Is the neutral disconnected in an AC disconnect?

[Carultch]

PG&E requires unfused disconnects? All my projects use fused disconnects because the unfused disconnects only have a 10kAIC rating and most of the time on a supply side interconnection in a C&I project the fault current is more than 10KA.

Here is a link to the PG&E spec for disconnect switches: http://www.pge.com/includes/docs/pd...lity/electrictransmission/handbook/060559.pdf

It states that you cannot switch the neutral (item 7).

One of the functional requirements is:

• A fusible ac disconnect switch is required for generators that do not have over-current protection (i.e., breakers, fuses) at the point of interconnection with the utility.

So if the generator does not have OCP you have to use a fused disconnect, for others it is an option.

An Eaton DG222URB or equal with the optional neutral kit would probably work for you if you don't need a higher AIC rating.

Well, unfused disconnects don't really have a KAIC rating, because they cannot automatically do the function which starts with the letter I in that initialism. They have an SCCR rating instead, which is usually 10 kA.

It is very counterintuitive that an unfused disconnect would have such a weak SCCR rating, because they are usually built with exactly the same main components as their fused counterparts, which have a 200kA SCCR rating.

 

[ggunn]

Well, unfused disconnects don't really have a KAIC rating, because they cannot automatically do the function which starts with the letter I in that initialism. They have an SCCR rating instead, which is usually 10 kA.

It is very counterintuitive that an unfused disconnect would have such a weak SCCR rating, because they are usually built with exactly the same main components as their fused counterparts, which have a 200kA SCCR rating.

Line side connections always need a fused disco and they need to be within 10' of the connection. Their KAIC rating must be higher than the available fault current at the location of the OCPD, which is a little less than the fault current available at the transformer.

Load side connections only need an unfused disco; the KAIC rating of the disco is a non issue because it is not subject to fault current in excess of the rating of the OCPD at the interconnection, and opening the switch is not a fault clearing action. Likewise, the OCPD at the connection needs only to withstand the fault current available from the OCPD upstream of it.

 

[Carultch]

Load side connections only need an unfused disco; the KAIC rating of the disco is a non issue because it is not subject to fault current in excess of the rating of the OCPD at the interconnection, and opening the switch is not a fault clearing action. Likewise, the OCPD at the connection needs only to withstand the fault current available from the OCPD upstream of it.

What if you have a commercial application with a 65 kaic breaker on both sides of the unfused disconnect, and greater than 10 kA incident on the unfused disconnect per your fault current calculations?

 

[ggunn]

What if you have a commercial application with a 65 kaic breaker on both sides of the unfused disconnect, and greater than 10 kA incident on the unfused disconnect per your fault current calculations?

That's why you typically do not use a breaker for your main OCPD on a supply side connection. Most breakers cannot interrupt the available fault current from a "large" transformer.

EDIT: By "most" breakers, I mean breakers that are cost effective to use in a given installation, of course.

 

[Carultch]

That's why you typically do not use a breaker for your main OCPD on a supply side connection. Most breakers cannot interrupt the available fault current from a "large" transformer.

EDIT: By "most" breakers, I mean breakers that are cost effective to use in a given installation, of course.

I never said it was a supply side connection. The example I gave, is one with a load-side connection off of the main panelboard in a large commercial application.

 

[ggunn]

I never said it was a supply side connection. The example I gave, is one with a load-side connection off of the main panelboard in a large commercial application.

In a load side interconnection the fault current available to your interconnection breaker is limited by the trip rating on the MDP OCPD. The fault current at your unfused disco is limited by the trip rating of your interconnection breaker. Both breakers are in the MDP so the disco cannot be between them, or am I missing something?

 

[Carultch]

In a load side interconnection the fault current available to your interconnection breaker is limited by the trip rating on the MDP OCPD. The fault current at your unfused disco is limited by the trip rating of your interconnection breaker. Both breakers are in the MDP so the disco cannot be between them, or am I missing something?

Breaker 1 is a branch breaker in the main distribution panel.
Breaker 2 is the main breaker in an AC Combiner, in the location of the inverters

The unfused disconnect is the utility-required open blade disconnect between Breaker 1 and Breaker 2.

Suppose both breaker 1 and breaker 2 have 65 KAIC, and the incident fault current is 20 kA throughout this circuit (neglect the effect of wire length). Is an unfused disconnect sufficient between these two breakers? Or is fault current alone, a reason why you would need a fused unit, even if the OCPD in it isn't required?

 

[GoldDigger]

Figure reduction in AIC based on added wire length, but do not count on main breaker for AIC reduction unless it is specifically series rated with your disconnect.

 

[iwire]

In a load side interconnection the fault current available to your interconnection breaker is limited by the trip rating on the MDP OCPD. The fault current at your unfused disco is limited by the trip rating of your interconnection breaker. Both breakers are in the MDP so the disco cannot be between them, or am I missing something?

Breakers do not reduce fault current.

 

[ggunn]

Breakers do not reduce fault current.

Why do you say that? If I have a backfed breaker in an MDP with a 200A main breaker in it, how can my breaker get more than 200A fault current? Any more than that and the main will trip.

