MASTER'S THESIS
Evolution of Clay Seals in Deep Boreholes
Case with compacted smectite-rich clay in air-dry and prewetted forms contained in perforated tubes.
Theoretical modeling and experimental study
Gudeta Hassen Dikelie
Master of Science (120 credits) Civil Engineering
Luleå University of Technology
Department of Civil, Environmental and Natural Resources Engineering
1
L
Evolution of clay seals in deep boreholes – Case with compacted smectite-rich clay in air-dry and prewetted
forms contained in perforated tubes
_________________________________________
Theoretical modeling and experimental study
GUDETA HASSEN
2
Acknowledgements
Sincere thanks go to the Almighty Allah for making this thesis a success. I would like to show my deepest gratitude to my supervisor professor Roland Pusch for the continuous support and guidance that he has offered me, and for providing support when I required motivation and encouragement. I would also like to thank my professor Sven Knutsson for facilitating me in acquiring a suitable master’s thesis project that best suited my area of interest. I would like to thank research engineers Thomas Forsberg and Kerstin Pousette for giving me assistance during the laboratory investigation of the study and for their invaluable experience, practical ideas and help.
Finally, I would like to dedicate all my life achievements, including this masters study, to my
family.
3
Sammanfattning
Fast, högaktivt radioaktivt avfall (HLW) och använt kärnbränsle isoleras från den omgivande miljön genom att förvaras i slutförvar under jord. Borrhål från markytan ned i förvarsberget måste tätas genom att installera pluggar av lera/betong för att förhindra snabb transport av möjligen frigjorda radionuklider från förvaret. Lerpluggar placeras där berget är sprickfattigt och lågpermeabelt medan betong gjuts där hålen genomsätts av sprickzoner. Risken för instabila förhållanden i deras tidiga mognadsskeden måste beaktas. För betongtätningarna kan erosion före stelnande, och förlust av cement i senare skeden, ge försvagning. För lertätningarna innebär dispergering och kanalbildning i det tidiga mognadsskedet den största risken.
Avhandlingen handlar om mognaden hos lertätningar i djupa borrhål i första hand genom att bestämma med vilken hastighet som lerpluggar av standardttyp, dvs bestående av pressade block av lufttorra repektive hundraprocentigt vattenmättade lergranuler, homogeniseras. I andra hand undersöktes risken för “piping” och erosion av den initiellt mycket lösa lergel som bildas mellan hålväggen och det perforerade rör som innehåller lerblocken.
En tredje uppgift var att undersöka den kemiska interaktionen mellan ler- och betongtätningar gjutna i borrhålen.
En fjärde uppgift vara att utarbeta konceptuella modeler för lertätningarnas mognad och göra ett försök att utveckla teoretiska fysikaliska modeller av mognadsprocessen.
Slutsatsen av studien var att från början vattenmättade lerblock gav långsammare bildning av täta lerpluggar än block bestående av lufttorra block, samt att den initiellt bildade lergelen mellan hålvägg och perforerat rör innehållande block med 1050kg/m
3densitet undergår piping och erosion vid en hydraulisk gradient lägre än 10 m/m (meter vattenpelare per meter srömningslängd).
Slutligen ges rekommendationen i avhandlingen att följande ytterligare undersökningar bör
göras: 1) Långtidsfunktionen hos tätningarna, samt 2) Utveckling av funktionsanalysen till att
omfatta samtliga relevanta faktorer, extern påverkan och yttre och inre processer, samt att
genomföra en mera detaljerad modellering av funktionen hos enskilda och samverkande
borrhål.
4
Summary
Solid high level radioactive waste (HLW) and spent nuclear fuel are isolated from the environment by being stored in underground repositories. Boreholes from the ground surface into the host rock need to be sealed by installing clay/concrete plugs for preventing quick migration of possibly released radionuclides from the repository. Clay seals are installed where the rock is poor in fractures and has a low hydraulic conductivity, while concrete will be cast where the boreholes are intersected by fracture zones . The risk of unstable conditions in the early stage of their maturation should be taken into consideration. For the concrete seals, erosion before stiffening, and dissolution and loss of the cement component at later stages can weaken them. For the clay, dispersion and channeling of the early formed clay seals are the greatest risks.
The thesis deals with evolution of clay seals in deep boreholes, focusing first on the determination of the rate of homogenization of clay plugs of standard type, i.e. with the clay blocks prepared by compaction of air-dry clay granules as well as granules pre-saturated to about 100 %. A second aim was to investigate the risk of piping of the clay formed between the perforated tube and the borehole walls by hydraulic gradients in the vertical direction of the borehole through flow tests for identifying critical states for piping and erosion.
A third issue was to investigate the chemical interaction of clay seals and concrete cast on them in simulated boreholes.
A fourth issue was to develop conceptual models for the maturation of clay seals and to make an attempt to work out theoretical physical models of the maturation process.
