"Steam Injection Technique for In Situ Remediation of Chlorinated Hydrocarbons from Low Permeable Saturated Zones - Experiment and Numerical Approach
"Steam injection is a commonly applied technique to enhance soil vapour extraction for the remediation of contaminants in the subsoil. However, steam propagation is highly dependent on the permeability of the soils. This thesis describes investigations regarding the heat distribution in a low permeable, saturated zone and the contaminant removal efficiency. To this end, a large-scale tank experiment with two distinct steam injection methods was conducted: 1. steam injection in the saturated zone below the low permeable zone (override) and 2. an additional injection into the unsaturated zone (sandwich). The observed heat distribution was compared to 1D and 2D numerical simulations to evaluate reproducibility and applicability for dimensioning of field applications.
The experiment demonstrated the successful heating of the low permeable, saturated layer (thickness of 1.5 m) to about 90°
C at the top of the layer after 16 days using the steam verride method. A constant steamed zone below the low permeable zone was observed. The additional steam injection in the unsaturated zone showed only a limited eff ect on the heat distribution in the low permeable layer and achieved no significant increase in contaminant removal. A total mass of 3 kg of contaminant (46% of initial amount) was removed from the subsoil. 88% was removed by the soil vapor extraction, 11% by the groundwater outflow and 1% by the extracted condensate from the unsaturated zone. Contaminant concentrations decreased by 99% in the soil vapor extraction and by 85% in the groundwater. A persisting contamination of the saturated zone was evident.
The 1D model was not able to simulate the predominated conductive heating correctly and further model modifications regarding the relevant processes for heat distribution were necessary. The 2D numerical model reproduced satisfactorily the heat distribution in the low permeable, saturated layer for an adapted set of parameters. However, shortcomings resulted from the negligence of the third dimension and the estimation of the heat losses via the boundaries. Hence, the model can not be applied for different configurations of parameters and further investigations and calibration efforts are required.