Limits and applications of the radon deficit technique for the study of two sites contaminated by NAPL (Non-Aqueous Phase Liquids)

Previous works (Schubert, 2015; De Miguel et al., 2020) show that the higher solubility of 222Rn in NAPLs (Non-Aqueous Phase Liquids) than in water or air is often used to localize the spatial distribution of a NAPL plume. However, the limitations and advantages of this technique in different real contexts have scarcely been addressed. The aim of this work is to present an evaluation of this method applied to different real situations. Two polluted areas in Rome have been studied for two years. They are both gas stations with different geological and spill characteristics. The first site is in a volcanic geological setting, with high natural radon concentrations. It is characterized by a deep aquifer with low variations in the water table level during the year and it is affected by an ancient spill. The other area is characterized by the presence of alluvial deposits, a shallow aquifer with significant oscillations of the groundwater level and a variable flow direction during the year. Furthermore, the spill on this site is much more recent than that of the other site. In both areas Rn-deficit was periodically measured in groundwater to monitor the distribution of the contamination. Several parameters have been evaluated for the interpretation of the data, such as radon and NAPL concentrations in sampled waters, precipitation rates and fluctuations in groundwater levels, as well the role of rainfalls and groundwater oscillations in the remobilisation of pollutants and consequently conditioning radon concentrations in both contexts. Multivariate statistical analysis is then applied to explain the variations and correlations between the parameters analysed, to observe the sensitivity and the validity of the technique, also in the light of the different remediation systems applied to the two sites. Multivariate statistical analysis (PCA) shows a strong negative correlation between water radon and residual NAPLs (total hydrocarbons, MTBE and ETBE) concentration and no dependence of NAPLs levels from rainfall rates in the first site affected by an old contamination, with limited fluctuations of groundwater levels. A negative covariance between radon and NAPLs is demonstrated also for the second site, even if seasonal reversal of the groundwater flow complicate NAPLs concentration and rainfall rates relationships. The role of water pumping on contaminants and radon distribution is more evident in this area affected by a much more recent NAPL spill. De Miguel E., Barrio-Parra F., Izquierdo-Díaz M., Fernández J., García-González J.E. & Álvarez R. (2020) - Applicability and limitations of the radon-deficit technique for the preliminary assessment of sites contaminated with complex mixtures of organic chemicals: A blind field-test. Environment International, 138. 105591. Schubert M. (2015) - Using radon as environmental tracer for the assessment of subsurface NonAqueous Phase Liquid (NAPL) contamination. A review. Eur. Phys. J. Spec. Top., 224, 717-730.