Live Ground Shocked 5 Year Old

[Hedges]

I made the mistake of reading this, and now I don’t know which of you is correct.

Breaker Voltage

www.electriciantalk.com

I think the comments laid it out pretty clearly. But commenters aren't authoritative, just as we forum members aren't.

"I've heard them called "slash rated". The lower rating is phase to ground voltage, the higher is phase to phase. As Bird dog said, the issue comes up in wanting to use the high leg on a 3P4W delta for a single phase 208v load; the breaker has to have a phase to ground rating of at least 208v, which a slash rated breaker does not."

"120/240V Slash rated breakers are suitable for 1Φ or 3Φ loadcenters, where the highest voltage to ground is 120V and the max voltage phase to phase is 240V. Also used in 120/208V 3Φ 4W Wye systems.



240V Straight rated breakers are suitable for 1Φ or 3Φ loadcenters, where the highest voltage to ground is 240V and the max voltage phase to phase is 240V. Such as in a Center tapped delta system, where the wild leg is 208V to ground.


Most 3P breakers come standard with a straight 240V rating.
2P units can come either way, normally slash rated, but straight rating is required if your connecting to a 208V wild leg.


BR240H is an example of a 2P unit with a straight rating, for use on a center tapped delta system."

 

[timselectric]

Which would you use for corner-ground 240V delta?

I would have no issue using either one. They are both rated at or above any voltage they could see.

Breakers have no idea if the short is coming from another leg, neutral, or ground.

 

[Mike 134]

Been awhile since I've done industrial work with Delta services but do remember if you have a high leg a plain old 120/240 breaker won't work for the high leg. I let Mr. google help me remember. So not all multi-pole breakers are rated for 240volts.

Which QO breakers can you use on a high leg delta system? | Schneider Electric USA

 

[Cmiller]

Power distribution blocks.

https://www.galco.com/1343350-52798.html?utm_source=adwords&utm_medium=&utm_campaign=[EM] PMax | Shopping | Z | Zombie | In Stock - G&utm_term=&utm_content=&utm_id=19750239001&utm_device=m&utm_network=x&utm_mt=&phone=&source=AdWords&gad_source=1&gclid=Cj0KCQjwn7mwBhCiARIsAGoxjaJ700CJt4dxnGsn-h-MVA6aOHADd2-_6PPGq9HL1IEcV56zZr6X4DkaAj6lEALw_wcB

What about when you have more than 2 inverters to parallel? My preference is to be able to shut down the output of each individual inverter (regardless of 1,2 ,3..... 12 units) for testing and troubleshooting purposes. Combining with power distribution blocks and no breakers eliminates that ability.

Do you have any thoughts on combining say, 3x Sol-Ark 15k? (Rated 62.5A @ 120/240V with solar, for a total rated output of ~187.5A)

If you don't have any output breakers and one inverter fails, you need to shut the whole system down in order to do any work on a unit. You can't temporarily run on 2 units (settings would need to be re-configured of course) while the 3rd is getting figured out, as that 3rd units output wires would be hot.

Also, you mentioned a risk of overloading the bus of the panel... I honestly think that in most cases it would be nearly impossible to actually overload the bus of the panel, if it is a 200A panel. I do realize that this statement has a million variables that play into it, but you also mentioned the inverter having a limited output, so looking at it that way, you would need to have 200A x 240V = 48kW of inverter to get to 200A, let alone surpass it to the point of overloading it.

 

[timselectric]

What about when you have more than 2 inverters to parallel? My preference is to be able to shut down the output of each individual inverter (regardless of 1,2 ,3..... 12 units) for testing and troubleshooting purposes. Combining with power distribution blocks and no breakers eliminates that ability.

I will have six combined with power distribution blocks.
Each does have its own breaker, just for a maintenance disconnect.
Located in an enclosure beneath each unit.

Do you have any thoughts on combining say, 3x Sol-Ark 15k? (Rated 62.5A @ 120/240V with solar, for a total rated output of ~187.5A)

Same answer.
They have built in breakers, I believe.

Also, you mentioned a risk of overloading the bus of the panel... I honestly think that in most cases it would be nearly impossible to actually overload the bus of the panel, if it is a 200A panel. I do realize that this statement has a million variables that play into it, but you also mentioned the inverter having a limited output, so looking at it that way, you would need to have 200A x 240V = 48kW of inverter to get to 200A, let alone surpass it to the point of overloading it.

