Study On The Total Pollutant Load Allocation And Its Margin Of Safty Under Uncertain Conditions

Total pollutant load is an effective means to meet the water quality standard. It is the technical basis to allocate the pollutant load to the pollutions sources or areas. Due to the allocation method, subjective decision methods and so on, the total pollutant load is affected by uncertainty. So it is very important to estimate the uncertainty in the total pollutant load to improve the scientificness and reliability of the allocation results.After review of the domestic and foreign research on the uncertainty analysis study on the total pollutant load, a multi-objective total pollutant allocation model of Bohai Sea was built, In the model, the wighting coefficients are the important factors affecting the allocation results. But the coefficients showed obvious uncertainty because they were usually determined by human judgement. To overcome the uncertainty in the weighting coefficients and ensure the reasonability and fairness in the allocation process, the HSY method was used to analyze the sensitivity of the model coefficients. Then the subjective uncertainty was analyzed on the basis expert survey. Thereafter the effects of the reasonability index and its construction factors on the allocation results were studied. Finally, the margin of safty in the allocation model coefficients was given based on the expert survey and Monte Carlo method. It showed that:(1)The sensitity of the6weighting factors were different. For TN, the sensitivity order was:Environmental capaicty> Field area> Population> Water resource> Presnet load>GDP; for TP, the order was:Field area> Population>GDP> Water resource> Presnet load> Environmental capacity.(2) The expert judgement can affect the allocation result. According to the weighting coefficient investigation from30experts, which were devided by3groups, the total loads were allocated and compared for the Bohai Sea. TN allocation amount average aws16400t/a with the the standard deviation of1600t/a and coefficient of variation of9%. The maximum and the minimum average relative deviation ratio were31%and-13%respectively.The average of the TP allocation amount is1200t/a with the standard deviation of140t/a and coefficient of variation of12%. The maximum and the minimum average relative deviation ratio were32%and-19%respectively. So the allocation loads were affected by the expert judgement.(3) The fairness index is negatively correlated with the economy index. When the index of fainess was changed14.26%and14.24%for TN and TP, the index of economy was changed47.89%and47.16respectively. More emphasis on the fairness would reduce the economy and increase the treatment cost.(4) Based on the expert survey and the Monte Carlo analysis, the margin of saftery introduced by the weighting coefficients was calculated. When the variation range for the coeffcients was±5%, the margin safety for TN and TP was1.1%and2.3%. When the variation range for the coeffcients was±10%, the margin safety for TN and TP was2.4%and4.6%. It showd that the uncertainty introduced by the weighting coefficients was under control and would not affct greatly the reaiablity of the allocation results.