Filtration Membranes for Drinking Water Applications
Any good overview of filtration membranes for drinking applications should include a summary of the advantages and drawbacks, some discussion of the current membranes market and technologies, and a crystal-ball view of what to expect in the near future.
By Jeff Neemann, Scott Freeman, Trevor Cooke, and Paul Delphos
Any good overview of filtration membranes for drinking applications should include a summary of the advantages and drawbacks, some discussion of the current membranes market and technologies, and a crystal-ball view of what to expect in the near future. Read on if you’re looking for this type of text, but don’t overlook the information about the move toward sustainable operations. Decision-makers increasingly will need to balance investment and capital costs against the benefits of greater treatment performance and operations and maintenance costs. Using membranes today requires a broader understanding of not only membranes but also specific application issues and trends.
Membrane filtration is typically used instead of conventional media filtration because it offers several advantages. Membrane filtration provides an additional barrier to pathogens, as evidenced by EPA’s Long-Term 2 Enhanced Surface Water Treatment Rule, which grants additional Cryptosporidium removal credit to membranes. Membranes also provide superior particulate removal compared to conventional media. Membrane filtration produces low-turbidity water at all times, even when the filter influent water contains a heavy load of particulates, whereas media filters can allow passage of more particulate matter under such conditions. Membranes also typically require a smaller footprint and are more amenable to automation and remote operation compared to media filters.
A potential drawback to membrane filtration is that when the membrane system influent particulate load increases, system performance decreases, possibly requiring more energy and/or more frequent cleaning for continued production of water. Although the quality of the water produced by a conventional media filter may decline, water will continue to be produced. Another drawback is that daily testing to monitor the integrity of the membrane is required in order to receive Cryptosporidium removal credit. In addition, the capital and operations and maintenance costs for a membranes system can be higher than those for a conventional media filter.
State of the Market
A number of factors have contributed to the growth of low-pressure membranes in the United States over the past 15 years. The first significant microfiltration (MF)/ultrafiltration (UF) water treatment plant began operation in 1993 in Saratoga, CA, for the San Jose Water Company (AWWA, 2005). The 3.6 million gallon per day (mgd) facility was nearly 4.5 times larger than any other MF/UF installation in operation at the time. The number of MF/UF installations in North America has continued to grow steadily since then, and today there are more than 250 MF/UF installations in operation, with a combined treatment capacity exceeding 1 billion gallons per day.
The Tuas Seawater Desalination Plant in Singapore is one of the largest and most energy-efficient seawater reverse osmosis facilities in the world. (Photo courtesy of Black & Veatch)
The continually decreasing cost of MF/UF systems has further expanded the potential size of membrane treatment facilities. While most of the initial facilities had relatively small capacities, typically less than 5 mgd, increasingly larger MF/UF facilities have been constructed since then. An example is Minneapolis, where the construction of a 70 mgd MF/UF facility was recently completed, and another MF/UF facility with a capacity of 90 mgd is scheduled to become operational by 2010. In 2004, the median capacity of new membrane facilities was 7.2 mgd whereas in 2000, it was only 0.9 mgd. Part of this increase is because MF/UF systems have become more cost-competitive with conventional water treatment technologies over a wide range of capacities.
According to one study, only approximately 2.5 percent of the drinking water in the United States is treated with MF/UF membranes (Mcllvaine Company, 2006). Therefore, this technology has significant market growth potential in the U.S. municipal drinking water industry alone. More importantly, the entire worldwide membrane market is predicted to grow from $7.6 billion in 2006 to over $10 billion in 2010, with a growth rate of 10 percent or greater for the foreseeable future. According to reports, 50 percent of the membrane market is associated with reverse osmosis (RO), which is used for desalination and producing ultrapure water for electronics, power, and pharmaceutical applications. The remaining portion of the market is split between MF and UF systems, including the rapid growth of membrane bioreactors for wastewater treatment.
According to some reports, the municipal sector component of the membrane industry was approximately $1 billion in 2006 (BCC, 2007), 15 percent of the entire market. Therefore, utilities considering membrane technology, particularly MF/UF systems, should take comfort in the vastness of the membrane marketplace and the associated membrane replacement market. This large market, of which municipal systems are only a small portion, provides assurance that MF/UF systems will be a viable and very well supported technology for many years to come.
Essentially all large-scale potable municipal plants use hollow-fiber membranes. There are two basic configurations – encased and submerged systems – in which the fibers are mounted inside pressure vessels or submerged in open tanks, respectively. Sometimes these are called pressurized or vacuum systems because the driving force for the encased type is provided by positive pressure on the feed side while the submerged systems is driven by negative pressure on the filtrate side.
Membrane filtration was selected for the Chaparral Water Treatment Plant in Scottsdale, AZ, for several reasons, including its high product water quality and small footprint. (Photo courtesy of Black & Veatch)
Although the early, mid-1990s installations were quite effective, manufacturers have been incrementally improving materials and other aspects of the products to yield greater resistance to fouling, increased chemical tolerance, and greater strength to prevent fiber breakage. Early membranes were made of cellulose or polypropylene, whereas polyvinylidene difluoride (PVDF) and polyether sulfone have recently become more popular as membrane materials. Membranes made of other materials, such as polyvinyl chloride (PVC) or nano-designed materials, are now being developed. Ceramic UF has been applied, mostly in Japan. While relatively expensive compared to PVDF and PVC materials, ceramic membranes offer potential benefits that include an extremely wide chemical and temperature tolerance that allows capacity maintenance with harsh cleaning regimes and reduces operating costs.
