of the same impedance would only give 4200 amperes at the transformer.

Both breakers (branch and main) should be able to clear the fault if they have the proper interrupt ratings. Better two devices properly rated than just one.

Interesting note: While service entrance and feeder breakers are sized for connected loads (per the NEC) utilities are free to use their judgment when applying load diversity factors to select transformer sizes. The transformer high side fuse will (of course) be sized to match the transformer rating.

It sounds like the fault you describe might have been an arcing fault with a high enough impedance not to trip the feeder protection. But the transformer might not have had the same capacity as the feeder, so the fault may have exceeded the transformer rating.

You select a panel and branch circuit breakers with the appropriate rating.

Interesting note: Our local utility (and probably many others) has a standard practice that requires fault currents to residential services of 200 and 400 A ratings to be kept under 10 kA.

For commercial services, its a whole different ball game. In fact, this is somewhat of an issue with another local utilities' network secondary systems. As loads grow, they increase the secondary buss capacities and add more transformer capacities feeding them. But many of the customers connected years ago were quoted lower fault duty ratings than what are available now. So, whose responsible for upgrading the service equipment or installing C/L fuses?

You mean one that's out of spec? Possibly. But poor protection coordination design is more likely. If the main is spec'd to be quick, the branch breakers should still be quicker.

At high levels of fault current, all of the breakers will be quick enough. Having the main beat a branch breaker is at minimum a nuisance and will make troubleshooting much more difficult.

| Actually, when you really study it, even 10kA is hard to deliver to a | residential service. I don't have the data at my fingertips, but, believe it | or not, the meter is a major consideration in calculating residential fault | current. The other two are the transformer and the length of the service | lateral. | | The only place where you routinely get into trouble in a residential setting | is in high-rise apartments, where there might be a 3000 amp bus duct up | through the middle of the building with each apartment having a tap. In that | case, you just can't use Home-Line or similar breakers. Don't rule out the | use of current-limiting main fuses in these cases | | I lived in a suburb, built in 1955. We had 8 houses on a 10 kVA pole | transformer. Everyone had a 60 or 100 amp service. About 1/3 had central | A/C. Line would dip to about 110 on a hot summer day. Nobody noticed except | me, and that's because I measured it! | | Full load secondary current = 42 amperes (hard to believe, but true!). | Assuming Z of 5% that would give a short circuit current, at the | transformer, in the neighborhood of 840 amperes! A 50kVA transformer of the | same impedance would only give 4200 amperes at the transformer.

So if you turned on your electric oven, all the stove top burners, along with the A/C to cool the kitchen ... and 2-3 of your neighbors would do the same, would there be an earth shattering boom?

I routinely see 100 and 167 kVA transformers in what few new neighborhoods have overhead wiring. No easy markings to read on padmounts. But it would seem fault currents might be hovering in the 10 kA range these days for some. I'm not sure where I'll be building my house in the next few years, but I'll probably be trying to make sure I get on my own transformer so it will be rated just for my demands alone. If not, I might hunt for a

50 kVA dry type unit on Ebay.

|> So how would you wire up a panel if the utility says that your available |> fault current is, for example, 15kA, when individual branch breakers are |> only rated to 10 kA? Two major product lines (Eaton/Cutler-Hammer CH and |> Square-D HOMELINE) have no branch breakers available above 10 kA. Even |> in other product lines, higher rated branch breakers are available only |> lesser choices. | | You select a panel and branch circuit breakers with the appropriate | rating.

You mean a $20,000 industrial panel?

My point is, the choices are rather slim above 10 kA. Not everyone needs to worry about 10 kA, yet, but it seems more and more are.

The CH line has MCB panels rated to 100 kA, but individual breakers do not go above 10 kA.

The QO line has MCB panels rated to 22 kA, and invididual breakers rated at

10 kA and almost all available for 22 kA. A few go above 22 kA. Only the monster 400 A panel could have a substituted L-frame breaker for very high interrupting capacity as much as 100 kA.

