Andra research on the geological disposal of high-level ... · Dossier 2005 Research on deep disposal of radioactive waste A general interest task Andra (French National RadioactiveWaste

Agence nationale pour la gestion des déchets radioactifs

Andra research onthe geological disposalof high-level long-livedradioactive wasteResults and perspectives

Re

po

rt

Se

rie

s

June 2005

The present English version is a translation of the originalDossier 2005 documentation written in French, which remains ultimately the reference documentation.

In order to be consistent through the various documents,while the word “storage” (“entreposage” in French) refersonly to temporary management (in terms of concept and facility), “disposal” (in terms of concept) and “repository” (in terms of facility or installation) refer to long term management of high level long lived radioactive waste.

Research on deep disposal of radioactive wastep.02 > A general interest taskp.02 > Legislative frameworkp.02 > Andra scientific objectivesp.03 > Inspections and assessments

Designing a safe and reversible disposal systemp.05 > Repository safetyp.06 > Reversibility: an essential requirement

Clay Research on a repository in a clay formationp07 > A long research programmep08 > Dossier 2005 Argile

Meuse/Haute-Marne site clayp10 > Expected properties of the rock formationp10 > Choice of argillitep10 > Meuse/Haute-Marne sitep11 > Conclusions from 10 years of research at the Meuse/Haute-Marne site

Repository installationsp.14 > Safe and reversible architecturep.15 > Disposal of B wastep.15 > Disposal of C wastep.15 > Possible disposal of spent fuel (CU)

The disposal facility in operationp.17 > From waste packages reception to their disposal in cells p.17 > Stages of the progressive closure of engineered structures

Reversible managementp.19 > Freedom of choice for future generationsp.19 > Various closure stages

Long-term evolution of the repositoryp.21 > Apprehending the repository complexityp.21 > Main evolutions expectedp.22 > Slow and limited release of radioactive substances

Repository safety and impact on manp.25 > Several evolution scenariosp.25 > Normal evolutionp.25 > Altered evolution

Granite Research on a repository in a granite formationp.28 > A global approachp.29 > Scientific co-operationsp.29 > Dossier 2005 Granite

Characteristics of French granite formationsp.30 > What properties are required for a repository?p.30 > Different types of granite formations

Repository installationsp.32 > Repository design adapted to granite fracturesp.32 > Clay seals to prevent water flowsp.32 > Waste disposal packages ensuring long-term leak-tightnessp.33 > Physical and chemical environment favourable for waste packagesp.33 > Architecture limiting the effects of heat

Results Status of progress and new perspectivesp.34 > Fifteen years of considerable progress in researchp.34 > The feasibility of a repository in a clay formation has been establishedp.35 > A repository in a granite formation is conceivablep.36 > After 2006: What are the perspectives for research on clay formation?

Contents

2

The French Atomic Energy Commission

(CEA) pursues two other avenues of re-search: partitioning of long-lived ele-ments, associated with the reduction ofthe lifetime of the most toxic ones (trans-mutation), and conditioning and long-termstorage (at surface or shallow depth).

The Waste Act stipulates the need “tocomply with nature, environment andhealth protection” and “to take the rightsof future generations into account”, i.e.,not leaving them with a pending problemwhile giving them the possibility tocontrol the process initiated. It also stipu-lates that at the end of a period not ex-ceeding fifteen years, the Governmentauthorities shall submit a global assess-ment report on these research activitiesto the Parliament, as well as a draft law.

Andra scientificobjectivesThe feasibility study for an undergroundrepository is intended to evaluate thepossibility of constructing, operating andmonitoring a reversible repository in com-plete safety for man and environment.

Within this scope, Andra mission entailsthe following roles:

• general management role, to orientresearch activities and organise thescientific and technical communityinvolved in this field.

• direct research role. For the claymedium, the Meuse/Haute-Marneunderground research laboratory loca-ted at Bure is available. For the granitemedium, since there is no undergroundresearch laboratory available in France,Andra carries out studies to assess the

ANDRA > Andra research on the geological disposal of high-level long-lived radioactive waste. Dossier 2005

Research on deep disposal of radioactivewasteA general interest taskAndra (French National Radioactive WasteManagement Agency) is the publicbody responsible for the long-termmanagement of all radioactive waste pro-duced in France. Under the supervisionof the French Ministries of Industry,Research and Environment, Andra opera-tes disposal facilities adapted for lowestlevel radioactive waste. In addition, itconducts scientific research programs tostudy the possibility of high-level or long-lived radioactive waste deep geologicaldisposal. Finally, it maintains an inventoryof radioactive waste and provides factualand verifiable information available to thepublic.

