By Larry Stenger, Sr.
Summary: The reverse osmosis (RO) industry has several pre-treatment scenarios it recommends for its systems including water softening to reduce mineral scale formation, multimedia filtration to reduce suspended solids, and a filtration method to reduce oxidizing agents.1 The process medium discussed here—KDF 55—fulfills all of the basic requirements for pre-filtration to various degrees to provide excellent water quality for RO system performance.
Since the discovery of reverse osmosis (RO) membrane technology some 50 years ago, biofouling continues to be a problem that interferes with the satisfactory operation of membrane systems.
Biofouling is a term to describe the growth of biological materials that block the filtration capabilities of a particular system. This material could be bacteria, algae, mold, etc., and is usually represented by the physical presence of a slimy substance or biofilm. The large surface areas of RO membranes with their web like spacers offer excellent opportunities for biological activity to take place. The flow of water over these surfaces can bring plenty of oxygen, food and minerals to the hordes of microorganisms reproducing inside such a system.
Many times both professionals and non-professionals in the industry assume that because the water supply is chlorinated or it’s pretreated with ultraviolet (UV) light or ozonation for disinfection that biological activity won’t occur or become a foulant issue within the RO system. What’s sometimes overlooked is nature seeks every opportunity to promote biological growth. This can occur in relatively short timeframes. Cases I’ve known include: newly installed RO systems that became biologically fouled within a few days of start up after they were let to sit idle only to grow internal bacterial or algae slimes; RO systems that ran great for eight months then, during warmer summer months, slowly started to lose permeate flow, with corresponding decreases in ion rejection; and systems that ran well for years with no problems and then, when membranes were changed, biological fouling rapidly shut them down.
The number of problems with RO systems increases when source water quality is varied, including whether it’s publicly or privately treated. Deep well water supplies have a tendency to have less biological activity due to lower temperature and oxygen levels. And depending on the size and length of the water’s storage and distribution system, type of piping, age and treatment methods, unusual levels of algae and other biofilms can grow in system supplies, also affecting RO membrane performance.
The current construction of RO membranes can contribute to the potential biological fouling problem because growth can take place on the outside of the membrane as well as inside the membrane envelope. Cases exist in which bio-sensitive high purity water applications suffered biological problems because the inside of the membrane envelope was contaminated during the manufacturing cycle. The inside of membrane envelopes have also been documented as contaminated from faulty permeate check valves and/or reject system designs that allowed backpressure to develop. Once biological activity reaches a certain population stage inside the membrane envelope (indicated by the loss of permeate flow), it is difficult to remove. Chemical flushing with caustic-type cleaning compounds can be ineffective, as occasionally many of the chemicals are rejected on the outer surface of the membrane before they can get inside the RO envelope.
There are a host of chemical companies and some RO membrane manufacturers who offer various types of cleaning compounds to help clean and restore membranes that are biologically fouled. Depending upon the size of the system, this can take anywhere from several hours to several days. In some cases, results still aren’t good enough and a new set of membranes must be ordered—an expensive lesson. Generally speaking, if the membranes are cleaned before pressure increases or TDS rejection falls by 10 percent from original values, then cleaning is considered successful.
Preventing biofilm growth in your RO systems can be a formidable task. Regardless of water source, though, use of activated carbon filters to remove chlorine can create a bioreactor within the carbon bed. The chlorine is removed within the top of the carbon bed while opportunistic algae and bacteria find ideal breeding grounds in the bottom. Over time, they create large colonies that then get flushed out of the bed into the RO cartridge prefilter and RO membranes themselves. Since most RO systems don’t run 24 hours a day, the off-time allows biological growth to concentrate and accelerate—which can cause poor performance in a few months to a matter of days.
One recommendation is that if an RO system has to be shut down for more than three days (especially on warmer surface waters) the system should have a preservative solution that removes oxygen and reduces aerobic activity—such as sodium metabisulphate—run through the membranes to reduce potential biological activity. This process will help but does little for controlling the day-to-day biofilm growth that takes place in running systems. Finding an agent that controls the daily growth of biological activity that doesn’t potentially cause damage the RO membrane has been a constant challenge to the RO industry.
