SALT Workshop

1. 1
Comments of the German Association for
Repository Research (DAEF)
Till Popp, Wolfgang Minkley
Institute for Geomechanics GmbH (IfG)
Washington, DC
September 7-9, 2016

2. 2
„Deformation assisted fluid percolation in salt“
(Ghanbarzadeh et al., 2015)

3. 3
Ghanbarzadeh et al., 2016 - What is the explosive?
 The study confirms already existing observations of Lewis
& Hollnes (1996), but …
However, deformation may be able to overcome this
threshold and allow fluid flow.
The observed hydrocarbon distributions in rock salt require that
percolation occurred at porosities considerably below
the static threshold due to deformation-assisted
percolation.
Therefore, the design of nuclear waste repositories in salt
should guard against deformation-driven fluid
percolation. In general, static percolation thresholds
may not always limit fluid flow in deforming
environments.
 The low permeability of static rock salt is due to a
percolation threshold.



 Sophisticated analysis is needed to proof if
the integrity of the geological barrier salt
is really violated according to this thesis
 Why are Hydrocarbons inside salt?


4. 4
Percolation threshold – Evaluation of salt barrier integrity
Percolation threshold is a mathematical concept
related to percolation theory, which is the formation
of long-range connectivity in random systems.
Below the threshold a giant connected component
does not exist; while above it, there exists a giant
component of the order of system size.
https://en.wikipedia.org/wiki/Percolation_threshold
 Permeability in a salt barrier can be only induced
under special mechanical or hydraulical
conditions which result from the same micro-physical
process, i.e. the percolation of flow paths along grain
boundaries after exceeding a threshold.
 This corresponds:
(1) at deviatoric conditions with the dilatancy
boundary and
(2) at increased fluid pressures with the
minimum stresses

5. 5
Geomechanical proof of the salt barrier integrity
Cap rock
GOK
drifts
Underlying bed
clay
anhydrite
salt
potash
f (I2)
Dilatancy
boundary
Dilatancy field
Compaction field
I1
Cap rock
GOK
drifts
Underlying bed
clay
anhydrite
salt
potash
Detail A
pL = gL ∙ h
s3 s3
Hypothetical
gradient
pL = Fluid pressure
s3 = minimum stress
 Generally accepted!
 Part of the integrity analysis, as
requested by the GERMAN SAFETY
REQUIREMENTS
(BMU, 2010)
Dilatancy criterion
Minimum stress
criterion

6. 6
A 3rd threshold ? The dihedral angle q for the salt-brine system
PT-dependent change of the
wetting properties between
brine and salt :
• Change of the wetting angle –
development of connected
pore channels along triple
junctions (Lewis & Holness,
1996).
• Increase of the permeability
up to 10-16 m2 (Schleder et al.,
2007).
q > 60°
Rock salt is tight
Increase of
permeability
is possible
q < 60°
Rocksalt is
permeable
 Well known theory, but
neglected because the pT-
conditions are not relevant
for a salt repository (e.g.
VSG)
Repository

7. 7
The material tested vs. natural salt
Table salt: Grain sizes: 200 – 400 µm
Composition: analytical grade halite
(99.9% pure)
For each experiment, about 150 mg of
halite and 7-15 mg of distilled water used
 Water will dissolve salt
 4,5 – 9 wt.-% correspond to
7 – 16 Vol.-% = brine filled
porosity of the condensed
material
Natural rock salt (e.g. bedded salt Harlingen)
Grain sizes: < 1 mm ... 10 mm … 1dm
Composition: > 90% Halite, anhydrite, clay, accessoric minerals
 Water content, usually 0,1 – 1 wt.-%
20 mm
 The grain scale distributions and the water
content are not realistic!

8. 8
The experimental approach – Undrained Hydrostatic Experiment
Hydro-thermal autoclaves with ovens
pmax = 400 MPa
External Furnace Tmax ≤ 850°C
 The teflon capsule is positioned
inside a platinum tube (5 mm
outer diameter) in a
 Pressure vessel:
20 to 100 MPa; 100 to 275°C
 Quenched to room
conditions within 1 minute, i.e.
putting the autoclave in water
Teflon capsule and
covered with a Teflon lid  Grain boundary opening due to
unloading effects respectively
thermal shrinking

