Solar edge 11.4 on high leg shutting down

[tortuga]

I would bet a milkshake no inverter manufacturer supports connecting an inverter across an open phase B-C of a open delta.
I think the inverter is measuring grid impedance (%Z) of the primary voltage source thru a secondary coil impedance.
Where the open phase (B-C) is likely two unequal transformers in series I think the inverter would get confused, or need to be specially designed for that scenario. I'd like to be wrong about it though.
Since there is no standard or NEC requirement for identifying the open set, like saying A= black , B = orange, C = blue and then require orange / blue (or B-C) is always the open set (if its an open delta), it seems difficult to even tell which set is the open phase.

 

[ggunn]

Would moving all the inverters to the A and C phases be feasible? If it could be done it would likely solve the problem and it would be a simple change to make if the combiner panel has enough A and C tabs.

 

[tortuga]

Would moving all the inverters to the A and C phases be feasible? If it could be done it would likely solve the problem and it would be a simple change to make if the combiner panel has enough A and C tabs.

I'd just start by moving the B-C one to A-C like the open delta diagram in post #29, that is if the open set it known.

 

[ggunn]

I'd just start by moving the B-C one to A-C like the open delta diagram in post #29, that is if the open set it known.

My opinion is that it is rarely if ever a good idea to connect a single phase inverter to the B phase of a 240 high leg delta service.

 

[jaggedben]

I would bet a milkshake no inverter manufacturer supports connecting an inverter across an open phase B-C of a open delta.
I think the inverter is measuring grid impedance (%Z) of the primary voltage source thru a secondary coil impedance.
Where the open phase (B-C) is likely two unequal transformers in series I think the inverter would get confused, or need to be specially designed for that scenario. I'd like to be wrong about it though.
...

Why should it be any more difficult for an inverter to push current into an open phase than for a load to draw current from an open phase? (As it were, so to speak. I realize in both cases it's two transformers on the other phases, but the question stands.) Impedance alone should not in general cause any issue with voltage rise that isn't similar to a voltage drop caused by the load; either the conductors and transformer are big enough or not, in either case. Again, it's something *other than impedance* at issue. POCO guy essentially told electrofelon impedance was okay. (Post #37.) And the inverter should have no way of 'knowing' it's an open phase purely from impedance. If the stinger pot is too small that could be a problem on the AB phase and not just the BC phase, but if the stinger pot is big enough I don't generally see why it should be an issue with either.

Also, to repeat, to my understanding the only way an inverter 'measures' grid impedance is by injecting current and measuring voltage change. From electrofelon's narrative it never actually did that.

I repeat that I think (in essence) a scope would be necessary to understand the problem.

 

[ggunn]

... POCO guy essentially told electrofelon impedance was okay....

If the POCO guy didn't know what he was talking about or knew nothing at all about PV systems, it wouldn't be the first time that has happened..

 

[jaggedben]

If the POCO guy didn't know what he was talking about or knew nothing at all about PV systems, it wouldn't be the first time that has happened..

Also POCO guy's responsibility is just to make sure their network would handle it okay, not to ensure the inverter likes it enough to work properly.

My opinion is that it is rarely if ever a good idea to connect a single phase inverter to the B phase of a 240 high leg delta service.

To be clear, I agree with this, too, somewhat on theory but actually more on experience. I am also just curious to understand it more deeply. Had a very similar experience to the subject of this thread a number of years back, but never understood the reason and actually don't know if it ever got resolved.

 

[electrofelon]

My opinion is that it is rarely if ever a good idea to connect a single phase inverter to the B phase of a 240 high leg delta service.

Why? Granted this open phase thing seems to be a problem but disregarding that, what if it is a closed Delta or the open phase is known and avoided? Have you had any problems in those cases, besides poco's telling you not to do it?

 

[electrofelon]

Would moving all the inverters to the A and C phases be feasible? If it could be done it would likely solve the problem and it would be a simple change to make if the combiner panel has enough A and C tabs.

I told them just avoiding the "open jaw" would be the easiest fix. Only downside is the feeder is not large enough even if the inverters were spread across AB and AC. Would be an easy upgrade though, the conduit is large enough.

 

[electrofelon]

Also, to repeat, to my understanding the only way an inverter 'measures' grid impedance is by injecting current and measuring voltage change. From electrofelon's narrative it never actually did that.

I suppose it is possible the inverter sends out a quick pulse of current that either I didn't notice or the meter didn't even pick up. Does anyone know if a inverter sends any current at all out in that 5 minute qualification period?

I don't really think inverters care about grid impedance though, as long as the allowable voltage window is maintained. My personal PV system at my house for example has probably a 15 volt rise. From 0 to full production. I have such high impedance because I have a step up step down transformer in addition to the utility transformer. My inverters have never cared whatsoever - that is after I lowered the taps to keep the voltage within range.

 

[electrofelon]

Based on jaggedben's proposition that it is probably the inverter's anti islanding scheme not playing nicely with the BC phases, this has some good information, probably the best I could find. Although it doesn't really explain anything about what might be going on, but it could well be over my head. Can any of the engineers among us explain how slip mode frequency shifting might somehow react with the open phases to trick the inverter into thinking it's an island?

