Limits of modeling approach for refrigerant charge assessment
Given their performance, heat pumps are a key technology to decarbonize the heating production in buildings. Their share in the residential heating market is increasing at a very fast pace, making preventive maintenance and, thus, automatic fault detection and diagnosis always more necessary, especially to maintain an optimum efficiency and lighten the load for installers-maintainers. Soft and progressive faults such as refrigerant leakages must therefore be detected before they deteriorate the performance of the system or even damage the components. Early detection of leakages and real-time quantification of the refrigerant charge is thus a major challenge. Several methods for assessing refrigerant leaks have been developed. Most rely on a statistical approach, where an empirical relationship is established between the refrigerant amount in the system and other features. Such methods can be rather accurate, but rely on a significant amount of experimental data for their calibration. Thus, developing a physical model of the system to generate the training data numerically seems to be a very promising lead. It also allows to represent the distribution of the refrigerant charge in the various heat pump components and thus better understand the system’s behaviour. Developing such a model is however not an easy task, as several components containing a significant amount of refrigerant are very difficult to model accurately due to the often unknown internal geometry. Many models neglect these components or make highly simplifying assumptions. Therefore, in a first part, this study aims to describe the problematic components, in particular the suction accumulator and the filter, to study their modeling and the associated uncertainty. In a second part, the objective is to assess the consequences of incorrectly taking the characteristics of these components into account on the quantification of the total charge in the heat pump.