Basic NMR Physics and MRIL

1
Basic NMR Physics and MRIL Tool Physics

• Outline
• Nuclear Magnetism
• Origin of the NMR signal
• Spin echoes and the CPMG pulse technique
• Relaxation times T1 and T2
• Commercial probe designs and investigation
  characteristics
• Experiment timing, nomenclature, and basic data
  flow

2
Nuclear Magnetism
spin moment charge Þ magnetic moment m

• Spin quantum no. I 0, 1/2 , 1, 3/2 , ...

• Nucleus has 2I 1 spin states

• Protons have I 1/2 Þ 2 spin states

Quantum Mech. View (Energy)
Classical View (Orientation)
Gyromagnetic Ratio (?)
high E
DE
Applied Magnetic Field, Bo
determines measurement frequency h Planks
constant I spin quantum number
low E
3
Nuclear Magnetism, cont.
Many Spins
Single Spin
z
z
Bo
M
Spins precess about Bo at frequency .....
Bo
m
y
q
x
y
x
At high temperatures, net magnetization (M)
the parallel - the anti parallel protons to the
Bo field. NMR is insensitive 1016 to 1018
protons required for measurable M
4
Pulsed NMR Log Measurement Principle
... the basic NMR experiment ...
N
1. Permanent magnet in tool polarizes hydrogen
nuclei
2. Transmit train of RF pulses and records
returning spin echo signals
signal
time
RF pulses
S
3. Wait for recovery hydrogen magnetization

• maximum signal amplitude µ fluid-filled
  porosity
• signal decay time µ pore size, fluid props,
  flow props

5
The Resonance Effect
z
Excite transitions between spin states by
irradiating at Larmor frequency
Bo
z
M
y
x
z
at equilibrium
y
x
90x pulse
y
B1
x

• rf pulse generates magnetic field B1
• B1 oriented normal to Bo
• B1 oscillates at Larmor frequency

precession in xy plane induces FID signal in coil
6
T1 Polarization
Relaxation Mechanisms for T1

• Bulk Relaxation - intrensic property of fluid

T1B f (temperature, little pressure effect
(liquids))

• Surface Relaxation - Fluid-Rock interface

T1S f (S/V ratio (pore size) , relaxivity )
7
T1 Polarization
1.00
0.95
Polarized _at_ 95 often estimated as a multiple of
3 X T1
M(t)/Mo
0
0 1 2 3
4 5 sec.
Polarization Time / T1
8
T1 and T2
ML
RF
MT
T1 characterizes the rate at which longitudinal
magnetization builds up
T2 characterizes the rate at which transverse
magnetization decays
9
A Single Spin Echo
90
180
RF field
time
signal amplitude
free-induction decay (FID) signal
spin-echo signal
time
0
t
2t
adapted, with permission, from Akkurt, 1990.
10
Carr-Purcell Gradient Field Relaxation Rate
11
Idealized CPMG Spin-Echo Train
envelope of spin-echo amplitudes µ
TE
time, t
2t
6t
8t
4t
90º pulse
180º pulse
180º pulse
180º pulse
180º pulse

• Increasing number of echoes ...
• increases signal-to-noise (SNR)
• improves resolution of long T2 components

• Shortening inter-echo spacing (TE) ...
• reduces diffusion-induced shortening of T2
• improves resolution of short T2 components

12
Data Acquisition ...
1. Record CPMG trains in phase-alternate pairs
(PAPs)
2. Stack adjacent echo trains to improve
signal-to-noise (SNR)

running average
.....
-

• alternate phase of first (p/2) pulse
• corrects for baseline offset, drift
• reduces interference from ringing
• one tool also alternates frequency

