This page was initiated (8-08-05) to offer water treatment performance enhancing tips to our industry. If you would like to comment on a Water Tip or add one of your own, please e-mail CEAcustomerService@DIwater.net.
Note: CEA postings are taken from the Encyclopedia of Water Treatment, Reverse Osmosis, A Practical Guide for Industrial Users, or as noted.
I.1. Filtration of fragile floc, such as aluminum based floc. Aluminum based floc will tend to work its way through most filtration media, even microfiltration membrane. When attempting to filter it with a granular media, an extremely slow flow velocity is best. Gravity filters operating at a flow rate of 1 gpm per square foot of media work well, but this slow rate would likely result in channeling if applied to a pressure filter. CEA 8/8/05
I.2. Tank corrosion in activated carbon filters. Corrosion of the inside surface of metal tanks used in activated carbon filtration is common. The activated carbon media creates an environment well suited for anaerobic biological activity that can aggressively attack metal surfaces. Not allowing the filter bed to sit stagnant will reduce some of this activity and its corrosion potential, especially if the feed water contains dissolved oxygen or a biocide. CEA 8/8/05
II.1. Using a hydrometer to check softener regeneration steps. You can use a hydrometer to check a number of softener operational parameters. The diluted brine should have a specific gravity of about 1.075 for a 10% NaCl concentration (1.000 when negligible salt is present). You can note how long it takes for the brine to pass through the tank as a way to verify regenerant flow rate and the tank’s lateral flow distribution system. You can verify that the effluent brine concentration starts to decline right at the end of the slow rinse step (as it should). If these events occur too quickly, there may be a problem with a lateral. CEA 8/8/05
II.2. Conserving salt. Salt may be conserved by re-using the softener effluent produced during the very end of the slow rinse step (regenerant displacement step). This water contains a high NaCl concentration with a limited concentration of hardness. If willing to accept a slight increase in the concentration of hardness shed by the softener, this effluent water can be used to dilute the salt in the brine tank (as an alternative to the addition of raw water to the tank) . CEA 8/8/05
III.1. Include phosphates in the scale inhibitor formulation when iron is present. The inclusion of a phosphate component (such as an organophosphonate) in a scale inhibitor formulation will sequester iron that might otherwise foul the reverse osmosis system. It will also tend to prevent polymeric scale inhibitors from coagulating with the iron and subsequently coming out of solution. CEA 8/8/05
III.2. Do NOT over-inject reducing agents. When attempting to achieve a near zero chlorine concentration by injecting a sulfite based reducing agent, a common approach is to increase the injected concentration of a sulfite/bisulfite reducing agent. But if the residual sulfite concentration is greater than 3 mg/L, there is a greatly increased chance of creating a reductive environment where anaerobic bacteria will thrive and rapidly reproduce. This type of bacteria has been known to permanently foul a reverse osmosis system within a weekend. CEA 8/8/05
III.3. Elimination of chloramines prior to an RO/EDI system. A common challenge for high-purity water systems is the elimination of chloramines prior to a reverse osmosis (RO) or electrodeionization (EDI) system. An RO can tolerate low concentrations of chloramines when iron or other transition metals are not present, but the anion resin and sheets in a downstream EDI system are much more sensitive. To protect these systems, as well as the process water quality, facilities are forced to find ways to eliminate the chloramines.
High pH and excess ammonia will reduce chloramine reactivity.
The complete elimination of chloramines is particularly difficult if the pH is relatively high (> 8) or if an excess of ammonia has been added by the water plant in the original creation of the chloramine residual. These conditions appear to drive the equilibrium reaction between free chlorine and ammonia more heavily toward the chloramine (NH2Cl) compound, which is less reactive. The result when relying on the injection of a reducing agent like sodium bisulfite, is that chloramine elimination often requires excessive bisulfite concentrations (> 2 mg/L residual bisulfite) or increased contact time (such as with longer pipe runs or storage).
Do not over-inject sodium bisulfite.
I have found numerous RO and EDI fouling problems to be directly related to the injection of an excessive bisulfite concentration. The bisulfite removes dissolved oxygen and creates a low Oxidation-Reduction Potential (ORP) environment that is ideal for the rapid growth of anaerobic bacteria. These bacterial species can create massive amounts of inert polysaccharide-based materials that readily foul RO/EDI systems and are difficult to clean. The problem is often magnified by warm water temperatures and long pipe runs after the bisulfite injection.
