gas hydrates in coarse-grained reservoirs interpreted from ... · 2 gas hydrates in coarse-grained...
TRANSCRIPT
1
GSA Data Repository 2018186 1
Gas hydrates in coarse-grained reservoirs interpreted from velocity pull 2
up: Mississippi Fan, Gulf of Mexico 3
Andrew S. Madof 4
5
METHODS6
Petrophysical calculations 7
Well-log interpretation was conducted on the GC 955-H well using commercial 8
petrophysical analysis software (Geolog® by Paradigm®). Porosity was calculated from density 9
via the following equation: 10
where is porosity, is the density of the matrix (2.65 g/cc – sandstone), is the 11
bulk density (measured from the density-logging tool), and is the density of the fluid (1.00 12
g/cc – water). Archie’s equation was subsequently used to solve for gas hydrate saturation via the 13
following relationships: 14
where Sw is water saturation, a is the Tortuosity coefficient (equal to 1), Rw is the 15
electrical resistivity of the formation water, is porosity (calculated from density), m is a 16
cementation exponent (equal to 2), Rt is the electrical resistivity of the formation (measured from 17
2
deep resistivity), and Sgh is gas hydrate saturation. Results of the calculations are displayed in 18
Fig. DR2. 19
Seismic interpretation 20
Reflections were mapped on full-stack zero-phase seismic data (two-way travel time) 21
using commercial seismic-interpretation software (SeisEarth® by Paradigm®). High-amplitude 22
peak reflections were used to trace the top of channelized systems P1-P3, as well as Reflections 23
1,2 and 4 (R1, R2, and R4). High-amplitude trough reflections were selected to map the base of 24
the P1-P3 systems, as well as Reflection 3 (R3 – see Fig. DR8B).25
Root mean squared (RMS) amplitude and spectral decomposition were calculated on the 26
3D seismic volume (Figs. 2C and DR7A). RMS was calculated on the interval from the base of 27
the gas hydrate stability zone to the sea floor using SeisEarth® by Paradigm®. Spectral 28
decomposition was calculated over the entire dataset and calibrated to the upper unit of the P3 29
system using commercial seismic-interpretation software (GeoTeric® by Foster Findlay 30
Associates). 31
Wavelength filtering 32
To isolate VPU, R1-R4 (Fig. DR8B) were filtered and stacked using commercial 33
geophysical-interpretation software (OasisMontaj® by GeoSoft®). Individual surfaces were 34
filtered via a high-pass Butterworth filter, with a wavelength of 75 km (46.6 mi) and a filter 35
order of 5. The average of the four surfaces is displayed as a single surface in Fig. 2D. 36
37
3
REFRENCES CITED 38
Dai, J., Xu, H., Snyder, F., and Dutta, N., 2004, Detection and estimation of gas hydrates using 39
rock physics and seismic inversion: Examples from the northern deepwater Gulf of 40
Mexico: The Leading Edge, v. 23, p. 60-66. 41
Guerin, G., Cook, A., Mrozewski, S., Collett, T., and Boswell, R., 2009, Gulf of Mexico gas 42
hydrate joint industry project leg II: Green Canyon 955 LWD operations and results: 43
https://www.netl.doe.gov/File%20Library/Research/Oil-44
Gas/methane%20hydrates/GC955LWDOps.pdf (April 2018). 45
Jones, I.F., 2012, Tutorial: Incorporating near-surface velocity anomalies in pre-stack depth 46
migration models: First Break, v. 30, p. 47-58. 47
Lee, M.W., Hutchinson, D.R., Collett, T.S., and Dillon, W.P., 1996, Seismic velocities for 48
hydrate-bearing sediments using weighted equation: Journal of Geophysical Research, v., 49
101, p. 20347-20358. 50
Kvenvolden, K.A., and McMenamin, M.A., 1980, Hydrates of natural gas: A review of their 51
geological occurrences: United States Geological Survey Circular 825, 11 p. 52
Madof, A.S., 2016a, Methods and systems for identifying a clathrate deposit, United States 53
Patent application 15/218910, filed July 25, 2016. 54
Madof, A.S., 2016b, Methods and systems for quantifying a clathrate deposit, United States 55
Patent application 15/218920, filed July 25, 2016. 56
Shaw, J.H., Connors, C., and Suppe, J., eds., 2005, Seismic interpretation of contractional fault-57
related folds: An AAPG Seismic Atlas: American Association of Petroleum Geologists 58
Studies in Geology 53, 157 p. 59
= Base of GHSZ
Figure DR1. Gas hydrate stability zone (GHSZ) in the marine setting, modeled with 100% methane (modified from Kvenvolden and McMenamin, 1980). A: Model is constructed using a sea-floor depth of 1,500 m (4°C at the sea floor) and geotherms of 60°C/km (B) and 20°C/km (C). The 60°C/km geotherm (B) yields a 210-m thick GHSZ, compared to the 20°C/km geotherm (i.e., 750-m thick GHSZ). The 20°C/km model (C) is used for the study area, and is displayed in Figs. 1B, 2B, DR8B, and DR10B.
