A quantitative correlation model for guiding VOC removal in internally-cooled liquid desiccant dehumidifiers

Volatile organic compounds (VOCs), which are widely present in indoor air, pose a serious threat to human health. Liquid dehumidification technology has gained increasing attention due to its ability to regulate humidity and remove VOCs simultaneously. Among these technologies, the internally-cooled dehumidifier addresses the performance limitations of adiabatic dehumidifiers caused by temperature rise by maintaining the solution at a lower temperature, demonstrating significant application potential. However, under specific dehumidification requirements in different environments, the quantitative relationship between its operating parameters and purification performance remains unclear, limiting optimal design for practical applications. To address this issue, a quantitative correlation model for benzene and formaldehyde removal is proposed to represent the relationship between purification performance and key parameters. Using a developed correlation method, the model identifies three critical parameters characterizing VOC transfer capacity: air outlet humidity, gas-liquid flow rate ratio, and number of VOC transfer units. These parameters are then integrated to predict overall purification efficiency. Results indicate that the model prediction aligns closely with existing simulation data, with deviations controlled within ±20%. By visually integrating performance and key parameters, the model allows VOC removal performance to respond rapidly to varying dehumidification requirements. Furthermore, the model clearly demonstrates that lower air outlet humidity, smaller gas-liquid flow rate ratio, and larger number of VOC transfer units synergistically enhance benzene removal. Finally, the correlation mechanism is quantitatively revealed based on the distribution of these key parameters. This research provides a convenient method for predicting dehumidifier performance in removing VOCs under varying operating scenarios.