Move Toward Sustainable Operations
While membrane systems have been growing in both size and numbers, the industry has made them more cost-effective by pushing the envelope on the flux rate, (rate of flow per unit of membrane area). Higher flux rates mean less membrane area, which reduces the capital cost of the system. However, designs with high flux rates have made operation of the membranes more challenging and time consuming, and in some cases have even led to systems that cannot meet their design capacities during periods of degraded feed water quality. High flux rates often result in membranes that foul more rapidly and require more frequent cleaning by backwashing with water and/or air, chemically enhanced backwashing (CEBW), or chemical clean-in-place (CIP) procedures. More frequent cleaning results in lower productivity, because the membranes are taken off-line; lower recovery, because more permeate is used for backwashing; and reduced membrane life and increased fiber breakage rates, because the more frequent and aggressive chemical cleanings can damage the membrane fibers.
Changes in water quality can also make operation of the membranes challenging and time consuming. Most membrane systems intended for treatment plant applications have undergone some form of pilot testing, prior to design. However such testing is often of short duration – typically less than six months – during which the water quality encountered may or may not be representative of seasonal changes or even from year to year. As a result, a “design flux” may be selected that may not be successful if the feed water quality degrades more than experienced during the pilot testing. Once such a system is operational and experiences changes in water quality, the cleaning frequency must be increased, in some cases beyond the so-called design frequency, just to maintain the system production. So the flux rate for which the system was designed was not sustainable based on the quantity of water required and the quality of feed water.
What is Sustainable Flux?
Sustainable flux is the flux at which the system can be operated long-term at the design flow rate, with acceptable energy use, recovery, cleaning frequency, fiber breakage, and membrane life. In essence, the sustainable flux seeks to strike a balance between capital cost and the operation and maintenance costs. While the flux rate is not the only design parameter for the membrane system, it is critical, and when changed can change many other parameters.
So what is acceptable? Acceptable frequency may vary from one utility to another, but some of the acceptable values are:
- water recovery of 90 to 95 percent;
- backwashing interval of 20 to 30 minutes;
- CEBW interval of 24 to 48 hours; and
- CIP interval of 30 days.
Acceptable energy use, fiber breakage, and membrane life are more dependent on the specific membrane and manufacturer and are difficult to generalize, but acceptable values can be determined for a given project.
How do you select a sustainable flux?
Selecting a sustainable flux requires testing and taking a reasonable approach to the system design. Once acceptable operating frequencies have been established, they should not be used as the goals in testing; instead, testing should be based on more conservative values to take into account the uncertainties in water quality. For instance, if the acceptable CEBW interval is every 24 hours, the goal of the pilot testing should be to successfully maintain the flux rate with a CEBW interval of not more often than 48 or even 72 hours. If the acceptable CIP interval is 30 days, the pilot testing goal should be 33 days or longer. While these differences seem small, it is desirable to select a flux at which it is still possible to meet acceptable cleaning frequencies in degraded water quality. If an acceptable CEBW interval is 24 hours and the pilot testing used that interval, the operators would be unable to effectively respond to water quality changes. However, if the testing interval was 72 hours but the system was designed to be able to conduct a CEBW every 24 hours, the operators would be able to increase the frequency to address water quality changes.
These concepts can also be carried over to the design of the system by providing spare space for additional future modules (which reduces the average flux), considering down time for cleaning when selecting the number of standby trains or skids/racks, or even considering and limiting the instantaneous flux rate instead of only the average or net flux rate. Another valuable tool that can be included during the design with minimal cost impact is programming to allow the membrane trains to be operated at different flow rates or flux values. Because most systems are not operated at the design capacity year round, this type of programming makes it possible to conduct a long-term test on one of the trains at the design flux to establish the appropriate operating procedures before challenges happen, not when the water is really needed.
What if conditions are better than expected?
While the concepts discussed above may be considered to be overly conservative by some, they present a more reasonable approach to the design of the membrane system to ensure that all expectations can be met. As with all design there is risk; in this case the risk is that the system will not produce the required quantity of water or that the cost to produce that quantity will be higher than anticipated. So what if both the conditions and the water quality are better than expected? The design should account for this condition by including a hydraulic overload factor of 1.3 to 1.5 times the design value. This means that the facility could produce more water than anticipated or that future expansions can cost less or even be deferred.
Sometimes the future is hard to predict, but in one respect membranes are easy – there will increasingly be more MF/UF water treatment plants. Certainly this includes traditional surface treatment plants but also more challenging applications, such as reclamation facilities that allow recycling of wastewater in water-poor areas. Membranes have been and will continue to be improved with the use of new materials, to provide better tolerance of cleaning chemicals, greater resistance to fouling (to maintain capacity with minimal chemical cleaning), and strength, as discussed above. In addition, analytical and design tools will be improved to select sustainable design conditions, such as flux, with fewer pilot tests. This will include a better understanding of filterability tests as well as fouling “fingerprints” based on water quality measurements.
About the Authors
- Jeff Neemann is assistant director of Water Treatment Technology based in the Kansas City, MO, office of Black & Veatch. He has more than 10 years of experience in pilot testing and designing membrane treatment systems.
- Scott Freeman is a membrane technology leader and Trevor Cooke is a process engineer with Black & Veatch, also based in Kansas City. Paul Delphos is a Black & Veatch membrane practice leader based in Virginia Beach, VA.
- American Water Works Association, Microfiltration and Ultrafiltration Membranes for Drinking Water, Manual of Water Supply Practices M53, First Edition, 2005
- The McIlvaine Company, RO/UF/MF World Markets, 2006
- Business Communications Company, Advanced Technologies for Municipal Water Treatment, 2006.