non-fused for AC disconnect supply side tap

[SolarPro]

4
The code requires OCPD within 10 feet for safety, but the exception requires not-an-OCPD (see jb, #2 above) within the j-box, and drops the whole OCPD within 10 feet thing.
It simply doesn't seem safe.

Perhaps you missed this: "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."

ggunn was correct when he said that the exception in 705.31 isn't much of an exception in practice. Granted, you don't have to install a fuse w/in 10', but you are basically still installing OCP (just not overload protection) within 10' .

 

[PVfarmer]

1 It was meant entirely seriously.

2 Yes it does. It's within the meter enclosure. Look above the meters in your drawing for the three way junction.

3 you either have space for a disconnect - which can be fused -

4 HOWEVER, I do not agree that this leads to ambiguity with regard to 705.31. The point of connection to the service is still the 3-way junction point.

5 Granted, you don't have to install a fuse w/in 10', but you are basically still installing OCP (just not overload protection) within 10'.

1- Well, seriously, the definition of "point of interconnection" isn't in the NEC or the dictionary.

2- See drawing below- there is no 3-way point one could attach limiters for PV.

3- This is interesting (to me)- if you have space for a j-box but not a switch, why not skip the j-box and use a small (say 8 circuit) load panel in the spot where the j-box would have gone. It becomes load side intead of supply side, but...it is also becomes simpler.

4- So you don't agree that- if there are no loads on the premises, just PV to grid, the PV disco switch is the main service disconnect?
What else is it then?

5- You end up with a totally unspecified/unregulated length of conductors (between limiter and remote PV OCPD) which have zero overload protection.
Not my rodeo.

 

[SolarPro]

5- You end up with a totally unspecified/unregulated length of conductors (between limiter and remote PV OCPD) which have zero overload protection.
Not my rodeo.

You do realize that a PV system is a current limited power source, right? Moreover, power electronics based interactive inverters can contribute very little during fault conditions.

Like the PV power source itself, an interactive inverter is a current-limited device. According to the National Renewable Energy Laboratory (NREL) technical report, “Understanding Fault Characteristics of Inverter-Based Distributed Energy Resources,” independent testing conducted at NREL suggests that “inverters designed to meet IEEE 1547 and UL 1741 produce fault currents anywhere between 2 to 5 times the rated current for 1 to 4.25 milliseconds.” The authors explain: “Inverters do not have a rotating mass component; therefore, they do not develop inertia to carry fault current based on an electromagnetic characteristic.”

 

[jaggedben]

1- Well, seriously, the definition of "point of interconnection" isn't in the NEC or the dictionary.

It really is though. See the 'circuit in science' section.

2- See drawing below- there is no 3-way point one could attach limiters for PV.

If I were an AHJ I'd let you put them on the premises side meter terminals, provided there was space and the appropriate lugs, because it's all in the same enclosure and I'd be primarily concerned about a fault on your field wiring outside the enclosure. Your AHJ might not agree. As I said above, nothing guarantees you can meet code with a given piece of equipment.

3- This is interesting (to me)- if you have space for a j-box but not a switch, why not skip the j-box and use a small (say 8 circuit) load panel in the spot where the j-box would have gone. It becomes load side intead of supply side, but...it is also becomes simpler.

Sure, that could be feasible in many situations. There might be reasons why not, but they'd be particular to the details on a given project. You could use service rated MLO panel, too, as long you're not in one of those AHJs that ggunn has told us prohibits them for no sensible reason.

4- So you don't agree that- if there are no loads on the premises, just PV to grid, the PV disco switch is the main service disconnect?
What else is it then?

I would actually agree with you, but others would question whether that is even considered a service. Compare the NEC definition of a service (and note what it delivers energy 'from' and 'to') to 230.2(A)(5) and see if you can judge them consistent with each other. There have been some long threads on this subject on this forum and I (somewhat like ggunn) have given up on there being a winnable argument with the current code. At one point I posted a dozen or two code sections that can be used to argue both ways. Search the forum and see if you can find that.

Ultimately it's the AHJ's opinion that matters.

 

[PVfarmer]

1 It really is though. See the 'circuit in science' section.

2 If I were an AHJ I'd let you put them on the premises side meter terminals, provided there was space and the appropriate lugs, because it's all in the same enclosure and I'd be primarily concerned about a fault on your field wiring outside the enclosure. Your AHJ might not agree. As I said above, nothing guarantees you can meet code with a given piece of equipment.

3 Sure, that could be feasible in many situations. There might be reasons why not, but they'd be particular to the details on a given project. You could use service rated MLO panel, too, as long you're not in one of those AHJs that ggunn has told us prohibits them for no sensible reason.