The HOM line is rather skimpy.

the BR line reverses what the CH line has ... panels don't go so high but it has branch breakers up to 42 kA.

The ideal would have both branch breakers and main breakers rated at various levels up to what is marketable to residential and small commercial. That should, IMHO, be around the 10 kA, 22 kA, and 42 kA levels.

|> Will people have to be demanding lower fault currents from the utilities, |> which otherwise seem to be increasing them as they deal with the increasing |> demand? Energy-efficiency is also contributing, by pushing down temperature |> rise ratings of transformers, which have some higher fault currents due to |> larger conductor sizes. | | Interesting note: Our local utility (and probably many others) has a | standard practice that | requires fault currents to residential services of 200 and 400 A ratings | to be kept under 10 kA.

Interesting. So keep the transformers from getting too large, and just have more of them?

| For commercial services, its a whole different ball game. In fact, this | is somewhat of an issue with another local utilities' network secondary | systems. As loads grow, they increase the secondary buss capacities and | add more transformer capacities feeding them. But many of the customers | connected years ago were quoted lower fault duty ratings than what are | available now. So, whose responsible for upgrading the service equipment | or installing C/L fuses?

And I assume they can't segment the network?

|> Should one install a transformer to reduce the available fault current |> (with a high interrupting main breaker on the primary side)? |> |> |> Though it has never happened to me, I've heard of cases where people have |> |> had short circuits that resulted in both the branch breaker and the main |> |> breaker tripping. While that can be annoying, obviously you don't want |> |> the main breaker to wait for too long before "deciding" that the branch |> |> breaker isn't able to interrupt the fault. |> | |> | This is a case of poor breaker coordination. It can result from the use |> | of breakers not rated for the available fault current. For example, the |> | branch breaker, with a lower overcurrent rating, would be expected to |> | act sooner for any given fault current. However, if it is incapable of |> | breaking too high a current quickly enough (or at all). Even as the |> | branch breaker is opening, sufficient current will pass, and the main |> | breaker will sense it, so that the main will reach its trip point. The |> | end result is that both breakers operate. |> |> Could it be an overly fast main breaker? | | You mean one that's out of spec? Possibly. But poor protection | coordination design is more likely. If the main is spec'd to be quick, | the branch breakers should still be quicker.

How quick, or slow, should they be under peak fault conditions?

|> I would think that a really serious fault approaching the available fault |> current would put things down in the flat line of the interruption curve, |> and both breakers should be trying within a fraction of a cycle, and at |> least the main should succeed by at least the next zero crossing. |> |> If given an available fault current in excess of available branch breakers, |> wouldn't you at least want your main breaker to be one that interrupts as |> rapidly as possible to minimize the downstream damage from such a fault? |> Or would it, because it trips too often in non-damaging cases, lead people |> to think that in a scenario where there is damage, to just reset it without |> investigating causes and damages? | | At high levels of fault current, all of the breakers will be quick | enough. Having the main beat a branch breaker is at minimum a nuisance | and will make troubleshooting much more difficult.

Which hopefully doesn't happen at home too often. It could be a mess in a commercial situation where fault currents could be larger and more loss incurred when a main breaker opens.

Nope. Just a low line for a while. After a few hours there might be an OT light on the transformer, but that's about it.

Sure. Again, that is at the transformer secondary. Add in the impedance of the service lateral and the meter and you are back at or below 10 kA.

All it takes is .024 ohms to limit the fault current to 10 kA! That would give 1% voltage drop on a fully loaded 100 amp service.

| Sure. Again, that is at the transformer secondary. Add in the impedance of | the service lateral and the meter and you are back at or below 10 kA. | | All it takes is .024 ohms to limit the fault current to 10 kA! That would | give 1% voltage drop on a fully loaded 100 amp service.

While dissipating 240 watts.

I guess as long as it's before the meter, why should I care?

I know, maybe I can get them to run the service lateral zigzagging under my driveway to help melt the snow in the winter.