| Front page | | Contents | | Previous | | Next |

Reactive zerovalent iron barrier for removal of TCE and Cr(VI) from groundwater

Summary and conclusions

A preventive project to reduce the secondary groundwater content of trichloroethylene (TCE) and chromium (Cr(VI)), i.e. mixed pollution, has been performed at a property situated at Fabriksvej 4 in Kolding, Denmark (cadastral register No. 128 n, Kolding Markjorder).

The project has been completed with financial support from the Danish EPA Technology Programme for soil and groundwater contamination. The protective technique applied is based on remediation / reduction of the pollution components at the passage of the groundwater through a reactive barrier consisting of zerovalent iron (without the so-called 'Funnel & Gate'). The project was carried out in the period of 1998-2001.

On the basis of the project the following may be concluded:

The preliminary examinations revealed a major contamination of the secondary aquifer beneath the property. This contamination is related to previous activities at the property. The highest concentrations were found in the southern part of the property, where there seems to be a reasonably hydraulically coherent aquifer.

The contamination is most likely spread at a very low rate in the secondary aquifer. Within the hydraulically coherent area, the main flow direction is to the west and a decline of the TCE concentration in the groundwater from east to west was observed. TCE concentration has been observed to a minor extent in the upper aquifer of the adjoining property situated to the west.

No products of decomposition (vinylchloride) have been observed in the upper aquifer and the estimated natural TCE decomposition of the upper aquifer is estimated to be at a minimum. Consequently, the contamination of the upper aquifer constitutes a risk to the surrounding environment as it is not decomposable and there is a clear risk for exposure.

The primary groundwater resource is not immediately threatened by the observed contamination.

The geological conditions on and around the property are highly complex and reflect several glacial pushes. As a result, the hydrogeological structure consists of several hydraulically connected aquifers. The structure of secondary aquifer on the property is very inhomogeneous.

Due to the complex geological and hydrogeological structure of the soil layers on the property, it was decided to extend the passive barrier with a recirculation system. Accordingly, water passing the barrier was accumulated in a drain and recirculated upstream the barrier. This implementation was performed in order to control the current against the barrier and force underwashing of the contamination's hot-spot areas.

During the monitoring it has been observed that this strategy was valid, and it was possible to control the natural groundwater current against the barrier by pumping and recirculating the pumped up, purified groundwater. As a result, sources of error occurred. Despite careful control, there is a risk of altering the flow speed in the barrier and thereby influencing the detention time of the contaminated groundwater in the barrier.

As it is one of the key parameters in this process, the detention time may mean that the transformation of the contamination components does not take place as required. It cannot be denied that this has actually happened in this project.

Due to large variations in the groundwater level of the secondary aquifer, where contamination is prevented, pumping up and the recirculation have not been possible for long periods due to a potential risk of lowering the groundwater level far below the barrier's top level. During these periods, mainly in the spring / summer, it was observed that the flow pattern in the aquifer was re-established, thus there is agreement with the discharge pattern prior to the start of the pumping up.

In relation to the sounding of wells in and close to the barrier, two years of operation indicate that a longitudinal flow (from south to north) in / by the barrier may occur. This can probably be ascribable to the micro variations of the hydraulic properties in the barrier and the surrounding soil layers.

Quality control comprising micro examination using iron granules will provide basic information regarding the quality of the iron. In relation to this project, German iron filings from Fa. Gothard Meier have been used.

At the delivery it was determined that the quality of samples from the same delivery varied, thus the filings of one sample appear more corrosive in comparison to another sample.

This aspect must also be taken into consideration when laying the filings in the barrier. The logistics must be planned so that the iron filings are laid immediately after reception. It is not recommended to store the German iron prior to laying.

As regards the assessment of the iron's quality during the operational period, it has been indicated that it is possible to take undisturbed core samples from the barrier with the aid of Geoprobe technology. At the same time it has been possible to filter new wells in the barrier and thus considerably improve the sampling possibilities in the barrier.

In general the microscopy examination of the core samples’ origin from the barrier showed no sign of iron filing corrosion selected in the interval beneath the groundwater level.

Core samples from intervals, which throughout the entire operational period were found close to or below the groundwater level (where there is every probability of finding oxygen), showed indications of corrosion.

In general it appears that it was possible to reduce the concentrations in the hot-spot area, both as regards to Cr(VI) as to TCE. In general it is estimated that the method of reactive barriers is a well-functioning method.