13
Relaxation TimesT1 , T2 , and T2
Pulse has two effects 1. Increases thermal
energy (spin temperature) 2. Introduces phase
coherence
at equilibrium
90 pulse
After pulse is switched off ... rapid loss of
phase coher- ence, time constant T2
recovery of longitudinal magnetization, time
constant T1
decay of transverse magnetization, time
constant T2
T2 can be measured faster and thus is more
practical for logging applications than T1
14
NMR Experiment Timing
Mo
T1 400 msec
M to Bo (longitudinal component)
0
TW
Mo
T2 250 msec
M to Bo (transverse component)
0
TE
TX
B1
RF field
0
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
time, seconds
adapted from Murphy, D.P., World Oil, April 1995
15
Data Acquisition
Tw
time
Tw wait time Te interecho time Ne Number
of echoes RA running average
time
Te
16
Effects of Pore Fluids on T1
Fluids
Solids
capillary bound water BVI
clay matrix
movable water
clay bound water
hydrocarbons
rock matrix
Variability in T1 due to Fluid Types
Invisible to NMR
Clay Bound Water
Capillary Bound
Movable Water
Polarization (ideal)
Hydrocarbons
Time in sec.
3
1
10
0.3
17
Effects of Chemistry and Texture on T1 and T2
(water filled)
Low Porosity Clean Cgr Sandstone
T1 Build-up
T2 decay
Low Porosity Shaly Fgr Sandstone
Higher Porosity Shaly Cgr Sandstone
Time, sec.
18
Rule of Thumb for T1 build-up
3 T1
95 Polarization
Must use the correct Tw (wait time) to see full
porosity
19
Type of Measurements by MRIL
f MRIL Porosity Effective porosity T2 Transve
rse Relaxation Time Differentiates capillary
bound water from free fluids important for
permeability estimate. T1 Longitudinal
Relaxation Time Identifies hydrocarbon fluids in
the non-wetting phase. D Fluid
Diffusivity Differentiates between gas phase and
liquid phase.
T1
20
Resonant Frequency
Field Strength diminishes with increased distance
from magnet. Proton spin rates are
proportional to Magnetic Field Strength. F g
B0 For Hydrogen g 4258 Hz / Gauss F
Frequency (Hz) B0 Field Strength
(Gauss) Applying a radio frequency signal equal
in frequency to the proton spin rate causes the
protons to resonate.
N
S
field strength
distance
21
MRIL Gradient Magnetic Field
Sensitive Volume
Formation
Borehole Wall
Magnet
Antenna
mud
Frequency Bandwidth
B0(r)
B0(r)
6
16
ã NUMAR Corp., 1996
22
Measurement Principle
Damaged Zone
Sensitive Volume
typ. 12 cm
Mud Cake
23
MRIL in Wellbore
MRIL Probe
Borehole
Sensitive Volume Cylinders (each 1 mm thick at 1
mm spacing)
24
16
ã NUMAR Corp., 1995
24
Using Multiple Frequencies
f1
f2
f3

• Improves Logging Speed
• Increases Signal to Noise

25
A New Multiband Generation of NMR Logging Tools
MRIL-Prime

• Features
• 9 discrete measurement volumes
• Accelerated polarization - 12 sec in 6 sec
• more robust electronics and new sonde

• Benefits
• Single-pass, fast, total porosity T2
  distribution
• 24 fpm logging speed or
• Data quality of C tool x 3
• Multi-parameter acquisition (T1 curve, etc.)
• Familiar outputs and interpretation

26
MRIL-Prime Shells
Borehole
9 Sensitive Volume Cylinders (each 1 mm thick at
1 mm spacing)
MRIL Probe
760kHz
580kHz
1
24
16 _at_ 250F
27
MRIL Diameter of Investigation6 Probe
20
50 F.
19
100 F.
50 F.
18
150 F.
100 F.
200 F.
17
150 F.
Diameter (Inches)
250 F.
16
200 F.
300 F.
15
250 F.
14
300 F.
13
12
500
550
600
650
700
750
800
850
900
950
1000
Frequency (kHz)
28
MRIL Calibration Tank
Side View
End View
Formation Chamber
Faraday Cage
Borehole Chamber
MRIL Tool
Sensitive Volumes
ã NUMAR Corp., 1996
29
T2 Relaxation
30
Idealized Echo Train
Amplitude
T2R
Time
ã NUMAR Corp., 1995
31
Hydrogen in Matrix Clay Bound Water
Hydrogen in Clay Bound Water
Hydrogen in Lithology
ã NUMAR Corp., 1995
32
NMR Relaxation in a Water-Filled Pore