= Stable free gas= Stable gas hydrates
1,500 m
1,710 m
1,500 m 1,500 m
2,250 m
1,500 m
20°C/kmCTemperature vs depthA 60°C/kmB0
2,000
2,500
3,000
3,500
1,000
500
1,500
4,000
4,500
5,000
0 10 20 30 40
20°C/km geotherm
20°C/km geotherm
60°C/km geotherm60°C/km geotherm
Sea floor
Dep
th (m
)
Hyd
roth
erm
Hyd
roth
erm
Methane stability
Methane stability
= Thickness of GHSZ (60°C/km geotherm) = Thickness of GHSZ (20°C/km geotherm)
Figure DR2. Petrophysical response of a gas hydrate-bearing sandy reservoir from the GC 955-H well in the northern Gulf of Mexico (modified from Guerin et al., 2009 - see Fig. 1A for location). Gas hydrate-bearing sediments are identified primarily based on elevated values of resistivity (>5 ) and P-wave velocity (>2,500 m/s) that coincide with depressed gamma ray measurements (<50 API). Although poor-hole conditions inevitably lead to erroneous measurements, velocities of non-gas hydrate-bearing sediments (i.e., background) are taken as 1,700 m/s (sand) and 2,100 m/s (mud); these values are used as input for Models 1-4 (see Fig. 3). See Methods section for calculations regarding porosity and saturation.
2,500
2,550
2,400
2,450
Density (g/cc) Porosity (v/v)Gamma Ray (API) Velocity (m/s) Saturation (v/v)Caliper (cm)
1.5 2.0 0 0.5 1.00 50 100 0 105 15 1,500 2,500 0 0.5 1.020 24 28
**
**
=Gas hydrate- filled sand
* = Poor hole conditions
Dep
th (m
)Petrophysical response of gas hydrates
1,7001,700 2,1002,100SandSand MudMud
Figure DR3. Crossplot of P-wave velocity (vgh) versus saturation (Sgh) for the 2,450-2,500-m depth interval in the GC 955-H well (see Fig. DR2). The data show a linear relationship, with gas hydrate-bearing sand corresponding to values of increased velocity (>2,500 m/s) and decreased porosity (<0.5). The relationship is used to calibrate saturation to velocity in Fig. 3.
4000
3800
3600
3400
3200
3000
2800
26002400
2200
2000
1800
1600
0.20.10.0 0.40.3 0.60.5 0.80.7 1.00.9Saturation (v/v)
Porosity0.7 =0.6 =0.5 =0.4 =0.3 =
Velo
city
(m/s
)
Water-filled sandWater-filled sand
Velocity vs saturation
Gas hydrate-bearing sandGas hydrate-bearing sand
r2 = 0.841r2 = 0.841gh = 2582.95 Sgh + 1356.29gh = 2582.95 Sgh + 1356.29
Fig.