In accordance with the principle “pollu-ters pay”, Andra is financed by radioactivewaste producers (nuclear power plants,reprocessing plants, research laborato-ries, hospitals, etc.) proportionally to thevolumes produced.

It thus assumes responsibility with regardto the national authority by protectingman and the environment from the risksassociated with such waste.

LegislativeframeworkThe French Waste Act of Parliament da-ted 30th December 1991 entrustedAndra

with the task of assessing the feasibilityof the deep geological disposal for high-level, long-lived waste (HLLL waste)based on a rationale of reversibility,notably through the construction ofunderground laboratories. Two geologicalmedia are considered: clay and granite.

Aerial view of the Meuse/Haute-Marne underground research laboratory

3

potential of French granite formations.Foreign underground laboratories inSwitzerland, Belgium and Swedenprovide significant contributions forboth media. In addition, Andra conductsengineering and safety-related studies.

Inspections andassessmentsThe French National Review Board

(CNE) created by the December 1991 Actis composed of French and foreign scienti-fic experts. It examines the researchconducted by the CEA and Andra and publi-shes an annual report. For the purpose ofthe 2006 parliamentary debate, it is cur-rently preparing a general assessment re-port on the scientific results achieved so far.In addition, CEA and Andra research activi-ties are monitored and coordinated by theMinistry of Research.

This reviewing process is completed withthe regulatory intervention of the

Nuclear Safety Authority (ASN) and itstechnical support, the Institute for

Radiological Protection and Nuclear

Safety (IRSN).

What are the types ofwaste involved?

There are two categories of high-levellong-lived waste (HLLL waste).

High-level waste (C waste) accounts for1% of the volume of radioactive wasteproduced in France, but 96% of total ra-dioactivity. This type of waste consists ofnon-recyclable materials resulting from

NPP (nuclear power plants) spent fuel re-processing and gives off large amountsof heat for several tens of years. A tempo-rary storage period is therefore requiredto allow the waste to cool down prior toits possible disposal in a repository.

C waste is incorporated in a glass matrixwith high confinement properties over se-veral hundreds of thousands of years andthen poured into stainless steel drums.

Intermediate-level long-lived waste

(B waste), more varied, mainly includesmetals (fuel claddings), effluent treat-ment sludges and nuclear plant operatingequipment, with a volume and radioacti-vity both amounting to approximately 4%of the total for all the radioactive wasteproduced. B waste gives off little heat. Itis either compacted or embedded (in bitu-men or concrete), and then placed inconcrete or steel containers.

Non-reprocessed spent fuels (CU) arenot considered as waste because they docontain recoverable materials (uranium,plutonium) which could be reprocessedand then recycled. Their possible dispo-sal is nevertheless taken into accountin the studies (in case they are no longerreprocessed in the future).


Research on deep disposalof radioactive waste

Diagram of an R7T7 vitrified C waste primary package

Standard vitrified waste

container (CSD-V)

Cut-away model of a CSD-C

container showing the stack of wafers

1

C.IM.ASTE.05.0122.A

Stainless steelprimary envelope

Glass containingthe radionuclides

4

What is currently donewith this waste?

The HLLL waste generated by EDF nu-clear power plants, COGEMA fuel repro-cessing plants, CEA research centres andNational Defence activities are currentlystored in the production facilities, at LaHague and Marcoule for the most part.Despite being stored in complete safety,their toxicity and lifetime call for a speci-fic solution. This temporary situation musttherefore be replaced with sustainableand safe management methods. That isthe objective of the research conducted wi-thin the scope of the 1991 Act.

How can protectionbe ensured?

Barriers must be installed between thewaste and the environment so as to ef-fectively isolate (confine) the radioactivesubstances and chemical elementscontained in the waste.

Each waste category is associated witha specific management method and amultibarrier system adapted to the levelof radioactivity and the potential duration


Research on deep disposalof radioactive waste

of toxicity. However, the waste reposi-tory principle is always based on multi-ple safety features: conditioning (wastedrums or packages), repository structu-res (disposal cell or vault), and geologicallayer. 90% of the radioactive waste

produced in France (very low level, lowlevel and intermediate-level waste) is al-ready disposed of according to this prin-ciple in the Manche and Aube dis-tricts.

What quantitiesof waste need

to be disposed of?

Andra has compiled an inventory of exis-ting B and C waste and has also conser-vatively evaluated future production bycurrent installations based on various hy-potheses for spent fuel reprocessing.The studies cover the full range of possi-ble situations, without privileging any fu-ture industrial choice in particular.According to the reprocessing scenarios,the volume of B waste amounts to bet-ween 70,000 and 80,000 m3, and that ofC waste between 2,500 and 6,300 m3.

Aube waste disposal facility for low and intermediate-level radioactive waste

Current temporary storage facility for C waste

5


Repository safetyThe repository must protect man and theenvironment against the possible hazardsassociated with radioactive waste. Itmust also reduce at best the possibleradiological impact.

Andra has therefore adopted a safety ap-proach that privileges the robustness ofthe repository over very long periods oftime. Certain types of waste will remainradioactive for tens of thousands of years,or even longer. Andra lays particular em-phasis on assessing the uncertainties as-sociated with the future of the repository.

Based on these principles, the repositorymust fulfil three functions:

• preventing water circulation,since watercan degrade the waste packages andtransport the radioactivity contained therein,

• limiting the release of radioactive

substances by the packages and immo-bilising them in the repository as long aspossible,

• delaying and reducing the migration

of radioactive substances beyond therepository or geological layer.

Designing a safe and reversible disposal system

> Basic Safety RuleBasic Safety Rule (RFS) III.2.f was issued in 1991 by the Nuclear safetyauthority. It sets out the main objectives for a deep repository site:

> absence of seismic risks in the long term,

> absence of significant water circulation inside the repository,

> rock suitable to underground installations excavation,

> confinement properties for radioactive substances,

> sufficient depth to keep the waste safe from potential aggressions,

> absence of nearby rare exploitable resources.

Surfaceinstallations

Shafts

500m

Zone B

Zone C0

Zone C

Underground installations

Disposalcells

Basic diagram of a repository layout during operation

1

C.IM.OSES.04.0263.B

6

Reversibility:an essentialrequirementThe 1991 Act refers to reversible or irre-versible disposal. Since then, it has beendecided to firmly adopt a reversibility ratio-nale.

This reversibility requirement calls for amodest approach with regard to thescientific knowledge available at a givenpoint in time. Associated with the imple-mentation of the 'precaution' principle, itmeans a cautious management, offeringthe possibility to modify previous choicesregarding radioactive waste manage-ment. It entails a progressive approach tothe design, construction, operation andclosure of the installations, includingmeans to retrieved emplaced wastepackages, would another decision betaken.

What would deepgeological disposal

consist of?

Deep geological disposal consists of em-placing the waste in a geological layer ata depth of several hundred metres. Theobjective is to isolate the waste from manand the environment for very long periodsof time, until the radioactivity has decrea-sed and no longer means a hazard for po-pulations. The repository therefore confi-nes the radioactive substances containedin the waste.

Designing a reversible disposal systemallows continuous control over the pro-cess. In particular, it allows for the possi-bility of retrieving the waste whenevernecessary or in the event that other mana-gement choices are made. Nevertheless,ultimately, the repository must progres-sively become a passive installation,without monitoring or human interven-tion, while maintaining the same perfor-mance as regards safety.


How long wouldthe disposal system

remain reversible?

Reversibility is possible for at least seve-ral centuries, with no intervention otherthan standard maintenance and monito-ring tasks.

The reversible disposal system can beinitially managed as a temporary storage,with waste emplacement and retrieval. Itcan also be closed progressively. Andrahas defined several levels of reversibility,i.e., closure in several stages. It has esta-blished simple and robust repositoryconcepts and has identified durablematerials. It has developed processes tofacilitate the possible retrieval of thewaste packages and has designed theunderground installations as independentmodules to allow flexible and open-endedmanagement. An observation programme(deformation, temperature and pressuremeasurements, implementation of datatransmission networks inside the engi-neered structures) has been developed toensure the technical feasibility of thereversibility process (backward). As alower level of reversibility is chosen, wasteretrieval operations are still possible butwill become more complex.

Designing a safe and reversible disposal system

Access shaft of the Meuse/Haute-Marne underground research laboratory

7


For nearly 15 years, Andra has beenconducting a major scientific programmeto acquire knowledge on a repository in aclay formation. The most important re-search tool is the Meuse/Haute-Marneunderground laboratory, located at adepth of 490 metres in the heart of a verystiff (indurated) clay formation (argillite).This geological layer, referred to as theCallovo-Oxfordian layer, has been verystable since its formation more than 150million years ago.

After drilling numerous boreholes since1994, Andra has studied the Callovo-Oxfordian layer (as its surrounding layers)within the rock formation (in situ), in thelaboratory shafts. Experimental drifts ata depth of 445 metres are in servicesince November 2004, with variousexperiments intended to confirm sitedata previously acquired.

In addition to boreholes drilled from thesurface, sample analyses and under-ground studies, the research programmerelies on the work conducted in several underground laboratories abroad, parti-cularly the Mont Terri laboratory inSwitzerland.

Andra has thus been able to reconstructthe geological history of the Meuse/Haute-Marne site so as to consider its futureevolution. This representation serves asthe basis for simulations to assess theperformance of the disposal system.

In addition to geological studies, theresearch programme covers four comple-mentary issues: waste packages and ma-terial behaviour to understand the reposi-tory evolution over very long periods oftime, repository design (waste conditio-ning, repository architecture, operating

Research on a repositoryin a clay formationA long research programme

> Over 10 years of research

> 1,300 km of seismic profiles studied

> 27 deep boreholes

> 4.2 km of cored boreholes

> 2.3 km of argillite cores

> 30,000 samples taken

> 100 m of horizontal drifts excavated

> 40 m of experimental drifts

> 40 boreholes drilled from the drifts andequipped with 350 sensors

2

CL

AY

Drilling platform

Geologists examining the wall of the

experimental drift at – 445 m

Seismic survey through deviated boreholes

Height abovesea level

(m)Depth

(m)

8

and reversible closure methods), reposi-tory evolution, and long-term safety.

Dossier 2005ArgileThe Dossier 2005 Argile submitted to theFrench authorities consists of five refe-rence knowledge documents containingall data currently available on respectivelythe geological medium and the biosphere,the materials (steel, concrete, etc.), the ra-dioactive substances, waste behaviour inthe repository and the inventory of HLLLwaste produced and yet to be produced byexisting nuclear facilities.


Vertical exaggeration : 4x

Research laboratory

Gondrecourt le Château

Joinville

Gondrecourt graben

Ligny-en-Barrois

Marne graben

Marne river

Ornain river

Saulx river

Savonnières-en-Perthois syncline

Poissonsfault

Cal

lovo

-Oxf

ord

ian

0 5 km Scale :

SURFACE GEOLOGY

LITHOLOGICAL CROSS SECTION

Lower CretaceousTithonianKimmeridgianOxfordianlimestone

Sandstone and evaporites (Trias)Sandstone and clays (Permian)Undifferentiated basement

LimestoneMarls and clays

Alluviums orPlio-Quaternary

Meuse/Haute-Marneunderground researchlaboratory

TERTIARY

UPPER CRETACEOUS

LOWER CRETACEOUS

JURASSIC

TRIAS

PERMIAN

PALEOZOIC BASEMENT

Scale

Research on a repository in a clay formation

Core sample library of the Meuse/Haute-Marne research laboratory

3D geological block diagram of the Meuse/Haute-Marne sector

D.PL.FSTE.04.0020.B

9


Research on a repository in a clay formation

Based on these data, Andra presents itsanalysis in three volumes:

• proposal of a repository architectureand management method gearedtowards safety, industrial feasibility andreversibility,

• analysis of the repository evolution,considering all thermal, hydraulic,mechanical and chemical phenomena inthe environment over a period of onemillion years,

• repository safety assessment and riskanalysis, in both normal and incidentalsituations.

Which scientific organi-sations has Andra colla-

borated with?

Andra has worked with a large number ofFrench partners notably: French GeologicalSurvey (BRGM), French Atomic EnergyCommission (CEA), French National Centrefor Scientific Research (CNRS), Paris Schoolof Mines (Ecole des Mines de Paris),French Petroleum Institute (IFP), NationalInstitute for Industrial Environment andRisks (INERIS), National PolytechnicInstitute of Lorraine (INPL), and approxima-tely some other 100 laboratories.

