Sol-Ark 12K shutting down on grid failure

[afblock]

Greetings all.

I recently installed my own grid-connected system using 20x 390W Helene panels, Ironridge racking, and a Sol-Ark 12K inverter. The battery is a BigBattery Kong Elite Plus. System drawings are attached below (though they don't show the battery, which is of course connected to the Sol-Ark's battery posts).

Since it went online a few months ago, I have never gotten the Sol-Ark to a state where it can reliably transition power during a grid failure (since my house was under construction I had many opportunities to test this). Regardless of the settings, at least half of the grid failures result in either an F56 DC Volt Low or F20 DC Overcurrent error, at which time the inverter shuts down for a minute or two before successfully switching over to the battery.

I have been around and around on this with Sol-Ark, who while suggesting various settings changes have not been able to explain why any particular change would fix the issue (and they didn't, though they did affect the frequency of the various error types). I'm now wondering if I got a lemon of a unit. Sol-Ark support has not replied to any messages for 10 days now.

I would very much appreciate any feedback from forum members who may have had similar issues, found solutions, or gotten redress from Sol-Ark for faulty units.

More details below of the various settings changes and their results.

Thanks!

/afb

Sol-Ark 12K support dialogue timeline

1. December 2022: Power up inverter and connect BigBattery using their recommend settings found in the spec sheet. Inverter always shutting down on grid failure.
2. March 30, 2023: File initial request for aid
3. April 4: After some back and forth, support notes that "You had the Batt Empty value greater than the Batt Shutdown and the Batt Low value, when the Batt Empty value should be less than the other values", and modifies the following values remotely:
   • Batt Empty: 43V (43%)
   • Batt shutdown: 45V (20%)
   • Batt low: 47V (35%)
4. April 17: Support remotely upgrades the firmware in the inverter
5. April 17: I test shutting off the grid input breaker and find I am still getting the F56 DC Volt Low errors
6. April 19: Support recommends the following settings changes (but requests that I confirm with BigBattery):
   • Battery Capacity: 332Ah, recommend changing to 300Ah
   • Batt Empty: 43V, recommend changing to 45V
   • Batt Shutdown: 45V, recommend changing to 46V
   • Batt Low: 47V, recommend changing to 47V
   • Batt Restart: 52V, recommend changing to 49V
7. April 23: I make the recommended settings changes. Still getting cutover failures, but now they are F20 (DC Overcurrent) rather than F56. I report this to support.
8. April 25: Support recommends the following settings changes:
   • Float: 58.8V changed to 55.8V
   • Absorption: 58.8 changed to 56V
   • Equalization: 58.8 changed to 56V
9. May 1: I make the recommended settings changes and am back to getting F56 errors on grid failure. I report this to support. No response.
10. May 4: I reach out again to support. No response.
11. May 10: I reach out again to support. No response.

 

[tve]

You are effectively connecting the Sol-Ark's grid input/output with its load output? Do you see the issues as well if you disconnect your load panel F from the 175A sub-panel?

 

[dcg9381]

Your total battery capacity is 332Ah and you have 48V batteries, correct?
That puts your battery capacity at: 16KW.

I looked at your battery spec, it's rated for Continuous Discharge Current of 150A.

That's 150A * 48V = 7200 watts absolute maximum (continuous).

How is this wired to the inverter? What gauge wire?

I'm "guessing" that what you've got going is a start up surge that isn't being handled by that battery.

Have any measurements of the kind of power your house is pulling?

I have "lesser" inverters capable of 13k, but similar battery capacity - around 15KW, 3 batteries. Hitting them hard, I've seen peak amps at over 200A DC. That's at around 9-10KW of output. My batteries are rated for 300A continuous (I'd never drive them past 80%).

 

[RCinFLA]

Both errors, F56 DC Volt Low or F20 DC Overcurrent error, are indicating an inverter DC input voltage slump. When inverter input DC voltage slumps down under load it draws more DC input current to keep the AC output regulated.

Either you have too much voltage drop between battery and inverter due to wiring and/or too small of battery and BMS to handle the large DC current.

 

[Ampster]

Either you have too much voltage drop between battery and inverter due to wiring and/or too small of battery and BMS to handle the large DC current.

