John Deceuster

Improving the delineation of hydrocarbon-impacted soils and water through induced polarization (IP) tomographies: a field study at an industrial waste land.

Without a good estimation of samples representativeness, the delineation of the contaminated plume extent and the evaluation of volumes of hydrocarbon-impacted soils may remain difficult. To contribute to this question, a time domain induced polarization (IP) field experiment was conducted on an industrial waste land. Boreholes were drilled to specify the local geological context. Cross-hole seismic tomographies were performed to extend borehole logs and to draw an interpreted geological cross-section. Soil samples taken during drillings were analysed in laboratory. A preliminary survey was conducted to locate the IP profile. The polarization signatures linked to the presence of clayey sediments were filtered out from the data set. Chargeability and resistivity depth soundings were computed and compared to mean concentrations of total organic products to overcome the data support issue between the geophysical models and the spot samples of soils. A logarithmic relation between chargeabilities and smoothed hydrocarbon concentrations in soils was found. Taking into account contaminants concentration thresholds defined in local codes and regulations allows defining chargeability classes to delineate hotspots on this site. This showed that IP tomography can be an accurate screening methodology. A statistical methodology is proposed to assess the efficiency of the investigation strategy.

A 3D Resistivity Tomography Study of a LNAPL Plume Near a Gas Station at Brugelette (Belgium)

A field experiment was conducted over a Light Non-Aqueous Phase Liquid (LNAPL) contaminated site located near a gas station where a tank had leaked its content. In this area, a dolomitic bedrock is overlain by 5to9m of clayey sands. In order to delineate the plume and measure piezometric heads and depths to bedrock, ten boreholes were drilled and four cone penetration tests were conducted. Soil and water samples were collected and analyzed. Although these samplings highlight the presence of hydrocarbons in some boreholes, the plume is poorly outlined due to the limited number of drillings. To predict the response of hydrocarbon-contaminated areas in this specific context, a laboratory study was undertaken. It shows that an increase in resistivities of about 40% should be found in highly polluted areas. To assess the contribution of geo-electrical investigations in delineating hydrocarbon-contaminated areas, a 3D cross-diagonal resistivity survey was performed using a roll-along technique. The electrical d...

A New Method to Quantify Carbonate Rock Weathering

The structure and composition of carbonate rocks are modified greatly when they are subjected to phenomena that lead to their weathering. These processes result in the production of residual alterite whose petrophysical, mechanical, and hydrological properties differ completely from those of the unweathered rock. From a geotechnical perspective, it is important that such changes are fully understood as they affect reservoir behavior and rock mass stability. This paper presents a quantitative method of calculating a weathering index for carbonate rock samples based on a range of petrophysical models. In total, four models are proposed, each of which incorporates one or more of the processes involved in carbonate rock weathering (calcite dissolution, gravitational compaction, and the incorporation of inputs). The selected weathering processes are defined for each model along with theoretical laws that describe the development of the rock properties. Based on these laws, common properties such as rock density, porosity, and calcite carbonate content are estimated from the specific carbonate rock weathering index of the model. The propagation of measurement uncertainties through the calculations has been computed for each model in order to estimate their effects on the calculated weathering index. A new methodology is then proposed to determine the weathering index for carbonate rock samples taken from across a weathered feature and to constrain the most probable weathering scenario. This protocol is applied to a field dataset to illustrate how these petrophysical models can be used to quantify the weathering and to better understand the underlying weathering processes.

The Role of Geophysical Investigations in Physical Planning over Covered Karstic Areas in Southern Belgium

Since 1999, maps (at 1:10,000 scale) of zones of karstic constraints have been established in the Walloon Region of Belgium (the southern part). These maps sort areas in terms of karstic hazards (negligible, low, moderate and high). They were drawn based on geological settings, field observations, background knowledge and geostatistical forecasts. Nowadays, these maps are used as background documents in physical planning over covered karstic areas. Indeed, the local government now requires to conduct geophysical and/or geotechnical investigations previous to building construction over moderate and high karstic hazard areas. In this context, we conducted several 2D, quasi-3D and 3D resistivity tomography surveys using a total number of electrodes ranging from 100 to 1000. 3D resistivities were measured using a cross-diagonal survey with a combination of dipole-dipole, pole-dipole and equatorial dipole arrays. The inversions were computed within Res2DInv and Res3DInv. The results of these investigations allowed pointing out resistivity anomalies interpreted as effects of karstic features. When possible, boreholes are then drilled over and away from the detected anomalies to confront geophysical results. Finally, geophysical interpretations are used to adapt construction projects.

Correlation between Inverted Chargeabilities and Organic Compounds Concentrations in Soils – A Field Experiment

Resistivity and induced polarization tomographies were performed along profiles on an industrial waste land to detect and delineate hydrocarbon-impacted areas. To confront geophysical results and to conduct cross-hole measurements, 4 boreholes were drilled. Soil samples were taken and analysed. Concentrations of organic compounds in soils were then compared to inversed chargeabilities obtained from surface measurements. This comparison shows that chargeabilities are linked to contaminant concentrations in soil samples except from areas where clays were found. A filtering technique based on normalized chargeability values was implemented and validated in order to discriminate chargeabilities linked to the presence of clayey sediments and contaminant concentrations in soil samples.

Management of Sinkhole Risks Using Long Term ERT Monitoring - A Laboratory Experiment

Accurate methodologies are required to manage risks linked to land-use planning in covered kart terrains, especially in densely urbanized areas. The main risk lies in the occurrence of sinkholes at the base of buildings or infrastructure. We conducted a laboratory experiment to evaluate the contribution of ERT monitoring in the long term management of such karstic risks. After presenting the design of the laboratory experiment, we detail the selected scenarios and the acquisition protocols tested. The methodology proposed to process the data and manage the inversion results relies on two steps: (1) we estimate the resistivity variations due to measurement and inversion errors based on Monte-Carlo simulations and (2) we define a resistivity changes index for every cell of the 3D model. The methodology is tested on a 3D surface survey including inline dipole-dipole, equatorial dipole and Wenner-Schlumberger arrays. A 10 cm in diameter plastic ball is used to model a 3.5 m sinkhole at depths ranging from the surface to 20 cm, using a water resistivity of 12 Ohm.m. Based on the proposed methodology, we show that this target can be suitably detected when its top reaches 10 cm or less in depth.

Detection and Mapping of a Palaeokarstic Network in a Quarry Work Using GPR

A Ground Penetrating Radar (GPR) survey was conducted on the bench of a limestone quarry to detect and map a network of weathering. The investigated area is a rectangle of 16.60 m by 10.25m on the Carboniferous limestone from southern Belgium. Profiles are parallel and equally spaced of 25 cm from each other and the distance between each trace along a profile was 2 cm. The GPR survey was conducted using a 250 MHz antenna. After filtering, 2D-sections show very clearly weathered features to at least 10m in depth. A 3D model of the area was created to identify the branches of the weathered network. On time slices, these features are well located throughout the entire model and anthropic cuttings lines are recognized on their total height (~6m).

Detection of a Fracture Network Near a Cluster of Sinkholes using 3D Electrical Resistivity Tomography

Over the last two decades, 2D electrical resistivity tomography (ERT) surveys have been conducted to detect and map karstic features as the expected electrical resistivity contrast is high between sound limestone and weathering residue as well as between sound limestone and cavities filled with clayey or sandy sediments. However, karstic features often show complex 3D geometries which are not well reconstructed through 2D surveys. In this paper, the efficiency of 3D ERT in the identification of a fracture network located at the top of limestone bedrock is assessed through a field experiment. The results of the 3D ERT interpretations were confirmed by visual observations of an outcrop located 100 meters away. Moreover, former sinkholes which occurred near the investigated area are aligned following the directions pointed out with the 3D ERT survey. This indicates that local karstification probably affect fractures located at the surface of the bedrock leading to the formation of a cryptolapiaz.

Determining Background Variations to Identify Significant Changes in Resistivity and IP for Time-lapse Experiments

Electrical Resistivity Tomography conducted in time-lapse mode is frequently used for the monitoring of time-varying processes. Nevertheless, it is tricky to identify significant changes in resistivity and chargeability when monitored processes vary slowly in time. Moreover, the errors on acquired measurement might not be negligible as well as the uncertainties on the blocks of the reconstructed models. The uncertainty distribution depends especially on selected arrays, resistivity and chargeability distributions and model blocks sensitivities. An estimate of the background variations for every block of the models is thus required prior to the mapping of resistivity and IP changes during the monitoring experiment. A new methodology is proposed to estimate these background variations and to point out significant changes in resistivity and chargeability. This methodology is based on a two steps approach. The first step consists in determining a sensitivity cut-off value to estimate the depth of investigation. The second step aims to estimate resistivity and chargeability confidence intervals of each model block based on Monte-Carlo simulations. This methodology was applied to a field monitoring experiment conducted on a site contaminated with chlorinated solvents where biodegradation remediation is performed to assess its efficiency.

A Bayesian-based Methodology to Detect Suspicious Electrodes for Long-term Permanent ERT Monitoring Experiments

During long-term ERT monitoring experiments, the data quality may reveal time-dependent mainly due to changes in galvanic contact resistance between buried electrodes and soil. Identifying suspicious electrodes in a permanent spread is of major importance as a faulty electrode may affect the quality of tens to hundreds of measurements on each time-slice. An automated methodology was developed to detect suspicious electrodes based on a Bayesian approach. This methodology allows pointing out faulty electrodes based on the analyses of temporal sets of measurements, each one containing multiple electrode arrays. Standard and studentized estimators of the influence of each electrode in the global data quality are computed for each time-slice based on the measurement quality factor Q given as a coefficient of variation of repeated measures. The automated detection of faulty electrodes is obtained by comparing the computed studentized estimators to values expected when every electrode can be considered as good for the given data set. These expected values are computed by Monte Carlo simulations using a distribution of Q factor values of quadripoles selected as good based on reciprocal errors. The efficiency of the proposed methodology is assessed on a field experiment.