 

[ggunn]

Breaker 1 is a branch breaker in the main distribution panel.
Breaker 2 is the main breaker in an AC Combiner, in the location of the inverters

The unfused disconnect is the utility-required open blade disconnect between Breaker 1 and Breaker 2.

Suppose both breaker 1 and breaker 2 have 65 KAIC, and the incident fault current is 20 kA throughout this circuit (neglect the effect of wire length). Is an unfused disconnect sufficient between these two breakers? Or is fault current alone, a reason why you would need a fused unit, even if the OCPD in it isn't required?

I don't understand. Your 20kA fault current is less than the 65 KAIC rating of the breakers, so the breakers (one or the other) shouldn't have any problem interrupting it.

 

[iwire]

Why do you say that? If I have a backfed breaker in an MDP with a 200A main breaker in it, how can my breaker get more than 200A fault current? Any more than that and the main will trip.

200 amp is the breaker rating it is not the level of fault current.

During the time between the fault and the time the contacts open and clear the only thing limiting the fault current the circuit impedance.

In your example above you might have 10,000 amps flowing through the circuit before the breaker actually responds and opens. That would be wire those breakers would need to have a 10,000 AIC.

 

[ggunn]

200 amp is the breaker rating it is not the level of fault current.

During the time between the fault and the time the contacts open and clear the only thing limiting the fault current the circuit impedance.

In your example above you might have 10,000 amps flowing through the circuit before the breaker actually responds and opens. That would be wire those breakers would need to have a 10,000 AIC.

Interesting. Are you saying that every breaker in an MDP and every subpanel fed by it must have a KAIC rating equal or greater to the fault current available from the transformer feeding the MDP, less impedance effects?

 

[iwire]

Interesting. Are you saying that every breaker in an MDP and every subpanel fed by it must have a KAIC rating equal or greater to the fault current available from the transformer feeding the MDP, less impedance effects?

Every breaker downstream must be rated for the availbe fault current at each location. (There are series rated combinations that change this, see here http://ecmweb.com/content/behind-2005-nec-changes-24086-series-ratings)

But putting aside the series rated combinations you need to make sure each OCPD has an AIC higher than the available fault current at that location.

So a 480 volt main might have 30,000 fault current, the breakers in that panel will need to be 30,000 rated, but now run a feeder from that 100' to another panel and the available fault current will be lower due the feeder conductors impedance. So those breakers might ony need to be 22,000 AIC.

 

[Carultch]

I don't understand. Your 20kA fault current is less than the 65 KAIC rating of the breakers, so the breakers (one or the other) shouldn't have any problem interrupting it.

You are missing my question. I know the breakers have no problem with this fault current in this example.

I'm asking whether or not the unfused disconnect that on its own has a 10 kA rating of SCCR, can be located between these two breakers. If the answer is no, then does this imply that a fusible disconnect could be required due to fault current alone?

 

[Carultch]

Why do you say that? If I have a backfed breaker in an MDP with a 200A main breaker in it, how can my breaker get more than 200A fault current? Any more than that and the main will trip.

Because whether or not a breaker trips, is a function of time and current. Not all currents in excess of the nominal trip rating will trip it immediately. High amounts of current trip it fast. Low amounts of current trip it slowly. 210A of current can sneak through that very same breaker before it has time to trip, and be incident on everything downstream.

 

[iwire]

Not all currents in excess of the nominal trip rating will trip it immediately.

Kind of nit picky here but regardless of the current no breaker opens immediately. Quickly but not before at least cycle or two passes I believe.

It takes time for the contacts to move and time for the arc to break.

 

[GoldDigger]

Why do you say that? If I have a backfed breaker in an MDP with a 200A main breaker in it, how can my breaker get more than 200A fault current? Any more than that and the main will trip.

The problem is that even the "instantaneous" mag trip will let a lot more current through for between a fraction of a cycle and several cycles.
Current limiting fuses can open at a low enough voltage in the first 1/2 cycle to limit the current to a particular specified value.
A breaker cannot limit it that well, but maybe enough to meet a 5kA or 10kA AIC rating for the disconnect IF it is specifically tested and listed.

 

[ggunn]

You are missing my question. I know the breakers have no problem with this fault current in this example.

I'm asking whether or not the unfused disconnect that on its own has a 10 kA rating of SCCR, can be located between these two breakers. If the answer is no, then does this imply that a fusible disconnect could be required due to fault current alone?

I think the answer to your question is that an unfused disconnect is not supposed to be used to interrupt a fault so its interrupt capability isn't an issue under fault conditions. I wouldn't design a system that way, anyway; the conductors from a backfed breaker typically pass through an unfused disco to an MLO AC combiner panel if more than one inverter is used.

 

[GoldDigger]

I think the answer to your question is that an unfused disconnect is not supposed to be used to interrupt a fault so its interrupt capability isn't an issue under fault conditions. I wouldn't design a system that way, anyway; the conductors from a backfed breaker typically pass through an unfused disco to an MLO AC combiner panel if more than one inverter is used.

An unfused disconnect does not have an AIC since it does not interrupt fault current. It does have an SCCR, which is the fault current it can withstand without damage while waiting for something else to interrupt it. That SCCR still needs to be larger that the fault current available at that point, IMHO.