The OP mentioned Big Battery and based on feedback I have seen on that company I would focus on the battery. I do not view it as a SolArk issue since SolArk is just reporting what is going on with the battery. The data has to be there in either the SolArk or the BMS to measure the voltage sag under load. If not available from either of those, a reliable VOM should be used.

 

[robby]

The OP mentioned Big Battery and based on feedback I have seen on that company I would focus on the battery. I do not view it as a SolArk issue since SolArk is just reporting what is going on with the battery. The data has to be there in either the SolArk or the BMS to measure the voltage sag under load. If not available from either of those, a reliable VOM should be used.

If I could have put 10 Thumbs up next to this answer I would have!
I am assuming you have proper sized wiring?
If so, from what I have read here in the past, Big Battery would be my first suspect!
Chances are that you have some bad cells in the pack that are just dropping in voltage the minute a sudden load is put on them.

 

[dcg9381]

I think we're all in agreement here..

 

[robby]

I think we're all in agreement here..

Yes but I should also add that the cells might not be bad but one or more of the packs may have cells that are completely out of balance. The Op needs to check the cell voltages when they under load.

 

[Nobodybusiness]

Greetings all.

I recently installed my own grid-connected system using 20x 390W Helene panels, Ironridge racking, and a Sol-Ark 12K inverter. The battery is a BigBattery Kong Elite Plus. System drawings are attached below (though they don't show the battery, which is of course connected to the Sol-Ark's battery posts).

Since it went online a few months ago, I have never gotten the Sol-Ark to a state where it can reliably transition power during a grid failure (since my house was under construction I had many opportunities to test this). Regardless of the settings, at least half of the grid failures result in either an F56 DC Volt Low or F20 DC Overcurrent error, at which time the inverter shuts down for a minute or two before successfully switching over to the battery.

I have been around and around on this with Sol-Ark, who while suggesting various settings changes have not been able to explain why any particular change would fix the issue (and they didn't, though they did affect the frequency of the various error types). I'm now wondering if I got a lemon of a unit. Sol-Ark support has not replied to any messages for 10 days now.

I would very much appreciate any feedback from forum members who may have had similar issues, found solutions, or gotten redress from Sol-Ark for faulty units.

More details below of the various settings changes and their results.

Thanks!

/afb

Sol-Ark 12K support dialogue timeline

1. December 2022: Power up inverter and connect BigBattery using their recommend settings found in the spec sheet. Inverter always shutting down on grid failure.
2. March 30, 2023: File initial request for aid
3. April 4: After some back and forth, support notes that "You had the Batt Empty value greater than the Batt Shutdown and the Batt Low value, when the Batt Empty value should be less than the other values", and modifies the following values remotely:
   • Batt Empty: 43V (43%)
   • Batt shutdown: 45V (20%)
   • Batt low: 47V (35%)
4. April 17: Support remotely upgrades the firmware in the inverter
5. April 17: I test shutting off the grid input breaker and find I am still getting the F56 DC Volt Low errors
6. April 19: Support recommends the following settings changes (but requests that I confirm with BigBattery):
   • Battery Capacity: 332Ah, recommend changing to 300Ah
   • Batt Empty: 43V, recommend changing to 45V
   • Batt Shutdown: 45V, recommend changing to 46V
   • Batt Low: 47V, recommend changing to 47V
   • Batt Restart: 52V, recommend changing to 49V
7. April 23: I make the recommended settings changes. Still getting cutover failures, but now they are F20 (DC Overcurrent) rather than F56. I report this to support.
8. April 25: Support recommends the following settings changes:
   • Float: 58.8V changed to 55.8V
   • Absorption: 58.8 changed to 56V
   • Equalization: 58.8 changed to 56V
9. May 1: I make the recommended settings changes and am back to getting F56 errors on grid failure. I report this to support. No response.
10. May 4: I reach out again to support. No response.
11. May 10: I reach out again to support. No response.

View attachment 148780

Sounds to me like you got a big load pulling batteries down below cutoff.

What gauge wire do you have connected to the inverters from the batteries?
What size is the battery?

