MLO Panel as a service.

[SolarPro]

Regarding the 120% rule, even with OCPD at the inverter and a 30A disconnect in the MLO panel, aren't we technically still on the supply side of all building loads? And if so, I don't see how the 120% rule applies.

Correct.

 

[Pharon]

MLO Panel as a service.

The OCPD needs to be located at the source of the fault currents, and a utility-interactive inverter itself is not a potential source of higher than expected currents.

Here's a relevant discussion from a recent article Understanding the NEC 2014 and Its Impact on PV Systems:

Thanks, SolarPro - I knew you'd come through. So if I'm reading that correctly, as long as there's proper fusing on the DC side in the combiner box, no OCPD is required on the AC side of the inverter? Good to know.

 

[SolarPro]

To be clear, you still need OCP. But it needs to be located at (or as close as practicable to) the point of interconnection or in an inverter aggregation panel, rather than at the inverter itself.

This 2014 NEC change is also relevant:

Section 705.31 ?Location of Overcurrent Protection?

This new section limits the length of unprotected conductors in supply-side interconnections to 10 feet, due to safety concerns. According to a substantiation provided by CMP No. 4 in the ROP: ?Often, when connections are made to, or ahead of, existing service-entrance equipment, space limitations do not allow for a disconnecting means with overcurrent protection to be adjacent to the service entrance equipment and still have proper working clearances. Consequently, many of the required disconnects with overcurrent protection are being installed remote from the service entrance equipment.?

Though subject to fault current from the utility, these conductors are not protected by the utility transformer?s primary OCPD(s). The CMP concludes: ?A fault on these conductors will likely result in a violent, explosive conductor vaporization and potential equipment damage or complete burn down.? The unanimous decision to limit the unprotected conductor length to 10 feet mitigates this safety hazard.

Where the overcurrent protection for the parallel power production equipment is located more than 10 feet from the point of connection, an exception allows the use of cable limiters to provide short-circuit protection. Note that these cable limiters must be installed ?at the point where the electric power production conductors are connected to the service.? A cable limiter is different than a fuse in that it isolates a conductor in response to short-circuit current only and does not provide overload protection. Due to space constraints, it may prove easier to install cable limiters at a service entrance in accordance with this exception rather than install a fused disconnecting means within 10 feet.

The addition of Section 705.31 brings the guidance for PV supply-side connections into closer agreement with requirements in Article 230 pertaining to service disconnecting means. For example, Section 230.70(A) requires that service disconnecting means be located at ?a readily accessible location outside of a building or inside nearest the point of entrance of the service conductors.? Section 230.91 further requires that overcurrent protection be integral to or immediately adjacent to the service disconnecting means. Note that where disconnecting means for PV power production sources are located within 10 feet of the main service disconnecting means, they could be used, in conjunction with contactor combiners or other equipment, to initiate rapid shutdown for PV systems on buildings.

 

[Pharon]

Okay now I'm confused again. Typically, a utility disconnect is required near the supply side connection point, but I'm not sure if it's always fused or not. It seems backwards that OCPD is necessary at the point of utility connection instead oat the source itself (i.e. the inverter).

 

[GoldDigger]

Okay now I'm confused again. Typically, a utility disconnect is required near the supply side connection point, but I'm not sure if it's always fused or not. It seems backwards that OCPD is necessary at the point of utility connection instead oat the source itself (i.e. the inverter).

The thing that makes this different, as mentioned earlier, is that although the inverter is a power source it is inherently incapable of delivering high current into a short in the branch/feeder wiring.
The utility "load" on the other hand is capable of supplying enormous current into either a failed inverter or shorted wiring.
OCPD at the inverter end protects only the inverter. OCPD at the interconnect end protects both at once.

Tapatalk!

 

[ggunn]

Okay now I'm confused again. Typically, a utility disconnect is required near the supply side connection point, but I'm not sure if it's always fused or not. It seems backwards that OCPD is necessary at the point of utility connection instead oat the source itself (i.e. the inverter).

It's because you size the conductors to handle the maximum current the inverter can produce so that the inverter could not possibly endanger them. Any fault current in excess of the rating of the conductors could only come from the service. Under normal conditions the current flows from the inverter to the service, but in a fault condition the current flows from the service to the fault.

 

[SolarPro]

Exactly. Study the requirements in 690.8(A)(3) and 690.9(B). The OCPD on the inverter output circuit is always at least 125% of the inverter continuous output current rating. However, utility interactive inverters have no surge capability beyond the rated continuous output current. Therefore, the inverter cannot be a source of currents in excess of what the conductor and OCPD are rated for.

 

[Pharon]

Interesting. So I guess the takeaway is to make sure to have OCPD (fuses or breaker) within 10 feet of the utility supply side connection (for those types of installs). I will take a look at those sections - thanks for the info.

 

[Smart $]

Funny you should bring this up, because I've seen several installations with no OCPD at the inverter, and for the life of me I can't figure out what's protecting the feeder. Is it something internal in the inverter itself? Maybe something on the DC side? I'm sure there's something about it in article 690 or 705 but don't have a Code handy right now.