The study demonstrated, in conclusion, that the maturation rate of the prewetted compacted clay was slower than the initially compacted air-dry sample and also that for clay gel with initial density of saturation 1050kg/m
3the critical hydraulic conductivity that causes erosion/piping is below 10 m/m (meter water per meter flow length).
Finally, the report recommends that further study should concern: 1) Long-term behavior of borehole seals, and 2) Performance assessment analyses should be extended to consider a complete list of relevant features, events and processes, and to incorporate more detailed modeling of the performance of single and interacting boreholes.
Key Words, Swelling pressure , Air dry/Prewetted clay block , Smectite-rich clay , clay gel
5
Contents
1 Introduction ... 7
2 Scope of study ... 10
2.1 Background ... 10
2.2 Comprehension and intention ... 10
3 Description of studies ... 11
3.1. Experiments with precompacted clay ... 11
3.1.1 Compacted Clay ... 11
3.1.2 Clay material ... 11
3.1.3 Confining copper tube ... 13
3.1.4 Confinement ... 13
3.2. Determination of the maturation rate of air-dry clay plugs of standard type ... 14
3.2.1. Performance ... 14
3.2.2. Test Results ... 14
3.2.3. Main conclusions ... 16
3.3 Determination of the maturation rate of initially saturated clay plugs. ... 17
3.3.1 Performance ... 17
3.3.2 Main conclusions ... 21
3.4 Investigation of the risk of piping and erosion of the clay formed between the perforated tube and the confining rock ... 21
3.4.1 General ... 21
3.4.2 Results ... 22
3.5 Investigation of the chemical interaction of clay seals and concrete in simulated boreholes 23 3.5.1 Plan ... 23
3.5.2 Performance ... 23
3.5.3 Outcome ... 24
4 Development of conceptual models for the maturation of clay seals and attempts to work out theoretical physical models of the maturation process. ... 26
4.1.1 General ... 26
4.1.2 Proposed model of formation of a clay gel between perforated tube and rock ... 26
5 Conclusions ... 31
6 References ... 32
7 Appendix ... 34
6
Legend and abbreviations
HLW High level radioactive waste ρ Bulk Density , (kg/m3)
ρd Dry density, (kg/m3)
ρsat Degree at saturation, (kg/m3)
w water content (%)
7
1 . Introduction
The present study is part of a more comprehensive investigation of the performance of clay/concrete plugs installed in deep boreholes for preventing quick migration of possibly released radionuclides from a repository to the ground surface along the holes (Figure 1).
Figure 1. Short-circuiting of fracture zones by open or inadequately sealed or leaching boreholes.
The principle of borehole sealing proposed and planned to be used by the Swedish Nuclear Fuel and Waste Management AB (SKB) was worked out early in the present century. It implies tight sealing of the parts of boreholes where the rock has few fractures and a low hydraulic conductivity, and filling the parts that intersect water-bearing fracture zones with physically stable material that does not need to be very low-permeable (Pusch & Ramqvist, 2007). The tight seals consist of smectite-rich clay in the form of highly compacted blocks, while the fillings separating them are proposed to be made of concrete cast on site (Figure 2).
GENERAL
Teknikrådet 2007-09-27 Ground surface
A B
C Zone II
Zone I
Zone III
If poorly plugged between theZones, Boreholes A, B and C form paths ( ) for radionuclides via all three Zones and directly to the ground surface via hole B.
Thick black=Clay plugged Thin black=Quartz/cement
Early structural/hydraulic modelling after flow meas.
in A and B while open
Repository
8
Figure 2. Schematic picture of borehole seals consisting of very tight clay where the rock is low-permeable and concrete where fracture zones are intersected.
Figure 3. shows the seal consisting of clay in a perforated copper tube. This metal is selected because of its corrosion resistance. The tubes to be inserted in a borehole are filled with well fitting highly compacted clay. The clay expands through the perforation by taking up water from the water-filled hole and creates a clay gel in the gap between the tube and the borehole wall. Clay continues to enter the gap by which the density is successively increased. The sealing potential increases and the seal is soon tighter than the surrounding rock.
Figure 3. The clay seal. The left picture shows jointed perforated copper tubes with tightly
fitting clay blocks inside. The detail shows how the plug units can be connected by
pressurizing (Pusch, 1983).
9
Figure 4 demonstrates the first stage of maturation of the compacted clay, which is smectite- rich and highly expansive. After a few hours a clay gel is formed between the borehole wall and the tube and after one or a few days the gel becomes coherent, tight and erosion-resistant.
This process is quick where the surrounding rock gives off sufficiently much water but can be slow where the rock is very tight.
Figure 4. Lab experiment showing formation of soft clay gel after 8 hours by migration of clay from the dense clay blocks through the perforation of a copper tube in an 80 mm diameter oedometer. The larger part of the clay block is still unaffected by water (Pusch &
Ramqvist, 2007).
