Experimental Study On The Combined Process Of Denitrifying Biological Filter-aquatic Plant Pond For Low Pollution Water Purification

With the continuous and rapid development of economy,people's requirements for high water quality of water bodies such as lakes and rivers are gradually increasing.The treatment of low-pollution water is an important aspect to improve the water quality of rivers and lakes.The existing technologies for low-pollution water treatment have their advantages and disadvantages.How to remove nitrogen and phosphorus from low-pollution water economically and effectively is the focus of current research.In this research background,the low-pollution water was selected as the research object in this paper,and the denitrification biological filter?DNBF?and aquatic plant pond process were used to treat the low-pollution water.In the DNBF experiment,the feasibility of new natural carbon sources as denitrification carbon sources was investigated,the effects of different filter and operating conditions on the treatment effect of DNBF were studied,and the composition of microbial community structure in DNBF was analyzed.In the study of aquatic plant ponds,the contribution rate of pollutants absorbed by submerged plants,the contribution rate of pollutants degraded by plant rhizosphere microorganisms,and the difference of microbial community structure were analyzed under dynamic conditions to explore the removal mechanism of pollutants in low-pollution water by aquatic plants.The treatment effect of DNBF and aquatic plant ponds combined process for low-pollution water was studied,and the concentration of pollutants in the effluent of the above process was predicted by artificial neural network?ANN?model.The main research contents and results are as follows:Corn cob,poplar branches,wood chips and straw were used as carbon source to explore its feasibility as denitrification carbon source.The effects of different filter materials?ceramsite,quartz sand and polypropylene?on denitrification were studied.The results showed that the denitrification potential of alkali treated corn cob was the highest,which was 143.07 mg NOx^--N/?g·material?.In the static denitrification experiment,the TN removal rate was 69.70%when alkali treated corn cob was used as denitrification carbon source.The carbon releasing capacity and denitrification potential of the alkali treated corn cob were greatly improved compared with that of the natural corn cob.When the dosage of alkali treated corncob was 20 g and HRT=2 h,the denitrification effect of DNBF with ceramsite as filter material was the best,the effluent TN could meet Class IV of the Environmental Quality Standard for Surface Water?GB 3838-2002,China??EQSSW?.relative abundance of denitrifying bacteria in the upper,middle and lower layers of DNBF using ceramsite as filter were23.13%,8.45%and 19.46%,respectively.The relative abundance of Thauera accounted for 19.78%,4.29%,and 13.25%,respectively,which was much higher than that of the other DNBFs.Among the three DNBFs filled with different filter materials,the relative abundance of denitrification bacteria was the highest in the lower biofilm of the filter,indicating that in the up-flow DNBF,denitrification bacteria were concentrated in the lower layer.Three kinds of submerged plants?Hydrilla verticillata,Vallisneria natans,and Potamogeton wrightii Morong?were used to construct plant ponds for the removal of nitrogen,phosphorus,and organic matter in low-pollution water.The bacterial community structure in these plant rhizospheres was analyzed.Results showed that the maximum removal rates of total nitrogen?TN?and total phosphorus?TP?were observed in the Hydrilla verticillata plant ponds at 92.83%and 82.66%,respectively.Under the conditions of three different hydraulic retention times?HRT??4 days,6 days,and 8 days?,the maximum removal rates of TN and TP in each of the three kinds of plant ponds increased with increasing HRT.The absorption contribution rates of the three plants were 16.22%?Vallisneria natans?,20.38%?Hydrilla verticillata?,and16.97%?Potamogeton wrightii Morong?for TN.For TP 19.16%?Vallisneria natans?,18.88%?Hydrilla verticillata?and 21.06%?Potamogeton wrightii Morong?.The relative abundances of Proteobacteria in the plant rhizosphere of Potamogeton wrightii Morong,Vallisneria natans,and Hydrilla verticillata were 59.70%,88.57%,and 68.57%,respectively.Proteobacteria played an important role in nitrogen removal for the three submerged plants.The relative abundances of heterotrophic denitrifying bacteria Rhodobacter for each of the rhizospheres were found to be 7.74%?Potamogeton wrightii Morong?,3.19%?Vallisneria natans?,and 13.87%?Hydrilla verticillata?,where denitrification was apparent.Due to the poor phosphorus removal effect and the high concentration of TP in the effluent of DNBF,the pollutants removal effect of low-pollution water treated by the DNBF and aquatic plant pond combined process was studied.The effluent water quality of COD_Cr,NH₄⁺-N,TN and TP of the low-pollution water treated by the combined process could meet Class IV of EQSSW.At the same time,artificial neural network?ANN?model was used to predict the effluent concentration of each pollutant in the treatment of low pollution water by the three processes mentioned above.ANN model for the treatment of low pollution water by the above three processes was established by using quasi-newton BP algorithm?trainbfg?,quantized conjugate gradient algorithm?trainscg?and L-M optimization algorithm?trainlm?,respectively.The training results showed that the training accuracy of L-M optimization algorithm?trainlm?was the highest.The L-M optimization algorithm?trainlm?was used to predict the effluent pollutants of the three processes.The mean square error?MSE?of TN and TP in the predicted results was less than 1.00.Therefore,the ANN model established in this paper had high prediction accuracy and could be used to predict the effluent pollutant concentration of the three processes.