Contact author wangqiqi@utexas.edu for details.

1. CHARACTERIZING BEDDING-PARALLEL
FRACTURES IN SHALE:
Aperture-size distributions and spatial organization
Qiqi Wang1,2 and Julia F. W. Gale2
1. Jackson School of Geosciences, University of Texas at Austin
2. Bureau of Economic Geology, University of Texas at Austin

2. Opening-mode
Bed-parallel
fracture with
fibrous calcite
cement
500 µm
Thickest fracture: 8.7 cm
Thinnest fracture: ~15 µm (SEM image) Greatest lateral extent: > 32.6 m
Calcareous
concretion
Bed-Parallel Fractures

3. Motivation
• Bed-parallel fractures are common in shales and may
impact fluid flow in hydrocarbon reservoirs and
propagation of hydraulic fractures
- hydraulic fracture height growth inhibition
- hydraulic fracture horizontal propagation
• No published systematic studies of size scaling and
spatial distribution of bed-parallel fractures.

4. • Knowing the aperture-size scaling and spatial
organization of bed-parallel fractures will contribute to
modeling of fracture networks
- They are important for quantifying the
mechanical behavior of fractured rock masses
- Fracture density, size, and orientation obtained
from field and core help defining fracture
distributions when simulating stochastic fracture
network
Motivation

5. Locations
Vaca Muerta outcrop and well
locations
- 1 field area including 3 Vaca Muerta
outcrop
- 5 wells; datasets including core scan
images, sampled cores, and well-logs
Field Area
50 Km

6. Marcellus Fm.
- 2 wells datasets including
continuous core
(2 vertical scanlines)
Locations
Wolfcamp shaleMarcellus shale
Wolfcamp Fm.
- 1 well dataset including
continuous core
(1 vertical scanline)

7. Scanline Method
Field data Direct core measurement
Core scan mosaic data

8. Spatial Organization
• Where do fractures occur within the rock column?
• Are they more likely to occur within certain
lithology than others?

9. Spatial Organization Hypotheses:
• Bed-parallel fractures are more intense in
organic-rich layers
• Bed-parallel fractures form preferentially along
mechanical interfaces

10. GR TOC-KerogenLab TOC
Fracture intensity
Fracs/m
Intensity - Core vs. Log
• Lab-TOC spikes
match with 4 major
fracture intensity
peaks.
• TOC-KER is also a
good match to
intensity, but
generally not as
well matched as
lab TOC.
TOC-Kerogen: Amount of TOC based on Kerogen volume. Estimated from GR, sonic, density,
neutron and NMR well log data
0 20

11. Another well:
At some depth high TOC
correlate to high intensity
but not always the case.
No lab TOC data for this
depth column.
Intensity - Core vs. LogTOC-Kerogen
Core 2

12. Photos of fractures along interfaces
BPF associated
with cncretions
BPF associated
with host-rock
lithology change
concretion
concretion

13. Spatial Organization and Interfaces
occurrence % of material
interfaces with
fracture
occurrence
wells Total # of
fractures
At material
interfaces
At lithology
change
At concretion
Vaca Muerta well #1 341 105 (30.7%) 35 71 64%
Vaca Muerta well #2 229 55
(24.0%)
44 11 76%
Vaca Muerta well #3 142 44
(30.9%)
44 N/A 70%
• Material interfaces includes abrupt lithology change and concretion
margins
• 25-30% of the bed-parallel fractures occur at material interfaces.
• ~ 65-75% of material interfaces have bed-parallel fractures

14. Aperture Size Population
1mm
5
c
m
0.05mm

15. Aperture-size scaling hypotheses
For bed-parallel fractures aperture-size scaling :
• is power-law
– Vertical fracture aperture-size distribution is
commonly power-law
• follows a different function (e.g. exponential)
• has preferred size(s)
• follows no pattern

16. VM_Well #1
340 fracs,
73.28 m
Best fit:
Negative Exponential
Correlation Coefficient:
R²=0.9923
From core scan mosaics
Results

17. Marcellus_Well #2
82 fracs,
163.46 m
Best fit:
Negative Exponential
Correlation Coefficient:
R²=0.9348
From direct observation of core
Results

18. VM_Field #2
54 fracs,
13.2m
Best fit:
Negative Exponential
Correlation Coefficient:
R²=0.9801
From outcrop
Results

19. Results
Core/Well # Scanline Length
(m)
# of fracs Best Fit Model
(N.E. or P.L)
Best Fit Equation Correlation
Coefficient
(R²)
VM_Well #1 73.28 340 N.E. y = 4.6538e-0.655x 0.9923
VM_Well #2 34.69 230 N.E. y = 6.2313e-0.465x 0.9744
VM_Well #3 28.00 142 N.E. y = 6.3e-0.663x 0.9939
VM-Well #4 18.99 12 P.L. (?) y = 0.3317x-1.143 0.9537
VM_Field #1 17.84 88 N.E. y = 3.5994e-0.056x 0.9507
VM_Field #2 13.20 54 N.E. y = 3.9522e-0.067x 0.9801
VM_Field #3 48.46 30 N.E. (?) y = 0.5092e-0.224x 0.9549
Marcellus Well #1 89.18 47 N.E. y = 0.3543e-0.825x 0.918
Marcellus Well #2 163.46 82 N.E. y = 0.3386e-0.492x 0.9348
Wolfcamp Well #1 141.88 68 N.E. (?) y = 0.3557e-0.185x 0.9165
Total 628.98 1093 N.E.

20. • Fracture Intensity:
VM>Wolfcamp>Marc
ellus
• Vaca Muerta core
data contains large
number of thin
fractures but lacks
thick ones.
• Vaca Muerta outcrops
have high number of
thick fractures
Discussion

21. 0.00001
0.0001
0.001
0.01
0.1
1
10
100
0.01 0.1 1 10 100
CumulativeFrequency(Fracs/m)
Fracture Aperture Size (mm)
Compiled Aperture Size Distribution Plot (10 datasets)
MarcellusWell
#1
MarcellusWell
#2
VM_Well #1
VN_Well #2
VM_Well #3
VM_Well #4
VM_Field #1
VM_Field #2
3/7/19
vertical E-W
fracturesin tuff
VM_Field_V-Frac
#1
VM_Field_V-Frac
#2
VM_Field_V-Frac
#3

22. • Data collected indicate bed-parallel fracture aperture size
distribution follows a negative exponential distribution.
• Studied bed-parallel fractures are more intense in organic-rich layers
in some cases, but not in others.
• 25-30% of the bed-parallel fractures occur at material interfaces.
• ~65-75% of material interfaces have bed-parallel fractures
Conclusions

23. Questions?

24. References
Bonnet, E., O. Bour, N. E. Odling, P. Davy, I. Main, P. Cowie, and B. Berkovitz, 2001, Scaling of
fracture systems in geological media: Reviews of Geophysics, v. 39, no. 3, p. 347–383.
Cobbold, P.R. and N. Rodrigues, 2007, Seepage forces, important factors in the formation of
horizontal hydraulic fractures and bedding-parallel fibrous veins (‘beef ’ and ‘cone-in- cone’):
Geofluids, 7, p. 313–332.
Gomez, L.A., Laubach, S.E., 2005, Rapid digital quantification of microfracture populations, Journal
of Structural Geology, 28 (2006) 408–420.
Gale, J. F. W., Ukar, E., Elliott, S., and Wang, Q., 2015, Bedding-parallel fractures in shales:
characterization, prediction and importance (abs.): AAPG Annual Meeting, Denver, Colorado.
Ortega, O., 2002. Fracture-size scaling and stratigraphic controls on fracture intensity. Ph.D.
dissertation, The University of Texas at Austin.
Ortega, O., Marrett, R., Laubach, S.E., 2006, A scale-independent approach to fracture intensity
and average spacing measurement, AAPG bulletin, V. 90, Issue. 2, p. 193-208.
Rodrigues, N., P. R. Cobbold, H. Loseth and G. Ruffet, 2009, Widespread Bedding-parallel Veins of
Fibrous Calcite (‘beef’) in mature source rock (Vaca Muerta Fm, Neuquen Basin, Argentina):
Evidence for overpressure and horizontal compression: Journal of the Geological Society,
London, v. 166, p. 695-709.

25. y = 0.3557e-0.185x
R² = 0.9165
y = 0.2069x-0.458
R² = 0.8824
0.01
0.1
1
0.01 0.1 1 10 100
CumulativeFrequency(Fracs/m)
Fracture Aperture Size (mm)
Aperture Size Distribution
Wolfcamp_Wiggo Expon. (Wolfcamp_Wiggo)
y = 0.2396x-0.44
R² = 0.9917
y = 0.4723e-0.201x
R² = 0.9978
0.01
0.1
1
0.01 0.1 1 10 100
CumulativeFrequency(frac/m)
Fracture Aperture Size (mm)
Aperture Size DIstribution
Wolfcamp Wiggo-Power Wolfcamp Wiggo-Exponential
Power (Wolfcamp Wiggo-Power) Expon. (Wolfcamp Wiggo-Exponential)