Chapter 2: Radioactive waste types and repository designs

Chapter 2: Radioactive waste types and repository designs

Chapter 2: Radioactive waste types and repository des ig n s Radioactive wastes are characterised by their cycle (the part before fuel reaches the nu...

9MB Sizes 0 Downloads 2 Views

Chapter 2: Radioactive waste types and repository des ig n s Radioactive wastes are characterised by their

cycle (the part before fuel reaches the nuclear

mode of formation, physico-chemical nature and level of radioactivity. Different repository designs are generally required to contain the individual

reactor), uranium mining and milling operations generate large masses of natural material, such as displaced rock and soils etc., and discarded waste,

waste types, although some repository concepts exist to house wastes with different levels of radioactivity. The generation of the different waste types and the repository designs developed to accommodate them are discussed in this chapter.

mainly in the form of mill tailings which are the residual materials from ore processing. On

Radioactive wastes are the unwanted by-products

average, 86 000 tonnes of uranium-depleted tailings are produced every year for each reactor (US National Research Council, 1990). These wastes are not normally considered for disposal in a repository and are treated separately. However, they emit levels of radiation above average background and care is required to ensure safe management of this processed material.

of civilian nuclear power (electricity) generation

Wastes are also produced during the uranium

2.1 The nuclear fuel cycle and radioactive wastes

programmes, military nuclear weapons const- conversion and enrichment processes, and during ruction and decommissioning strategies and, to a fuel fabrication, and some of these wastes may be lesser extent, industrial, medical and scientific scheduled for disposal in a repository, depending research activities. The majority of radioactive on their levels of radioactivity. wastes from around the world are produced by The majority of radioactive wastes which will be nuclear power generation although, in countries sent for disposal in repositories are generated by with an extensive nuclear weapons programme the operation of nuclear power reactors. These such as the United States, the national inventory of wastes fall into three principal categories: weapons waste can be larger than the inventory from power generation.

used irradiated fuel and the wastes derived from its reprocessing;

The nuclear fuel cycle, shown in Figure 2.1, is the term used to encompass all activities associated

operational wastes created by the day-to-day

with the production of nuclear power. Wastes are

running of nuclear reactors; and

generated at each stage and these wastes are variously radioactive. At the front end of the fuel

9

reactor decommissioning wastes.

29

The geological disposal of radioactive wastes and natural analogues

Figure 2.1: The nuclear fuel cycle from mining of uranium orebodies to the final disposal of reactor wastes. These wastes are described below, together with some additional wastes from other sources that potentially may also be disposed of in

while others, such as Sweden, do not and are currently selecting sites for its direct disposal in deep geological repositories (see Box 1).

repositories.

Spent fuel reprocessing is a complex process that involves dissolving fuel elements using aggressive inorganic solvents and then employing a sequence 2.1.1 Used fuel and reprocessing of organic and inorganic reactants to separate wastes uranium and plutonium from the other unwanted Most nuclear reactors burn a uranium dioxide fuel radioactive products. The whole operation is (UO2) enriched in 235U that, after a few years in the performed on an industrial scale and produces reactor core, becomes poisoned with fission large volumes of wastes with a broad spectrum of products and transuranic elements produced as a physical and chemical compositions, and varying result of the nuclear reactions. These elements levels of radioactivity. The most radioactive reduce the efficiency of the fuel and, consequently, wastes are the liquids remaining after dissolution it has to be replaced periodically. This used, and separation of the uranium and plutonium. irradiated fuel, known as spent fuel, once removed These liquids must be solidified, usually in a glass from the reactor, is stored for at least 6 months to matrix, prior to final disposal. allow the substantial radioactive decay heat to Other less radioactive wastes generated from the diminish to a level which allows safe handling and reprocessing operation comprise a range of transportation, materials such as Zircaloy cladding from the fuel After cooling, the spent fuel assemblies may be stored prior to direct disposal in a repository or they may be reprocessed to extract the unburnt uranium for incorporation in new fuel elements,

assemblies, ion exchange resins used in the chemical separation process, swabs used for cleaning laboratory benches and disposable clothing. These lower activity wastes contain a

Some countries, such as the United Kingdom and France, reprocess the majority of their spent fuel

high proportion of solid, organic material and, to facilitate disposal, they are compacted and

30

Radioactive waste and repository types

immobilised in either a cement, bitumen or resin

reprocessing systems. Various decommissioning

matrix prior to disposal,

options are possible but all involve partial" or complete dismantlement of the facility and

2.1.2

associated buildings. Some of the materials generated by this process will be radioactive. For

Operational wastes

The routine operation of a nuclear power plant

example, in a nuclear power plant, the reactor

generates many different types of waste in both

vessel and the surrounding reinforced concrete

liquid and solid form. The most significant waste,

structure

both in terms of volume and activity, are ion exchange resins which are used to recover radionuclides from liquid wastes. These liquid

radioactive due to neutron irradiation by the reactor core, while other components, such as primary coolant piping, may contain residual

wastes are generated during

radioactive liquids or solids.

tamination

procedures,

routine

shield) will

become

chemical

Depending on the nature of the dismantled

processes (e.g. primary coolant and fuel storage

components and their levels of activity, some

pond clean-up) or from detergent waste (e.g. laundry and personnel decontamination), Additional radioactive liquid streams are produced

components will be scheduled for disposal in a deep geological repository after suitable packaging or will be routed for shallow burial.

during

Many of the dismantled components will not,

solid

waste

either from

decon-

(the biological

treatment

and

volume

reduction processes,

however, be radioactive and can be disposed or

Within the nuclear plant, raw solid wastes are

recycled as normal industrial wastes.

classed either as 'wet' (e.g. sludges, ion exchange resins) or 'dry' (e.g. rubber gloves, paper tissue). The ion exchange resins are generally dehydrated and powdered, and the final powder is solidified in either a bitumen or cement matrix. Other noncombustible solid wastes are compacted to reduce their volume and are then variously packaged in steel or cement containers before being solidified in cement or bitumen. The low activity, combustible solid wastes may be incinerated to reduce their volume, creating an ash

2.1.4

Other wastes

Radioactive wastes are generated outside the civilian nuclear fuel cycle in small amounts (by medical, industrial and research activities) but these still require careful handling and packaging, and most are scheduled for disposal in a repository. Larger volumes are also created by nuclear weapons programmes in some countries.

which is then immobilised in a cement matrix, Incineration itself generates secondary wastes such as the scrubber chemicals and particulate air filters which are used to clean the incinerator flue

In medicine, the radioisotopes used, particularly for clinical and diagnostic purposes, tend to be short-lived with non-penetrating radiation and do not pose a significant radiological hazard. However, sources of high activity are used in

gases. These secondary wastes must also be processed and become part of the total inventory

radiotherapy leading to an important long-lived waste component. In addition, medical and

of wastes to be disposed,

biological research experiments sometimes make use of 3H (tritium) and 14C and these provide a

2.1.3

Decommissioning wastes

significant contribution to the activity of medical radioactive wastes. Occasionally, in medical

