Medium-deep or very deep disposal of highly radioactive waste?

tickbitewryΜηχανική

30 Οκτ 2013 (πριν από 3 χρόνια και 10 μήνες)

135 εμφανίσεις

1


Medium
-
deep or very deep disposal of highly radioactive waste
?



Roland Pusch
1
*
,
Gunnar Ramqvist
2
,

Sven Knutsson
1
, Mohammed Hatem Mohammed
1

1.

Department of Civil, Environmental and Natural Resources Engineering, Luleå University of
Technology, Luleå, Sweden;
*
Corresponding Author

Email
: drawrite.se@gmail.com

2.

Eltekno AB,

Oscarshamn
,

Sweden


Abstract
:
Several of the commonly proposed concepts for disposal of highly radioactive
waste (HLW)
imply construction at medium depth (400
-
6
00
m) in granitic
rock
, which

is
excellent
for constructing a stable repository since it provides effective mechanical protect
ion
of the waste
. A drawback is that m
ajor water
-
bearing fracture zones are
frequent
and must be
avoided in the site selection process since they can undergo large deformations caused by
seismic and tectonic events and
cause failure of waste containers
located in or near them
.
The
effect of such events can be minimized by surroun
ding them
with
ductile
“buffer” clay that
retards groundwater
-
driven
adflow.
An alternative concept i
s

placement of HLW in very deep
boreholes
(VDH)
where the rock is much less p
ermeable and where the very salt, heavy
groundwater is stagnant. The
boreholes are proposed to be 4

km deep and grouped in a small
number of sites. The upper 2 km
parts
, with temperatures lower than about 100
o
C,
are
sealed

by being filled
with
perforated s
upercontainers with dense clay blocks,
while the lower part

contain
s

supercontainers with
wast
e c
anisters

and dense
clay blocks, raising

the temperature
between 2 and 4 km to 100
-
150
o
C.
The
holes are kept filled with clay mud into which the
supercontainers

are inserted

where
the rock contains few fractures
,

while concrete is
cast
where the rock is fracture
-
rich.
In the upper part c
lay migrates through
the
perforated

supercontainers
and consolidates the mud. In the lower part clay
the same process takes plac
e
where the
clay block in each supercontainer
is located,
while the rest of the mud retains its
original

low

density but undergoes stiffening. In the upper, sealed part of the hole, the
consolidated clay will
be
much
tighter than the surrounding rock, whil
e
in
the lower
part the
mud will be more permeable but still capable of
limit
ing

wa
ter circulation within the hole
.
The paper compares the two
repository
principles and
recommends closer examination of the
very deep hole concept, which has obvious
advantages respecting both performance and cost.


Keywords:

Canisters, Deep boreholes, Groundwater, Radioactive w
aste
, Site selection,
Supercontainers, Waste d
isposal
.


1.

Background


Several comprehensive studies like the CROP project [1] have been
conducted for
assessing the capacity of proposed repository concepts for isolating highly radioactive waste
(HLW) from the biosphere. T
he major conclusions from them
were that different geological
media
-

crystalline rock, salt, and argillaceous rock
-

hav
e advantages and disadvantages in
various respects and that disposal in crystalline rock requires more effective engineered
barriers than
the others
because of its higher permeability. H
owever,
crystalline rock provides
more

stable conditions and
is presen
tly a primary
candidate
for hosting repositories in
Sweden, Finland and Canada. Underground laboratories (URLs) for testing the performance
of rock and engineered barriers (EBS) have
shown that they
perform according to predictions
and plans and represent
a large enough variety of conditions for providing credible generic
2


information on repository concepts. Commonly proposed depths for locating a repository of
one
-
level type is
4
00

600 m. Access to the repositories is through
shafts
and shafts.

As to the
h
ost rock three major factors

are essential: 1) the structure on large and small
scale because it controls the impact on the repository by earthquakes and tectonic movements,
2) the groundwater flow, and 3) the impact on the physical

and chemical performanc
e of

engineered barriers

(EBS)
. These are metal containers
(canisters)
containing the waste (spent
fuel or vitrified, processed waste), and clay
-

termed buffer
-

surrounding the containers for
minimizing groundwater flow in their vicinity and for
providing ductile embedment of the
canisters (Fig
.
1). The waste generates a heat pulse
that lasts

for several hundred years
and
can strongly affect the
function of
the
barriers
and the groundwater flow, as well as the stress
conditions in the rock, and th
ereby its physical stability.


Fig.
1

Schematic section of a tunnel with deposition holes with clay
-
embedded canister
s

containing heat
-
producing high
-
level waste

according to the KBS
-
3V
1

type
[1
]
.


D
eep h
ole concept
s

(VDH)
imply waste placement in
the lower part of 3
-
5 km deep holes
and effective
sealing
of the upper par
t

(
Fig.

2)
, [2
,3
,4,5
]
.
Several holes will be bored in
slightly
different directions from a chamber at a depth of some tens of meters below the
ground
surface, by which the distance betwe
en the waste
-
bearing parts can

be sufficiently
large to avo
id interference and superposition of

temperature fields, and sufficiently small to
establish a confined common space for rational establishment of the boring sit
e.
In Sweden,
e
arlier concepts suggested a hole diameter of 1.675 m from its upper end to 2000 m depth and
a diameter of 0.8 m down to 4000 m,
or
0.76 m diameter down to 3000 m
,

and 0.375 m down
to 5500 m depth.
A recent version, discussed in this paper, h
as a constant diameter of 0.8 m
below a few hundred meters.
While it may well be possible to make stabl
e large
-
diameter
boreholes it
is
judged
necessary to provide them with extra support in the form of casings




1

V is for vertical

3


over the whole borehole length

except
where c
oncrete seals are constructed. The casings must
be chemically compatible with the supercontainers and canisters
.






Fig
.
2 Schematic view of a VDH with constant diameter below the uppermost
0.5 km
concrete fill
, below which
the hole is tightly sealed with clay to 2 km depth. The
“disposal” zone with sets of connected HLW canisters separated by clay blocks is shown
here in a hole with 0.8 m diameter

[5
]
.


2.

H
ost rock


2.1

Structure



For
technical/scientific purposes and for predicting the thermal, mechanical and hydraulic
performances of the host rock in bulk one needs to use definitions of discontinuity elements
with typical properties respecting the hydraulic function and rheological pe
rformance. Major,
water
-
bearing structural elements, represented by
dominant
fracture zones termed 1st and 2nd
order discontinu
ities [6
,

7
], should not intersect
waste
-
containing parts of a mined
repository
and are hence of primary importance in the
site
-
selection process (
Fig.

3
). They can be
identified by examination of cores from deep, vertical and graded boreholes and by in
-
situ
logging using geophysical methods
.
Finer ones, of which 3rd order discontinuities,
i.e.
minor
fracture zones, and those
of 4th

order being

discrete, persistent

fractures, determine the
performance of the rock where waste containers are placed. They remain unidentified until
construction work has started.
Even finer discontinuities, i.e. those of 6
th

and 7
th

orders,
belong to the rock matrix and
represent
its
content of
por
es

and fine fissures
.


