Having excavated 66,000m 2of diaphragm wall trench, Italian contractor Rodio filled it all in, only to dig it out again.And it was all part of the plan.Paul Wheeler finds out why.
Italian contractor Rodio's project to contain heavily polluted soil and groundwater at a petro-chemical plant in Brindisi on the Adriatic coast looks remarkably simple.
Its two year, E10M contract with Enichem is to isolate two, 16ha and 5ha, zones which are contaminated by non-volatile hydrocarbon, solvents and heavy metals.
Perimeters around each are 1,550m and 850m respectively. But it is the depth of the cut-off that makes the project stand out.
A suitable impermeable layer to create a basal cut-off at the site is between 25m to 32m deep. Even so, a mineral barrier should be able to achieve the 1x10 -9 m/s permeability required in the technical specification. Why then has Rodio excavated 66,000m 2ofdiaphragm trench and installed the same quantity of HDPE membrane within a self-hardening mud-filled trench.
First, it is not just about permeability. The cut-off's specification is also very severe in terms of deformability and homogeneity.
Rodio's Raffaella Granata also believes that: 'Although conventional cut-off walls constructed around waste disposal or polluted areas can reduce leachate migration, over time construction defects and residual permeability mean contaminants can seep through.'
In Rodio's experience of this type of work, which stretches back to the early 1990s, a reliable way to reduce the hydraulic conductivity and chemical diffusion is a composite barrier of HDPE membrane and a self-hardening mud diaphragm wall.
And in any case, Granata maintains, since March 2003 Italian law, in accordance with a European directive, requires 'dangerous wastes' to be contained by a composite/mineral lining system.
That the wall is deep would push the experience of the few European contractors that operate in this field. But that is before accounting for the geology.
Lurking above the clay layer is an ultimately much more significant problem - a hard calcareous sandstone layer (known as Calcarenite in Italy).
This layer, which appears at between 5m and 15m depth, is up to 8m thick and has a strength of up to 25Mpa. There was no way it could be excavated using a conventional grab.
It was however within the capabilities of Rodio's hydromill diaphragm rock-cutters, but this in turn became incompatible with the requirement to install the HDPE liner in a continuous production process in a self-hardening mud.
Granata explains: 'The hydromill works under bentonite, and we realised replacement of the bentonite with the self-hardening mud was 'mission impossible'.'
As installation of the wall was into fresh mud, 'this could only be achieved if the trench was excavated without interruption, ' continues Granata, rather than separated into 'panels' as in a conventional diaphragm wall installation.
Replacing the bentonite with self-hardening mud would have to have been carried out for the entire wall 'all at the same time'.
And there are other factors that make this approach impossible, even where a geomembrane is not needed and where, in theory, the diaphragm could be built in panels.
As Granata points out: 'The replacement would still not be possible because the density of the self-hardening mud is very light and about the same as that of bentonite, so you wouldn't get the mass displacement achieved with concrete.'
And to cap it all, the specification for the self-hardening mud at Brindisi was so onerous in terms of permeability, deformability, and homogeneity that it could not be achieved with any form of remixing.
This all added up to very convincing arguments for Rodio's two-stage approach. The firm excavated the trench in panels under bentonite using the hydromill and then backfilled it with a plastic concrete. The properties of the mix for this 'preliminary diaphragm' were selected so that it could be re-excavated by grab under self-hardening mud, but would not mix and contaminate the mud during the process.
With this in place, Rodio was in a position to install the composite diaphragm in a continuous process. 'Secondary panels' were excavated before the retarded selfhardening mud of adjacent primaries had started setting.
By 28 days, tests on samples taken both from the mixing plant and trench surpassed the technical specification, with permeability values of 10 -9 m/s to 10 -10 m/sand an unconfined compressive strength of 200kPa to 400kPa.
With the whole point of the complex preparation to enable the HDPE geomembrane to be installed in fresh mud, once started work continued round the clock, seven days a week, until the job was done.
The final task was to seal the surface with an engineered capping and install leachate monitoring - but that's another story.