 

[afblock]

You are effectively connecting the Sol-Ark's grid input/output with its load output? Do you see the issues as well if you disconnect your load panel F from the 175A sub-panel?

Because of the connection between subpanels E and F? There is an interlock on the breakers in subpanel E that allows it to power the inverter or subpanel F (bypassing the Sol-Ark) but not both at once.

 

[afblock]

Your total battery capacity is 332Ah and you have 48V batteries, correct?
That puts your battery capacity at: 16KW.

I looked at your battery spec, it's rated for Continuous Discharge Current of 150A.

That's 150A * 48V = 7200 watts absolute maximum (continuous).

How is this wired to the inverter? What gauge wire?

I'm "guessing" that what you've got going is a start up surge that isn't being handled by that battery.

Have any measurements of the kind of power your house is pulling?

I have "lesser" inverters capable of 13k, but similar battery capacity - around 15KW, 3 batteries. Hitting them hard, I've seen peak amps at over 200A DC. That's at around 9-10KW of output. My batteries are rated for 300A continuous (I'd never drive them past 80%).

The wiring between battery and inverter is via the cable supplied by BigBattery, which they rate at 150A. It appears to be welding cable.

I have not put a clamp meter on the battery cable when the grid power is cut off, but that is a good idea. Will do a test tomorrow and report back.

But note that when I was initially testing, the house loads on subpanel F were only 1 – 2kW.

 

[afblock]

The OP mentioned Big Battery and based on feedback I have seen on that company I would focus on the battery. I do not view it as a SolArk issue since SolArk is just reporting what is going on with the battery. The data has to be there in either the SolArk or the BMS to measure the voltage sag under load. If not available from either of those, a reliable VOM should be used.

As I mentioned above, I was switching with relatively small loads on subpanel F — 1 to 2 kW. There was no visible voltage drop shown on the BigBattery display at the time, but its possible a transient voltage happened too quickly for the display to update. I will try a test tomorrow of both voltage and current draw at the time of cutover.

Thanks all for the feedback so far!

 

[Ike Isaac]

Good morning all

I having a problem with Megarivo inverter, it sometimes showing the low voltage and low frequency, and it is also showing the overload, but after I restart the system, it works, but keeps on doing several times,

It is also showing the BMS COM FAILL, yet the battery is full

What might be the problem ?

Thank you

 

[dcg9381]

The wiring between battery and inverter is via the cable supplied by BigBattery, which they rate at 150A. It appears to be welding cable.

Ok, we'll assume they have that right. At 150A, I went to parallel cables, but welding copper is rated to 90C usually. We'll assume this isn't the issue as this came with the batteries.

I have not put a clamp meter on the battery cable when the grid power is cut off, but that is a good idea. Will do a test tomorrow and report back.

I'm (now) not concerned about amps as your cable set came with the battery. Your inverter is indicating that your batteries are dropping below an expected voltage (likely due to surge). It'd make more sense (to me) to log voltage at battery output / inverter input. You need something that will log fast.. Maybe a scope.

There's an easier way to make an "educated guess". Simulate a grid failure. Vary your panel loads. Bet it works with less load. Bet you see more failures with higher loads.

But note that when I was initially testing, the house loads on subpanel F were only 1 – 2kW.

I have an Emporia Vue just after my power meter. It's normal for my home to be in the 1-2kW range. It can easily spike to 6+ kW, with this months high being 15kW and all time high of 17kW. I can't get a screen shot tonight, but the jumps are huge. And they likely do not register the inductive start up loads of HVAC.

If you watch some of the tests on the EG-4 inverters, you'll see that they often fail for loads that are well within inverter range on single 5kWh batteries capable of 100A. Those inverters are in the range of 6.5kW. Testing showed that to run something like an air compressor (not a big one) 2 batteries were needed.. 2 of those batteries are rated at 200A. I think you have a "surge" problem. EG-4 is likely inferior, but the concepts are all the same.

 

[afblock]

Ok, we'll assume they have that right. At 150A, I went to parallel cables, but welding copper is rated to 90C usually. We'll assume this isn't the issue as this came with the batteries.


