Fast model for air source heat pump frosting and defrosting behavior

The most popular heat pump system is the air-to-water heat pump, known for its lower investment cost and easier installation. However, in cold climates where outdoor temperatures drop to near 0°C and below, moisture condensation freezes on the air-refrigerant heat exchanger surface, negatively impacting the system’s energy performance until the ice is removed. Common methods involve heating the surface to melt the ice, a process referred to as the ‘defrost cycle’.

Yearly simulations should account for this effect to correctly assess the impact of the defrosting cycle on the building and the seasonal performance of the heat pump. Accurate models for frost formation and defrosting process are present in literature. However, they tend to suffer from high computational requirements. Another approach is having a static model that adjusts the instantaneous Coefficient of Performance (COP) as a function of ambient temperature to account for a defrost cycle.

However, these simplified models do not account for the dynamics of the frost formation and defrost cycle, which can lead to inaccurate results because they do not take into account the dynamic response of the building to the defrosting cycle. In this work, we present a simplified Modelica model that can estimate frost accumulation and defrost times based on external air conditions and the power extracted from the evaporator. The model has low computational demand and a small set of tuning parameters.

The model was validated with experimental data from a small capacity heat pump and compared to the results from a detailed physical model. The results show an average percentage error with respect to the final value of less than 5.6% for ice formation predictions and 10% on the starting time for defrosting trigger events.