The Startup Of Anaerobic Baffled Reactor And Research Of Microbial Ecology In ABR

This paper analyzed the characteristics and performance of the startup of anaerobic baffled reactor(ABR) under high organic loading rate(OLR), and the microbial ecology developing in every compartment was investigated simultaneously. The major research in this paper involved the characteristics and water quality in the startup of ABR under high organic loading rate, the shift of microbial ecology in the startup of ABR and the establishment of mathematical model by artificial neural network. The startup of ABR under high OLR. The reactor configuration of ABR confers considerable resistence to organic shock loads and results in high removal efficiencies even at high organic loading rates, however, many literatures are concerning of the startup of ABR under low OLR. This paper investigated the performance of the reactor and the profiles of water quality index such as COD,pH and VFA in every compartment during startup. After several months of hard work, the configuration, manufacture and startup of ABR was accomplished. The result demonstrated that ABR could startup successfully at high OLR by feeding granular sludges which had good quality. The COD removal rate could reach more than 50% at the 9th day after inoculation, and maintained 80% after 2 weeks, which indicated the startup was successful. Meanwhile, the anaerobic digestion system could proceed an acidification process if the OLR was too high. This paper demonstrated that if the loading rate was between 3.5 kgCOD/m~3 and 6.5 kgCOD/m~3, the removal efficiency was good enough and the acidification process could be avoided. (2) the microbial ecology of ABR The hydraulic characteristics of ABR drive the acidogenic bacterial and methanogen distributed appropriately in different compartments. By far the most common observation of microbial ecology in such reactor was based on the optical microscope and electronical microscope. For accurately describing microbial populations, recent advanced molecular biology technique was employed to monitor the shift of Archaeal community in every compartment and provided detailed descriptions of the complex Archaeall populations present.9 operational taxonomy units were obtained . The 3rd and 4th compartment incorporated the whole OTU, while the Archaeal biodiversity of the first compartment was the least, and the second compartment was second to the least. The dominant Archaeal in every compartment is the same. (3)the establishment of BP neural network model The BP model was established by using the water quality index of influent and effluent as input and output functions. The convergency network model was obtained after several times of trainings. The result of simulation revealed that the relative error between predicting data and actual data was in the range of 1.23%～4.93%,the network reflected the condition of influent and effluent. The network is stable and the result is credibility.