[FR] Price-aware solar surplus absorption — only absorb during the cheapest hours

[joostlammers-web]

Summary

On a fully dynamic (hourly) tariff where export is paid at the same rate as import, absorbing solar surplus the moment it appears is not free — it costs the export revenue of that hour. Charging from surplus and charging from the grid are economically identical; only the timing matters.

Omnibattery already reasons about price for grid charging (min_arbitrage_margin, max_price_threshold, negative-price charging), but surplus absorption is pure load-following: the PD controller charges whenever there is export, regardless of what that hour is worth.

Measured over 31 days on a single Marstek Venus E: charging the same daily volume in each day's cheapest hours instead would have cut the average charge price from € 0.213 to € 0.162 /kWh, roughly doubling the battery's net return (€ 0.30 → € 0.55 per day). The loss comes from absorbing surplus from mid-morning onwards, when export is still worth € 0.23–0.28, instead of waiting for the € 0.16–0.18 window a few hours later.

Measurement detail

Single battery, single phase, NL hourly-priced tariff, 31 full days, 149.4 kWh charged.

Average price by hour of day over the period (€/kWh):

00-05   0.32  0.31  0.30  0.30  0.30  0.31
06-11   0.33  0.33  0.31  0.28  0.23  0.19
12-17   0.17  0.16  0.17  0.18  0.21  0.27
18-23   0.32  0.36  0.37  0.36  0.35  0.33

Actual charging cost € 31.77 (€ 0.213/kWh). Re-allocating the same per-day volume to
that day's cheapest hours, capped at the configured max charge power, gives € 24.22
(€ 0.162/kWh). Discharge value is unchanged at € 42.24, so the net goes from € 9.37 to
€ 16.91 over the period. Day-ahead prices are published, so this is schedulable rather
than a hindsight-only optimum.

What I'd like

Goal-seeking behaviour rather than a rule: the battery decides when to take its energy, given how much it needs and what each hour costs.

• Surplus exports while the current hour is expensive relative to the cheaper hours still ahead that could meet the same target.
• The day's target follows from expected consumption and the SoC floor, not a fixed time window.
• It re-plans during the day. If a cloudy cheap window delivers less than expected, absorption moves to the next-cheapest hours instead of the day ending short.
• It honours a deadline. If surplus alone won't meet the target, it grid-charges in the cheapest remaining hours before the evening peak rather than silently under-charging.

I'm describing the outcome rather than the mechanism, but one thing stands out: the predictive planner already holds every input this needs — solar forecast, expected consumption, price curve, SoC floor, safety margin. Today it answers "should I grid-charge tonight?"; the same inputs answer "when today should I absorb surplus, and do I need to top up if that falls short?". In winter, with the cheapest hours at night and no surplus, the behaviour collapses back into the grid charging that already exists — so this may be a widening of the planner's remit rather than a new subsystem.

Why the current settings don't cover it

• charge_delay ("holds back grid charging while solar can still cover the required energy") acts on grid charging, not surplus absorption.
• max_price_threshold and min_arbitrage_margin gate predictive grid charging only.
• system_max_charge_power can't serve as an external gate: 0 W means disabled (no cap), not no charging.

Gating the battery's own battery_allow_charge from an automation leaves the PD controller commanding charge power the hardware refuses, and desynchronises Omnibattery's energy accounting. Driving set_charge_power externally appears to hit the same conflict as #334.

Versions

Omnibattery v1.3.0 (HACS), Home Assistant 2026.8.3, one battery, AC-coupled, single phase.

[github-actions[bot]]

Thanks for the suggestion! 🙌

Here's how this works:

• Vote by clicking 👍 on the original post — not in the comments
• The more votes an idea gets, the higher it moves in the priority queue
• I review feature ideas periodically and implement the ones that benefit the most users

If your idea is very specific to your setup, it may not make it into the core integration — I'll try to suggest a workaround in that case.

Feel free to discuss details or alternatives below.

[syphernl]

This should address this feature request (alongside a similar one): #355

[joostlammers-web]

Update: this is now implemented by PR #355 (merged Sep 4), tracked in #237 where @ffunes committed it for v1.5.0 — so this request is effectively a duplicate of #237. Anyone landing here: please add your use case / votes there. Marking this as answered; looking forward to the release and happy to test on my Venus E (NL, symmetric feed-in tariff).

[ffunes]

Implemented in v1.5.0 as Price-aware solar surplus absorption (Dynamic Pricing, opt-in): the battery lets surplus export while a clearly cheaper export window is still ahead and takes the day's charge in the cheapest hours instead. Turn on its switch on the Control tab and, optionally, set an export price sensor. Thanks @joostlammers-web.