hems script negative prices

[BelhsanHmida]

Purpose

Evaluate how HEMS behaves when electricity prices become negative, and specifically whether PV curtailment occurs in this scenario.

Test

I ran the HEMS script with explicitly negative electricity prices for a limited time window:

2030-01-15T11:00:00+01:00,-0.282600
2030-01-15T12:00:00+01:00,-0.244270
2030-01-15T13:00:00+01:00,-0.230340

results:

Screencast.from.2026-02-06.22-42-37.webm

[BelhsanHmida]

@Flix6x , what so you think we see total curtailment, expect for a peak from 12:00 -> 12:15

[Flix6x]

Nice. We definitely see planned curtailment. That 15-min peak is unexpected. I'm not sure what might have caused that. We don't have different prices for consumption and production in this script, right? It would be also be an interesting case to have negative production prices and positive consumption prices during the same period. Then you'd expect the scheduler to maximize self-consumption, instead of maximizing consumption.

But first, could you simply repeat the experiment with simulate_live_corrections=True? I'd like to know what it does to that mystery peak.

Please post the full dashboards, too, including the aggregate power. From those it is sometimes easier to see the scheduler's intentions.