Testing of reactive iron barrier for groundwater purification
Summary and conclusions
In 1976 Vapokon Petrokemiske Industri A/S established a business recovering solvents from used liquids, paints, etc. The company went bankrupt in 1996-97 and the property was taken over by Municipality of Søndersø.
In 1989 an investigation showed that the property was contaminated. Since then a range of investigations have been carried out and in 1998 a preliminary design describing possible remedial actions was carried out. Based on the preliminary design, and environmental and economical considerations the decision was made to install a reactive barrier in a Funnel &Gate system complemented with a groundwater drainage system as well as digging up contaminated soil in the hotspot area. This solution will prevent further spreading of the contamination. The reactive barrier consists of iron chips. Anaerobe corrosion of iron causes dechlorination of the chlorinated substances in the water.
The remediation project covers preliminary investigations, elaboration of the detailed design, modelling, installation of remedial measures as well as monitoring. Monitoring is carried out in order to document the efficiency of the remedial actions and in order to avoid exceeding limits set for discharging drainage water.
The remedial project is conducted by RAMBØLL on behalf of Fyns Amt. The project is a Technology development project under the Danish EPA.
Previous investigations
The previous investigations showed among others:
 | Heavy soil contamination with chlorinated solvents and BTEX. |
| The plume of contamination in the groundwater has moved 200 meters downstream from the property. The width of the plume is 80 meters. |
| Due to the groundwater flow direction existing drinking water intakes are not at risk of contamination. However, the stream (Holmebækken) Southeast of the site, and the groundwater resource are affected by contamination. |
Preliminary investigation
Prior to elaboration of the detailed design of remedial actions a range of preliminary investigations have been conducted in order to:
| Map the horizontal and vertical distribution of the groundwater contamination |
| Investigate clogging of the reactive barrier |
| Investigate the vertical distribution of the groundwater flow |
| Investigate the geotechnical conditions for installation of the Funnel & Gate system |
Results of preliminary investigations have not given cause for changing the location of the reactive barrier as decided in the preliminary design. It is concluded that the barrier should have the same thickness all over. The clogging up of the barrier has also been estimated.
Based up on the preliminary investigation, discharge criteria, and considerations concerning geotechnical and construction matters the Funnel (sheet piling) and drainage system have been repositioned.
Modelling and detailed design
A model simulating groundwater flow and transport of chemical substances has been set up in order to optimise the dimensioning of the reactive barrier and the drainage system. The thickness of the barrier has been calculated to 0,8 meter which have been verified using a detail model.
The ramming of sheet piles, excavation within the case of sheet piles, and installation of the barrier have been projected. Finally the drainage system has been dimensioned and planned.
Conduction of remedial measures
The production equipment from the former activities have been removed, and following 4,600 tons of soil and 13 tons of waste have been removed from the hotspot area.
The drainage system has been established with 7 manholes, which is installed with changeable drainage levels.
The sheet piling has been rammed down to 12 meter below ground level, corresponding to a minimum of one meter into the till underneath the aquifer.
The barrier is installed by first ramming down a sheet piling case. The materials within the case are removed and the barrier is installed. Before the case is removed a filter consisting of high permeability material is installed on each side of the barrier.
Monitoring
A comprehensive preliminary monitoring program was carried out and based upon these results further monitoring was conducted. Monitoring of the effect of the remedial actions has been carried out from February 2000 to December 2001. A total of 7 monitoring rounds have been conducted.
Drainage water
The concentrations in the drainage water have been exceeding discharge criteria all through the monitoring period. Water samples taken from the 7 manholes have been analysed and due to the results and measurements of the groundwater level the drainage levels have been changed twice. In order to ensure better control of the drainage water the drain has now been separated into 7 individual drainage pipes, and tests of the flow to each of the pipes have been conducted as well as analyses of water samples from each manhole. Based upon the results new drainage levels will now be reestimated.
Distribution of contamination in and around the barrier
Analyses show that there is an uneven distribution of contaminants in and around the barrier. Slugtests show an uneven distribution of hydraulic conductivity in and around the barrier, and it is assessed that the uneven water movements result in the uneven distribution of the contaminants.
Variation in the concentrations over time
The concentrations of the contaminants have gone up a lot from the time of first monitoring round to September 2000. It is assumed that the effect of installing the Funnel & Gate system has resulted in this rise as well as the fact that the site has been unpaved from December 1999 to December 2000.
Treatment in the reactive barrier
A steep reduction in concentration of chlorinated solvents takes place across the barrier. Near the back of the barrier most parameters meet the criteria set for remediation. However the concentration of chlorinated degradation products is still quite high here. Since the monitoring wells near the back of the barrier is placed approximately 0,1 meter from the back, the concentrations are expected to be lower when the water leaves the barrier.
Dichloromethane and 1,2-dichloreoethane as well as BTEX are not degraded by corrosion of iron. However, the results show a considerable reduction of these parameters. This is considered to be due to the biological degradation taking place in the barrier.
Clogging of the reactive barrier
Hydraulic conductivity within the barrier shows a tendency of being reduced with time. Development in pH and ions also indicate that precipitation takes place.