Parametric influence and optimization of the dehumidification performance of a variable-frequency dehumidifier.

This paper optimizes a variable-frequency dehumidification system (VFDS) based on the principle of vapor compression refrigeration (VCR), wherein the compressor frequency and fan speed can be changed. The effect of the compressor frequency and fan speed on the operating parameters of the VFDS, including the evaporating temperature (<italic>T</italic>e), condensing temperature (<italic>T</italic>c), superheating (Δ<italic>T</italic>sh), and subcooling (Δ<italic>T</italic>sc), is studied experimentally under three operating conditions. The effect of variable-frequency on the dehumidification performance is analyzed at the mechanistic level through two dehumidification performance indicators, namely, the dehumidification capacity (DC) and dehumidification energy efficiency (DEE). Δ<italic>T</italic>sh is primarily controlled by the capillary throttling effect, which always starts to cross the “0” point when the fan speed is 600 rpm under different conditions. An increase in the compressor frequency leads to the opposite trend for DC and DEE, but their corresponding optimal fan speeds are relatively close. At 27 °C/60%, the optimal fan speed intervals with DC and DEE as single optimization objectives are both [600 rpm, 700 rpm]. However, at 30 °C/80%, the overall difference between the optimal fan speed intervals with DC and DEE as optimization objectives is 50–100 rpm. Importantly, the risk of frost at 18.3 °C/60% can lead to a significant extension of the optimal fan speed interval, which is [500 rpm, 900 rpm], making it more difficult to control. The concept of the dehumidification operating temperature difference (Δ<italic>T</italic>) is proposed to comprehensively evaluate the dehumidification capacity and as a control index for variable frequencies. The relationships among the optimal Δ<italic>T</italic> and the ambient dew point temperature (<italic>Td</italic>) and compressor frequency are established through response surface methodology (RSM) with a maximum deviation of 5.97%. This work can provide an optimized variable-frequency solution for the VFDS and some guidance for the control logic development of the VFDS. [ABSTRACT FROM AUTHOR]