10
2 . Scope of study 2.1 Background
The principle of sealing boreholes with intermittently installed clay and concrete plugs has been described in the literature and successfully applied in practice. The maturation process leading to homogenization and strengthening of the clay is demonstrated by the successively higher force required to move the plug axially in the hole (Pusch & Ramqvist, 2007, 2011).
Laboratory experiments have indicated that the erosion caused by moving the clay plugs on site in more than 500 m deep boreholes may reduce the clay mass in the plugs too much and that a technique for delaying the expansion is needed. This matter was investigated in the present study.
It is estimated that hydraulic gradients caused by water pressure differences in the vertical direction of deep boreholes may cause piping and erosion of the maturing clay between the perforated tube and the borehole wall. This matter was also investigated in the study.
Field experiments have shown that smectite clay contacting concrete with Portland cement and Glenium as superplasticizer causes degradation of both to a few centimetres depth in a about three years. This finding led to the desire of checking the chemical interaction between smectite-rich clay and concrete based on low-pH cement (Merit 5000). The matter was investigated in the study.
2.2 Comprehension and intention
The present investigation comprised four parts:
1. Determination of the maturation rate of clay plugs of standard type, i.e. with the clay blocks prepared by compaction of air-dry clay granules, and with clay pre-saturated to about 100 %. The latter was expected to give slower expansion and hence less risk of being exposed to erosion in real borehole sealing projects,
2. Investigation of the risk of piping of the clay formed between the perforated tube and the borehole walls by hydraulic gradients in the vertical direction of the borehole, 3. Investigation of the interaction of clay seals and concrete cast on them in simulated
boreholes,
4. Development of conceptual models for the maturation of clay seals and attempts to
work out theoretical physical models of the maturation process.
11
3 . Description of studies
3.1. Experiments with precompacted clay
3.1.1 Compacted Clay
Compacted clays are commonly proposed as primary sealing materials for nuclear waste repositories and have been extensively investigated against rigorous performance requirements (e.g., Van Geet 2007). Advantages of clays for sealing purposes include: low permeability, demonstrated longevity in many types of natural environments, deformability, sorptive capacity, and demonstrated successful utilization in practice for a variety of sealing purposes. Compacted clay as a borehole sealing component functions as a barrier to water flow and radionuclide movement and possibly to gas flow is one of the applications.
The exact specification for compacted clays used in borehole sealing will depend upon site- specific details such as water chemistry, but an extensive experimental data base exists for the permeability of a variety of bentonite clays under a variety of conditions. Bentonite clay, a highly plastic swelling clay material dominated by smectite clay minerals (Mitchell 1993), is chosen here because of its positive sealing characteristics. Compacted clay of this type can generate a substantial swelling pressure and seal any space that it is placed in. In boreholes it expands and fills them, ensuring conformance between the clay seal component and the borehole walls.
Smectitic clays have been widely used in field and laboratory experiments concerned with radioactive waste disposal. Verification of engineering properties such as density, moisture content, permeability, or strength of compacted clay seals can be determined by direct and indirect measurement during construction (Pusch, 1994; Pusch, Yong, Nakano, 2011).
Figure 5. Air dry compacted clay.
3.1.2 Clay material
The clay material used in the present study was “Saline seal 100”, a Polish clay with a
smectite (montmorillonite) content of about 90% by weight of the total mineral mass
according to the manufacturer. The clay material had an initial water content of 7.7 percent by
12
weight. The blocks prepared by compacting air-dry clay granules had a dry density (ratio of mineral mass dried at 105
oC) of 1600 kg/m
3, which corresponds to a density of complete water saturation of 2008 kg/m
3. Blocks prepared to be fully water saturated had the same dry density and density at complete water saturation. The swelling pressure can be estimated from the diagram in Figure 6. Figure 7 gives the hydraulic conductivity for testing under a hydraulic gradient of 30-50.
Figure 6. Typical relationship between swelling pressure and density of smectite-rich clay.
The curve starting at “origo” represents Na montmorillonite (MX-80) saturated with low- electrolyte water and the other Ca montmorillonite saturated with low-electrolyte water.
Typical values for lower densities of Na montmorillonite are: 100 kPa for 1400 kg/m
3, 20 kPa for 1200 kg/m
3and 8 kPa for 1100 kg/m
3density at water saturation.
Density at saturation with distilled water and 3.5 % CaCl
2solution, respectively
1500 1600 1700 1800 1900
3.0
2.5
2.0
1.5
1.0
0.5
0
MX-80 Swelling pressure
2000 3.5
Swelling pressure, MPa
3.5% CaCl2 solution Distilled
water
4.0
13
Figure 7. Typical relationship between density at water saturation and hydraulic conductivity of smectite-rich clay. Measurements made under hydraulic gradients of 30-50.
