CARBON DIOXIDE INJECTION TRIALS AT NELSON WATER TREATMENT PLANT, NEW ZEALAND - IMPACT ON PH, COAGULATION AND MEMBRANE FILTRATION PERFORMANCE

摘要

The Tantragee water treatment plant in Nelson, New Zealand was built to provide upgraded 'A' standard drinking water quality for a population of 40,000 people. The new NZ$26M plant, commissioned in 2004 provides coagulation, flocculation, ultrafiltration and sodium hypochlorite disinfection. The plant treats water from three sources. Colour levels from the River sources are typically low. However, colour from the Dam source can be high, requiring chemical precipitation. A key feature of the raw water quality from the Dam source is the combination of low alkalinity and high pH. When raw water is sourced from the Maitai Dam, ferric chloride is dosed at the plant inlet for both coagulation and pH adjustment to achieve a pH of 6.2 prior to the membranes. The ferric chloride dose is flow paced at a typical dosing rate of 20 to 30 mg/L as FeCl3. When ferric chloride dosing is practiced, the final water is pH corrected to 7.9 after the membranes by the addition of soda ash. Iron precipitation in the treated water reticulation has been reported, believed to be associated with the residual iron remaining in the permeate water. Preliminary calculations indicated that the cost of ferric chloride dosing may be reduced if an alternative chemical, such as sulphuric acid or carbon dioxide is used for pH adjustment and the ferric chloride is used only for coagulation and flocculation. Estimates suggested that operating cost savings could be made as well as reducing the residual iron concentration remaining in the permeate water. It was decided to undertake a full scale trial of carbon dioxide dosing to demonstrate the predicted reduction in ferric chloride dose and examine concerns about the potential effect on the performance of the flocculation and membrane filtration stages. A skid mounted carbon dioxide dosing system was sourced and installed at the plant inlet and a series of trials undertaken with varying carbon dioxide and ferric chloride dosing rates. Plant performance was monitored using on-line instrumentation and sampling. This paper describes the background to the project, the operation of the carbon dioxide full scale trial and presents results including the impact on the flocculation stage pH, ultra-violet transmissivity and membrane performance.