Vacuum multi effect membrane distillation (VMEMD) is a thermal separation process with microporous hydrophobic membranes that is able to concentrate liquids with high salt concentrations close to saturation. This document describes the technology, test runs with high concentrated salt solution, observed phenomena and experiences gathered from the testing with VMEMD modules by memsys. Salt influences the distillation process in direction to lower distillate output due to boiling point elevation of the salt water. The process proofs reliability in a 365 days long-term testing with feed at conductivities between 130′000 μS/cm and 240′000 μS/cm (approximately 9.5–22wt% salt in solution) and distillate output with less than 10 μS/cm. The testing program delivers experience for the layout of modules with high recovery ratio.
Concentrated brine from reverse osmosis (RO) or other desalination plants can not always be returned into the sea due to regulations of environmental protection or for economic reasons and the disposal of low concentrated waste waters is uneconomical (Xiaosheng et al. 2010). Produced water or flow back water from conventional or unconventional oil and gas production contain often very high salt concentrations. The disposal of these liquids is often regulated by law, what means that it is transported by pipelines or trucks from bore holes to disposal sites where the water is dumped into pits or deep wells. The efforts may even rise in future if governments adopt more stringent regulations and if there is a limited availability of disposal sites (Cooley H. & Donnelly K. (2014)). These are reasons for the further development of technologies for brine concentration and zero liquid discharge (ZLD).
Physical boundaries limit the applicability of RO at high salt concentration and conventional evaporators with metal surface heat exchangers are expensive and endangered by corrosion. The technology membrane distillation (MD) offers a possibility to overcome these disadvantages. It is applicable for high salt concentrations and other thermal separation challenges (Tomaszewska 2000). MD is a separation process based on evaporation with microporous hydrophobic membranes. A liquid feed solution is in direct contact with the membrane and partially evaporated due to a negative vapor pressure difference over the membrane. The company memsys developed and patented a thermal separation process based on MD called vacuum multi effect membrane distillation (VMEMD) (Heinzl 2010). The process is comparable with multi effect distillation (MED) however flat sheet membranes are used for evaporation and thin polypropylene foils for condensation. Owing to the selection of plastic materials, VMEMD modules are resistant to corrosive media. The modules have a light weight, are easily transportable and produced in an automated production process with many equal parts what keeps investment costs low. High thermal efficiency of the process is reached by internal energy recovery. The distillation process is located in a compact casing: a total surface of approximately 60 m² for each, evaporation and condensation, is provided per cubic meter module volume.
There is a wide range of scientific papers available about small laboratory MD setups which treat high concentrated salt solutions with the result that MD is able to handle this task and most of the authors have developed model equations for heat and mass transfer inside their setups (Gryta (2001), Mericq et al. (2009), Nghiem et al. (2011), Ramon et al. (2009), Weckesser (2008), Yun et al. (2006)). This fact highlights the relevance of the technology from a scientific point of view.
Main objective of this document is to provide information about experiences at the treatment of high concentrated salt water solutions with the commercially available VMEMD technology of memsys. Several modules are operated at equal conditions with pure water and salt water to demonstrate the influence of salt on the distillation process. A long-term test run is shown in which a module is operated for more than 365 days with high concentrated feed solution. Finally, the key experiences from the salt testing are summarized.
TECHNOLOGY AND METHODS
The driving heat for the process is transferred to the first effect by a heating system and is used for partial evaporation of a feed liquid mixture of components with different boiling temperatures. In case of saltwater being the feed, water is evaporated, condensed and gathered as distillate whereas the salt remains in the brine. The usage of multiple effects by condensing the generated vapor and reusing the latent heat for once more evaporation at a lower temperature and pressure level in each effect increases the system efficiency as thermal energy that is once brought into the system can be reused several times depending on the available number of effects. Beside thermal energy, VMEMD needs auxiliary electrical energy for its controlling, pumps, sensors and valves.
VMEMD unit setup
Feed is sucked into the module and separated into distillate and brine. Pumps are required to transport the product liquids brine and distillate from low module pressure to ambient pressure. A heating loop is driven by a pump that transports a heating liquid through a heat exchanger and through a steam raiser, the internal heating system of the module, where the heating liquid is partially evaporated through membranes. The generated steam condenses in the first effect and is fed back into the heating loop by a recirculation pump. This system is used to decouple the low module pressure from potential pressurized external heat sources like solar thermal collectors or waste heat of industrial processes or engines in a temperature range between 50 and 90 °C. On the other module side, a cooling loop provides cold temperatures which are necessary to generate the driving temperature difference for the process. Its main component is an internal condenser in which the vapor of the last effect is condensed. The cooling loop works analogously to the heating system without recirculation. A VMEMD unit requires a vacuum system to keep the module pressure low and non-condensable gases out of the process. Modules can also be operated with a thermal or mechanical vapor compressor.
