← Back to the optimizer

The RKV method

Energy. Working together.

Plan heating, charging and storage using solar forecasts, electricity prices and your needs.

Suitable devices, measurements and authorised interfaces are required. We agree the scope for your installation; savings are not guaranteed.

48 hours as an example.

Interactive example plan

Synthetic winter values and simplified rules. No product-solver run, live forecast or device control.

WeatherOutdoor temperature · solar forecast
Electricity tariffPrice windows
Inverter & batterySolar power · state of charge
EV & chargerCharge target · departure time
Heat pumpPower + heat · COP learning
Building & roomsRoom temperature · heat demand
Energy Hubconnects measurements and forecasts
Optimizerplans around cost, comfort and device limits
48-hour outlookHover or tap chart: home data
T0 · now+24 hT+48 h
Included consumers Switches change the example calculation

48 h in 15-minute steps; other rows use 4-hour blocks. Heating: deeper blue = higher electrical power, gaps = pauses. Long operation at low power is normal on cold days. Battery: green = charge, blue = discharge.

Grid exportGrid import after battery

Switches include or exclude loads. Different demands or storage end states are not a fair savings comparison.

Heating demand and efficiency.

A building can store heat. When to heat depends on demand, comfort and the cost of producing that heat.

Building and hot-water storage

The model balances a faster room node and a slower building-mass node: stored heat changes with supplied heat, internal and solar gains, and heat lost outdoors. Room measurements and calibration are needed for an actual installation.

Hot water has a separate heat store and demand. The example prioritises space heating if the next 15-minute interval would otherwise fall below its 20.4 °C threshold. This simplified guard is not a promise of real-world comfort or a drinking-water hygiene model.

COP and heating costs

COP is useful heat divided by electrical energy: a dimensionless ratio. Heating and hot-water cycles are assessed separately, including start-up energy. Meter boundaries, sampling and missing operating ranges limit what can be inferred.

Heat-generation cost = required heat ÷ COP × the relevant electricity price. Direct solar electricity has a cost through forgone export revenue; battery losses and wear also matter. The example's “direct heat price” assumes a 2 kW heat pump supplied by residual solar after the base load and the grid, without a battery or optional consumers.

NIST background on transient heat-pump operation

Electricity balance and the product optimizer

At the AC connection: solar + grid import + battery discharge = consumers + battery charge + export. Power is measured in kW; energy in a time window is mean power × duration, in kWh. Battery state also accounts for charge and discharge losses.

Planning costs = grid-import costs − export revenue + assumed battery-wear costs. Comparisons require the same covered demand and comparable thermal and battery end states. Unmet heating or an empty store is not a saving.

The product optimizer searches permitted profiles using mixed-integer optimisation. The graphic above uses simplified rules to explain the relationships; it does not run that solver. Method reference.

Planning is limited to shared, gap-free coverage of prices, solar, weather and consumer requirements. Missing measurements mean unknown, not zero demand.

Higher electricity price. Lower heating cost.

12 kWh of heat in both periods · a fictional example.

Same heat, different electricity costsSchematic comparison
Cold night −2 °C
Electricity price
100%
COP
2.5
Electricity use
4.8 kWh

Heat cost 100%

Warmer afternoon +8 °C
Electricity price
120%
COP
4.0
Electricity use
3.0 kWh

Heat cost 75%

How the 25% difference is calculated

With COP 2.5, 12 kWh of heat needs 4.8 kWh of electricity; with COP 4.0 it needs 3.0 kWh. A 20% higher electricity price gives 1.20 × 3.0 ÷ 4.8 = 0.75 of the generation cost.

The comparison assumes grid electricity, equal useful heat, comfort and a comparable thermal end state. Additional losses, defrosting and auxiliary loads can remove the advantage: producing 16 instead of 12 kWh in the warmer period would give equal costs here. This is not a measured customer result or a promised saving.

Assumptions of the example.

Synthetic winter days; no customer data or live forecast.

Parameters and limits
Parameters of the 48-hour example
PartAssumptions
Time and weather15-minute steps; a stable starting time in Germany. Outdoor temperature −8 to +4 °C; 10 kWp solar, zero at night, up to 5.3 kW by day.
BuildingBoth nodes start at 21 °C. Heat capacities 2 and 18 kWh/K; outdoor losses 0.08 and 0.04 kW/K; coupling 0.7 kW/K. Heating target 20.8–21.6 °C.
Heat pumpUp to 2 kW electrical; heating modulates from 0.6 kW. Assumed COPs at 35 °C heating flow / 50 °C hot water, without individual device maps or defrost cycles. A final hot-water interval can represent a shorter run at its interval-average power.
Run timesHeating aims for at least 2 hours running and 1 hour paused; the model's temperature guard and hot-water priority can override these. Changing heating mode is not itself a compressor restart.
Hot water10 kWh heat store, initially 7.5 kWh. Demand 6 kWh/day plus up to 1.44 kWh/day standing loss; charging towards 9 kWh after a lower threshold. No temperature stratification or hygiene simulation.
Electricity and battery0.8 kW base load. Battery 10 kWh, initially 5 kWh; usable state 1–9 kWh, up to 3 kW each way. About 93.8% efficiency each way, 88% round trip. No simultaneous charge/discharge or grid import/export within one step.
EVConnected 18:00–07:00. Each complete session requests 12 kWh at the vehicle battery; charging up to 3.6 kW with 90% efficiency. An initial partial session uses achievable remaining demand.
Optional consumersPool filter 0.7 kW; pool heat pump 1.8 kW, only with the filter and outside temperature at least +2 °C. Washing machine: a complete 2-hour, 1.1 kWh cycle with a variable profile up to 2 kW.
ValuationSolar export value 8 ct/kWh; battery wear 4 ct/kWh delivered. These are assumptions, not a household cost or savings calculation.

The battery charges from residual solar and discharges to loads. Optional consumers change the remaining solar and grid demand; they do not jointly reschedule heating or EV charging. Pool rows describe electrical loads only and are not a winter-operation, frost-protection or temperature recommendation. Actual use needs suitable interfaces, flow, safe start permission and commissioning.

See the cost example on the homepage →