4 I would actually agree with you, but others would question whether that is even considered a service. Compare the NEC definition of a service (and note what it delivers energy 'from' and 'to') to 230.2(A)(5) and see if you can judge them consistent with each other. There have been some long threads on this subject on this forum and I (somewhat like ggunn) have given up on there being a winnable argument with the current code. At one point I posted a dozen or two code sections that can be used to argue both ways. Search the forum and see if you can find that.

5 Ultimately it's the AHJ's opinion that matters.

1 You could have posted that link a while ago! Ok- we're talking 2 power sources (grid, inverter(s)) connected to one component- is it the main house breaker...or j-box?
Circuits in which a power source is directly connected to two or more components...2 I actually think it's great NatGrid doesn't allow supply-side j-boxes anymore around here. So the dual meter goes where the single meter was, and you can put the PV disco where the j-box would have gone! How could it be simpler?

3 Right, with a MLO panel it would be supply side, but with breaker instead of lugs it's load side. Seems like it'd be a lot easier in general to get load side permitted?

4 Consider this- there's a single inverter connected to the grid, no loads or house involved...so at night!
The inverter might be drawing 3 or 5 watts from the grid. Therefore, it's a service.
Or...is it?
If a tree falls and no one hears it....? :huh: ( https://www.physicsforums.com/threads/solar-grid-tie-inverter-night-time-loss.669824/ )

If the GTI is "rectifying" the Utility AC - you may be loosing the 45W in the inverter, they have bleed down resistors. The current on the AC line may be very non-linear, basically just a little current pulse at the peak of the AC Waveform. This current pulse may be so fast that the electrical mechanical meter does not move at all.

5 Yes, but like you said before, some AHJs say "I've never seen that before..."

 

[GoldDigger]

4. The standard electromechanical (spinning disk) meter uses a mechanical integrating scheme. A very small pulse of current should still apply a force to the disc causing it to spin. As long as there is enough total force to overcome the very small friction the waveform should not really matter.
I would be more concerned about an electronic meter doing sampling via an A to D converter missing small pulses.

mobile

 

[PVfarmer]

A very small pulse of current should still apply a force to the disc causing it to spin.

I'm saying if the disc spins in a way where power enters the premises, it is a service.
Even if a week of total darkness would cost a nickel.

 

[ggunn]

I'm saying if the disc spins in a way where power enters the premises, it is a service.
Even if a week of total darkness would cost a nickel.

That distinction is just silly, IMO. Most AHJ's take a more pragmatic view; whether they consider a connection a service is a function of the physical wiring, not whether or not it delivers half a kWh over the course of a week.

 

[PVfarmer]

You do realize that a PV system is a current limited power source, right? Moreover, power electronics based interactive inverters can contribute very little during fault conditions.

That distinction is just silly, IMO. Most AHJ's take a more pragmatic view; whether they consider a connection a service is a function of the physical wiring, not whether or not it delivers half a kWh over the course of a week.

Yes, I know how inverters work.
I also think quite simply that inside a house does not equal "readily accessible", and the way the term is used in 690.13 contradicts the definition in 100.

Also, if supply side PV isn't a load, service or a separately derived system... what is it?
It's the same wiring when 10kW of power is going out to the grid or

What's wrong with requiring the main PV disco to be adjacent to the meter and outdoors (unless it is impossible to set it up that way),seeing as it is simpler and safer?

690.4 General Requirements
(D) Multiple Inverters. A PV system shall be permitted to
have multiple inverters installed in or on a single building
or structure. Where the inverters are remotely located from
each other, a directory in accordance with 705.10 shall be
installed at each dc PV system disconnecting means, at
each ac disconnecting means, and at the main service disconnecting
means showing the location of all ac and dc PV
system disconnecting means in the building.

705.21 Disconnecting Means, Equipment.
Means shall be provided to disconnect power production equipment,
such as utility interactive inverters or transformers associated
with a power production source, from all ungrounded
conductors of all sources of supply.


III. Disconnecting Means
690.13 Building or Other Structure Supplied by a PhotovoltaicSystem.
Means shall be provided to disconnect
all ungrounded dc conductors of a PV system from all other
conductors in a building or other structure.
(A) Location. The PV disconnecting means shall be jnstalled
at a readily accessible location either on the outside
of a building or structure or inside

705.22 Disconnect Device. The disconnecting means for
ungrounded conductors shall consist of a manually or
power operable switch(es) or circuit breaker(s) with the
following features:
(1) Located where readily accessible.