Fig
.

4 shows a structural model
derived on the basis of core mapping and
geophysical
investigations including
in
-
situ hydraulic measurements. The discontinuities
represent fracture
zones of 2
n
d

order
forming distorted and curved
,

more or less
orthogonal patterns.
Fig.

5
illustrates
the pattern of
actually identified 4th order fractures
at 360 m depth in
another area,
showing similar winding nature of discontinuitie
s also on this smaller scale. The orientation
4


of the major principal plane follows approximately the orientation of one of the major sets,
the
highest
horizontal pressure ranging between 20 and
28 MPa.


Fig
.
3

Simplified model of a repository site

derived from comprehensive
core mapping

and geophysical measurements
. The light green area is the ground surface and the blue
plates 2
nd

order discontinuities with
about
100 m width. The red plates are
large
3
rd

order discontinuities with 10
-
30 m width. Black panels are
systems of
deposition tunnels

[7
].


Fig
.
4

Perspective view of
16 km
3

rock volume in southeastern Sweden

with
33

steep
2
nd

order
fracture zones intersected by up to 1 km deep boreholes:
19 zones
by KLX04, 4
zones by KLX06, and 10 zones by KLX10. The spheres indicate the upper ends of the
holes

[
8
].


5



Fig
.
5
Structural winding and v
ariation in orientation of the major horizontal stress (fat
arrows) in the system of 4
th

order discontinuities in the Stripa URL at 360 m depth.

The
spacing of the lines is 30 m
[
9
].



2.2

The
waste
-
isolating
role of the host rock


2.2.1

Major criteria



Safe disposal of HLW requires that there is
nearly
no dissemination of possibly released
radionuclides in at least 100

000 years as stipulated by Swedish authorities and
those of
several other countries.

While the function and role of engineered barriers, p
rimarily the waste
canisters

and their
clay
embedm
ent
,

are well defined and predictable, the
properties and
function of
rock hosting repositories
at a few hundred meter depth vary so much that its role
as barrier
is being questioned. I
n recent time
it has
been degraded to be a “mechanical
protection

of th
e chemical apparatus” [5
].

The major reason for this is the difficulty in
working out
reliable models of

its
hydraulic

performance
, especially in a long
-
term

perspective,
because of the very limited information on the structure, which is a
consequence
of the winding and
unknown interconnectivity and persistence of both low
-

and high
-
order
discontinuities.
A second reason is that the degree of utilization
, expressed as the number of
acceptable positions of HLW containers per
unit length of a
KBS
-
3 deposition tunnel or per
meter length of

the deployment segments of VDH
s, depends on the frequency of intersected
2
nd

and 3
rd

order discontinuities and of the stability of the rock around the deposition holes
.


The role of the host rock is
fundamentally different for the two concep
ts
KBS
-
3V, taken
as a representati
ve of medium
-
deep repositories,
and VDH
.

For KBS
-
3V it is to provide
mechanical protection of the waste while f
or VDH

it is to prevent groundwater at the waste
level to move up to shallow, permeable rock. This is fulfilled by
the high salt content that
makes possibly contaminated groundwater stay at depth.




6


2.2.2

Usefulness




(1)
KBS
-
3V

For a KBS
-
3V repository the
possibility
to locate

deposition holes, which are 1.95 m in
diameter and about 8 deep, cannot be predicted until rock excavation has reached the planned
repository level

and the position and orientation of transport
and deposition tunnels have

been
decided

[1]
.
After constructing these tunnels, which will provide rich information on the rock
structure and
proba
bly
cause
changes of the plans,
slim pilot holes
will
be drilled in different
directions for
general
exploration and
in the direction of
tunnel positions
p
roposed
on the
basis of gained information on structural features. A few holes with 3
-
400 m length drilled
within the 20 m2

section of
forthcoming tunnel
s

will reveal

the presence and orientation of
3rd order zones
and give

a first indication of whether
th
e plan is
worth pursuing.
A

large
number of
deposition hole
positions mu
st be investigated

for making sure that none will be
intersected by other than 4th an
d higher order discontinuities
before the entire repository
design can be decided.
Preparative work comprises boring of pilot holes for determining the
uniaxial compressive strength and the shear strength of intersected single fractures of 4th
order type
for predicting the degree of mechanical stability of the rock.

The
site investigatio
n and design work
will take at least a decade

for a repository with
6000
-
7000 canisters
but will

ultimately provide

plans of the type shown in ge
neralized form
in
Fig.

6. They
ind
icate the degree of utilization, which is commonly 50
-
60 %
, representing
the
percentage of useful hole positions.

Full
-
face boring of deposition holes
will be made by
using TBM
-
type technique.


Fig
.
6

Schematic plan view of a
KBS
-
3V
repository with a transport tunnel connected
to deposition tunnels in rock with typical frequency of

3
rd

order fracture zones (
D3
).
Tight bulkheads are keyed into the rock for isolating the waste
-
containing parts of the
deposition tunnels

[7
]
.


(2)
VDH

For a VDH repository the
basis of evaluation
of the degree of utilization for waste
placement is
much simpler since it is
confined to identifying 2nd and 3rd discontinuities that
will intersect the deep holes.
T
his
only
requires a prospec
tive investigation by
slim
core
7


drilling,
rock
characterization and
hydraulic and
strength testing

for deciding
how

frequent
ly

stabilization and placement of waste canisters
shall
be made.
Subsequent to this
,

detailed
design and boring of a full
-
size VDH can start.
The matter of where excavated rock shall be
transported
and stored
is an important
environmental issue for mined
KBS
-
3V
repositories but
is
of no concern for VDH.

An example of what the conditions are with respect to the degree of utilization

in a 1000
m deep hole at the HLW disposal site proposed by the Swedish Fuel and Waste Handling C
o
(
SKB) is illustrated in Table 1.
The e
xample indicates that the parts

useful for waste
placement represent

about 50 %. There are
, however,

strong
indications that the spacing of
water
-
bearing fracture
-
rich zones is larger at depth

as indicated by the Gravberg
-
1 boring
in
crystalline rock
[4
]
, which
showed extensive fracturing down to 1200 m depth but
only 2 to
20 m fracture zones with a spacing of at least 200 m deeper down
2
. This
would imply a
percentage

of 90 % or more
.

A cautiou
s estimate would be 75 %.


Table 1

Major structural features
i
n 1000 m deep borehole KFM07A [8
]

Depth, m

Structural features

Useful for
placing
waste canisters

0
-
210

Major 2
nd

and 3
rd

order fracture zones (DZ1, DZ2)

Unsuitable

210
-
260

Fracture
-
poor

Suitable

260
-
290

Fracture
-
rich

Unsuitable

290
-
325

Fracture
-
poor

Suitable

325
-
345

Fracture
-
rich

Unsuitable

345
-
415

Fracture
-
poor

Suitable

415
-
435

Major 2
nd

order fracture zone (DZ3)

Unsuitable

435
-
495

Fracture
-
poor

Suitable

495
-
505

Fracture
-
rich

Unsuitable

505
-
535

Fracture
-
poor

Suitable

535
-
545

Fractures

Acceptable

545
-
790

Fracture
-
poor

Suitable

790
-
1000

Major 2
nd

order fracture zone (DZ3)

Unsuitable


2.3

Constructability
and short term performance


While adaption of the deposition holes and HLW
canisters
to the rock structure are
major
geometrical
task
s

at the planning stage,
the final decision of the localization and design of
a
repository

and
start of
construction
requires that two more criteria are fulfilled, i.e.