DR
6Figure DR4. Diagram illustrating velocity pull up (VPU) in two-way travel time (TWTT - modified from Madof, 2016a and 2016b). A: Gas hydrate-bearing coarse-grained turbidites (blue - high velocity) are encased in background sediments (gray - low velocity) and create travel-time deficits resulting in convex reflections at depth (orange). Major erosional surfaces defining slope valleys are shown as solid red lines; time-equivalent non-erosional components are illustrated as red dashed lines. Solid black lines associated with Reflections 1-4 show surfaces in TWTT; dash-dot lines under VPU show the surfaces in depth (i.e., velocity effect removed). B: Two idealized well logs showing gamma ray (GR) and P-wave velocity (v) trends. Well x encounters gas hydrate-bearing sediments and results in values of decreased gamma ray and increased velocity. Well y is located lateral to gas hydrate-bearing system, and shows no change in petrophysical character. C: Map-view image of two slope valleys and underlying VPU. D: Perspective view of the gas hydrate system; thicker gas hydrate-bearing sediments correspond to higher values of VPU.
y
Slope valley
VPU
xD
Map viewCx y
Slope valley
VPU
Perspective viewDx y
TWTT Base GHSZBase GHSZ
Sea floor
Coarse-grained turbidites
Slope valley
Reflection 2
Reflection 3
Reflection 4
Reflection 1VPU
Section viewAx
GR v
y
GR v
WellsB
Creating a VPU surface from seismic data
Laterally filter each time-structure surface to isolate short-wavelength
features (VPU)
Pick laterally continuous reflections at depth (TWTT)
Fig. DR8B (see picked red, orange, yellow, and green surfaces)
Figure DR5. Flowchart showing steps necessary to create VPU surface from seismic data (TWTT). Short-wavelength features (VPU) on time-structure surfaces are isolated via a high-pass filter using geophysical software (e.g., OasisMontaj by GeoSoft® - see Methods section). Magnitude is measured directly from VPU surface (see Fig. 2D) and used as input to solve for P-wave velocity of gas hydrate-bearing sediments (see Fig. DR6).
Stack multipe surfaces to remove noise and display as single VPU surface (TWTT)
Fig. 2D (see map view of VPU surface)
Use VPU surface as input to constrain P-wave velocity of gas
hydrate-bearing sedimentsFig. 3 and Fig. DR6
Step 2: Express gh in terms of bStep 1: Estimate Th(t)
Step 4: Solve for b
Note that in order to solve for gh (Eq. 1) based on b (Eq. 4), three quantities are needed:
(i) b - estimated from checkshot, velocity function, or analog
(ii) Th(t) - measured from seismic (TWTT) and multiplied by net:gross
(iii) VPU(t) - calculated from stacked and filtered time-structure surfaces
(Eq. 4)b =VPU(t)Th(t) + 1
Calculating P-wave velocity of gas hydrate-bearing sediments using VPU
Where:Th(t) = Net thickness of gas
hydrate-bearing sediments (TWTT)*
* = Calculated by multiplying gross thickness (from seismic) by net:gross (from analog)
*** =
(Eq. 1)gh = b
**** = Estimated from checkshot, velocity function, or analog
** = Unknown
Although there exists a complex three-dimensional relationship between vb and vgh, the one-dimensional model presented here provides a simplification by treating values as linear
Where:
b = P-wave velocity of background sediments****
b = Scalar relating velocities of background sediments to gas hydrate-bearing sediments***
gh = P-wave velocity of gas hydrate-bearing sediments**
Step 3: Relate Th(t) to VPU(t)
VPU(t) = Velocity pull up (TWTT )† Where:
(Eq. 2)VPU(t) = gh
b( )- 1 Th(t)
(Eq. 3)VPU(t) = b
b( )- 1 Th(t)
Substitute Eq. 1 into Eq. 2:
Gas hydrate-bearing sediments
not present
Gas hydrate-bearing sediments
present
Th(t)
VPU(t) = 0 when gh = b
Th(t)
VPU(t)
VPU(t) = Th(t) when gh = 2 b
Sea floor
Base GHSZ
Reflection
TWTT TWTT † = To calculate a VPU(t) surface, see Fig. DR5
†† = For detailed graphs, see Fig. DR9
Step 6: Substitute Sgh for vgh
To relate gh to Sgh (i.e., saturation of gas hydrate-bearing sediments), use empirical relationships†††, rock-physics models, or theoretical correlations:
