A new method for pore structure quantification and pore network extraction from SEM images

Chenhui Wang*, Kejian Wu, Gilbert George Scott, Alfred R. Akisanya, Quan Gan, Yingfang Zhou

*Corresponding author for this work

Research output: Contribution to journalArticle

Abstract

The structure of porous media is complex yet fundamental to the understanding of transport processes including single and multi-phase fluid flow. Pore network modelling is a valuable tool for studying these processes in many diverse porous media. This paper presents a new method of discretizing the pore space into connected pore elements which capture the pore geometry and topology. This method discards the difference between pore bodies and throats, instead treats them equally as pore elements which captures the features of pore structure from the widest to the narrowest part of the pore space (pore body to pore throat), honours the topology and does not over-segment the pore space. The method is a higher level discretization of the pore space compared to other traditional methods which construct pore throats explicitly. Pore networks were extracted from three test images using the method and the computed flow properties are compared with finite element simulation in terms of absolute permeability. A reasonable agreement with finite element simulation is achieved, demonstrating the value and functionality of the new pore structure quantification method.
Original languageEnglish
Pages (from-to)82-94
Number of pages13
JournalEnergy & Fuels
Volume34
Issue number1
Early online date9 Dec 2019
DOIs
Publication statusPublished - 16 Jan 2020

    Fingerprint

Keywords

  • ALGORITHM
  • DRAINAGE
  • POROUS-MEDIA
  • PREDICTION
  • SANDSTONE
  • SCALE
  • SINGLE

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Chemical Engineering(all)
  • Fuel Technology

Cite this