question

joshuagoshert avatar image
joshuagoshert asked

Quattro Overload Alarm

I have two Quattros, in parallel, which have been in place since around 2018'ish. In the alarm log there are random Overload Alarm logs, maybe 2 or 3 each year, which seem like fluke incidents. But over the past week I have had two such errors. And the not-cool part of this overload is that the whole system freezes up, despite if the grid is still up, or if the PV panels are producing, or if the batteries are up and running. I get that it's a preventative and protective measure, but I'd really like to prevent any overloads. Any ideas what might cause this and how to fix it? Both times was in the middle of the day, with strong sunshine. Fronius and MPPT were working fine. The batteries were at 90-95% SOC, and the grid was up. After restarting the Quattros, I got the standard lights, the Primary's LED show "Feed-in with PV power - bulk" and the Slave's LED shows "Slave," and everything was back to normal operations.

From what I've searched for in the forum, I see two main possibilities, firmware that isn't up to date, or wiring issues. Before I try updating the firmware (which will require an intimidating complete re-setup), I was hoping someone might have some good suggestions for what else to check on. Nothing has changed much with wiring, is that really a possibility given that it works 99.999% of the time?

layout.jpg

MultiPlus Quattro Inverter Charger
layout.jpg (76.0 KiB)
6 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

knownot avatar image knownot commented ·
More experienced folks will likely chime in but I suspect they'll have more questions first - like are the alarms on both devices, battery and solar size. I might wonder is it possible that a few automatic loads (freezers on defrost cycle, heat pump water heater, etc etc) could occasionally line up and equal an overload?
0 Likes 0 ·
joshuagoshert avatar image joshuagoshert knownot commented ·
Thanks for the questions, hopefully my answers will help.


The two Quattros are in parallel mode, the error is on AC Out Line 1, which connects to both units. Nothing in the system indicates that one unit is at fault, but it does appear that resetting the slave unit can reset everything (not sure that matters much).


The batteries can hold 200 Ah, and the solar generation is a total of 10 kWh (half to a Fronius Inverter connected to AC Out Line 1, and half to an MPPT connected to batteries).


With 6 buildings supplied by the system it is very possible that multiple devices pulled heavy loads at the same time. There are no heaters or air conditioners currently on the system. The most pull comes from electric kettles for tea and coffee. There are 12 refrigerators and 1 chest freezer, all rather inefficient.


This error only happened 3 times in all of 2021, and then twice in this past week.
0 Likes 0 ·
kevgermany avatar image kevgermany ♦♦ commented ·

This may help

https://youtu.be/4SXtGIx0x5w

But what's running when it happens? And have you recently added new loads?

0 Likes 0 ·
joshuagoshert avatar image joshuagoshert kevgermany ♦♦ commented ·

Thanks for sharing the video, very informative. Our system does not attempt a reset 30 seconds after an overload stop. Is that only for the Multis and not for the Quattros?

I can't pinpoint what was running when the overload happened. I can only tell the system was operating at normal levels. During the first overload the solar was generating about 6 kWh and the loads were averaging 3 kWh, with occasional spikes to 5 and 6 kWh. At about 10 AM the grid dependency had dropped to 0 kWh, and the batteries reached 100% SOC. During the second overload, at 9 AM, the solar was only generating 4 kWh, and the loads totaled about 7 kWh on average with spikes up to nearly 10 kWh. The grid dependency covered the gaps that the solar was unable to produce for.

Interestingly, the battery voltage and current seem a bit wild at those times. During the first overload the battery voltage was up to 55.91, and during the second overload, the battery current was down to -48.9 A. But the error is an overload on L1, and doesn't seem to be anything related to charging/discharging the batteries.

We have not added new loads.

0 Likes 0 ·
kevgermany avatar image kevgermany ♦♦ joshuagoshert commented ·
Which Quattros are installed?

Hopefully someone with more knowledge will chime in




0 Likes 0 ·
Show more comments
7 Answers
Paul B avatar image
Paul B answered ·

some suggestions that may help

whats type of batteries do you have as if they are lead acid then they are properly collapsing and can no longer deliver the required voltage.

so test your batteries


if lithium is the bms turning the batteries off due to high cell voltage


changing the firmware is unlikley to help but its worth a try. you can use the vrm to update the firmware and thus not have to do a full reinstall

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

joshuagoshert avatar image joshuagoshert commented ·

Paul, thanks for your suggestions and questions.

img-3739.jpg

The batteries are lithium iron phosphate. 4 x 48V 50ah in parallell, so 48V 200ah going into the BMS.

Readings are 53.71V, 53.79V, 53.70V, and 53.72V.

img-3740.jpgimg-3741.jpgimg-3742.jpgimg-3743.jpg


Where would I see if the BMS has been turning off the batteries? From the graphs?

And I agree that in general, updating firmware usually only fixes specific issues targeted by that specific update, and I have no idea what the issue would be since I can't find what the updates were targeting. I just keep finding two suggestions, update firmware and/or check all wiring (which hasn't changed in many months). I do have VRM updates on, but the Quattros are in parallel, and apparently I cannot update the firmware through the system.

firmware-update-page.png

0 Likes 0 ·
img-3739.jpg (95.7 KiB)
img-3741.jpg (127.3 KiB)
img-3742.jpg (89.1 KiB)
img-3743.jpg (86.8 KiB)
firmware-update-page.png (40.8 KiB)
joshuagoshert avatar image joshuagoshert commented ·

Here are some screen grabs of the battery before, during and after the last overload event (alarm occurred at 9:19).

