Experimental based multi-objective optimisation for structured packed bed liquid desiccant dehumidification systems

In last few decades, a substantial advancement has been observed in dehumidifier modules for their energy-conservative and environment-friendly nature; however, still uncertainty in decision-making is of concern in selection of optimum input parameters to attain an effective performance. In this perspective, an experimental study is performed in an adiabatic packed bed, counter-flow liquid desiccant dehumidifier fuelled with the blend of lithium bromide and calcium chloride solution at various operating conditions with different control settings. It is observed that air mass flux rate (F_a), solution concentration (ζ) and air specific humidity (ω_a) have significant impact on condensation rate (CR), whereas, air mass flux rate (F_a) and solution temperature (T_s) are the most effective parameters for moisture effectiveness and latent heat factor, respectively. The experimentally obtained results for multiple response performance characteristics are represented based on inlet process parameters using gene expression programming (GEP). The proficiency of developed GEP meta-models is provided by statistical parameters, Taylor diagram and Theil uncertainty. Subsequently, GEP meta-model based fuzzy logic is developed for optimising dehumidifier inlet process parameters in terms of multi-responsive performance characteristics using Genetic Algorithm. The optimisation results showed that dehumidifier performance characteristics have a tendency to become optimal at F_a = 0.766 kg/m²-s, T_a = 30.745 °C, ω_a = 0.023 kg_wv/kg_da, F_s = 1.812 kg/m²-s, T_s = 24.01 °C, and ζ = 48.1%. Finally, experiments were conducted at optimum conditions and observed proximity between the predicted and experimental outcomes, which further validates the current optimisation approach.