13.1.1 Nanofiltration membranes as a water treatment solution

Nanofiltration membranes are defined as having a pore size in the order of nanometers (nm) (1×10⁻⁹ m). As a comparison, the atomic radius of a sodium ion and a chlorine ion is about 0.97 nm (0.97×10⁻⁹ m) and 1.8 nm (1.8×10⁻⁹ m), respectively. This demonstrates that nanofiltration membranes are near the range to remove rather small ions.

However, the term nanofiltration is really a misnomer. As the nanofiltration membranes are charged, the removal mechanism is not purely filtration as with ultrafiltration membranes, but also osmotic. This makes them a true hybrid, bridging ultrafiltration and reverse osmosis membranes in the range of membrane treatment options.

In general, the primary factors that affect the performance of the membranes include the membrane material (charge of the membrane), concentration polarization at the membrane face (buildup of concentration at the membrane face), and fouling of the membrane to name a few. As such, pore size alone does not predict the removal of constituents. Adding more complication to the problem, every manufacturer’s membranes are slightly different, meaning there is no simple method for predicting removals. Pilot testing of the nanofiltration membranes is imperative when designing a system to target certain constituents.

Membrane manufacturers do discuss ranges for constituent removal based on atomic mass. Most indicate that nanofiltration membranes remove compounds/ions with a molecular weight greater than 300–400 g/mol. It should be noted that this number represents the size for complete (or nearly so) removal. For partial removals, the range extends down to less than 100 g/mol. This is evidenced by a figure published by Koch Membranes on their website that shows the range of nanofiltration membranes from 100 to 20,000 g/mol. In addition, testing completed by the Long Beach Water Department (Long Beach, California) confirms the smaller molecular weight as they have shown significant removal (up to 90 percent) of aqueous salts, which range in molecular weight from about 60 to 500 g/mol. With this range of removals, nanofiltration membranes provide engineers and scientists with a unique opportunity to produce customized water products.

Currently, there is a wide range of nanofiltration membranes available in the market, providing numerous options for targeting mass removals of specific or groups of constituents. Although current reverse osmosis technology can remove a larger mass of the same constituents as compared to nanofiltration membranes, this may not be an ideal solution. Reverse osmosis membranes remove not only the constituents of concern, but also dissolved minerals and hardness to such an extent that the product water is aggressive toward metal and concrete conduits. Treatment systems must then add some of these minerals and hardness back. The result is “wasted” effort to overshoot the water quality objective and then bring it back to a level that can be used.

With new breakthroughs in membrane technology, constituent specific membranes are a possibility. Membrane companies have developed nanofiltration membranes capable of 95 percent salt removal, increasing from the 20 percent that the early nanofiltration membranes had. In addition, research in the use of embedded “nanoparticles” in reverse osmosis membranes has shown promise in decreasing the transmembrane pressure, thus increasing the flux. This type of technology may provide similar improvements to nanofiltration flux and removal efficiencies.

13.1.2 Nanofiltration of freshwater sources

Current treatment of freshwater sources has focused on removal of particles, microbes, and viruses. As much of our potable water sources double as our drainage systems (stormwater, agricultural, and wastewater effluents), more concern is being raised about pollutants such as pesticides, heavy metals, endocrine disrupters, and pharmaceuticals. Conventional treatment methods do not remove these types of constituents at appreciable levels. Membrane technologies can remove all of these to very low levels. As discussed earlier, reverse osmosis membranes can remove these, but will generally overshoot the goals and require significant chemical stabilization downstream. Nanofiltration membranes can achieve the same results with a potentially significant increase in process efficiency. Table 13.1 shows some of the constituents of concern for water systems and their molecular weights.

Membrane manufacturers produce a variety of nanofiltration membranes that target different molecular weights. As an example, Dow Filmtec offers nanofiltration membranes to target approximately 90, 200, and 270 g/mol. This provides design engineers with many options for single pass applications. However, when one includes multiple pass applications, the options are endless.

Consider the following scenario as an example: A constituent of concern (Compound A) needs to be removed to meet drinking water standards. Based on testing of the source water, 85 percent (by mass) of Compound A needs to be removed to meet the requirements. Pilot testing is conducted resulting in the following single pass removals: reverse osmosis membrane, 99.9 percent; nanofiltration membrane no. 1, 75 percent; nanofiltration membrane no. 2, 61 percent; nanofiltration membrane no. 3, 40 percent. Based on this type of result the options include single pass reverse osmosis or the two pass combination of nanofiltration membranes as shown in Fig. 13.1.

