Assessing the Link Between Electromagnetic Induction Data and Hydrocarbon Pollution at a Former Refinery in Rouen (FR)

Brownfields are an important real estate reserve inside cities. However, contaminant sources generated by former activities, notably hydrocarbon pollutants, must be characterized to access the site redevelopment costs. Environmental assessment techniques (soil sampling) are the classical way to gather soil contamination data, but they offer punctual information and are costly and time-consuming. Electromagnetic induction methods (EMI) allow on-the-go mapping of the soil apparent electric conductivity and in phase component with a higher spatial density that could complement the local chemical analyses. To investigate the correlations between EMI data and contamination levels, an EMI survey and 25 boreholes were realized on a former 3 000m² gas refinery site at Rouen, France. Eighty-four soil samples were analyzed for total hydrocarbon (THC). From a log-log plot between the measured apparent electric conductivity and THC levels, the results show a linear trend. Four classes were individualized and showed that THC levels, the urban fill heterogeneities and the in phase component contribute differently to the EMI signals.