Experimental Study On Improvement And Regulation Effect Of Bio-retention System Packing

With the rapid development of urbanization,on the one hand,the original hydrological balance has been changed,on the other hand,the problem of water environmental pollution caused by source pollution has become increasingly prominent.As a typical low impact development(LID)technology,bioretention system has been widely used to absorb and purify rainfall runoff,especially in regulating water quality and groundwater recharge.Through hydraulic water quality design and different mixing of fillers,the pollution of pollutants in rainfall runoff can be controlled from the source.In this study,a large number of literatures were consulted and the existing research results of the research group were used for reference.Four kinds of modifiers,fly ash,water treatment residue(WTR),peat soil and coconut bran,were selected.Six small-scale biofilter columns(1#Soil,2#BSM,3#BSM+fly ash,4#BSM+WTR,5#BSM+peat soil,6#BSM+coconut bran)were built by mixing four single modifiers with bioretention soil media(BSM)as the basic medium.Through nine water distribution experiments,the water reduction rate and pollutant load reduction rate of biological filter column were explored,and the orthogonal test difference analysis and correlation analysis were carried out on the test results.The main conclusions of the study are as follows:(1)When the rainy peak comes,the 3#(BSM+fly ash)and 4#(BSM+WTR)biofilters have about 10 cm of accumulated water respectively.The remaining biofilters have no obvious accumulated water phenomenon,and can exert the maximum water retention effect under the condition of saturated infiltration in 9 simulated rainfall runoff experiments.Among them,4#(BSM+WTR)has a higher average water reduction rate but a slightly larger fluctuation,ranging from 21.34%to 52.36%,and the water reduction rate is the best under three different concentrations of high,medium and low.(2)The removal rates of COD loads by the five biofilters with 2#?6#ranged from 40%to 40%,fluctuating greatly and the effect was not good,mainly due to the addition of POPs;the removal rates of TN pollutants were good,and the removal rates of TN pollutants were above 54%;the removal rates of TP loads were very stable,all above 99%.3#(BSM+fly ash)and 4#(BSM+WTR)biofilter columns showed the best load removal rate for general pollutants.(3)Among the 2#?6#,3#(BSM+fly ash)and 4#(BSM+WTR)had the best removal efficiency for Cu,Zn and Cd,while 6#(BSM+coconut bran)had the highest removal efficiency for Cu(95.52%),Zn(91.94%)and Cd(87.10%)with relatively poor removal efficiency that in each filter column is relatively poor.(4)Orthogonal test difference analysis showed that water volume and COD,TN,TP,Cu,Zn had significant relationship with pollutant water distribution concentration in 2#,but Cd had no significant relationship with the three factors;in 3#,water volume,TN and three factors were non-significant factors,while COD,TP and heavy metals had significant relationship with water distribution concentration;in 4#and 5#,water removal had no relationship with the three factors.In addition,COD,TN,TP and heavy metals were significantly correlated with the factors of water distribution concentration of pollutants;in 6#,water volume,Cd had not significant with all three factors,while COD,TN,TP,Cu,Zn were significantly correlated with the factors of water distribution concentration.(5)The cumulative input of water,COD,TN,TP,heavy metals and cumulative reduction in 9 orthogonal experiments were modeled and analyzed.It was found that there was a strong correlation between cumulative input and cumulative reduction.Then,the first seven sub-water quality and water quantity test results are used to model and the remaining two sub-water quality and water quantity test results are used to predict.The results show that the cumulative reduction can be effectively predicted by the cumulative input of water volume,COD,TN,TP and heavy metals.The overall prediction results are good,and the relative error is between 0%and 12%.But according to the forecasting results of the load reduction estimation model,the reliability of the model decreases gradually with the increase of the number of tests.