Use cases
Each one is the same loop — collect, predict, dispatch, verify — pointed at a different goal. What changes between them is which devices are in the cabinet and what the household or the site is trying to achieve.
What the market takes for granted
The eight scenarios below are answers to these seven expectations — the ones a German or European installation is measured against before anyone asks what else the system can do.
Control that keeps running when the internet does not.
Measurements stay on the device unless someone chooses otherwise.
Devices from different manufacturers on one bus, not one vendor’s ecosystem.
Hourly and quarter-hourly prices as an input to the plan, not a display.
§14a EnWG: follow the operator’s command without shutting the building down.
Energy is part of the building, not a second system standing beside it.
A cabinet wired today is expected to still work in fifteen years.
Generation, consumption, tariff and device state are solved together rather than by four independent controllers arguing with each other. The result is a single committed plan the battery, heat pump and wallbox all follow.
A COP model of the heat pump and a thermal model of the building say how much heat the house will need and how long it holds it. The pump then stores heat ahead of the expensive window instead of reacting once it arrives. Hot water keeps priority.
You state the energy you need and when you are leaving. Surplus PV goes in first, the rest is pulled from the cheapest hours left before departure, and charging power follows what the house is doing at the time.
When the grid operator calls for a reduction, the cut is distributed locally across battery, heat pump and wallbox rather than applied bluntly. Base load, comfort and a departure deadline are all still honoured, and the response is verified back.
Charge through the cheap and negative hours, discharge into the expensive ones or export, and always serve local load first. Every decision lands in the journal with the price curve that produced it.
The billing peak is predicted from the site’s own load history, then storage discharges or flexible load steps back before the peak is set. On a demand-charge tariff a peak avoided is the whole saving.
Available power is measured, not assumed, then shared across the charge points by priority and departure time. More vehicles fit behind the same grid connection instead of a connection upgrade.
PV serves the load first, the heat pump runs in the high-generation and low-price hours, surplus goes to storage, and storage covers the expensive peak. One optimiser, not three timers.
When goals disagree
Every scenario above can produce two answers at once — the tariff says charge, the grid operator says do not. The controller does not improvise: it applies a stated order, and the journal records which rule decided.
A curtailment instruction or a protection limit is never overridden by a cheaper hour.
Room and hot-water setpoints and a stated departure deadline hold against the tariff.
Within what safety and comfort allow, the cheapest hours are used.
All else equal, own generation is used before the grid.
Tell us what is in the cabinet and what you are trying to achieve. We will show you the schedule OptiBee Energy would commit for it.