Petrophysical Data Integration for Improved Reservoir Description
Rahul Dastidar, Carl H. Sondergeld, Chandra S. Rai, 2006. "Petrophysical Data Integration for Improved Reservoir Description", Reservoir Characterization: Integrating Technology and Business Practices, Roger M. Slatt, Norman c. Rosen, Michael Bowman, John Castagna, Timothy Good, Robert Loucks, Rebecca Latimer, Mark Scheihing, Hu Smith
Download citation file:
Integrated reservoir characterization relies increasingly on vastly improved log-based results from new technologies such as Nuclear Magnetic Resonance, NMR. Our experimental study is designed to extract more petrophysical information from NMR for reservoir characterization.
We compared the empirical permeability estimation based on NMR with direct measurements; evaluated the use of NMR observations in providing capillary pressure estimates; and the use of NMR to classify rock types. Using a 2MHz NMR spectrometer, we analyzed 90 clastic cores from five different wells. Cores were measured at 100% brine saturated and at irreducible saturation achieved through centrifuging the core plugs at 5800 rpm. High pressure mercury injection was performed on parallel samples from the same plug. The porosity of the cores studied ranged from 4% to 23% while measured permeabilities ranged from 0.01 md to 900 md.
The measured T2 cutoffs (i.e., the boundary between free and bound water) ranged from 6 ms to 100 ms, which represents significant departures from the typically assumed 33 ms cutoff for clastics. Mineralogy appears to have an influence on the T2 cutoff value. In general the permeability estimation based on the weighted geometric mean of the T2 time is better than the model based on the ratio of free fluid index to bound volume index. Additionally, mapping NMR and mercury measurements provided estimates of surface relaxivities, which ranged from 16 to 50 μm/sec.
Measurements of surface relaxivity allow the empirical mapping of NMR data to capillary pressure data. The mismatch between the cumulative NMR and mercury data at lower and higher T2 times reflects differences in how the pore space is accessed between the NMR and the Hg measurements. The applicability of NMR T2 distribution for rock typing is discussed. It is observed that NMR is more sensitive to subtle pore characteristics (dimension, shape, and composition, etc.) as compared to Levertt J-function derived from capillary pressure and may provide an alternative method for rock typing.