Cascade deep air dehumidification with integrated direct-contact cooling and liquid desiccant absorption

Deep dehumidification of the feed air is expected for the energy saving and stable operation of the air compression process in large-scale industries. Liquid desiccant-based dehumidification technology provides an alternative solution mainly featuring the continuous operation and low regeneration temperature, but the conventional systems still face a challenge to achieve the deep air dehumidification at high air-to-solution flow ratios, especially under high humidity conditions. Based on the cascade matching of driving forces, this study first presents a cascade deep air dehumidification system integrating a direct-contact cooling module (DCCD) and multi-stage internally-cooled liquid desiccant module (MILDD). In particular, the DCCD uses a novel internal matrix structure where the moist air is in direct contact with the cooling water with low hygroscopic. To evaluate the proposed system, a complete thermodynamic model is developed based on the newly-fitted general empirical correlation for mass transfer coefficient between the moist air and cooling water. The numerical model is validated against the initial experiment of the DCCD–MILDD test bench with a mean relative discrepancy of 6%–9%. The supplied air states and energy performance of the cascade system are evaluated at low humidity and high humidity conditions. The cooling water utilization is further improved by two series-looping schemes of cooling water allocation. Numerical results show that the DCCD–MILDD cascade system can achieve deep air dehumidification with the outlet air humidity ratio lower than 6.2 g/kg and the air-to-solution flow ratios over 4.0 even at high humidity conditions. This attributes to an improvement of the coefficient of performance and exergy efficiency, respectively 1.90 and 0.57. This work may provide an effective design of the liquid desiccant-based deep air dehumidification systems applied to the large-scale air compression process.