AbstractOne of the biggest challenges in characterizing clandestine groundwater pollution sources is correctly estimating the number of such sources and their physical locations. Indiscreet underground disposal of toxic waste leading to groundwater pollution is widespread, and it is extremely difficult to detect the number of such sources and their locations. In this study, a real-life scenario of groundwater pollution in suburban New South Wales, Australia, is solved using a simulated annealing (SA)–based linked simulation optimization (LSO) technique with Benzene, Toluene, Ethylbenzene and Xylene (BTEX). The developed LSO considers the number of pollutant sources and their locations as explicit unknowns. This overcomes a crucial limitation in the earlier methods that limited the search to a few known potential source locations. The results show the applicability of the developed methodology to real-life groundwater pollution problems without making limiting assumptions about the number of sources or potential source locations being known, which has not been addressed to date.Practical ApplicationsThe developed methodology can potentially improve the process of remediation of polluted groundwater aquifers by simultaneously identifying the number of pollution source locations and release flux histories. The results of the study can be used to assist remediation engineers in developing an appropriate reclamation approach for a polluted aquifer, where knowledge of groundwater pollutant source characteristics is vital. The suggested method will aid in holding polluters accountable, contributing to predictions about the fate and movement of pollution, and developing a strategy for monitoring compliance. When implementing the results from the modeling technique to real-life scenarios of groundwater pollution, the equifinality of the solution cannot be entirely ruled out. The accuracy of the results is governed by assumptions and simplifications made in the model to simulate the aquifer response. Even the model calibration is dependent on the availability of data, which may be noisy and erroneous, resulting in nonunique solutions.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *