Environmental Engineering Reference
In-Depth Information
13.8. Conclusions
The basic phenomena of transfer and retention of soluble chemical pollutants
through the underground are now fairly well known. Summaries more complete
than the one proposed in this chapter may be found in [BEA 87] and [ZIE 89].
Numerical methods developed over the past two decades allow us to model these
phenomena using the finite element method. A number of questions remain open
and are now discussed.
The evaluation of dispersivity remains difficult in relation to transfers. Its
relationship with the flow scale considered must be further investigated, especially
in relation to the spatial variability of the hydraulic and mechanical properties of the
soil.
The retention of pollutants with complex chemical or biological structures must
be better understood by taking into account their evolution based on chemical
balance equations and on equations describing the chemical or biological evolution.
When the amount of dilute pollutant becomes relatively large, the flow of the
transporting fluid can be altered, as for instance in the case of aquifers in contact
with seawater with a salt concentration close to 3.5%. At this level, the viscosity and
density of water are sufficiently modified to affect the flow. Salt transport modeling
in coastal aquifers must therefore take into account coupling between pollutant
transfer and water flow [GAL 92, HAS 88]. It should be noted that the transport of
heat in a saturated porous medium obeys the same laws. This problem is being
studied in greater depth in research into transport phenomena around underground
radioactive waste storage [ALO 99].
The transfer and retention of pollutant in the unsaturated zone have been
relatively little studied so far. The main concepts are briefly mentioned in [BEA 87].
Experimentation appears to be particularly difficult.
With regard to numerical methods, the association of the characteristics method
with Galerkin's finite elements method is potentially the most elegant and powerful
approach, once on-going research allows us to determine the optimal compromise
between the numerical code complexity and the precision and stability of the
calculations.
13.9. Acknowledgments
The authors thank the FNRS, the French Community of Belgium and the
European Commission for the support received in the framework of their research.