At the end of a nuclear facility's operational life, it

applications,

must be decommissioned. These facilities include

with pathogenic substances which create special

radioisotopes may be associated

nuclear power plants, and fuel fabrication and

31

The geological disposal of radioactive wastes and natural analogues

waste handling and treatment concerns (Savage, for plutonium treatment being considered are 'burning' in civilian nuclear power reactors in the

1995). In industry, radioactive sources are used for a range

of applications

such

as radiography,

material thickness and density gauging, well logging, moisture detection, and food sterilisation and preservation. These sources are relatively short-lived but may be intensely radioactive, posing a significant radiological hazard and

form of a mixed uranium-plutonium oxide (MOX) fuel and direct disposal to a repository. However, not all stockpiles of plutonium and other military radioactive materials are currently classified as 'wastes' and thus are excluded from some inventories of radioactive wastes currently scheduled for disposal in repositories.

requiring specialist disposal. They are increasingly located in less developed countries with no nuclear facilities and no experience in radioactive

2,2 Classification of radioactive

wastes

waste management. Studies are currently in hand to evaluate the disposal of these very small volumes in deep boreholes. At nuclear research centres, various radioisotopes are produced in research reactors, particle accelerators and cyclotron facilities. In general, such facilities generate relatively small amounts of waste containing long-lived radioisotopes, especially accelerator and cyclotron facilities which do not possess nuclear fuel. For the latter, the main source of waste is derived from liquids produced during chemical processing or etching of target materials (Savage, 1995).

When defining a classification for radioactive wastes, it is important that the classification is linked to the planned method of disposal, including waste conditioning, to ensure adequate isolation from the surface environment. Although the classification of wastes tends to vary from one country to another, there is Some commonality in approach. In most countries, the wastes described earlier are generally classified according to their levels of radioactivity and three categories of radioactive wastes are usually referred to: 9

Lastly, military applications generate large volumes of radioactive materials. In broad detail, the reactors used to produce the nuclear component in weapons and to power some submarines and surface navel vessels are similar to civilian power plants. Thus, many of the radioactive materials produced as a result of military applications are similar to those from commercial nuclear power plants. However, many nuclear submarine enriched

fuel,

reactors

which

use very highly

presents

its

own

is

being

stockpiled

due

to

the

dismantling of large numbers of thermonuclear weapons. Plutonium disposal raises particular safeguards issues because of concerns regarding the hazards of theft and incorporation into new nuclear weapons (e.g. Garwin, 1996). Two options

32

waste stream generated during reprocessing of spent fuel. These wastes are characterised by high-levels of radioactivity, have a component of very long-lived waste nuclides and are heat generating. The activity of these wastes generally ranges from 1016 to 1018 Bq/t. 9

Intermediate-level waste (ILW)which includes a

diverse

range of materials

such as ion

exchange resins and metal wastes from normal reactor operations and spent fuel

management issues. Plutonium

High-level waste ( H L W ) w h i c h includes spent fuel and the solidified forms of the liquid

reprocessing. These wastes are generally solidified in either a cement or bitumen matrix. They are characterised by significant levels of radioactivity, have a component of long-lived waste nuclides but are not heat generating. ILW containing a high proportion of actinides is sometimes referred to as

Radioactive waste and repository types

transuranic containing

(TRU) waste but is

different physical and chemical characteristics. In

generally treated and packaged in the same manner as all other ILW. The activity of these

particular, the ILW contains large volumes of cement as an immobilisation matrix which will

wastes is generally higher than 10 9 Bq/t but

cause the

below the limit for HLW.

all other wastes that have a radioactivity content below the threshold for ILW coming normal

reactor

operations,

fuel

fabrication and reprocessing. These typically includes the 'dry' operational

wastes (e.g.

paper tissues and disposable clothing) which are compacted and packaged cement containers. Wastes

from

the

nuclear

in steel or

industry

with

radioactivity content below exemption

a

limits

(sometimes called very low-level wastes, VLLW) do not require disposal to a repository and can be treated

as

normal

industrial

wastes.

These

exemption levels vary from country to country but are generally around 1 Bq/g, which is broadly consistent with average background levels and the radioactivity contents of many natural materials such as soils and rocks. Not all countries operate this basic three category system" for example, in the United States, commercial wastes from nuclear reactors generating electricity are designated only as HLW or LLW. However, irrespective of the terminology used, the principal objective of any categorisation system is to ensure that the timescale over which any waste is isolated from the surface environment is compatible with the radionuclide content of that waste. This generally means that the HLW and ILW are destined for deep geological disposal, while the LLW may be disposed of in near-surface facilities. Most deep geological repository designs preclude the co-disposal of HLW with ILW, where co-disposal is taken to mean that the two waste types would be

placed

underground

in

the

same

excavations.

to

become

alkaline:

interaction between this alkaline groundwater and

Low-level waste (LLW) which essential includes

from

groundwater

or

closely

This

is

to

spaced avoid

complex interactions between wastes with very

HLW would

be

undesirable.

An

alternative

engineering option is co-location which involves building two different, separated repositories at one site, which share common surface facilities, shafts and access tunnels. The distinction with codisposal is that in co-location, the HLW and ILW are separated by a distance large enough to avoid any significant physical between them.

or

chemical

interactions

LLW may be co-disposed with ILW because they both have similar characteristics and are generally both solidified in cement, as shown in Figure 2.2. However, separate disposal facilities are often chosen for these two waste categories due to the additional costs of building a deep repository to house both ILW and LLW. Nonetheless, some combined L/ILW repositories have been built: examples include the Finnish VLJ repository at Olkiluoto and SFR repository at Forsmark in Sweden (see Box 3). In many countries, however, two or more different repositories are planned to be built for the individual waste categories.

2.3

Repository designs

There are basically two types of engineered repositories which are either being built or planned for the disposal of radioactive wastes. These are deep, mined repositories in the subsurface rock which potentially could house any waste category and near-surface repositories which would be used only for LLW. The exact design details of any repository will be dependent on the nature and volume of the waste it is planned to contain, the geological environment

at

the

disposal

site

and

engineering

constraints imposed by the host rock. However, all repository designs are based on the multibarrier

concept whereby the wastes are emplaced inside a

33

The geological disposal of radioactive wastes and natural analogues

concept is common to them all. A conceptual set of multiple barriers for a repository, starting innermost, includes the solid wasteform and the container (together referred to as the waste package), a low permeability backfill or buffer and finally the surrounding rock mass. A generic system showing these components was illustrated in Figure 1.2. The man-made barriers (wasteform, canister, buffer material) are generally referred to as the engineered barrier system and the rock as the geosphere. Clearly, in a deep repository for HLW, the extent of the geosphere will be considerably larger than that surrounding a surface or near-surface repository for LLW. Disposal sites will

be chosen such that the

geosphere provides a stable physical and chemical environment to protect the engineered barrier system. The surface environment, populated by humans and other animals and plants, is generally referred to as the biosphere, see Section 1.5.4. Construction of a deep repository is likely to employ either smooth-wall blasting techniques or tunnel boring machines. Either method will result Figure 2.2: Most ILW and LLW will be immobilised in a cement matrix, as shown here in a demonstration waste drum. Several drums may be placed in larger reinforced concrete boxes to create a stable waste package which can be stacked in the disposal vaults in the repository.

