污染控制技术及原理

  • 郝晓地,胡沅胜,王克巍.生物营养物去除(BNR)工艺模型参数校正与验证方法[J].环境科学学报,2010,30(3):463-469

  • 生物营养物去除(BNR)工艺模型参数校正与验证方法
  • Methodological study on calibrating and validating model parameters of biological nutrient removal(BNR) processes
  • 基金项目:国家高技术研究发展计划(863)项目(No2006AA06Z320);北京市属高等学校人才强教深化计划——高层次人才资助项目(NoPHR20100508)
  • 作者
  • 单位
  • 郝晓地
  • 北京建筑工程学院可持续环境生物技术研发中心, 北京 100044
  • 胡沅胜
  • 北京建筑工程学院可持续环境生物技术研发中心, 北京 100044
  • 王克巍
  • 北京建筑工程学院可持续环境生物技术研发中心, 北京 100044
  • 摘要:研究了结合活性污泥系统物料平衡的生物营养物去除(BNR)工艺模型参数校正与验证方法.首先采用代尔夫特理工大学开发的TUD模型及其参数缺省值进行初步模拟,并将模拟结果与实测数据进行对比以明确模拟偏差.同时,对原始运行数据进行物料平衡衡算,以消除测量误差并获得TP排放量、反硝化量及氧的净消耗量等平衡计算值,为模型校正提供依据.初步模拟结果显示,TP排放量(3913mg·d-1,以P计)和反硝化量(16170mg·d-1,以N计)均超出了平衡计算值的误差范围(分别为(3762.0±43.1)mg·d-1和(14150±237)mg·d-1),且沿程溶解性COD(SCOD)、PO43--P、NH4+-N、NO3--N不能完全与实测数据吻合,表明工艺模型参数的默认值需得到必要的校正.结合TP平衡,先修正污泥龄(SRT)的计算误差;应用平衡后的运行数据,将SRT由初始值10.3d调整为10d.SRT调整后,再次模拟得出的TP排放量为3750mg·d-1,已在平衡计算结果的范围之内,TP平衡得到了闭合.将发酵过程最大反应速率常数qfe从3d-1提高到15d-1后,沿程及PO43--P拟合结果同时得到了显著改善.通过调整DO及NH4+-N的亲和系数KNH4(从1mg·L-1至0.2mg·L-1),沿程NH4+-N得到了拟合.出水及沿程NO3--N的拟合通过调整氧的亲和系数KO2从0.7至0.1mg·L-1)得以实现.最后校核反硝化负荷及氧的净消耗量,模拟计算的反硝化负荷为14090mg·d-1,氧的净消耗量为126900mg·d-1,均在平衡计算值的误差范围之内(分别为(14150±237)mg·d-1和(129700±2910)mg·d-1),说明校正过程合理、可靠.
  • Abstract:Asystematic method for calibrating and validating model parameters of biological nutrient removal(BNR) processes was developed.The TUD model and its default parameters were applied for initial modeling,and comparisons were made between the simulation results and the experimental data.The raw operational data were checked with mass balance calculations to eliminate errors caused during measurements and to exactly calculate the amounts of discharged TP,denitrified Nand net consumed O2.The initial simulation results indicated that both the simulated amounts of discharged TP(3913 mg·d-1,as P)and denitrified N(16170 mg·d-1,as N) were beyond the ranges of measuring errors(3762.0±43.1) mg·d-1 and(14150±237) mg·d-1,respectively).But the simulated profiles of soluble COD,PO43--P,NH4+ -Nand NO3--Nwere not in accordance with the measured ones.The disagreements indicate that the default parameters of the TUDmodel need to be adjusted.First,the calculated error in the initial SRT was calibrated in combination with the TPbalance.When the initial SRT(10.3 d) was adjusted to 10 d,the TPbalance was closed.Both the simulated profiles of PO43--Pand soluble CODwere significantly improved with qfe(maximal fermentation rate constant) adjusted from the default value of 3 d-1 to 15 d-1.The simulated profile of NH4+ -Nwas matched with DO and KNH4 adjusted from 1 mg·L-1 to 0.2 mg·L-1).Moreover,the simulated profile of NO3--Nwas also matched with KO2 adjusted from 0.7 mg·L-1 to 0.1 mg·L-1.After these adjustments,the simulated denitrified Nand net consumed O2 were at 14,090 mg·d-1 and 126,900 mg·d-1 respectively,falling in the ranges of balanced errors(14150±237) mg·d-1 and(129700±2910) mg·d-1,respectively).The results demonstrate that the calibrating processes were reasonable and reliable.

  • 摘要点击次数: 3216 全文下载次数: 2839