What’s the answer?
In one facility that suffered from these problems, the replacement of the bisulfite system with conservatively sized activated carbon (AC) filtration appears to have completely eliminated the RO fouling. Yet, I have worked with AC filters that also seemed to cause downstream RO fouling. I believe these particular problems tended to occur if the ORP of the AC inlet water was not kept high (i.e., well chlorinated), and/or if bacteria were able to penetrate the lining of steel AC vessels.
Anaerobic bacteria have an impressive ability to find the weakness in the lining of steel vessels, especially when they contain activated carbon. If the AC is not frequently sanitized or replaced, bacteria may find their way through the lining of a steel vessel and corrode the steel. In the process, the bacterial activity is stimulated, which increases the production of materials that can foul downstream RO systems. Using fiberglass pressure vessels reduces the potential for these problems.
Chlorine-destruct UV systems are potentially a “clean” alternative to bisulfite injection and AC filtration, but may suffer from a similar challenge to that of bisulfite injection when attempting to break down stable chloramines (less reactive as a result of high pH or excessive ammonia). Their complete elimination requires such over-sizing that the application may become cost prohibitive. A possible way to reduce the cost might be to combine the UV with a minimal concentration of bisulfite. CEA 9/29/06
IV.1. Temporarily improve RO permeate quality. If the RO high pressure pump is driven by a variable frequency drive (VFD) or if the pump is throttled, increasing the feed membrane pressure will create additional permeate flow. The rate of salt passage is mostly independent of the applied pressure so the end result is an improvement in permeate quality. However, this increased permeate flow will also cause an increase in the rate of membrane fouling if suspended solids are present in the feed water. CEA 8/8/05
IV.2. Improve RO energy efficiency. If the raw water is heated (not by a waste heat source) as a means of increasing the RO permeate flow rate, and if the RO high pressure pump is driven by a variable frequency drive (VFD) or if the pump is throttled, greater energy efficiency and improved performance is gained by operating at higher pressures. Permeate quality is better and there will be reduced biological activity. CEA 8/8/05
IV.3. Fouling by Bacteria. Most heavy biological fouling problems with reverse osmosis and electrodeionization systems are the result of uncontrolled anaerobic bacteria growth, which can be prevented. CEA 9/13/05
IV.4. Frequent Shutdowns. If frequent shutdowns are necessary, designing the RO system to automatically rinse out its raw water with purified water has been found to dramatically reduce biological and scale formation problems. CEA 9/26/05
IV.5. Reduce Fouling. High permeate flux rates in an RO system often contributes to high fouling rates localized in the lead-end membrane elements. CEA 1/6/06
IV.6. Prevent RO Draining. RO systems are often plumbed with their concentrate or permeate outlets located below the top RO pressure vessel. If not protected by isolation valves on these lines, this often results in draining when the RO shuts down, which leads to other problems. CEA 1/26/06
IV.7. Better Monitoring. Monitoring individual ion concentrations, such as sulfate and chloride, in the RO feed and permeate streams results in more accurate rejection monitoring, and easier identification of O-ring problems. CEA 3/6/06
V.1. Distribution problems with uniformly sized resin beads. Uniform bead sizing greatly improves the ability to separate cation and anion exchange resins in a mixed bed during regeneration. Their use will also result in a decrease in pressure drop across the bed. Unfortunately, this can result in greater channeling in beds that have poorly designed distribution laterals, since the greater pressure drop across the resin aids in distributing the flow across the bed. Posted by CEA 8/8/05 based on information provided by Ken Frederick of Ion Exchange Associates, Inc.
V.2. Reducing the creation of colloidal silica in mixed beds. If mixed beds are fed an acidic feed water (as is common when reverse osmosis is used for the make-up water), the anion resin will tend to exhaust first. This results in extremely acidic conditions at the bottom of the bed that will tend to polymerize any previously removed silica into colloidal silica particles. These are shed into the effluent water and are difficult to remove by downstream beds. This often occurs without noting any change in the effluent water resistivity. To prevent the shedding of silica (and boron), mixed beds should be regenerated prior to demonstrating any drop in effluent resistivity. CEA 8/8/05
V.3. Fouling by Bacteria. Most heavy biological fouling problems with reverse osmosis and electrodeionization systems are the result of uncontrolled anaerobic bacteria growth, which can be prevented. CEA 9/13/05
