Font Size: a A A

Study On Dairy Wastewater Treatment By AMBR And Effluent Quality Forecast By BP Neural Network

Posted on:2018-06-18Degree:MasterType:Thesis
Country:ChinaCandidate:Y J YanFull Text:PDF
GTID:2321330518979618Subject:Engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of large-scale livestock farming,livestock and poultry breeding industry has to become the major sources which cannot be ignored.The excessive fecal urine concentration and the increasing number of waste flushing water problems,bring great pressure to the local ecological environment.Usually,the Effluent indicators are all use the manual detection,and the detection time is longer and has lags,leads to substandard sewage discharge,it is not conducive to the optimization of sewage treatment system parameters and the improvement of process.Aiming at the above problems,treatment of livestock wastewater by AMBR process(the combination process of Anaerobic Baffled Reactor and Membrane Bio-reactor),to analysis and research the pollutant removal characteristics and removal efficiency,to optimize the parameters of aquaculture wastewater treatment system and improve the treatment efficiency.Using BP neural network to simulate and predict the water quality,and change the input parameters(DO,MLSS,pH)to predict the effluent quality.To realize the dynamic detection of the aquaculture wastewater and optimize the parameters of aquaculture wastewater treatment system,for reduce the environmental pollution of livestock and poultry breeding.The main results are as follows:(1)The average total removal rates of CODcr,NH3-N,TN,TP by using AMBR system are 94.17%,84.17%,83.37%,41.63%and 98.33%,respectively.According to the size of the removal rate is:SS>CODcr>NH3-N>TN>TP.The AMBR system has a good effect on the treatment of the fluctuation of influent concentration(CODcr is 1298.29mg/L-3372.42mg/L,NH3-N is 79.35mg/L-207.6mg/L,SS is 809.9mg/L?32396.4mg/L).AMBR system reflects the strong impact resistance and adaptive adjustment and recovery capabilities(2)The removal of CODcr,NH3-N and TN by AMBR system mainly depends on biodegradation,the contribution rates of biodegradation were 60%?85%,82%?90%and 69.96%?90.44%,respectively.Because NH3-N and TN molecules are smaller,they can pass through the microporous membrane module and discharge with water,the removal of NH3-N mainly depends on nitrification of nitrobacteria,the anaerobic ammonia oxidizing bacteria use nitrite as electron acceptor to convert ammonia nitrogen into nitrogen.The removal of SS depends on the membrane system,and its contribution rate is 61.26%.Mainly due to the use of AMBR membrane module is the membrane pore size of 0.1?0.2 microns ultrafiltration membrane,the particle size of colloid,suspended substance SS and high molecular organic matter in aquaculture wastewater were greater than 0.2 microns,all of them were trapped by the physical filtration.(3)there are two reasons of the TP removal rate was significantly lower than other water quality indicators,on the one hand,the internal environment of the system was destroyed by high concentration organic matter,the carbon source supply of phosphorus accumulating organisms is inhibited by biological competition,so the amount of gasification phosphorus is reduced.On the other hand,the DO concentration in anoxic zone was about 0.45?0.6 mg/L,higher DO concentration than anaerobic phosphorus removal conditions(DO=0.2mg/L),the growth of non polyphosphate bacteria inhibited the denitrifying effect of phosphorus accumulating bacteria,the removal efficiency of TP was inhibited(4)The mean relative error of effluent CODcr,and NH3-N,TN,TP and SS were 0.99%,1.67%,0.08%,0.02%and 0.03%respectively.And the absolute mean error rate is 1.01%,2%,8.1%,2.12%and 4.99%.It can be concluded that the BP neural prediction model has good adaptability and accuracy for those indicators,and the best performance is the CODcr prediction model.(5)When the DO concentration increased,the pH value increased(not more than pH=8),the MLSS is increased the removal of TN,NH3-N,and SS was enhanced.When the operating conditions are adjusted to DO=3mg/L,MLSS=16000 mg/L,pH= 7,the livestock wastewater treatment system to maximize the degree of purification,reach the best treatment effect The removal effect of CODcr,NH3-N and TN increased slowly during the course of MLSS from 10000 to 16000.For saving money and ensure the effect of sewage treatment,without considering other pollutants water quality parameters,MLSS to 10000?12000 to CODcr and the best removal effect of nitrogen;it can achieve the best removal effect of CODcr and nitrogen by MLSS only needs to control to 10000?12000.
Keywords/Search Tags:AMBR, aquaculture wastewater, BP neural network, Water quality prediction
PDF Full Text Request
Related items