Simultaneous Gaussian and exponential inversion for improved analysis of shales by NMR relaxometry.

Autor: Washburn KE; Ingrain, Inc., 3733 Westheimer Road, Houston, TX 77027, United States. Electronic address: washburn@ingrainrocks.com., Anderssen E; Laboratory of Molecular Medical Research, Institute of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway; Weatherford International, 6550 West Sam Houston Tollway, Houston, TX, United States., Vogt SJ; Chemical and Biological Engineering, Montana State University, Bozeman MT 59717, United States; School of Mechanical and Chemical Engineering, University of Western Australia, Crawley, WA, Australia., Seymour JD; Chemical and Biological Engineering, Montana State University, Bozeman MT 59717, United States., Birdwell JE; U.S. Geological Survey, Denver Federal Center, Box 25046 MS 977, Denver, CO 80225, United States., Kirkland CM; Chemical and Biological Engineering, Montana State University, Bozeman MT 59717, United States., Codd SL; Mechanical and Industrial Engineering, Montana State University, Bozeman, MT 59717, United States.
Jazyk: angličtina
Zdroj: Journal of magnetic resonance (San Diego, Calif. : 1997) [J Magn Reson] 2015 Jan; Vol. 250, pp. 7-16. Date of Electronic Publication: 2014 Nov 10.
DOI: 10.1016/j.jmr.2014.10.015
Abstrakt: Nuclear magnetic resonance (NMR) relaxometry is commonly used to provide lithology-independent porosity and pore-size estimates for petroleum resource evaluation based on fluid-phase signals. However in shales, substantial hydrogen content is associated with solid and fluid signals and both may be detected. Depending on the motional regime, the signal from the solids may be best described using either exponential or Gaussian decay functions. When the inverse Laplace transform, the standard method for analysis of NMR relaxometry results, is applied to data containing Gaussian decays, this can lead to physically unrealistic responses such as signal or porosity overcall and relaxation times that are too short to be determined using the applied instrument settings. We apply a new simultaneous Gaussian-Exponential (SGE) inversion method to simulated data and measured results obtained on a variety of oil shale samples. The SGE inversion produces more physically realistic results than the inverse Laplace transform and displays more consistent relaxation behavior at high magnetic field strengths. Residuals for the SGE inversion are consistently lower than for the inverse Laplace method and signal overcall at short T2 times is mitigated. Beyond geological samples, the method can also be applied in other fields where the sample relaxation consists of both Gaussian and exponential decays, for example in material, medical and food sciences.
(Copyright © 2014 Elsevier Inc. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: