An assessment of the possible occurrence of riverbank filtration in selected municipal water supplies within New Brunswick

Abstract

It is very important to identify all water sources of a particular water supply system. With this knowledge an appropriate water supply treatment plan can be implemented since all potential sources of contamination are identified. If this strict classification system is not followed, possible sources of contamination may be missed, possibly leading to water quality being compromised for a whole water supply network. However, on the other hand, the process of riverbank filtration can be used as an effective pre-treatment process that can significantly reduce the costs incurred by treating raw water through traditional chemical, and filtration treatment facilities. When a groundwater flow system is hydraulically connected with a surface artificial recharge source, a whole new list of potential sources of contamination must be identified, and adequate emergency systems must be implemented. Water treatment processes may also change for the municipality since new possible sources of contamination can be identified. However, water treatment costs may be reduced since riverbank filtration is proven to provide an effective means of raw water pre-treatment. This project outlines the method of creating and using a Comparison Matrix. The Comparison Matrix is the medium used to compare and evaluate the water supplies for 13 different municipal water supply systems within New Brunswick. This process has helped to give an optimum comparison since, only municipalities with sufficient municipal water supply data was selected for comparison. All of the data used for the project was taken from the New Brunswick Department of Environment and Local Government. Therefore the quality of the project results is highly dependent upon the accuracy of these data. All of the selected municipal groundwater flow systems have been compared with the Fredericton South water supply, and a "Base Case" synthetic scenario. These systems were chosen since they both represent examples of known, or highly probable riverbank filtration systems. As the similarity to these cases increase, so does the probability of riverbank filtration for the municipality being compared. Two separate Case Based Reasoning formulae were used to determine the similarity to each case, giving a relative ranking of similarity. This has allowed for a comparison of the two CBR formulae used; The Similarity Factor and the Similitude Function. The five factors that were examined, and compared against the Fredericton, and Base Cases included; 1) Aquifer Type 2) Aquifer Transmissivity 3) Production Well Depths (to top of screen) 4) Average Daily Yield from Production Wells, and 5) Horizontal Distance of Production Well from surface water source. The water supply assessment based on the Similarity Factor was found to be insensitive to well distance data since the same sequence of similarity was found using the average well distance to the surface water source, and the minimum distance to surface water source. For this reason, and the fact that weighting factors were not well defined, the Similarity Factor data was not considered suitable for this project. Therefore, the overall assessments were based on Similitude Function results. From the results obtained from the Similitude Function, a tangible assessment of each municipal water supply was accomplished. From a numerical and more subjective approach to data analysis, a subsequent percent probability, and likelihood of riverbank filtration was established, respectively. It has been found that the Grand Falls, Edmundston, Sussex and Woodstock water supplies are 'very likely' to be operating with riverbank filtration. Perth/Andover, Hartland, Baker Brook, St. Leonard, and Ste. Anne-de-Madawaska are 'likely'\o be operating with riverbank filtration. It is 'possible' that the water supplies of St. Basil and Fredericton Junction are operating with riverbank filtration. For St. Francois-de-Madawaska and Sussex Corner, it is 'unlikely-but-possible' that riverbank filtration is occurring within their water supplies. From the Case Based Reasoning process used to complete this project, and the respective municipal water supply assessments, numerous recommendations are established accordingly.