3.1.3 Confining copper tube
The copper tube had a perforation ratio of 50 %, an outer diameter of 88 mm and a wall thickness of 2 mm. The diameter of the perforation holes was 10 mm.
3.1.4 Confinement
The perforated tube was placed in an outer plexiglass tube with 100 mm inner diameter. The average gap between the plexiglass tube and the perforated tube was 12mm. The perforated tube was prevented from being lifted by water and swelling pressures by a stiff yoke (Figure 8).
E-13 E-12 E-11 E-10
1600 1700 1800 1900 2000
Density at water saturation, kg/m
3K, m/s Aqua dest
3.5 % CaCl
214 Figure 8. Test arrangement.
3.2. Determination of the maturation rate of air-dry clay plugs of standard type
3.2.1. Performance
The clay sample was prepared by uniaxial compression of air-dry clay granules with a water content of 7.7 % by weight. The 86 mm diameter block was inserted in the perforated tube that was placed in the tap-water filled plexiglass cylinder with 100 mm inner diameter.
3.2.2. Test Results
Figure 9 illustrates the successive migration of clay from the dense clay block through the perforation of the tube made of copper.
a) After 10 minutes from start b) After 60 minutes
15
c) After 2 hours d) After 3 hours
e) After 24 hours
After 5 days
Figure 9) a,b,c,d,e. Rates of maturation of compacted smectite-rich clay in air-dry form as
viewed from ouside the confining plexiglass tube.
16
After 5 days the content of the plexiglass cylinder was pushed out and samples taken for determination of the water content and density distribution (Figure 10).
Figure 10. Upper: Distribution of water content and dry density after 96 hours. Lower:
Distribution of dry density after 96 hours.
3.2.3. Main conclusions
The test showed that the maturation after 5 days had led to an average dry density of the clay between the perforated tube and the plexiglass cylinder of about 496.3kg/m
3and that the dry
0 5 10 15 20 25 30 35 40
0 50 100
sorbed water vs time
sorbed water vs time
sor bed wa ter (m l)
Time in hr
0 0,5 1 1,5 2 2,5
0 50 100 150
density vs sorbed water graph
density vs sorbed water graph
de nsit y in g /cm3
Sorbed water with time in cm3
17
density within 10 mm distance inside the perforated tube had dropped from initially 1600 kg/m
3to 1232 kg/m
3. No or very slight changes were found in the clay block for larger distance from the perforated tube than 10 mm.
3.3 Determination of the maturation rate of initially saturated clay plugs.
3.3.1 Performance
The samples were prepared in different ways for testing how homogeneous wetting can be achieved. A first attempt involved placement of air-dry granules at the bottom of a stiff steel tube and evacuating the interior through a piston used for compacting the granular fill, while letting water in from the base. This technique, proposed by Dr Leo Christensen, technical director at Lolland community, Denmark, gave dry densities of up to 1750 kg/m
3but the wetting, which was complete only for the lowest 1-2 cm part, decreased upwards and left half of the fill with a very low water content (4.4%), cf. Figure 11. The technique can probably be improved but this was out of the scope of the present study.
A new technique, being patented, was then applied, implying mixing the granular fill with very fine droplets of water so that uniform wetting was reached. The clay granules had a water content of 7.7 % by weight and adding water gave a water content of 25.4 % and a dry density of 1600 kg/m
3in the compaction state, requiring a compressive force of about 60 kN (density at saturation 2008 kg/m
3). The 86 mm diameter block was inserted in the perforated tube that was placed in the tap-water filled plexiglass cylinder with 100 mm inner diameter.
The maturation and successive migration of clay from the compact clay block into the gap between the perforated tube and the plexiglass cylinder that represented rock, is illustrated by the series of pictures in Figure 12.
Figure 11. Appearance of sample of dense smectite clay, originally air-dry, after compression
with concomittant hydration. Only the lowermost part became largely wetted.
18 i) At start
ii) After 10 minutes iii) After 60 minutes
iv) After 2 hours v) After 3 hours
19
vi) After 24 hours
Figure 12. Photos illustrating the successive formation of a clay gel between the perforated tube and the plexiglass cylinder.
After 5 days the content of the plexiglass cylinder was pushed out and samples were taken for determination of the water content and density distributions using the graphs in Figure 13.
The results are shown in Figure 14.
Figure 13. Bulk density as a function of the water content at different degrees of water
saturation for the mineral density ρ
s= 2700 kg/m
3.
20
Figure 14. Distribution of water content and dry density after 96 hours.