series of nested engineered structures and natural barriers which act in concert to control the rate of release of radionuclides over long periods. These barriers are generally devised such that they have no common failure mechanism. This means that, if some event or process were to cause one barrier

in the development of fractures in the rock around the excavations, although the extent to which such fractures might form depends on the physical characteristics of the rock, the construction technique used and on the design of the facility. These newly formed fractures will extend a certain distance out into the rock and define the engineered damaged zone (EDZ) which will affect both the physical and hydrogeological conditions around the repository. According to standard terminology, the combination of the engineered

to fail, this should not lead to the other barriers

damaged zone plus the engineered barrier system

also failing. In some sense, this implies a degree of redundancy in the disposal system but no

is referred to as the near-l~eld and the physically undisturbed, intact rock between this and the

individual barrier is actually designed in such a

surface is referred to as the far-l~eld.

way that it, alone, would necessarily ensure the

In addition to using deep engineered facilities,

safe isolation of the waste,

radioactive wastes can be disposed of in liquid

The specific barriers employed in the repository designs for the different waste types vary from

form by injection into deep rock formations. This

design to design but the fundamental multi-barrier

facilities (polygons) in Russia since the 1960s (Rybalchenko, 1998). This disposal method

34

method of disposal has been carried out at three

Radioactive waste and repository types

involves injection of liquid wastes into aquifers

design features for a HLW repository include a

(permeable rock formations), known as 'collector

series of tunnels in which the waste will be

galleries,

layers', at depths of up to 1.5 km for HLW and up

emplaced, sometimes referred to as

to 400 m for L/ILW. The wastes are assumed to be

which are excavated at depths in excess of 100 m

isolated from the surface by aquitards (low

(typically between 500 and 1000 m) and a number

permeability rock formations) above and below

of vertical or inclined shafts which connect the

the collector layer. Very large volumes of liquid

repository to the surface. An example is the

waste are injected by this method: for example,

proposed Swedish spent fuel repository design

150 000 m3/yr of liquid is injected at the

which is shown in Figure 2.3.

Dimitrovgrad facility in Russia under an injection pressure of 5 MPa. This method of disposal is relatively cheap and conceptually easy, although practical problems related to the precipitation of solids

in

the

injection

wells

have

created

operational difficulties. Due to the wastes being in liquid form, long-term safety cannot be guaranteed with confidence, although recent safety assessments of some of these injection facilities indicate that the wastes are well confined ona3OOyeartimescale(Hoek, 1998). Disposal by liquid injection is planned to be discontinued in Russia and is unlikely to be practiced in other countries, thus this method of disposal will not be considered any further here.

2.3.1

Deep repository designs for HLW

At the time of writing, no engineered repository for solid HLW (spent fuel and solidified reprocessing wastes)has yet been built. However, a number of conceptual repository designs have been developed and subjected to detailed safety analysis at a generic level. Furthermore, several countries are now involved in the site selection and site characterisation activities necessary to build such a repository. Examples include the Swedish and Finnish designs for spent fuel repositories, and the Swiss design for a repository for vitrified (glass) reprocessing wastes: these are presented in Boxes 1 and 2.

The waste itself will be placed in large metal canisters and these will be located in the disposal tunnels. All the spaces between the canister and the host rock will be filled with a buffer comprising compacted bentonite clay. Several geometric options for locating the canisters in the tunnel are under consideration. The simplest has the canisters arranged axially at intervals along the tunnel. This design has the advantage that a circular section tunnel can be driven easily into the rock with a tunnel boring machine but the disadvantage that the space around each canister must be backfilled at the time of emplacement before the next canister can be brought into position. Alternative designs locate the canisters in individual disposal holes drilled either into the floor or the walls of the tunnels. In this design, the tunnels generally have to be both wide and tall to allow long canisters to be manoeuvred into the disposal holes. This makes excavation of the tunnels a more difficult procedure. However, the advantage is that only the immediate space around each canister in the disposal hole needs to be backfilled at the time of emplacement. The main part of the tunnel can be left open, if so desired, for monitoring or other purposes until all canisters are in position. The canisters will be massive, thick-walled metal structures designed to isolate the waste for long periods of time. The canister metal may either be

Most deep repository designs for HLW currently

iron or steel, or a less reactive metal such as copper or titanium, or a combination of both. Iron

under development have a great deal of similarity,

and steel corrode faster than other possible

although alternative designs have been considered

canister metals but, in so doing, will buffer the

and evaluated over the last two decades. The basic

redox conditions by scavenging free oxygen from

35

The geological disposal o f radioactive wastes a n d n a t u r a l analogues

F Box 1" The proposed Swedish and Finnish spent fuel repositories The proposed Swedish repository for spent fuel is based on a design presented in 1983 in an early performance assessment known as KBS-3 (KBS, 1983) which has been modified and optimised over the last two decades. The basic design of the repository is shown in Figure 2.3. Disposal would occur in several arrays of tunnels excavated at depths in excess of 500 m. A very similar design is also being developed in Finland and POSIVA, the Finnish implementing agency, are seeking approval to construct a repository in granitic rocks close to the Olkiluoto nuclear power plant.

Figure B 1.1: Artist's impression of the proposed Swedish repository for spent-fuel. The repository will be excavated at depths in excess of 500 m. Illustration courtesy of SKB. The repository will be sited in the fractured crystalline rocks which predominate in Sweden and Finland. These rocks frequently contain large-scale fractures and fracture zones which are the dominant control on the groundwater flow system. The repository will have to be sited and designed to fit within the largest, stable blocks of rock defined by these fractures. The disposal tunnels have a distinctive 'key-hole' geometry in cross-section due to the individual disposal holes which are drilled into the floor of the galleries approximately 6 m apart, although this spacing will be modified to avoid smaller faults and fractures in the rock. It is not planned to emplace canisters along the axis of the tunnels. No tunnel liners are required because of the inherent strength of the crystalline rock. Repository designs with both one and two levels have been considered. The general design for a one level repository has tunnels with a diameter of 3.3 m placed approximately 25 m apart. The spent fuel will be encapsulated in bimetallic canisters which have a copper outer shell surrounding an inner cast iron (or steel) vessel. Copper is chosen for the shell because it is essentially inert in most groundwater systems (Section 4.4) and, thus, will corrode slowly providing very long canister life-times: some estimates put the life-time of these canisters at 106 years or longer. The inner cast iron vessel provides mechanical support for the canister and a large redox buffering capacity if the copper outer shell is perforated.

j

36

Radioactive waste a n d r e p o s i t o r y types

f

"x Single canisters will be placed in the disposal holes with all the spaces around them filled with machined blocks of compacted bentonite. The space in the tunnels above the disposal holes will be backfilled with a mixture of sand and bentonite. A number of recent performance assessments have been undertaken for this general repository design (SKB, 1992; SKI, 1994; SKB, 1999; Vieno and Nordman, 1999) which have used proxy data from real crystalline rock sites in Sweden and Finland to improve the level of realism in the assessments. These assessments and the research undertaken to support them highlighted a number of issues which have required more detailed investigation by natural analogues, these include: 9

dissolution processes and rates for the spent fuel go 2 matrix (Section 4.2);

9

general and Iocalised corrosion of copper and steel (Section 4.4);

9

radionuclide transport and retardation in crystalline rock (Section 5.1);

9

matrix diffusion in the rock adjacent to fractures (Section 5.3);

9

radiolytic decomposition of groundwater (Section 5.4); and

9

radionuclide behaviour at redox fronts (Section 5.5).

