Predictive charging¶
Predictive charging is an optional feature that charges batteries from the grid when the expected energy balance for today is negative.
Decision logic¶
If (Usable battery + Solar forecast) < Expected consumption:
Charge from grid the exact deficit
Else:
Do not charge (cost saving)
- Usable battery: energy currently stored above the configured min SOC.
- Solar forecast: estimated production for today (Solcast/Forecast.Solar sensor).
- Expected consumption: 7-day rolling average. See Daily consumption estimate.
Charge target¶
When charging is triggered, the integration does not charge all the way to max_soc from the grid. Instead it calculates a grid-only target SOC — enough to cover only what solar will not be able to provide during the day:
solar_surplus = max(0, solar_forecast − estimated_consumption)
grid_charge = max(0, gap_to_max − solar_surplus)
target_soc = current_soc + grid_charge / capacity × 100
gap_to_max is the kWh distance from the current SOC to max_soc. Solar output in excess of household demand charges the battery the rest of the way during the day.
Example: the battery needs 5 kWh to reach max_soc. Solar forecast is 13 kWh, expected consumption is 10 kWh — a surplus of 3 kWh available for the battery. The integration charges only 2 kWh from the grid; solar handles the remaining 3 kWh during the day.
Grid charge margin¶
The grid-charge calculation trusts the solar forecast. When the forecast is optimistic — or the weather turns out worse than predicted — solar may not deliver the expected surplus and the battery ends the day below max_soc. The optional Predictive Grid Charge Margin (%) hedges this by topping up the grid amount:
Continuing the example above, a 2 kWh grid need with a 50 % margin charges 3 kWh from the grid instead. The result is capped at gap_to_max, so the margin can never charge past max_soc. Default is 0 % (off); it also applies to the dynamic-pricing evening re-evaluation. Set it in the setup wizard, the options flow, or via the number.*_predictive_grid_charge_margin_pct slider on the dashboard Control tab.
Multi-battery systems¶
In systems with multiple batteries at different SOC levels the grid charge is distributed proportionally to each battery's individual gap to max_soc. A battery further from full receives a larger share; a battery already close to full relies mostly on solar for its remainder. This prevents overcharging any single unit from the grid and minimises total grid import.
Guaranteed minimum SOC floor¶
Predictive charging only grid-charges when the day nets to a deficit. On a sunny day the whole-day balance can be positive even though the battery is near empty at dawn — leaving the morning gap (before solar ramps up) covered from the grid at full price, or the battery drained.
The optional Guaranteed Minimum SOC slider (Control tab, 0 = off) forces an overnight grid charge to reach at least that SOC by the end of the charging window, regardless of the day's net balance. It sizes the deficit to the floor, so it flows through the per-battery target SOC and the dynamic-pricing slot sizing unchanged — the cheapest slots are still chosen to reach it.
It re-triggers with hysteresis: once SOC recovers to the configured floor, charging stops if the floor was the only reason to charge; it re-arms when SOC drops to floor − 5 %. Set it via the number.*_predictive_min_soc_floor slider, paired with the Guaranteed Minimum SOC switch.
Available modes¶
| Mode | Description |
|---|---|
| Time Slot | Charges during a fixed window (e.g. overnight off-peak tariff) |
| Dynamic Pricing | Automatically selects the cheapest hours of the day |
| Real-Time Price | Activates/deactivates charging based on the current price |
Notifications¶
The integration sends Home Assistant notifications:
- 1 hour before the slot starts: energy balance analysis and charging decision.
- When the slot starts: confirmation that charging has begun.
Use the Override Predictive Charging switch to cancel predictive charging at any time.