It doesn't matter where the ocpd is located, with the inverter output conductor sized at 125% rated output, it will be protected. Any faults on the inverter side of the ocpd will create a utility-side voltage drop and the inverter is required to automatically disconnect.

Looked up ocpd required within sight of inverter. Only applies to an inverter mounted in a not readily accessible location (such as on a roof) [690.15(A) {2014}].

Regarding the 120% rule, even with OCPD at the inverter and a 30A disconnect in the MLO panel, aren't we technically still on the supply side of all building loads? And if so, I don't see how the 120% rule applies.

While I agree, contrary to SolarPro's firm assertion, not everyone is on the same page...

If you have a 30A ocpd at the inverter and a 30A ocpd at the MLO service panel, don't you have feeder conductors in between? Wouldn't the ocpd in the MLO service panel thus be a service disconnect by definition? The feeder is on the load side of this service disconnecting means, right? What part of 690 and/or 705 covers this feeder's protection?

 

[Pharon]

While I agree, contrary to SolarPro's firm assertion, not everyone is on the same page...

If you have a 30A ocpd at the inverter and a 30A ocpd at the MLO service panel, don't you have feeder conductors in between? Wouldn't the ocpd in the MLO service panel thus be a service disconnect by definition? The feeder is on the load side of this service disconnecting means, right? What part of 690 and/or 705 covers this feeder's protection?

I don't think putting a disconnect at the MLO panel makes it a feeder - at least not in that direction.

The normal power flow is from the inverter to the panel. As long as the OCPD is satisfied on the inverter side of the feeder, then all is well. At no point do I see it being a feeder in the opposite direction, though I do acknowledge the potential fault current contribution/protection in that direction.

Bottom line is that it's either a supply side connection or a load side connection. It can't be both. At least not from an overload standpoint.

 

[Smart $]

I don't think putting a disconnect at the MLO panel makes it a feeder - at least not in that direction.

The normal power flow is from the inverter to the panel. As long as the OCPD is satisfied on the inverter side of the feeder, then all is well. At no point do I see it being a feeder in the opposite direction, though I do acknowledge the potential fault current contribution/protection in that direction.

Bottom line is that it's either a supply side connection or a load side connection. It can't be both. At least not from an overload standpoint.

I'm not an adversary of your logic, but can you cite reference to the section(s) which covers these conductors' protection and rating requirements?

Is there anything in the code which prevents these conductors from being spliced or tapped to supply a load?

Under the NEC, there are only three categories of current-carrying system conductors: 1) power supply source (service, separately derive system, or other), 2) feeder, and 3) branch circuit. Power source conductors only exist on the line side of their respective ocpd's. You can look up the definition of the other two, then tell me which of the three apply to these conductors...

 

[Pharon]

MLO Panel as a service.

705.12(A) vs. 705.12(D). And yes, the supply side conductors per 705.12(A) cannot be tapped or spliced except as permitted in 230.82. As for what type of conductors they are in that scenario... wouldn't they be a power supply source?

 

[SolarPro]

If you have a 30A ocpd at the inverter and a 30A ocpd at the MLO service panel, don't you have feeder conductors in between?

If you have a 30A OCPD at the MLO service panel, why would you put another 30A OCPD at the inverter where it can't protect anything. Maybe you put an equipment disconnect at the inverter; there are any number of scenarios where that might be required or useful. But the OCPD goes at the end of the inverter output circuit at the point of interconnection with the utility.

 

[jaggedben]

*Without looking it up, aren't inverter output OCPD's required to be within sight of the inverter?

Nope. The only requirements of that nature are for a disconnect only, not OCPD. (e.g. 690.14(D)(2) and 690.15 in the 2011 code.)

Looked up ocpd required within sight of inverter. Only applies to an inverter mounted in a not readily accessible location (such as on a roof) [690.15(A) {2014}].

Don't have the 2014 yet. (Ordering it now!) But I think that's just a disco requirement, not OCPD.

...Is there anything in the code which prevents these [inverter output circuit] conductors from being spliced or tapped to supply a load?

Yes, the 'Dedicated OCPD and disco' requirement. 705.12(D)(1) in 2011.

Under the NEC, there are only three categories of current-carrying system conductors: 1) power supply source (service, separately derive system, or other), 2) feeder, and 3) branch circuit. Power source conductors only exist on the line side of their respective ocpd's. You can look up the definition of the other two, then tell me which of the three apply to these conductors...

Branch circuit, if we accept your premise. 'Inverter output circuit' if we don't accept your premise, and look to 690 definitions.

 

[Smart $]

705.12(A) vs. 705.12(D). And yes, the supply side conductors per 705.12(A) cannot be tapped or spliced except as permitted in 230.82. As for what type of conductors they are in that scenario... wouldn't they be a power supply source?

If you have a 30A OCPD at the MLO service panel, why would you put another 30A OCPD at the inverter where it can't protect anything. Maybe you put an equipment disconnect at the inverter; there are any number of scenarios where that might be required or useful. But the OCPD goes at the end of the inverter output circuit at the point of interconnection with the utility.

