HNordic's AI energy management system holds four inputs simultaneously — electricity prices, generation forecasts, frequency signals, and tenant demand. A battery managed on a fixed schedule captures one of these. The system captures all four.
References to "AI energy management" on HNordic's site describe a multi-variable optimisation engine — not a scheduling system, a smart thermostat, or a monitoring dashboard. The distinction is commercial: the four-axis simultaneous optimisation is why HNordic's energy income projections are not replicable by a property owner purchasing a battery from a hardware vendor and managing it manually.
The system accesses the day-ahead electricity price forecast from the relevant spot market — Nord Pool for Swedish properties, the GB day-ahead market for UK properties. These markets publish hourly price forecasts for the following day by mid-afternoon.
The system uses this forecast to plan the battery's charge-discharge cycle: charge from the grid (or from on-site generation) during the low-price hours; discharge to supply the building's load (or export where available) during the high-price hours. The margin between the charge cost and the discharge value is the energy arbitrage return.
In Nordic spot markets, the spread between daily low and high regularly exceeds 10:1. A battery that consistently charges at €8/MWh and discharges at €80/MWh earns materially more than one following a fixed overnight-charge/daytime-discharge schedule that does not adapt to the actual price curve.
Where wind and solar generation are installed, the system receives weather-based forecasts of expected generation output — wind speed at the turbine location and solar irradiance at the panel plane.
The generation forecast adjusts the battery's charge cycle: when significant generation is forecast, the system reduces or postpones grid charging, preserving charge capacity for generation capture. When generation is forecast to be low, the system draws from the grid during the cheapest hours available. The result is maximum self-consumption of on-site generation — the battery stores what the building generates rather than letting excess generation export at low prices.
The system reserves a defined proportion of battery capacity for frequency-response obligations — FCR-N, FCR-D, or aFRR in Sweden; Dynamic Containment in GB. This reserved capacity earns the grid-services capacity fee from the TSO regardless of dispatch volume.
The allocation between frequency-response reservation and energy arbitrage is dynamic. When frequency-response capacity fees are high relative to the arbitrage spread available, more capacity is committed to frequency response. When arbitrage spreads are wide and frequency fees are comparatively low, more capacity is available for energy optimisation. This real-time allocation is not possible with a fixed schedule; it requires continuous reassessment against live market data.
The system models the property's historical demand pattern — which hours of the day, which days of the week, create the peak demand events that generate the capacity charge component of the electricity bill.
Commercial electricity contracts typically include a demand charge based on the property's highest 30-minute peak in a billing period — often 20–40% of the total electricity cost. The system anticipates these peaks and pre-charges the battery to cover them, smoothing the metered demand profile. The grid sees a flatter draw; the demand charge falls.
This optimisation requires knowledge of the specific demand pattern for the specific property — not a generic profile. The system learns the property's patterns in the commissioning period and refines the model continuously as patterns change.
A battery managed on a fixed schedule — charge overnight, discharge during the day — captures a partial version of Input 1 only. It does not:
A fixed-schedule battery earns from one axis at a fraction of the available opportunity. The AI system earns from four axes simultaneously — and the margin between the two is the difference between a battery that reduces energy costs and one that generates a contracted income stream.
See also: Battery storage: backup power versus revenue asset · What are FCR-D, FCR-N, aFRR, and mFRR? · Full FAQ