Question on NEC 2017 NEC 705.12(B)(2)(3)

Status
Not open for further replies.
M

milkyway

Member
Location
95112
Occupation
Engineer
How does this code affect upstream panels from the interconnection point? For instance, if PV is tied into a load panel which meets NEC 2017(B)(2)(3), does the main switchgear panel which feeds this load panel also need to comply?
 
Any distribution equipment that can simultaneously receive power from more than one source (e.g. utility and PV) needs to comply.

Cheers, Wayne
 
wwhitney said:
Any distribution equipment that can simultaneously receive power from more than one source (e.g. utility and PV) needs to comply.

Cheers, Wayne
Click to expand...

Thanks. This opens up a new question. When doing calculations for upstream distribution equipment, what do I use as my power source current? I assume the smaller of the following: 125% power source output current OR rating of OCPD protecting the downstream panel.

For example, if the OCPD on the downstream distribution panel is smaller than 125% of my output current, do I still have to use 125% output current when doing compliance calculations for upstream panel(s)?
 
I think you may have exceeded my knowledge, so we'll see what others have to say.

But I'm having trouble thinking of a case where the OCPD could be smaller than 125% of the inverter output current. If this is grid-tie, the intervening panelboards might not have any active loads, so there needs to be a feeder path from the inverter to the grid where everything is rated at least 125% of the inverter output current.

Cheers, Wayne
 
wwhitney said:
I think you may have exceeded my knowledge, so we'll see what others have to say.

But I'm having trouble thinking of a case where the OCPD could be smaller than 125% of the inverter output current. If this is grid-tie, the intervening panelboards might not have any active loads, so there needs to be a feeder path from the inverter to the grid where everything is rated at least 125% of the inverter output current.

Cheers, Wayne
Click to expand...

Thanks again Wayne,

I don't want to get too off the question, but this is for a microgrid design which contains both PV and energy storage with emergency backup. I know this is probably very uncommon design and requires some degree of custom engineering.

A new critical load panel is required to backup specified loads. These specified loads are currently fed from the main panel but will be moved to this new critical load panel. This critical load panel will couple the non utility power sources (PV, energy storage). Since the critical load panel is new, we have the ability to size it properly to satisfy this code requirement.

In order to not go overboard with the bus rating on the critical load panel, the OCPD is significantly smaller than the bus rating of the critical panel, something that you don't see very often. Given that that the PV and energy storage will feed a decent size amount of loads in the critical load panel, the OCPD can be sized smaller than 125% of the output power source current. Further, energy storage current and PV current will never output into the circuit at the same time (energy storage charges from PV and only outputs after the sun goes down). Although NEC does not care about this fact, this is how we can get away with a smaller OCPD on the critical load panel compared to 125% of total inverter output current.

I am not sure how to satisfy this code requirement for this use case in regards to the upstream panel. I can see only the following options:

1. We are stuck with abiding by 2017 NEC 705.12(B)(2)(3)(c) for the upstream breaker
2. Based on info above, we can use size of OCPD of critical load panel as output of inverter current for calculations for NEC 705.12(B)(2)(3)(a) and NEC 705.12(B)(2)(3)(b) [this is doubtful but better option for us]
3. We have to use 125% full output power of all power sources for doing cellulations on upstream breaker, regardless of OCPD size on critical load panel (this would be unfortunate in our case)
4. Some other option someone can point me to?

Thanks again for the help, I hope this gives more context for folks to help out here.
 
milkyway said:
Further, energy storage current and PV current will never output into the circuit at the same time (energy storage charges from PV and only outputs after the sun goes down). Although NEC does not care about this fact, this is how we can get away with a smaller OCPD on the critical load panel compared to 125% of total inverter output current.
Click to expand...
So, my first instinct is that the NEC doesn't let you down-size the OCPD, unless the system is listed as a Power Control System under 2020 705.13 (or you get permission from the AHJ if you are under an earlier version of the NEC). I.e. even though it will work technologically, the regulatory aspect is lagging.

A one line diagram would be helpful for thinking more about this.

Cheers, Wayne
 
wwhitney said:
So, my first instinct is that the NEC doesn't let you down-size the OCPD, unless the system is listed as a Power Control System under 2020 705.13 (or you get permission from the AHJ if you are under an earlier version of the NEC). I.e. even though it will work technologically, the regulatory aspect is lagging.

A one line diagram would be helpful for thinking more about this.

Cheers, Wayne
Click to expand...

Hmmm. I'm going to provide a one-line as you suggested.

By OCPD I do mean a breaker. Maybe I don't fully understand, why can't you derate an OCPD? Conductor sizes remain the same, so everything is still protected.

For example, if a main breaker was 600A, can I not replace that breaker with a 400A to free up bus space for say a PV installation?

Thanks again!
 
Well, (2020) 705.30 basically says that the OCPD shall be sized at 125% of the continuous output current rating of the power production equipment, unless elsewhere required or permitted in the NEC. I expect earlier versions have the same language, perhaps in 690 or a different part of 705. So absent the allowance in (2020) 705.13, you have to size the wiring and the breakers for the full output rating of the inverters.

