NEC PCS rule for uncontrolled/unmonitored sources

H

hitehm

Senior Member
Location
Las Vegas NV
All - Please see the attached SLD of the backfed PW3 system we're installing with an existing 20A backfed unmonitored system and refer to 705.12 and 702.13 for this discussion. Our issue is as follows: The engineering team is setting the PW3 Panel Limit feeding the 100A main panel bus to 60A. They limited the max monitored and controlled pcs sources (the pw3 and solar) to 80% of the bus rating. This is based on Tesla's own doc that states (in so many words) you must consider the bus to be 80% of its rating because of the way manufacturers test their equipment. That's fine, so that would make the max PCS output 80A. However, they are also subtracting the 20A unmonitored existing system from the PCS making the max PCS panel limit 60A. Although this seems logical in several ways, we can find absolutely no code or AHJ rule that says to do this. In all of 705.13 the only language that deals with unmonitored sources on the busbar is 705.13A that states:

705.13A Any busbar or conductor on the load side of the service disconnecting means that is not monitored by the PCS shall comply with 705.12


There is no mention of subtracting this value from the monitored controlled source. The engineering team highlight 705.13B as their reasoning but it vaguely at best only states the PCS can be no more than the OCP it's connected to if connected to an unmonitored bus. Tesla considers this bus monitored if it has CTs on the utility consumption input to the bus and of course its own output feeding the bus.

705.13B - Where the PCS is connected to an overcurrent device protecting any busbar or conductor not monitored by the PCS, the setting of the PCS controller shall be set within the ratings of that overcurrent device.

Therefore, we believe by the CODE, the PCS Panel Limit setting should be 80A not 60A and let the unmonitored existing system follow the 120% rule of 705.12. But this brings up the obvious problem that the existing system can still apply 20A in addition to the 80A pcs from the PW3 which combined is more than 80% of the bus rating. Also, without monitoring the existing system, even though you are monitoring the main bus source inputs from both the utility and the PW3, is the bus FULLY monitored if you don't monitor ALL the other source on that bus?

So the question for all is, how is this interpreted correctly and should the unmonitored source be subtracted from the PW3 panel limit?


pcs w 2 sources.png
 
The PCS + Existing Solar would need to met 705.12(B). I don't know if I am understanding the situation correctly, but you can not treat each separately. That is why the code uses the word "source(s)". Implying if there are more than one source you need to consider that when doing the calculation on the service.

P.S. Is there no production meter on the existing? The utility isn't going to mind having 3 meters?
 
Elect117 said:
The PCS + Existing Solar would need to met 705.12(B). I don't know if I am understanding the situation correctly, but you can not treat each separately. That is why the code uses the word "source(s)". Implying if there are more than one source you need to consider that when doing the calculation on the service.

P.S. Is there no production meter on the existing? The utility isn't going to mind having 3 meters?
Click to expand...
So by what you're saying, having any unmonitored system on the main bus would negate the ability to use PCS on any system since it would force you to use 705.12 rules for all the systems connected to the common bus. Respectfully, I can't see that being correct although I might be misunderstanding you. 705.13 allows you use PCS in place of 705.12 for PCS controlled sources as long as you follow its rules. If not, you must refer back to 705.12 for bus protection. In fact, this in itself is a grey area they should've been more clear on but here is an article where the author states that 705.13 can effectively bypass 705.12
www.linkedin.com

Tech Guide: Designing with panel-level PCS in Powerwall Systems

Adrian Hood, May 2024 Part 1 - What is panel-level PCS? POWER CONTROL SYSTEM (PCS) technology represents one of the most significant breakthroughs for installers and homeowners in the residential solar and battery industry in recent years, overcoming one of the biggest restraints that kept solar fro
www.linkedin.com www.linkedin.com

I assume the existing system has production metering, not sure why the utility would mind.
 
