地下油罐泄漏区污染源的自动识别

Automated identification of the unknown contaminant source in groundwater at a leaking underground storage tank site

地下油罐泄漏隐蔽性大,不易被及时发现,而且泄漏出的烃类化合物毒性强,对地下水会造成严重污染.为了对某一地下油罐污染源位置和强度进行准确识别,本研究结合现场调查结果,建立了该地下油罐泄漏区的石油污染物反应运移模型.在基本确定污染源泄漏区范围的前提下,进一步建立地下油罐泄漏区污染源的识别模型,并利用遗传算法对泄漏区污染源的位置和泄漏强度进行了自动识别,进而通过参数敏感性分析来评价和比较不同参数对污染物运移结果的影响.结果表明,运移模型中与地下水流场变化相关的特征参数(如开采量、水力坡度、渗透系数等)对污染物运移结果的影响相对比较明显.对于污染源来说,研究不同污染组分在含水层中的化学和生物化学变化显得尤为重要.

The leaking of underground storage tank is hard to be discovered by humans, and groundwater resources can be seriously polluted by the highly toxic hydrocarbon compounds from leaking. In this study, in combination with the outcome of spot field investigation, a reactive transport model of groundwater and petroleum contaminant typical component was built in the groundwater system at a leaking underground storage tank （LUST） site to accurately identify the location and leak intensity. On the premise of identifying the spatial range for the contaminant source in general, the model in identifying contaminant source in the LUST groundwater system was built, and a genetic algorithm was used to automatically identify location and amount of leakage from contaminant source. Afterwards, the effect of parameters on the results was estimated by parameter sensitivity analyses. All above results suggested that the parameters related to groundwater flow model, e.g. exploitation rate, hydraulic gradient, and hydraulic conductivity, have a comparatively remarkable effect on the contaminant transport process. In particular, the biochemical degradation of contaminant in groundwater aquifer plays an indispensable role in the transport model.