unwanted passthru causing AC-IN consumption despite battery full

Question

question

jeanmarie asked May 13, '24

unwanted passthru causing AC-IN consumption despite battery full

Dear Victron Forum, I hope some user or Victron staff will be able to help me in understanding (and maybe fixing it) the reason why my ESS system is going from time to time into passthru causing AC-IN consumption.

My system :

MPII 10kVA + CerboGX

RS450/100 + Fronius Primo 3kW each with (roughly) 5kWc and 3kWc respectively

6 Seplos batteries

the problem can be found in the graph below, it summarize well the issue : sytem is going time to time into passthru, without any reason (that I can understand to be able to correct it).

The graph is during the night with, of course no DC neither AC production but battery almost full (SOC around 94%) and with a stable consumption (EV charger). As you can see the system switch alone the consumption from batteries to grid and back (3 times during that night). This occurs also during the day, some time, most often when the AC OUT load is high (40-60% of the capacity of the MP2).

I noticed that it correspond to an "Overload L1" alarm but without being able to distinguish if this alarm is the cause or the consequence.

For those focusing on the not validated batteries, if working in inverter only mode, I experience absolutely no disturbance of any kind whatever the consumption or battery level is, so I don't see how a BMS communication issue could cause this in ESS mode and nothing in inverter only mode.

And yes I have read the passthru chapter and I have no grid meter, no error of any kind reported by my Cerbo.

Thank you in advance for any help on this topic

ESS

capture-du-2024-05-13-09-51-14.png (140.1 KiB)

2

2 |3000

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

dognose commented · May 13 at 09:15 AM

Check the Graph for the BMS Charge Voltage Limit (CVL) and Charge Current Limit (CCL) - it seems like the battery is temporary setting this values to "0" for reasons. Maybe internal temperature critical?

At least that would tell you, if the batteries are causing the limitation, or if the multiplus are doing it.

1 ·

Alexandra ♦ dognose commented · May 13 at 10:05 AM

Unfortunately with the Seplos being unsupported officially weigh in from Victron cannot be expected.

However it is likely that the battery is the right track here. It seems to be trying asking for a charge.

Do you have DVCC enabled?

Do you have a min soc set in ess and then is peak shaving on or off at that level.

Can i suggest a graph with one scale -5000 to positive 5000.

System - battery watts

Inverter power in

Inverter power out

Solar if you want but unnecessary as it does not seem to happen with solar.

0 ·

Modifications Space

Node Red Space

Answer 1 · 2024-05-13T12:34:54Z

jeanmarie answered · May 13 at 12:34 PM

Hi @dognose & @Alexandra , thank you for replying so fast to my SOS and your suggestions and questions

Regarding CVL/CCL and internal temperature, batteries and inverter are in a well ventilated place, batteries temperature are around 18-20°C and neither CVL nor CCL can be seen going to 0, at least on VRM portal, and neither by looking at CerboGX interface battery's parameters

for DVCC : yes it is activated and for ESS, min soc, it is set to 20% and peak shaving is set to always (off position doesn't exist for this parameter, or?) as shown in the screen dump

and for the suggested graph, here below is an attempt of my understanding

capture-du-2024-05-13-13-30-10.png (58.2 KiB)

capture-du-2024-05-13-13-28-47.png (157.6 KiB)

2

2 |3000

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

Alexandra ♦ commented · May 13 at 05:59 PM

@jeanmarie

Ok. One adjustment to the graph - set it to one scale. It makes it easier to see how things are working in all the watts.

It is the (DCL) Discharge Current Limit that is important for load support. Usually passthrough indicates the battery is saying no to the discharge. Or there is loss of communication with the battery.

Peak shaving to always means under 20% it will aid big load start up.

The overload warnings could be related to lower voltage from grid maybe? Inverters derate for that reason as well.

0 ·

jeanmarie Alexandra ♦ commented · May 14 at 08:43 AM

Hi Alexandra, thank you very much for your help, I appreciate

I changed scales as advised in the screen dump below and added DCL and a view on current & Voltage

Grid voltage is not really stable but is quite high, never got any low voltage to explain this Overload L1 alarm, so I don't know where this alarm could come from, neither do I know if this is a cause or a consequence, is there a way to diagnostic this more precisely?

As said for weak BMS communication, which is a possible cause, I have difficulties in understanding how this could have absolutely no consequence neither alarm of any kind nor even shutdown of the multiplus while I'm on battery. Beside that, when the issue occurs, it takes a part only from the grid and the rest from batteries.

Is there an accessible way to log the BMS communication? or to better diagnostic this "Overload L1" alarm?

0 ·

capture-du-2024-05-14-08-18-03.png (190.8 KiB)

Answer 2 · 2024-05-14T09:33:34Z

jeanmarie answered · May 14 at 09:33 AM

Just one point that pop up to me : I already read that, sometimes the current sensor that may be connect to a socket on the MP had problems even if nothing is plugged in.

Could this be a reason for my problems?

2

2 |3000

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

dognose commented · May 14 at 02:21 PM

Here is an interesting comment from Guy Stewart about Overload Warnings:
https://community.victronenergy.com/questions/38954/understanding-overload-warnings-on-multiplus-ii.html

And he actually describes how and why using a external meter (which you don't) could eliminate those warnings.

In a nutshell: Without meter, the inverter will try to flatten out all (unlogged) spikes as well, while with a meter he will act "slower" and therefore don't run into warnings due to millisecond-spikes.

0 ·

jeanmarie dognose commented · May 15 at 09:13 AM

Do you think installing a meter could solve my issue?

0 ·

Answer 3 · 2024-05-14T13:49:52Z

jeanmarie answered · May 14 at 01:49 PM

One more information, maybe Victron would be interested in helping

today I got a new Overload L1 alarm despite a consumption of less than 2000W at the time of the error.

As I was in inverter only mode, of course nothing else happened regarding AC-IN consumption or any other item

So where could this Overload L1 alarm come from ?

capture-du-2024-05-14-14-47-30.png (148.9 KiB)

2 |3000

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

question