These performance assessments have all highlighted the potential very long life-time of the canister due to its unreactive copper outer shell. In the absence of any manufacturing defects or unexpected early perforation, the canister alone can provide adequate long-term isolation of the spent fuel assuming a stable disposal environment. However, absence of defects cannot be guaranteed at the current pilot stage of canister manufacturing development and thus the performance assessments have investigated the consequences of various canister failure scenarios. The results of these calculations show that continued safety of the repository, after canister failure, is controlled largely by the groundwater flow barrier provided by the bentonite buffer and by radionuclide retardation in fractures in the crystalline host rock.

Figure B1.2: The cross-sectional geometry of the Swedish spent fuel repository tunnels. Individual canisters are placed in disposal For all probable scenarios, the holes drilled into the floor of the galleries with surrounding spaces performance assessment calcufilled with blocks of compacted bentonite. Canister spacing would lations indicate that no unsafe be modified to avoid any large fractures or faults in the rock. releases to the surface environment Illustration courtesyofSKB, would occur. These scenarios include evaluation of the consequences of future glaciation and permafrost development which are expected to occur in Sweden and Finland over the next 105 years due to natural climate changes.

37

The geological disposal of radioactive wastes and natural analogues

properties of plasticity and expansion in the presence of water (Section 4.5). Once the repository has been closed and sealed, groundwater will begin to flow back into the excavations causing it to hydraulically resaturate. The bentonite will adsorb this water and expand to fill all remaining void spaces, creating a material with a very high swelling pressure, and very low porosity and hydraulic permeability. As a result, groundwater flow will not be able to occur through the buffer between the rock and the canister; movement of water, canister corrosion products and released radionuclides could only then occur by diffusive processes which are very slow. Other advantages of the bentonite are that it will sorb some

Figure 2.3: An example of a deep repository for spent fuel or HLW, showing an array of disposal tunnels at depth, linked to the surface by a number of shafts. In this case, the design is for a Swedish repository for spent fuel and shows the typical 'key-hole' near-l~eld design described in Box 1. Illustration courtesy of SKB. the near-field, thus maintaining a strongly chemically reducing environment (see Section 4.4). Chemically reducing environments are favourable because most of the radionuclides of concern are poorly soluble under such conditions, Copper and titanium, in contrast, will corrode much more slowly than steel, providing a potentially longer canister lifetime but less capacity for buffering the redox conditions. Both 'corrosion allowance' and 'corrosion resistant' metals offer different containment advantages,

proportion of the released radionuclides to the clay particle surfaces, act as a barrier to colloid movement (see Section 5.6) and evenly

distribute water around the canister surface, lessening the chance of Iocalised corrosion. Sodium-bentonite is presently favoured as the buffer material in most repository designs due to its superior swelling capacity. This will react with the groundwater, buffering pH to mildly alkaline conditions (around pH 8). This buffering capacity provided by the large volume of bentonite means that the porewater composition close to the canisters is, to some extent, independent of the original groundwater composition in the far-field.

The bentonite clay buffer placed around the canister plays a very important role in ensuring repository safety. Bentonite exhibits special

38

Sodium-bentonite can undergo ion exchange with any dissolved

calcium

or potassium

in the

Radioactive waste and repository types

trans-

long boreholes with closely spaced containerised

formation to either a calcium-bentonite or illite (see Section 4.5). These clays have a lower

groundwater,

causing

mineralogical

waste at depths of around 500 m (e.g. SKB, 1990; Gibb, 1999). Some of these options are shown in

swelling capacity than sodium-bentonite and,

Figure 2.4. Although some alternatives have been

thus, these transformations potentially may affect

investigated in detail, they generally have been

repository performance. However, given the very

found to be less suitable than the 'standard' HLW

slow rate of these transformations, coupled with the almost stagnant groundwater flow conditions

design for a number of reasons, such as the potential requirement for future retrievability of

in the repository near-field and the massive

the waste.

amounts of bentonite present, these reactions are not expected to be significant,

A conceptually very different HLW repository design is being developed in the United States and

One possible disadvantage of bentonite as a buffer

is unique in that it is located above the water table

is its load-bearing capacity and there is the possibility that the heavy canister may slowly sink

in unsaturated rocks (volcanic tuffs). This is in contrast to all other current HLW and spent fuel

through the bentonite if it were to act as a viscous fluid. If the canister did sink completely through the bentonite and come into contact with the rock, this would effectively negate the bentonite's barrier capacity. This is an issue which has not yet been fully resolved, although most assessments

repository designs which will both be located below the water table and, thus, will be water saturated, such as the Swedish and Swiss designs detailed in Boxes 1 and 2. The United States repository is planned to be built at Yucca Mountain in Nevada, and the preferred site is the focus of

indicate that it is unlikely to be a serious problem, Another drawback of bentonite is its generally low thermal conductivity, in contrast to the

intense site characterisation (see Section 1.5.1). If an operation licence is obtained, this repository

surrounding bedrock, which means that the spacing of canisters and their waste loading must be carefully planned to avoid overheating in the near-field.

will house spent fuel and some military wastes. The subsurface layout of the Yucca Mountain

crystalline rocks, various types of sedimentary rocks or evaporite (salt) deposits, all of which vary in their physical and chemical characteristics, in terms of groundwater flow, chemistry, thermal conductivity and physical strength. As a consequence, the design of a HLW repository will

repository consists of a series of disposal tunnels connected to the surface by two inclined ramps leading to the lower slopes of the Yucca Mountain ridge. Waste will be encapsulated in large containers manufactured from an outer corrosion allowance metal, such as mild steel, and an inner corrosion resistant metal, such as Inconel. The containers will be located along the centre lines of the disposal tunnels. It has not yet been decided if the void spaces between the containers and the

require optimisation for the particular host rock and geological environment chosen, although the basic elements of the near-field design would

rock will be backfilled immediately after the containers are put in place. The repository could be left without a backfill for a considerable period

not change significantly.

of time to allow monitoring and waste retrieval, if

The host rock for a HLW repository could be hard

Potential host rocks

and geological environments are discussed in

required.

Section 2.4.

Although the volcanic tuffs at Yucca Mountain are

Alternative HLW repository designs which have

technically hydraulically unsaturated, they do

been proposed in the past include disposal of

contain significant amounts of groundwater in the

containerised solid waste in very deep boreholes,

pore spaces and some perched water tables have

drilled to depths ranging from 3 to 4 km, or in very

been encountered in the tuffs during the site

39

The geological disposal o f radioactive wastes and n a t u r a l analogues III

I

II

I

II

f

Box 2: The proposed Swiss repository for vitrified HLW The proposed Swiss repository for vitrified (glass) spent fuel reprocessing wastes is based on a design presented in 1985 in an early performance assessment known as Project Gew~hr (Nagra, 1985) which has been modified and optimised over the last fifteen years. Alternative host rocks which have been considered for this repository include crystalline basement rocks in which the repository would be sited at a depth of around 1200 m, and argillaceous (clayey) sediments in which the repository would be sited at a depth of around 850 m. An artists impression of this repository design was shown in Figure 1.1.