hitehm said:
So by what you're saying, having any unmonitored system on the main bus would negate the ability to use PCS on any system since it would force you to use 705.12 rules for all the systems connected to the common bus. Respectfully, I can't see that being correct although I might be misunderstanding you. 705.13 allows you use PCS in place of 705.12 for PCS controlled sources as long as you follow its rules. If not, you must refer back to 705.12 for bus protection. In fact, this in itself is a grey area they should've been more clear on but here is an article where the author states that 705.13 can effectively bypass 705.12
www.linkedin.com

Tech Guide: Designing with panel-level PCS in Powerwall Systems

Adrian Hood, May 2024 Part 1 - What is panel-level PCS? POWER CONTROL SYSTEM (PCS) technology represents one of the most significant breakthroughs for installers and homeowners in the residential solar and battery industry in recent years, overcoming one of the biggest restraints that kept solar fro
www.linkedin.com www.linkedin.com

I assume the existing system has production metering, not sure why the utility would mind.
Click to expand...


I got busy yesterday so I couldn't respond timely, but let me try to stab at it.

PCS monitors the amps seen at the service or feeder to ensure that that panel never exceeds it's rating.

If you have more than one source your concern is now the direction of current flow and if any direction could exceed the panel's rating and the PCS would not catch it because it only checks the "Main breaker" side.

I wonder, then, can the 20A of uncontrolled solar + the PCS controlled solar ever exceed the panels rating with 100A of current going to the loads. I believe the answer is yes.

The PCS will not see the contribution of the uncontrolled solar. It will say, "ahhh there is only 40A of load on this board from the utility. So the PCS will say we can supply 40A of continuous or 60A noncontinuous". But it doesn't know whether or not the uncontrolled solar is offsetting an exisitng 20A already. So your current seen near the uncontrolled and PCS controlled could accidentally exceed the panel boards continuous limit.

That is just how I see it.
 
I should add. If the PCS is smart enough to monitor both existing sources to ensure it never exceeds their combination than you should be fine.

You would essentially have 3 interconnected sources at different locations in one panelboard.
 
Elect117 said:
I got busy yesterday so I couldn't respond timely, but let me try to stab at it.

PCS monitors the amps seen at the service or feeder to ensure that that panel never exceeds it's rating.

If you have more than one source your concern is now the direction of current flow and if any direction could exceed the panel's rating and the PCS would not catch it because it only checks the "Main breaker" side.

I wonder, then, can the 20A of uncontrolled solar + the PCS controlled solar ever exceed the panels rating with 100A of current going to the loads. I believe the answer is yes.

The PCS will not see the contribution of the uncontrolled solar. It will say, "ahhh there is only 40A of load on this board from the utility. So the PCS will say we can supply 40A of continuous or 60A noncontinuous". But it doesn't know whether or not the uncontrolled solar is offsetting an exisitng 20A already. So your current seen near the uncontrolled and PCS controlled could accidentally exceed the panel boards continuous limit.

That is just how I see it.
Click to expand...
To all your points, I couldn't agree more as you're arriving at the same conclusions we did. The code only states that if a bus/conductor is not monitored it must follow 705.12 for interconnected bus protection. Well just because it must revert to the 120% rule doesn't mean it's not adding current to the bus and the code doesn't subtract that current from the PCS max current at all. Therefore, as you mention, it has the potential to exceed the bus rating. As mentioned originally, our engineering team did subtract the 20A unmonitored from the PCS max setting because it seemed logical to do so but they weren't required to. The intent of my post was to get clarity on the strict rules of the code but this might be one of the most flawed codes I've dealt with.
 
hitehm said:
To all your points, I couldn't agree more as you're arriving at the same conclusions we did. The code only states that if a bus/conductor is not monitored it must follow 705.12 for interconnected bus protection. Well just because it must revert to the 120% rule doesn't mean it's not adding current to the bus and the code doesn't subtract that current from the PCS max current at all. Therefore, as you mention, it has the potential to exceed the bus rating. As mentioned originally, our engineering team did subtract the 20A unmonitored from the PCS max setting because it seemed logical to do so but they weren't required to. The intent of my post was to get clarity on the strict rules of the code but this might be one of the most flawed codes I've dealt with.
Click to expand...