It would depend on the size of the inverters, plus pass through if applicable.
Two Sol-ark "15k" units could easily overload a panel.

The first time I heard anyone mention an "AC combiner panel". Was Signature Solar, with the Growatt SPF-5000-ES.


Here's a professional installation of 3 "15k" units.
You can skip to 10:55 to see the power distribution blocks in the trough.

 

[Cmiller]

If 25A 2-pole breaker feeds 240V L1/L2 of auto-transformer like I show in this thread, it allows 50A 120V 6000W load between N and L1 (or N and L2).
My measurement showed approximately 3.5A in each of L1 & L2, 7A in N, confirming this.

The Growatt transformer in question spec's only 3000W imbalance.

So I disagree 25A on L1/L2 provides protection.

With 240V feeding the L1 and L2 of the autotransformer, the Neutral will only see the imbalance between L1 and L2...

In order for the neutral to see the rated 3000w, there needs to be 3000w imbalance between L1/L2. That imbalanced neutral current will be drawn from the HIGHEST leg, as that additional current needs to flow through the neutral to "balance out" the total. Therefore if L1 = 25A and L2 = 0A, then your neutral will be at full rated current of 25A.

@Hedges You have been saying the 25A on L1 plus 25A on L2 = 50A. This is incorrect because while L1 is +120V to N, L2 is -120V to N. Remember, this is SPLIT PHASE that we are talking about! 1 high plus 1 low does not equal 2! Instead 1 high plus 1 low equals zero! (or maybe we should say (+25A) + (-25A) = 0A)

In summary, a 25A 2 pole breaker (the key here is that we are discussing SPLIT PHASE!! 180 degrees phase angle) will never allow more than a 25A imbalance, as any current in the second leg will zero neutral current from the first one!


Take a look at your statements concerning multi-wire branch circuits. If you were indeed correct that a 25A 2 pole breaker could result in 25A + 25A = 50A total neutral current, then a multi-wire branch circuit using 12/3 romex and a 20A 2 pole breaker would be a recipe for disaster!

Again, we are discussing 120/240V North American split phase. Where we have 2 legs that are at a 180 degree phase angle to each other. Therefore the neutral only carries the differences between the 2 legs. And therefore also, we only have 120V to ground on either one of these 2 legs! Much safer than EU 240V single phase, if you ask me. But also much more complicated to understand, unfortunately....

Multi-wire branch circuit, so long as neutral doesn't get open-circuit (something I experimented with as a child), the neutral always guarantees each non-ganged single pole breaker only sees 120V max. A multi-wire branch circuit is just 2 or 3 120V circuits.

When loads are balanced, zero current flows in neutral. 2x 1800W 15A 120V space heaters, one on L1, one on L2.
If fed by ganged breaker, when it opens there is 240V across two poles in series.
If fed by individual breakers, when one pole opens it interrupts current to one space heater, 15A 120V. Voltage never rises above 120V because neutral wire picks up the 15A of space heater on other leg.

I can install two single-pole breakers and add a handle tie, then feed a 240V baseboard heater. If I trip one pole, it pulls handle of other pole and opens it too. Everything OK. If I remove the handle tie and trip one pole, that one pole has to interrupt 240V, which I say it is not rated for.

We could set up an experiment with single pole QO15 in a box, L1 from grid to L1 of the box, L2 of grid to N of the box (because I don't have 240V single phase handy.) I would do this through a 70A QO270 from my breaker panel. Likely the QO15 would successfully interrupt 15A 240V, which could be 2x 1800W space heaters in series (something I have), because current isn't high.

If we overloaded it to say 200A at 240V (I don't have any suitable loads), then at such high current and 2x rated voltage I suspect QO15 would burn. After a short while at 3x rated current, QO270 will end the fireworks.

 

[Hedges]

I would have no issue using either one. They are both rated at or above any voltage they could see.

Breakers have no idea if the short is coming from another leg, neutral, or ground.

I disagree. QO230 is only rated up to 120V. QO230H is rated up to 240V.
"120/240" apparently refers to 120/240V split phase, NOT either 120V or 240V.

Been awhile since I've done industrial work with Delta services but do remember if you have a high leg a plain old 120/240 breaker won't work for the high leg. I let Mr. google help me remember. So not all multi-pole breakers are rated for 240volts.

Which QO breakers can you use on a high leg delta system? | Schneider Electric USA

Thanks, I had read that before but couldn't seem to find it with Schneider's search.
Google to the rescue.