1)
acceptance of the predicted

performance of the repository, and 2) demonstration that it is
constructable

and stable
.
The stability conditions are that the rock strength must be sufficient
for
preventing initiati
on of

practically important
fracturing
an
d
fall of rock fragments
from the
borehole walls
.





2

Hydraulic conductivity E
-
10 to E
-
9 according to the investigators

8



2.3.1

KBS
-
3V


For the KBS
-
3V depo
sition holes with 6 m spacing, 1.95

m diameter, and 8 m depth in a
repository at 400
-
500 m depth in rock with the typical horizontal rock stresses 20 and 40 MPa
and vertical stress 10 MPa, the theoretical maximum hoop (tangential)
stress
is about 120
MPa assuming the least critical orientat
ion of the tunnel [
10
].

Using the unconfined compressive strength,
which is commonly
150
-
350 MPa

for granitic
rock
, as practical parameter for assessing hole positions, application of statistical methods
can
be made for finding the degree of utilization.
For this purpose one needs to apply a lower
value of the compressive strength since initiation of macroscopic failure in fact takes place at
about 50 % of the conventionally determined strength value

[
11
].
Adopting this principle it
has been found that
a f
ew percent of the positions

have to be a
bandoned
in the site selection
process

as exemplified by
the ongoing construction of SKB’s HLW repository at Forsmark in
Sweden
.
Considering the winding
of fractures, fracture zones and major horizontal stress
,

a
certain fraction of the deposition tunnels will be less favourably oriented in the stress field
and
, realizing that

there will be superpositioning of the stress
es

from neighbouring deposition
holes
,
some
additional positions have to be given up
. T
he major
ity of the
planned
holes
w
ould still
be acceptable
taking
also

the risk of activation
and prop
agation
of natural, partly
hidden discontinuities of 5th and 6th order discontinuities
into consideration
[
7
]
.

However,
the situation will be totally
different w
hen the heat pulse

emerges

from the radioactive decay
.
T
he hoop stress w
ill rise

and
spalling
occur in
most
of
the
deposition holes, causing
a very
substantial increase in vertical hydrau
lic conductivity along them

[10]
.

Rock fall in
to

a
n empty

KBS
-
3V hole in the canister
-
placement phase can be very
problematic since a wedge of any size coming off from the
rock
wall
can
hinder further
deposition

or upward pulling
of the canister
(
Fig.

7
).







Fig
.
7

Wedge formed by
intersection of three
critically oriented 4
th

order discontinuities.

2.3.2

VDH

For VDH the matter of
stability is different
because of the higher stresses.
Usin
g
data
bases and generalized relationships between depth and stress states
[
6
]
the maximum and
mini
mum horizontal
stresses
can be estimated at
56

MPa and 39

MPa at 2

km depth
,
but
higher pressures have been
recorded
at this depth

as in
the
6 km deep
German KTB hole

[12
].
Here
,

the
maximum and minimum horizontal stresses were
found to be 7
0 and 30 MPa at 2
km depth
, and 120 and

70 MPa

at
4

km depth
. High pressures
have also been
reported from
measurements at 3 km depth in
crystalline rock as in
the Nojima fault in Japan
where the
maximum and minimum horizontal stresses were 88 and 64 MPa, respectively
[13
]
.

Assuming
,

for a forthcoming VDH
in Sweden
,

t
he
same
stresses
as in the German
KTB
hole

the
hoop stress
around a bored hole
would be
up to
10
0 MPa
at 2 km depth and 19
0 MPa at

4

km depth.
No correction for t
he impact of winding stress directions
would be necessary
9


below 2

k
m
since the variation in orientation of principal stresses
is reported to vary only
within 10
o

[14
].

As for the KBS
-
3V the thermal stresses will add to the static pressure in the deployment
part
but to a lesser extent because
of the
much
smaller
temperature rise
.

At 4 km deep holes
the
ambient temperature
is expected to be about 64
o
C

[
4
]
. Two possible canister
configurations
have been
considered, one with 4 BWR
spent
fuel
elements, and one with 2
BWR an
d 1 PWR.
The heat production of either of them would give a net temperature
of
90
o
C at 2 km depth and 150
o
C
at
4 km depth
assuming generally accepted data of the heat
generation and thermal properties of the
engineered barriers
and the rock [3
,4
].



2.4

Waste placeability


The
impact
of
borehole anomalies on the
placement of waste packages is
of
great
importance. For KBS
-
3V
ins
ertion of the 246
00 kg canister

in the center of the
about
7 m
high column of
clay
blocks

will be difficult because the space is only
10 mm
wide
. The
difficulty has not to do with
the straightnes
s of the deposition holes but with

the
verticalness
and the
pr
e
cision
with which the blocks
can be

placed
.
The tolerances are very small and the
technique of installation
, made remotely,

is
deemed
not practical

[7
]
.

For VDH t
he

p
laceability is determined b
y the straightness of the holes
while the degree
of verticalness

is
not important
. If
the waste packages
are too long they can be
come

stuck
in
curved borehole
s [8
].
For 800 mm borehole diameter
and an inner diameter of the rock
-
supporting casing
of
750 mm
,

supercontainers
with
700

mm outer diameter
and 25 m length
can be
moved down in the clay mud without difficulties
.
Fig.

8

illustrates the variation in
curvature of slim boreholes, which is believed to be
valid also for
a full
-
size VDH
using a
slim hole for guidance.


Fig
.
8

Example of
logging of
cored hole in the Forsmark region used in SKB’s current
project on sealing of
deep investigation holes

[8
]
.

All supercontainer units rest on previously inserted sets except where fracture zones are
intersected. Here, concrete is cast
,

after
local
reaming
of
the hole,
up to the level where the
fracture zone ends. The next sets of can
ister
-
containers are inserted when the concrete has
matured
sufficiently
for getting the required bearing capacity. Schematically, a VDH unit for
placement in the deploymen
t zone would look as in
Fig.

9
. In the overlying sealed zone
with
no waste
the supercontainers look the same but contain only dense clay blocks.
In the
deployment zone supercontainers filled with only clay blocks can be incorporated between
those with waste for avoiding criticality.


10



Fig
.
9

Princ
iple of composing supercontainers

in the deployment zone.
The
y sink

down in
clay mud with 16
00 kg/m
3

density

[
5
]
.

3.

Engineered barriers

3.1

Types and roles


Multibarrier concepts like KBS
-
3V
and
VDH
rely on the rock as
natural barrier and on
man
-
made barriers:

canisters with HLW and clay surrounding them
,

and
clay and concrete for
sealing shafts and holes
that lead

from the ground surface to the waste.
The temperature, to
which the barrie
rs will be exposed, can be up to

100oC for KBS
-
3V and 150oC for VDH
,

and
c
hange their waste
-
isolating capacity
,

requir
ing

assessment of their role in the
respective
repository.


