Mouhamadou Sow

Developed Python-based workflow to homogenize well markers across multiple sources and local basins. Developed Machine Learning workflow for markers depth prediction. Synthesized stratigraphic charts, marker data, and well logs to establish a set of trusted markers per well and basin. Automated calculation of petrophysical properties (PHIE, VCL, Net, NTG) per interval for multiple wells using log data. Applied to key regions: CCS studies in Norway North Sea and UK Exploration in Lower Congo Basin, Oman, and Malaysia. Enabled basin-scale reservoir/source rock summaries across multiple wells and saved time by 75%. Performed source rock characterization in the Precambrian intervals of the Oman Basin. Tested classical TOC estimation methods: Passey method (TOC_DlogR), Schmoker method (TOC_RHOB). Utilized Geolog software to analyze log signatures: GR, Neutron-Density, Sonic, and Resistivity. Generated and refined crossplots iteratively to calibrate and interpret TOC estimates. Created a benchmark dataset for future TOC predictions using machine learning.

Automated seismic-to-well calibration with Python. Data export from SISMAGE, preparation/cleaning/Visualization. Automatic seismic trace extraction at well location with Python. Wavelet extraction and Tdlaw generation. QC on SISMAGE by comparing new Tdlaw to existing and visual control on markers placement on seismic. Interaction with teams and technical presentations. The added value at the end was that part of the code for automatically extract seismic traces was used in Namibia exploration wells for further ML workflows.

3D Seismic interpretation on West African Atlantic margin (ROP block, Senegal). Horizon tracking and picking/Computation of RGT models. Seismic stratigraphy approach and well data integration (markers and logs). Determined hydrodynamic and geological processes for sediment waves formation. Realized Geological and depositional models. Computed attribute maps with Paleoscan like RMS, Coherency, Sweetness, Spectral Decomposition. Channel system and associated levees detection on attribute maps to understand sediment waves formation. Interaction with teams and technical presentation. The added value was to better characterize these bedforms for potential reservoir/traps/seal for exploration and ccs.