Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Antoine Bouziat has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Deep Learning Applications to Unstructured Geological Data: From Rock Images Characterization to Scientific Literature Mining

    by Antoine Bouziat, Sylvain Desroziers, Mathieu Feraille, Jean-François Lecomte, Renaud Divies, François Cokelaer


  • Open Access

    An Assessment of Stress States in Passive Margin Sediments: Iterative Hydro-Mechanical Simulations on Basin Models and Implications for Rock Failure Predictions

    by Antoine Bouziat, Nicolas Guy, Jérémy Frey, Daniele Colombo, Priscille Colin, Marie-Christine Cacas-Stentz, Tristan Cornu

    Capturing the past and present hydro-mechanical behavior of passive margin sediments raises noticeable interest, notably in geo-hazard risk assessment and hydrocarbon exploration. In this work, we aim at assessing the stress states undergone by these sedimentary deposits through geological time. To do so, we use an iterative coupling between a basin simulator and a finite element mechanical solver. This method conciliates a computation of the full stress tensors with a dynamic and geologically detailed modelling of the sedimentation. It is carried out on a dedicated set of 2D synthetic basin models, designed to be representative of siliciclastic deposition in passive margins and integrating variations in their geological history. Contrary to common assumptions in operational basin modelling studies, our results imply that passive margin sedimentary wedges are multidimensional mechanical systems, which endure significant non-vertical stress without external tectonic input. Our results also highlight the variability of the stress states through space and time, with a strong control from the geometry and lithological heterogeneities of the deposits. Lastly, we used the simulation results to predict a location and timing for the development of weakness zones in the sedimentary stacks, as privileged areas for rock failure. The outcome underlines the influence of the basal tilt angle, with a slight tilt impacting the wedges stability to a similar extent as a substantial increase in sedimentation rate. Altogether, this study emphasizes the need for careful consideration of non-vertical stresses in basin simulations, including in passive tectonic contexts. It also suggests that the iterative coupling method employed is a promising way to match industrial needs in this regard.


  • Integrating Active and Passive Seismic Data to Better Predict Hydraulic Fracturing

    by Gabino Castillo, Kevin Chesser, Antoine Bouziat, Guy Oliver, Chi Vinh Ly, Lih Kuo, Eric Bathellier

    Although drilling has slowed substantially from its peak in 2010, steadily improving natural gas prices coupled with the promise of demand from liquefied natural gas and gas to liquids facilities have renewed interest in the prolific Haynesville shale gas play in NW Louisiana. A consensus of opinion among operators in the field will agree that only a fraction of the Haynesville potential has been developed to date. As is the case in most shale plays, production from wells has been highly variable, leading to the use of 3D seismic reservoir characterization studies for the determination of sweet spots, well placement and completion strategies where seismic anisotropy has been proven to be an important factor in understanding the shale plays. This paper illustrates a workflow (Figure 1) integrating reservoir and geomechanical properties obtained from pre-stack seismic inversion and incorporating stress and fracture information extracted from azimuthal analysis of the seismic data. Eight wells in the area targeting the Haynesville and mid-Bossier reservoirs were used for calibration of surface seismic measurements of reservoir and geomechanical properties. A variety of seismically derived attributes are used to estimate production potential in the field. This paper shows the application of global azimuthal inversion, a technology for extracting the azimuthal anisotropy. Above all, the workflow makes quantitative use of microseismic and SEM (Scanning Electron Microscope) derived mineralogy data to validate the seismic-derived attributes. Figure 1 Integrated geoscience workflow for a shale play, incorporating numerous discipline to high-grade the survey area to identify ‘sweet spots’ and optimize drilling locations and completions.


  • Seismic-Based Stochastic Porosity Prediction in a Giant Field: A Fine-Scale Quantitative Approach in Carbonates.

    by Shubhabrata Samantaray, Remi Moyen, Jean-Marc Michel, Antoine Bouziat, Ashraf Yahia, Naema Al Zaabi

    Stochastic inversion of 3-D seismic data is used increasingly for reservoir characterization. It provides information on the reservoir at a much finer scale than deterministic inversion and delivers multiple scenarios for uncertainty analysis. In this work, a stochastic seismic inversion workflow has been developed to characterize porosity variations in the Thamama reservoir of the giant Field, onshore Abu Dhabi. In contrast to traditional band-limited inversion, this stochastic inversion workflow first generates high frequency models of acoustic impedance which can be used directly for geomodelling without downscaling. Next, a collocated co-kriging sequential Gaussian simulation technique has been applied to generate fine scale 3-D porosity realizations constrained by the impedance stochastic models and by log porosity data. In the example, post-stack stochastic inversion is combined with stochastic porosity modeling to characterize the uncertainty in the spatial distribution of thin, low porosity / permeability intra-Thamama layers, which adversely affect the field water flood performance. These thin layers have been mapped using seismic-constrained stochastic workflow. P10, P50 and P90 porosity realizations have been generated which represent more or less pessimistic scenarios of lateral extent of the tight zone along the flank of the field. A number of blind wells demonstrate that the seismic-based workflow provides more accurate porosity predictions than a purely well-based reservoir model. Specific technical contributions of the work include:Demonstrate the value of seismic information for characterizing mature carbonate reservoirsImplement field-specific stochastic workflow to characterize uncertainty in spatial extent of thin reservoir flow barriers from 3-D seismic dataPerform seismic inversion directly in fine-scale stratigraphic grid to facilitate integration with the field geomodel


  • Integrating Surface Seismic, Microseismic, Rock Properties and Mineralogy in the Haynesville Shale.

    by Neil Peake, Gabino Castillo, Norbert Van de Coevering, Simon Voisey, Antoine Bouziat, Kevin Chesser, Guy Oliver, Chi Vinh Ly


  • A parametric method to model 3D displacements around faults with volumetric vector fields

    by Gautier Laurent, Guillaume Caumon, Antoine Bouziat, Mark Jessell


  • Vector Field Based Fault Modelling and Stratigraphic Horizons Deformation

    by Antoine Bouziat

    Consequent choices during hydrocarbon exploration and field developement depend on 3D numerical models of the main geological interfaces and the major faults in the area of interest. Such structural models provide decisive informations like reservoir rock volume or compartmentalization, and are unavoidable stages to build the reservoir grids used for resources estimation, flow simulations and well planning. However, these models need to be continually modified, either to integrate new information from well drilling and production history or to consider several hypotheses about the most uncertain features of the fault network. To do so, petroleum engineers involved in reservoir modeling need flexible and efficient tools to easily edit parts of structural models while preserving their general consistency. In a first step to handle structural model edition challenges, we propose a new way to numerically model a fault object and deform previously modelled stratigraphic horizons. This method is inspired by the "Vector Field based Shape Deformations" (VFSD ) techniques used by the Computer Graphics community. Our approach considers a fault as a 3D vector field, interpolated thanks to several scalar fields linked to the fault geometry and few scattered data points if available. This vector field is then integrated into path lines able to drive the deformation of the surrounding objects in a purely geometrical manner. The so-obtained deformation is limited to an influence region neighboring the fault and can affect several horizons simultaneously in a consistent and reversible way. Moreover, self-intersections can be prevented and both the first order smoothness and the fine-scale features of the deformed horizons are conserved. The vector field interpolation methodology we present is based on a user-defined conceptual model for deformation attenuation and a direction constraint on the displacement. It was implemented in common geomodelling software. The resulting tool is applicable on a wide range of fault geometries, and can be added as a new exit condition for fault network stochastic simulation algorithms. Last, some derived applications, related to salt diapir modelling or paleotopography restoration, are also conceivable.