3.2

Canisters



For the KBS
-
3V
and VDH concepts metal
waste
canisters
have been proposed as
major
engineered barriers. The presently favoured KBS
-
3V canister consists of cast iron
with 12
BWR

elements
of spent fuel, lined by 50 mm copper
,

and the latest VDH canister type,
favoured by the present authors, is
of the same type
but with only 4
BWRs
.
The
outer
diameter of the KBS
-
3V canister is 1050 mm and the height 4850

mm, its weight being 24500
kg.

The VDH
canister
, fitted in perforated supercontainers,

is
proposed here
to have
50
0

mm
diameter and
4850 mm

height
, weighing about 6000 kg.

The casing is proposed to be made of
an alloy of
Navy Bronze, consisting of
more than 90 % copper and
about
10 % nickel for
fulfilling the crit
erion of
mechanical strength and
chemical compatibility

[6
].

A small fraction of the canisters will have defects

from start and leak, which contaminates
the surrounding clay and the surrounding rock

with radionuclides
. Depending on the degree of
ductility of the clay
,

thermally or seismically generated displacements in the rock will be
transferred to the canisters
that may ultimately fail
[7
]. Water will migrate into the canisters
and make the iron core corrode and produce highly pressurized hydrogen gas that penetrates
11


the surrounding clay channe
l
-
wise and
migrate
further along
the rock contact
or within

the
rock.
The canisters in a KBS
-
3V repository will be exposed to tension caused by upward
expansion of the buffer clay

[7]
. This movement can be very significant depending on the
compressibility of the overlying
tunnel
backfill and can cause critical stresses and b
reakage at
the junction of copper liner and lid. VDH canisters of SKB type will be under high
isostatic
pressure but there will be no tension.


3.3

Clay

3.3.1

Alteration

Clay serves as a very effective

sealing component in both concepts
by retarding
groundwater flow around and along the canisters
and
by providing a homogeneous
ductile
embedment of the canisters

[1]
. It must be tighter than the surrounding rock and stay ductile
during the
entire operational time,
which is
taken as 100000 y
ears

in Sweden and Finland
.
This
requires use of expandable clay

of smectite type although other minerals like
palygorskite and mixed
-
layer types may turn out to be acceptable or
even
favourable

[
2
]
.

For
the KBS
-
3V concept, the clay surrounding the waste containers, termed buffer, is presently
proposed by SKB to be a smectite
-
rich clay consisting primarily of montmorillonite. For the
VDH concept there are two clay components of the same type as for KBS
-
3V but placed with
different densities, i.e. the mud, and the compacted blocks of KBS
-
3V buffer type.

The maturation of the buffer clay in KBS
-
3V and of the mud an
d dense clay components
in VDH
is quite different

and
affects
their long
-
term chemical
stability

differently
. For the
firstmentioned, the water pressure will be
relatively
low
and the inflow into the deposition
holes
low, which
can delay
water saturation
by
tens to hundreds of years
.
During this period
the desiccated clay is exposed to at le
ast 100oC and salt enrichment, dissolution and
precipitation processes on the microstructural scale can reduce the expandability and self
-
healing ability of the ultimately water saturated c
lay, and also make it stiffer [6
,
15
].

In the
upper,
sealed part of

the VDH the temperature will range from about 15oC to 60oC
causing no heat
-
induced changes and providing total tightness early after installation. In the
deployment part, which will be heated to 150oC

in its lowest part
, mineralogical changes and
less goo
d isolation are expected
as indicated by
Fig.

1
0

[16,17,
18]
.
Montmorillonite is
converted to (non
-
expansive) illite via mixed
-
layer smectite/illite minerals
or precipitation
,

and quartz
is
formed at a rate determined by the access
t
o potassium and
temperature [7,
19
].
Cementation is caused by neoformed quartz and illite. Natural analogues from various parts of
the world

indicate that Tertiary and Ordo
vician bentonites exposed to about 130
-
160oC for a
few thousand years have a significant part of thei
r montmorillonite content preserved, and a
common belief is that this will
also be the case for KBS
-
3V
and VDH
[7
].

12



Fig
.
1
0

Schematic diffractograms of a

montmorillonite
-
rich
reference
(
MX
-
80
)

sample
(20
o
C) and
of the most heated part of a

hydrothermally tested sample (130
o
C).
Feldspars, amphibole, some of the quartz and smectite disappeared in the h
ot part [15
].


3.3.2

Chemical evolution

In all parts of a

VDH the clay
blocks are

quickly water saturated under the prevailing high
water pressure
.
Th
e ultimate density of the
clay
blocks
in both KBS
-
3V and VDH repositories
will be 1900 to 2000 kg/m
3
, which gives the clay a hydraulic conductivity of no more than E
-
11 m/s at saturation with Ca
-
dominated water, and a swelling pressure of at least 1 MPa

[6]
.
Over the larger part of the length of the supercontainers
contain
in
g HLW canisters
the density
of the mud
will remain unchanged at 1600 kg/m
3

(dry density 950 kg/m
3
) but the physical
properties will change, implying an increase in hydraulic conductivi
ty and a drop
in swelling
pressure [15
].

The chemical evolution of the clays in both repository concepts depends very much on the
temperature and on the
salt concentration in the groundwater
as well as
on the interaction with
the canisters, supercontainers and casings.
Some minor exchange of the initially sorbed
sodium ions by copper ions will take place but
Ca, being the dominant cation in the strongly
brackish KBS
-
3V groundwater and in the very salt gr
oundwater (>10 g/l)
in the deeper parts
of a VDH [
6
], will control

the microstructural constitution and thereby the physical properties.


3.3.3

Impact of gamma radiation

Exposure of water saturated montm
o
rillonite
-
rich clay to strong gamma radiation and 90
-
135
o
C temperature has no degrading impact according to French investigations but speeds up
the release of Fe from metal iron like
iron or s
teel canisters [15
]. However, creep testing of
samples from 1 year
long
hydrothermal test gave witness of significant sti
ffening.


3.3.4

Modelling

The

model of mineralogical changes in montmorillonite proposed by Grindrod and Takase
[20
] considers dissolution and precipitation of phyllosilicates by taking O
10
(OH)
2

as a basic
unit and defines a general formula for smectite (S) and illite (I) as:

X
0.35

Mg
0.33

Al
1.65

Si
4
O
10

(OH)
2

and K
0.5
-
0.75
Al
2.5
-
2.75

Si
3.25
-
3.5

O
10
(OH)
2



(
1
)

13


where X is the interlamellar absorbed cation (Na) for Na montmorillonite. According to
the model the rate of the reaction r can be expressed as:



r
=A

exp (
-
E
a
/RT)(
K
+
)
S
2





(
2
)

where:

A
=coefficient
,
E
a
=activation energy for
the conversion of montmorillonite to illite
(
S/I
),

R=universal gas constant
,
T=absolute temperature
,
K
+
=potassium concentration in the
porewater
, and
S
=specific surface area for
reaction
.


