Load Management - Charger current exceeds max current limit on switching on

[michi-mr]

Describe the bug

When switching on the charger (changing from off to fast or by starting a plan), the charger systematically starts at high if not full power triggering the warning "over current detected". After some regulation cycles, the charger adjusts its power correctly to match the max current limit.
I cannot follow the calculation from the log... current in circuit is measured at 3.2 A (18:17:56), and with a limit of 10 A the max current set point should be 10-3.2 = 6.8 A, but lp-1 sets the limit to 13.8 A (18:17:57).
If this would be at the real limit of the circuit, this behaviour would burn the main fuse.

Steps to reproduce

1. set current limit to 10A
2. start charging (fast or plan)
3. observe first power setting and warning in the log
   ...

Configuration details

site:
  title: home
  meters:
    grid: my_grid
    pv: my_pv
  residualPower: -1000 # negativ erlaubt Netzbezug positiv erzwingt Einspeisung

interval: 10s

circuits:
- name: main
  maxCurrent: 10
  maxPower: 21000
  meter: my_grid
  parent:

loadpoints:
  - title: Chargeur Garage
    charger: Warp3
    mode: pv # nach Ausstecken
    circuit: main
    enable:
      threshold: 0 # positiv um noch bei Netzbezug zu starten
      delay: 1m
    disable:
      threshold: 0 # positiv um noch Netzbezug zu erlauben
      delay: 3m


chargers:
  - name: Warp3
    type: template
    template: tinkerforge-warp3-smart
    host: 192.168.1.40
    port: 1883
    timeout: 30s
    topic: warp3/2haD

meters:
  - name: my_grid
    type: template
    template: solaredge-inverter
    usage: grid
    modbus: tcpip
    id: 1
    host: 192.168.1.24
    port: 1502

  - name: my_pv
    type: template
    template: solaredge-inverter
    usage: pv
    modbus: tcpip
    id: 1
    host: 192.168.1.24
    port: 1502

Log details

[site ] DEBUG 2026/01/02 18:17:56 ----
[lp-1 ] DEBUG 2026/01/02 18:17:56 charge power: 0W
[circuit-main] DEBUG 2026/01/02 18:17:56 power: 855W
[circuit-main] DEBUG 2026/01/02 18:17:56 current: 3.2A
[site ] DEBUG 2026/01/02 18:17:56 grid power: 809W
[site ] DEBUG 2026/01/02 18:17:56 pv 1 power: 0W
[site ] DEBUG 2026/01/02 18:17:56 !! solar production: accumulate set my_pv 68876.248kWh meter total (was: Accumulated: 0.000kWh updated: 2026-01-02 18:17:46 +0100 CET meter: 68876.248kWh)
[site ] DEBUG 2026/01/02 18:17:56 !! solar production: accumulate moved my_pv from 0.000 to 0.000
[site ] DEBUG 2026/01/02 18:17:56 grid powers: [613 72 121]W
[site ] DEBUG 2026/01/02 18:17:57 grid currents: [3.2 1 1.3]A
[site ] DEBUG 2026/01/02 18:17:57 site power: -191W
[lp-1 ] DEBUG 2026/01/02 18:17:57 charger status: C
[lp-1 ] INFO 2026/01/02 18:17:57 start charging ->
[lp-1 ] DEBUG 2026/01/02 18:17:57 wake-up timer: stop
[circuit-main] DEBUG 2026/01/02 18:17:57 validate current: 3.2A + (7A -> 16A) > 10A capped at 13.8A
[lp-1 ] DEBUG 2026/01/02 18:17:57 set charge current limit: 13.8A
[site ] DEBUG 2026/01/02 18:18:06 ----
[lp-1 ] DEBUG 2026/01/02 18:18:06 charge power: 9522W
[circuit-main] DEBUG 2026/01/02 18:18:06 power: 10533W
[circuit-main] WARN 2026/01/02 18:18:06 over current detected: 11.9A > 10A
[site ] DEBUG 2026/01/02 18:18:06 grid power: 10536W
[site ] DEBUG 2026/01/02 18:18:06 pv 1 power: 0W
[site ] DEBUG 2026/01/02 18:18:06 !! solar production: accumulate set my_pv 68876.248kWh meter total (was: Accumulated: 0.000kWh updated: 2026-01-02 18:17:56 +0100 CET meter: 68876.248kWh)
[site ] DEBUG 2026/01/02 18:18:06 !! solar production: accumulate moved my_pv from 0.000 to 0.000
[site ] DEBUG 2026/01/02 18:18:06 grid powers: [3836 3329 3375]W
[site ] DEBUG 2026/01/02 18:18:06 grid currents: [11.9 9.6 9.7]A
[site ] DEBUG 2026/01/02 18:18:06 site power: 9536W
[lp-1 ] DEBUG 2026/01/02 18:18:06 session energy: 0.332kWh
[lp-1 ] DEBUG 2026/01/02 18:18:06 charger status: C
[circuit-main] DEBUG 2026/01/02 18:18:06 validate current: 11.9A + (13.8A -> 16A) > 10A capped at 11.9A
[lp-1 ] DEBUG 2026/01/02 18:18:06 set charge current limit: 11.9A
[site ] DEBUG 2026/01/02 18:18:16 ----
[lp-1 ] DEBUG 2026/01/02 18:18:16 charge power: 8211W
[circuit-main] DEBUG 2026/01/02 18:18:16 power: 9184W
[circuit-main] WARN 2026/01/02 18:18:16 over current detected: 14.5A > 10A
[site ] DEBUG 2026/01/02 18:18:16 grid power: 9176W
[site ] DEBUG 2026/01/02 18:18:16 pv 1 power: 0W
[site ] DEBUG 2026/01/02 18:18:16 !! solar production: accumulate set my_pv 68876.248kWh meter total (was: Accumulated: 0.000kWh updated: 2026-01-02 18:18:06 +0100 CET meter: 68876.248kWh)
[site ] DEBUG 2026/01/02 18:18:16 !! solar production: accumulate moved my_pv from 0.000 to 0.000
[site ] DEBUG 2026/01/02 18:18:16 grid powers: [3378 2871 2923]W
[site ] DEBUG 2026/01/02 18:18:16 grid currents: [14.5 12.2 12.4]A
[site ] DEBUG 2026/01/02 18:18:16 site power: 8176W
[lp-1 ] DEBUG 2026/01/02 18:18:16 session energy: 0.355kWh
[lp-1 ] DEBUG 2026/01/02 18:18:16 charger status: C
[circuit-main] DEBUG 2026/01/02 18:18:16 validate current: 14.5A + (11.9A -> 16A) > 10A capped at 7.4A
[lp-1 ] DEBUG 2026/01/02 18:18:16 set charge current limit: 7.4A
[site ] DEBUG 2026/01/02 18:18:26 ----
[lp-1 ] DEBUG 2026/01/02 18:18:26 charge power: 5106W
[circuit-main] DEBUG 2026/01/02 18:18:26 power: 5968W
[circuit-main] WARN 2026/01/02 18:18:26 over current detected: 10.1A > 10A

