This study develops physics-informed reduced-order models for a residential air-to-water heat pump (AWHP) system coupled with both sensible and latent thermal energy storage. The objective is to obtain fast, control-ready models that preserve key physical behaviour across operating modes and under frequent on–off cycling.
Three one-year datasets at 30-minute resolution are generated using a validated TRNSYS simulation, each driven by a different weather year for the same place. The system comprises a dwelling model, an AWHP, a domestic hot water tank, and thermal storage. Operating data cover direct supply by AWHP, indirect operation via AWHP and a buffer tank, and a combined mode in which the AWHP charges storage overnight and prioritises storage discharge during daytime before restarting the AWHP.
The AWHP is modelled using a control-oriented reduced-order formulation that maps air-side conditions, water-side inlet temperature, mass flow control, and setpoint tracking to heating output and electrical power, with explicit attention to low part-load operation and start-up transients. Water tank and PCM tank models are derived from an enthalpy-based multi-node formulation and reduced to a small number of stratification states that capture mass-flow energy transport, inter-layer heat exchange, and ambient losses. Latent storage is represented through an invertible enthalpy–temperature relation with hysteresis switching between heating and cooling curves, based on an experimentally tested 382 L tank containing water and 154 kg of sodium acetate trihydrate PCM.
The modelling framework is validated against TRNSYS outputs. It reproduces key dynamic responses in heat pump capacity, electrical power, buffer tank temperatures, and charging and discharging heat rates, while reducing simulation time by approximately 5 times for the water tank case and by 10 to 30 times for the PCM tank case compared with the TRNSYS reference. The resulting models provide a practical route to physics-informed, control-ready models for residential heat pump systems with water and PCM thermal storage.
Previous Article in event
Next Article in event
Physics-Informed Reduced-Order Models for Air-to-Water Heat Pumps Coupled with Sensible and Latent Thermal Storage
Published:
22 June 2026
by MDPI
in The 1st International Online Conference on Inventions
session Energy system analysis and modelling
Abstract:
Keywords: air-to-water heat pump; reduced-order model; physics-informed modelling; thermal energy storage; phase change material