(1)
KBS
-
3V

Applying

t
he
Grindrod/Takase
model to the KBS
-
3V case silica will be released and
transported from the hottest to the coldest part of the clay buffer

[
6
]
. Thus
, assuming
a
linear
temperature drop with time to 25
o
C after 10000 years
,

s
ilicification and illite formation
would
be
initiated after 500 years, the firstment
ioned
withi
n about 0.1 m from the rock
,

and the latter
occurring in the hotter part (60
-
100
o
C).


(2)
VDH

As to the m
ineralogical stability
of th
e VDH clay components
the same applies for
the
sealed
upper
part as for KBS
-
3V, while the mineral content of the mud and clay in the
deployment part will be transformed and serve differently with time. The high temperature
-

100 to 150
o
C
-

is believed to
give
essentially th
e same effects as those
predicted by using the
model by Grindrod and
Takase but the rates of creation of illite and
stiffening
will be much
higher than for KBS
-
3V and the
colder
part of VDH.
Thus, u
sing Eq.3 one finds that an
increase in clay temperature f
rom 100 to 150
o
C speeds up the rate

of converting
half
the entire
original content of montmorillonite
to illite
by about
100 times, assuming the activation
energy to be 27 kcal/mole

and all other factors being the same
.
T
his significant loss of
effective

s
ealing will take place in about 100 years but the practical importance is small since
the
isolating ability
of the re
maining clay is still considerable.
Thus, for pure illite with a
density of 1600 kg/m
3

the hydrauli
c conductivity is lower than E
-
8

m/s
[
21, 22
].

Total
conversion to non
-
expandable illite
in a dominant part of the dense clay
in the supercontainers
would in fact take 100000 years

because the controlling mechanism is diffusive transport of
potassium from the surrounding rock

[7]
.
Total
degrad
ation
by illitization
will occur at
approximately
the time
whe
n the
canisters
are through
-
corroded and
start leaking
.

The VDH concept discussed here implies that the boring mud is the same as the
deployment mud. For serving in the boring phase it must have the required ability to support
the hole being bored and to bring up the debris from the boring head.
T
he supercontai
ners
with waste canisters and dense blocks of expansive clay
shall

sink down or be pushed
through
the mud to the predetermin
ed levels
,
which
requ
ires that its

shear resistance
is
sufficiently
low. After placement the mud will be consolidated by the expanding dense blocks
in the
supercontainers,
intermittently seal
ing

off the holes in the deployment zone, and completely
tighten
ing

the
holes in the upper, sealed

part. The performance

of the mud in the two parts
will be largely different because of the difference in temperature and salt conditions. For the
deployment part, where the role of the heavy salt water is most important for isolating
possibly released radionuclides, the only c
riterion is that the consolidated mud should stay in
contact with the rock and canisters for minimizing convective water flow. For the dense clay
in the supercontainers in the 2

to 4 km deep deployment part

and in the sealed part down to 2
km, it is requir
ed that the expanded clay stays less permeable than the surrounding rock,
which is conservatively taken as E
-
11 m/s.


14


3.4

Installing
canisters in the mud


For
making
a set

of
superc
ontainers

weighing
about
20000

kilograms

sink or be pressed
down in the mud
its
shear resistance must
give it a lower
bearing capacity than the
load
exerted by the set
.

The resistance to moving the supercontainers is caused by the point
resistance and the
friction at the slip

along
the supercontainer.

The theoretical
point
-
bearing
capacity of clay
is 5
-
10 times the shear strength
3

[
23
]. The shear strength required for carrying
a 20000 kg
set of
supercontainer
s

exerting a
downward
pressure of
about
0.5

MPa
is
hence
on
the order of
5
0 to
10
0 kPa.
It can be estimated from the
swelling pressure and the angle of
internal friction, which can be taken as 15
o
, giving the approximate relationship between the
density and shear strength

in Table
2
.
It shows that the highest acceptable mud density
for
making
such supercontainer sets

start
sink
ing

without
adding loads
is
about
1600 kg/m
3
.
Once
failure has
been initiated
and the
supercontainer has started slipping down
,

the wall fri
ction
along its outer surface
is mobilized and
provides

additional
resistance. It is the product of the
sheare
d surface area and the residual

shear strength, which, for smectite clay
,

can be a fraction
of the shear strength at rest. Depending on the length of the containers t
he
y
will slip down
without additional force
or
require pushing
.
As at piledriving in

such soils a
vibrator with
variable frequency

(“vibroflotation”
)

may have to be used

if the
super
container sets are lighter
than 20000 kg.

The
movement is associated with upward flow of
remoulded
mud
along them.


T
able 2

The shear strength of
Na
smectite
-
rich (
montmorillonite
)

clay of different
densities.

Density, kg/m
3

Swelling pressure,
kPa

Shear strength, kPa

1300

45

12

1400

100

2
7

1500

200

5
4

1600

35
0

95

1700

5
50

149

1800

800

216


3.5

C
oncrete components


Concrete will not be used in a
KBS
-
3V repository

except in investigation boreholes

and
temporary bulkheads in tunnels
. In VDH, representing
,

in principle, a large
-
scale
analogy of a
borehole
to be sealed
,


dense clay will provide tightness
where the holes are located
in low
-
permeable ro
ck
,

and
concrete, serving as filter for avoiding loss of clay,

where they intersect
major discontinuities like fracture zones of 2
nd

and 3
rd

orders

[8
]
.

The
main
criteria for the concrete seals
in VDH
are:

1)
s
ufficient fluidity at

casting
, 2)
s
ufficient
bearing capacity for

carry
ing

the overlying
super
containers
, 3)
l
ower hydraulic
conduct
ivity than of the fracture zone
, 4) i
nsignificant chemical impact on contacting clay
seals
. They are fulfilled by
using
talc
-
based
concrete
with a density of
2100 kg/m
3

and a low
content of
very fine
-
grained low
-
pH
cement

(<
8 weight percent of the
solids),

[
24
]
.


4.

Compar
ison

of SKB
-
3V and VDH functions

4.1

Stability issues

The essential difference between the two concepts is that the stability of the deep vertical
VDH
with 800 mm diameter
becomes

acceptable by using a dense deployment mud, serving
also as drilling mud
,

while that of KBS
-
3 holes
is
insufficient

after placement of the canisters




3

The short period of time of installation means that the undrained shear strength is the relevant measure of

strength

15


and cannot be improved by technical means
. The impact on the maximum radial
stress by the
mud is demonstrated by inserting the depth
-
dependent mud pressure in the classical Kir
sch
equations [6
], leading to the expression for the hoop stress:




= (

H

+

h
)


2(

H




h
) cos

g

z




(
3
)

where

H

and

h

are the major and minor horizontal stresses,

the angle between the
considered plane and the plane in which the major horizontal stress is located,


the density of
the mud, g the coefficient of gravity, and z the depth.