"

Which QO breakers can you use on a high leg delta system?​

Issue:
QO breaker for a high leg delta system

Product Line:
Circuit Breakers

Resolution:
A high leg delta system is 240 V phase to phase, 120 V to ground on the center tapped phase (normally designated A to C) and 208 V to ground from the high leg (B phase).
Therefore, you can use a standard breaker with a 120/240 V rating, such as QO220, only on phases A-C and not on B, since the maximum voltage to ground for this breaker is 120 V.
A breaker connected to B phase (the high leg) requires a breaker with a straight 240 V rating such as QO220H. Standard 3 pole breakers such as QO320 or QO320VH type breakers may also be used.

Released for:Schneider Electric USA​

Published on:6/12/2019Last Modified on:6/13/2022
"

There you go, Tim. Couldn't be any clearer, and straight from the horse's mouth.

 

[Cmiller]

I will have six combined with power distribution blocks.
Each does have its own breaker, just for a maintenance disconnect.
Located in an enclosure beneath each unit.

I personally prefer having all "units'" breakers side-by-side with labels (e.g. inv1, in2, etc.) Keeps it way more cleaned up on bigger systems! For instance, I really like the Schneider PDP and how the individual inverter breakers are all right there next to each other!

Same answer.
They have built in breakers, I believe.

Yes, 200A breakers.... for an inverter that has a rated output of 62.5A. In reality, if you don't have loads that are higher than what the inverter is rated (again, 62.5A) then feeding 125A passthrough is enough to both charge and feed loads....

So for one, the 200A passthrough and output breaker size is really quite bigger than what it probably should be.... And for two, when you have multiple inverters it is soon quite a bit cheaper to run #2 AWG copper, rather than 2/0 copper...... and then those 200A breakers aren't protecting the wire, so you need other breakers....

It would depend on the size of the inverters, plus pass through if applicable.
Two Sol-ark "15k" units could easily overload a panel.

2 Sol-Ark 15Ks would be rated for a combined total continuous current of 125A. A 200A panel is rated for..... well...... 200A....

The first time I heard anyone mention an "AC combiner panel". Was Signature Solar, with the Growatt SPF-5000-ES.

Ever heard of Schneider XW inverters...??? They have been around for quite a number of years now. Their PDP is basically a combiner panel.

Here's a professional installation of 3 "15k" units.
You can skip to 10:55 to see the power distribution blocks in the trough.

I'll take a look at it. However, as I mentioned, the inverters are really only rated to output 62.5A each. So personally I see no benefit in running more than 125A passthrough, which brings us back to the whole 200A output breaker dilemma again.

 

[Hedges]

With 240V feeding the L1 and L2 of the autotransformer, the Neutral will only see the imbalance between L1 and L2...

In order for the neutral to see the rated 3000w, there needs to be 3000w imbalance between L1/L2. That imbalanced neutral current will be drawn from the HIGHEST leg, as that additional current needs to flow through the neutral to "balance out" the total. Therefore if L1 = 25A and L2 = 0A, then your neutral will be at full rated current of 25A.

@Hedges

I disagree. Not possible to have 25A flow in L1 and 25A flows out of center-tap N, while 0A flows through L2. Inductance doesn't allow it.

I challenge you:
Create a spice deck of two coupled inductors and try this.
Or, connect a transformer as auto-transformer and measure this.

A clamp ammeter applied to one leg at a time is sufficient. It won't reveal phase relationship, just magnitude, but the results will be distinct and clear (Only if load is many times the no-load current of the transformer.)

An oscilloscope can show relative phase, if current probes used on all three wires.

I did. The results match what I am saying:

Auto-transformer L1/N/L2 current

Transformers aren't well understood by most people who use them. It is well known that centertap "Neutral" connection carries the difference between L1 and L2, but that is not at all true for autotransformers. It is only true for the case of an isolation transformer where all power is delivered...

diysolarforum.com

Neutral carries the sum of L1 and L2. at any given instant, current flows IN L1 and L2, OUT N. Then AC reverses, current flows OUT L1 and L2, IN N.

 

[Cmiller]

I disagree. Not possible to have 25A flow in L1 and 25A flows out of center-tap N, while 0A flows through L2. Inductance doesn't allow it.

I challenge you:

Challenge accepted.

Create a spice deck of two coupled inductors and try this.
Or, connect a transformer as auto-transformer and measure this.