The tendencies are, however, not quite unambiguous, which is probably due to variations in precipitation and consequently the recirculation load. Relatively large percolation will cause larger underwashing. Therefore, after periods of heavy precipitation (thus a relatively large recirculation) the amount of especially Cr(VI) in the groundwater will increase.

The results of the inorganic analyses indicate lime and ferrous carbonate deposits in the barrier. No deposits were observed on the iron granules, however these presumably exist in the water-filled voids in-between. The results also indicate that the upstream barrier varies, and in time this may cause variations of the barrier's efficiency in the longitudinal direction.

Nitrate was observed in well MW1 (located in the southern part of the barrier), indicating that the ferrous / iron capacity of the southern part of the barrier is critically low - alternatively, that the retention time of liquid is too low.

In spite of reasonably good access, it has not been possible to estimate the original source intensity of the contamination, neither on the basis of the historic information, nor the investigations performed. This is mainly due to the following three conditions:

	Lack of adequate accident descriptions (causes of spillage)
	Spillage in connection with production has resulted in a diffuse contamination.
	The complex geological / hydraulic conditions in the geological bed sequence have caused an uneven distribution of contaminants in the secondary aquifer and in the underlying stratum.

Consequently, there is no starting point to compare with when the efficiency of the barriers is to be described.

However, rough estimates of quantities can made by using the average of the water's content of TCE and Cr(VI) in the monitoring wells at the front barrier shortcut channel, and by comparing these with the pumped up water, which during this period was 6,150 m³ water.

An average estimate of the theoretical calculated quantity of the two substances is:

The monitoring well results at the shortcut channels show that the concentrations at wells MK1 and MK2 are considerably higher than at MK3 and MK4. This indicates that the load factor on the barrier, as anticipated, is higher in the southern part.

If, with a traditional viewpoint, we assume that the above approx. 50-100 kg Cr(VI) are removed through passage in the southern part of the barrier, approx. 50-100 tonnes of zerovalent iron will be used, corresponding approximately to half of the total capacity of the barrier's resistance to Cr(VI). As the maximum load on the barrier is on the southern part, a possible future CR(VI) break through in the southern barrier cannot be ruled out.

In fact the conditions are more complex; the inhomogeneity of the geological bed sequence and the barrier presumably require a configuration of flow that provides another load on the barrier. At the same time moving the load on the barrier from south to north by regulating the percolation has been tested. Naturally, this causes elements of uncertainty in connection with the calculation of estimates as mentioned above.

The remaining lifetime of the wall is estimated to be 2 years, which correspond to the previous period during which half of the capacity of the barrier was used.

Main conclusions:

The following conclusions can be drawn from the project:

The principle of preventive measurements by reducing the dehaloganation of chloride solutions, at the same time reducing the Cr (VI) solutions, is successful and could be used in Denmark under certain conditions.

It is, however, not recommended to use this methodology in areas of high geological heterogeneity where the natural discharge in the aquifers is not completely defined. Avoid as far as possible (apart from the so-called "Funnel & Gate Systems") introducing further methods to lead the groundwater. This could result in a reduced effect of the principle reduction of the contaminants.

It is possible to analyse the quality of the iron filings delivered by means of a microscope analysis of the samples. The quality can vary considerably, as parts of one delivery of iron granules could have different corrosion qualities.

Logistics are an important parameter when delivering iron filings for the establishment of reactive barriers. Temporary storage of iron filings should be avoided.

At barriers established at groundwater level, it is possible to take out drill cores of the barrier after it has been established. This enables more effective regeneration of selected areas of walls in other projects. It is also possible to place a sampling filter in the barrier by using the so-called Geoprobe® technology.

A mixture of the contaminants TCE and Cr(VI) - Cr(VI) will probably use the capacity of the barrier prior to TCE and accordingly provide an option for a breakthrough of the barrier.

Deposits of lime and ferrous carbonate in the barrier may occur. No deposits were observed in the iron filings, however these may exist in the water-filled voids.

If compared to construction costs, the following key figures for the barrier effectiveness can be given:

DKK 13,000 - 26,000/kg of removed Cr(VI)
DKK 214/m³ of cleansed water

An educated guess at the treatment price at a facility is DKK 2,000-3,000, exclusive of transportation costs. The barrier as such has given a substantial cost reduction as compared to traditional treatment.

Initially a series of technical specifications (success criteria) for the barrier was stated. In the following table these criteria are evaluated:

The remedial project as carried out thus meets 80-90% of the success criteria.

| Front page | | Contents | | Previous | | Next | | Top |