For the same rock properties as f
or the KS
-
3V and taking the uniaxial compressive
strength according to SKB, i.e. 150
-
350 MPa, a
very significant fraction of

2
-
4 km deep part
of a water
-
filled VDH
would fail while the part down to
2 km part would stay largely intact
since the hoop stress would not exceed
about
1
00 MPa here. At the bottom of the
4 km deep
hole the hoop stress would be

208

MPa but f
illing it with mud weighing 16
00 kg/m
3

wou
ld
reduce the hoop stress to 168 MPa at this
level and to 85

MPa at 2 km depth. One hence finds
that boring of a 4 km deep VDH without risk of spalling and fall of rock

wedges would hardly
be problem
-
free when

using water or very dilute boring mud but that a mu
d with
about
16
00
kg/m
3

density make
s

it

possible.


4.2

Sensitivity to
thermally, seismically, and tectonically generated strain


The performance
in the first thousand years after closing the repository is
mainly
controlled by
waste
-
generated heating
, and later

by

exogenic processes like
large
-
scale
tectonics in the form of earthquakes, shear displacements, and rotation of regional stress
fields.
In countries
like Sweden
the cooling phase of a 100000 year glaciation

cycle
will have
obvious temperature
and stress effects to a depth of half
a
kilometre [1]
.

The heat production of the HLW in a KBS
-
3V repository with a lateral extension of a
square kilometre and located at 450 m depth has been found to give an accumulated
maximum rise of the ground surface of a couple of decimetres after 1600 y
ears [
25
]
and
significant shearing of discontinuities of 2
nd

and 3
rd

orders. This can affect the hydraulic
performance of the repository rock and make predictions thereof very uncertain for both near
-

and far
-
fields. For
a
VDH
repository
such impact is negligible since
the heat source
s
,
i.e.
the
waste in the deployment
part
s of a

group of
VDH holes

at the respective site
,
will be at least
200 m apart.

A possible
impact of high
deviatoric stresses is propagation of subhorizontal natural
we
aknesses of 2
nd

or 3
rd

orders through a KBS
-
3V repository
,

triggered by increased pore
pre
ssure in the deposition tunnels

that can

reduce

the effective vertical rock stress.
The
pro
cess is illustrated in
Fig.

1
1

(cf.
Fig.

3)
and
can be imagined as
growth of a
large flatlying
weakness in the rock mass

causing breakage and shear failure of the affected part of the host
rock.

The total absence of this risk for the VDH concept makes it more attractive than KBS
-
3V
.


16



Fig
.
1
1

Large horizontal break formed by propagation of an embryonic
weakness
extending from a natural major discontinuity of 2
nd

or 3
rd

order (A)
,

[
7
]



Fig
.
1
2

Perspective view of a one
-
level repository representing the KBS
-
3V concept (the
part termed “ignored” is a
possible pilot or experimental
repository),
[1].


5.

Possible improvement

of the concepts

5.1

KBS
-
3V

The
major problem with this concept is the overstressing caused at the boring of a certain
number of deposition holes
and
of all the holes whe
n the thermal pulse is generated
. A

recently proposed
version

(KBS
-
3i) has

two deposition holes
with
every second one
oriented
45o
off the vertical to one side
,

and the others to the opposite

side

[
7
], the purpose being to
use the rock more effectively
while reducing

t
he hoop stress
.
Both effects are achieved by
such an arran
gement, illustrated by
Fig.

1
3
. The
extension of the zone with
critical hoop
stresses of a standard vertical deposition hole for the rock stress conditions assumed in
Section 2.2.3
will be significantly
reduced as shown by
Fig.

1
4
.

It also exhibits an
improvement
by using the
supercontainer
-
in
-
mud concept

used for VDH
, which makes
placement of the clay
-
surrounded canister

in supercontainers

much simpler.
Some further
hoop stress reduction
is
, however,

desired
for
compensating

the effect of temperature rise. It
would naturally h
ave the form of
reduction
of
the amount of HLW
.


17



Fig
.
1
3

The KBS
-
3i concept with 45
o

dip and +/
-
45
o

deviation from the tunnel axis. The
clay granulate is tentative

[
7
]



Fig
.
1
4

Plotting of von Mises stresses for sections through deposition holes normal to the
hole axis. Left: KBS
-
3V to the left and inclined hole (“KBS
-
3i”) to the right. Red zones
representing critically high stresses are much larger for t
he left than for the rig
ht [
10
].


5.2

VDH

N
o improvements of th
e

proposed VDH will be
suggested
here but a number of
recommendations
to
future investigators
are motivated
:




preparation of the boreholes
for providing
acceptable conditions with respect to
stability and placement of
supercontainers,



pre
paration of
supercontainers with canisters and
dense clay blocks
for
safe handling
at the
installation
,



preparation of mud and of means for pumping and refining it.


5.2.1

Preparation and stabilization


The matter of preparing the deep boreholes includes rinsing and cleaning and stabilization
of the parts where the holes intersect fracture zones. Where rock fragments
have come off
from them the walls are irregular and the diameter varying, which requires
restoration by
using the proposed

technique in
Fig.

1
5
.

18



Fig
.
1
5

Technique for stabilizing boreholes. Left: Borehole intersecting fracture zone,
Center: Reamed hole filled with concrete between packers, Right: Re
-
boring giving a
stabilized hole [
8
].

5.2.2

Ways
of retarding expansion of the clay blocks in supercontainers

The clay expands out through the perforation and ultimately forms a tight “skin” around
the containers
. The

rate
of expansion of clay through the perforation of the supercontainers
depends on the perforation geometry and the density of the mud into which they are brought
down. The very strong
hydrophilic nature of dense smectite clay makes it suck water from the
mud and move out
from the containers
thereby starting consolidation and
densification of the
mud. The supercontainers become surrounded by a successively widening annulus of dense
mud
which
cause
s

a successively i
ncreased penetration resistance

to insertion of them in the
mud
.
For minimizing this effect the uptake of water fro
m the mud
by
the dense clay blocks in
the supercontainers should be retarded
,

which can be made by coating them with
a mixture of
montmorillonite clay and talc

(1/3 clay and 2/3 talc)
. When smeared on the supercontainer
containing dense clay the latter
fir
st
sucks water from the
thin
coating, which dries and
temporarily
forms
thin
solid plugs in the perforation holes

[
26
]
. When the containers are
moved down in the mud
-
filled hole
,

water sucked by the dense clay has to pass through the
coatings, which delay
s

hydration and expansion of the dense clay in the
m.
After several hours
expansion starts and the talc

skin in the plugs is interrupted and disintegrated. The properties
of the finally matured clay between the
borehole
walls and the containers

is
not affect
ed by
the talc additive because the coating
only
makes up an insignificant part of it.

Another method is to saturate the dense clay blocks with water before installing them in
the supercontainers since this reduces the suction potential of the blocks and
strongly retards
their expansion and migration through the perforation

[28
]
. The ultimate density and
properties of the finally matured clay between the rock walls and the containers will be the
same as when talc
-
based coatings are used.