0 5 10 15 20 25 30 35 40
0 20 40 60 80 100
sorbed water vs time
sorbed water vs time
sor bed wa ter (m l)
Time in hr
0 0,5 1 1,5 2 2,5
0 50 100 150
density vs sorbed water graph
density vs sorbed water graph
de nsit y in kg /cm3
Sorbed water with time in cm3
21 3.3.2 Main conclusions
The test showed that the maturation after 5 days had led to an average dry density of the clay between the perforated tube and the plexiglass cylinder of about 496.3 kg/m3 and that the density within 10 mm distance inside the perforated tube had dropped from initially 1600 kg/m3 to 1232 kg/m
3. The innermost core had retained its density. The maturation rate of the clay had hence proceeded slower than for the initially air-dry sample.
The slower formation of clay outside the tube must have been caused by a lower driving force for the case with initially water saturated clay, Here, this force was caused only by the difference in concentration of the clay diffusive migration of clay, while for the initially air- dry clay block this force was the sum of diffusive migration of clay and the expansion caused by the air pressure in the voids, which made up about 50 % of the total volume of the block.
3.4 Investigation of the risk of piping and erosion of the clay formed between the perforated tube and the confining rock
3.4.1 General
The hydraulic gradient between different levels in the rock surrounding a sealed borehole may amount to several meter water height and can theoretically generate piping and erosion (channeling) in the initially soft clay formed between the perforated tube and the confining rock. This was studied in a pilot project in which clay gels representing different stages of maturation of the ultimately dense clay were exposed to different hydraulic gradients.
The clay formed between the perforated tube containing clay in boreholes and the borehole walls is very soft in the earliest hours and if sufficiently high hydraulic gradients in the axial direction of the holes exist or are generated there can be breakthrough (piping) and erosion.
The critical gradient is in the range of 30-50 according to earlier investigations (Pusch et al, 1986). When the density has increased to about 1400 kg/m
3the risk is believed to be insignificant.
The experimental study was made by preparing clay of water-saturated smctite-rich clay confined in glass tubes that are exposed to a hydraulic gradient i=(H
1–H
2)/L where H
1and H
2are water heads at each end of the tube in meters, and L the sample length in meters (Figure 15).
The gradient was increased in steps from 5 to 10 to 20 to 30 m/m etc until piping took place.
The densities of the clays were 1050, 1100, 1150, 1200, 1250, 1300, 1350 and 1400 kg/m
3.
22
H2 H1
L
Figure 15. Test arrangement for determining the critical hydraulic gradient.
3.4.2 Results
The experiments comprised determination of the hydraulic conductivity of the clay for two densities, the outcome being shown in Figures 16 and 17, using Darcy’s law (Eq.1).The risk of piping can be evaluated from these tests by recording whether the evaluated conductivity tended to increase with time or not. Steady flow would correspond to stable microstructural conditions.
v=Kxi (1) where v=flow rate
i= hydraulic gradient (m/m)
Figure 16. Evaluated hydraulic conductivity of clay with initial saturated density of 1100 kg/m
3for a successively increased hydraulic gradient from 5 to 30 m/m. The conductivity was nearly constant and reached the value 3.5E-6 m/s, indicating no piping.
1E-07 1E-06 1E-05 0,0001
09:36 12:00 14:24 16:48
P er meabilit et [m/s]
time
23
Figure 17. Evaluated hydraulic conductivity of clay with initial saturated density of 1200 kg/m
3for a successively increased hydraulic gradient from 0 to 100 m/m. The conductivity was as high as for the softer clay and tended to increase, indicating a tendency of piping.
3.4.3 Main conclusion
Measurement of the hydraulic conductivity of the clay paste formed between tube and confining tube, representing the condition of successive maturation, indicated that the hydraulic gradient at which piping initiatesm, is in the range between 30 to 50 m/m.
3.5 Investigation of the chemical interaction of clay seals and concrete in simulated boreholes
3.5.1 Plan
Preceding experiments with smectite clay of MX-80 type in contact with on-site cast concrete with Portland cement and organic superplasticizer Glenium 51 had indicated dissolution of both the concrete and clay to a distance of a couple of centimetres from the contact in a three year field experiment in boreholes in crystalline rock. The present study comprised a pilot study of a recently developed cement-poor concrete at the geotechnical department of LTH.
3.5.2 Performance
Concrete of a new type
1with Portland cement and talc was cast in ordinary 50 mm oedometers to about 10 mm height and clay filled on the concrete and compacted to a density of 1986 kg/m
3after 2 days of rest (Figure 18). After 3 weeks, representing the chemically most important period for both materials, the cells were opened and thin slices cut for
1
The concrete, which has quartzite as major aggregate component contains about 6 % cement and a superplasticizer of silica-rich minerals, was worked out in the course of the current R&D work at the department.
1E-07 1E-06 1E-05 0,0001
09:36 12:00 14:24 16:48
P er meabilit et [m/s]
time
24
chemical and mineralogical analyses as well as for physical investigations. The firstmentioned could not be made in time for the project while determination of the hydraulic conductivity could be performed. It was found to be 1.3E-11 m/s, which is about 100 times higher than the conductivity of virgin Na-smectite (cf. Figure 7), hence indicating considerable microstructural reorganization associated with ion exchange from Na to Ca provided by the cement.
