I have observed several times that the FEMS Dynamic Electricity Tariff app does not seem to execute the charging power calculated in its own schedule, resulting in the actual SoC falling behind the scheduled SoC.
I documented one occurrence today (28 Sep 2026) in some detail.
System
-
FENECON battery: 22.4 kWh
-
Maximum grid power for battery charging configured in the app: 12 kW (previously 15 kW)
-
Much higher charging power is technically possible: when charging is forced manually via the emergency reserve setting, I observe up to approx. 10 kW battery charging power.
-
FEMS version: 2026.9.3
Schedule
The FEMS schedule predicted approximately:
-
13:00: 24% SoC
-
13:15: 27% SoC
-
13:30: 34% SoC
For the 13:15–13:30 CHARGE_GRID interval, the schedule specified approximately 6.7 kW charging power.
Actual behaviour
At 13:00 the actual SoC was 23%, i.e. almost exactly on schedule.
At 13:16, during the scheduled CHARGE_GRID interval:
-
SoC: 26% (already below schedule)
-
PV production: 2.9 kW
-
House consumption: 0.8 kW
-
Grid import: 2.1 kW
-
Actual battery charging power: 4.2 kW
-
Scheduled battery charging power: ~6.7 kW
The energy balance is consistent:
2.9 kW PV + 2.1 kW grid − 0.8 kW consumption = 4.2 kW battery charging.
Therefore, this does not appear to be a display issue. FEMS actually requested only about 2.1 kW from the grid, resulting in ~4.2 kW total battery charging instead of the scheduled ~6.7 kW.
At 13:30:
The resulting ~3 percentage-point SoC deficit corresponds to approximately 0.67 kWh for a 22.4 kWh battery.
Interestingly, the discrepancy can almost completely be explained by the lower charging power during the 15-minute interval:
(6.7 − 4.2) kW × 0.25 h = 0.625 kWh
There was also more PV production than FEMS had forecast, not less. This should have made reaching the scheduled SoC easier and should, if anything, have reduced the required grid import while maintaining the scheduled battery charging/SoC trajectory.
I have seen substantially larger discrepancies on other days (several kWh, in one case >10 kWh), but did not have screenshots of the schedule at the time.
My question:
Is it expected behaviour that the actual CHARGE_GRID power can be substantially lower than the charging power calculated in the schedule, even when the actual SoC is already below the scheduled SoC?
If so, what determines the actual charging-power request, and why does the controller not increase charging power or extend the charging interval to catch up with the scheduled SoC?
I have screenshots of the schedule/cockpit and the corresponding 15-minute FEMS export if these would be useful.