300A breaker tripping when backfed

Status
Not open for further replies.
synchro said:
In the diagram you attached, the CT's will only measure the current going from the inverters to the grid. But it would not include the current supplied to the loads by the inverter's load output terminals, or supplied by the grid directly to the loads in bypass mode. Wouldn't you want the CT's to measure the total net current flowing to or from the grid?
Click to expand...
jaggedben said:
I don't see the need. The inverters know how much current they are outputting, and they can add or subtract what they measure at the CTs to determine what is being consumed by the loads. The diagram is how most manufacturers do it.
Click to expand...

On page 55 of the Sol-Ark-15K manual at the link below, they show the CT's placed right after the meter.
This diagram has a bypass mode like in the OP's diagram where the loads can be connected directly to the grid and not to any terminals on the inverters. And so in bypass mode, I believe the inverters would not know what the net current through the connection to the grid would be because of a lack of information about the load current. But I could be wrong.

https://www.solar-electric.com/lib/wind-sun/Sol-Ark-15K-Manual.pdf
 
synchro said:
On page 55 of the Sol-Ark-15K manual at the link below, they show the CT's placed right after the meter.
This diagram has a bypass mode like in the OP's diagram where the loads can be connected directly to the grid and not to any terminals on the inverters. And so in bypass mode, I believe the inverters would not know what the net current through the connection to the grid would be because of a lack of information about the load current. But I could be wrong.

https://www.solar-electric.com/lib/wind-sun/Sol-Ark-15K-Manual.pdf
Click to expand...

I should have said: "And so in bypass mode, with the OP's location of the CT's the inverters would not know what the net current through the connection to the grid would be because of a lack of information about the load current."
 
synchro said:
On page 55 of the Sol-Ark-15K manual at the link below, they show the CT's placed right after the meter.
This diagram has a bypass mode like in the OP's diagram where the loads can be connected directly to the grid and not to any terminals on the inverters. And so in bypass mode, I believe the inverters would not know what the net current through the connection to the grid would be because of a lack of information about the load current. But I could be wrong.

https://www.solar-electric.com/lib/wind-sun/Sol-Ark-15K-Manual.pdf
Click to expand...
Yes, they wouldn't know the actual grid export if the system was switched to bypass. But it wouldn't matter as long as they limited their output to the max allowed export of 160A. Or, more precisely, they would always think there is zero load, and the grid export would always be their output minus the actual load, never more. Something else is causing the breaker to trip.
 
jaggedben said:
They've been configured to limit their total output to 160A, or to limit exports to the grid to 160A? Different things.
Click to expand...

Yes. The inverters should not need the CT's in order to properly limit their own output current. But they would need the CT's in order to measure and limit any export current.

Perhaps what is happening is that the CT's have been placed in the wrong orientation around the conductors. As a result, the inverters would be interpreting the CT outputs as a net load current instead of an export current. And therefore the inverters would not be limiting their output current based on these CT measurements because they would be interpreted as having no net export current.
 
synchro said:
Yes. The inverters should not need the CT's in order to properly limit their own output current. But they would need the CT's in order to measure and limit any export current.

Perhaps what is happening is that the CT's have been placed in the wrong orientation around the conductors. As a result, the inverters would be interpreting the CT outputs as a net load current instead of an export current. And therefore the inverters would not be limiting their output current based on these CT measurements because they would be interpreted as having no net export current.
Click to expand...
Yeah, I have a feeling about the CT direction too. Partly because CT orientation is often screwed up in these systems. But by itself that's not a complete explanation why the breaker reportedly trips when it thinks there's a really small export.
 