Art. 100
Accessible, Readily (Readily Accessible). Capable of being
reached quickly for operation, renewal, or inspections
without requiring those to whom ready access is requisite
to actions such as to use tools, to climb over or remove
obstacles, or to resort to portable ladders, and so forth.

 

[PVfarmer]

It's the same wiring when 10kW of power is going out to the grid or

That was supposed to end - "going out to the grid or the inverter is consuming 5 watts at night."

whether they consider a connection a service is a function of the physical wiring,

I certainly understand why you don't want to agree with me, but what about an expert?

---
When the new PV service disconnecting means is not in the same enclosure as the existing service disconnect means, the PV circuit neutral should be bonded to ground and a grounding electrode conductor originating from the PV service disconnect location must be routed to the grounding electrode (250.24).

Even where the PV inverter connection does not have a neutral connection, the utility neutral should be routed to at least the new PV service disconnect and any PV production meter. The meter may require the neutral for proper operation.

http://iaeimagazine.org/magazine/2013/11/16/supply-side-pv-connections-a-closer-look/

 

[ggunn]

That was supposed to end - "going out to the grid or the inverter is consuming 5 watts at night."



I certainly understand why you don't want to agree with me, but what about an expert?

---
When the new PV service disconnecting means is not in the same enclosure as the existing service disconnect means, the PV circuit neutral should be bonded to ground and a grounding electrode conductor originating from the PV service disconnect location must be routed to the grounding electrode (250.24).

Even where the PV inverter connection does not have a neutral connection, the utility neutral should be routed to at least the new PV service disconnect and any PV production meter. The meter may require the neutral for proper operation.

I don't know what qualifies as an expert in your eyes, but I have designed hundreds (literally) of commercial and residential PV systems that have been successfully installed, and a fair portion of them have been supply side interconnected. I submit that that gives me some degree of expertise in the area; I don't know everything but I know a lot.

I defer to the wishes of the AHJ whenever I design a system, and most of them I deal with (with one notable exception - CPS in San Antonio) do not see it the way that you do. Personally, I don't care; I only wish the NEC would clarify it so everyone would be on the same page. You said earlier that I don't care what you say about this. That is a very true statement; I only care what the AHJ who has the power to fail me at my inspection says.

 

[PVfarmer]

I don't know what qualifies as an expert in your eyes

The person who wrote this article, for one...

---
"The methods I recommend below result in the PV disconnecting means being installed under the same rules as service disconnects. These methods are not direct Code requirements, but rather what I consider best practices using existing rules from Code that result in a safe installation utilizing supply-side connections, and they will satisfy most AHJs."

Grounded conductor bonding. The grounded current-carrying conductor on the supply side of service equipment for PV systems should be bonded to the PV disconnect in accordance with Section 250.24.
Raceway bonding. Raceways containing the supply side of service equipment conductors for PV systems should be bonded in accordance with Section 250.92(B).
Wiring methods. The conductors and methods used for supply-side connected PV systems should be limited to those identified in Section 230.43.
( http://solarprofessional.com/articl...utility-interconnections?v=disable_pagination )


250.24 Grounding Service-Supplied Alternating-Current Systems.
(A) System Grounding Connections. A premises wiring
system supplied by a grounded ac service shall have a
grounding electrode conductor connected to the grounded
service conductor, at each service, in accordance with
250.24(A)(1) through (A)(5).

250.92 Services.
(B) Method of Bonding at the Service.

230.43 Wiring Methods for 1000 Volts, Nominal, or Less.
Service-entrance conductors...

 

[ggunn]

The person who wrote this article, for one...

... who is...?

But it makes no difference what he thinks, what you think, or even what I think, for that matter. The AHJ is the final word; what they think is all that matters. I choose my battles with AHJs very carefully and this is not one I choose to have. Whichever AHJ's territory I am working in, I do it their way.

Frankly, I think the issue is splitting hairs in the NEC and I don't think it makes the slightest difference in the safety or efficiency of a PV system either way. Maybe that's why the NEC doesn't unambiguously declare it one way or the other.

And, by the way, whether you have noticed or not, I never said that I think that either way is right or wrong. My only disagreement with you is with your assertion that you know the absolute right way this must be done to be NEC compliant. You don't, because the NEC is itself ambivalent on the issue.

 

[PVfarmer]

1 ... who is...?
But it makes no difference what he thinks,

2 The AHJ is the final word; what they think is all that matters. I choose my battles

3 I don't think it makes the slightest difference in the safety or efficiency of a PV system either way. Maybe that's why the NEC doesn't unambiguously declare it one way or the other.

