Kent may be known as the Garden of England but that does not mean it escaped Britain's industrial legacy. Damon Schünmann reports on Rochester's big clean up.
Victorian gasworks were not famed for cleanliness and the one at Rochester Riverside in the Thames Gateway is no exception. The site in south east England needs remediation before being redeveloped for commercial and residential use.
The site is highly permeable, with groundwater contamination from the gasworks within the river terrace gravels (RTG) and within perched water in made ground at shallow depths. The removal of the dense non-aqueous phase liquid (DNAPL), which is a source of dissolved phase contamination, protects an important drinking water resource and adjacent aquatic habitats.
The project is a non-integrated joint venture (JV) between Edmund Nuttall and Van Oord. Nuttall is undertaking the contamination excavation, land raising, construction of a sheet piled river wall, installation of sewers and supervising of subcontractors. Van Oord is responsible for dredging material for the landraise and putting in band drains.
Funding for the engineering works at the 32ha site includes £38M from the government and £16M from the South East England Development Agency (SEEDA).
The JV won £32M of this and is working for joint developers Medway Council and SEEDA.
Royal Haskoning designed the remediation scheme with Halcrow acting as Medway Council's representative.
Several areas are contaminated but chief offender is land around three former gas tanks.
'There was one above ground tank, which had fantastic integrity and was built in the 1950s. But the two below ground tanks were from the 1890s and we think one of these contributed to the majority of the contamination, ' says Nuttall remediation agent David Bugden.
The company's project manager, Steve Horton adds: 'The bulk of the contamination is hydrocarbons and there is also DNAPL, which is basically tar.'
Areas of contamination fall into several categories, one of which is unrelated to the hydrocarbon pollution. Some unforeseen asbestos was discovered on the site in a 3,000m 2 area. It is being removed with partial bulk excavation and hand picking.
Two subcontractors are tackling contamination in perched groundwater, made ground and an underlying gravel aquifer.
Biogénie won the contract to deal with 80,000m 3 of made ground and is using biopiles to decontaminate it. It installed a waterproof surface where the soil is deposited, and an air extraction system to ensure optimal treatment conditions in the soil pile.
There are two treatment methods for biopiles. Biostimulation stimulates the naturally occurring microorganisms in the soil, while bioaugmentation uses a combination of biostimulation and inoculation of soils with strains of micro-organisms to enhance the treatment.
Biogénie chose biostimulation for Rochester Riverside and has created favourable soil conditions to allow natural bacteria to enhance biological degradation of the organic contaminants so that concentrations are suitable for onsite reuse.
It installed perforated pipes within soil stockpiles with secondary pipes connected to a blower to enhance air circulation. It then mixed organic additives with the soil to increase the rate of pollutant degradation.
Biogénie hopes to reduce leachate concentrations to 0.0862mg/l for PAH (polyaromatic hydrocarbons) and 0.517mg/l for TPH (total petroleum hydrocarbons), in made ground and alluviums. Some PAH concentrations have exceeded 9.6mg/l in leachate, while other samples have presented soil PAH concentrations up to 5500mg/kg. By achieving this target, Nuttall aims to reuse about 70,000m3 of treated soil on site.
The company's project manager Claire Bourgouin says: 'As well as PAH and TPH there is blue billy [ferro and ferric cyanides] in the gasworks area.'
She says blowers draw air through the excavator-turned biopiles which assists their aeration and the moisture drawn off from the heaps can be cleaned if contaminated.
Another contractor, QDS Environmental, has won a contract worth about £750,000, and is using several distinct in-situ groundwater remediation approaches. These include DNAPL recovery, chemical oxidation, ow path management (FPM) and surfactant ushing through the perched groundwater within the made ground.
The choice of remediation was dictated by the geology, speci cally the high permeability of the RTG and the depth of the contamination.
QDS was forced to discount using surfactant in the RTG aquifer as well as thermal or steam techniques.
Although they would have increased the DNAPL's mobility, there was a risk that it would migrate through fractures within the chalk below, rendering it unrecoverable.
To remove DNAPL from the aquifer, site workers have installed a network of 24, 150mm boreholes in the impacted area around the former gas holder bases. Even though groundwater abstraction is at very high ow rates to entrain the product, the abstractions occur intermittently.
This allows recovery of both groundwater and pollutant but reduces the likelihood of creation of preferential w paths encouraging contamination recovery throughout the matrix.
Pumps transfer the contaminated uid to an oil water silt separator to remove the free product after which the water gets treated using sand and activated carbon lters.
QDS project manager Alistair Ray says that because the contaminants are in the underlying gravel aquifer, it is unlikely that the entire free product can be removed in any practically achievable timescale. The volume of free product will be plotted on a recovered volume against time graph until a steady state has been reached and signi cant volumes are no longer being recovered. At this point known as the asymptote, the recovery ef ciency will be too low to cost effectively continue with the RTG pumping system.
Once the asymptote has been reached, the subcontractor has proposed a programme of chemical oxidation, with potassium permanganate, to reduce the risk of residual contamination. This will be mixed with treated ef uent and reinjected via the 24 boreholes. QDS says the remediation period to ensure maximum economic recovery of DNAPL will take about two years.
Ray says the aim is to recover between 20,000l and 50,000l of DNAPL sludge.
But there is little point in cleaning the aquifer with a source of contamination sitting over it so the pollution must be removed at source. Tar that was left in pits has penetrated the normally impermeable alluvium and Nuttall is removing this to prevent re-contamination.
For the shallower perched groundwater DNAPL contamination, QDS is using ow path management (FPM), involving trenches and surfactant ushing. This will clean more than 100m of made ground that would otherwise have needed bulk excavation, and Ray believes using this application in such a way may be a UK rst.
Site workers will re-inject some of the recovered and treated water from the RTG work. This is for the remediation of the shallower perched groundwater within the made ground.
It involves using parallel trenches with water ushing from an injection trench to an abstraction trench, cleaning the ground in between.
Site workers are using surfactants in the FPM to accelerate recovery of the residual DNAPL. The treated efuent from the abstraction trench will be dosed with surfactant and the solution put back into the ground through the injection trenches. The surfactants will reduce the surface tension and enhance the ow of residual DNAPL back again to the abstraction trench.
The re-circulated pumped water will be sent to the central water treatment plant in between each circuit.
Alongside groundwater work, three mobile pumps are transferring contaminated surface water for cleaning in the system of sand and carbon filters.
The majority of the project should be complete by December with the aquifer cleaning finishing in mid-2008. The more polluted areas will be handed over in 2015, after settlement there has run its course.
The land raise
'The land raise will protect the site and its mixed use development from a one in 200 year ood with 400,000m 3 sand dredged from the Thames Estuary, ' says Edmund Nuttall project manager Steve Horton.
The effective volume of land raise is 300,000m 3 which is about 2m, with the 100,000m 3 difference being mainly accounted for by settlement.'
There will be a 1m thick capping layer to protect human health from the former use of the site and also 70,000m 3 of cut and ll to pro le it.
'The sand is regarded as waste by the Environment Agency and was treated (the water was removed from it) under a waste management license. Van Oord dredged using a Dredger Trailing Suction Hopper and has largely nished this work, ' he says.