Low-velocity ( b) model
Step 5: Correlate VPU(t) to Th(t)††
High-velocity ( b) model
TWTT
gh b
Th(t)
b = 2
Th(t)
gh b
b = 1
Th(t)
(Eq. 1)
NetNet
TWTT
Th(t)
Seismic model
+-
Top
TWTT
Base
= Gas hydrate-bearing sediments
= Background sediments
Geologic model
Gross
b
200
400
600
01.21.0 1.4 1.6 1.8 2.0 2.2
1,800 2,200 2,600 3,000 3,400 3,800vgh (m/s)
Th(t)Thin Thick
200
400
600
01.21.0 1.4 1.6 1.8 2.0 2.2
1,800 2,200 2,600 3,000 3,400 3,800
b (vgh / vb)
VPU
(t)VP
U(t)
b
†††From Fig. DR3 (GC 955-H well): gh = 2582.95 Sgh + 1356.29
†††From Fig. DR3 (GC 955-H well): gh = 2582.95 Sgh + 1356.29
Degree of lithificationLow High
2000
2500
1500
3000
3500
0.20.0 0.4 0.6 0.8 1.0 Sgh (v/v)
gh (m
/s)
From Lee et al. (1996)
Figure DR6. Schematic diagram showing derivation of equations used to solve for P-wave velocity of gas hydrate-bearing sediments. Velocity is solved for using VPU(t) and Th(t) (Steps 1-5), and is related to saturation using empirical relationships, rock-physics models, or theoretical correlations (Step 6 - see Lee et al., 1996 and Dai et al., 2004 for details). An example (in the absence of well control) is provided to further illustrate the concept (Step 7).
Step 7: Example
Inputs:
b = 1,700 m/s
Th(t) = 200 ms (TWTT) x 0.50 (net:gross) = 100 ms
VPU(t) = 50 ms
Solution:
b = 1.5 (Eq. 4)
gh = b b = (1.5) (1,700 m/s) = 2,550 m/s (Eq. 1)
Sgh = 0.46 (Eq. from Fig. DR3)
Sei
smic
(T
WTT
)
Th(t)
b
VP
U
(TW
TT)
VPU(t)
Consolidated sediments
Consolidated sediments
Unconsolidated sediments
Unconsolidated sediments
NN
Figure DR7. Map-view image (time slice) of the upper unit of the P3 system. A: Spectral decomposition of zero-phase full-stack seismic data (TWTT) showing sinuous high-amplitude high-frequency reflections. B: Interpretation showing variably saturated gas hydrate accumulations contained within the meandering P3 slope valley system. Deposits with the highest saturation (qualitatively based on amplitude and frequency) are located towards the north, and are situated westward of a gas hydrate-free slope valley system.
P3 system (upper unit) - uninterpreted P3 system (upper unit) - interpretedA B5 km5 km
Frequency(Hz)
30
20 40Frequency
(Hz)Inset High Low
Relative gas hydrate saturation
P3 reservoir
Slope valley system
Mud-filled valley above P2 system
Channelized sandy turbidite
Gas hydrate-free slope valley system
Erosional terrace
Gas hydrate- bearing slope valley system
UninterpretedA BNE
-1
+1
Am
plitu
de
SW
200
ms
5.0 km
Figure DR8. Seismic section and geometric models of a portion of the central Gulf of Mexico, showing convex reflections at depth. Features are interpreted to be VPU associated with gas hydrate accumulation in the P3 reser-voir. A: Uninterpreted zero-phase full-stack seismic section (TWTT). B: The section is interpreted to contain gas hydrate-bearing slope valley systems, resulting in underlying VPU that extends to the basement (i.e., 5,000 ms below the deposit). Because the average widths of P1-P3 systems (see Fig. 2C) are greater than that of the seismic streamer length (i.e., 6,000 m/19,685 ft), acquisition undershoot does not preclude imaging the accumulations. Yellow-black dotted line delineates the base of the GHSZ. C (top): Model showing gas hydrate-bearing coarse-grained reservoir responsible for underlying convex reflec-tions (i.e., VPU). Apparent axial surfaces converge upward, intersect layers 1-10, and extend from the base of the deposit to the basement. The model is consistent with observations from (A), particularly in that convexity increases in magni-tude and width with depth; this trend is a diagnostic feature of shallow high-velocity deposits (see Jones, 2012). Models showing reflection geometries associated with a drape onto a basement high (C - middle), and growth away from a basement high (C - bottom) (see Shaw et al., 2006). In these scenarios, axial surfaces associated with deformation are discontinuous above layer 5, and are therefore inconsistent with observations from (A).
