| Low-pressure membranes are now an interesting and proven alternative for water treatment. Those are particularly efficient for particulate pollution removal including parasites removal. However, they are poorly efficient at removing dissolved contaminants. In order to overcome this weakness, powder activated carbon (PAC) can be added directly into the basin in which hollow-fiber membranes are installed. This work focus on the study of such a process in which PAC acts under biological mode. The process is referred as a high concentration powder activated carbon membrane bioreactor (HCCAP-MBR). The process performance concerning water quality (TOC, DOC, BDOC, NH4+, UV254, THM, HAA) and impacts on operation (transmembrane pressure, chemical cleanings) were investigated at a pilot scale and compared to a membrane bioreactor without PAC (MBR) as well as to a biological granular activated carbon filter.;The HCCAP-MBR was studied at a stable PAC age of 30 days (carbon usage rate (CUR) = 16.1 mg/L), a stable PAC age of 60 days (CUR = 8.1 mg/L) and at an aging PAC (without addition or purge) of over 200 days (CUR < 2.4 mg/L). Overall, the treatment performance of HCCAP-MBR at an age of 30 days or 60 days was equivalent to the performance obtained by the biological GAC filter. Water quality obtained at a PAC age over 200 days is lower, which seems to justify usage of PAC at a controlled age. At a PAC age of 30 days, the average reduction of THM-UFC and HAA-UFC is respectively 15% and 65%. This reduction is equivalent to that obtained by the biological filter (23% and 50% respectively) and significantly higher than the reduction achieved by the MBR without PAC (-27% and 35% respectively). DOC and BDOC reductions reach respectively 32% and 60% under inter-ozonation condition during the summer (about 25°C). Under moderate temperature condition (about 15°C) at a PAC age of 60 days, DOC and BDOC reduction decreased to 22% and 33% respectively.;High concentration PAC in a MBR led to a much faster fouling as compared to the MBR without PAC. Biological PAC also caused a higher filtration resistance proportional to the flux which accounted for 0.4 psi at 25 LMH. The selected operational flux of 25 LMH seemed to be higher than the optimal flux and resulted in the need of a chemical cleaning each three months. The chemical cleaning was preceded by a physical cleaning which allows the removal of the PAC cake layer. This physical cleaning alone allows a low flux recovery, which accounted for less than 1 psi. inter-ozonation proved to be an efficient strategy to mitigate fouling, which appeared to lower its progression by an approximate factor of 2.4. Addition of fresh PAC on a daily basis also helps to reduce fouling compared to a biological PAC of the same age without addition or purge.;In summary, the HCCAP-MBR has demonstrated the quality of its effluent. However, further strategies must be developed in order to mitigate fouling, such as filtration cycle optimization, usage of specific membrane materials combined with specific PAC types and usage of specific water pretreatments. It remains that the process could be an interesting alternative to the biological GAC filters.;The HCCAP-MBR was studied at an hydraulic retention time of 79 minutes using a PAC concentration of 10 g/L for 300 days. In warm water, the HCCAP-MBR as the MBR allowed full nitrification. High concentration PAC has accelerated the development of nitrifying biomass and extended nitrification during cool water conditions, much longer than it was possible without PAC. However, in cold water, at 5°C, the HCCAP-MBR was unable to remove the NH4 +. |
PDF Full Text Request