Summary and conclusions, Miljøstyrelsen

Spreading and transport of MTBE in groundwater

Summary and conclusions

Under the Technology programme for soil and groundwater pollution, the Danish Environmental Protection Agency has in co-operation with Fyn County initiated this project concerning the spreading and transport of MTBE in groundwater. The project is one of several projects, which have been initiated to elucidate different aspects of soil and groundwater pollution with MTBE, and to clarify the magnitude of the risk that this contaminant poses to groundwater resources and future drinking water supplies.

The project site, Vestergade 81, Særslev, 4571 Søndersø, is a former gas retail station that has been found to be heavily contaminated with gasoline containing MTBE.

The project is carried out in three phases; two investigation phases (a preliminary characterisation and a detailed field investigation) and a reporting phase.

The preliminary characterisation was carried out using traditional well drilling and sampling techniques to describe the geological layers including installation of screens in the top of the secondary aquifer. The detailed field investigation was carried out by collecting discrete depth level water samples and performing gamma logs and conductivity logs using the GeoProbe-system.

All water samples have been analysed for MTBE and BTEX on a field-GC. On the basis of these results, a number of water samples were selected for laboratory analysis (Purge & Trap) for MTBE, BTEX and degradation products of MTBE. A good correlation between the MTBE-concentration was found for the two methods, which makes it possible to combine these methods to save time, improve economy and facilitate a flexible investigation strategy.

The structure of this report is that the results from the two investigation phases are presented and described chronologically in order to follow the process as realistically as possible. Thus no overall evaluation is made until all the results are presented.

All figures are presented in appendix 13 in A4-size to allow a more detailed study of these.

Preliminary characterisation

Based on the geological descriptions for the drilled wells carried out in phase 1, a conceptual geological model has been made. The overall geology of the site consists of sandy clay till to a depth of approx. 10 m below ground level, where a water-bearing layer of sand till as well as coarse sand/gravel was found.

There is a good correspondence between the different methods used for the assessment of the hydraulic conductivity (slug tests, pump tests and assessments based on grain size distribution). There is also a good correlation between the geological composition of the soil samples and the estimated hydraulic conductivity for the actual section.

In phase 1, the hydraulic conductivity in the top of the aquifer was assessed to be relatively constant across most of the site, and therefore a general rate of flow could be estimated. However, a relatively large vertical variability of the hydraulic conductivity is found.

The flow direction in the secondary aquifer is assessed to be north-eastern, with a general pore water flow of approx. 40 m/year.

High concentrations of contaminants, such as total hydrocarbons as well as BTEX, MTBE, and degradation products of MTBE, are found in the hot spot (B5). Furthermore, reduced redox conditions were found around the hot spot, which is likely due to the degradation of contaminants.

Low contents of contaminants were found in all other wells within a radius of 40 m from the source area. The presence of degradation products from MTBE indicates that there is an ongoing degradation of MTBE in the ground water at the site.

The measurements to date do not reveal a well-defined horizontal MTBE plume. However, the contaminant distribution and the flow direction indicate that a plume must be present at the site, and that it is likely to spread in a north-eastern direction, between the wells B6 and B8. Thus B6 is positioned at the western fringe of the plume. The concentration of pollutants in the ground water (BTEX and MTBE) indicates that a plume is formed in the upper part of the aquifer, downstream of the source of pollution.

Detailed field investigation

The discrete depth level water sampling in phase 2 has given rise to a substantial revision of the assessment of both the contaminant distribution and the hydrogeology of the site.

The MTBE contamination at the site is found to have spread to a great depth (13 m below groundwater level) immediately downstream of the source, and from here it spreads across the entire depth of the secondary aquifer.

The vertical spreading is likely due to a downward pressure gradient (1.2 m/m) between the secondary and the primary aquifer, causing a vertical flow component and hence the downward spreading of pollutants.

A considerable reduction in the concentrations over a short distance downstream the source of both MTBE and BTEX has been observed. Furthermore, MTBE degradation products are found in many of the water samples. The redox conditions at the site indicate that there is an ongoing degradation of organic components, in that nitrate, iron and sulphate reduction is observed. Thus there is a clear indication of an ongoing degradation of MTBE as well as BTEX, but it has not been clarified whether the degradation is solely responsible for the removal of the contaminants or whether other processes contribute.

The precise distribution pattern in and around the source area (horizontal vs. vertical) has not been fully delineated. This investigation has not clarified to which extent the observed reduction in concentrations is due to degradation, or which alternative processes might have caused this reduction.

The overall purpose of the project was to describe the spreading of the MTBE plume under "typical" Danish conditions, and to estimate general degradation rates, which could then be applied to other Danish sites.

Due to the local hydrogeological conditions at the actual site, it is not possible to produce representative degradation rates, and therefore it has been decided not to carry out phase 3 of the project.

The conceptual model, which formed the basis of this project, was based on a 2-dimensional model for a plume formation in the upper part of the aquifer. Therefore, the model did not allow for a potential vertical spreading of the contaminants. The investigation strategy was prepared on the basis of the model, and the wells carried out in phase 1 (screened in the very top of the aquifer) did not reveal any spreading of pollution over depth. The flexible strategy of phase 2, on the other hand, made it possible to adjust the investigations during the course of the project, in order to take into account the 3-dimensional contaminant distribution observed.