Characterization of hydrocarbon spreading in an alluvial aquifer by cross correlation study of precipitation and contaminant content data

Abstract

An accidental spillage of gasoline in a sector of the ‘Vega de Granada’ alluvial aquifer (Spain) affected a number of wells in the area. Hydrocarbon concentrations in ground water were monitored over a period of two years and five months. Monitoring indicated that the contamination levels tended to decrease, although sharp fluctuations were observed. Correlation and spectral analysis were used, among other methods, to construct a model of the hydrocarbon concentrations' evolution. We established the existence of maximum values of cross-correlation between the time series for hydrocarbon content and rainfall, with different lags for each well. Cross correlation is most clear in several wells located along a paleochannel in which the contamination tended to flow. Quantification was therefore possible of both the contaminant propagation rate and the permeability of the aquifer in that area. Although more detailed analysis showed that the high correlation values were due to a single data set (corresponding to the maximum concentrations observed in each well after a period of exceptionally heavy rainfall), the above conclusions may still be considered valid for periods of intense recharge of the aquifer, as demonstrated by an artificial recharge test.