Optimal defrost initiation for air-source heat pumps: Evaluating the improvement potential of common defrosting controllers
Jonas Klingebiel, Christoph Höges, Stephan Göbel, Paul Bannmüller, Thies Boelsen, Valerius Venzik, Christian Vering, Dirk Müller
Abstract
The timing of defrost initiation for air-source heat pumps (ASHPs) significantly impacts operational efficiency, especially in cold climates. While literature studies demonstrate the existence of optimal defrost initiation time ( t opt ), there is limited knowledge how operational parameters (e.g., air temperature, air relative humidity and heating capacity) influence t opt . Additionally, the performance gap between conventional defrosting controllers and the theoretical optimal defrost strategy remains uncertain, leaving the improvement potential for further refinements of defrost controllers unknown. Hence, this study investigates the impact of defrost timing on ASHP efficiency through experimental tests and simulations. First, experiments are carried out with a variable-speed ASHP and t opt is quantified across different operational parameters. We show that a 20 min deviation from t opt reduces efficiency by up to 9.1%. Second, the experiments are extended through simulations, providing a comprehensive defrost map that correlates t opt for a broad range of operational parameters. The results show that t opt is highly dependent on the heating capacity and thus, t opt varies depending on the application (e.g., building type). Moreover, the simulation compares common time-based (TBD) and demand-based (DBD) defrost controllers with the optimal defrost strategy. TBD shows efficiency losses of up to 16.0% compared to the optimal strategy, while DBD reduces this gap to a maximum loss of 7.0%. We show that optimizing the constant initiation threshold for DBD slightly improves performance; however, trade-offs between severe and mild frosting conditions remain, limiting overall efficiency. Hence, our results indicate that threshold correlations or adaptive control algorithms could further improve operational efficiency. • Optimal defrost initiation timing depends on ambient conditions and heating capacity. • A 20 min deviation from optimal defrost initiation timing reduces efficiency by up to 9.1%. • Common controllers are compared to the optimal defrost strategy across various conditions. • Time-based defrosting shows up to 16% efficiency gap to the t opt -strategy. • Demand-based defrosting shows up to 7% efficiency gap to the t opt -strategy.