Geostatistical modeling of permeability and application to geothermal reservoir simulation
This study employs a geostatistical methodological approach to evaluate the permeability of the reservoir, which is first analyzed through statistical characterization, experimental variogram construction, and theoretical model fitting to establish the continuity and variability of the formation. Ordinary kriging is applied to generate deterministic permeability estimates, while Sequential Gaussian Simulation is used to produce stochastic realizations that capture the intrinsic heterogeneity of the geological medium. These geostatistical results provide the permeability inputs required for the subsequent thermal reservoir simulation. The permeability fields obtained from SGeMS were incorporated into a geothermal reservoir model developed in OPM Flow, representing a configuration of injection and production wells. In this stage of the study, the primary focus is the evolution of temperature within the reservoir, particularly the thermal front propagation between the injector and producer wells over long-t erm operation. Post-processing and visualization were conducted in ReInsight. The integrated workflow demonstrates the value of combinin g geostatistical modeling with thermal reservoir simulation to better assess formation heterogeneity, predict temperature distribution over time, and support data-driven decision- making in reservoir management.
Geostatistical modeling of porosity for CO₂ storage assessment in a reservoir
This study presents a geostatistical analysis of porosity in a reservoir, aiming to understand its spatial distribution within the study area and to evaluate its potential for CO₂ storage applications. Descrip- tive statistical analysis and histograms were used to assess the characteristics of data distribution and their suitability for geostatistical modeling. An experimental variogram was constructed and fitted with a spherical theoretical model to describe the spatial correlation structure. Ordinary kriging was applied to estimate porosity and to generate spatial distribution maps, while kriging variance and Gaussian sequen- tial simulation were used to evaluate spatial uncertainty. The resulting porosity model was subsequently employed to simulate CO₂ storage under different injection configurations over a three-year period. The results indicate a well-defined spatial correlation structure, reliable porosity estimation, and a meaningful uncertainty analysis, providing a solid basis for reservoir evaluation and the feasibility of CO₂ storage.