Nope. The only requirements of that nature are for a disconnect only, not OCPD. (e.g. 690.14(D)(2) and 690.15 in the 2011 code.)

Don't have the 2014 yet. (Ordering it now!) But I think that's just a disco requirement, not OCPD.

Yes, the 'Dedicated OCPD and disco' requirement. 705.12(D)(1) in 2011.

Branch circuit, if we accept your premise. 'Inverter output circuit' if we don't accept your premise, and look to 690 definitions.

Yes, the requirement is for disconnect only, not ocpd.

However, the scenario being questioned is 30A ocpd at MLO service panel and 30A ocpd at inverter (breaker used as disconnecting means {...and/or AC aggregation panel}).

It's not a branch circuit unless its splice ot tap conductors directly supply a load... and its not a power source conductor because it is on the opposite side of the ocpd's of both power sources. What does that leave???

 

[SolarPro]

Inverter output circuit: "the conductors between the inverter...and another electric power production source, such as a utility"

 

[Smart $]

Inverter output circuit: "the conductors between the inverter...and another electric power production source, such as a utility"

Of interest in the definition is the demarcation point, otherwise known as the Point of Connection. 705.12 errantly assumes one source is a service and completely disregards the possibility of multiple other sources excluding a service. However, our discussion is regarding interconnection of a service (utility provided by definition) and one or more utility-interactive inverters.

The issue being debated can be reduced to the following:

• If you have one ocpd between an inverter and service conductor, you have a supply-side connection.
• If there are two or more ocpd's between inverter and service conductor, you have a load-side connection.

705.12 Point of Connection. The output of an interconnected
electric power source shall be connected as specified
in 705.12(A), (B), (C), or (D).

(A) Supply Side. An electric power production source shall
be permitted to be connected to the supply side of the service
disconnecting means as permitted in 230.82(6). ...

?

(D) Utility-Interactive Inverters. The output of a utility-
interactive inverter shall be permitted to be connected to
the load side of the service disconnecting means of the
other source(s) at any distribution equipment on the premises.
Where distribution equipment, including switchgear,
switchboards, or panelboards, is fed simultaneously by a
primary source(s) of electricity and one or more utility-
interactive inverters, and where this distribution equipment
is capable of supplying multiple branch circuits or feeders,
or both, the interconnecting provisions for the utility-
interactive inverter(s) shall comply with 705.12(D)(1)
through (D)(6).
?

 

[jaggedben]

Of interest in the definition is the demarcation point, otherwise known as the Point of Connection. 705.12 errantly assumes one source is a service and completely disregards the possibility of multiple other sources excluding a service. However, our discussion is regarding interconnection of a service (utility provided by definition) and one or more utility-interactive inverters.

The issue being debated can be reduced to the following:

• If you have one ocpd between an inverter and service conductor, you have a supply-side connection.
• If there are two or more ocpd's between inverter and service conductor, you have a load-side connection.

Each connection is only one connection and the rules which apply to it apply only to that connection. An inverter output circuit may have more than one connection with different rules applying to those connections.

Example: An inverter is tied into a subpanel, which is in turn supplied by one of two breakers in an MLO service panel. The connection in the subpanel is load side and subject to the 120% rule, etc. etc. The connection in the service panel is a supply side connection and not subject to those rules.

I actually referenced just such an example above. The full setup was 40A solar into 125A sub supplied by a 70A breaker. The 70A sub breaker was one of two service disconnects in a 100A service panel.

It's not really necessary to have a demarcation point between the inverter output circuit and the rest of the distribution system. They need not be mutually exclusive. Parts that serve both purposes just need to meet the rules for both purposes.

 

[SolarPro]

The issue being debated can be reduced to the following:

• If you have one ocpd between an inverter and service conductor, you have a supply-side connection.
• If there are two or more ocpd's between inverter and service conductor, you have a load-side connection.

Talk about a minority opinion. :happyno: The Article 100 definition of a PV system is broad enough to include an inverter aggregation panel. In effect, each subarray has an inverter dedicated to it. The output of these inverters are combined at the inverter accumulation panel. And the output of this inverter combiner panel can be supply- or load-side connected. This whole system?all of the subarrays and inverters in aggregate?meets this definition:

"The total components and subcomponents that, in combination, convert solar energy into electric energy suitable for interconnection to a utilization load."

 

[GoldDigger]

But IMHO the inclusion of "suitable for interconnection to a utilization load" actually muddies the issue.
Once the energy leaves the inverter output terminals it is "suitable" for load connection.
But even after AC combiners and disconnects it may not yet have reached a point where utilization loads are actually connected even after it has passed the interconnection point to POCO.
So either the term includes parts of the utility network too or it stops at the inverter output.
None of the upstream wiring changes the *form* of the energy, just the wiring that is delivering it and the fusing and metering involved.
To me this actually has the opposite result to what you propose.

Tapatalk!