Share

Development, calibration and validation of a dynamic modeling framework for air-source heat pumps under cycling of frosting and reverse-cycle defrosting.

Author(s) : MA J.

Type of monograph: Doctoral thesis

Summary

Frost accumulation on evaporator coil surfaces can significantly degrade the performance of air-source heat pump (ASHP) systems in winter operations. The continued buildup of
frost eventually necessitates a defrosting mode to remove the accumulated frost and return the system to its normal operating characteristics. With the emerging demand for
improving energy efficiencies and reliability of ASHP systems that operate in cold climates, a simulation tool capable of capturing the system dynamics with continuous mode-switching between heating and defrosting operations is extremely useful in the development and evaluation of advanced control and fault detection and diagnostics (FDD) algorithms. However, first-principles simulation of complicated thermo-fluid behaviors involving reverse-cycle defrost imposes a variety of numerical challenges including mode switching, flow reversal and multi-scale transients. Moreover, the inherent complexity of first-principles models that are represented by large systems of nonlinear differential-algebraic equations hinders calibrating such models with classical gradient-based approaches, as it is typically essential to calibrate model parameters with available measurements for achieving reliable predictions in presence of uncertainties.

Despite the abundance of modeling and experimental studies involving ASHP defrosting performance, there has not been an experimentally validated simulation tool that is capable
of predicting transients of general ASHP systems under cycling of frosting and reverse-cycle defrosting operations in the literature. Almost all of the previous efforts in modeling defrost cycles omitted the transition stage from nominal heating modes to defrost modes and, as a result, are not applicable to control and FDD purposes. This thesis presents a comprehensive dynamic modeling framework for reversible ASHP systems incorporated with non-uniform frost formation and melting. A complete first-principles ASHP cycle model is described where frost dynamics are integrated into a finite-volume evaporator model to characterize overall system behaviors under frosting-defrosting cyclic operations. Due to the intricate underlying cycle dynamics coupled with those of the frost and rapid state evolution triggering reverse refrigerant flow, this type of model can suffer from a lack of robustness which is crucial for failure-free simulations of complex thermo-fluid systems. In the present work, robust formulations are adopted across component models to improve model robustness. An experimental setup of a 2-ton residential ASHP unit is described with testing procedures for transient validation data collection. A systematic approach based on Bayesian optimization to simultaneously calibrate a large set of parameters associated with the dynamic thermofluid component models is then presented and demonstrated utilizing load-change transient measurement.

The developed model is simulated to investigate the transient behaviors of the ASHP system under multiple cycles of frosting and reverse-cycle defrosting. Comparisons between the model predictions and measurements demonstrate the predictive capabilities and numerical reliability of the developed modeling framework to capture system characteristics with cyclic frost buildup and removal through reversing the refrigerant flow. Additionally, predictions of the evolution of frost properties over time provide insights into the non-uniform frost formation and melting phenomena, which are typically difficult to measure and characterize for an ASHP system. The proposed models are attractive for development and evaluation of improved control and FDD designs for ASHP systems with frosting and defrosting taken into consideration.

Available documents

Format PDF

Pages: 146 p.

Available

Free

Download

Details

  • Original title: Development, calibration and validation of a dynamic modeling framework for air-source heat pumps under cycling of frosting and reverse-cycle defrosting.
  • Record ID : 30032828
  • Languages: English
  • Subject: Technology
  • Publication: Purdue University
  • Publication date: 2024
  • DOI: http://dx.doi.org/https://doi.org/10.25394/PGS.25521136.v1

Indexing