Glass matrix (in steel mould)

e L . . . . . . . . i. . . . teofglass 9 High resistance to radiation damage 9 Homog . . . . . . . .

dionuclidedistribution

Steel canister

9 Completely isolates waste for > 1000 years 9 C . . . . sion products act as a chemical buffer 9 Corrosion products take up radionuclides

Bentonite backfill

9 Long

. . . . turationtime 9 Low solute transfer rates (diffusion) 9 Retardation of radionuclide transport (sorption) 9 Chemical buffer

9 Low radi . . . . lide solubility in leachate 9 Colloid filter 9 Plasticity (self-healing following physical disturbance)

The disposal tunnels have a circular cross-section and canisters will be emplaced axially along the tunnels. Due to the different physical and thermal characteristics of the two types of rocks, a repository in argillaceous sediments would have smaller tunnel diameters (2.5 rn instead of 3.7 m) and tunnel spacings (25 rn instead of 40 m). For stability, a tunnel liner would be required if the repository was built in sediments. For both the crystalline and sedimentary options, the vitrified HLW will b e encapsulated in massive, 25 cm thick steel canisters. Canisters will be spaced along the tunnel axis and all the spaces around them filled with machined blocks of compacted bentonite. Steel is reactive in groundwater and will corrode to form amorphous iron oxyhydroxides. However, the benefits of this reaction are that iron corrosion

will buffer the geochemistry to maintain chemically reducing reactions and the Geological barriers iron oxyhydroxides will provide a large Repository zone: 9 Low water flux capacity for radionuclide sorption (see 9 Favourable geochemistry 9 Mechanical stability Section 4.4). In addition to the iron, the Geo, p. . . . . bentonite buffer will help to control the 9 Retardation of radionuclides (sorption, matri x di f fusi o n) groundwater chemistry such that the 9 Reduction of radionuclide concentration (dilution, radi o acti v e decay) near-field chemistry will be very similar 9 Physical protection of the engineered barriers (e.g.fromglacialerosion) in both the crystalline and sedimentary I options. The largest differences betFigure B2. 1" The series of engineered and natural barriers ween the two options relate to

which isolate the waste from the surface environment in the Swiss HLW repository design. The repository design was shown contrasts in the thermal conductivities in Figure 1.1. Illustration courtesy of Nagra. and mechanical strengths of the rocks.

40

Radioactive waste a n d r e p o s i t o r y types

f

Both the thermal conductivities and heat capacities of the sediments are lower than for the crystalline rocks, leading to higher predicted near-field temperatures. However, the temperature can be controlled by operational measures (e.g. lower waste Ioadings or longer storage prior to disposal) or repository design (e.g. larger tunnel diameter, higher backfill conductivity or greater spacing between canisters). The mechanical strength of the sediments is considerably less than that of the granite. Therefore, tunnels in the granite would be self-supporting while those in sediments may require steel liners. These liners have advantages (e.g. enhanced near-field redox buffering capacity) and disadvantages (e.g. delayed hydraulic resaturation) for repository safety and thus the consequences of leaving steel liners in place need to be evaluated. A number of performance assessments (e.g. Nagra, 1994) and supporting research studies have been undertaken on these designs and they have highlighted a number of issues which have required more detailed investigation by natural analogues, these include: dissolution processes and rates for the glass wasteform (Section 4.1); corrosion rates and products from steel (Section 4.4); radionuclide transport and retardation in crystalline rock (Section 5.1); radionuclide transport and retardation in argillaceous rock (Section 5.1); matrix diffusion in the rock adjacent to fractures (Section 5.3); and radionuclide behaviour at redox fronts (Section 5.5). With conservative assumptions for canister lifetime, performance assessments for this repository design indicate that the most important issues for safety are the geochemical controls on radionuclide solubility in the near-field and retardation in the farfield. In particular, they show that the advective flow path Figure B2.2: Canisters will be placed axially along the tunnels with a separation must be specified in of 5 m. All spaces will be backfilled with machined blocks of compacted detail at the smallbentonite. All void spaces between the bentonite blocks will disappear as the scale to calculate bentonite resaturates and swells. Illustration courtesy of Nagra. solute transport. From this point of view, homogeneous sediments have advantages over fractured, crystalline rock because the flow path will be easier to define. However, the performance assessments indicate that both potential host rocks will be adequate to guarantee repository safety j

41

The geological disposal of radioactive wastes and natural analogues

receiving

waste.

In

Scandinavia, both the repositories at Forsmark in

Sweden,

and

at

Olkiluoto and Lovisa in Finland take both ILW and LLW. The Forsmark repository is described in Box 3. In the United States, the WlPP (Waste Isolation Pilot Plant) facility in New Mexico, which is excavated from bedded evaporite (salt) deposits, has recently begun receiving military derived TRU wastes. Figure 2.4: Artist's impression of the Swedish spent fuel repository concept Other ILW repositories (left) and two alternatives: the very long hole (middle) and the very deep hole are planned in other (right). The size and spacing of canisters varies between concepts although, in each case, the canister is surrounded with compacted bentonite blocks, countries. Illustration courtesy ofSKB. Most ILW repositories investigations. There are significant conceptual uncertainties about the mechanisms controlling

are designed to be located at shallower depths than proposed HLW repositories. Typically, ILW

the behaviour of water in unsaturated porous rock and the transport of radionuclides away from the

repositories will be at depths of around 100 m below the ground surface, although some designs

repository. Consequently, natural analogues for the Yucca Mountain design are required to address a different series of transport issues to those of relevance to other HLW repository concepts, such as radionuclide migration in a two-phase (water and air) system.

are deeper. For example, the proposed British ILW repository was to have been built at a depth of around 750 m below the Sellafield site in northwest England, although plans for this repository have now stalled.

2.3.2

Deep repository designs for ILW

Designs for deep ILW repositories depend on the nature of the waste to be disposed of, in terms of its volume, radioactivity and physical nature (i.e. if it is immobilised in cement, bitumen or other material).

However, all designs for deep ILW

The excavations for an ILW repository generally consist of a number of large disposal caverns (sometimes referred to as vaults), rather than the disposal tunnels planned for a HLW repository. In most designs, the waste will be placed in large reinforced concrete boxes and stacked in the caverns, with spaces between them backfilled with cement

or

concrete.

