Predicting capillarity of mudrocks

This study provides a compilation, evaluation and correlation of published petrophysical datasets determined for 233 rock samples (165 mudrocks, 27 sandstone, 25 carbonate, 11 anhydrite and 5 marlstone datasets). With predominant focus on mudrocks a review of the methods used for determination of capillary breakthrough and snap-off pressures is given. Additionally, based on more recent data, previously published empirical correlations are critically investigated. Knowledge about these two critical pressures is important for both, the prediction of the capillary sealing capacity of natural gas reservoirs or CO2 storage sites, but also for production estimates from tight gas or shale gas plays. Capillary pressure experiments, when performed on low-permeability core plugs, are difficult and time consuming. Laboratory measurements on core plugs under in-situ conditions are mostly performed using nitrogen, but also with methane and carbon dioxide. Therefore, mercury injection porosimetry (MIP) measurements are preferably used in the industry to determine an equivalent value for the capillary breakthrough pressure. These measurements have the advantage to be quick and cheap and only require cuttings or trim samples. When evaluating the database in detail we find that (1) MIP data plot well with the drainage breakthrough pressures determined on sample plugs, while the conversion of the system Hg/air to gas/brine (e.g. CH4, CO2) using interfacial and wettability data does not provide a uniform match, potentially caused by different wettability characteristics; (2) brine permeability versus capillary breakthrough pressure determined on sample plugs shows a good match and could provide a first estimate of P-c-values since permeability is easier to determine than capillary breakthrough pressures. For imbibition snap-off pressures a good correlation was found for CH4 measured on sample plugs only; (3) porosity shows a fairly good correlation with permeability for sandstone only, and with plug-derived capillary break-through pressures for sandstones, carbonates and evaporates. No such correlations exist for mudrocks; (4) air and brine-derived permeabilities show an excellent correlation and (5) from the data used we do not infer any direct correlations between specific surface area (SSA), mineralogy or organic carbon content with permeability or capillary pressure. However we were able to better predict permeabilities using a more sophisticated model that relies on a combination of these parameters. (C) 2013 Elsevier Ltd. All rights reserved.