The use of borehole radar tomography to monitor

The use of borehole radar tomography to monitor a steam injection pilot study in a contaminated fractured limestone (Maine, USA) C. Grégoire, J. W. Lane and P. K. Joesten U. S. Geological Survey, Branch of Geophysics K. U. Leuven, Department Civil Engineering

Outline • • • Steam-enhanced remediation (SER) technique Modeling: effect of heating on the medium properties Fluid-temperature logs Borehole radar tomography Conclusions

Objectives • SER pilot project in an old limestone quarry, in Maine, USA • Evaluation of radar methods for monitoring the movement of steam and heat through fractured bedrock

SER Technique • Remediation of NAPL contaminated sites in porous media • Heating, vaporisation and removal of volatile and semi-volatile contaminants from the vadose and saturated zones • At the Loring AFB, ground water underlying the bedrock is contaminated with chlorinated solvents and petroleum products

SER Technique at the Loring AFB • Several injection wells on the Northeast of the site • Number of extraction wells across the entire site Injection at depths from 20 m to the bottoms of the boreholes (30 -40 m) • Start of injection and extraction September 1, 2002 • Injection continued until November 19, 2002 • Extraction until November 26, 2002

Location of the boreholes at the Loring Air Force Base Site (Maine, USA)

Modeling: Effect of heating on the medium properties Heating effects: • Heating of water present in the fracture system, with eventual replacement by steam - change of fluid permittivity and conductivity • Variation of limestone matrix conductivity • Variation of velocity and attenuation of radar waves

Modeling: Temperature effect • Temperature 20ºC to 100ºC: Small increase of velocity increase of attenuation (porosity 2%) • Steam replaces water: Increase of velocity and decrease of attenuation (porosity 2%)

Modeling: effect of limestone matrix conductivity

Modeling: effect of water conductivity

Modeling: effect of water permittivity

Modeling: results Heating of medium affects (results porosity 2%): (1) Limestone conductivity increases (0. 1 -0. 5 S/m): v decreases (106 - 95 m/us) a increases (6 - 26 d. B/m) (2) Water conductivity increases (0. 050 -0. 200 S/m): v decreases (106 - 105. 5 m/us) a increases (6. 2 - 8 d. B/m) (3) Water permittivity decreases (80 -50): v increases (107. 5 - 106 m/us) a increases (6. 25 - 6. 30 d. B/m) (1) has the largest effect

Field Measurements • Surveys conducted before start of injection (August 2002), 10 days after start of the injection (September 2002) and near the end of the injection (November 2002) • Casing: ~7 m steel casing at the top, solid fiberglass casing, capped at the bottom, sealed over the entire length by grout; filled with clean water • Borehole radar reflection • Borehole radar tomography (+ levelrun acquisition) • Fluid-temperature logs before each radar survey • Deviation logs

Fluid-temperature logs

Fluid-temperature logs • Measure of water temperature in boreholes: for the safety of the equipment and personnel. • September 2002 (10 days after start of steam injection): temperature was unchanged in well JBW-7816; small increases in well JBW-7817 A. • November 2002 (near the end of steam injection): increase of a max. of 10°C in well JBW-7816; increase above 40°C in well JBW-7817 A. • Larger temperature changes were expected in JBW 7817 A because closer to the steam injection well I-4.

Borehole radar tomography • Data collected between wells JBW-7816 and JBW-7817 A • Identification of velocity and attenuation changes • Acquisition: intervals of 20 cm (after 1 m) • Levelrun acquisition every 20 cm used for data calibration • Inversion of travel-times differences due to the very small changes between the 3 datasets • SIRT algorithm; inversion of levelrun data as starting model

Data calibration Calibration of levelrun data • Time-zero correction • Amplitude correction due to change in antenna power Calibration of the tomography data • Deviation in times: correction based on the levelrun data • Amplitudes variation: correction based on the levelrun data

Calibration of amplitudes (levelrun)

Calibration of traveltimes (tomography data)

Inversion of difference time data

Inversion of difference attenuation data

Results • Increase of slowness (0. 0002 usec/m) at the end of the injection, at larger depths (below 22 m); velocity decreases • Increase of attenuation (0. 7 d. B/m) at the end of the injection, at larger depths (below 22 m) • Consistent with the increase of temperature at these depths • Consistent with the modeling (largest effect is due changes in the limestone matrix conductivity because of the low porosity)

Conclusion Borehole radar tomography: • Calibration of levelrun and tomography data was necessary. • Inversion of difference time data: small increase of slowness (0. 0002 usec/m) associated with heating. • Inversion of difference attenuation data: increase of attenuation (0. 7 d. B/m) associated with heating. • The radar tomography detected thermally induced variations in the fractured limestone.

• The study was supported by the U. S. Geological Survey (USGS) and the KULeuven (University of Leuven, Belgium). • The study was funded by the U. S. Environmental Protection Agency (USEPA) Office of Solid Waste and Emergency Response, Office of Superfund Remediation and Technology Innovation and by the USGS Toxic Substances Hydrology Program.