5.2.3

Longevity of clay

in the
dense clay blocks

An early survey of possible candidate clay materials included
montmorillonite (Al
-
smectite)
clay (MX
-
80), saponite (Mg
-
smectite), palygorskite, kaolinite, and mixed
-
layer
smectitic clay

(Friedland)
.
Various investigations have
shown that for any given density
montmorillonite has the lowest hydraulic conductivity followed by saponite, mixed
-
layer

smectite/illite/muscovite
, and palygorskite. The expandability
,

expressed as swelling pressure,
was highest for palygorskite, followed
by mon
tmorillonite, saponite, and
mixed
-
layer clay

[
7
]
.

As for the KBS
-
3V buffer clay a

long operational lifetime of the
VDH
clay
s is desired

and

s
everal studies
involving hydrothermal treatment of m
ontmorillonite
-
rich clay

that had been
saturated with 10% and 20% NaCl solutions
and then isothermally heated at
110
o
C under
closed conditions for 30 days
, or heated during saturation with such solutions
,

have
given
comprehensive information. They have
demonstrated
that the hydra
ulic conductivity and

19


swelling pressure of the hydrothermally treated clay samples were
nearly the same
as for
untreated clay

and that t
he mineral composition
was not changed
except for
very slight
neoformation of

illite
-
smectite mixed layers
phases
and

irreversible partially collapsed phases
in the 20% NaCl solution

[29]
. For the clay heated during saturation
,

Na
-
illite and

fully
contracted
stacks of smectite lamellae were identified
and Mg was concluded to have
migrated

from octahedral lattice positions

to interlamellar sites, indicating

partial dissolution.
The particle size had increased indicating growth of particles.


Quite different results

have been obtained from experiments with open systems
, like in a
real VDH,

exposed to thermal gradients
with
the maximum temperature being 95
o
C
[30
].
They demonstrate reduction of the swelling pressure by more than 50 % and tenfold increase
of the hydraulic conductivity. The

changes were explained by coagulation and cementation by
silicious precipitates.
The

fact

that
saponite was much less affected by the hydrothermal
treatment
makes it
a possible
clay candidate for the dense blocks in VDH.


In conclusion, several
clay materials can

be considered for preparing the dense clay blocks
in the supercontainers. While
montmorillonite
is presently
taken as a number one candidate
the
chemical

stability of saponite
and the expandability
of palygorskite
make
them
competitor
s
.
Further hydrothermal testing
under open conditions
are required before
a

decision is taken.



5.3

Both concepts

5.3.1

Waste containers


The Swedish organization (SKB) that is responsible for deposition of radioactive waste,
and corresponding organizations in Finland (POSIVA) and Canada (OPG) rely on corrosion
-
resistant canisters for the spent fuel, while less durable
metal
canisters represent an option for
several other organizations. The present Swedish canister concept
appears in
Fig.

1
6
, which
also shows an alternative in the form of the “HIPOW” canister that is superior
with respect to

mechanical performance a
nd chemica
l integrity [7
]. It would make disposal of HLW
absolutely safe but is presently not a primary candidate for Sweden mainly
for
cost

reasons
.




Fig
.
1
6

Copper canisters. Left: version with insert of cast iron containing space for spent
fuel (SKB). Right: canisters of 100 % copper with simulated fuel rods by hot isostatic
comp
ression (HIPOW) technique [7
]


20


For VDH, which has the stagnant
deep
-
water as maj
or waste isolation component, simpler
and cheaper canisters and supercontainers can be considered. Thus, iron or steel may be
acceptable and would also make it possible to use these metals for preparing the casing. The
oxygen
-
free water in the deployment z
one would make the corrosion of such canisters
very
slow and the larger volume of the corrosion products, mainly hydroxides, than of the installed
canisters

and containers

would densify the mud

and make it less permeable. Cementation
would, however, make it less ductile.



5.3.2

Clay materials


Regardless of the type of clay mineral that will be selected for preparing the big clay
blocks for KBS
-
3 and VDH they have to be prepared by compaction of granulated
clay
material under high pressure. For KBS
-
3 the density is critical, the finally matured
blocks
must not have higher density than 2100 kg/m
3

or lo
w
er than 1900 kg/m
3

for avoiding damage
of the canisters or
causing
insufficient bearing capacity. For VDH the density should be as
high as possible for compensating the loss in density caused by the larger expansion. This can
be achieved by selecting a very low water content, suitably below 6 % by weight.


5.3.3

Placement of w
aste and clay



The impractical technique for installing clay blocks according to the KBS
-
3V concept can
lead to poor quality respecting the straightness and verticality of the stacked blocks
. They

will cause risks for getting the canisters stuck at lower
ing into the tight space in the stack

of

clay blocks
, which has to be made remotely (
Fig.

1
7
).


Fig
.
1
7

Left:
Buffer block with 1.85 m diameter and 0.4 m height weighing about 2000 kg
ready for installation in vertical
deposition hole. Right: Cani
ster installation
[1
,7
].

Filling VDH holes with clay mud or replacing it with water or other muds, all under
constant water pressure conditions, is made in the way that has been practiced by the oil and
gas companies for many years. Placement of the superc
ontainers is made by use of big drill
rigs with a capacity of handling heavy objects and with resources of installing or pulling up
the containers, and with constant access to equipment for driving them down, like
“Vibroflotation” if required. Shielding of

waste
-
containing supercontainers is required. A
number of pretests using dummies of relevant size and weight are foreseen.


21


6.

Conclusions


T
he KBS
-
3V and the VDH concepts
both
require
three well
operating
barriers

for
isolating the biosphere from radionuclides
, i.e. suitable

host rock,
tight
canisters
,

and
waste
-
embedding
clay
,

but to different extents
. For the firstmentioned the mechanical behaviour of
the rock is most important while for the latter

it
it’s the
hydraulic performance
. As to the
canisters they are the most important waste
-
isolating barrier for KBS
-
3V but of less
importance for
VDH because of the
radionuclide
-
confining role

of the salt
water

at depth
.

The
waste
-
embedding clay, finally, is strongly n
eeded for preserving the KBS
-
3V canisters by
being
tight and
ductile, while it plays a minor role for the VDH. The backfilled deposition
tunnels in a KBS
-
3V repository provide very limited hindr
ance of possibly released
radion
uclides to move to the biosphe
re while the very long, dense clay
columns
in the sealed
parts of VDH effectively
prevent possibly released radionuclides to reach up to the biosphere.

Comparison of each function of KBS
-
3V and VDH shows that the latter concept is
superior but that certain issues still remain to be worked on for accepting it as number one
candidate. The major positive and negative properties are listed in Table 3.


Table

3 Assessment of properties and function

for the same amount of HLW
.

Prope
rt
y
/function

KBS
-
3V


VDH

Note

Constructability

Good

Good

VDH
needs study

of casing

Rock stability

Insufficient

Good

VDH needs study of mud performance

Installation of waste

Difficult

Simple

KBS
-
3V needs improvement
, VDH needs
validation

Isolation of possibly
released radionuclides
from
the
biosphere

Poor

Good

Possibly c
ontaminated groundwater moves up
in
KBS
-
3V
host
rock but
stays at depth in
VDH host rock

Time for
construction

>20

years

<2
0 years

Parallel construction of several VDHs at each
site reduce
s

the time

further

Estimated cost

Very high

High


Waste retrievability

D
ifficult

Possible

Very long time required for KBS
-
3V

Long
-
term performance

Uncertain

Good



One concludes from the comparison of the concepts
that
KBS
-
3V, representing various
versions of reposit
ories located at medium depth,
and VDH
,

rep
resenting HLW placement at
very

large depths, that both are feasible but that VDH has the potential of being a
primary
repository concept for crystalline rock.


