Removal Characteristics Of Pollutants In Landfill Leachate By Two-stage Dtro And Simulation Research Based On Neural Network

Landfill leachate is a kind of high concentration organic wastewater,which is produced by sanitary landfill.It is closely related to the urban production and life.It's characteristics is large amount of production,high content of pollutants and maladjustment of nutrition ratio.If it is not treated properly,it will cause serious pollution in the environment and damage the health of human being.The traditional landfill leachate treatment process mainly used“pretreatment+biochemical treatment”,because the requirements for the biodegradability and stability of treatment system,its development was limited.Membrane technology has been widely applied in the field of leachate treatment because of its stable operation,strong impact resistance,and no requirement for biodegradability.This paper studies the two-stage DTRO process which is used in the leachate treatment station of Anyue landfill in Sichuan.The structure and characteristics of the DTRO are introduced.based on the engineering example,we analyzed the composition of DTRO membrane system and the working principle and operation parameters of the main equipment.During the operation of two-stage DTRO,CODcr,ammonia nitrogen,conductivity and pressure were monitored during November 14^th and December 7^th.The operation situation of the two-stage DTRO system and the removal rate of the pollutants were analyzed through the monitoring data.During the monitoring period,the CODcr average removal rate was99.85%,the ammonia nitrogen average removal rate was 99.65%,the average conductivity removal rate was 99.52%.The CODcr and ammonia nitrogen of effluent water is meets the requirements in the table 3 of the leachate discharge standard in"pollution control standard of domestic waste landfill site"?GB16889-2008?.In this paper,we use the monitored data to establish the BP neural network model to predicte the the concentration of CODcr and ammonia nitrogen in effluent.Among the 44selected samples,35 were training samples,and 9 were test samples,which were used to verify the reliability of prediction results.According to the results of simulation and prediction,the average relative error ofammonia nitrogen is less than 10%.The BP neural network model has better prediction ability.The changing rule of ammonia nitrogen concentration in effluent is discussed in this paper when we change water quality parameters in influent.According to the forecast results the ammonia nitrogen in the effluent from the two-stage DTRO system is affected by the influent ammonia nitrogen,CODcr and conductivity under certain work pressure.When the ammonia nitrogen concentration of the influent increase,the ammonia nitrogen concentration in the effluent increase.When the CODcr concentration of the influent increase,the ammonia nitrogen concentration in the effluent decrease.When the water conductivity of the influent increase,the ammonia nitrogen concentration in the effluent increase.