Study On Characteristic Of Urban Runoff And The Technology Of Constructed Wetland For Urban Runoff Treatment

Urban runoff treatment is of great importance for non-point pollution prevention, conservation of urban water resource, enhancement of stormwater utilzation and alleviation of municipal water crisis. Constructed wetland is a treatment technology fit for the characteristic of urban runoff, and its application in urban non-point source pollution control and stormwater utilization is urgently expected. However, present researches on constructed wetland for urban runoff treatment is scarce, expecially the research based on the characteristics of running, hydrology and hydraulics of this treament system. As there is lack of basical theory and patameter for design and runing, the development and application of urban runoff treatment constructed wetland are restricted. Precipitation hydrology for different urban cover in Chongqing has been studied and the Characteristics of water qualtity and quality of urban runoff are monitored and analyzed in this study. Demonstration project of palmspring constructed wetland for treatment of urban runoff is under long-period investigetion and its performance trait and problems are discussed. A pilot constructed wetland treating modeled urban runoff is built for experiment study, and performance, mechanism and effecting factors of pollutant removal under intermittent and constant inflow condition is investigated. Under the condition of modeled storm shock loads, tracer study and contaminants removal performance study have been carried out to investigate the hydrodynamics and pollution removal potentials of urban runoff treatment constructed wetlands. Nomalized RTD model and pollutant removal model based on residence time distribution for constructed wetland under non-steady-flow conditions.Research on hydrology of urban storm runoff indicate that characteristic of catchment cover is one of the most important factors which affect the qualtily of urban runoff, and control of organic matter and suspended solid should be given priority in urban runoff pollution prevention and utilization. Rank of average mean event concentration(EMC_A) of COD_Cr, NH₄⁺-N, SS, TP, turbidity and Pb in different catchment are: roof runoff-1 respectively, and annual load of SS are 118.8, 1058.7 , 2658.5and 7410.6 Kg.ha^-1. yr-^-1 respectively. And EMC_A of COD_Cr are 28.86, 110.59, 179.44 and 394.68mg. L^-1 respectively, and annual load of COD_Cr are 352.0, 714.1, 1673.5 and 4530.3Kg.ha^-1.yr^-1 respectively. Though stochastic, pollution process of urban storm runoff showes very obuvious first flush effect, which is correlated with, rainfall, precipitation type and pollutant type. First flush rainfall is a parameter characteristic of storm runoff pollution, which based on first flush theory and focused on totall pollution control. A definition of first flush rainfall based on approximate maximal precipitation event is elicited as H₀/Y%/H_max, that is for precipitation event with rainfall less than H_max (mm), the first rainfall of H0 (mm) would carriy pollutants more than Y%of the total pollutants in the whole storm event. This definition of first flush rainfall, which has taken total rainfall into consideration, is more definite campared with conventional denfinitionsExperimental study in intermittent and constant inflow condition indicate that with constant inflow load of 0.1114m3.m^-2.d^-1(HRT 36h)and water depth of 0.4m, the pilot constructed wetland showed a good pollutants removal performance for COD_Cr,SS,Tp and Pb. Annual removal loads and rates of these pollutants are 2.128kg.m^-2.yr^-1and 79.6%, 3.581kg.m^-2.yr^-1and 94.3%, 13.826g.m^-2.yr^-1and 59.3% and 2.101g. m^-2.yr^-1and 76.8% respectively. Impact of temperature and season on removal performance of COD_Cr, SS, NH₄⁺-N and Pb is not obvious except for TP. Influence rank of running parameters investigated is: HRT>water depth>resting period, and the most optimal running condition is under HRT of 36h and water depth of 0.4m. Total solid accumulation in substrates of constructed wetland after one year increase about 2.06 times and average 10.02mgTSS.g^-1substrate. Nitrification is the most critial process in nitrogen removal, which is affected by ammonification of organic compounds and compitition of DO between nitifiying bacteria and heterotrophic bacteria. Removal rate and load of ammonia are 38.4% and 21.632g. m^-2. yr ^-1. Removal of TP and Pb by plants assimilation are 1.816g.m^-2.yr^-1and 16.579 mg.m^-2。yr^-1, which contribute to 13.2% and 0.79% of the total removal respectivly. TP and Pb sink into solid accumulation in substrates of constructed wetland are 5.372 and 1.586g.m^-2.yr^-1, which contribute to 38.9% and 75.02%of the total removal respectivly. In resting period of intermittent running condition, concentration of COD_Cr in decreased with increasing resting days, the value of UV₂₅₄/COD_Cr , however, increased. DO concentration increased during resting period, and the concentration of ammonia decresed rapidly, which indicate that with no organic pollutant inflow and natural restoration of DO in resting period, competition potential of nitifiying bacteria and nitrification are enhanced. pH value decrease with increasing of resting days, and phosphors release is observed in constructed wetland and have adverse effect on phosphors removal in subsequent influent.Residence time distribution of constructed wetland is unstable in different hydrologic conditions such as constant flow and pulsed flow. Time variableφ( (φ= V _sys/1∫^t_λQ(t)dt )based on flow rate is used to normalize RTD, and the normalized precedure remove the effects of unstable flow rate, isolating the dipersive and mixing characteristics of the system. Tacer experiment indicated that normalized RTD is relatively stable and not sensitive to effects of flow rate and tracer injection amount. The normalized RTD is modeled with a series of continuous stirred tank in parallel with a plug flow tank with dispersion(CSTRs+PFD). There are four paramters in this model, which can regulate peak time, peak height, lean dgreen, mean residence time and distribution variance of RTD.