In this post we use a simulation model to analyze the influence of frosting on the cycle and the performance. The knowledge gained forms the basis for the development of defrosting strategies.
Air source heat pumps use outside air as heat source to evaporate the refrigerant. They are typically used to provide hot water for heating of buildings and are the most common type of heat pumps in Germany (according to BWP/BDH sales statistics). In winter there is a risk that the evaporator freezes up leading to a drop in performance of the heat pump. In extreme cases, the cycle can come to a complete standstill.
Outside air is moved by a fan over the heat pump’s evaporator releasing heat to the refrigerant in the process. This reduces the temperature of the outdoor air. At low temperatures and high humidities, the air can be cooled down to such an extent that water condenses and freezes on the cold surface of the evaporator. Over time, a layer of ice forms on the evaporator.
In the following, we use a 1D model of the heat pump to show the influence of evaporator frosting on the cycle and thus the performance of the heat pump. We have modeled the heat pump with the model library TIL in the modeling language Modelica. TIL is a specialized library for modeling thermal systems.
In our example we simulate a typical winter day in Germany: The outside air temperature is 0°C with a humidity of 95%. The heat pump uses CO2 as refrigerant. The heat pump’s expansion valve is controlled in such a way that a condensation pressure of 100 bar is achieved. The fan is not controlled and operates at constant speed. In normal operation the selected operation conditions lead to an evaporation pressure of 27 bar. This corresponds to an evaporation temperature of about -9 °C. The cycle is shown in the p-h diagram.
The heat pump cycle with a frosted evaporator reveals a significantly lower evaporation pressure compared to the heat pump cycle in normal operation. The reduced evaporation pressure is characteristic of frosted evaporators and is mainly due to two effects:
To still provide the required heat for the building, the evaporation temperature and thus the evaporation pressure of the refrigerant must drop. The increased thermal resistance of the evaporator and the reduced heat capacity rate of the outdoor air are thus compensated by an increased temperature difference between refrigerant and outdoor air.
More detailed analyses of the simulation results show that the air-side pressure drop has a significantly higher influence on the drop in evaporation temperature than the increased thermal resistance of the evaporator. This is mainly due to the fact that ice has a relatively high thermal conductivity and therefore the thermal insulation effect of the ice layer is not as important.
A small degree of frosting does not lead to a noticeable loss of performance of the heat pump: the COP (coefficient of performance) remains almost constant at 2.95 up to approx. 100 g of ice. Only when larger amounts of ice have formed on the evaporator the COP drops visibly. With 350 g of ice, the COP has finally decreased by approx. 25 % compared to normal operation. At this point at the latest, the evaporator should be defrosted in order to prevent the cycle from collapsing.
By simulating a heat pump, we were able to identify the main effects of evaporator frosting on the cycle and the performance. We will discuss possible defrosting strategies for the evaporator in a later post.