Figure 18. Test arrangement for investigating the interaction of contacting dense smectite clay and talc concrete with 6 % Portland cement.
3.5.3 Outcome
Although the analyses with respect to mineral changes could not be performed in time to be reported in this document the appearance of the sections cut for analysis were virtually unaffected (Figure 19). One can therefore assume that the mutual impact of smectite clay and concrete was the same as in a preceding experiment from which the following conclusions were drawn Pusch & Ramqvist, 2011:
Significant chemically induced changes in mineralogy and physical performance will occur within a few millimeters in laboratory experiments running for 3 weeks and to a few centimetres distance from the clay/concrete contact in 3 years under field conditions, indicating that the processes are of diffusive type,
Laboratory studies including X-ray diffraction (XRD), X-ray fluorescence (XRF) and electron microscopy study (SEM and TEM) and also dual beam (combined ion and electron) microscopy show that clay can infiltrate the contacting concrete plug, since clay has been detected both along the contact between the plug and the borehole wall as well as interleaved within the concrete fill,
The observed increase of the hydraulic conductivity by about two orders of magnitude
of the clay can be ascribed to cation exchange from originally sodium to calcium,
causing coagulation of the softest parts in the clay, leading to widened voids,
25
Chemical alteration of the cement mineral phases at the clay-concrete contact releases some quantities of Ca and K that had partly replaced Na in the interlayer (interlamellar) space of the clay. Precipitation of gypsum and halite occurs as well as chemical modification of the clay. The cement will be altered in contact with the clay plug and lose part of its mechanical strength. Neocrystallization of a fibrous Ca-Si phase will occur along with the formation of some amorphous components.
Dissolution in the saline water is the probable major mechanism.
Figure 19. Clay (upper) and cement sample before slicing, 3 weeks after preparation. No
visible changes in colour or structure.
26
4 . Development of conceptual models for the maturation of clay seals and attempts to work out theoretical physical models of the maturation process.
4.1.1 General
The experiments for determining the rate of maturation of the clay seals deliberately ran for a short time, because only the first few days or weeks are of interest for finding out which of the two ways of preparing clay blocks is preferable, and what the major processes are in the formation of an ultimately homogeneous clay seal.
The dominant process was the early release of clay material through the perforation of the tube and formation of a clay gel in the gap between tube and confinement. Both clay types gave similar rates of maturation as concluded from watching the confining plexiglass cylinder but the analysis of the samples taken at the extraction and dissection of the various parts of the clay seals showed that the prewetted clay matured more slowly than the initially unsaturated clay. This is advantageous for practical borehole sealing since the risk of losing clay by erosion in the course of the installation of deeply located clay seals can be reduced by applying the latter technique. A further improvement, that will be investigated, is to fill the borehole with Na-smectitic drilling mud before inserting the clay seals.
4.1.2 Proposed model of formation of a clay gel between perforated tube and rock Principle
The process of migration of clay from the dense clay core in the perforated tube and the borehole wall can be visualized as in Figure 20.
Rock Rock
Center
Rock
Center
Preparation 1st stage Final stage
Dense clay core Coating
Dry condition Installation
Center
Water pressure
The dense clay core fills the perforated tube Perfor.
tube
Water keeps coating pressed against the dense clay but percolation initiates wetting and expansion of the core
The water saturated clay plug ultimately becomes homogeneous but a little less dense than the original clay core
Maturation
Water
Air Water
saturated clay
Dense clay starts expanding
Figure 20. Steps in the evolution of the clay seals. The clay “skin” formed early in the gap
between tube and rock is initially very soft but consolidates under the pressure provided by
clay moving out through the perforation.
27 Water uptake and saturation
A proposed simplified maturation process is shown in Figure 21. It implies that water does not primarily enter the borehole in the form of uniform radial flow through the rock but through fractures intersecting the hole from which it migrates to and through the clay. The condition for this type of wetting is that a sufficiently high water pressure prevails in the fractures to make them control the hydration of the clay plug, implying that water flow through the rock crystal matrix is negliglible. This model fits the clay plug test in the present project.
Figure 21. Conceptual model of the inflow path of water from water-bearing fracture. Least resistance to the flow is through the clay “skin” around the perforated tube since it is more permeable than the dense clay core for several weeks after installation.
Processes
The model implies that the supply of water for saturation and maturation of the clay”core” in the perforated tube takes place by flow from rock fractures in the initially soft but successively denser clay “skin” that is formed by material moving out through the perforation of the tube.