4 whether you have noticed or not, I never said that I think that either way is right or wrong. My only disagreement with you is with your assertion that you know the absolute right way this must be done to be NEC compliant. You don't, because the NEC is itself ambivalent on the issue.

1- Oh golly. Big hints here- initials are MH, is an internationally renown electrical instructor??
Beyond that, if you can't click the link, that is not my problem.

2- If the AHJ says you don't need a switch, but the POCO says you do, the AHJ's word is worthless because POCO won't let you connect the PV!
They won't mind some PV being installed, just connecting to grid won't be an option...so...?

3 and 4- Those are both your opinions. I think it declares it, but way too ambiguously. People who know far more than I seem to agree with me.

Also, "doesn't unambiguously" is the same as "does ambiguously".
Hence us here, trying to make it less ambiguous...? :huh:
Two things being equally safe, I'd go for simpler...

 

[ggunn]

3 and 4- Those are both your opinions.

Yes, and I stated them as such. I don't know what you think we are arguing about. I design systems to pass inspection; it's as simple as that. The AHJs I deal with are just as convinced that their interpretation of the Code is correct as you are. It makes no difference to me; I am convinced that the systems are safe either way.

 

[ggunn]

1- Oh golly. Big hints here- initials are MH, is an internationally renown electrical instructor??
Beyond that, if you can't click the link, that is not my problem.

It might come as a shock to you, but I don't slog through everything you cut and paste here and click on every link.

 

[ggunn]

Yes, and I stated them as such. I don't know what you think we are arguing about. I design systems to pass inspection; it's as simple as that. The AHJs I deal with are just as convinced that their interpretation of the Code is correct as you are. It makes no difference to me; I am convinced that the systems are safe either way.

BTW, Mike Holt also says that if there are 6 handles already on a service, a supply side connected PV system does not count as a seventh handle.

 

[PVfarmer]

BTW, Mike Holt also says that if there are 6 handles already on a service, a supply side connected PV system does not count as a seventh handle.

I know. The article says you could have 6 handles for a service and another 6 for PV...
but do NOT DO THAT!

Because a PV system is outside that definition, the PV disconnecting means is not governed by the six-handle rule. Section 690.14(C) (4) requires that PV system disconnecting means must have no more than six switches or six circuit breakers. Given the Code definitions and requirements, a supply-side connection could have up to six PV-specific disconnects in addition to the utility service disconnects.
This conclusion contradicts solar integrators’ traditional line of thinking,

The point I get is that even if PV discos aren't technically called service discos, they are wired the same, and therefore bonded to a GEC.

The methods I recommend below result in the PV disconnecting means being installed under the same rules as service disconnects.

 

[PVfarmer]

i feel so much better now.

i feel so much better now.

even if PV discos aren't technically called service discos,

A big thanks to the NREL!!

Utility-Interconnected Photovoltaic Systems:
Evaluating the Rationale for the Utility-Accessible External Disconnect Switch
M.H. Coddington, R.M. Margolis, and J. Aabakken
( http://www.nrel.gov/docs/fy08osti/42675.pdf )

This answers the OP's question, and mine.

Well, of course I think a non-fused EDS is redundant because I don't see why having a fused PV disco adjacent to an outdoor meter is either optional or a problem.


They are TECHNICALLY CALLED:
The utility-accessible alternating current (AC) external disconnect switch (EDS) for
distributed generators, including photovoltaic (PV) systems, is a hardware feature that
allows a utility’s employees to manually disconnect a customer-owned generator from
the electricity grid.

Put simply, the utility-accessible EDS is increasingly viewed as redundant and
unnecessary for residential and small-commercial PV systems with UL-listed inverters.
Eight state PUCs (i.e., Arkansas, Delaware, Florida, Maryland, Nevada, New Jersey,

New Hampshire, and Utah) have reached this conclusion and eliminated their EDS
requirements for systems that meet criteria, and nine state PUCs have decided to leave
the EDS decision up to individual utilities. In the states with utility choice, at least five
utilities have eliminated the EDS requirement. These states and utilities accounted for
more than 80% of total installed PV capacity in the United States in 2006.

 

[ggunn]

The person who wrote this article, for one...

I get it. I really do. I got it way back there. You are of the opinion that every AHJ in Texas save CPS in San Antonio is wrong. Feel free to contact ONCOR, Austin Energy, the Pedernales Electric Co-op, Bluebonnet Electric, et al. and tell them so. Let me know how it turns out; if you succeed in getting them to change their interconnection guidelines, I will design my systems to fit them. Otherwise, this conversation is pointless.