Using a multi-method approach based on soil radon deficit, resistivity, and induced polarization measurements to monitor non-aqueous phase liquid contamination in two study areas in Italy and India

Abstract

Geochemical and geophysical surveys employing radon deficit, resistivity, and induced polarization (IP) measurements were undertaken on soil contaminated with non-aqueous phase liquids (NAPLs) in two different sites in India and in Italy. Radon deficit, validated through the comparison with average soil radon in reference unpolluted areas, shows the extension of contamination in the upper part of the unsaturated aquifers. In site 1 (Italy), the spill is not recent. A residual film of kerosene covers soil grains, inhibiting their chargeability and reducing electrical resistivity difference with background unpolluted areas. No correlation between the two parameters is observed. Soil volatile organic compounds (VOCs) concentration is not linked with radon deficit, supporting the old age of the spillage. NAPL pollution in sites 2a and 2b (India) is more recent and probably still active, as demonstrated by higher values of electrical resistivity. A good correlation with IP values suggests that NAPL is still distributed as droplets or as a continuous phase in the pores, strengthening the scenario of a fresh spill or leakage. Residual fraction of gasoline in the pore space of sites 2a and 2b is respectively 1.5 and 11.8 kg per cubic meter of terrain. This estimation is referred to the shallower portion of the unsaturated aquifer. Electrical resistivity is still very high indicating that the gasoline has not been strongly degraded yet. Temperature and soil water content influence differently radon deficit in the three areas, reducing soil radon concentration and partly masking the deficit in sites 2a and 2b.

Appendix.

Starting from Eq. (1), proposed by Schubert (2015) to calculate soil radon concentration in the pore space (C_∞), we have developed a new calculation, Eq. (2), where radon concentration in the polluted sites is firstly compared to soil radon values in close unpolluted areas to calculate the radon deficit (Di). Equilibrium radon concentration in the background is determined using a simplified version of Eq. (1):

$$ {C}_{\infty }=\varepsilon\ {A}_{\mathrm{Ra}}{\rho}_d/n $$

(3)

From the comparison of Eqs. (1) and (3), it turns out that

$$ {D}_i=1/\left(1-{S}_{\mathrm{F}}+{K}_{\mathrm{W}/\mathrm{SG}}{S}_{\mathrm{F}}\ \left(1-{X}_{\mathrm{NAPL}}\right)+{K}_{\mathrm{NAPL}/\mathrm{SG}}{X}_{\mathrm{NAPL}}{S}_F\right) $$

(4)

Now, the total fluid saturation of the pore space (S_F) is the sum of S_NAPL and S_W, which are the pores fractions occupied by NAPL (S_NAPL) and water (S_W), respectively, and these two terms can be expressed as follows:

$$ {S}_{\mathrm{NAPL}\mathrm{i}}={\mathrm{NAPL}}_{\mathrm{Pi}}/\left(\left({\gamma}_{\mathrm{NAPL}}/{\gamma}_{\mathrm{di}}\right)\hbox{--} \left({\gamma}_{\mathrm{NAPL}}/{\gamma}_{\mathrm{s}}\right)\right) $$

(5)

$$ {S}_{\mathrm{W}\ \mathrm{i}}={W}_{\mathrm{Pi}}/\left(\left({\gamma}_{\mathrm{W}\mathrm{i}}/{\gamma}_{\mathrm{di}}\right)\hbox{--} \left({\gamma}_{\mathrm{W}\mathrm{i}}/{\gamma}_{\mathrm{s}}\right)\right) $$

(6)

or simplifying, S_NAPL i = NAPL_Pi/A and S_W i = W_Pi/B

where

$$ A=\left(\left({\gamma}_{\mathrm{NAPL}}/{\gamma}_{\mathrm{d}i}\right)\hbox{--} \left({\gamma}_{\mathrm{NAPL}}/{\gamma}_{\mathrm{s}}\right)\right) $$

(7)

$$ B=\left(\left({\gamma}_{\mathrm{Wi}}/{\gamma}_{\mathrm{di}}\right)\hbox{--} \left({\gamma}_{\mathrm{Wi}}/{\gamma}_{\mathrm{s}}\right)\right) $$

(8)

then,

$$ {S}_{\mathrm{F}\ i}=\left({\mathrm{NAPL}}_{\mathrm{Pi}}/A\right)+\left({\mathrm{W}}_{\mathrm{Pi}}/B\right) $$

(9)

If parameters defined in equations from (5) to (9) are inserted in Eq. (4) and then the latter is developed and solved with respect to NAPL_Pi, we obtain Eq. (2), where the following polynomials are defined:

$$ C=\left({W}_{\mathrm{Pi}}/B\right)\hbox{--} \left({D}_i{W}_{\mathrm{Pi}}/B\right)+\left({D}_{\mathrm{i}}{W_{\mathrm{Pi}}}^2/{B}^2\right)\hbox{--} \left({D}_{\mathrm{i}}{K}_1{W_{\mathrm{Pi}}}^2/{B}^2\right) $$

(10)

$$ D=\left[{D}_{\mathrm{i}}{W}_{\mathrm{Pi}}/ AB\right)\left]-2\ \right[{D}_{\mathrm{i}}{K}_1{W}_{\mathrm{Pi}}/(AB)\Big]+\left(1/A\right)\hbox{--} \left({D}_{\mathrm{i}}/A\right) $$

(11)

$$ E=\left({D}_{\mathrm{i}}/{A}^2\right)\hbox{--} \left({D}_{\mathrm{i}}{K}_1/{A}^2\right) $$

(12)

$$ F=\left[{D}_{\mathrm{i}}{K}_2{W}_{\mathrm{Pi}}/(AB)\right]\hbox{--} \left[{D}_{\mathrm{i}}{K}_1{W}_{\mathrm{Pi}}/(AB)\right] $$

(13)

$$ G=\left({D}_{\mathrm{i}}{K}_2/{A}^2\right)\hbox{--} \left({D}_{\mathrm{i}}{K}_1/{A}^2\right) $$

(14)

$$ H=E-G $$

(15)

$$ J=D-F $$

(16)

Note that K_W/SG and K_NAPL/SG are labeled as K₁ and K₂, respectively.

The solution to the following calculation:

$$ \mathrm{H}\ {{\mathrm{NAPL}}_{\mathrm{Pi}}}^2+\mathrm{J}\ {\mathrm{NAPL}}_{\mathrm{Pi}}+C=0 $$

(17)

is the above mentioned Eq. (2):

$$ {\mathrm{NAPL}}_{\mathrm{Pi}}=\frac{-J\pm \sqrt{J^2-4 CH}}{2H} $$

where actually two solutions, one positive and one negative, are provided, but only the positive one (deriving from the negative sign in front of the square root) is meaningful. Calculated fractions of NAPL in the pore space (NAPL_Pi) of sites 2a and 2b are reported in Table S1 and S2 for all the stations of the grids. Table 1 lists other parameters used for calculation.