The

proposed

Swiss

repository design is shown in Figure 2.5. repositories currently under development differ significantly from proposed designs for deep HLW The number and dimensions of caverns in an ILW repositories, will vary from design to design but can be very large. As an indication of the upper size range, the In contrast to HLW repositories, several deep ILW proposed British ILW repository was to have been repositories have actually been built and are based around a number of caverns 25 m wide,

42

Radioactive waste and repository types

16 m high and 265 m long. To reduce groundwater flow through the cavern walls and to support the excavations, some caverns may be lined with concrete and steel reinforcements. In addition to the caverns, certain ILW repository designs also include a silo structure to contain the higher activity wastes. A silo is a reinforced concrete, cylindrical structure built within a tall cavern. These typically may have a diameter of 20 to 30 m and a height of several tens of metres. The spaces between the reinforced concrete silo shell and the rock may be backfilled with bentonite so that groundwater flow through the silo will be substantially less than through the other caverns. The silo structure at the Finnish ILW repository at Olkiluoto is shown in Figure 2.6. As discussed in Section 2.2, ILW is heterogeneous, comprising a mixture of materials arising from the routine operation or decommissioning of nuclear power plants and fuel processing plants. These wastes will be solidified (immobilised) in either a cement, bitumen or resin matrix, although cement will

be used most abundantly. The solidified

wastes will be contained in either thin-walled metal drums or large concrete boxes, depending on the waste type and repository design, to provide a stable waste package. In most designs, these waste packages are largely designed to aid handling and transport to and within the repository, and are not intended to provide longterm isolation of the waste from the groundwaters in the repository. In a typical repository, these waste packages will be stacked in the caverns and in the silo, and all residual spaces between them will be backfilled with cement or a cementitious mortar. Wastes with different levels of radioactivity or chemical form may

be segregated

and

placed

in separate

caverns, with the highest activity wastes located in the silo (if present). If LLW is also emplaced in the repository, this will be located in separate caverns

Figure 2.5: The series of engineered and natural to the ILW. barriers which isolate the waste from the surface environment in the Swiss ILW repository design. Illustration courtesy of Nagra.

43

The geological disposal o f radioactive wastes and n a t u r a l analogues

F

Box 3: The Swedish L/ILW repository at Forsmark

The Swedish combined repository for ILW and LLW (known as the SFR repository) has been built on a coastal site at Forsmark, adjacent to an existing nuclear facility to minimise radioactive waste transportation. It is one of the few examples of an operating geological repository, currently accepting wastes. The repository has been excavated at a depth of 60 m beneath the bed of the Baltic Sea in gneissose rock with a low hydraulic conductivity of 10.8 to 10-7 m/s. Locating the repository beneath the seabed ensures a very low hydraulic gradient and, as a consequence, low groundwater flow rates. At the current rate of isostatic uplift of the land of around 6 mm/yr (due to glacial rebound), it will take around 1000 years for the repository area to become dry land. In the short to medium-term, however, while the site is covered by the sea, the possibility of inadvertent human intrusion is clearly minimised. Currently, the repository consists of one silo and four caverns, each with a different barrier system, designed to hold different waste streams. The disposal volume in the repository is 60 000 m3, although there are plans to extend it with a second silo and two additional caverns to provide a total disposal volume of 90 000 m3.

Figure B3.1: Artists impression of the Swedish L/ILW repository at Forsmark (the SFR repository). The silo houses the highest activity wastes, while lower activity wastes are placed in the four caverns. Illustration courtesy of SKB.

x..

44

J

Radioactive waste and repository types

The silo is constructed from a reinforced concrete shell, 90 cm thick. It is 50 m high, 28 m in diameter and is located within a 70 m high cylindrical rock cavern. The void between the rock and the concrete silo is backfilled with bentonite, approximately 1.3 m thick. At the top and bottom of the silo, a sand/bentonite mixture (90% sand) is used as the buffer material to provide greater bearing strength (at the bottom) and gas permeability (at the top). The silo is receiving ILW in the form of ion exchange resins solidified in cement and bitumen matrices, and packaged in steel drums and concrete moulds. After this silo is filled, all spaces will be backfilled with cement. The four caverns are each 160 m long and their walls are lined with shotcrete to provide mechanical support. The caverns hold LLW and lower activity ILW immobilised in cement and packed in steel or concrete containers. After they are filled with waste, the spaces above and between the waste packages will be backfilled with either cement or crushed rock. No low-permeability buffer material is used in the caverns. Near-field groundwater chemistry in the repository will be both very alkaline and reducing. The large volumes of cement and concrete will buffer the pH, and the iron present in the canisters, reinforcing rods etc., will buffer the redox potential in the repository. On closure of the repository (scheduled for 2010), drainage pumping will end, the repository will be sealed and hydraulic resaturation will commence. The present conceptual hydrogeological model anticipates groundwater flowing horizontally through the caverns, leaching radionuclides from the waste form, and then flowing vertically through fissures in the rock to reach the seabed, where the Baltic Sea will ensure large dilution of releases. Radionuclides leached from the waste in the silo would first diffuse through the bentonite buffer before being advectively transported with the slowly moving groundwater to the seabed. This situation will change with time due to ageing and weakening of the engineered barriers and continued uplift of the seabed. Several performance assessments for the repository consider separately the period when the repository remains beneath the Baltic Sea when recharge waters are saline and the period, after uplift, when the repository area dries out and a freshwater system is created. These performance assessments have highlighted a number of issues which have required more detailed investigation by natural analogues, these include: 9

degradation of cement (Section 4.6)

9

degradation of bitumen and organic waste (Sections 4.7 and 4.8);

9

radionuclide solubilities in high pH environments (Section 5.1);

9

colloidal activity (Section 5.6);

9

microbial activity (Section 5.7); and

9

gas generation and transport (Section 5.8).

A particular issue which is more important for this repository than for HLW repository designs is gas generation from the degradation of organic materials. Preliminary assessments showed that a build-up of gas pressure potentially can affect the structural integrity of the engineered barriers and the rock. Consequently, the repository was designed to allow gas to escape freely frorr, the silo and caverns. For all probable scenarios, the performance assessment calculations indicate that no unsafe releases to the surface environment would occur, including changes that result from land uplift and retreat of the sea.

45

The geological disposal of radioactive wastes and natural analogues

Over a long time period, the high pH groundwater will migrate from the repository into the host rock as a hyperalkaline 'plume' which may result in slow changes to the composition of the cement and the pH of the near-field. The high pH groundwater could influence the host rock and, depending on the nature of any reactions, may cause both physical changes, thus creating a around the repository. slow groundwater flow

and chemical disturbed zone The expected rate will allow

chemical equilibrium to be maintained in the near-field and along the migration path from the near to far-field. The high initial pH in the near-field will gradually

Figure 2.6: Artist's impression of the silos in the Olkiluoto L/ILW repository in Finland. These silos are constructed from reinforced concrete, surrounded by a bentonite based buffer between the concrete and the rock. Waste is packaged in large reinforced concrete or steel boxes and stacked in the silos. Illustration courtesy of Nagra. The large volumes of cement in a typical ILW repository have an important safety role. As groundwaters infiltrate the near-field after closure, they will chemically equilibrate with the cement and concrete, buffering the pH of the porewaters to around pH 13 or 14. Such hyperalkaline conditions are expected to persist for 104 years or more, depending on the mass of the cement, and the chemistry and flow rate of the groundwaters.

drop

to

12 over

a few

thousand years but will then be maintained for some 10 s years by the slow dissolution of the cement minerals (see Section 4.6). After about a million years the pH will have dropped to around 10, depending on the groundwater flow rate. This is much in excess of the expected life span of the

engineered barriers and indicates that the conditions in the near-field should act as a chemical buffer even after the physical integrity of the concrete has been lost.