I'll use a transformer connected as auto-transformer.

A clamp ammeter applied to one leg at a time is sufficient. It won't reveal phase relationship, just magnitude, but the results will be distinct and clear (Only if load is many times the no-load current of the transformer.)

An oscilloscope can show relative phase, if current probes used on all three wires.

I did. The results match what I am saying:

Auto-transformer L1/N/L2 current

Transformers aren't well understood by most people who use them. It is well known that centertap "Neutral" connection carries the difference between L1 and L2, but that is not at all true for autotransformers. It is only true for the case of an isolation transformer where all power is delivered...

diysolarforum.com

Neutral carries the sum of L1 and L2. at any given instant, current flows IN L1 and L2, OUT N. Then AC reverses, current flows OUT L1 and L2, IN N.

I will need to do this tomorrow, as I'm out of time today. I'll report back with results and pictures of clamp meter(s).

I will also make a diagram of exactly how I wire it up, to make sure we are all clear on the "flow of power". I am beginning to think that maybe some of us aren't thinking of the exact same thing as the rest of us.

(By the way, just to be clear, I am not in any way trying to be cocky here! It'll truly be interesting to see exactly what happens!)

 

[timselectric]

Split-phase is single phase.
L1 and L2 are not two phases, or out of phase.
Current flows from L1 to N to L2, then reverses to flow from L2, to N, to L1. (Alternating)
If the same amount of loads are on L1 and L2, they operate in series at 240v. And nothing flows on the neutral.
If the loads are not equal, the current above the balance is carried back on the neutral.

 

[Hedges]

Challenge accepted.

...

(By the way, just to be clear, I am not in any way trying to be cocky here! It'll truly be interesting to see exactly what happens!)

Agreed. Just like Tim and I are respectfully discussing QO breaker voltage rating.

I do make mistakes, and can be confident of myself until convinced otherwise.

Regarding breakers, I'm going by interpretation of manufacturer documentation.

Regarding transformers, I've been thinking deeply about them for some time, also performing simulations (including extracting measured parameters for lossy saturable hysteretic models) and configuring bench test measurements.

I've gone quite far down the rabbit hole at work, finding the differences between different types of cores for chokes, measuring leaked magnetic fields, etc.

Most things, breakers and magnetic cores included, have only partial documentation. Few breakers actually quote what inrush they can ignore in excess of 5x rated current. I've measured transformer and SMPS inrush. Magnetic cores need at least three parameters to describe BH curves, and I don't find all of those in documentation. Have to make test jig to measure.

 

[timselectric]

2 Sol-Ark 15Ks would be rated for a combined total continuous current of 125A. A 200A panel is rated for..... well...... 200A....

If using the pass through, it can be as much as 200a.
It just depends on what you feed it from.
You could limit the input with a reduced feed.
And this could keep you from overloading the panel.
Every situation is different. And anything can be made to work.
I'm not saying that you can't make it work. I'm saying that I wouldn't recommend doing it that way. And passing inspection with it has been difficult for people.
Because there are no provisions in the NEC for it. So, most inspectors don't accept it. If you are not getting it inspected, then you can of course do whatever you want. It can be done safely.

 

[timselectric]

Ever heard of Schneider XW inverters...??? They have been around for quite a number of years now. Their PDP is basically a combiner panel.

Specifically designed for a purpose.
Not a standard loads panel, pulled from a shelf.

 

[Hedges]

Split-phase is single phase.
L1 and L2 are not two phases, or out of phase.
Current flows from L1 to N to L2, then reverses to flow from L2, to N, to L1. (Alternating)
If the same amount of loads are on L1 and L2, they operate in series at 240v. And nothing flows on the neutral.
If the loads are not equal, the current above the balance is carried back on the neutral.

Did you post this regarding breakers in a split-phase panel, possibly MWBC? Or regarding auto-transformer?

In an auto-transformer, only low (magnetization?) current flows from L1 to L2. Whatever inductance allows.

With a 120V load between N and one of the two Ln, current flows from L1 to N and into the load, and simultaneously flows from L2 to N (swimming upstream) and into the load. When polarity of AC voltage reverses, current flows from load and then from N to L1, and same/similar amount of current flows from load and then from N to L2.

Kirchhoff's current LAW holds true. Two coupled windings and the center tap N join at a node. Sum of currents into that node is always zero. If 25A flows out of N, then 12.5A flows into L1 and 12.5A flows into L2.

