Largely due to the advances in membrane separation processes, the price for desalinated water has decreased significantly over the last few years to less then $1 per cubic metre in production costs for large-scale desalination operations. Modern low-energy membrane elements require as little as 3-4 kWh per cubic metre total energy for seawater desalination (Lattemann et al. 2009, Fritzmann et al. 2007). These developments drive water desalination projects to more remote locations, integrate seamlessly with renewable energy sources to supply its energy demand and allow highly flexible and mobile solutions to provide clean water anywhere, at any time. Desalination processes are on their way to adopt a level of maturity which has been championed by the internal combustion engine for decades: large scale diesel motors power ocean vessels or supply grid electricity, while small generators can also power a single households in case storm-related power outages. In the following, recent trends in the water desalination industry and research are highlighted, with a particular focus on membrane technology. Furthermore, the connection between energy systems and water desalination are showcased and how to realise synergistic effects by incorporating energy storage systems and approaches for brine utilization.
Traditional methods of water desalination, such as multi-stage flash and multi-effect distillation or Reverse Osmosis (RO) membrane processes, have been largely singular in terms of their energy requirements. Recent technological advances have succeeded through hybridization of energy sources, combining different sources and adjusting to local supply. State-of-the-art Membrane Distillation (MD) processes employ waste heat sources and electrical energy from renewable sources.
Membrane distillation (MD) is primarily a thermal process. In contrast to other distillation processes, it employs porous hydrophobic membranes to allow the permeation of vapour while preventing liquid water from flowing across the membrane. The advantage is that the feed stream is heated to medium temperatures significantly lower than the boiling point of water, which makes using waste heat or solar thermal energy as heat supply possible, and then brought in direct contact with the hydrophobic membrane surface. The hydrophobicity of the surface prevents the liquid to penetrate the pores but allows vapour to flow through the pores to the condensate side. The pores of MD membranes can be considerably larger than the pores of a RO membrane and thus allow higher flux at lower effective transmembrane pressures differences. It is crucial to the process that the membrane is impermeable for the feed solution under the given operating conditions. Membranes made from hydrophobic polymers such as polypropylene/polyethylene or its fluorinated equivalents can be used as membrane materials. Alternatively, surface coatings can also be employed to render hydrophobic pores on the feed side of the membrane. A lower vapour pressure on the condensate side needs to be generated by applying a vacuum or condensing the permeating vapour on a condenser sheet. Over the last ten years, improvements in surface chemistry, module design and flux decay have helped overcome energetic and economic uncertainties in MD desalination (El-Bourawi et al. 2006). By adopting sophisticated design strategies, heating and cooling efforts can be optimized and energy losses minimized. Recently, Membrane distillation (MD) processes have shown to be successful in treating highly contaminated industrial waste water streams, replacing energy intensive distillation processes in this field of application (The Economist, 30 Nov 2013).
Significant barriers of adoption for membrane processes have been associated with fouling, reducing the long term flux to a fraction of what is achieved during start-up. Better water pretreatment, innovative coatings of the feed spacer have mitigated these issues, but more significantly for mid and small-scale operations, using membrane distillation processes, and therefore a hybrid energy source of heat and electricity, seems to have given an edge to the hybrid technologies. This is the area in which Serene Enterprise Alliance partner SolarSpring, based in Freiburg, Germany, are specialised in.
Other arguably more exotic membrane desalination approaches include forward osmosis and electrodialysis, and even more sophisticated strategies stemming from a convergence with nanotech manufacturing, namely nanofluidic devices and graphene membranes.
Forward Osmosis (FO) suggests a connection to osmosis phenomena. As opposed to RO where the effect is reversed by applying a pressure, forward osmosis processes use an osmotic pressure gradient created by differences in concentration to trigger a flux. The osmotic pressure of the draw solution needs to be higher than that of the feed in order induce the net flux across the membrane. Subsequently, the draw solution needs to be treated, often by also using low-grade head, to produce a fresh water product stream.
Water desalination through Electrodialysis uses a very different type of membrane mechanism: ions are transported through the membrane instead of water molecules. The feed stream is sweeping a stack of cation and anion exchange membranes, sandwiched between a cathode and an anode, which generate an electric field within the stream. As ions cross through the stack of ion exchange membranes, enriching the concentrate stream in ions, they are depleted from another, the purified stream. The primary energy requirement for this desalination approach is electricity. Advances in micro- and nanotechnology pave the way for further innovation in the field of membrane desalination through electrodialysis, such as building membranes that generate concentration polarization near an ion exchange membrane which enhances the ion depletion. Moreover, using different effective sizes of anion and cation exchange membranes leads to different current density levels of the two membranes, further enhancing the effect. Using this synthesis of membrane electrodialysis and concentration polarization creates the additional benefit of purifying the product water stream of non-ionic pollutants.
Recent years have also seen the development of Nanofluidic Devices for seawater desalination, similarly exploiting ion concentration polarization to an impressive degree of efficiency. The efficiency of the process is reported to be at only about 5 kWh/m³, which is remarkable for a small lab-scale device (Kim et al. 2010). These results hint towards applications where small but very robust and versatile processes are required. Research endeavours are multi-facetted and more significant advances in membrane technology will be seen in the future, with some as much futuristic as optimistic agents envisioning a single molecular layer making up the membrane (Lockheed Martin Corp. 2013).
