WaterSubject

In United States, chlorine dioxide (ClO2) was first used as a disinfectant to control taste and odor problems in the 1940s. However, because of high chemical cost and failure and inefficiency of the generation equipment, ClO2 has not become a widely used primary disinfectant in drinking water treatment facilities during the 1960s and 1970s. Due to the concerns about disinfection byproducts resulting from the use of chlorine in drinking water treatment, ClO2 is increasingly being considered as an alternative disinfectant for drinking water treatment. The selectivity and oxidation potential makes ClO2 a desirable alternative to free chlorine. Its biocidal efficiency is equal to or superior to chlorine. Moreover, ClO2 is effective over a wide pH range and is very effective for removing iron and manganese.Current applications of ClO2 in drinking water treatment include the use of ClO2 for secondary disinfection, nitrification control, oxidation of the cyanobacterial hepatotoxin microcystin-LR (MC-LR) and bromate control in the desalination process.

New electrochemical generation systems use electrochemical cassettes and membrane technology to generate stock solution with approximately 500 mg/L ClO2 from 25% sodium chlorite solution. The equipment is easy to install and safe to operate in the institutional plumbing system. Several studies have been conducted to evaluate the efficacy and safety of ClO2 generated by this electrochemical process for controlling water-borne pathogens in hospital water systems. These studies showed that ClO2 residual in hot water was significantly lower than in cold water. It is possible that faster reaction of ClO2 with organic compounds in hot water and high organic load in the hot water contributed to such observations.

However, the loss of ClO2 due to corrosion scales has not been studied in detail. Reactions between free chlorine and iron corrosion scales in distribution system have been reported to account for the significant free chlorine loss in the distribution system. Loss of chlorite in cast-iron pipe loops and full-scale drinking water distribution systems containing cast-iron pipes has also been reported.

Iron corrosion scales in water distribution systems have been investigated extensively because iron pipes are commonly used for distributing drinking water. The compounds usually found in iron corrosion scales include goethite (?-FeOOH), lepidocrocite (?-FeOOH), magnetite (Fe3O4), siderite (FeCO3), ferrous hydroxide (Fe(OH)2), ferric hydroxide (Fe(OH)3), ferrihydrite (5Fe2O3·9H2O), green rusts (e.g., FeII4Fe2III(OH)12(CO3)) and calcium carbonate and. Previous studies showed that iron corrosion scales generally contain reduced iron, which can react with oxidative disinfectants. ClO2 is a strong oxidant and will oxidize ferrous compounds in iron corrosion scales. The reactions of ClO2 with corrosion scales will lead to undesirable losses in the disinfectant residual.

In this study, the corrosion scales from a galvanized iron pipe and a copper pipe that have been in service for more than 10 years were characterized by energy dispersive spectroscopy (EDS) and X-ray diffraction (XRD). The impact of corrosion scale materials on ClO2 decay was investigated in DI (de-ionized) water at 25 and 45 °C in a batch reactor to simulate the application of ClO2 in hot and cold water systems. In addition, ClO2 decay was also investigated in a specially designed reactor made from the iron and copper pipes to obtain more realistic reaction rate data.

A 30-inch long, 4-inch diameter galvanized iron pipe and a 15-inch long, 2-inch diameter copper pipe that have been in service for at least 10 years were obtained from a local hospital water system and used for this study. The iron pipe was covered with deposits of corrosion products and heavily tuberculated as show. The copper pipe was comparatively clean and only a thin film of corrosion scale was detected by a scanning electron microscopy (SEM) image.

Inner surface of the galvanized iron pipe, (b) inner surface of the copper pipe from a local hospital water system and (c) SEM image of the copper pipe wall.

The scales were scraped from the top, middle and bottom layers of the tubercles close to the end of the iron pipe and ground into powder. The scales from the copper pipe were scraped from the copper pipe wall from both ends of the pipe. The elemental composition of the scales was analyzed by EDS using a Philips XL Series 30 scanning electron microscope and an X-ray energy dispersive spectrometer (Philip Analytical Inc., Natick, MA). The XRD patterns of the samples were obtained with a Philips X’PERT diffractometer (Philip Analytical Inc., Natick, MA) using a standard Ni-filtered Cu K? radiation source operating at 40 kV and 30 mA X-ray patterns were analyzed using pattern processing software based on the latest Joint Committee on Powder Diffraction Standards (JCPDS) files. Powder samples were also sent to Materials Characterization Laboratory (Pennsylvania State University, University Park, PA) for X-ray photoelectron spectroscopy using a Kratos Axis Ultra X-ray photoelectron spectrometer (Kratos Analytical Inc., Chestnut Ridge, NY) with an X-ray source of monochromatic Al K? (1486.6 eV).