At least, that's my understanding. I know that Tesla Powerwall, for example, now supports limiting the output current in various ways, but I don't know if Tesla has gotten the Powerwall listed as a Power Control System per (2020) 705.13.

Cheers, Wayne
 
wwhitney said:
Well, (2020) 705.30 basically says that the OCPD shall be sized at 125% of the continuous output current rating of the power production equipment, unless elsewhere required or permitted in the NEC. I expect earlier versions have the same language, perhaps in 690 or a different part of 705. So absent the allowance in (2020) 705.13, you have to size the wiring and the breakers for the full output rating of the inverters.

At least, that's my understanding. I know that Tesla Powerwall, for example, now supports limiting the output current in various ways, but I don't know if Tesla has gotten the Powerwall listed as a Power Control System per (2020) 705.13.

Cheers, Wayne
Click to expand...

Thanks Wayne,

Understood, OCPD for each inverter will not be derated, I think that was a misunderstanding. As you mentioned, a one-line would help. Will provide tomorrow if you are still interested!
 
Sure.

When you say microgrid, do you mean there won't be any utility connection? That would put a different complexion on things. If it is grid connected, and the microgrid occurs only when the grid is down, I think that the 705.30 required sizing has to be maintained to the grid.

Cheers, Wayne
 
wwhitney said:
Sure.

When you say microgrid, do you mean there won't be any utility connection? That would put a different complexion on things. If it is grid connected, and the microgrid occurs only when the grid is down, I think that the 705.30 required sizing has to be maintained to the grid.

Cheers, Wayne
Click to expand...

The latter is correct, it is grid-tied and microgrid islands during a utility outage event.
 
milkyway said:
How does this code affect upstream panels from the interconnection point? For instance, if PV is tied into a load panel which meets NEC 2017(B)(2)(3), does the main switchgear panel which feeds this load panel also need to comply?
Click to expand...

If you interconnect at a subpanel, assume the worst case scenario that all local load in the subpanel diminishes to zero, while the interconnected source is at full current. Carry that full current upstream to all panelboards between it and the service point. They all need to comply with a 705.12 rule for busbar protection. This is one reason why it is best to aim to interconnect at the most-main panel that is practical. Especially if your plan is to use a larger busbar than main breaker (like 600A main on 800A busbar), in order to comply.

Bear in mind that the interconnected source current is going to be a smaller and smaller fraction of the busbar rating, the higher upstream you go through the panelboard/switchboard hierarchy. As an example, suppose you interconnect with a 40A breaker at a 200A main/bus subpanel, which is the maximum that meets the 120% rule. This subpanel is fed from a 400A main/bus main panelboard. If 40A worth of 125% current is compliant to interconnect on the 200A panelboard, it definitely is compliant on the 400A panelboard.
 
milkyway said:
Hmmm. I'm going to provide a one-line as you suggested.

By OCPD I do mean a breaker. Maybe I don't fully understand, why can't you derate an OCPD? Conductor sizes remain the same, so everything is still protected.

For example, if a main breaker was 600A, can I not replace that breaker with a 400A to free up bus space for say a PV installation?

Thanks again!
Click to expand...
It is my understanding that you are allowed to derate breakers provided that it is behind a locked door or locked equipment panel, so it is only accessible to qualified personnel. It is advisable to apply a label to equipment where code-compliance depends on a field-derating of a breaker, so future site personnel know the correct setting and why.
 
IMHO, the specific case of the upstream breaker from the critical loads panel (which contains the PV and storage backfeed) is pretty clear. It must be rated for 125% of inverter or more. There is no allowance in the NEC for reducing it based just on the fact that there are local loads in the panel. Except the mentioned managed system exception.

Reducing the upstream breaker to satisfy the 120% rule is fine as long as you do not drop below 125% of inverter output.

Sent from my Pixel 4a using Tapatalk
 
Thank you for the responses thus far. I am attaching a sketch for more clarity.

DWGA.PNG

The critical load panel as I have drawn it should meet 2017 NEC 705.12(B)(2)(3)(b)
The main panel should meet with 2017 NEC 705.12(B)(2)(3)(c)

As mentioned, the energy storage and PV will not be running at the same time. I am not sure if that matters regarding the code.

125% PV inverter output current = 312.5A
125% energy storage inverter output current = 312.5A

Questions:
1. Any code compliance issues with this sketch?
2. What this lowest rating I am allowed to put on breakers 1,2, and 3? And what code section does this correspond to?
 
You haven't drawn in the transfer switch you would need to allow island operation when the grid is down.

Also, I take your drawing to mean that there are PV inverters with a maximum continuous output of 250A, as well as separate storage inverters with a maximum continuous output of 250A

Let me answer your question under two different scenarios (this is all covered under 705.12(B)(2)(3)):

Option A) The PV and storage inverters are interlocked in a way that your AHJ will recognize (e.g. a 2020 Power Control System), and so the combined maximum inverter output current is 250A.

