Nuclear Magnetic Resonance Permeability Estimation in FineGrained Sediments

1
Nuclear Magnetic Resonance Permeability
Estimation in Fine-Grained Sediments

• Hugh Daigle
• Brandon Dugan
• Department of Earth Science
• Rice University

2
Motivation
-Flow and pressure in sedimentary
basins -Geohazards -Gas hydrates -Typically
70 of sediments in basins are clays -Even more
in the Gulf of Mexico -Measuring clay
permeability is time-consuming -NMR permeability
relationship (SDR Equation) -Can we extend
the SDR equation for use in clay-rich sediments?
3
Gulf of Mexico Gas Hydrate JIP -Keathley Canyon
151-2 logged with LWD -Keathley Canyon 151-3 cored
Dugan (in press)
500 m
TWTT (ms)
Conte Bloys (2006)
392 m
4
Porosity ()
0 50 67 75
Possible hydrate zone
5
NMR transverse relaxation time (T2)
-T2B is on the order of seconds for
seawater. -T2D is on the order of seconds. -T2S
for a typical sandstone is on the order of
milliseconds.
6
Pore size distributions, T2LM, and the SDR
equation
A is typically equal to 3.9x10-15 m2/ms2 (4.0
mD/ms2) in sandstone reservoirs (Kenyon et al.,
1995)
7
Laboratory Experiments
Controls Strain Rate 0.35-0.5 /hr Cell
Pressure 0.39 MPa Zero Lateral
Strain Measurements Sample Height Vertical
Stress Pore Pressure
8
(No Transcript)
9
-Linear (elasto-plastic) portion of consolidation
curve isolated, extrapolated back to in situ void
ratio
Permeability (mD)
0.001
0.01
e0 0.73 k0 1.07x10-17 m2 0.0108 mD
10
Permeability (mD)
0.001 0.01
0.1
1.0
10
11
Log analysis procedure
-Determine T2 from NMR data -Determine porosity
from density data -Determine gamma ray value
from log data -Calculate A using
consolidation-derived permeability -Compare A
to gamma ray value (a proxy for lithology)
12
0.2
A 6.08x10-17 m2/ms2 0.0616 mD/ms2
millidarcies
0.1
0
68 confidence interval is 2.23x10-17 m2/ms2
13
0.2
mD/msec2
0.1
0
68 confidence interval is 1.97x10-17 m2/ms2
14
How do the two models compare?
Constant A
Variable A
Permeability (mD)
Permeability (mD)
0.001 0.1
10
0.001 0.1
10
Goodness-of-fit for each model
s 1.63x10-17 for variable A s 2.32x10-17 for
constant A
15
Conclusions
-A variable coefficient A gives a more accurate
estimation of permeability than a constant
coefficient in the SDR equation -Using A4.0
mD/ms2 will overestimate permeability by 2
orders of magnitude -A varies with lithology,
and can be calculated from the gamma ray log -A
exhibits lower values with higher gamma ray
values, reflecting a change in the pore
space-permeability relationship with increasing
clay content
Areas of Future Research
-NMR measurements on permeability
specimens -Specific surface area and
porosimetry -Synthetic samples -Measurement and
quantification of permeability anisotropy