Citation:

Haught*, R C., C Patterson*, T F. Speth*, R. Sinha, G. A. Sorial, AND B. Ramakrishnan. ADVANCED OXIDATION PROCESSES (AOPS) FOR DESTRUCTION OF METHYL TERTIARY BUTYL ETHER (MTBE -AN UNREGULATED CONTAMINANT) IN DRINKING WATER. Presented at WSWRD Peer Review, Cincinnati, OH, September 23, 2004.

Impact/Purpose:

To inform the public

Description:

Advanced oxidation processes (AOPs) provide a promising treatment option for the destruction of MTBE directly in surface and ground waters. An ongoing study is evaluating the ability of three AOPs; hydrogen peroxide/ozone (H2O2/ O3), ultraviolet irradiation/ozone (UV/O3) and ultraviolet irradiation/ozone/hydrogen peroxide (UV/O3/H2O2) to destroy MTBE in drinking water on a pilot scale. The study investigated drinking water levels of MTBE around 300 g/L. Experiments were conducted in the continuous flow mode with two recirculation ratios of 2.5: 1 and 5: 1 and different oxidant concentrations to study the impact of these factors on the treatment efficiency. A low-pressure mercury vapor lamp was used in the studies involving UV irradiation. Experiments with H2O2 alone showed no effect on MTBE destruction. On the other hand, the use of O3 alone at 4.5 mg/L was able to achieve a maximum treatment efficiency of 70%. Increasing the O3 concentration to higher levels did not appear to further improve the treatment efficiency. The use of UV/O3 resulted in a considerable increase in the treatment efficiency (93% at an O3 dosage of 4.5mg/L). However a high concentration of TBF, TBA and acetone (oxidation by-products) were present in the treated water for both O3 alone and UV/O3. Experiments showed that the use of H2O2/O3 is a more effective treatment process than UV/O3. Non-detectable levels of MTBE were recorded using an O3 dosage of 4.5mg/L and a H2O2 to O3 molar ratio of 1.4. It was found that the by-product concentrations were still high and an applied O3 dosage of greater than 5.5mg/L was necessary to maintain the by-product levels below 20 g/L. The combined treatment process of UV/O3/H2O2 showed no remarkable increase in the treatment efficiency of MTBE as compared to the H2O2/O3 process.

Batch/Continuous Flow Studies with UV/Ozone Oxidation (2001)
A study was conducted in 2001 to evaluate a pilot-scale UV/ozone system's ability to destroy MTBE in drinking water supplies (e.g., surface water, groundwater). The UV/ozone system was tested as a closed-loop recirculating batch reactor at 10 gallons per minute (gpm) for 61 minutes to determine MTBE and MTBE by-product degradation with time. The UV/ozone system was also tested in single-pass mode with partial recirculation at 2.5, 4, and 5 gpm for up to 30 minutes to determine destruction efficiency with continuous flow. UV/ozone system design accommodated several treatment options for comparison: UV/ozone treatment, ozone treatment, UV treatment, and no treatment. Several waters with unique water qualities (dechlorinated tap water, Mill Creek water, East Fork Lake water, and C.M. Bolton well water) were tested at initial MTBE concentrations of 50, 100, 300, 1200 and 3000 micrograms/liter (mg/L) and turbidities of <0.5, 1, 2, and >15 nephelometric turbidity units (ntu). The combined UV/ozone treatment process demonstrated the greatest destruction for all initial MTBE concentrations. Laboratory analysis confirmed the formation of MTBE oxidation by-products such as tertiary butyl formate (TBF), methyl acetate, acetone, butene, and acetaldehyde. Closed-loop UV/ozone treatment oxidized MTBE and MTBE by-products at initial concentrations of less than 1000 mg/L in waters with turbidities of less than 3 ntu. Closed-loop ozone treatment oxidized MTBE at a slower rate than UV/ozone treatment. UV treatment alone resulted in poor MTBE destruction. Continuous flow treatment provided the most cost-effective destruction of MTBE at low initial MTBE concentrations of less than 100 mg/L. Onsite production of UV light and ozone incurred electrical costs ($0.50 to $0.70 per 1000 gallons), but did not require the purchase and handling of liquid oxidants.

Air Stripper Off-Gas Adsorption Studies (2001)
A study was conducted in 2001 to evaluate MTBE removal by the combined treatment technologies of air stripping followed by treatment of the off-gas using adsorption. A pilot-scale counter-current packed tower air stripper was designed and fabricated for this study. Six test runs were conducted on the air stripper with air-to-water ratios ranging from 100:1 to 200:1 and with influent MTBE concentrations ranging from 80 to 250 ?g/L. The MTBE removal efficiency achieved in each of these runs was greater than 98%. The performance of the air stripper was evaluated based on the overall mass transfer coefficient of liquid concentrations determined from performance data.

The air stripper off-gas was treated using two absorber beds in parallel, one containing granular activated carbon (GAC) and the other containing a carbonaceous resin, in order to compare the performance of these adsorbents. Four test runs were conducted with influent off-gas MTBE concentrations ranging from 1.0 to 2.4 ppb and with relative humidity values ranging from 20 to 55 percent. The breakthrough behavior of each bed was monitored with time, and later modeled with the plug flow homogeneous surface diffusion model. The kinetic parameters (surface diffusion coefficient and film diffusion coefficient) were obtained using empirical equations (Sontheimer correlation and Gnielinski correlation). The equilibrium parameters were determined by conducting isotherms of MTBE at the relative humidity and temperatures of the four test runs.

The GAC results showed that the adsorption of MTBE is adversely impacted (reduced) by relative humidity below 50 percent. MTBE is similar to water in that it is a very polar molecule. Water molecules tend to adsorb on the basil plane edges of the GAC because there are oxygen containing functional groups associated with these surfaces (e.g. carboxylic, lactone, and carbonyl groups). Most likely, even at low relative humidity, both water and MTBE will compete for these adsorption sites resulting in lower MTBE capacities. The carbonaceous resin's adsorption capacity for MTBE was not impacted by relative humidity. However, the resin's capacity was lower than what was expected.

Record Details: