Groundwater denitrification by fluidized bioelectrochemical systems

Abstract: Groundwater (GW) accounting for most of the freshwater available around the World, finding sustainable techniques to depollute it is of crucial importance for safe drinking water supply. The extensive use of fertilizers in the agriculture, as well as other anthropogenic activities, are contributing to the excessive nitrate levels in some aquifers. These levels need to be reduced to obtain potable water. Bioelectrochemical systems (BES), using microorganisms to catalyze a desired electrochemical reaction, recently proved to be a very promising technology for water remediation. Groundwater denitrification using Microbial Electrolysis Cell (MEC) needs to be improved for further scaled-up on-site system. The advantages conferred by fluidized bed reactor (FBR), as well as the outstanding electrochemical properties of reduced graphene oxide (rGO), are two potential enhancements of such bioelectrochemical denitrification system that were investigated in this thesis. Some essential parameters could be determined during the preliminary steps' experiments. The fluidization trials gave us a clear insight that Coconut-based Activated Carbon (CAC) particles were resistant carrier particles, nicely fluidized within a 39.27cm3 circular cathodic chamber for a flow rate ranging between 450ml/min to 590ml/min. For the same flow rate of 500ml/min, we could obtain CAC particles fluidization for the upstream fluidized configuration, and still bed particles for the fixed bed downstream configuration, which would be very useful for later unbiased comparison. The denitrifying bacteria showed during their enrichment, a nitrate removal rate of up to 1.986ppm NO3-N/h in serum bottles, with an average of 0.38ppm NO2-N/h accumulation. The parallel running of fixed bed versus fluidized bed denitrifying reactor in order to compare their denitrification performances, was planned, but could not be performed due to COVID-19. The graphene oxide (GO) batch experiments showed a good biocompatibility between GO/rGO and our autotrophic denitrifying bacteria. A change of morphology within about 20 hours was observed, probably suggesting the reduction of GO to rGO by the bacteria. During a first test, the presence of GO led to a 2.7 folds less efficient denitrification performance as compared with the GO/rGO-free condition, likely due to the competition between nitrate and GO for being reduced. However, the denitrification rate in presence of GO/rGO increased up to 1.873ppm NO3-N/h after the second pulse of groundwater and flush with H2/CO2 gas, which is almost 2.3 folds higher than initially in the same condition. This suggests that GO needs some time to get fully reduced to rGO, and the denitrification rate might reach the same or higher levels as in the GO/rGO-free conditions, when GO is fully reduced. Improved denitrification would indicate that rGO facilitates the electron transfer between bacteria and nitrate, as it can be expected from its electrochemical properties previously studied. This would be worth being investigated in the scope of a longer experience.