less remediated than with oxidant injection. A detailed data analysis and a simplified
model suggested that a 99% removal could be reached. However, this is not sure due
to the uncertainty on the potential toluene removal from the low permeability lenses.
Sparging efficiency was shown to depend largely on the spatial distribution of the
air channels, which cannot be predicted. In the tank experiment, the air was provided
through a spiral tube at the bottom of the tank. In the field, this is not possible and the
air is provided through injection wells that may require a very dense injection well
network to reach a similar injection rate as in the tank. The duration of the air
sparging to reach a significant removal in the low permeability lenses was estimated
by considering that the diffusion process that was present at the end of the tank
experiment would remain stable. This point is an assumption that cannot be validated in the present experiments and which has not been validated at this scale. The
flux decreases for surfactant and sparging techniques thus remain with a high
uncertainty on the prediction.
It is not possible to strictly compare the studied methods in terms of cost because
the man power and analytical costs are generally the highest and cannot be calculated
easily. Concerning the setup of injecting and pumping wells, the cost may be close as
it is now observed that the well density is similar for a lot of treatment techniques, the
5 m radius of action being often used. In a first approach, sparging seems to have the
lowest costs as it requires only air injection and gas-phase treatment, often realized
with activated carbon. If the potential use of treatment agent provided in Fig. 4.13 is
used as a comparison basis for the cost, surfactant shall be more efficient in terms of
agent consumption than persulfate oxidation. However, using surfactant requires a
complex treatment of pumped water and takes much longer time, which may lead to
costs much higher than the product itself. All tank experiments were conducted on
approximately one cubic meter of soil. It would thus be tempting to simply use the
numbers provided in this study to upscale them for any source of contaminant. This
would be meaningless as this study emphasized that the efficiency depended on the
hydraulic conductivity of the lenses containing the contaminant, and it shall also
depend on their size and the value of this hydraulic conductivity.
Heterogeneity of the remaining contaminants was reported elsewhere (Atteia
et al. 2017) and was shown to be very high. This effect could not be specifically
linked to one treatment technique and can therefore be originating from the conditioning of the pilot with high-velocity water flush. As the heterogeneity leads to more
difficult treatment, it may be useful to use directly a treatment instead of water
washing and also to treat the contamination as fast as possible to limit this effect.
Further research should explore combinations of the different treatments, in order to
understand whether potential improvements in efficiency could compete with an
increase in treatment costs.
4 Comparing the Efficiency of Oxidation, Sparging, Surfactant Flushing, and. . .
235
model suggested that a 99% removal could be reached. However, this is not sure due
to the uncertainty on the potential toluene removal from the low permeability lenses.
Sparging efficiency was shown to depend largely on the spatial distribution of the
air channels, which cannot be predicted. In the tank experiment, the air was provided
through a spiral tube at the bottom of the tank. In the field, this is not possible and the
air is provided through injection wells that may require a very dense injection well
network to reach a similar injection rate as in the tank. The duration of the air
sparging to reach a significant removal in the low permeability lenses was estimated
by considering that the diffusion process that was present at the end of the tank
experiment would remain stable. This point is an assumption that cannot be validated in the present experiments and which has not been validated at this scale. The
flux decreases for surfactant and sparging techniques thus remain with a high
uncertainty on the prediction.
It is not possible to strictly compare the studied methods in terms of cost because
the man power and analytical costs are generally the highest and cannot be calculated
easily. Concerning the setup of injecting and pumping wells, the cost may be close as
it is now observed that the well density is similar for a lot of treatment techniques, the
5 m radius of action being often used. In a first approach, sparging seems to have the
lowest costs as it requires only air injection and gas-phase treatment, often realized
with activated carbon. If the potential use of treatment agent provided in Fig. 4.13 is
used as a comparison basis for the cost, surfactant shall be more efficient in terms of
agent consumption than persulfate oxidation. However, using surfactant requires a
complex treatment of pumped water and takes much longer time, which may lead to
costs much higher than the product itself. All tank experiments were conducted on
approximately one cubic meter of soil. It would thus be tempting to simply use the
numbers provided in this study to upscale them for any source of contaminant. This
would be meaningless as this study emphasized that the efficiency depended on the
hydraulic conductivity of the lenses containing the contaminant, and it shall also
depend on their size and the value of this hydraulic conductivity.
Heterogeneity of the remaining contaminants was reported elsewhere (Atteia
et al. 2017) and was shown to be very high. This effect could not be specifically
linked to one treatment technique and can therefore be originating from the conditioning of the pilot with high-velocity water flush. As the heterogeneity leads to more
difficult treatment, it may be useful to use directly a treatment instead of water
washing and also to treat the contamination as fast as possible to limit this effect.
Further research should explore combinations of the different treatments, in order to
understand whether potential improvements in efficiency could compete with an
increase in treatment costs.
4 Comparing the Efficiency of Oxidation, Sparging, Surfactant Flushing, and. . .
235