Performance analysis for a membrane-based liquid desiccant air dehumidifier: experiment and modeling

2019-01-16T21:01:22Z (GMT) by Xiaoli Liu
Liquid desiccant air dehumidification (LDAD) is a promising substitute for the conventional dehumidification systems that use mechanical cooling. However, the LDAD system shares a little market because of its high installation cost, carryover problem, and severe corrosion problem caused by the conventional liquid desiccant. The research reported in this thesis aimed to address these challenges by applying membrane technology and ionic liquid desiccants (ILDs) in LDAD. The membrane technology uses semi-permeable materials to separate the air and liquid desiccants, therefore, the solution droplets cannot enter into the air stream to corrode the metal piping and degrade the air quality. The ILDs are synthesized salts in the liquid phase, with a large dehumidification capacity but no corrosion problems. In order to study the applicability and performance of these two technologies, both experimental and modeling investigations were made as follows.
In the study, experimental researches and existing models on the membrane-based LDAD (MLDAD) was extensively reviewed with respects of the characteristics of liquid desiccants and membranes, the module design, the performance assessment and comparison, as well as the modeling methods for MLDAD.
A small-scale prototype of the MLDAD was tested by using ILD in controlled conditions to characterize its performance in Oak Ridge National Lab. The preliminary experimental results indicated that the MLDAD was able to dehumidify the air and the ILD could be regenerated at 40 ºC temperature. However, the latent effectiveness is relatively lower compared with conventional LDAD systems, and the current design was prone to leakage, especially under the conditions of high air and solution flow rates.
To improve the dehumidification performance of our MLDAD prototype, the two-dimensional numerical heat and mass transfer models were developed for both porous and nonporous membranes based on the microstructure of the membrane material. The finite element method was used to solve the equations in MATLAB. The models for porous and nonporous membranes were validated by the experimental data available from literature and our performance test, respectively. The validated models were able to predict the performance of the MLDAD module and conduct parametric studies to identify the optimal material selection, design, and operation of the MLDAD.