Project Body:


California’s current population of 35 million is expected to increase by approximately 12 million by the year 2030, which will impact the state’s water demands significantly (Karajeh et al 2005). Agricultural irrigation represents a considerable portion of fresh water demand, with an estimated 65% global water demand and nearly 90% of the water demand in California (Abu-Zeid 1998). This creates a challenge for water supply reliability and availability as technologies shift from the construction of new dams, reservoirs and conveyance canals, and move toward water conservation and reclamation. As a result, saltwater and brackish water reverse osmosis desalination is becoming of greater interest, largely because technological advances have caused the cost of membranes to decrease dramatically (Karajeh et al 2005). In the same time that desalination costs have been declining, the costs of surface and groundwater have been increasing, making desalination a more competitive source of water for both municipal and agricultural purposes (Beltrán and Koo-Oshima 2004).

A study by Sorour et al (1992) investigated various desalination technologies for agricultural drainage water applications. The study ranked reverse osmosis (RO) desalination as the highest performing desalting technology, when compared to ion exchange, electrodialysis and vapor compression. This high ranking was due to lower desalting costs, higher tolerance to changes in salinity, and ability to remove dissolved organics. Per the results of the study, reverse osmosis desalination became the recommended technology for agricultural drainage desalination (Sorour et al. 1992).

The process of reverse osmosis is, simply stated, the removal of contaminants by pushing water through a membrane with the use of hydraulic pressure (Solt and Shirley 1991). RO membranes are generally nonporous and will pass water, while retaining most solutes, including ions. The separation of salts and other minerals from the water is achieved by reversing the natural osmotic flow with the application of pressure to the side of the concentrated solution as illustrated in Figure 1 (Vigneswaran et al 2004).

Figure 1. Reverse osmosis principle (Fritzmann et al. 2007).


Reverse osmosis is one of many processes—including electrodialysis, ultra-filtration, micro-filtration, etc.—that can be used to purify water for a wide number of applications. Compared to other processes, RO has proven to be much more successful for water purification, and has thus become the most widely used desalting process for both seawater and brackish water sources. Depending on the specific membrane used, RO is able to remove 90-98% of all dissolved salts, organic molecules, microorganisms, colloids and suspended matter (Solt and Shirley 1991), and is capable of rejecting particles with diameters as small as 0.0001 µm (Taylor and Jacobs 1996).

Reverse osmosis membranes are engineered into a single operation unit, referred to as a module, of which there are several different types. Four major types of modules are produced: plate and frame, spiral wound, tubular and hollow fiber (see Figure 2). Plate and frame modules are

Figure 2. Schematic representation of four membrane modules: (a) plate and frame; (b) spiral module; (c) tubular module; (d) hollow fiber (Aptel and Buckley 1996).


composed of stacked flat-sheet membranes and support plates. These modules are designed to work similar to a filter press as they circulate feed water between the membranes of two contiguous plates. Spiral wound modules consist of two flat-sheet membranes enveloped and enclosing a flexible porous sheet, or permeate collector. The open end of the envelope is connected and rolled spirally around a perforated tube, where permeate is collected. One of the simplest reverse osmosis configurations is the tubular module, where a membrane is cast on the inside wall of a porous support tube. Permeate flows through the membrane and out of the module while the concentrate is flushed through the tubes. Finally, hollow fiber modules typically consist of bundles of several thousand (or up to several million) fibers through which feed water is conveyed (Aptel and Buckley 1996).

The transfer of water over the membrane in a reverse osmosis module is governed by Fick’s First Law of Diffusion (Equation 1), which provides that the diffusion flux across a membrane is

directly proportional to the concentration gradient, or (Vigneswaran et al 2004).



J = the diffusion flux

(amount of substance per unit area per unit time)

Di= the diffusion coefficient or diffusivity

φ = concentration

x = length

It can be seen that an increase in the concentration gradient would result in a decrease in the flow across a membrane (Vigneswaran et al 2004). Additionally, the pressure required by the RO system must be great enough to first overcome the osmotic potential pressure (caused by the difference in concentrations of the solutions on either side of the RO membrane), and then to drive the flow of feed water through the membrane. From this, it can be determined that the higher the concentration of the solution to be purified, the greater the osmotic potential; and the greater the osmotic pressure, the higher the pressure requirement will be. This relationship explains why the pressure requirement is lower for brackish waters, which have lower solute concentrations than seawater (Solt and Shirley 1991).

Fick’s First Law of Diffusion (Equation 1) also provides that the diffusion flux across a membrane is directly proportional to the surface area of the membrane. Therefore as the area of the membrane decreases, the diffusion across a membrane will decrease. By increasing the surface area of the filtration membrane, the diffusion across the membrane can be maximized (Vigneswaran et al 2004). The spiral wound RO modules (Figure 3) have become common due


to their expansive surface area which not only increases the diffusion across the membrane, but also provides easier access for cleaning agents (Taylor and Jacobs 1996).

Figure 3. Diagram of flow through spiral-wound reverse osmosis membrane module (Fritzmann et al. 2007).

Although spiral-wound reverse osmosis membranes are widely used and highly effective, they are unable to treat turbid feed water. The spacers implemented in these membrane modules make them more susceptible to clogging, or blocking, and pre-treatment of the feed water is required (Taylor and Jacobs 1996). If care is not taken to properly prevent dissolved, colloidal, or biological matter from entering RO modules, accumulation will occur that will inhibit diffusion across the membrane. There are two types of blockages, which are distinguished by the particulate build-up type. The first of these is scaling, which is characterized by the precipitation of inorganic material on a membrane surface. A super-saturation of concentrate on the feed side of a membrane can precipitate if not properly treated, and can dramatically reduce permeate flux. Scaling can be treated, and precipitate can be released from the membrane surface, but transporting the crystalline mud out of the module can be extremely difficult. Therefore, scaling should be avoided at all costs; antiscalants (pre-treatment) are available that adjust the pH of the feed water in an effort to prevent precipitation.

The second type of blockage is referred to as fouling, and is characterized by the accumulation of particulate matter or biological growth on a membrane surface. Mechanical pre-treatment of feed water for prevention of particulate fouling can be performed by use of screens, sand filtration, cartridge filtration, or membrane pre-treatment. Chemical pre-treatment of feed water is required for biological fouling, which, if left untreated, can produce a gel-like layer of microorganisms. This will drastically reduce the permeate flux through the module, and a chlorination of the feed water during pre-treatment is recommended. Of course, fouling can never be fully prevented, and build-up will occur. Regular maintenance and membrane cleaning can be performed to manage potential fouling problems (Fritzmann et al. 2007).

Useful Links:

Related Projects