• Water in rocks relaxes 101 to 104 faster than
  bulk fluid value (T1 3.5 sec at 25C)
• Enhanced relaxation on pore surfaces due to
  additional relaxation mechanisms ...
• - paramagnetic centers (T1 and T2 )
• - hindered molecular motion (T1 and T2 )
• - fluid/solid bulk magnetic susceptibility
  contrast (T2 only)
• Overall relaxation rate on the surface
  characterized by relaxivity r
• Diffusion mixes slow-relaxing H spins in V with
  fast relaxers on S

slow-relaxing spins in pore volume, V
1H
fast-relaxing spins in thin film on pore surface
S, ... ... film thickness l 10 Å
Diffusion and surface relaxation in an isolated,
water-filled pore
33
Pore Size and T2 (Water)
T2
time
T2
time
time
T2
T2
time
T2
time
34
Coarse Grain Response
MPHI 36, MFFI 30, MBVI 6, MPERM 4200 md
40
35
30
25
Porosity
20
15
10
5
0
0
50
100
150
200
250
Time (ms)
ã NUMAR Corp., 1995
35
Fine Grain Response
MPHI 36, MFFI 6, MBVI 30, MPERM 6.7 md
40
35
30
25
Porosity
20
15
10
5
0
0
50
100
150
200
250
-5
Time (ms)
ã NUMAR Corp., 1995
36
MAP Processing
ã NUMAR Corp., 1995
37
Data Processing
MAP Inversion Processing
Spin-echo data
T2 Spectrum
NMR porosity
Free Fluid - FFI
2.00
16
multiexponential fit to spin-echo amplitudes
14
1.50
12
Clay Bound Water - CBW
Capillary Bound Fluid - BVI
large-pore (mobile fluid) signal
10
Incremental Porosity pu
Cumulative Porosity pu
Invisible Region
1.00
8
6
0.50
4
2
0.00
0
time
0.1
1
10
100
1000
10000
small-pore (irreducible fluid) signal
T2 msec
clay-bound water
Water-saturated rock rT2 V/S Invisible
Region is a function of signal to noise and echo
spacing
38
NMR - Porosity Model
Neutron ?
Density ?
Resistivity ?Sw
capillary bound water BVI
clay matrix
movable water
clay bound water
rock matrix
hydrocarbons
NMR BVI
NMR FFI
Integration of MR Log and Resistivity Log
Interpretation
MR porosity (effective)
MR porosity (total short TE)
Producible hydrocarbon
nonmovable water
will produce some water
39
Spin Echo Attenuation by Diffusion in a Gradient
Magnetic Field

• only stationary spins are completely rephased by
  p pulses in a CPMG expt
• spins diffusing in a gradient magnetic field
  undergo unrecoverable dephasing... Þ echo
  attenuation Þ transverse relaxation mechanism

..... two sources of magnetic field gradients
.....
B0
B0d
B02d
B03d
cfluid
B0
Grain
Rock Grain
Pore
cgrain
Rock Grain
2r
Rock Grain

• Applied ... MRIL uses strong gradient magnetic
  field to perform slice selection ...
• known, well-defined gradient gives predictable T2
  shifts that depend only on diffusion

• Natural ... grain scale gradients arising due to
  magnetic susceptibility c contrast between
  minerals and pore fluids ...
• randomly varying at grain scale
• pore size and mineralogy dependent

40
Diffusion and T2D

• only effective for T2 relaxation (not for T1)

