Impacts des conditions d'exploitation sur les performances des filtres au charbon actif biologique

The water treatment plant of Sainte-Rose, Laval, QC, has been using biological filtration since 1986. In 1997, chlorination has been introduced upstream of six biofilters in order to reduce maintenance costs of the pumping station, which follows post-ozonation. Typically, a dose of 0.3 mg Cl2 / L is added at the output of the ozone contact towers and the granular activated carbon (GAC) in the filters have not been changed for many years, as many other treatment plants where biological filtration is used.;The pilot used for these tests was made of four columns (2 in.) with an empty bed contact time (EBCT) of 11 min. Three granular activated carbons of different ages (1, 15 and 20 years) were tested in filtration columns in this pilot unit to compare their biological performances. The performance of a column supplied with a chlorinated influent was also followed. The main parameters taken into account were total organic carbon (TOC), biodegradable dissolved organic carbon (BDOC), ammonia nitrogen, chlorine demand (24) and THM-UFC. The analytical monitoring that were also made included solid analysis of the carbon colonization with the measurement of the potential nitrifying and heterotrophic activities. These measurements were made in order to determine which carbon gives the better water quality and to evaluate the effects of chlorination on the filtration performances.;The comparison between a column and an industrial filter gave the following results: the difference between both concerning TOC removal remained relatively constant during the monitoring period (between 0.03 and 0.09 mg / L) but TOC removal of the full-scale filter was slightly higher. The difference between the EBCT of the column (11 min) and the filter plant (17.9 min) could be the origin of this difference. No significant difference between both for the BDOC removal (> 40%) was observed. The removal of ammonia nitrogen was about 100% for the column and for the filter. A slightly better reduction of THM-UFC was observed for the full-scale filter.;The main results obtained by comparing a column with a non-chlorinated affluent and a chlorinated affluent with 0.3 mg Cl2 / L did not showed large differences in their biological performances. The average removal of TOC is similar (10%) for both and the removal of ammonia nitrogen is also above 90%. Then, an influent with increasing concentrations of total chlorine (0.3 to 1.4 mg Cl2 / L) was applied. Results showed that for total chlorine concentrations below 0.5 mg Cl2 / L, the residual chlorine was under the form of mono and dichloramine. When the concentration of residual chlorine was above this value, we observed the presence of free chlorine. The higher the concentration of total chlorine in the influent was, the more the reduction of ammonia nitrogen decreased inside the column. The final reduction is about 90% for all concentrations of chlorine in the influent. Moreover, the biomass density decreased in the surface of the column for a TCFV between 1 and 2 min when the concentration of total chlorine in the influent increased from 2.82 to 0.65 mg N/h*cm3. No difference appeared for EBCT included between 7 and 8 min. The concentration of nitrifying biomass of between 5.17 and 5.60 mg N/h*cm3, was similar for each test.;The main objective of this project was to study the impact of various operating parameters on biological performances of the filters in order to understand why CAG filters efficiency has declined in recent years in the Sainte-Rose water treatment plant. The effects of a pre-chlorination of the biofilters on their biological performance and the effects of the carbon's age on its efficiency for the removal of biodegradable compounds were also studied in this project. A comparison between a full-scale biofilter and a filtration column in a pilot unit has been performed in order to validate the results.;Carbons aged of 15 and 20 years are statistically more effective in the removal of TOC with a respective average of 9.9 and 10.1% than carbon aged of one year (7.6%). However, the concentration differences are still low, less than 0.09 mg / L for the three effluents. The three columns have a nitrogen ammonia removal around 100% in warm waters. For a water temperature of 11°C, the removal of ammonia nitrogen at the end of the column containing the media of 20 years was only 62%. Finally, regarding the experience of measuring the maximum capacity for biomass attachment, it appeared that the carbon aged of 1 year was more colonized by nitrifying biomass (25,3 microg N/h* cm 3) than the carbon aged of 20 years (4,2 microg N/h*cm3). (Abstract shortened by UMI.)...