1) Looks OK to me.

2) Breaker 3 is redundant and can match breaker 2 or be eliminated. Breaker 2 needs to be the larger of (a) 125% * 250A (power export) and (b) the NEC calculated load for Load A, Load B, and the storage inverter when it is charging. If the storage only charges from PV and is interlocked in a way that your AHJ will recognize, then that last contribution can be taken to be 0. The bus in the critical panel can be downsized to (80% of Breaker 2) + 250A.

Breaker 1 can be sized the same way as Breaker 2: the larger of (a) 125% * 250A (power export) and (b) the NEC calculated load for all the loads.

Option B) You need to consider that the PV and the storage inverters are both producing 250A simultaneously.

1) You have a number of problems. Breakers 1, 2, and 3 have to be at least 125% * 500A, to start.

2) Sizing is the same as in Option A), but you need to use 500A instead of 250A. The main panel bus needs to be larger.

Cheers, Wayne
 
P.S. A more efficient connection topology can often be had using (2017) 705.12(B)(1). For example, under Scenario (B), one possibility depending on the load calc, where the vertical line is a 500A Feeder:

Code: 
Utility
  |
500A 100% rated service disconnect
  |
  |--- Feeder Tap based on 1000A Upstream -- Non-Critical Main Breaker Panel
  |
Transfer Switch
  |
  |--- Feeder Tap based on 1000A Upstream -- Critical Loads Main Breaker Panel
  |
Generation Panel with 500A Continuous Inverter Output
 
Late to the thread, and I haven't read every word of it, but a couple clarifications...

First, the requirement to use 125% of inverter output is stated plainly near the beginning of 705.12. For example in the 2020 NEC it is 705.12(B) "The power source output circuit current multiplied by 125% shall be used in ampacity calculations for 705.12(B)(1) through (B)(3.)". Similar language can be found back to the 2014 NEC.

Second, I highly doubt that any AHJ (unless they are simply incompetent) will allow for less than the full output of both PV and storage, unless you have a listed Power Control System. Further, they may not even recognize Power Control Systems, since they aren't mentioned before the 2020 NEC and many places aren't on that yet. So if I were you and the customer does not have the budget for a 700A instead of 400A feeder to the micro-grid I would tackle that issue early on.
 
wwhitney said:
You haven't drawn in the transfer switch you would need to allow island operation when the grid is down.

Also, I take your drawing to mean that there are PV inverters with a maximum continuous output of 250A, as well as separate storage inverters with a maximum continuous output of 250A

Let me answer your question under two different scenarios (this is all covered under 705.12(B)(2)(3)):

Option A) The PV and storage inverters are interlocked in a way that your AHJ will recognize (e.g. a 2020 Power Control System), and so the combined maximum inverter output current is 250A.

1) Looks OK to me.

2) Breaker 3 is redundant and can match breaker 2 or be eliminated. Breaker 2 needs to be the larger of (a) 125% * 250A (power export) and (b) the NEC calculated load for Load A, Load B, and the storage inverter when it is charging. If the storage only charges from PV and is interlocked in a way that your AHJ will recognize, then that last contribution can be taken to be 0. The bus in the critical panel can be downsized to (80% of Breaker 2) + 250A.

Breaker 1 can be sized the same way as Breaker 2: the larger of (a) 125% * 250A (power export) and (b) the NEC calculated load for all the loads.

Option B) You need to consider that the PV and the storage inverters are both producing 250A simultaneously.

1) You have a number of problems. Breakers 1, 2, and 3 have to be at least 125% * 500A, to start.

2) Sizing is the same as in Option A), but you need to use 500A instead of 250A. The main panel bus needs to be larger.

Cheers, Wayne
Click to expand...

Understood and very helpful. Thank you for the detailed reply. This back and forth has essentially answered the bulk of what I needed to know.

I didn't show the disconnecting point for microgrid isolation for simplicity, but essentially breaker 2 will be an advanced stored energy breaker which can be opened/closed on command from power control system and interconnection relay. Had done a project a few years back with an oversized inverter for a battery stack, and AHJ was ok with ensuring PCC would not output more than xxKW, though each AHJ is a new adventure full of unknowns.
 
wwhitney said:
P.S. A more efficient connection topology can often be had using (2017) 705.12(B)(1). For example, under Scenario (B), one possibility depending on the load calc, where the vertical line is a 500A Feeder:

Code: 
Utility
  |
500A 100% rated service disconnect
  |
  |--- Feeder Tap based on 1000A Upstream -- Non-Critical Main Breaker Panel
  |
Transfer Switch
  |
  |--- Feeder Tap based on 1000A Upstream -- Critical Loads Main Breaker Panel
  |
Generation Panel with 500A Continuous Inverter Output
Click to expand...

I think I follow somewhat. In this topology, would utility feed critical load panel during normal operation? That would be a requirement here. I don't like the word transfer switch because my brain thinks of a sitting diesel genset isolated from grid and not called upon until asked and transfer switched engaged.
 
Status
Not open for further replies.
Top