As most industrial activities water desalination has environmental implications and the size of its footprint depends on a variety of factors, ranging from energy expenditure, carbon emissions to the accumulation of waste streams. One unavoidable feature is the production of concentrated brine as a product of the desalination process. The salinity of brine varies depending on the desalination process. The recovery ratio, that is the ratio of feed water to each unit of product water, is the decisive measure determining the salinity of the brine. Processes with high recovery ratios produce more concentrated brine. Desalination plants often produce brines that contain as much as 70 000 ppm total dissolved solids (Ahmed et al. 2000). The highly concentrated brine disposal needs to be managed, often by discharge into surface water bodies, which constitutes the least expensive disposal method today. The discharge of concentrated brine into the open sea may significantly affect marine environments. Meanwhile, a significant amount of energy has been expended to separate salt water into drinking water and concentrated brine. From a thermodynamic vantage point, the osmotic pressure of seawater against fresh water is about 28 bar. The mechanical energy required to pump water through a semi-permeable membrane is at least 2780 kJ per cubic metre, or 0.77 kWh per cubic metre. This is the theoretical limit of how much energy is required to desalinate a cubic metre of sea water. Losses of mechanical and electrical nature are rendering real desalination processes considerable less efficient. As opposed to handling concentrated brine as harmful waste, it should be utilised as a useful resource.
A handful of sophisticated ways to reverse the desalination process and recuperate some of the energy exist. One such way to make use of the brine is via Pressure Retarded Osmosis (PRO). PRO membranes are closely related to FO membranes, because there is a net flux of water from a more dilute stream into a more concentrated stream across PRO membranes. The technique was conceived in the 1970s, closely related to the development of RO membranes, but attempts for commercialization took until the 2000s (Statkraft 2009). The hydraulic head generated from osmosis is astounding, which hints to the amount of energy stored in concentration gradients, dwarfing the pressure head of most hydroelectric dams. The limiting factor is the water flux as it is highly correlated with the membrane area. The hydraulic head from concentrated brine to sea water is not as high as that between fresh water and salt water, but it is still considerable, in the order of magnitude of 10 bar. Finding the optimal working pressure of the process, which would in fact be lower, is the key challenge as well as fitting the suitable power generation equipment around the PRO process. A high flux is crucial as the amount of power generated depends on the hydraulic head generated by the osmotic pressure as well as the volume flux. By increasing the membrane area, the power production can be scaled to the desired power requirements, but it is also the limiting cost factor.
Reversed electrodialysis is an attempt to generate an electrical current directly from two streams of water of different salinity. A stack of alternating cathode and anode exchange membranes use the gradient in ion concentration to generate a voltage of 80 mV across one single membrane stack (Nijmeijer et al. 2010). By connecting membranes in series, the potential differences across all membranes can be summed to a large voltage. Similarly using the energy captured in a concentration gradient can be used to desalinate more water. Very innovative approaches use a double stack of electrodialysis membranes and the concentrated saline solution to deplete a product water stream (Saltworks Technologies 2012).
Given the vast amount of research and technological development in the area of membrane desalination, desalination processes will become more efficient and more flexible in terms of their energetic, infrastructure and capital requirements. The immediate future holds extensive potential for extending the processing capabilities of membrane-distillation solutions by combining it with thermal storage and electric battery systems. Since the turn of the millennium, the renewable energy industry has made leaps in technology and especially electricity generation from photovoltaic panels has seen unprecedented cost reductions. This opens further opportunities for desalination plants to become less dependent on external energy sources and acts as buffers for electricity grids as well as consume and produce energy according to local demand. The advantages this can bring to remote arid regions, which will benefit most from differentiating approaches to providing energy requirements for solar desalination solutions. Likewise, mobile and flexible solutions for on-demand water treatment, especially in situations of need, are dire.
By Hendrik Frentrup, Ph.D Candidate, Imperial College London
S. Lattemann et al. “Global Desalination Situation” in Sustainability Science and Engineering - Sustainable Water for the Future (eds. I. C. Escobar and
A. I. Schäfer) Vol. 2 Elsevier (2010)
C. Fritzmann et al. “State-of-the-art of reverse osmosis desalination” Desalination 216 (2007) 1-76
M. S. El-Bourawi et al. “A framework for better understanding membrane distillation separation process” Journal of Membrane Science 285 (2006) 4–29
The Economist “Clean that up” 30 November 2013
S. J. Kim et al. “Direct seawater desalination by ion concentration polarization” Nature Nanotechnology 5 (2010) 297-302
Lockheed Martin Corporation “Perforated graphene deionization or desalination” Patent No. US 8361321 B2 (2013)
Ahmed et al. “Use of evaporation ponds for brine disposal in desalination plants” Desalination 130 (2000) 155-168
Statkraft Development AS “Use of a semi-permeable membrane in pressure retarded osmosis for providing electric power, and plant” Patent No. EP 1315554 B1 (2010)
K. Nijmeijer et al. “Salinity Gradient Energy” in Sustainability Science and Engineering - Sustainable Water for the Future (eds. I. C. Escobar and A. I. Schäfer) Vol. 2 Elsevier (2010)
Saltwork Technologies Inc. “Method And System For Desalinating Saltwater Using Concentration Difference Energy” Patent No. US 20120279861 A1
Figure 2: Reverse electrodialysis employs cation and anion exchange membranes (CEM/AEM) to generate a current as water streams of different ion concentrations sweep the membrane.
Figure 1: The energy from the concentration gradient between sea water and brine, exhibited in the hydraulic head H, can be recuperated via a pressure retarded osmosis process.