Pore space characterization in carbonate rocks — Approach to combine nuclear magnetic resonance and elastic wave velocity measurements

• A new workflow for pore space characterization in carbonate rocks is presented.
• NMR and elastic wave velocity data are combined to derive average pore dimensions.
• An alternative method to derive Spor from NMR data is tested.
• Experimental data of micritic and oomoldic carbonates are studied.
• Carbonate fabric types are related to distinct ranges of petrophysical parameters.

Pore space features influence petrophysical parameters such as porosity, permeability, elastic wave velocity or nuclear magnetic resonance (NMR). Therefore they are essential to describe the spatial distribution of petrophysical parameters in the subsurface, which is crucial for efficient reservoir characterization especially in carbonate rocks. While elastic wave velocity measurements respond to the properties of the solid rock matrix including pores or fractures, NMR measurements are sensitive to the distribution of pore-filling fluids controlled by rock properties such as the pore-surface-to-pore-volume ratio. Therefore a combination of both measurement principles helps to investigate carbonate pore space using complementary information.In this study, a workflow is presented that delivers a representative average semi-axis length of ellipsoidal pores in carbonate rocks based on the pore aspect ratio received from velocity interpretation and the pore-surface-to-pore-volume ratio Spor as input parameters combined with theoretical calculations for ellipsoidal inclusions. A novel method to calculate Spor from NMR data based on the ratio of capillary-bound to movable fluids and the thickness of the capillary-bound water film is used. To test the workflow, a comprehensive petrophysical database was compiled using micritic and oomoldic Lower Muschelkalk carbonates from Germany. The experimental data indicate that both mud-dominated and grain-dominated carbonates possess distinct ranges of petrophysical parameters. The agreement between the predicted and measured surface-to-volume ratio is satisfying for oomoldic and most micritic samples, while pyrite or significant sample heterogeneity may lead to deviations. Selected photo-micrographs and scanning electron microscope images support the validity of the estimated representative pore dimensions.