Example: Installation #1
Summary of Results
Introduction (Back to Top)
Chlorine is the most frequently used chemical for water disinfection. However, many industrial and commercial processes cannot tolerate chlorine in the process because of chlorine's reactive nature. For example, chlorine affects the flavor or smell of drinks and liquids. In pharmaceutical processing, chlorine oxidizing agents can disrupt important chemical reactions and destroy products. High chlorine levels also accelerate corrosion on process vessels and piping, and are generally considered damaging to process equipment. For example, reverse osmosis (RO) membranes and deionization (DI) resin units can be damaged by continual exposure to chlorine.
Granular activated carbon (GAC) filters or the addition of such chemicals as sodium metabisulfite are the most common methods of removing both free chlorine and chloramines (combined chlorine) within a water treatment system.
Sodium sulfite or sodium metabisulfite is purchased either in premixed solution or as a dry powder and then mixed on site. It is commonly injected in front of RO membranes so that the sodium combines with the chlorine to make harmless salt. This practice is common in the pharmaceutical and semiconductor industries. One common problem with this approach is the resulting salt solution becomes an incubator of bacteria, causing biofouling of the membranes. Bisulfite only adds to the complex chemistry of the incoming water, and must be removed during the water purification process. Bisulfite must also be documented in use, handling and storage for regulatory agencies such as the US Environmental Protection Agency (EPA) and the Occupational Safety and Health Administration (OSHA). Then there is the human factor, by which over- and under-mixing of the chemical may occur.
Granular activated carbon (GAC), common in the beverage and pharmaceutical industries, has an imperfect structure as a "cleaning up" agent. GAC's high porosity and propensity to promote intermolecular attractions enable it to remove volatile organic compounds, trihalomethanes, and other halocarbons from process streams. But GAC's benefits are also its downfall: Because of the porous structure and nutrient-rich environment, GAC beds become prime incubators for microorganisms.
Long ago, it was discovered that natural sunlight oxidizes free chlorine in water. This natural phenomena forces owners of outdoor pools to routinely add chlorine to their pools in order to maintain a disinfection residual. Municipal water districts using surface water sources must also compensate for the sun's effect when chlorinating. This phenomenon can be duplicated on a commercial level using a concentrated source of UV light emitting specific wavelengths. This article reviews different field trials conducted to evaluate the effectiveness of UV oxidation of free chlorine.
High-intensity, medium pressure UV arc tubes emit a broad spectrum of UV wavelengths. The light emitted by these arc tubes dissociates free chlorine through a photochemical reaction, forming chloride ions and/or hydrochloric acid. The peak wavelengths for dissociation fall above 300 nm (see Figure 1). Low pressure UV lamps are not as effective as medium pressure arc is removing free-chlorine.
Example: Installation #1 (Back to Top)
A large west coast semiconductor manufacturer recirculates 80 gpm of RO water through an air scrubber to wash isopropyl alcohol (IPA) out of the exhaust air coming from one of their manufacturing processes. After removing the IPA from the exhaust air, the water is run through RO to reject the alcohol. Approximately 8-10gpm of city water is continually added as makeup. (see Diagram 1). The RO membranes are thin film composites which achieve high levels of IPA rejection. Since the warm summer months will promote biofouling within the packed tower air scrubber, a powerful biocide was selected.
After operating for several years this way, the very high cost and extreme hazards of working with this biocide prompted a search for a reliable, cost-effective alternative. After considerable review, the decision was made to inject 1.0 ppm of sodium hypochlorite in place of the biocide. A special medium pressure UV arc tube system was installed inline just ahead of the RO membranes for dechlorination. The design target was to reduce free chlorine levels of 1.0 ppm to less than 0.01 ppm, with 500 ppm of IPA present in the water. The outlet limit of 0.01ppm of free chlorine was chosen based on a "1,000 ppm-hours" rule-of-thumb. This "rule" set a guideline for a thin film RO membranes exposure limit to free chlorine. It says that an RO membrane can tolerate 1 ppm of free chlorine for 1,000 hours before "significant" damage has occurred. Therefore, based on our limit of 0.01 ppm free chlorine, the end users membranes won't reach the 1,000 ppm-hours limit until after 11.4 years of exposure, an acceptable timeframe.
The end user installed a single, 3.5 kW medium pressure arc tube system just prior to RO as shown in Diagram 1. In RO quality water, this medium pressure UV system provided at UV dose of 420,000 µW-sec/cm2 (14X). Testing was conducted over the period of two days in August 1998. Free chlorine and combined chlorine measurements were taken just prior to and immediately following the UV unit. Each time a measurement was to be taken, two water samples were taken. One sample was measured with a "PALINTEST Micro1000 Chlorometer" and the other sample was measured with a "HACH Pocket Colorimeter." This was done as a check for measurement accuracy. If there was a difference of 0.02ppm or less, it was attributed to equipment resolution and the higher value was recorded. If the difference between the two sample readings was greater than 0.02ppm, this was attributed to sampling error and new samples were taken and measured.
Summary of Results (Back to Top)
Medium pressure UV was effective in this trial at reducing free chlorine concentrations by > 96% minimum. Average reduction was approximately 98%. Testing conditions did not permit evaluating what minimum dose would be required under these conditions to guarantee 100% free chlorine reduction. Variables not within our control which affected dechlorination performance were quality of makeup water and concentration of IPA coming from scrubber. A change in one or both of these variables would have a direct impact (positive or negative) on dechlorination performance. Combined chlorine (as a percentage of total chlorine) was relatively low in this pilot test. Based on additional testing not documented in this article, it is believed that for a given concentration of total chlorine, the UV dose required for total chlorine destruct increases as the percentage of combined chlorine increases.
Conclusions (Back to Top)
Medium pressure UV is a feasible alternative for the destruction of incoming chlorine. The industry-standard UV disinfection dose applied in process waters is 30µW-sec/cm2. Since this alternative requires a much higher UV dose be applied to the process water, the user also benefits from a high log reduction of micro-organisms.
Close attention must be given to the possible variations in the quality of the water to be dechlorinated. Depending on a users objective and any problems with an existing dechlorination method, this alternative could be a more cost effective and reliable method.
For more information: Brad Shipe, Pharmaceutical Industry Sales Manager, Aquionics Inc., 21 Kenton Lands Rd., Erlanger, KY 41018. Tel: 606-341-0710. Fax: 606-341-0350.