AWWA WQTC58825

The City of Phoenix (City) currently operates four conventional water treatment plants (WTP). The City is updating its Water Quality Master Plan (WQMP). The initial phase of the WQMP update study recommended modifying the current treatment processes to achieve enhanced coagulation (EC) and incorporating additional treatment processes such as granular activated carbon (GAC) filter adsorbers and ozone at all the City WTPs. In order to develop the design and operating criteria for GAC filter adsorbers and intermediate ozonation (ozonation before filtration), the City funded a pilot study that was performed over a period of seven months from December 2002 to June 2003. This paper summarizes the key findings of this pilot study. The key objectives of the pilot testing were: can GAC filter adsorbers assist in complying with the Stage-2 D/DBP Rule requirements; what is the acceptable filter media depth for the GAC filter adsorbers; what are the benefits of backwashing with unchlorinated water; and, what are the benefits of applying ozone prior to the GAC filter adsorbers? To address the above objectives, a pilot facility was constructed and operated adjacent to the Salt River Project's Arizona Canal. This pilot facility had 11 filters; each filter was 3-inches (in) in diameter and 15 feet (ft) tall. Six of the 11 filters received plant settled water (SW) and the rest of the five filters received ozonated settled water (OSW). The pilot filters were filled with different media that included anthracite (A)/sand (S), bituminous GAC (GAC-B) and lignite GAC (GAC-L). The GACs that were used in the pilot study were of 8x20 mesh size (US Sieve) with an effective size of 1.0-1.2 millimeters (mm) and uniformity coefficient of less than 1.5. Virgin reagglomerated GAC media obtained from a leading national manufacturer was used in this study. The SW and OSW served as the feed waters to the pilot filters. Ozone was applied at a mass ratio of 0.5 mg of ozone per mg of TOC in the settled water. The average TOC of the SW ranged from 1-3 mg/L and therefore the ozone dose ranged from 0.5-1.5 mg/L. At the ozone dose of 0.5-1.5 mg/L, no significant TOC removal was observed. Ozonation lowered the average specific ultraviolet absorbance (SUVA) from 1.9 in SW to 1.4 in OSW. The bromide concentration in SW was between 80 and 110 Au/L. The low ozone dose and the moderate ozonation pHs resulted in bromate formation at less than the method detection limit (MDL) of 5 micrograms/liter (Au/L). The pilot influents and filtered waters were routinely monitored for field parameters such as pH, turbidity, temperature, ultraviolet (UV) absorbance at 254 nanometers (nm) wavelength, ozone residual in OSW and chlorine residual in backwash water. Additionally, samples were collected and sent to certified laboratories for analysis of TOC, heterotrophic plate counts (HPCs), bromide, bromate, THMs, HAAs and taste and odor (T&O) compounds. The simulated distribution system (SDS) THMs and HAAs were measured by chlorinating the samples to targeted distribution system chlorine residual levels of 0.5-1.5 mg/L at the end of 2-days and 5- days. The 2-days and 5-days correspond to the average and maximum water ages in the City's distribution system. The results and discussion are presented in four sections that correspond to the four objectives. Includes tables, figures.