12
T2D

D . ( G . ? . Te )2
D Diffusion Coefficient of Fluid (cm2/sec)
depends on Temp. (K) Viscosity G Magnetic
Field Gradient (Gauss/cm) depends on
Tool Freq. Temp. ? Gyromagnetic Ratio
(Hz/Gauss) 4258 for Hydrogen Te
Inter-Echo Spacing (sec.)
41
Effect of Oil on T2 Distribution
Incremental Porosity
Sw 56.9
Sw 65.4
Sw 84.3
Sw 100
T2 Distribution
ã NUMAR Corp., 1995
42
Partially Saturated Water Wet Rock Behavior
Sw 100
T2
43
Viscosity Diffusion vs T2_at_ 80 F.
ã NUMAR Corp., 1995
44
T2 Distributions Incomplete Recovery
The long end of T2 distributions - indicative of
hydrocarbons - can be associated with very long
T1 times. If standard recovery times TR are
used, that portion of a T2 spectrum will be
depressed by 50 or more.
correct T2 spectrum
incomplete recovery
P (T2)
T2
ã NUMAR Corp., 1995
45
Permeability Chart E-4
0.5
E - 4
0.4
0.3
Porosity
(f x Swirr) increases
0.2
0.1
k (md)
0
0
0.2
0.4
0.6
0.8
1
Sw
irr
ã NUMAR Corp., 1996
46
Permeability from Porosity and Water Saturation
40
35
30
k, Permeability (md)
25
Phi x Swirr
20
Porosity
15
10
5
0
0
10
20
30
40
50
60
70
80
90
100
Sw
ã NUMAR Corp., 1996
irreducible
47
T2 and Pore Size Distributions
Coarse
Clay
Silt
Fine
Sand Domain r2 5mm/s
Clay Domain r2 1 mm/s
48
MRIL Technology

• NMR As Formation Evaluation Tool
• Measurement Model
• Formation Evaluation Parameters
• Laboratory Modeling
• The MRIL Logging Tool
• Measurement Principle
• Examples
• MRIL vs. CMR
• The MRIL As MWD Device
• Technological Basis
• Tool Configuration
• Wireline Compatibility And Replacement

49
MRIL Inversion Processing
Log
50
Viscosity vs. T2 for oils
10000
1000
T2 (ms)
100
100 F
300 F
T2cutoff
T2 Oil
adopted from Looyestijn et al, 1996
10
0
10
20
30
40
50
Viscosity (cp)
51
Viscosity vs. D0 for oils
Diffusion coefficient (x10e -5 cm2/s)
100 F
300 F
adopted from Looyestijn et al, 1996
0
10
20
30
40
50
Viscosity (cp)
52
MRIL Diameter of Investigation4.5 Probe
14
50 F.
100 F.
13
50 F.
150 F.
100 F.
12
200 F.
150 F.
250 F.
Diameter (Inches)
11
200 F.
300 F.
250 F.
10
300 F.
9
8
500
550
600
650
700
750
800
850
900
950
1000
Frequency (kHz)
53
MRIL Response
54
MRI LogProcessing Interpretation
Conductive Fluids
Matrix Dry Clay
ClayBoundWater
Capillary BoundWater
Moveable Free Water
Oil
Gas
MPHI
MCBW
MFFI
MBVI
MRIL
PHIT
MRIL Porosity
?MRIL ? . HI . ( 1 etw/T1)
?MRIL ? . Sxo . HIw . ( 1 etw/T1w) ? . ( 1
- Sxo) . Hlh . ( 1 etw/T1h)
55
MRI LogProcessing Interpretation
Conductive Fluids
Matrix Dry Clay
ClayBoundWater
Capillary BoundWater
Moveable Free Water
Oil
Gas
MPHI
MCBW
MFFI
MBVI
MRIL
PHIT
MRIL Permeability
56
Direct Hydrocarbon TypingDifferential Spectrum
Method
Brine Gas Oil
Long Recovery Time (TR)
Short Recovery Time (TR)
Difference
1 10 100
1,000 10,000
T2 Time (ms)
ã NUMAR Corp., 1995
57
MRI LogProcessing Interpretation
T.D.A.MRIL Time Domain Analysis
Correcting MRIL Porosity for T1 effects
Unique Liquid Phase
Porosity
?TwLu
Long Tw Short Tw
?TwSu
1
10
10,000
100
1,000
T2 Time (ms)
58
MRI LogProcessing Interpretation
TDA_COMP
The Porosity Measured by MRIL is Subject to
Hydrogen Index HI and Polarization T1 effects.
? Tw ? . HI . ( 1 - e -Tw/T1)
? True Hydrocarbon Porosity ?Tw Measured
Hydroc. Porosity Tw Wait Time HI Hydrogen
Index ?TwLu Porosity from TwL of
unique Fluid phase ?TwSu Porosity from TwS
of unique Fluid phase TwL Long Wait
Time TwS Short Wait Time
T1 estimation is based on the Ratio r ? TwLu
/ ? TwSu
? . HI . ( 1 - e -TwL/T1 )
r
? . HI . ( 1 - e -TwS/T1 )
1 - e -TwL/T1