One definite possibility might be that one CT is flipped around. That could lead the inverter to believe that there is close to zero export when export is actually quite high. The inverter may therefore think there is a lot of load, and increase output, which doesn't get measured, until the breaker trips. One way to check for this is to see if the monitoring ever shows negative consumption. However in an unlucky situation (load is unbalanced in just the wrong way) you might also never see that. Lower load than expected, or just weird load data in general, could be another telltale, if you know how much load you expect.
 
wwstrick said:
The main breaker at the service feed is tripping and only when backfed. It will eventually trip even with the backfed current limited to 20A
Click to expand...
jaggedben said:
Yeah, I have a feeling about the CT direction too. Partly because CT orientation is often screwed up in these systems. But by itself that's not a complete explanation why the breaker reportedly trips when it thinks there's a really small export.
Click to expand...

The CT's are only on the inverter outputs in the OP's schematic and therefore they would not be responding to any load currents. And so if both CT's are flipped, the inverters would always interpret the CT readings as a net load current no matter what the backfeed current limit is set to, because it would never see a current that looks like it's going in the export direction (since inverters only source current and not sink current). Therefore the CT's would not be limiting the output current of the inverters, but perhaps some settings could limit it when in the export mode, like the inverter is doing in the non-export mode.
 
Last edited:
synchro said:
The CT's are only on the inverter outputs in the OP's schematic and therefore they would not be responding to any load currents.
Click to expand...
Only in bypass mode. In regular operation the CTs catch the net current from the grid for both load and inverter output, as the loads are connected through the inverters.

And so if both CT's are flipped, the inverters would always interpret the CT readings as a net load current no matter what the backfeed current limit is set to, because it would never see a current going in the export direction (since inverters only source current and not sink current).
Click to expand...
I'm not 100% sure about Sol-Ark but the convention I've see everywhere else is that the CTs are installed in the normal load direction and exports are treated as negative load. In any case, if they were both backwards, the system would definitely be confused and it could keep increasing output to compensate for the increased 'load' (which is actually export). You might be right, here, but...

Therefore the CT's would not be limiting the output current of the inverters, but perhaps some settings could limit it when in the export mode like the inverter is doing in the non-export mode.
Click to expand...

Assuming the bypass is not enabled...
In a non-export mode the system would think the load was export, and reduce output to zero to avoid exporting. In a 'sell' mode, the system would think its own output is load, and increase output to the max trying to export. This seems to match the OP's description except for a couple things. One, he seemed to say the total output current (regardless of CT measurement) was limited to 160A in both cases, in which case neither of these situations should trip a 300A breaker. Second, he didn't indicate any problem with the system in non-export mode.
 
jaggedben said:
Only in bypass mode. In regular operation the CTs catch the net current from the grid for both load and inverter output, as the loads are connected through the inverters.


I'm not 100% sure about Sol-Ark but the convention I've see everywhere else is that the CTs are installed in the normal load direction and exports are treated as negative load. In any case, if they were both backwards, the system would definitely be confused and it could keep increasing output to compensate for the increased 'load' (which is actually export). You might be right, here, but...



Assuming the bypass is not enabled...
In a non-export mode the system would think the load was export, and reduce output to zero to avoid exporting. In a 'sell' mode, the system would think its own output is load, and increase output to the max trying to export. This seems to match the OP's description except for a couple things. One, he seemed to say the total output current (regardless of CT measurement) was limited to 160A in both cases, in which case neither of these situations should trip a 300A breaker. Second, he didn't indicate any problem with the system in non-export mode.
Click to expand...
There is no problem with anything in non-export mode. The CT's are catching all current where they are and I have verified that they are reading correctly when current is flowing in as well as out. The problem tripping occurs without regard to the amount of current, only if we are exporting. It has tripped when exporting 20A and it has tripped when exporting 150A. We set the export limit at 160A at the request of the electric company. The limit is maintained by the readings of the CTs.
 
wwstrick said:
There is no problem with anything in non-export mode. The CT's are catching all current where they are and I have verified that they are reading correctly when current is flowing in as well as out. The problem tripping occurs without regard to the amount of current, only if we are exporting. It has tripped when exporting 20A and it has tripped when exporting 150A. We set the export limit at 160A at the request of the electric company. The limit is maintained by the readings of the CTs.
Click to expand...

