Process intensification:integrated membrane operations for brackish and seawater desalination

Abstract

The present research study is focusing on the evaluation of the integrated membrane system which merges the membrane contactor technology such as gas-liquid membrane contactors (GLMC) and membrane distillation /crystallization (MD/MDC) with the conventional pressure-driven membrane operations such as micrfiltration/ultrafiltration (MF/UF), nanofiltration (NF) and reverses osmosis (RO) within the logic of Process Intensification (PI) strategies in order to redesign the desalination plants to be cheaper, safer and sustainable. The importance of applying the PI strategies in the desalination industry is presented in chapter 1. In addition, this chapter gives the research project objectives and activities. The optimization and the feasibility of using the GLMC in the proposed integrated membrane system were discussed in chapter 2. Simulation model for the GLMC was implemented by computer and the results were verified by experimental tests. The results showed that there was a good agreement between the simulation and experimental results with less than 10% differences. In terms of CO2 transfer rate, the results showed that higher transfer rates were obtained at higher liquid flow rates and higher pH values due to lower mass transfer resistance and higher reaction rates, respectively. The feasibly study showed that using GLMC is more economically feasible since the cost of the NaOH used in the GLMC after reacting with CO2 to produce Na2CO3 was less than the cost of using Na2CO3 directly from the market in order to precipitate Ca+2 as CaCO3. Moreover, the GLMC will contribute to the reduction of CO2 emission from desalination plants and reduce their environmental impact. Since so far there are no membrane modules especially made for MD, the aim of this study was to provide optimization guidelines for materials and methods for using MD in desalination. Therefore, in chapter 3, comprehensive theoretical analysis have been carried out and simulation model was developed to describe the mass flux and heat efficiency in MD processes considering transport phenomena, membrane structural properties and most sensitive process parameters, with the aim to investigate the effects of the membrane properties on the MD performance and to set some criterions to optimize these properties in order to obtain the best performance. Experimental tests were conducted in order to validate the results obtained by the computer simulation and the results showed that the computer simulations were able to estimate the MD performance with errors not exceeding 5%. The results showed that an increase of the temperature gradient resulted in the enhancement of both transmembrane flux and thermal efficiency. On the other hand, feed concentration had low effects in flux reduction even at high values close to saturation which contribute to only 30-50% flux reduction. This makes the MD ii process attractive technique for seawater desalination especially when integrated with RO in the logic of the ZLD concept and satisfying the process intensification goals. The investigation of the effects of membrane properties confirmed that better MD performance was achieved when using polymeric membranes characterized by low thermal conductivity (flux and thermal efficiency declined by 26% and 50%, respectively, when increasing thermal conductivity from 0.1 to 0.5 W/m K), lower thickness (increasing the membrane thickness from 0.25 to 1.55 mm resulted in a flux decay of about 70% without a significant improvement in thermal efficiency), and high porosity. The investigation of the complex correlations between physico-chemical properties of the membrane and MD performance confirms the need for a customized hardware, i.e. high porosity hydrophobic membranes with appropriate thickness and made by low-heat conductive polymers in order to reduce the amount of wasted energy. The basic mechanisms and kinetics of crystallization were considered in chapter 4 in order to accomplish the modeling and simulation for the membrane crystallizers. The computer simulation of the MDC was similar to the one of the MD presented in chapter 3 with addition of crystallization kinetics calculation. The simulation model was used in parallel with the experimental tests in order to improve the design and performance of the crystallizers. The results showed that it was possible to obtain NaCl crystals from the NF retentate at a good quality and narrow crystal size distribution (CSD). The effects of the concentration polarization in the transmembrane flux were very limited; however, there was an unexpected flux decline after the formation of the crystals in the system. This was due to the deposition of the salts crystals on the membrane surface which caused pore blockage and hence flux drop. The design improvement of the MDC suggested to introduce another opening at the bottom of the crystallizer tank for removing crystals, and to install a filter in the suction side of the feed pump in order to avoid crystals for recirculation inside the membrane module with the feed. Exergy analysis, economical investigation and sensitivity study were carried out in chapter 5 to evaluate the feasibility of the integrated membrane system. The exergy analysis showed that the highest work input was for the plant which involved the pressure-driven membranes UF-NF-RO due to the high pumping and pressurizing energy requirement especially in NF and RO pumps. On the other hand, the highest heat energy input was associated with the membrane distillation plant as a stand alone process. The exergy efficiency was generally higher in case of pressure-driven operations than thermal processes. In addition, the performance of plants with energy and heat recovery systems was always better than the ones without energy and heat recovery systems. Economical study and cost evaluation for several configurations showed that the lowest total water costs were 0.51 and 0.29 $/m3 when using UF-RO plant with energy recovery system for seawater and brackish water desalination, respectively. In case of the integrated system which contained both pressure and thermal processes, the best combination was obtained when using the pressuredriven membranes combined with a membrane crystallization unit operating on the NF concentrated stream and a membrane distillation unit operating on the RO brine stream. The total water cost in this case was 1.27 $/m3 and 1.10 $/m3 for seawater and brackish water, respectively. Moreover, the combination of membrane crystallization units is very attractive especially if the salt crystals produced by the crystallization process are considered. This means that the desalination plant will produce both water and salt crystals. In this case, the price of the salts can cover the whole expenses of the desalination process. Besides, the problems related to brine disposal were minimized when using the integrated membrane system. The sensitivity analysis revealed that the pressure-driven membrane operations were very sensitive to the feed concentration and the cost of electricity. On the other hand, MD processes were not sensitive to the variation on the feed concentration or the electricity costs. The most sensitive parameter in the total water cost of the MD plant was the cost of steam which contributed to values as high as high as 11.4% in case of MD without heat recovery system. The best tolerance to the variation of these parameters was obtained when using the integrated membrane system of pressure-driven membranes and MD/MDC processes. The realization of the semi-pilot plant of the integrated membrane system was covered in chapter 6. The semi-pilot plant of the integrated membrane system was designed and assembled based on the results obtained by the computer simulations and the preliminary experiments done for each unit individually in the previous chapters. It consisted of UF-NF-RO as the pressuredriven membrane operations with the GLMC for Ca+2 precipitation and an MDC unit which can be operated as an MD or as a membrane crystallizer. The semipilot desalination plant of the integrated membrane system was operated using synthetic and real seawater in order to confirm the performance and process stability. The transmembrane flux was stable during the operation. The MDC was able to produce salt crystals from the NF retentate and the RO brine streams. The CSD of the crystals obtained by the MDC operating on the RO brine showed sharper distribution trends than the ones obtained from the MDC when operating on the NF retentate. In addition, the MD unit was operated as a standalone desalination process using real seawater and the results showed that it was stable and the membrane did not loss its hydrophobicity during the operation