Release Time:2019-03-12  Hits:

Indexed by: Journal Article

Date of Publication: 2017-08-01

Journal: HYDROGEOLOGY JOURNAL

Included Journals: SCIE

Volume: 25

Issue: 5

Page Number: 1465-1476

ISSN: 1431-2174

Key Words: Laboratory measurements; Pore geometry; Unconsolidatedsediments; Petrophysical property; Modeling

Abstract: Permeability variation in reservoir rocks results from the combined effects of various factors, and makes porosity-permeability (phi-k) relationships more complex, or, in some cases, non-existent. In this work, the phi-k relationship of macroscopically homogeneous glass-bead packs is deduced based on magnetic resonance imaging (MRI) measurement and Kozeny-Carman (K-C) model analysis; these are used as a frame of reference to study permeability of reservoir rocks. The results indicate: (1) most of the commonly used simplified K-C models (e.g. the simplified traditional (omitting specific surface area), high-order, threshold, and fractal models) are suitable for estimating permeability of glass-bead packs. The simplified traditional model does not present obvious dependence on rock samples. Whether for the glass-bead packs or clean natural sandstones, the sample coefficients almost remain invariant. Comparably, the high-order, the fractal, and the threshold models are strongly sample-specific and cannot be extrapolated from the glass-bead packs to natural sandstones; (2) the I-k relationships of quartz sands and silty sandstones resemble those of the glass-bead packs, but they significantly deviate from the K-C models at low porosities due to small pore entry radius; (3) a small amount of intergranular cements (< 10%v) does not affect the general variation trend of permeability with porosity but can potentially increase predictive errors of the K-C models, whereas in the case of more cements, the phi-k relationships of sandstones become uncertain and cannot be described by any of these K-C models.