Velocity effect: pull up
Basement
Apparent axial surfaces converge upward
Apparent basement high
Structural effect: drape
Axial surfaces converge upward
Fault plane
Convex reflection
Basement high
Structural effect: growth
Axial surfaces diverge upward
1
2
3
4
5
6
10
7
8
9
10
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
10
7
8
9
Flat reflection
Flat reflection
Basement high= Reflection 1 (R1)
= Reflection 2 (R2)
= Reflection 3 (R3)
= Reflection 4 (R4)
Interpreted Geometric models
Gas hydrate- bearing slope valley system
Velocity pull up
Flat reflection
Fault plane
Convex reflection
Convex reflection
Convex reflection
Gas hydrate-bearing coase-grained reservoir
C
= Transect
P3
VPU vs velocityLow- b model High- b model
A Model 1 Model 2
Model 3 Model 4
b
Th(t)
VPU
(t) (m
s)
1.21.11.0 1.41.3 1.61.5 1.81.7 2.01.9 2.22.1 2.30
100
300
200
400
500
700
600
b ( gh b) b ( gh b)
1,800 2,000 2,200 3,400 3,6002,400 2,600 2,800 3,000 3,200 3,800 4,000gh (m/s)
VPU
(t) (m
s)
0
100
300
200
400
500
700
600
1.21.11.0 1.41.3 1.61.5 1.81.7 2.01.9 2.1
1,800 2,000 2,200 3,400 3,6002,400 2,600 2,800 3,000 3,200 3,800 4,000gh (m/s)
0.90.8
0.9 1.00.80.70.60.50.40.30.2Sgh (v/v)
0.9 1.00.80.70.60.50.40.30.2Sgh (v/v)
Figure DR9. Inputs for calculating the P-wave velocity of gas hydrate-bearing sediments (see Fig. 3). A: VPU versus velocity of gas hydrate-bearing sediments using a low- and high-velocity background. Colored lines represent iso-values of reservoir thickness and show a linear relationship. Background velocity values are calibrated to the GC 955-H well (Fig. DR2). B: Two models for the P3 reservoir with a net:gross (N:G) thickness of 0.35 (left) and 0.65 (right); the gross thickness of the reservoir ranges from 150 ms (TWTT - NW) to 450 ms (TWTT - SE).
Reservoir thickness vs distanceLow-Th(t) model
BHigh-Th(t) model
100200300400500
0
Th(t)
(ms)
0 0140 16012010020 40 60 80 140 16012010020 40 60 80Distance (km) Distance (km)
SENW
100200300400500
0
0.65 N:G = 1.00 N:G =
0.35 N:G = 1.00 N:G =
Th(t)
(ms)
SENW
500=400=300=200=100=
Th(t) (ms)
350=
150=50=
450=
250=
500=400=300=200=100=
Th(t) (ms)
350=
150=50=
450=
250=
b = 2,100b = 1,700
NW
P3 system - uninterpreted
P3 system - interpreted
SE
x zy500
ms
10.0 km
2D data3D data
A
B
= Transect
P3
Channel-levee complex
Channel-levee complex
SaltSalt
Slope valley system
Slope valley system
Reservoir boundaries= Top gas hydrate (peak)
Base gas hydrate/top free gas (trough)=
= Base free gas (peak)
= Free gas= Gas hydrate
Reservoir type
Figure DR10. Seismic traverse along the thalweg of the P3 system, showing basinward-thickening wedge-shaped reservoir. A: Uninterpreted zero-phase full-stack seismic section (TWTT). B: Interpreted section showing gas hydrate-bearing deposits of the P3 system located well above the base of the GHSZ. Towards the northwest (left), the P3 system is intersected by a fault associated with the top of a salt diapir. In medial locations, the P3 system is interpreted to consist of upper and lower reservoir units surround-ing a gas hydrate free zone (GHFZ). Towards the southeast (right), the top of the P3 reservoir is situated approximately 100 ms (TWTT) below the modern sea floor.
Fault plane
Base GHSZBase GHSZ
Longitudinal extent of Fig. DR7
Sea floorGHFZ
Upper unit
Lower unit
w
x zyw
x
z
y
w
w
x
y
z