Figure 22 illustrates the first phase of maturation of an initially not water saturated clay seal, involving formation of clay columns (“plugs”) growing from the dense clay core through the perforation of the tube and forming, in the first phase, an embedment of the tube of different densities. After about 24 hours the clay components in the space between the tube and the rock wall have become relatively homogeneous but it takes weeks and months for them to reach a high degree of homogeneity and a density that approaches that of the successively
Dense clay core
Water-bearing fracture
Perforated tube Borehole perimeter
Clay skin formed by migration of material from the dense core. Flow from the fracture follows the route of least resistance, i.e. the clay skin Major inflow
Very small inflow through the rock crystal matrix
Major inflow
28
softening central clay core. From the point of maturation and performance they initially make up a very heterogeneous, soft and permeable “clay skin” that successively becomes denser and less permeable. The wetting and maturation of the initial not fully water-saturated clay core takes place as indicated in Figure 21.
For the case of an initially fully water saturated clay core in the perforated tube the whole maturation takes place by exchange of porewater in the core with water in the borehole, making the entire maturation process one of diffusion. The diffusion coefficient can be estimated at E-9 m
2/s to E-10 m
2/s.
Figure 22. Schematic picture of the earliest stage in the maturation of a clay seal..
One can estimate what the hydraulic conditions, particularly the magnitude of the pressure in a fracture intersecting a plugged hole, must be to provide the skin zone with sufficiently much water for not limiting the maturation rate. This can be made by using experimental laboratory data and a simple theoretical model by which one can calculate the suction of water transported from the skin zone into core and comparing it with the rate of axial inflow into the zone from an intersected fracture (Pusch & Ramqvist, 2007). The results are compiled in Table 1.
Table 1. Amount of water for unlimited hydration of a radially expanding 2.5 m long clay core as a function of time. Initial dry density 1670 kg/m
3. Strength data are included. For clay plugs confined by a perforated tube the rate of maturation is at least 50 % lower.
Time, hours
Density of “clay skin”, kg/m
3Amount of sorbed water, liters
Hydraulic conduct- ivity of “skin zone”, m/s
Shear strength of
“skin zone”, kPa
6 1100 10 E-9 50
12 1150 15 5E-10 200
24 1325 33 5E-11 400
48 1400 40 2E-11 550
96 1700 70 E-12 700
Perforated tube
Moving boundary of unaffected central clay core
Softened peripheral part of central clay core
Soft clay gel
Densest parts of clay columns(”plugs”)
Softest parts of clay columns forming the initial
”clay skin”
29
In practice, the expansion of the clay core is limited for geometrical reasons (the “gap” is thin!) meaning that the “clay skin” starts to consolidate early under the compressive force exerted by the dense clay core. This means that only the amounts of water required for maturation in the first 12 hours are relevant since consolidation will then be the dominant process. One can estimate that about 10 liters need to enter the holes in the first 6 hours and another 5 liters in the subsequent 6 hours for avoiding retardation of the maturation of the clay plugs. This demonstrates that the inflow of water through rock fractures is a determinant of the maturation of the clay skin of initially unsaturated clay cores.
For a 2.5 m long clay seal consisting of a perforated plug with very dense unsaturated clay core only about 3 liters would hence be needed for formation of the “skin” in the first 6 hours and somewhat more than 2 liters in the subsequent 6 hours. Using these data for comparison with the results from the present test with a unsaturated clay core, being 6 cm high, the required amount water for formation of a clay skin in 6 hours would be 3/25 liters, or about 120 ml. The actual volume was 90 ml, implying reasonable agreement.
Role of the rock structure
Focusing on clay seals of initially unsaturated type one estimate that the rate of migration of water into the core takes place by diffusion from the interface of skin and core
2. For 12 hours it is estimated that complete hydration of the dense core has taken place to 0.3 cm radial distance from the wet boundary, hence requiring about 5 liters of water for a 2.5 m long clay core with 80 mm diameter and 1650 kg/m
3dry density and an initial degree of water saturation of 50 %. For checking if the plug has access to this amount of water over its entire periphery the issue is to find out whether such inflow can take place and maintain complete water saturation of the gap between rock and perforated tube under the prevailing conditions.
Assuming a fracture spacing a=1 m, and considering 0.5 m long elements consisting of clay
“skin and core” in Figures 21 and 22 it is realized that water from the fracture needs to be driven into and through the successively formed clay skin by flow, which requires that the hydraulic gradient operating over the 0.5 m length and the average conductivity of the clay skin are sufficiently high to bring water all the way to the mid distance between the two neighbouring fractures. For fulfilling the condition that the flow must be sufficient to feed the dense clay core also at the opposite end of the 0.5 m long element, the flux must hence be about 1/12 liter per hour, i.e. 0.2 liters for the 2.5 m long plug. The cross section area of the
“clay skin” is on the order of 10 cm
2, meaning that the minimum flow rate must be at least 0.6 m per day. Assuming zero water pressure at the inner end of the element and taking the water pressure in the fracture as driving force, this would, for a pressure of 500 kPa (50 m water head), require a hydraulic conductivity K of the skin of K>3E-10 m/s, which is approximately that of MX-80 clay with a density of about 1100 kg/m
3, at percolation with low-electrolyte water. Taking instead the distance of neighbouring fractures to be 10 m, the minimum conductivity of the skin to serve as required would be roughly 3E-9 m/s, which would imply insufficient supply of water the clay core at mid distance. One hence finds that the rock structure is of major importance for the maturation rate of the plugs.
