Specific surface area versus porosity from digital images

By computing the total porosity φ and specific surface area S for a number of segmented digital volumes of natural sandstones, carbonates, and granular samples, as well as their subvolumes, we observe fairly tight trends between those two variables. The emerging picture is different for low-to-mediu...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of petroleum science & engineering 2021-01, Vol.196, p.107773, Article 107773
Hauptverfasser: Hussaini, Syed Rizwanullah, Dvorkin, Jack
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:By computing the total porosity φ and specific surface area S for a number of segmented digital volumes of natural sandstones, carbonates, and granular samples, as well as their subvolumes, we observe fairly tight trends between those two variables. The emerging picture is different for low-to-medium porosity rocks from that for high-porosity granular samples. In the former, S increases with increasing φ, while the behavior is opposite in the particulates. We explain these trends by invoking simple theoretical derivations where the consolidated low-to-medium porosity samples are modeled as solids with inclusions, while the particulates are represented by packs of grains. While in the former S is linearly proportional to φ, it is linearly proportional to 1−φ in the latter. The digital data are fairly accurately matched by the respective theoretical curves with the pore- and grain-size statistics extracted from the digital volumes. This fact arguably means that the trends obtained here from microscopic digital volumes are valid at a much coarser core and reservoir scale. •Specific surface area versus porosity transforms depend on rock type, consolidated versus granular.•These transforms are obtained from digital pore-scale images on a large number of natural rock samples.•Simple theoretical equations describe the phenomena observed and match the data.
ISSN:0920-4105
1873-4715
DOI:10.1016/j.petrol.2020.107773