9. 9
The texture investigation method - Pore-Scale Imaging
Equilibrated salt-brine structure, reconstructed from X-ray micro-
tomographic images of synthetic salt sample scanned by 1.1µm
resolution. Each side of the cube corresponds to 660µm. left: Salt
grain separation using watershed algorithm. right: medial-axis or
skeleton of pore space. (source:
https://sites.google.com/a/utexas.edu/ghsoheil/research)
P=100MPaandT=275°CP=200MPaandT=100°C
University of Texas High-Resolution X-ray Computed
Tomography Facility
Zeiss (formerly Xradia) microXCT 400 scanner:
 3D resolution to ca. 1.1 μm pixel
 Image analysis:
 Reduction of noise level
 Converting grayscale image
data into segmented images
 Filtering the segmented data
 Quantification and post
processing
 Pore space topology and
connectivity
 Estimate of the Dihedral Angle
 It’s an indirect method which
quality depends on the data
evaluation …

10. 10
The original data sets from Lewis and Hollnes (1996)
 The pore property changes
described by the dihedral
angle are small !!
With respect to experimental artefacts due to
grain boundary opening ..
 What is the reliability of the estimate
of the Dihedral Angle?

11. 11
The static pore-scale theory - influence of porosity /water content?
 Under consideration of natural
salt properties the observed
phenomena are probably not
realistic, mainly due to the
unusual high water content
realized in the experiments
(source: https://sites.google.com/a/utexas.edu/ghsoheil/research)
Modeling of Textural
Equilibrium using synthetical
3D-networks indicates that
independently from the
dihedral angle connectivity of
pores depends on porosity
(respectively the amount of
brine)
Soheil Ghanbarzadeh
dihedral angle q
Porosity

12. 12
Experiment vs. Nature: salt from drill holes
q > 65°
60° < q < 65°
q < 60°
Petrophysical observations in wells -
Occurence of HC in salt
Low rock resistivities and
occurence of HC may give hints for
fluid-connected pore space due to
percolation

13. 13
UV-LightArtifical light
2 m 2 m
Occurence of hydrocarbons in the salt dome Gorleben (1)

14. 14
Occurence of hydrocarbons in the salt dome Gorleben (2)
 Conclusion in Greenpeace (2010) –
Hydrocarbons below a salt deposit
contradict against a repository because
the fluids will migrate through the salt 
no long-term safety
Presentation of J.
Hammer (BGR)

15. 15
Observations on ARA salt cores from the OMAN Salt basin
Schoenherr, J., J.L. Urai, P. Kukla, R. Littke,
Z. Schleder, J.-M. Larroque, M. Newall, N. Al-
bry, H. Al-Siyabi, and Z. Rawahi (2007): Limits
to the sealing capacity of rocksalt: A case
study of the Infra-Cambrian Ara Salt from
the South Oman Salt Basin: AAPG Bulletin,
v. 91/11, p. 1541-1557.
 Salt is impregnated by oil / bitumen
 contact between carbonate stringers / salt

16. 16
Cartoon of the mechanism of diffuse dilatancy of ARA salt
Stage 1: Diffuse dilation of
salt
• Evidence for decompaction and
local damage during uplift
• Over-pressurisation of Stringer
fluids
 violation of the minimal stress
criterion
 Pressure-driven percolation
Stage 2: Re-sealing
• After the oil pressure dropped to
values equal to s3 in the rock salt
(initial situation)
 The micro-scale deformation
mechanisms (dislocation creep
and fluid-assisted grain-boundary-
migration recrystallization) results
in re-sealing of the fluids
from Schoenherr et al., 2007
Stage 1:
Stage 2:
 Violation minimum
stress criterion Pressure-driven
percolation

17. 17
Is the PT-dependent dihedral angle q – threshold real?
IfG Kármán pressure cell:
 s3 – max = 1000 bar
 T up to 120°C
 Hydrostatic / deformational
conditions
 Sample size (natural salt)
 Length 200 mm
 Diameter 100 mm
 Permeability testing with
 Gas, brine and oil
 Preliminary test results are
available

18. 18
Preliminary lab test results at T = 97°C and increasing confinement
 No gas flow was detected!
 Resolution better than 10-20 m2.

19. 19
„Deformation assisted fluid percolation in salt“
(Ghanbarzadeh et al., 2016)
Conclusions
My personal opinion,
it is a well written and interesting paper but the authors wanted to amplify the
public interest by including aspects of nuclear waste storage.
However, there are some remarks regarding the conclusions:
• Sample size and water content are not appropriate to natural salt
• The thesis, that salt becomes permeable if the dihedral angle becomes
lower than 60° is only based on theoretical models.
• Porosity resp. fluid content is an important parameter for the porespace
network, but not discussed by the authors in the paper.
• The occurence of hydrocarbons is not always an indication of permeability
(autochthonous origin is possible)
• It is a well known fact that salt can become permeable so that fluids can
migrate through it (especially during the diagnesis or salt dome uplift), but
the dilatancy and minimal stress criterion are sufficient to explain the
acting processes.
As summary, from our point of view, the integrity of salt as host
rock for storage of radioactive waste is out of question.