Quicker densification of the skin, larger fracture spacing, or lower fracture pressure would delay maturation of the core and make it heterogeneous since only the part of the skin close to the fracture would be fully mature in the early stages. Such delay is exemplified by field
2
The code ANADIFF (Trygve Eriksen, KTH) is useful for this
30
experiments in which short holes extending from a drift were equipped with clay seals (Pusch
& Ramqvist, 2007).
It is obvious that the maturation of clay seals of this type consisting of prewetted clay blocks with 100 % degree of saturation will not depend at all on the access to water from the rock.
Hence, they will mature even in extremely “dry” rock.
31
5 . Conclusions
In maturing clay seals of the investigated type, the very dense clay in perforated tubes
expands through the perforation and embeds the tubes. If the clay core is initially not water
saturated, water will be taken up by the clay from the water-filled hole and surrounding rock
by which a clay gel (“skin”) is formed in the gap between the tube and the borehole wall. Clay
continues to enter the gap by which the density is successively increased. The sealing
potential increases and the seal is soon tighter than that of the surrounding rock. If, on the
other hand, the clay core is fully water saturated from start, the only process that results in the
formation of the clay gel is coupled transport of clay from the core out though the tube and
transport of water from the borehole into the core. This is a hydraulically closed system. The
latter case is advantageous because the migration of clay through the perforation is slower and
hence reduces the risk of loss of clay from the clay seal in the installation phase, which can be
4-12 hours depending on the depth of the borehole.
32
6 . References
Newman, J., Seng Choo, B., 2003 . Advanced concrete technology , concrete properties.
Elsevier , Britain (ISBN 0 750 6 5104 0)
Paulini,P, 1993. A through solution model for volume changes of cement hydration , Pergamon , cement and concrete research , vol. 24 , no. 3 , Elsevier Scince
Powers,T.C, Copeland,L.E.,hayes J.C., Mann,H.M,1955. Permeability of Portland cement paste , journal of American concrete institute ,Bulletin 53
Powers,T.C., 1958. Structure and Physical Properties of Hardened Portland Cement Paste, , journal of American concrete institute ,Bulletin 94, VOL.41
P-Selecta , Viscometer manual, (Spain)
Pusch, R., 1983. Borehole sealing for underground waste storage. ASCE Proc. J. Geotechnical Engineering, Vol.109, No.1 (pp.113-119).
Pusch,R , 1994. Waste Disposal in Rock , Elsevier , Netherland , (ISBN 0-444-89449-7) Pusch,R, 2006. Clays and Nuclear Waste Management,eodevelopment AB, Ideon, S-22370 Lund, Sweden , Handbook of Clay Science , Developments in Clay Science, Vol. 1 , Elsevier Pusch,R, Ramqvist,G, 2007. Borehole Sealing Project, Final report of Phase 3, SKB , Sweden.
Pusch,R, 2008. Geological storage of highly radioactive waste , Springer , Sweden (978-3- 540-77332-0)
Pusch,R, Ramqvist,G, Bockgård ,N , Ekman,L, 2011. Sealing of investigation Boreholes, Phase 4 final report , SKB , Sweden.
Pusch, R., Yong,R , Nakano,M , 2011. High level radioactive waste disposal, WITpress (ISBN 978-1-84564-566-3)
Shinji Kobayashi, 2007. Design for rock grouting based on analysis of grout penetration Verification using Äspö HRL data and parameter analysis , SKB report , R-07-13
Vervoort,R.W., Cattle,S.R., 2006. Linking hydraulic conductivity and tortuosity parameters to pore space geometry and pore-size distribution , Journal of Hydrology 272 (2003) 36–49 , Elsevier
VWR company , Talc Catalogue
Warin, A., 2003. Desing of borehole seals in rock formation using expert system. Suranaree
University of Technology (ISBN 974-533-311-5)
33
Widmann,R, 1993 . Grouting in rock and concrete , Balkema , Netherland .(ISBN 90 5410 350 7)
Wildenborg,T, Lokhorst,A, 2005. Introduction on CO
2Geological Storage.Classification of Storage Options, Oil & Gas Science and Technology – Rev. IFP, Vol. 60 (2005), No. 3, pp.
513-515 , Netherland
34
7 . Appendix
1.
degree of saturation 2.
.