Anti-Islanding Protection with Grid-Tied PV Inverters

Anti-islanding protection is a commonly required safety feature which disables PV inverters when the grid enters an islanded condition. Anti-islanding protection is required for UL1741 / IEEE 1547. Knowledge of how this protection method works is essential for today’s PV system designers. We...

solectria.com

 

[ggunn]

Why? Granted this open phase thing seems to be a problem but disregarding that, what if it is a closed Delta or the open phase is known and avoided? Have you had any problems in those cases, besides poco's telling you not to do it?

I would think that the evidence provided in this thread would be sufficient, but yes, I have. It's been years, so I don't recall the details. If the center tapped transformer can handle the PV output, I always recommend connecting A - N - C and ignoring B. If nothing else, connecting single phase inverters three phase to a 240 delta high leg in many cases makes the interconnection unnecessarily complicated.

 

[electrofelon]

I would think that the evidence provided in this thread would be sufficient, but yes, I have. It's been years, so I don't recall the details. If the center tapped transformer can handle the PV output, I always recommend connecting A - N - C and ignoring B. If nothing else, connecting single phase inverters three phase to a 240 delta high leg in many cases makes the interconnection unnecessarily complicated.

Well I did qualify that with if it was a closed Delta or the open phase is known. I don't see what the issue would be. Note in this case the only phase combination we are having trouble with is the open one.

 

[electrofelon]

And of course YMMV, but I would say a POCO is just as likely to want the inverters balanced vs feeding everything on A-C, so not sure I would agree that A-C is the way to go.

 

[ggunn]

And of course YMMV, but I would say a POCO is just as likely to want the inverters balanced vs feeding everything on A-C, so not sure I would agree that A-C is the way to go.

That has not been my experience. Why would a POCO want the inverter output to be balanced over A, B, and C when they do not expect or want the load to be balanced? On every 240V delta high leg service I can remember dealing with, the B phase transformer, fuse, and conductors have been significantly smaller than those on the A and C phases. Everyone's experience is different, for sure, but in recent years I have always connected A - N - C with these services and I have never had any pushback from the POCO. Some POCOs will not even allow feeding the B phase. We can go round and round about what should work, but ignoring the B phase does work.

 

[electrofelon]

That has not been my experience. Why would a POCO want the inverter output to be balanced over A, B, and C when they do not expect or want the load to be balanced? On every 240V delta high leg service I can remember dealing with, the B phase transformer, fuse, and conductors have been significantly smaller than those on the A and C phases. Everyone's experience is different, for sure, but in recent years I have always connected A - N - C with these services and I have never had any pushback from the POCO. Some POCOs will not even allow feeding the B phase. We can go round and round about what should work, but ignoring the B phase does work.

Well I don't know how they do things in Texas, but this POCO wanted them balanced. Can you give them a call and tell them balancing this system in the grand scheme of things on their distribution system does not matter? Thanks!

 

[ggunn]

Well I don't know how they do things in Texas, but this POCO wanted them balanced. Can you give them a call and tell them balancing this system in the grand scheme of things on their distribution system does not matter? Thanks!

Ignoring your hyperbole for a moment, are the transformers themselves balanced? Have you asked the POCO if they would allow the system to be connected to only the A and C phases? Are you sure that whoever at the POCO told the solar contractor to balance the PV system across all three phases knew what they were talking about?

 

[tortuga]

Can any of the engineers among us explain how slip mode frequency shifting might somehow react with the open phases to trick the inverter into thinking it's an island?

I got a lot of help with open deltas on here in the past from @synchro and @gar but they have not been around in a while.

Well I don't know how they do things in Texas, but this POCO wanted them balanced.

Here is my very uneducated explanation that's hopefully so bad a real EE will come along and correct me;
Since the primary side of an open delta is a open wye, then we're pushing power back into 2 primary phases (to primary neutral), call them coils L1 and L2, so my guess is the best we could is balance 2 primary phases.
Think of the primary phase coils like kitchen sinks and your PV is water you want to pour down the sink.

Since you have two primaries there are two sinks, each sink can hold 5 gallons of water.
Each inverter is a 5-gallon bucket of water.
First inverter A-B dumps a 5 gallon bucket down sink L1,
Next inverter B-C (the open jaw) dumps 2.5 gallons of water down between each sink most goes in but but some spills
Then inverter A-C dumps 5 gallons of water in sink L2
and the cycle repeats.
There is enough time for the water to drain between each cycle.

Now if the transformers are unequal size imagine sink L1 only holds 1 gallon and sink L2 still holds 5 gallons.

 

[Birken Vogt]

Well I don't know how they do things in Texas, but this POCO wanted them balanced. Can you give them a call and tell them balancing this system in the grand scheme of things on their distribution system does not matter? Thanks!

What would happen if you just moved the troublesome inverter to a more stable phase and did not tell anyone?

It's a mistake anybody could make, wrong phasing of a single phase device.

 

[ggunn]

Ignoring your hyperbole for a moment, are the transformers themselves balanced? Have you asked the POCO if they would allow the system to be connected to only the A and C phases? Are you sure that whoever at the POCO told the solar contractor to balance the PV system across all three phases knew what they were talking about?

Moreover, did the POCO tell you personally that the PV system needs to balanced over all three phases, or is that what the solar guys are telling you that the POCO said? This is an unusual configuration for a system, it could be that the solar guys have got it wrong. I would check out the simpler solution that we know would work (connecting to the A and C phases) thoroughly before discounting it.