The presence of bromide ion in raw water can lead to the formation of brominated disinfection by-products (DBPs) such as bromate during water treatment with ozone. The formation of bromate depends on water quality parameters such as influent bromide ion concentration, pH, total organic carbon (TOC) and various treatment conditions (O3 dose, dissolved O3 and contact time). The US Environmental Protection Agency (USEPA) has proposed a maximum contaminant level (MCL) for bromate in drinking water of 10 µg/L.

Many studies on bromate formation during ozonation have been conducted on a bench-scale under batch or semi-batch conditions. A pilot scale application is considered to be closer to a full scale operation.

The NSERC Industrial Research Chair in Drinking Water Treatment at the University of Waterloo and partners from various segments of the water industry, are conducting research to investigate important issues in drinking water treatment. Three state-of-the-art drinking water treatment pilot plants were built to conduct these studies.

The Windsor Utilities Commission, Windsor, Ontario (one of the NSERC Chair partners), and the Civil and Environmental Engineering Department, the University of Windsor, conducted a pilot scale study for the formation of bromate during a pre-coagulation ozonation process. The Windsor Water Treatment Centre serves three municipalities whose total population is approximately 214,000. Raw water is drawn from the Detroit River which connects Lake St. Clair and Lake Erie.

The Windsor Water Treatment Centre was upgraded in 1994 to increase total capacity from 227 ML/D to 454 ML/D with the construction of a second drinking water treatment plant (A.H. Weeks Plant) at the site of the original plant. Provisions were made for the addition of ozone at some point in the future.

The pilot plant used in this study had two identical process trains constructed of organically inert materials (stainless steel, glass, or fluorocarbons). Each side of the pilot plant shared identical physical characteristics which allowed for direct comparison between the two sides of the plant with common raw water quality.

Aluminum sulfate (alum) was used as the coagulant, and a cationic coagulant aid (Percol LT 24), was applied to both sides of the pilot plant. Ozone was applied prior to coagulation on side 2 of the pilot plant (Figure 1) at a dose of 1.3-1.5 mg/L. Flow rate through the anthracite/sand filters was maintained at 3.25 L/min. (10.7 m/h).

Samples were taken from the pilot plant influent (raw water) for bromide analysis, while samples for bromate analysis were taken from the pre-ozone contactor and anthracite/sand filter effluent.

Bromide concentration in raw water (Detroit River water), was 21-23 µg/L during the experiment; bromide concentration did not exceed 1.1 µg/L after ozonation (samples were collected from the ozone contactor), while bromate was always lower than the detection limit (1 µg/L) for filtered water. (Figure 2)

Future experiments would investigate bromate formation at pre-coagulation ozonation compared to post sedimentation ozonation, and pre-coagulation ozonation compared to multi-stage ozonation.

The findings of this study would be of particular significance to the Windsor Utilities Commission in future consideration for the implementation of ozonation as a process to be considered at the A.H. Weeks Water Treatment Plant and potentially at other locations.

Dr. Saad Y. Jasim is conducting studies on water quality for the Windsor Utilities Commission. He also serves as an Adjunct Assistant Professor in the Civil and Environmental Engineering Department, University of Windsor, Windsor, Ontario.