Comment évaluez-vous l’efficacité d’une barrière réactive perméable pour l’assainissement des sols et des eaux souterraines ?

Groundwater flow is a vector - flow rate and direction. One of the key design parameters is how much these could change as a result of climate change. As groundwater flow rate increases, the PRB needs to be thicker to allow the residence time for treatment to be effective. As groundwater flow direction changes, there is a need for the barrier to be able to still intercept the plume and ensure it passes through the treatment zone. Assuming business as usual - i.e. today's conditions will also be tomorrow's - could result in only partial treatment of contaminated groundwater or the PRB being partly or even wholly by-passed by the plume.

The site geology is key. A PRB is most suitable for a shallow aquifer otherwise excessive cost may be encountered during the installation. So it is important to assess the depth and structure of the underlying aquifer and the suitability of the underlying strata to act as an aquitard. Modelling of the site groundwater is key to understanding the movement of contaminants and whether these are at a concentration that can be treated. The concentration would also allow an understanding of the longevity of the treatment material used. The main issue often with a PRB is the long term operational cost, which often is unfavourable to site owners who wish to develop the site. It can be a favourable solution to an operational site.

Think about your true source term. Groundwater concentrations will contain only a small percentage of soluble contaminants. 80-90% of the mass will reside in the soil. Maximum flow in the formation will likely hold the minimum concentrations in groundwater. The life of the PRB will be driven by contaminant flux through the barrier but likely for a much longer period resulting from the contribution of the residual source mass in the soil. Measure total soil mass with a carefully designed high-resolution survey including a statistically significant number of soil cores across the full profile of the planned PRB.

It seems, which is more important to control or to eradicate the contamination, so its important to adopt this geochemical with specific chemical applications

The reactive media used is a key consideration this may include chemical or biological based media which can degrade or adsorb contaminants of concern. For example contaminants such as phenols, BTEX and hydrocarbons could be biodegradable under the right conditions whereas heavy metals could be accumulated on reactive media such as zero valent iron. The risk may exist where the concentration of metals builds up to such an extent that the “roll up effect” means that a higher concentration could be discharged by the reactive media, causing a potentially higher concentration of heavy metal leaving in the water leaving the PRB.

A PRB is most likely to be considered on a site where operational limitations prevent access to areas of the site where the contamination may be present and the PRB can manage the risk of off site migration. The will incur a long term cost for maintaining and monitoring the system to ensure it is meeting the regulatory requirements. This long term cost can be significant but it may be the most cost effective way to maintain an operational site until an appropriate point when the contaminant sources could be removed, should that be feasible. A passive system where by the groundwater passes through the reactive media can be more cost effective than some active systems, but requires very good characterisation of the site hydrogeology.

So often considered secondary to the design of PRBs is the expected total lifecycle cost when considering the many variables discussed elsewhere in this article. The value of high-resolution definition of each of the hydro-geo-bio-chemical elements of the existing conditions is critical to a cost-effective design. Then one needs to consider what range of uncertainty remains with these parameters along with the physical dimensions of the PRB and the probable cost of these uncertainties. Lastly, how can these uncertainties and their associated costs be reduced with better characterization and pilot testing?