Thin membrane test
In early 1991, a new oxidation/reduction (redox) filter media—utilizing high purity copper-zinc formulations—was installed as pretreatment in a small 15,000 gallon per day (gpd) RO system using thin film composite membranes for a St. Petersburg, Fla., municipal underground water supply noted for warm temperatures and biofilms in its distribution system. This media is registered as a bacteriostatic agent with the U.S. Environmental Protection Agency (USEPA) and holds NSF Standard 61 and 42 validations.
Seven years later to date, the same membranes and redox media are being employed. The RO membranes are still in use, although they’ll need to be replaced in the near future, since they’re showing signs of losing normal mineral rejection. The interesting part of this experiment was that, when the redox media filters were taken off line for several weeks, the RO system lost over 10 percent of the flow and rejection, requiring cleaning of the membranes.
The recommended design for an RO pretreatment system to provide biological control is as follows: for continuous operation, a twin water softener for hardness removal (rated so each tank can handle the required flow) followed by twin automatic backwashing redox filter media sized to the manufacturer’s specifications, followed by a polishing activated carbon filter to remove any remaining chlorine (if the water is chlorinated). If the RO system can be shut down during the softener regeneration and redox media backwash cycles, then a single-tank design will work instead of the twin filters. Five-micron prefilters are then used before the water enters the RO membrane pumping system. A very important factor in using the granular redox media is to insure you have a good backwash design to prevent amalgamation or clumping of the media.
If you start off with a clean system—because the copper/zinc redox media is considered bacteriostatic by the USEPA—the media keeps the system resistant to bacterial growth. Part of the interesting phenomenon about this method of pretreatment is control of biological activity passes through the RO membrane such that the inside envelope of the membrane is also receiving the benefit of biological control performance. It’s believed the redox reaction—the gaining and transfer of electrons between dissimilar metals in an aqueous solution—causes low levels of hydroxyl radicals to form, which pass through the membrane carrier. This is a major issue in high purity water production where even very low levels of biological activity are unacceptable.
The results are extremely interesting from both a performance and an economic standpoint. The savings in time and money in cleaning membranes over the past seven years is estimated at over $750 for the small system mentioned above, not accounting for downtime in the rest of the plant due to cleaning requirements. Savings in the cost of replacing 14 thin film composite membranes every three years is $8,400 so far. The cost of providing the redox media system was approximately $1,500. This is a very good investment by any standard. And the quality control factor of having high quality water offers a great operational advantage.
Biofilms are natural formations that can interfere with many water treatment processes. Reverse osmosis membranes offer unique places for biofilms to grow, which can impede flow and purity of the water produced by the RO system. The materials used to manufacture thin film RO membranes are very sensitive to normal agents such as chlorine, chloromines and ozone—disinfectants generally used to control biological growth. The use of redox media offers an effective alternative method of extending the performance of RO membranes.
- Harfst, W.F., “Back to Basics: Pretreatment Requirements for Reverse Osmosis Systems,” Ultrapure Water, Nov. 1994, pp. 42-44.
- Martin, David H. “Creative Marketing: Lessons Learned—Launching New Technologies,” WC&P, Jan. 1999, pp.18-20.
- Stenger, Larry, D. Heskett and W. Ball, “Using an Alloy to Remove Metals and More,” WC&P, June 1990, pp. 56-58.
About the author
Larry Stenger is the founder and former president of Water Equipment Technologies (WET) Inc., now part of Waterlink, Inc. WET was the first company to apply thin film composite RO membranes in drinking water markets in 1977. He’s a graduate of the University of Miami business school and holds several patents on water treatment devices. KDF Fluid Treatment Systems Inc. of Three Rivers, Mich., the manufacturer of the redox media in this article, can be reached at (616) 273-3300, (616) 273-4400 (fax) or email: [email protected]