Simulation and experimental investigation of a two-stage oil-free magnetically levitated centrifugal chiller with R1234ze(E) refrigerant

Driven by carbon neutrality targets and the Kigali Amendment, oil-free magnetically levitated centrifugal chiller (MLCC) with low-global-warming-potential (GWP) are expected to become a core technical solution for high-efficiency building cooling systems. Yet conventional simulation methods rely on fixed compressor efficiency or empirical correlations, which induce considerable performance prediction errors ranging from 5% to 10%, failing to support precise off-design performance evaluation. To investigate the operating characteristics of MLCC and verify the reliability of the simulation method, a Dual-domain Coupled Simulation (DCS) method for the refrigeration system was developed by combining numerical simulation and cycle model, which can accurately capture the off-design aerodynamic characteristics and achieve superior prediction accuracy for the full-range performance of an R1234ze(E) chiller. And a 2800 kW-class oil-free magnetically levitated centrifugal chiller test rig with R1234ze(E) refrigerant was built to validate model accuracy. The performance of the compressor and the refrigeration system was systematically analyzed under various operating conditions, including constant evaporating temperature under constant speed and variable speed, constant condensing temperature under constant speed and variable speed, and full-range operating conditions. The results indicate that the relative deviation between simulated and experimental results remains below 2.5% except for a few operating conditions such as constant condensing temperature under constant speed. Under both constant evaporating and constant condensing temperature conditions, reducing the compressor rotational speed can maintain a constant cooling capacity while achieving a higher system COP. Nevertheless, the speed adjustment is constrained within a narrow range of ±10% of the rotational speed corresponding to the optimal efficiency. The DCS method proposed in this paper can provide a theoretical and technical reference for the performance prediction, operational parameter optimization and energy-saving strategy formulation of similar units.