2.3.3

Near-surface repository designs f o r LLW

Progressive corrosion of the steel canisters and

The disposal of LLW in near-surface or shallow repositories is practiced routinely in many

steel reinforcing rods will occur, and the corrosion

countries. These repositories are located at or

products (principally hydrogen and dissolved iron)

close to the land surface because the low activity

will control the redox environment, creating and

and short half-lives of the waste they contain

maintaining chemically reducing conditions. High

means that they do not require the very long

pH, chemically reducing environments are favourable because most of the radionuclides of

isolation times necessary for HLW or long-lived ILW.

concern are poorly soluble under such conditions.

The earliest LLW disposal designs were little more than landfills, based on simple trenches into which

46

Radioactive waste and repository types

the waste was tipped and then backfilled with earth.

More

recent

designs employ a more robust,

multi-barrier

engineering

approach

and it is these that are considered

here,

e.g.

Figure 2.7. The basic design involves a series of reinforced

concrete

bunkers or trenches into which packaged waste is

emplaced.

structures

These

may

be

located on the ground surface or they could be fully or partially set into the ground, with even- Figure 2.7: Photograph of the Centre de I'Aube surface repository for LLW in tual closure beneath a France. Wastes packaged in steel or concrete boxes will be placed in the large thick earth cap. engineered compartments which are each 24 m long. When the repository is 191led, the compartments will be sealed with a concrete 'lid' and the entire As discussed in Section facility will be covered with an earth and clay cap, to form a small hill. Illustration courtesy of Nagra. 2.2, LLW is heterogeneous, comprising a mixture of materials from completed, the entire repository system will be normal

reactor

reprocessing

operations,

and

fuel

fabrication,

decommissioning.

covered with a low permeability earth and clay cap

These

to provide an outer protective barrier. The earth

typically include items such as cleaning materials

cap may form a mound or hill. This cap will often

and disposable clothing but may also include large

comprise layers of low permeability clay or an

items from decommissioning activities. These materials will be compacted and solidified (immobilised) in cement. Most solidified wastes will be contained in large concrete boxes or steel

impermeable membrane to reduce water percolation into the repository. Specially chosen vegetation is often planted on top of the earth cap, again to limit water percolation.

containers, depending on the waste type and

The various combinations

repository design. These waste packages are not intended primarily to provide any long-term isolation of the waste.

concrete structures and earth caps used reflect the

of waste packages,

volume and characteristics of the LLW and the geology and climate at the repository site.

In a typical repository, these waste packages will be stacked in the bunkers and trenches and all with earth, clay or cement. After each trench or

2.4 Geological disposal environments

bunker is filled, it may be closed with an 'anti-

The

residual spaces between them will be backfilled

fundamental

requirement

of a suitable

intrusion' slab of reinforced concrete designed to

geological environment for a repository is that it

prevent inadvertent excavation into the wastes in

should

the

engineered barriers and its behaviour should be

future.

When

all

disposal

activities

are

be

relatively

stable

to

protect

the

47

The geological disposal of radioactive wastes and natural analogues

adequately predictable. The requirement for

performance assessment will be provided by site

predictability arises from the need confidently to

characterisation programmes (see Section 1.5.1)in

demonstrate the long-term behaviour of the

an

repository system in performance assessment, as

heterogeneity and spatial variability of the system,

attempt

to

understand

the

nature

of

discussed in Section 1.3. In this regard, geological

especially for fractured crystalline rocks in which

stability relates not only to the physical features of

the heterogeneity of the hydraulic system can be

a site but also to the geochemical and hydrochemical aspects. This is particularly

particularly marked (Hodgkinson and McEwen, 1991).

important

for

deep

spent

fuel

and

HLW

repositories designed to contain the longest-lived

Nonetheless, natural analogues can provide useful

radioactive wastes,

generic information in this regard, by identifying and quantifying the most important processes

Put simply, the basic requirement of any suitable

which control transport and retardation in the rock

host rock is that it should provide a stable cocoon for the repository in which the rates of all physical and chemical processes which might disturb the engineered barrier system are very slow and are not likely to be subject to significant disturbances or modifications over the time period of containment.

types listed above. This is particularly so for natural analogue studies which look at sites on a scale large enough to examine an entire flow system and can show how radionuclide transport is affected along the flow paths. A good example is the investigation of colloidal transport at the Morro do Ferro site, part of the Polos de Caldas analogue study (see Box 14).

The search for suitable host rocks for deep repositories has concentrated on four principal

A lot of the information that comes from the

rock types:

natural analogue studies which is useful for

9

crystalline basement rocks"

9

extrusive volcanic

characterising disposal environments is not strictly analogue information. Rather, it is geoscientific data that is collected about the

rocks (e.g. lavas and

pyroclastics);

analogue

9

low-permeability sedimentary sequences; and

9

thick or diapiric evaporite (salt) deposits,

analogue system, and to constrain and interpret other data. For example, analogue information on colloidal populations cannot be usefully

sites

in order

to

understand

the

Some geological environments which have been

interpreted with regard to radionuclide transport,

considered for a repository combine one or more of these rock types, as with basement rocks under sedimentary cover, the so-called 'BUSC' environment. The fundamental characteristics of these

if the hydrogeological parameters of the system are not known. As a consequence, there is a considerable volume of field data obtained from analogue studies which can be used to improve

different

later,

our understanding of flow and transport in

together with the transport and retardation issues

different rock types, which is generally not given

specific to these environments which have been

an analogue label.

environments

are

described

addressed by natural analogue studies.

Most rocks considered as host rocks for a

In any geological environment, the greatest

repository will have low hydraulic conductivities

uncertainty with regard to natural conditions and

and low hydraulic gradients, and it is these

processes will probably be associated with the

features that are largely important for determining

flow of groundwater and radionuclide transport in

groundwater flow velocities. In rocks with plastic

the far-field rock. Much of the relevant information

behaviour (e.g. argillaceous rocks and evaporites)

on the transport pathways at a site needed for

the hydraulic conductivities

48

may be so low

Radioactive waste and repository types

(< 1010 m/s) that groundwater and contaminant

the measurement of representative transport

movement occurs predominantly by diffusion,

properties of such pathways is inherently difficult.

especially at depth. The hydraulic conductivities of the matrices of rocks with brittle behaviour (e.g. crystalline and volcanic rocks, and some hard

It is only by linking hydrogeological measurements and modelling with geochemical and structural geological studies that sufficient

sediments) may also be very low particularly in the rock mass chosen to locate the deposition tunnels. At these locations, some diffusion may also be

understanding is likely to be obtained (Nagra, 1994; Mazurek et al., 1992a) and natural analogue studies have helped in this regard.

expected to occur.