I've demonstrated this on the bench. Cmiller is going to replicate or refute my results. You previously refuted them, but I need to see your no-load current (center tap N disconnected from load) to see if you just had a very non-ideal transformer and too light a load.

I think we will all agree about Kirchhoff's current law. Exact values of L1, L2, N current will vary. The fundamental behavior of coupled windings is what I have been arguing about.

These minutiae (or fundamental operating principles) about auto-transformers belong in my thread on the topic, not this thread about shock hazard from hot chassis.

 

[timselectric]

Not possible to have 25A flow in L1 and 25A flows out of center-tap N, while 0A flows through L2. Inductance doesn't allow it.

We are in total agreement on this.

 

[ULR]

I disagree. You put 12.5A in L1 of an auto-transformer. That flows and supplies 12.5A to center tap N of transformer and load.
Magnetic coupling sucks another 12.5A in L2 of auto-transformer (note I said "in", not "out", at any given instant it is flowing in opposite direction not same direction as current in L1), and supplies a second 12.5A for 25A in N.

The key here is to think in terms of current which is not "in phase" as you like to say about split-phase systems, but rather current at 0 degrees phase or 180 degrees phase. Same direction or opposite direction.

There is never 25A flowing in L1 and out L2. It is one big inductor, can't get more than a fraction of an amp to flow at 60 Hz. The only way to have 25A flow in L1 is to have 25A flow in L2. They are flowing in opposite directions. When the meet in the middle, they both flow out N for 50A. And that is 2x the transformer's rating.

12.5A I say it is.

This is NOT an isolation transformer, where current flowing in secondary allows current to flow in L1, out L2.

Well... The conclusions are right but I'd phrase it differently.

Confusion arises because the two lines are at opposite phase, so if power IN on L1 is current IN, power OUT on L2 is ALSO current IN. Total current on L1, L2, and N has to be zero, so the current contributions from load-balancing power being transferred between L1 and L2 are twice on N what they are on each of L1 and L2.

The other current contributions come from losses, reactive magnetization currents, less than perfect balance between the two halves of the autotransformer winding, and less that perfect "mutual inductance" coupling between the two halves. These currents only flow between L1 and L2, not through N. But these currents are a drop in the bucket compared to those from correcting the maximum rated load imbalance. So let's ignore them for now. Without them the size of the L1 and L2 currents would be precisely equal and half the size of the N current.

But an imbalance of, say, 3000 watts means Lx is being loaded with 3000 watts more than Ly. To correct it the transformer needs to take half that, 1500 watts, from Ly and give it to Lx, letting Lx provide the other 1500. So the size of the current in L1 and L2 corresponds to the 1500 / 120 = 12.5A, in N it's 25A.

Worst case imbalance could be generated by a single 120V load pulling >25A and thus, with the autotransformer, putting a 240V load of >12.5A (plus a tad for transformer losses, etc.) on the inverter. So a dual-gang 13A breaker on the feed from the combiner breaker box to the 120V split-phase breaker box should protect the autotransformer - at the cost of limiting the total load on the 120V box to about 3000 watts (equivalent of about a single 25A 120V circuit).

It would be nice to have a remote trip breaker driven by a current sensor around the autotransformer's N wire and set to trip at 25A there. That would let you run all the loads the inverters would power and still drop power to the 120 and 240-with-neutral loads if the imbalance became excessive. But in a few minutes searching I didn't find anything ready-to-go.

I note that the only place a greater-than-L1-or-L2 current appears in the autotransformer is from the center-tap tie between the two coils out through the N terminal. Everything else carries either L1's or L2's current. Since the two coils seem to be joined AT the N terminal, and N's current is essentially JUST proportional to the imbalance, which is rated, I don't see any need for special thinking about protecting N's current specifically, rather than just meeting the imbalance specification.

I also note that this is basically exactly a transformer. Like a motor winding, it has some thermal mass and takes a bit of time to warm up. So motor-rated breakers that allow a half to a few seconds of moderate overload should be fine.

 

[timselectric]

Did you post this regarding breakers in a split-phase panel, possibly MWBC? Or regarding auto-transformer?

Just split-phase in general. Because people think that it's two phases 180 degrees out of phase with each other.

In an auto-transformer, only low (magnetization?) current flows from L1 to L2. Whatever inductance allows.

