A new scheduling model targets the hard part of electrified heating

Researchers at Cranfield University in the United Kingdom have developed a scheduling model for residential heat pumps designed to reduce electricity costs while preserving thermal comfort in homes with rooftop solar.

The work, reported by pv magazine, addresses a practical challenge that is becoming more important as households combine heat pumps, solar panels, and electricity tariffs that change over time. A heat pump can shift some demand away from expensive hours, but only if it can do so without leaving occupants too cold or too warm. Rooftop photovoltaic generation adds another variable because solar output is available at some times and uncertain at others.

The Cranfield model is intended to coordinate three resources at once: grid electricity, rooftop PV generation, and the building’s own thermal flexibility. That means deciding when to draw power from the grid, when to use solar generation directly, and when to rely on stored heat in the building fabric or indoor environment.

Dynamic tariffs create an opening for smarter control

Time-varying tariffs change the economics of heating. Electricity may be cheaper during some periods and more expensive during others, creating an incentive to move flexible loads into lower-cost windows. Heat pumps are a major candidate for this kind of load shifting because they consume electricity and can often preheat a home slightly before prices rise.

Banu Yektin Ekren, the corresponding author, told pv magazine that rooftop PV strengthens the heat pump’s load-shifting potential under dynamic tariffs because it gives the scheduler a low-cost source of electricity beyond the grid. The optimization can coordinate when electricity is cheap, when PV is available, and how much thermal flexibility the building can provide.

This is a more complex problem than simply running a heat pump when solar panels are producing. Solar generation is uncertain, household comfort has limits, and electricity prices may not align neatly with peak PV output. A useful scheduler has to make tradeoffs among those factors rather than optimize for one variable in isolation.