in the last weeks I saw a strange behavior with my system: suddenly 11kW feed into the grid for a couple of seconds, then normal operation again… it happened once yesterday and once today for example (spikes can’t be seen in the graphs on VRM, i do record my own data and there it shows)
for reference, here is a screenshot, times were 2025-12-05 04:07:45 and 2025-12-06 13:51:00
Currently my SOC is basically at 15% most of the time, not going above 50%. And the strange thing is not some spikes that last „a second“ but spikes that last up to 30 seconds. So it’s not some kind of inertia or reaction time in the system.
Just now I witnessed another spike, no big loads before or after, just a 11kW feed into the grid for about 30-45 seconds. Can anyone from Victron chime in or take a look at my system? or are there better ways to contact them?
seeing you are german? Check your Grid-Input-Frequency. The German Grid-Code foresees that when Frequency drops to 49.8 Hz, any ESS is supposed to feedin with upto 100% nominal power to stabilize the grid.
That may indicate a grid issue in your neighborhood. (or eventually just a faulting grid-meter)
Also Construction Sites nearby with some intense startup-current-equipment may pull frequency down, triggering your System to “Power-Assist” the grid.
E VDE-AR-N 4105:2024-10, Abschnitt 5.7.4. „Wirkleistungsverhalten bei Über- und Unterfrequenz“
So I just had another one of those spikes and i was saving the historical data of the grid frequency (if I’m correct about the right one, (seems to be Modbus adress 2644 called “Grid AC Frequency”), and it didn’t drop, not going below 49.94 Hz in the last hour… so this seems to be a dead end.
Is this also your minimum soc? Because at the Minimum Soc, there will be a mechanic ensuring batteries dont discharge deeply.
That may match 20,30 second spikes as well, because a tiny soc drop is enough to trigger it amd the bms then quickly reports a value where it stops again.
If you could give me your site id (numeric one in the url) I could also look if it is eventually related to the DynamicESS Schedule. (Send you a pm)
just had a look at the DESS-Schedule. There are no anomalies, these spikes appear independent. Even in the middle of just “self-Consumption”, where DESS is basically doing nothing at all.
Seeing you are on v3.70~61. Try upgrading to v3.70~69 and see if the problem remains. It may be coincidence, but a colleague here had several issues with his ~61 upto a complete unresponsive and bricked Erkano. Maybe that particular version really has a hickup.
From the graphs in VRM I couldn’t find anything else looking suspicious, but as you already noted, VRMs reporting interval makes that impossible for short spikes.
I updated to ~74 a couple of days ago, and just had another one of those spikes. since then, i also updated to Multiplus firmware from v506 to v558 (to hopefully lessen the fan noise).
So the update has not helped, since then I also changed the house fuses to 35A and now I’m running the system with 32A limit, which might have changed the spike to go up to 16kW into the grid.
Whats interesting: while the Grid output should max out to about 22kW (32Ax230Vx3), it only goes up to 17kW, maybe because the House is using 3kW at the moment? Battery output matches the 22kW Grid limit pretty close:
then we may be able to narrow down if these spikes are caused by the inverters internally, or by “issued” values from the gx / venus side. These values are for example used by DESS to control the inverters. I didn’t see any indication for DESS causing it, but our remote logging here is ofc. also limited to the reporting intervals precision.
How you could track them best depends on what possibilities you have to watch dbus / mqtt.
I have seen DESS spikes when changing between charge, idle, discharge etc, due to the order of setting ForceCharge (async) before the dvcc charge power (sync). Probably depends how busy the gx device is at the time.