7.

References


[1] Svemar, C.,

2005.
Cl
uster Repository Project (CROP),

Final Report of European
Commi
ssion Contract
FIR1
-
CT
-
2000
-
200
3
, Brussels, Belgium.

[2] Yong, R. N., Pusch, R., Nakano, M.,

2010.

Containment of High
-
Level Radioactive and
Hazardous Solid Wastes with Clay Barriers
, Spon Press, London & New York.

[3] UK Nirex Ltd, 2004. A review of the Deep Borehole D
isposal Concept for Radioactive
Waste.
NirexReport no. N/108.

22


[4]
Grundfelt,
B.,

2010.

Jämförelse mellan KBS
-
3
-
metoden och deponering i djupa borrhål för
slutligt omhändertagande av använt kärnbränsle, Swedish Nuclear Fuel and Waste
Management Co (SKB),

R
-
10
-
13,

Stockholm.

[5
] Pusch, R, Ram
qvist, G, Knutsson, S,
H
atem, M
. The concept of highly radioactive waste
(HLW) disposal in very deep boreholes in a new perspective.
Journal of Earth Sciences and
Geotechn
ical Engineering, Vol. 2, no. 3

(
2012
) pp

1
-
24.

[6
] Pusch, R., 1995. Rock Mechanics on a Geological Base, Developments in Geotechnical
Engineering, 77.
Elsevier Publ. Co., Amsterdam, (p.
158
).

[7
] Pusch, R.,

2008.

Geological storage of radioactive waste.
Spri
nger
-
Verlag, Berlin,
Heidelberg
.

[8] Pusch, R., Ramqvist, G..
Bockgård, N., Ekman, L., 2007.
Sealing of investigation
boreholes. Final Report of Phase 4, SKB Technical Report (In print), Swedish Nuclear Fuel
and Waste Managem
ent Co (SKB), Stockholm.

[9
] Pusch, R., Practical visualization
of rock structure. Eng. Geol., Vol.49
,

Årtal
,

pp.231
-
236.

[10] Pusch, R., Weston, R.,

2011.

Superior techniques for disposal of highly radioactive waste

(HLW).

[11] Martin, D.,

1995.

Brittle Rock Strength Failure: Laboratory and In Situ, 8th ISRM
Congress, September

25
-

29, Tokyo, Japan.

[12
] Zoback, M.D., Apel, C., Baumgärtner, J., Brudy, M., Emmermann, R., Engeser, B.,
Fuchs, K., Kessele, W., Rischmuell
er, H., Rummel, F., Vernik, L
. Upper crustal strength
inferred from stress measurements to 6 km depth in the KTB borehole
, Nature
, 1993
.

[13
] Tadokoro, K., Structure and physical properties of fracture zone derived from seismic
observations at the Nojima fault and the western Tottori

eartquake fault, Japan. Bull.
Earthquake Res. Inst. Univ. Tokyo, Vol.78

(2003)

pp.67
-
74.

[14
] Hickman, S., Zoback, M., Stress orientations and magnitudes in the SAFOD pilot hole.
Geophys. Res. Letters, Vol. 31

(
2004
)
.

[15] Pusch, R., Karnland, O.,
Lajudie, A., Decarreau, A.,

1993.

MX
-
80 exposed to high
temperatures and gamma radiation. SKB Technical
Report TR
-
93
-
03. SKB, Stockholm.

[16]

Herbert, H.
-
J., Kasbohm, J., Sprenger, H., Fernández, A.M., Reichelt, C.
,

Swelling
pressures of MX
-
80 bentonite in solutions of different ionic strength, Physics and Chemis
try
of the Earth, vol. 33, (2008)

pp. 327
-
342.

[17] Kolaříková,
I.,
Švandová,
J.,
Přikryl,
R.,
Vinšová,
H.,
Jedináková
-
Křížová,
V.,
Zeman,
J.
,
Mineralogical
changes in bentonite barrier within Mock
-
Up
-
CZ experiment,

Applied Clay
Science, vol. 47 (
2010
)

pp.10
-
15.

23


[
18] Pusch, R., Yong, R.N., Nakano, M., 2012.
High
-
level Radioactive Waste Disposal, A
Global Challenge. WIT Press, Southampton, Boston.

[19]

Pusch, R., Yong, R.N., Nakano, M., 2010.
Stiffening of smectite buffer by hydrothermal
effects. Eng
ineering Geology, Vol. 110

(2010)
pp.
21
-
31.

[20
] Grindrod, P., Takase, H.,
Reactive chemical transport within engineered barriers.

In:
Proc. 4
th

Int. Conf. on the Chemistry and Migration Behaviour of Actinides and Fission
Products in the Geosphere, Charleston, SC USA, 12
-
17 Dec.
Oldenburg Verlag 1994 (pp.
773
-
779).

[21] Dixon, D.A., Graham, J., Gray, M.N.,
Hydraulic conductivity of clays in confin
ed tests
under low hydraulic gradients. Can. Geotechnical Journal, Vol. 36

(1999)

(pp. 815
-
825).

[22] Anderson, D
.M., Pusch, R. Penner, E.,

Physical
and T
her
mal Properties of Frozen
Ground in:

Geotechnical Engineering for cold regions., McGraw
-
Hill, UK.

(1978)

(pp.44
-
49).

[23]

Meyerhof, G.G., Bearing capacity and settlement of foun
dations
, ASCE, JGED, Vol.102

(1976)

(pp.195
-
228).


[24]

Pusch, R., Warr, L., Grathoff, G., Pourbakhtiar, A., Knutsson, S.,
Mohammed, M. H.,
“A
talc
-
based cement
-
poor concrete f
or sealing boreholes in rock”,
accepted

journal of
Engineering, January
, 2013.

[25
] Pusch, R. Waste Disposal in Rock, Developments in Geotechnical

Engineering. Elsevier
Publ. Co
.

[26
]
Pusch, R., A technique to delay hydration and maturation of borehole
seals of expansive


clay. Engineering Geology, Vol. 121

(2011)

pp.1
-
6.

[28
]
Forsberg, T., Pusch, R., Knutsson, S. 2013.
Rational method for preparing dense clay
blocks with desired degree of water saturation. Engineering Geology, Technical Note (In
preparation).

[29
]

Kasbohm,
J.,

Pusch,
R.,

Nguyen,
T.L, Hoang, T.M.T.,
“Labscale performance of selected
expandable clays under HLW repository near
-
field view”. Accepted for publication in
Environmental Earth Sciences (2085), DOI: 10.1007/s12665
-
012
-
2085
-
1
, 2012.

[30] Xiaodong, L. Prikryl, R., Pusch, R., ”THMC
-
testing of three expandable clays of
potential use in HLW repositories”, Applied Clay

Science, vol.52

(
2011
)

pp. 419
-
427.