As can be seen, there are a large number of combinations that can be made to provide the required results. Pilot testing of these combinations would allow for the determination of the most consistent, energy efficient, and easily maintainable configuration, thus providing the best system for the application.

13.1.3 Nanofiltration for seawater desalination

Of all membrane applications for purifying water, seawater desalination is the most complex. In recent years, many communities have constructed seawater desalination facilities using conventional seawater reverse osmosis membrane processes. The pressures required for desalination using these conventional reverse osmosis membranes are extreme—55 to 83 bars. In addition, the raw seawater and brine are highly corrosive requiring extensive use of austenitic and duplex stainless steels. The result is an energy consumptive, maintenance intensive, and costly process.

One community, however, decided to take a different proactive approach toward supplying their future water needs. The Long Beach Water Department in Long Beach, California, began a testing program in order to identify potential methods and processes to simplify and reduce the cost of seawater desalination. As a result of this testing, they developed a novel dual-stage (or pass) process using nanofiltration membranes.

With the support of their board of directors, the Long Beach Water Department staff built and tested a pilot scale plant with a capacity of 34 m³/day. The results of this testing proved the concept and showed significant potential.

As significant funds for research are generally not available in the municipal setting, Long Beach Water Department entered into a partnership with the United States Bureau of Reclamation and the Los Angeles Department of Water and Power to continue the research through the construction and operation of a 1,136 m³/day prototype (side by side with a reverse osmosis system) testing facility. Having been completed in 2007, the preliminary results at the facility have shown some potential advantages to using a dual-stage nanofiltration process over a traditional reverse osmosis process. These include the following:

More information on the results and progress of the Long Beach Water Department testing can be obtained from their website (www.lbwater.org), from the American Water Works Association Research Foundation, or the United States Bureau of Reclamation.

13.2.1 Nanofiltration for wastewater treatment and reuse

Wastewater treatment for discharge or reuse has a different set of issues than does drinking water treatment. Historically, wastewater treatment has focused on removal of solids, organics, and microorganisms. In the near future, regulations may no longer be limited to just these types of contaminants. As an example, the United States Environmental Protection Agency is now implementing Total Maximum Daily Loads (TMDL) for water bodies within the United States. The TMDL studies have resulted in more limits for compounds such as ammonia, nitrates, and nitrites. In the future it is possible that further study may lead to limits for contaminants such as heavy metals, pharmaceuticals, and endocrine disrupters, just to name a few. Table 13.2 lists the sizes of some of the contaminants of concern for wastewater. Although there is some improvement that can be achieved through source control prior to waste discharge, it is likely that much of these compounds will need to be removed at the wastewater treatment plant. These types of limits may not be restricted to water that will be reused, but also for water that is discharged.

As with drinking water, reverse osmosis is an option for meeting these increasingly stringent requirements. However, there are several problems with this approach. First, it is likely that reverse osmosis will overshoot the requirements for significant stabilization, not just for protection of pipes and channels, but also for protection of habitat. Second, it is very costly, both in terms of capital and operation. Producing such extremely high quality water for discharge can be considered by some to be inefficient. However, use of this type of water for replenishment of potable sources, although technically feasible, has proven to be difficult from a public perception standpoint.

Ultrafiltration membranes provide another alternative and are being used at many wastewater treatment plants to meet the tertiary filtration requirements. Ultrafiltration membranes do not, however, help much with any but the very large contaminants as they do not have any charge and thus osmotic removal capabilities.

As with water treatment, the broad range of nanofiltration membranes available, all with different characteristics, can provide a much better set of solutions to meet the more stringent discharge requirements. In single or multiple passes, nanofiltration membranes can be configured to target certain contaminants more efficiently than reverse osmosis. In addition, this type of configuration can prove to be the ultimate in flexibility to provide designer waters for reuse customers. Consider the following example: A wastewater treatment plant has installed a “loose” nanofiltration membrane (target for constituents at 270 g/mol) to meet effluent and reuse requirements. This level of water quality is sufficient for all of their reuse customers, except for two that need enhanced hardness removal to prevent scaling. As a solution, the wastewater treatment plant could install a second pass nanofiltration membrane unit for the reuse portion of the effluent stream or could install small versions of these units at the point of delivery.

In general terms, this flexibility will result in cost savings and improved efficiency. In the future, this increased efficiency will allow more municipalities to provide cleaner effluents and reuse larger quantities for a palatable cost.