Absolutely

With a 120V load between N and one of the two Ln, current flows from L1 to N and into the load, and simultaneously flows from L2 to N (swimming upstream) and into the load. When polarity of AC voltage reverses, current flows from load and then from N to L1, and same/similar amount of current flows from load and then from N to L2.

Kirchhoff's current LAW holds true. Two coupled windings and the center tap N join at a node. Sum of currents into that node is always zero. If 25A flows out of N, then 12.5A flows into L1 and 12.5A flows into L2.

I've demonstrated this on the bench. Cmiller is going to replicate or refute my results. You previously refuted them, but I need to see your no-load current (center tap N disconnected from load) to see if you just had a very non-ideal transformer and too light a load.

I think we will all agree about Kirchhoff's current law. Exact values of L1, L2, N current will vary. The fundamental behavior of coupled windings is what I have been arguing about.

This is my understanding.
240v Input current flows between L1 and L2.
120v Output current flows between either L1 or L2 and N.
I will describe one side (L1) and just say that the other is happening in reverse at the same instance.

On the first half of a cycle.
Current flows in L1 and out N.
Half of the load current is physically flowing through the L1 coil, and the other half is induced from the L2 coil.


On the second half of the cycle. (Reverse flow)
Current flows in N and out L1.
Half of the load current is physically flowing through the L1 coil, and the other half is induced from the L2 coil.

When the loads are balanced, an autotransformer is just idle.
No current flows in or out of L1, L2, or N.
Other than a small amount of saturation current.

We agree that the 120v loads current inside the autotransformer is divided in half on the 240v feeding the autotransformer.

Were we disagree on the feed breaker size, is on how the actual loads will be presented to the breaker.

I say that it depends on how the autotransformer is connected.
In line with the feeder, I say protect the feeder. (As it will be carrying the most current)
Or "T" tapped at the loads, I say protect the neutral. (As it will be carrying the most current)
You say protect at the neutrals rating in both cases.

 

[ricardocello]

Like a motor winding, it has some thermal mass and takes a bit of time to warm up. So motor-rated breakers that allow a half to a few seconds of moderate overload should be fine.

This is why the Victron Autotransformer has a solenoid trip for its main breaker based on toroid temperature, not Neutral current.
Short excessive unbalances over 30A do not trip it. They claim 28A continuous depending on ambient temp.

 

[Cmiller]

If using the pass through, it can be as much as 200a.

Correct. The issue lies in a power outage scenario. I don't believe that it is best practice to have more passthrough loads than what an inverter can handle during a grid outage. You run the risk then, of having a complete power failure when grid is down, due to inverter overload. Kinda defeats the whole point of a "hybrid" inverter with batteries.

Just because it's possible doesn't mean its practical, advisable, or best practice.

We do, by the way, from time to time run 200A passthrough on the 15ks. In fact today we did a 15k "whole house backup" setup. We also "locked out" the aux heat coils (during grid outage) on the furnace to avoid overloading the 15k!

It just depends on what you feed it from.
You could limit the input with a reduced feed.
And this could keep you from overloading the panel.

You are right that it depends on what you feed it from. I would put money on it, that the average double or triple 15k install does NOT have more than a 200A electric service to start with. So then you may as well only run 100A through each inverter... and in the downstream panel you won't want more than 200A anyway, unless you are splitting to 2 panels.

Every situation is different. And anything can be made to work.
I'm not saying that you can't make it work. I'm saying that I wouldn't recommend doing it that way. And passing inspection with it has been difficult for people.
Because there are no provisions in the NEC for it. So, most inspectors don't accept it. If you are not getting it inspected, then you can of course do whatever you want. It can be done safely.

If I had an issue with an inspector, I would bring up the NEC articles on how to size breakers for PV inverters. (Rated output x 1.25 and round up to nearest breaker size.) Sorry I don't know where it's found offhand. Don't have my code book in front of me. That would make the relevant breaker size for a 15k an 80A. (62.5 x 1.25 (78A) == 80A breaker)

Shouldn't make a difference if you are feeding "forward" or "backward" for lack of better terminology.

I have run into where the engineers said we need to install (2) 15Ks because the total loads were ~22kW. Since we felt confident that the loads wouldn't exceed one 15Ks capabilities (they overrated load draw for a number of items, plus not everything is on at the same time, ever...) we got them to combine the 15K rating with the standby gen rating (14kW) to get the number we needed. (Gen assist via grid peak shave....) I should note, this is a complete off grid install.