What type of operating system or environment does evcc run on?

Docker container

External automation

• I have made sure that no external automation like HomeAssistant or Node-RED is active or accessing any of the mentioned devices when this issue occurs.

Nightly build

• I have verified that the issue is reproducible with the latest nightly build

Version

v0.300.2 nightly.20260102-a861b11

[mfuchs1984]

You should first try a longer interval. Apart from that, there is an issue with meterless chargers and load management, see #21513. A longer log that starts earlier would tell if this is the problem in your case. See also #26270.

[Jrlohni]

I also have problems with the loadmanagement.
If two loadpoints are charging they don't use the full amps wich are usable.

Usable are 17 amps. First loadpoint (lp-1) "Hof" use 6 amps and the second loadpoint (lp-2) use 7,5 amps. That are just 13,5 amps (rounded 14 amps).
The loadmanagement calculates 13,9 A + (7,64 A -> 16A) > 17A Capped at 10,8 A
I think don't know at what point the calculation is wrong.

Don't wondering about the 17 A the meter always shows 0,2 A more then from the loadpoints used.
I also use a really short interval (2 seconds) but it always worked before update on 300.x. For me is just a 2 second interval usable because of the votaility of our energy load.
At first i thought the problems might be caused by the power limit (11500W) I commented it out, but first took screenshots of the log file. The behavior remained the same.

I use the go-e Charger (lp-2) with the meter included and a shelly meter directly after the fuse. The loadpoint "Hof" is a JuiceCharger (lp-1) without a meter. The shelly meter is the main loadmanagement meter.

[mfuchs1984]

@Jrlohni this is off topic here. Create a separate discussion with a complete debug log file (not screenshots) if you want help.

[michi-mr]

You should first try a longer interval. Apart from that, there is an issue with meterless chargers and load management, see #21513. A longer log that starts earlier would tell if this is the problem in your case. See also #26270.

Good point, thanks! I played around with the interval in the last days, but I cannot always reproduce the same behavior. The goal should be to switch on the charger, either by a plan or by clicking fast in the UI, without exceeding the max current limit reliably.

Similar as discribed in #21513, I observed at 10s and 15s interval that the car is not fast enough to start charging, leading to a wrong calculation of the initial max current. The first calculation is correct, and the second calculation assumes the charger is already running at the calculated current while it is not in reality. Thus, the second calculation gives a too high max current, and the initial current is too high during the first interval. Also at 10s, the regulation oscillates for a few cycles by max 1 A before finding the correct max current.

Interestingly, 20 s, 25 s, and 30 s intervals sometimes work, but sometimes don’t.
I think I found why:

The success depends on the moment within the interval when the charge is switched on.
If it’s switched on early in the interval, the car has enough time to react before the end of the interval; the first current is correctly calculated and the regulation is spot on, without oscillation.
If it’s switched on late in the interval --> [lp-1] charger enable, the car does not have enough time before the next interval starts. In this case --> [lp-1] start charging occurs before the car has actually started.

This then can lead to two behaviors, and I could not find out when which behavior is triggered.
Path 1: A second calculation is run, leading to a wrong result as described for the short intervals above.
Path 2: A second calculation is not run, and the current is set to the default wallbox limit of 16A.

I have attached logs of 10 s interval, and of paths 1 and 2 for 30 s interval.

Maybe a potential workaround for the wrong calculation in path 1 is to use [lp-1] charge power/3/230V, and not the previously calculated capped current? No idea for path 2...

evcc-20260103-210230-debug.10s.log.txt

evcc-20260107-234227-debug.30s-path1.log.txt

evcc-20260108-194217-debug.30s-path2.log.txt