Does the tripping happen regardless of the amount of solar illumination during the day, or only when there are higher illumination levels?

Do you know approximately what the expected range of load currents would be?
 
synchro said:
Does the tripping happen regardless of the amount of solar illumination during the day, or only when there are higher illumination levels?

Do you know approximately what the expected range of load currents would be?
Click to expand...
I have not noticed any relationship with solar illumination. Loads range from 10kW to 45kW
 
wwstrick said:
There is no problem with anything in non-export mode. The CT's are catching all current where they are and I have verified that they are reading correctly when current is flowing in as well as out. The problem tripping occurs without regard to the amount of current, only if we are exporting. It has tripped when exporting 20A and it has tripped when exporting 150A. We set the export limit at 160A at the request of the electric company. The limit is maintained by the readings of the CTs.
Click to expand...

How did you verify the amps? Note that a clamp on ammeter can verify the amplitude but not the direction of current. Can you confirm the monitoring doesn't show anything weird for load currents? Both CTs are in the right direction?

All that asked, I'm currently leaning against the CT installation being the issue.

Do you know the response time of the Sol-Ark PCS? How sure are you that it is limiting export to 160A within a second or two after load changes? Does the tripping correspond to instances of high load or fluctuating load?

Your 10-45kW load represents a variance of something like 145A. Supposing 35kW of load dropped all at once, and this happens at the same time the inverters were exporting 160A, you are now export 305A. Or maybe more on one line? If the Sol-Ark response to this isn't faster than the breaker trip time then that could be causing the trips, given you apparently have 375A of total inverter output available.

I also wonder if the breaker trip setting is correct, as someone asked above. A lower actual breaker trip point would exacerbate what I just described.

Is it possible to limit the total output of all inverters to less than 300A (or whatever the breaker trip setting really is) regardless of load and export? That would ensure that the inverters never output enough current to trip the breaker at 300A, even if all 45kW of max load drops at once. If the issue persists after that, I think the breaker is somehow to blame, although I can't explain how if it doesn't have GFP and is otherwise suitable for backfeed.
 
wwstrick said:
The main breaker is an ABB XT4N 300A. I have not yet seen it, but it does not have GFP or RCD. The attached diagram is for an installation with 3 inverters. This one is just like it, but with 6 inverters.
Click to expand...
Where in this drawing is the ABB main breaker that is tripping?
it would be really helpful to have the actual drawing that reflects your system, rather than this sample dwg.
 
Last edited:
jaggedben said:
"at the spot below the meter labeled '200A Knife Disconnect, Fused'" Confirmed in post #29.
Click to expand...
There must be a loadcenter that the breaker is in, right? what else is connected there and how is it connected in relation to the CT's and the feed to the inverter combiner? Just hard to visualize.
 
synchro said:
Yes. The inverters should not need the CT's in order to properly limit their own output current. But they would need the CT's in order to measure and limit any export current.

Perhaps what is happening is that the CT's have been placed in the wrong orientation around the conductors. As a result, the inverters would be interpreting the CT outputs as a net load current instead of an export current. And therefore the inverters would not be limiting their output current based on these CT measurements because they would be interpreted as having no net export current.
Click to expand...
The CTs are in correctly and SolArk runs a routine where they check the CT installation by dropping solar production and then increasing it. The inverter will then make any needed adjustment to the calculations based on the CT readings. So even if they are on the wrong cables or in the wrong orientation that would be taken into account.
 
solarken said:
Where in this drawing is the ABB main breaker that is tripping?
it would be really helpful to have the actual drawing that reflects your system, rather than this sample dwg.
Click to expand...
The truth is that our installation is identical with the only real changes being a breaker instead of a fused disconnect and 6 inverters instead of three..
 
Status
Not open for further replies.
Top