Equipment for test runs
All mentioned test runs of this document have been performed with units that have the described kind of setup and a module which consists of a steam raiser, two or four effects and a condenser. The total internal membrane area of the two modules is 2.6 m² (2 effects) and 6.4 m² (4 effects). There are integrated sensors to measure temperatures, pressures and volume flows of the module.
Units for salt concentration in water like gram per liter water (g/l), gram per liter salt solution (g/l) and gram per kilogram solution (g/kg) are often mixed up as its difference is small at small concentrations and ambient temperature but there is a significant difference at high concentrations. In this document, concentrations of the used sodium chloride water solutions are determined with a conductivity measurement cell with temperature correction to 25 °C. Water is saturated with NaCl at approximately 254 mS/cm what corresponds to 260 g NaCl per kg solution or 26wt% in the temperature range of the conducted experiments as the solubility of NaCl in water does not show strong temperature dependence (M. W. Kellogg (1976)).
RESULTS AND DISCUSSION
Comparison of module operation with pure water and salt solution
The highest flux of 9.2 kg/(m²h) is measured at 80 °C heating temperature, 100 l/h feed volume flow and pure water as feed. It can be read from the diagram that the distillate output is lower the more salt is dissolved in the feed, what can be reduced to the effect of boiling point elevation (BPE), what means that salt water starts boiling at a higher temperature at constant pressure or at lower pressure at constant temperature compared to pure water. BPE is dependent on temperature level and salt concentration and must be considered for module layout calculations as at all other evaporative desalination technologies too (Gebel & Yüce 2008).
‘Vapor 1’ is used to heat the first effect with latent heat as it is described in Figure 1. The heat flow brought into the module was reduced from 4.8 to 3.2 kW and the specific distillate output per membrane area decreased from 4.5 to 2.4 kg/(m²h) at operation with salt water. The heat and mass transfer dependent process is dependent on driving temperature and pressure gradients which are decreased the more salt is dissolved in the feed. Higher distillate output can be achieved by a higher heating or lower cooling temperature for the module.
Long-term operation of a VMEMD module with high salt concentration
Heating and cooling temperatures, feed volume flow and concentrations can be adjusted as input variables for distillation operation. The temperature difference between hot and cold side determines the resulting flux of the module which was initially not designed for high salt concentrations. The output brine concentration is a result of input feed flow, feed concentration and the distillate output.
The testing can be divided in three main sections: the first section started with high concentration tests in which it was examined how high the input feed concentrations for this specific module can be. This interval was only interrupted due to summer holiday season. The module was operated with a maximum of 22wt% feed and produced a 24wt% brine (saturation at 26wt%) in this section. After the first section was stopped for more than 1 month because of reconfiguration works at the external heating system in the laboratory, the second section with some more defined testing at 10wt% feed concentration started. This section was only interrupted by Christmas time. The third section is an interval with 15wt% NaCl in solution as feed. The module was able to produce distillate at less than 10 μS/cm conductivity over the whole testing time. The maximal adjusted flux was 7 kg/(m²h).
Key experiences from salt water testing
The long-term test demonstrates that VMEMD is a reliable technology to concentrate saline waters and many valuable experiences could be gathered within the testing. Even the small two effect module that was initially not designed for salt testing can treat high concentrated feed. Very high concentrations close to saturation were tried to keep the effort for a final crystallizer in case of a ZLD application as low as possible. Within these trials, the brine concentration ended up in saturation many times what happened carefully controlled or even unintended due to handling errors. It was found that membranes can partially be wetted due to salt crystals at saturated feed. This phenomenon is also described in literature (Gryta (2001), Weckesser (2008)). In case that wetting has happened, salt liquid is not mixed with distillate within a VMEMD module and the distillate quality remains high. An in-situ flushing and drying procedure was developed for VMEMD to recover membranes completely without the need to disassemble the module. Extensive studies and test runs provide a good knowledge base to avoid this concentration polarization induced phenomenon by balancing feed volume flow, flux and salt concentration within a module and to design modules for high concentration factors.
The VMEMD technology of memsys is capable to face the challenge of brine concentration in direction to ZLD. It is a vapor pressure and temperature difference driven process what means that the flux, the specific distillate output per membrane area, shows a strong dependence of the available temperature difference of hot and cold side of the module. Compared to pure water the distillate output is reduced at operation with saline water as at every desalination technology. This reduction due to BPE can be overcome by adjusting a higher temperature difference for the module. An ongoing continuous long-term test run of more than 365 days duration with various high salt concentrated feed and distillate at conductivity below 10 μS/cm proofs the long-term reliability of the process. It was found that saturation of dissolved salts must be avoided however lots of experience could be gathered from the salt testing to design and operate modules with very high RRs close to saturation by balancing feed flow, flux and concentration inside a module.