Wan Yuwen, Guo Changqiang, Mao Zhi, Li Xinjian, Cui Yuanlai, Zhao Shujun. Sewage purification effect of multi-series surface flow constructed wetland[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(3): 220-227. DOI: 10.11975/j.issn.1002-6819.2016.03.032
    Citation: Wan Yuwen, Guo Changqiang, Mao Zhi, Li Xinjian, Cui Yuanlai, Zhao Shujun. Sewage purification effect of multi-series surface flow constructed wetland[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2016, 32(3): 220-227. DOI: 10.11975/j.issn.1002-6819.2016.03.032

    Sewage purification effect of multi-series surface flow constructed wetland

    • Abstract: Rural domestic sewage has become one of the main sources of agricultural non-point source pollution. In order to develop effective treatment of domestic sewage in less development rural areas with low cost in construction and simple in maintenance, we studied a multi-series system of surface flow constructed wetland to treat rural sewage in Qingshitan irrigation district in Guilin City, Guangxi Province, China. The purification effect of different subsystems and the nutrient concentration variation at each stage were studied and analyzed. The results showed that the nutrient removal efficiency of the overall wetland system presented a trend of gradual decline over time. At the same time, the greater the surface area of the wetland was, the higher the removal rate of nitrogen and phosphorus were. However, the differences of removal rate of nitrogen and phosphorus affected by temperature could be ignored because the change of daily temperature was relatively stable during the test period (April to September). Total nitrogen (TN) and NH4+-N (ammonium nitrogen) was high at earlier stage and low at late period of the inlet for the whole study period. The average removal rate decreased from 83.8% and 84.7% to 65.0% and 57.0%, for TN and NH4+-N, respectively. There was no significant correlation between inlet concentration and removal rate (P>0.05) for nitrogen. The inlet concentration of TP (total phosphorus) showed no obvious change, but its removal rate presented sharply decline from 82.2% at the inlet to 21.2% at the outlet. This was caused by a strong phosphorus adsorption by sediment and phosphorus use by microorganisms. Nevertheless, there was still 21.2% TP was not removed. This demonstrated limitations of TP removal capacity on surface flow constructed wetlands in comparison with the subsurface flow constructed wetlands. The average removal rate of TN, NH4+-N and TP were 76.3%, 70.8% and 60.5%, respectively of the whole system. There could be a better purification result with a larger aspect ratio and surface area. This study also revealed that this wetland system displayed a fast and stable purification effect for high nutrient influent concentration, but less effect for removal low nutrient concentration. From the experiment, we received ten sets of concentration data, including the data of each wetland bed. Of them six groups of measured data were used for the curve fitting of nitrogen and phosphorus concentration. We established the exponential, the linear, and the quadratic polynomial model which related to the inlet concentration of the wetland system through the monadic linear regression analysis of model coefficients. Then these three models were used for prediction and evaluation on the basis of the other four groups of testing data. Correlation analysis revealed that exponential decay model had a smaller relative root-square-error value with the best performance. The wetland system showed good purification effect during the whole test period, reducing the emissions of nitrogen and phosphorus of rural domestic sewage greatly, and mitigating the damage of domestic sewage to freshwater environment. The establishment of exponential attenuation model provided reference for the prediction of concentration changing along the multistage tandem constructed wetland system. The model can be used to easily predict nutrient concentration in any wetland subsystem when the influent concentration of the system is known.
    • loading

    Catalog

      /

      DownLoad:  Full-Size Img  PowerPoint
      Return
      Return