It does. Your PCS install still needs to meet 705.12(B) which would put the PCS output + Other non controlled sources in the calculation.

You have 3 sources. You do not meet 705.12(B)(3)(2). You would need to chose a different method. "Where two sources, one a primary power source and the other another power source, are located at opposite ends of a busbar that contains loads,.. ". You would meet it if you put the PCS source and other other source on a new panel then connected to the service panel.

Your calculation based on the set up above would fall under:

125% of PCS rating + 125% of other source rating + Service OCPD must be less than the rating of the busbar.

or

Engineering supervision.
 
Systems utilizing PCS do not need to meet 705.12. PCS is an alternate method of qualifying a bus or conductor when the system design does not meet any of the options of 705.12. I have yet to design a system that simultaneously utilizes 705.12 and 705.13 to qualify a panel bus, so I find this post rather interesting. After reading the code again, I agree with the OP that there is no provision to account for the 705.12 system current in limiting the PCS system's current contribution, but I also can't see why it would create an unsafe situation.

Prior to the PW3 installation, the system was considered safe since it met the 120% rule. The utility could supply 100A from the top of the bus, and 20A could be supplied by the PV system, for a total of 120A on a 100A rated bus, which would not be overloaded since the currents originated from opposite ends of the bus. Of course, in order for this worst case situation to arise, the loads would have to draw 120A.

Now we add the PW3 system with its PCS capability, and let's consider how the situation changes if we set the PW3's PCS limit to 80A. Under the worst case when the loads draw 120A, the PCS system will see 100A on the utility conductors at the top of the bus, there will be 20A of unseen PV system current on the bottom of the bus, and the PCS controller will throttle its output to 0A because the utility current already exceeds the PCS setting. This is exactly the situation that was considered safe prior to the PW3 installation. In fact, it would still be the same safe situation if the PCS current setting was set to 100A.

Unless the PW3 breaker was at the bottom of the bus next to the PV system breaker. In that case, you would not want 20A of PV current plus 100A of PW3 current traveling up the bus to supply the 120A of worst case load, so limiting the PCS setting to 80A would render the design safe regardless of the location of the PW3 breaker.

But limiting the PCS current setting to 60A seems overly conservative to me.
 
Last edited:
PWDickerson said:
Systems utilizing PCS do not need to meet 705.12. PCS is an alternate method of qualifying a bus or conductor when the system design does not meet any of the options of 705.12. I have yet to design a system that simultaneously utilizes 705.12 and 705.13 to qualify a panel bus, so I find this post rather interesting. After reading the code again, I agree with the OP that there is no provision to account for the 705.12 system current in limiting the PCS system's current contribution, but I also can't see why it would create an unsafe situation.

Prior to the PW3 installation, the system was considered safe since it met the 120% rule. The utility could supply 100A from the top of the bus, and 20A could be supplied by the PV system, for a total of 120A on a 100A rated bus, which would not be overloaded since the currents originated from opposite ends of the bus. Of course, in order for this worst case situation to arise, the loads would have to draw 120A.

Now we add the PW3 system with its PCS capability, and let's consider how the situation changes if we set the PW3's PCS limit to 80A. Under the worst case when the loads draw 120A, the PCS system will see 100A on the utility conductors at the top of the bus, there will be 20A of unseen PV system current on the bottom of the bus, and the PCS controller will throttle its output to 0A because the utility current already exceeds the PCS setting. This is exactly the situation that was considered safe prior to the PW3 installation. In fact, it would still be the same safe situation if the PCS current setting was set to 100A.

Unless the PW3 breaker was at the bottom of the bus next to the PV system breaker. In that case, you would not want 20A of PV current plus 100A of PW3 current traveling up the bus to supply the 120A of worst case load, so limiting the PCS setting to 80A would render the design safe regardless of the location of the PW3 breaker.