I'm (now) not concerned about amps as your cable set came with the battery. Your inverter is indicating that your batteries are dropping below an expected voltage (likely due to surge). It'd make more sense (to me) to log voltage at battery output / inverter input. You need something that will log fast.. Maybe a scope.

There's an easier way to make an "educated guess". Simulate a grid failure. Vary your panel loads. Bet it works with less load. Bet you see more failures with higher loads.


I have an Emporia Vue just after my power meter. It's normal for my home to be in the 1-2kW range. It can easily spike to 6+ kW, with this months high being 15kW and all time high of 17kW. I can't get a screen shot tonight, but the jumps are huge. And they likely do not register the inductive start up loads of HVAC.

If you watch some of the tests on the EG-4 inverters, you'll see that they often fail for loads that are well within inverter range on single 5kWh batteries capable of 100A. Those inverters are in the range of 6.5kW. Testing showed that to run something like an air compressor (not a big one) 2 batteries were needed.. 2 of those batteries are rated at 200A. I think you have a "surge" problem. EG-4 is likely inferior, but the concepts are all the same.

I will try and capture the voltage with a meter that can track peaks.

But note that there are no HVAC loads on subpanel F. In fact, at the time of initial testing the only connected loads were lighting, electronics, and small refrigerators.

Also I still don’t have a good sense of why changing inverter settings changed the mix of under voltage and over current faults.

 

[Quattrohead]

It sure does sound like a battery or battery cable issue, you can read pretty much full voltage but still have a higher resistance connection so your voltage will sag under a decent load.

 

[afblock]

Okay, so I just made a change (on guidance from Sol-Ark support) which seemed to "fix" the problem. They recommended bringing the charge voltages back to 58.8V per the BigBattery specs, and then reducing the max discharge rate from 150A (which is the BigBattery rated current) to 100A. After I did that, I tested grid failure four times with no issue — not even a flicker.

I am not clear, however, on why this cleared up the situation. At the time I did the tests I was running 2.4 kW of house loads, with the largest fan load being 200W (rated; it was probably actually drawing a lot less). Everything else was LEDs and electronics.

Of course the other issue is that reducing my discharge capacity to 100A means I am only able to supply 5.5 kW from the battery, which is not sufficient to cover all of the F subpanel (my data shows peaks draws of 7.5 kW in the evenings when the resistive floor heaters are on. That would interfere with my plans to reduce grid loads to ~zero during the 3pm to midnight peak billing period.

I'm continuing the dialog with Sol-Ark to see what we can conclude from this result.

 

[dcg9381]

I am not clear, however, on why this cleared up the situation.

Educated guesses:

• Bringing up your battery voltage drops your amp draw by the same ratio. IE, you bring voltage up 10%, you need 10% less amps.
• I suspect that your battery is struggling to provide 150A... I'd bet it's getting "surged" higher, which is causing a voltage drop at your inverter and the inverter faults...

Really need some instrumentation here. Did you clamp on and measure peak DC amps?

I margin "safety factor" on my batteries, try not to drive them beyond 80% of max continuous discharge.... It's a design thing.

 

[Ampster]

I am not clear, however, on why this cleared up the situation.

My guess is the voltage sag on your battery was less at 100 Amps versus 150 Amps. You either need a bigger battery or more efficient heat. Until then, cutting your peak consumption by 2/3rds is a start. Are there loads you can shift? Can you preheat before 3PM? Are those heaters in the floor or at the floor level?

 

[afblock]

Sorry for being dense but I'm still very confused:

1. If my loads are ~2 kW when the grid fails, that's only about 40A out of the battery (confirmed by the inverter and the battery display). Why would switching over to the inverter cause a spike near 150A? Any motors should not experience startup current spikes in this situation, right?
2. Why would limiting the current to 100A fix anything? If there are in fact motor loads spiking large currents over 100A, wouldn't capping the total current in fact cause those loads to shut down altogether?

I did put an inrush clamp meter on the battery and it only recorded 3A which obviously doesn't make sense. The same Fluke meter accurately measured steady state loads (or at least matched the battery display).

As for the resistive heating loads, those can be managed. I'm just trying to understand the inverter/battery interrelationship at the moment.

Thanks for the continued help.