摘要:
随着北京市再生水补给河湖规模扩大,污水处理厂出水中雌激素活性物质引起的受纳水体生态风险日益受到关注。以雌二醇为例,利用物种敏感度分布(species sensitivity distribution,SSD)模型和联合概率曲线(joint probability curve,JPC)方法开展北京市污水处理厂出水的概率生态风险评价研究。通过文献调研整理了北京市约430个物种,利用美国环境保护署ECOTOX数据库获取了其中7个物种的雌二醇毒性数据,构建了正态分布、对数正态分布、Logistic分布、对数Logistic分布、Weibull分布、Burr III型分布和Gumbel分布等7个SSD模型,评价了北京市污水处理厂二沉池出水以及"混凝-沉淀-过滤-臭氧"、"超滤-臭氧"和"超滤-活性炭-臭氧"3种深度处理工艺组合出水的生态风险。结果表明,利用北京市7个物种雌二醇毒性数据构建的SSD模型具有合理性,SSD模型选择对生态风险评价结果的影响较大,对数正态分布、对数Logistic分布、Weibull分布和Burr III型均是可接受的SSD模型,其中拟合效果最佳的Burr III型分布模型给出了最保守的风险估计。Burr III型分布模型的模拟结果显示,北京市污水处理厂二沉池出水以及3种深度处理工艺组合出水的总体风险期望值分别为0.070、0.040、0.036和0.026,该结果可以为北京市未来水生态保护目标的设定以及污水处理工艺的升级改造提供决策参考。
Abstract:
With the augmentation of rivers and lakes using reclaimed wastewater expanding in Beijing, there have been growing concerns about the ecological risk caused by estrogenic endocrine-disrupting chemicals in the effluents of wastewater treatment plants (WTPs). The ecological risk of estradiol, as an example, in WTPs'effluents was assessed with species sensitivity distribution (SSD) models and joint probability curves (JPCs). About 430 native species in Beijing were collected through literature review, among which 7 species found their toxicity data associated with estradiol in the ECOTOX database developed by US EPA. Seven SSD models were established to fit the available toxicity data including Normal distribution, Log-normal distribution, Logistic distribution, Log-logistic distribution, Weibull distribution, Burr Type III distribution and Gumbel distribution, and then the validated SSD models were applied to assess the ecological risk of estradiol caused by the effluents of secondary sedimentation tanks and 3 advanced treatment processes, i. e. coagulation-sedimentation-filtration-O3, ultrafiltration-O3, and ultrafiltration-activated carbon-O3, from typical WTPs in Beijing. The SSD models based on the toxicity data of estradiol with respect to the 7 native species of Beijing could be reasonably interpreted, and different types of SSD models had significant impact on the results of ecological risk assessment. Log-normal distribution, Log-logistic distribution, Weibull distribution and Burr Type III distribution were acceptable SSD models, and the best fitted model, Burr Type III distribution, gave the most conservative risk estimates. With Burr Type III distribution model, expected total risk of estradiol from the effluents of secondary sedimentation tanks and 3 advanced treatment processes was estimated to be 0.070, 0.040, 0.036 and 0.026, respectively. The results could support future decision-making when defining protection goals of aquatic ecosystems and upgrading and retrofitting wastewater treatment processes in Beijing.