But limiting the PCS current setting to 60A seems overly conservative to me.
Click to expand...
I think you nailed it. If (a big if) I'm understanding this correctly, it's the function of the PCS to drop the PCS source output to 0A when the load demand exceeds the combined PCS limit that is key to understanding it all. In summary, Once the loads draw more than 100A (lets say 101) with the fixed unseen source supplying it's full 20A, there's no combined sum of the PCS sources to supply 81A for the remaining load demand w/o exceeding the panel limit. You can't decrease one source w/o increasing the other so either way the sum of both will exceed the 80A PCS limit. Therefore, the PCS must basically drop to 0A and let the utility and the unseen source handle the loads, which then frees up the utility to supply a full 100A.

Interestingly, the 20A unseen source actually moved the total load demand where the PCS shuts off to 100A instead of 80A so it aided not hindered the PCS by siphoning off 20A the utility doesn't need to supply. Also interesting is as long as the panel limit is not exceeded, you can take advantage of the max PW3 output or any fraction thereof during charging, TOU strategy, Export/Import rules etc, which is basically the point of it all.
 
hitehm said:
This is based on Tesla's own doc that states (in so many words) you must consider the bus to be 80% of its rating because of the way manufacturers test their equipment.
Click to expand...
This peaked my interest, so I checked UL 67 for the testing procedure. The temperature rise test (Section 21 of the standard) is done with a specified total loading while the temperatures at various locations are monitored until they stabilise, with specific limits for temperature rise over ambient depending on the location. If the panelboard is MLO, the specified loading is 100% of the panelboard rating. If the panelboard has a main breaker, the specified loading is indeed only 80% of the main breaker rating. I assume that any convertible panel does get tested in its MLO configuration with 100% of the panelboard rating.

So Tesla's guidance does seem appropriate for any panel that is main breaker only. But if the panel is MLO or convertible, it seems to me reasonable to use a Panel Limit of 100% of the busbar rating.

As to the OP, the Powerwall3 PCS can measure other sources. So unless there is some reason it can't be done (did I miss it?), the best solution is to instrument the existing PV with CTs and set the Panel Limit according to the above. If there is an unmonitored source, it seems to me that the source's rating would indeed need to be subtracted from the Panel Limit one would use if all sources were monitored.

Cheers, Wayne
 
wwhitney said:
This peaked my interest, so I checked UL 67 for the testing procedure. The temperature rise test (Section 21 of the standard) is done with a specified total loading while the temperatures at various locations are monitored until they stabilise, with specific limits for temperature rise over ambient depending on the location. If the panelboard is MLO, the specified loading is 100% of the panelboard rating. If the panelboard has a main breaker, the specified loading is indeed only 80% of the main breaker rating. I assume that any convertible panel does get tested in its MLO configuration with 100% of the panelboard rating.
Click to expand...
I was told the exact same thing from a Siemens tech about the MLO vs Main Breaker percent testing. I agree the MLO should allow a panel limit of 100%, unless Tesla has an 80% policy on this regardless.
wwhitney said:
So Tesla's guidance does seem appropriate for any panel that is main breaker only. But if the panel is MLO or convertible, it seems to me reasonable to use a Panel Limit of 100% of the busbar rating.

As to the OP, the Powerwall3 PCS can measure other sources. So unless there is some reason it can't be done (did I miss it?), the best solution is to instrument the existing PV with CTs and set the Panel Limit according to the above. If there is an unmonitored source, it seems to me that the source's rating would indeed need to be subtracted from the Panel Limit one would use if all sources were monitored.
Click to expand...
See the post from PWD above and my summary of his take on it. At first it seems logical to subtract the unseen source if it can't be added to the monitored sources but I think he's spot on as to why it's not needed, at least for PW3 systems that drop the PCS to 0A if the panel limit is exceeded. I wonder, the code does not require this method nor does it have that level of detail so do all PCS systems work in this manner?
wwhitney said:
Cheers, Wayne
Click to expand...
 
wwhitney said:
This peaked my interest, so I checked UL 67 for the testing procedure. The temperature rise test (Section 21 of the standard) is done with a specified total loading while the temperatures at various locations are monitored until they stabilise, with specific limits for temperature rise over ambient depending on the location. If the panelboard is MLO, the specified loading is 100% of the panelboard rating. If the panelboard has a main breaker, the specified loading is indeed only 80% of the main breaker rating. I assume that any convertible panel does get tested in its MLO configuration with 100% of the panelboard rating.