Natural analogue studies have tended to focus on

Contrastingly, in the far-field environment, movement of groundwater predominantly occurs through the fracture networks which are ubiquitous. Consequently, for these rock types, it is the hydraulic and transport properties of the fractures rather than of the matrix that are important and these can vary considerably, Crystalline rocks generally have very low total porosities which comprise the porosity of the fracture network and the porosity of the dead end pores which are connected to the flowing fractures,

determining the retardation capacity of the fracture surfaces in these rocks because it is predominantly these surfaces which are in contact with mobile groundwater. The low effective porosity also means that the surface area available for sorption is much less than in permeable sedimentary sequences, lnaddition, in contrast to the potentially long pathways which are possible in sedimentary environments, pathways in crystalline rocks are invariably short owing to the frequent presence of near-vertical fractures or faults (SKB, 1992; Mazurek et al., 1992b).

Crystalline rocks

Natural analogue studies in hard, fractured rocks have examined a number of processes which might affect flow and transport, and have provided important information on:

Crystalline rocks form at high temperatures and pressures deep beneath the surface either by cooling and crystallising from a magma (in the case of igneous rocks) or by solid state recrystallisation of existing rocks (in the case of metamorphic rocks). These rocks are very hard but tend to be brittle and, thus, are usually cut by faults and fractures. Several countries are considering crystalline basement rocks for deep repositories, including Canada, Czech Republic, Finland, France, Japan, Spain, Sweden, Switzerland and the UK. The critical

feature

of these

rocks

is that

groundwater flow and, thus, radionuclide transport will be concentrated in the fracture network, with essentially no advection occurring in the intact matrix of the rock, which can act to retain radionuclides which diffuse into it from mobile water in the fractures: a process known as

the groundwater chemistry in crystalline rocks and radionuclide solubilities in these hydrochemical environments (Section 5.1); 9

9

the retardation processes which operate in fractures and quantitative information on the sorption capacity of fracture lining and fracture filling minerals (Section 5.2); the potential for matrix diffusion and the depth of interconnected porosity adjacent to the fractures (Section 5.3); the development and movement of redox fronts in fractured systems (Section 5.5); and the populations and movement of colloids, organics and microbes through the fracture network (Sections 5.6 and 5.7).

matrix diffusion (Section 5.3). The detection and

49

The geological disposal of radioactive wastes and natural analogues

Extrusive volcanic rocks

9

Extrusive igneous rocks are formed by the cooling and solidification of volcanic lavas and ash

radionuclide speciation in an unsaturated, oxidising environment (Section 5.1); and

9

the transport and retardation processes which

deposits, in some cases, the ash deposits are

operate in an unsaturated fractured rock

deposited at very high temperatures which 'glues'

(Section 5.2).

together the individual particles to form a massive rock formation known as a welded tuff. These rocks tend to have low matrix porosity and, after

Sedimentary sequences

cooling, are hard and brittle, and thus can have physical characteristics similar to the crystalline rocks, although their mineral grain size tends to

The sedimentary rocks which may provide potentially suitable repository host rocks are clays, mudstones, marls and shales. These are all fine

be much smaller. In geochemical terms, welded

grained rocks which form by the accumulation of

tuffs have similar ranges in bulk compositions to crystalline rocks, although they are often altered by hydrothermal fluids associated with their volcanic origin, leading to the formation of zeolite

sediment underwater in marine or freshwater environments. Due to their mode of formation, they can be laterally extensive and homogeneous although, in both the vertical and horizontal

and clay minerals in fractures and pore spaces,

directions, sedimentary sequences can change

Geologically old volcanic rocks may be located

rapidly from fine to coarse grained material.

below the surface under more recent sedimentary sequences and form potentially suitable repository host rocks. Some countries are considering these

As sediment accumulates, it becomes compressed and its porosity is reduced. In some cases, the individual grains become cemented together

rocks for deep repositories, including Japan and the UK. Due to their hard, fractured characteristics, when these rocks occur below the water table they

(lithified) making the sediment mass brittle and liable to fracture. The majority of shales and many mudstones are fractured and can have quite

have quite similar transport characteristics to

similar flow characteristics to fractured crystalline

crystalline rocks and, hence, analogue study objectives for the two rock types would be similar,

rocks and, hence, analogue study objectives for the two rock types can be similar.

Young welded tuffs are frequently found at or close to the surface in volcanic regions. Such rocks occur at the site of the proposed US repository at Yucca Mountain, described in Section 2.3.1. The US repository would be excavated in the upper parts of the welded tuffs, above the water table, and hence the host rock environment would be

In some clay-rich sedimentary rocks, the clay particles do not cement together and the sediment mass can retain its plasticity. Thick clay sequences are potentially suitable for deep disposal because they tend to have low hydraulic conductivities, are associated with low solute transport rates and can have discontinuities which are self-sealing. Transit

unsaturated. Although the basic geochemical and

times for water from within a relatively deep

mineralogical characteristics of this rock would be

clay horizon to the surface can also be extremely

similar to other welded tuffs, the unsaturated

long. Several countries are considering clay-rich

nature of the tuffs at Yucca Mountain mean that analogue studies would need to address specific issues, such as:

sediments for deep repositories, including Belgium, France, Spain and Switzerland.

9

However, there is concern that the plastic nature of the degradation of UO2 (spent fuel) in an clay-rich sediments could be lost in some unsaturated, oxidising environment (Section circumstances and that radionuclides might be 4.2);

50

able to move more freely through them than

Radioactive waste and repository types

predicted (e.g. Gera et al, 1992). Natural analogue

radionuclides within salt is very slow for most

studies in sedimentary sequences have examined

situations which can be envisaged. Salt has higher

a number of processes which might affect flow and transport in these rocks types and have

thermal diffusivity than crystalline or argillaceous rocks and, therefore, temperature rises due to heat

provide important information on:

from the waste can be lower. Due to their plastic behaviour, these rocks are also essentially self-

the groundwater chemistry in sediments and radionuclide

solubilities

in

these

hydro-

chemical environments (Section 5.1 ); 9

the retardation processes that operate in sediments and quantitative information on the sorption capacity of clay minerals (Section 5.2);

9

diffusion coefficients in compacted sediments (Section 5.2);

9

matrix diffusion in sedimentary rocks (Section

5.3); 9

sealing.

Several

evaporates

for

countries deep

are

repositories,

considering including

Belgium, France, Germany, Spain, Switzerland and the US. However, despite the interest in evaporites, very few natural analogue studied have examined them in detail and relatively little analogue information has been obtained for them. Field studies at the WlPP repository site in New Mexico, now receiving military TRU wastes, have revealed that evaporites can be associated with high pressure brines and the presence of local deposits of both oil and gas.

the development and movement of redox fronts in sediments (Section 5.5); and the populations and movement of colloids through sediments (Section 5.6).

Evaporite deposits Evaporite deposits form by the evaporation of closed water bodies to leave behind an accumulation of salts. They tend to be located in sedimentary basins, in association with sedimentary rocks and, hence, can be interstratified with clays and other sediments. A wide range of minerals can be found in evaporites, typically chlorides and sulphates of sodium, magnesium, calcium and strontium, although most interest for waste disposal is focussed on large masses of relatively pure rock salt (sodium chloride). Thick beds of evaporites (bedded evaporites) may remain in horizontal formation but, if they are covered by large thicknesses of sediment, they may become unstable, in terms of density and plasticity, and rise upwards to form salt domes. Both bedded evaporites and salt donees appear to provide some of the most suitable environments for deep disposal because the transport rate of

51