1 - e -TwS/T1
59
MRI LogProcessing Interpretation
TDA_COMP
?gas is calculated by correcting for HIgas T1gas
?g Gas Porosity ?g Apparent Gas Porosity
seen by echo difference HIg Hydrogen Index
of Gas ?o Oil Porosity ?o Apparent Oil
Porosity seen by echo difference HIo Hydrogen
Index of Oil T1h T1 of either Gas or Oil
?oil is calculated by correcting for HIoil T1oil
60
MRI LogProcessing Interpretation
TDA_COMP
Fully Polarized Liquid Porosity from Long Tw
PhiFPL MPHIA -

?g . HIg . ( 1 - e -TwL / T1g ) ?o . HIo . ( 1
- e -TwL / T1o ) ? w . HIw . ( 1 - e -TwL / T1w
)

Corrected Porosity for T1 and HI
TDAMPhi PhiFPL ?g ?o ? w
61
MRI LogProcessing Interpretation
T.D.A. MRIL Time Domain Analysis
Long Tw
Short Tw
-
Porosity
Porosity
1
10
100
1,000
10,000
1
10
100
1,000
10,000
T2 Time (ms)
T2 Time (ms)
Difference
Water Gas Oil
?o
?g
Porosity
1
10
100
1,000
10,000
T2 Time (ms)
62
Direct Hydrocarbon TypingShifted Spectrum Method
Short Echo Spacing
Brine Gas
Porosity
Long Echo Spacing
Porosity
1 10 100
1,000 10,000
T2 Time (ms)
ã NUMAR Corp., 1995
63
Direct Hydrocarbon TypingShifted Spectrum Method
Short Echo Spacing
Brine Oil
Porosity
Long Echo Spacing
Porosity
1 10 100
1,000 10,000
T2 Time (ms)
ã NUMAR Corp., 1995
64
MRI LogProcessing Interpretation
DIFANMRIL Diffusion Analysis

• MRIL data acquired using Dual Te
• Solves for Oil in the slice under investigation
• Does not solve for VERY Heavy Oil

65
MRI LogProcessing Interpretation
DIFAN MRIL Diffusion Analysis
Long Te Short Te
Water signal
Oil signal
FFI
BVI
Porosity
Gas
1
10
10,000
100
1,000
T2 Time (ms)
Diffusion
Water

• Shift in Spectrum Data is due to Diffusion
• Calculating the Geometric Mean of the Short and
  Long Spectra as a functin of the area under the
  curve, helps identify the amount of shift due to
  Diffusion

Viscosity
Oil
66
MRI LogProcessing Interpretation
DIFAN
Based on the Thermal Diffusion properties of
Fluids in the pore space
D Diffusivity of Fm. Fluid DW Diffusivity of
Water at Fm. Temp. Press. 12.5 at surface
Temp. press.
T2irr Lower T2 boundary of Free Fluid
67
MRILProcessing Interpretation
DIFAN
Based on the Thermal Diffusion properties of
Fluids in the pore space
Swapp L2 / L1
1 / T2irr
L2
100 Water Saturation
1 / T2int
Sw
((MPHIA - BVIA) . Swapp ) BVIA
L1
X
50 Water Saturation
MPHIA
0.0
1.0
RDDW
1 / T2Hy
DIFSW Sw DIFBVW MPHIA . Sw
T2irr Lower T2 boundary of Free Fluid
68
MRI LogProcessing Interpretation
DIFAN
Based on the Thermal Diffusion properties of
Fluids in the pore space
1
1
1
1



T2
T2Bulk
T2Surf
T2Diff
? Relaxation Constant S Surface V Volume
D Diffusion Coefficient of Fluid (cm2/sec)
depends on Temp. (K) Viscosity G Magnetic
Field Gradient (Gauss/cm) depends on
Tool Freq. Temp. ? Gyromagnetic Ratio
(Hz/Gauss) 4258 for Hydrogen Te
Inter Echo Spacing time (sec.)