Cheers, Wayne
Click to expand...
Can you provide a web where you found this? I would love to have something official discussing this. Can't find it.
 
Notice that 705.13 was changed significantly between the 2020 when it was introduced and 2023 because the wording was poor.
If the PCS monitors the main and the BESS contribution and is set to 80A max then the most the panel bus will see is 20A from the main, 80A from the BESS, and 20A from the PV, That's 120A, the bus limit under 705.12. As the main increases the PCS will decrease the BESS contribution. At 100A from the main the BESS is 0A and the PV can be 20A. Still 120A on the bus.
 
hitehm said:
At first it seems logical to subtract the unseen source if it can't be added to the monitored sources
Click to expand...
It is certainly necessary if the goal is to get behavior no worse than if all sources were monitored. But it sounds like the argument is that that PCS behavior with all sources monitored is unnecessarily conservative.

pv_n00b said:
If the PCS monitors the main and the BESS contribution and is set to 80A max then the most the panel bus will see is 20A from the main, 80A from the BESS, and 20A from the PV
Click to expand...
If the PCS is monitoring just the main and the BESS, and is set to 80A max, then with 20A on the main, it will limit the BESS to 60A.

Cheers, Wayne
 
wwhitney said:
It is certainly necessary if the goal is to get behavior no worse than if all sources were monitored. But it sounds like the argument is that that PCS behavior with all sources monitored is unnecessarily conservative.


If the PCS is monitoring just the main and the BESS, and is set to 80A max, then with 20A on the main, it will limit the BESS to 60A.

Cheers, Wayne
Click to expand...
Why would it limit to 60A if it can't measure the PV output? If all the PCS sees is 20A coming into the panel on the main and it is programmed to output up to the bus rating of 100A then 100-20=80A.
 
Last edited:
wwhitney said:
This peaked my interest, so I checked UL 67 for the testing procedure. The temperature rise test (Section 21 of the standard) is done with a specified total loading while the temperatures at various locations are monitored until they stabilise, with specific limits for temperature rise over ambient depending on the location. If the panelboard is MLO, the specified loading is 100% of the panelboard rating. If the panelboard has a main breaker, the specified loading is indeed only 80% of the main breaker rating. I assume that any convertible panel does get tested in its MLO configuration with 100% of the panelboard rating.
Click to expand...
While the temperature test in section 21.3 does load a panel with a main breaker to 80% for the test the standard does not require a panelboard to be marked to limit the load to 80% of the main breaker rating. Section 11.3.2 even provides an alternate testing protocol to the section 21 temperature test. All the OCPD in a panel are limited to the usual continuous current limits if not 100% rated. So a standard main breaker is limited to carrying 80% of its rating of continuous current. This is not a bus limit, it is a current through the breaker limit.
 
pv_n00b said:
While the temperature test in section 21.3 does load a panel with a main breaker to 80% for the test the standard does not require a panelboard to be marked to limit the load to 80% of the main breaker rating.
Click to expand...
That's correct, but it means the panelboard assembly has not been tested for temperature rise with higher currents. It is certainly possible that if you took the standard test arrangement, and added another 20% source to the busbar, along with another 20% of load, one of the temperature limits would then be exceeded. So the conservative approach apparently advocated by Tesla is to set the panel limit to 80% of the main breaker rating.

pv_n00b said:
Why would it limit to 60A if it can't measure the PV output? If all the PCS sees is 20A coming into the panel on the main and it is programmed to output up to the bus rating of 100A then 100-20=80A.
Click to expand...
I took your statement "the BESS contribution is set to 80A max" to mean an 80A panel limit, not a 100A panel limit. Also the OP is about the suitability of an 80A panel limit.

If the panel limit is 100A, then I agree with your math.

Cheers, Wayne
 
You must log in or register to reply here.
Top