A cell culture flask (Wheaton Science Products, Millville, NJ) capable of holding 2500 mL of liquid was used as the batch reactor. A floating glass cover was used to prevent exchange of gases between the headspace and room air and to minimize the volatility of ClO2. Single-port 45-mm red rubber stoppers with a 0.25-inch hole (Wheaton Science Products, Millville, NJ) were used on the side arms as a temperature monitoring port and a sample withdrawal port. The stoppers were sealed gas tight using a 45-mm inlet cap (Wheaton Science Products, Millville, NJ). The reactor and all parts were autoclaved prior to each experiment. The reactor was soaked overnight in 50 mg/L ClO2 solution to satisfy disinfectant demand of the reactor material and rinsed with DI water before use.

The batch reactor with the floating glass cover.

The experiments were carried out at room temperature, which varied in a very narrow range of 25±2 °C. Hot water temperatures were maintained by heating the flask on the hot plate with a temperature probe feedback. Temperature monitoring was performed using the temperature probe of the hot plate (PMC Industries Inc., San Diego, CA).

The corrosion scale material for these experiments was sampled from the entire scale of the corroded iron pipe and ground to powder without sieving before adding to the batch reactor. Commercial cuprite (Cu2O) and Fe3O4 powder (particle size <5 ?m; Sigma-Aldrich, St. Louis, MO) were also used in these batch reactor experiments.

A ClO2 generator (Diox, Klenzoid Inc., Conshohocken, PA) provides a concentrated ClO2 stock solution. The effluent solution was discarded until the generator achieved a ClO2 concentration of approximately 500 mg/L. The ClO2 concentration of the concentrated stock solution (fresh stock solution was prepared for each experiment) was monitored using the Hach Method 8138 (0-700 mg/L). The appropriate amount of the ClO2 stock solution was pipetted into the batch reactor to achieve the target ClO2 concentration of 1.0 mg/L. The decay of ClO2 was monitored after the addition of the corrosion scales using the Hach Method 10101-DPD Method for ClO2 (0.00-5.00 mg/L) utilizing a glycine reagent and Hach DPD Free Chlorine Reagent (Hach Company, Loveland, CO). The colorimetric measurements were made using the Hach DR/2010 Spectrophotometer (Hach Company, Loveland, CO).

Samples for chlorite, chlorate and chloride analyses were collected at the beginning and at the end of each experiment. Chlorite, chlorate and chloride were measured by ion chromatography (DX-500, Dionex, Sunnyvale, CA) equipped with a suppressor and conductivity detector according to USEPA Method 300.1. Chlorite, chlorate and chloride concentrations produced through ClO2 reaction with corrosion scales were determined as the difference between the final and initial concentrations to eliminate the interference due to the presence of these ions in the stock solution.

pH of the solution was buffered with 0.1 M phosphate buffer and it was adjusted by the addition of 0.1 M NaOH.

Copper and galvanized iron pipes were used to set up the experimental system shown. Three holes were drilled in the corroded iron and copper pipes and a 0.25-inch plastic tubing was attached as sampling ports. The pipe reactor was flushed with tap water for 24 h before the experiment to re-wet the pipe surface and flush out any easily dislodged tubercles. The flow rate of 1.0 mg/L ClO2 stock solution through the pipe was adjusted to achieve the retention time of 10, 20 and 30 min for sampling ports 1, 2 and 3, respectively. The ClO2 residual at each sampling port was measured during the experiment until stable levels were achieved. These experiments were conducted in duplicate to provide statistical validity of the results.

Goethite (?-FeOOH) and magnetite (Fe3O4) were identified as the main components of iron corrosion scale. Cuprite (Cu2O) was identified as the major component of copper corrosion scale. The reaction rate of ClO2 with both iron and copper oxides followed a first-order kinetics. The estimated first-order reaction rate constant for ClO2 reaction with iron corrosion scales and Fe3O4 ranged from 0.0251 to 0.0829 min-1. The estimated first-order reaction rate constant for ClO2 reaction with Cu2O was much smaller and it ranged from 0.005 to 0.006 min-1. Fe3O4 and Cu2O were likely the main compounds in the scales that caused ClO2 loss in this study through a one-electron-transfer mechanism. The loss of ClO2 in the corroded iron pipe is most probably dominated by the reactions between ClO2 and this ferrous compound present in the corrosion scale. Based on these results, it can be concluded that the corrosion scale will cause much more significant ClO2 loss in corroded iron pipes of the distribution system than the total organic carbon that may be present in finished water. The application of ClO2 in the water distribution system using cast-iron pipes is not recommended unless measures to prevent corrosion are fully implemented. Although ClO2 loss caused by corrosion scale was much slower in the copper pipe than in the iron pipe, it may still be necessary to prevent the corrosion and unnecessary loss of disinfectant due to the corrosion scale in the copper pipe distribution system to maintain effective disinfectant residual.