9-18.png

9-19.png

9-20.png


0 Likes 0 ·
9-18.png (158.3 KiB)
9-19.png (161.9 KiB)
9-20.png (154.2 KiB)
joshuagoshert avatar image
joshuagoshert answered ·

Are there some classic troubleshooting steps I can do to find what is going wrong? If the overload is something that takes 30 minutes to occur, is there a telltale sign I could look for beforehand?

And one other thing, anyone have any ideas why the system won’t try to reboot after 30 seconds?

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

joshuagoshert avatar image
joshuagoshert answered ·

I just had another overload warning that cleared within a minute, but it is early enough that more items can be reduced from the complexity of the problem; so I'm hoping that sharing this specific warning will help in identifying the issue. I was beginning to suspect the rapid drop in Battery Amps as somehow connected to the issue, but it continued down for the next minute or two as the system cleared the warning. Does anyone have any ideas what might be causing this, or what steps I need to take to further identify the source of these overloads?

To help explain the graphs...

1) Grid was up and being relied on for the most part because not much solar generation was happening yet.

2) Consumption was averaging around 6 kW, with spikes not exceeding 7.5 kW.

3) Genset is not in use.

4) Fronius not yet online.

5) MPPT is up to 1200 W of generation.

6) Fronius not yet online.

7) Battery Watts drop sharply as system begins using batteries to cover AC Load.

8) Battery Volts and Amps drop sharply with the use of the batteries.

9) Battery SOC is min 90% if on Grid, the minute before it reached 93.8% and switched to discharging to help with AC Load.

overload-warning.jpg


overload-warning.jpg (66.1 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Paul B avatar image
Paul B answered ·

You need to contact your battery manufacturer and or ghet the BMS software for you batteries and then monitor the cell voltages.

OR in your GX unit under the device list drill down into the BMS/battery list and check what the highest or lowest cell is. if you can do this then on the vrm Under Advanced then select the widjets and then the batterery/BMS device then one of the graphs also records hi and low voltages. for each cell as well.

This can only happen if the batteries come up as a device on the gx unit

1651451861798.png

1651451806205.png


1651451806205.png (28.7 KiB)
1651451861798.png (29.5 KiB)
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

joshuagoshert avatar image joshuagoshert commented ·

What does a significant drop in voltage (like on my graph) indicate?

And no, my BMS does not show high or low voltages for individual batteries, just the total.


0 Likes 0 ·
80380d63-7fbe-4228-b156-e80de5a757e0.png (295.3 KiB)
Paul B avatar image
Paul B answered ·

It indicates total battery voltage, the BMS systems turn the battery on and or off depending on cell voltage and you have 16 cells I am presuming. so if one cell goes hi (ie depending on BMS setting ) over 3.65 then the BMS turns that battery off, or it tells the Inverter charger to stop charging and or depending on connections and assistant setup it can tell the inverter to turn off - or reduce its max allowed to draw current to say 20 amps then you can get a overload.

I have been suggesting to you to analyze your battery and check what its doing at the cell level. you need to do this so you can rule out other issues.


1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

joshuagoshert avatar image joshuagoshert commented ·
Ok, great explanation of the potential error. I will see what I can do with the BMS. It is not a Victron device, perhaps it doesn’t communicate any more information. I will test with my meter too. Would I see a cell problem by testing each full battery? Or do I need to figure out how to test each cell?
0 Likes 0 ·
joshuagoshert avatar image
joshuagoshert answered ·

I think I’ve stumbled upon the problem. While testing a power point to see if was hot, I was seeing readings over 260V. I even had 267V flash once, and this was with a true RMS meter. Moments later the Overload L1 alarm went off again. I started reviewing the backlog of Grid L1 input voltages, and every time the overload warning occurred there were voltage spikes around the 255V range, almost every time the overload alarm occurred there were spikes near or exceeding 265V. Apparently our provider keeps it hot in the mid mornings.

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

joshuagoshert avatar image joshuagoshert commented ·
Would a voltage regulator on the line in be a good move at this point?
0 Likes 0 ·
Chris avatar image
Chris answered ·

Whens an overload not an overload... oh yeah.. when its a voltage spike.
I get (I suspect) similar issues when running on a generator - Overload alarms.... that arent actually overloads. I suspect more like frequency / waveform issues. The data resolution isnt fine enough to pick up the nuances of the events to see whats actually going on. Gotta love Victron programming :/ Its a shame there isnt a function to have very high logging frequencies when there are alarm events - Like a rolling 30s at maybe 1s intervals (or less) - that can be constantly overwritten, so you capture whats happening just before and during an event. IF it occurs. Otherwise the regular logging resolution persists for the 'normal' history (Suspect the NVM might not like this long term though)

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nickdb avatar image nickdb ♦♦ commented ·

If you install grafana you can get much, much higher resolutions, generally though, still tough to catch a transient on a chart.

Many things can trigger misleading alarms. They recently updated the battery page to actually detail a lot more about these warnings that actually are caused by the BMS, though not so relevant for you.

0 Likes 0 ·