The implementation of genetic algorithm for the identification of DNAPL sources

Contamination caused by dense non-aqueous phase liquids (DNAPLs) is a serious peril to the groundwater supply quality. The DNAPLs are largely undetected and yet are likely to be a significant factor limiting site remediation. It is necessary to identify the sources of pollution and control pollution in the early stage. The primary objective of this research was to investigate the pollution source and intensity of migrating DNAPLs. This objective was accomplished through the combination of genetic algorithm (GA) and T2VOC numerical experimental simulation. This study considers two synthetic cases of numerical modeling based on the T2VOC from the TOUGH2 code family to assess and evaluate the efficacity of the algorithm. The first case identified concurrently the intensity as well as the location of the one unknown contaminated water source. The second case distinguished various contaminated water sources. For the single source pollution identification, the estimation of intensity error was 0.08%, and the error of GA at second and third true value calculation was 0.05% and 0.03%, respectively, achieving the goal of pollution identification accurately. While for the multiple sources, the deviation of intensity was ranging between 1.02 and 5.65%, the results suggested that source No.2 was mainly responsible for the leakage, and source No.3 adopted the secondary responsibility among others. These two cases suggested that GA established by optimal search strategy could be adopted for effective identification of DNAPL pollution sources. This study provides information that could facilitate the development and implementation of innovative remediation strategies.

Automated summary

The research investigates pollution sources and intensities of migrating DNAPLs using genetic algorithms and numerical simulations, demonstrating the effectiveness of GA in identifying DNAPL pollution sources accurately.