Advanced Petrophysical Characterization of Thin-Bedded Reservoirs Through Integrated Laboratory, Well Logging, and Seismic Data

Thin-bedded Miocene formations of the Carpathian Foredeep represent a major challenge for reliable petrophysical characterization due to strong lithological heterogeneity, high clay minerals content, and the vertical resolution of standard well log interpretations. Accurate assessment of porosity and permeability in such reservoirs is essential for hydrocarbon exploitation, geothermal applications, and the evaluation of Carbon Capture and Storage (CCS) potential.

This study presents an integrated pore-structure modeling workflow applied within a multi-well correlation framework, allowing the transfer and validation of petrophysical models across laterally variable, thin-layered deposits. The methodology combines multiscale laboratory measurements (NMR, nitrogen adsorption, MICP, X-ray CT, and FIB-SEM) with machine-learning–assisted interpretation of well log data to generate high-resolution continuous profiles of porosity and permeability. Models calibrated on core-scale laboratory data are propagated between correlated wells, enabling consistent characterization of reservoir properties beyond a single well.

To increase the geological credibility of the multi-well interpretation, seismic data are incorporated as an independent constraint. Seismic attributes support stratigraphic correlation, identification of thin-bed architecture, and the lateral continuity of petrophysical units. This integration facilitates the upscaling of pore-scale information from laboratory and well log data into a seismic framework, reducing uncertainty related to heterogeneity and thin layering.

The results indicate that the combined use of artificial neural networks, advanced statistical methods, and seismic support significantly improves both vertical and lateral resolution of petrophysical properties in thin-bedded reservoirs. The proposed workflow enables reliable application of pore-network–based models within a multi-well context and provides a scalable approach for reservoir characterization in complex clastic systems. The methodology is particularly relevant for unconventional reservoirs and mature fields considered for CCS or geothermal repurposing, where accurate representation of thin-layered architectures is critical for realistic resource assessment.