Modeling and multi-objective optimization of vacuum membrane distillation for enhancement of water productivity and thermal efficiency in desalination

Abstract

Water productivity and thermal efficiency in membrane distillation (MD) have been the main research targets, for the aims of commercial application in desalination. The comprehensive understanding of the influence of module configuration parameters, operating conditions and their interaction on MD performance is the key for MD commercialization. In this paper, the multi-objective modeling and optimization in the vacuum membrane distillation were performed by response surface methodology and desirability function approach. A series of PVDF hollow fiber modules of different scale were used to provide the essential data and to verify the modeling program. The multi-objectives including water permeate flux (J), water productivity per unit volume of module (Pv), gained output ratio (GOR), and a comprehensive index (Dm) assessing the 10 desired MD performances were predicted and experimentally verified. The influence of operating 11 parameters (temperature, velocity, and concentration of feed) and membrane module parameters 12 (membrane packing density and length-diameter ratio of module) and their binary interactions on the multi-objectives was investigated. It is found that among the investigated factors, feed inlet temperature and its interaction effect with module parameters play dominant roles on MD performance. Under the multi-objective optimum conditions, 4.85×10 3 kg/(m 3 •h) of Pv and 0.91 16 of GOR were achieved within the investigated range. Water productivity and thermal efficiency 17 can be simultaneously enhanced by optimizing operating and module conditions with the approach 18 developed in this study. 19

Review of Transport Phenomena and Popular Modelling Approaches in Membrane Distillation

“…The RSM modelling approach is a visualisation technique based on mathematical statistics and can plot a 3D response surface with two variables [ 27 , 28 , 31 , 119 ]. In this approach, polynomial fits are conducted for sets of experimental data to form an adequate representation of the experimental response.…”

“…The experimental data are foundational for RSM. The data used in the RSM normally include solute concentration, stream velocity, temperature, flux, GOR, vapour pressure and energy input [ 27 , 29 , 31 , 121 , 122 ]. However, all these data should be from the identical systems, and the predicted results should be strictly applicable to these identical systems.…”

“…However, CFD can be used to study the hydrodynamic conditions, heat and mass transfers in MD modules for optimisation purpose [ 25 , 26 ]. Mathematical statics approaches, such as response surface methodology (RSM) [ 27 , 28 , 29 ]. RSM is a collection of mathematical and statistical methods based on fitting a polynomial equation to the experimental data and is able to predict the behaviour of a data set with the objective of making statistical previsions [ 30 ].…”

A general optimal operating strategy for commercial membrane distillation facilities

“…However, not only the temperature affects the performance of MD modules but also the feed water flow rate [26]. The optimal management of this variable is specially relevant because a tradeoff solution must be adopted to maximize both thermal efficiency and distillate production in current commercial-scale MD modules [19,27], thus requiring properly formulated optimization problems.…”

Modelado y control automático en destilación por membranas solar: fundamentos y propuestas para su desarrollo tecnológico

“…En la mayoría de ellos se utiliza la metodología RSM para predecir el flujo de destilado, como es el caso de Fadhil et al(2019); Elzahaby et al (2016); Khalifa and Lawal (2016); Bouguecha et al (2016); Mohammadi et al (2015); Boubakri et al (2014); Khayet and Matsuura (2011); Khayet and Cojocaru(2012a);Khayet et al (2007). Por el contrario, solo unos pocos predicen aparte del flujo del destilado algúníndice de desempeño relacionado con el consumo térmico del módulo como el STEC o el GOR(Gil et al, 2018c;Ruiz-Aguirre et al, 2018;Cheng et al, 2018;Ruiz-Aguirre et al, 2017;He et al, 2014). Además, en casi todos los trabajos se utilizan como variables de entrada la temperatura y caudal del agua de alimentación y la temperatura a la entrada del canal de evaporación del módulo, y solo enGil et al (2018c);Mohammadi et al (2015);Khayet et al (2007) se incluye la salinidad del agua de alimentación como entrada del modelo.…”