Home | Sitemap | Imprint | Privacy Statement | Deutsch

1.1.08 Direct oxygen injection – measuring and modelling dynamic gas accumulators

Many organic pollutants in groundwater are biodegradable under aerobic, or oxygenated, conditions. Adding oxygen can accelerate the degradation process; a relatively cost-effective way of doing this is to use direct gas injection. A project team investigated the processes determining the efficiency of such direct gas injections in porous media (aquifers), and developed forecast models for use within remediation. Several demonstration projects have already been successfully completed on the basis of the research results.

One way to accelerate the degradation of pollutants underground is to use direct gas injections – an in-situ remediation procedure. Injection lances moving both horizontally and vertically are used for the targeted introduction of oxygen into the stream of contaminated groundwater, where it establishes itself in the aquifer in the form of small, finely distributed bubbles. Low-permeability layers of sediment prevent the gas from dissipating upwards; sideways-expanding oxygen-flow capillary networks form instead. The immobile gas phase acts hydraulically and biologically like a reactive oxygen wall that cleans the groundwater flowing through it. This means the bubbles dissolve slowly and add oxygen to the groundwater passing by, supporting the process of degrading the pollutants it contains, while new oxygen is constantly fed in from above.

Controlling the spatial effect of the gas accumulator

Between 2000 and 2010, the BMBF financed the SAFIRA, PROINNO 1 and 2 and ZIM demonstration projects in order to research this procedure. The research sites included the site of a former chemical plant in Leuna and the Auensee recreation area in Leipzig, which is under threat from the groundwater being contaminated with PCE-TCE. Conventional technologies for direct gas injection are based on homogeneous gas distribution through lances. The distribution itself is not measured, meaning that these technologies operate “blind”. Basically, the gas distribution at the injection lances behaves like a dynamic gas accumulator comprising branch-like and inter-related (coherent) channels of gas and non-related (incoherent) gas clusters. The injection or release processes cause these either to expand or draw closer together.

3D representation of gas distribution underground with LPI and HPI direct gas injection

3D representation of gas distribution underground with LPI and
HPI direct gas injection
3D representation of gas distribution underground with LPI and HPI direct gas injection

The heart of the innovation and the scientific challenge of the projects lay in the small-scale metrological recording of the gas dispersion and accumulation processes in the heterogeneous conditions underground and the interpretation and control of these using computer models. The aim was to control the spatial effect of the gas accumulation. As well as conventional low-pressure injection (LPI), the scientists also experimented with high-pressure injection (HPI) for the first time.

Improved monitoring

Sensatec in Kiel worked together with the UFZ centre for environmental research at Leipzig-Halle to develop a new systems technology for coupled LPI-HPI direct gas injection. This involved both a reliable in-situ gas measurement system and a dynamic gas dispersion model on the basis of which the new injection process can be controlled and optimised. The measuring system features an innovative set of sensors (sensor array), which enables it to measure and store large amounts of data extremely quickly. This makes it suitable for recording changes in the gas transport and accumulation processes in the otherwise heterogeneous conditions underground. As this is much faster than typical groundwater transport processes, standard systems are unsuitable for this type of monitoring. With regard to the measurement data, the scientists used suitable geostatistical processes to calculate intermediate values (interpolation) and visualised all the data in 3D. These 3D data fields formed the basis for developing the gas dispersion model.

Example 3D computer simulation of gas distribution with LPI and HPI direct gas injection

Example 3D computer simulation of gas distribution with LPI and HPI direct gas injection
Example 3D computer simulation of gas distribution with LPI and HPI direct gas injection

Technologically relevant results

The research projects produced the following results:

  • Independent measuring of horizontal and vertical hydraulic conductivity is essential for model-supported forecasting of dynamic gas accumulation underground. Thin clay layers in particular can significantly influence the dynamics of the gas flows reaching vertically upwards. In existing groundwater models, the vertical conductivity was calculated from the horizontal conductivity using an empirical factor (< 1). However, this basic approximation breaks down when it comes to predicting the dispersion behaviour of dynamic gas accumulators.
  • A fine-scale site survey is required to ensure the optimum vertical and horizontal positioning of injection lances and sensors. As a minimum, monitoring must involve dynamic drilling and injection logs in order to determine hydraulic conductivity and geoelectric profile recordings.
  • The gas must be measured in a sealed in-situ sensor net adapted to the local conditions (approx. 60 measuring points per injection lance; 1 sensor per m3), which is essential for successful predictions of gas dispersion in heterogeneous sediments.
  • The experts derived from the laboratory experiments a working hypothesis that low-pressure injections lead to incoherent gas transport and high-pressure injections lead to coherent gas transport. This finding is important when it comes to ascertaining the correct dimensions for the gas barriers.
  • A sensor system must be able to measure and store lots of data in a short period of time and to differentiate between coherent and incoherent gas transport.
  • LPI scenarios and combined LPI/HPI scenarios with an extremely low injection rate of 0.18 cubic metres an hour lead to an incoherent accumulation of gas. Conversely, LPI scenarios with an injection rate ten times higher than this and pulsed HPI scenarios lead to a coherent accumulation of gas. This means that it is clearly the injection rate and not the injection pressure that predominantly affects the various gas distribution patterns. The only significant difference between the two injection methods is that high-pressure injection achieved a higher level of gas saturation in the lower range.
  • A 3D gas dispersion model for optimising the coupled LPI/HPI gas supply procedure must:
    a) be a multi-phase model,
    b) factor in heterogeneous horizontal and vertical permeability and capillary pressure fields,
    c) use parameter fields conditioned for the measurement data.
  • The experiments regarding the gas dispersion processes conducted in the lab and the field are also of great interest to CCS technology, i.e. the underground accumulation of greenhouse gases.

Project website http://www.ufz.de/index.php?en=1623

Helmholtz Centre for Environmental Research – UFZ
Hydrogeology department

Prof. Dr. Helmut Geistlinger
Theodor-Lieser-Straße 4
06120 Halle, Germany
Tel.: +49(0)3 45/5 58-52 20
Fax: +49(0)3 45/5 58-55 59
E-mail: helmut.geistlinger@ufz.de
Funding reference: 02WT9947/8
Ressource Wasser
Quick view
Project website