Seven groups of laboratories have beenestablished according to Andra researchthemes: metal corrosion, clay, concrete,thermo-hydromechanical coupled pheno-mena, radioactive substances, geome-chanics and bio-geoprospective studies.Likewise, three research groups havebeen created within the scope of theelectronuclear cycle backend programmeconducted by the CNRS (FORPRO,PARIS, MOMAS).

At the international level, Andra has colla-borated with its Swiss, Spanish, Germanand Belgian counterparts and participatedin joint programmes with international or-ganisations such as the EuropeanCommission, the IAEA (InternationalAtomic Energy Agency)…

Does Andra submit itsresults to international

experts?

In 2001, Andra prepared a first synthesisof its research activities and results. Thisreport was submitted for critical reviewby a group of independent internationalexperts under the OECD/NEA aegis, whostressed the relevance of the knowledgeacquired and the interest of the results.A second review is scheduled in 2005.

> Meuse/Haute-Marne laboratory:chronological milestones

1992 - Work on repository design and identification of knowledge to beacquired.

1994-96 - Geological survey work on two clay sites: the first one strad-dling the Meuse and Haute-Marne districts and the other one in theGard district.

1997 - Initial selection of repository concepts.

1998 - Selection of Meuse/Haute-Marne site by Government decisionand definition of the experimental programme, selection of conceptswith a broad range of technical solutions.

1999-2001 - Acquisition of further knowledge concerning the Callovo-Oxfordian layer and start of laboratory shaft sinking.

Late 2001 - Dossier 2001 Argile providing an intermediate synthesisof knowledge acquired.

2002 - Based on the Dossier 2001 Argile, revision of scientific programmefor 2002-2005 and selection of repository concepts (waste packages anddisposal cells).

2003-2004 - Borehole drilling on and around the laboratory site.

October 2004 - Auxiliary shaft reaches - 490 metres.

November 2004 - Availability of the experimental drift at - 445 metres inthe main shaft.

Since February 2005 - Excavation of experimental drifts at the bottomof the auxiliary shaft.

2

CL

AY

10

Expectedproperties of theformationThe geological medium must be very sta-ble in the long term, i.e., with limited ex-posure to earthquakes and erosion.

While being deep enough in order toavoid surface disturbances, the clay layermust present a homogeneous geologi-cal structure and mineralogical composi-tion. Water flow in the rock must be low,as it constitutes the main alteration fac-tor and the major transport vector for ra-dioactive substances. Finally, chemicalstability over time and suitability for exca-vation are also two essential criteria.

Choice of argilliteArgillite has excellent properties. It is astiff (indurated) sedimentary rock withvery low permeability. Radioactive or non-radioactive elements dissolved in watermove very slowly through this rock be-cause their migration is mainly due totheir own motion (diffusion), not to theirtransport by flowing water (advection). Inaddition, argillite has the ability to retain alarge number of chemical elements. Itprovides a stable chemical environmentand presents a good capacity to absorbchemical perturbations. Finally, argillitehas a good mechanical resistance whileremaining sufficiently deformable in thelong term to adapt to movements that occur very slowly over time.

Meuse/Haute-Marne siteCallovo-Oxfordian argillites

The sector north of the Haute-Marne andsouth of the Meuse constitutes a geologi-cally simple domain of the Paris basin,with a succession of horizontal layers of

limestone, marl and clay rock depositedin ancient oceans.

The layer studied is a clay rock 155 millionyears old, at least 130 metres thick andlocated at a depth of between 400 and600 metres, referred to as the Callovo-Oxfordian argillites.


Meuse/Haute-Marnesite clayThe geological medium is at the core of the repository system.It must ensure the very long-term confinement of radioactivesubstances to prevent their migration into the environment.

Geological map of the Meuse/Haute-Marne sector

KEY

Geology

District boundaryFault

Tithonian and Kimmeridgian limestone terminalUpper and Lower KimmeridgianundifferentiatedLower Kimmeridgian andOxfordian limestoneArgillaceous Callovo-Oxfordian

Callovian and Dogger limestone

Recent and old alluvium

Argillo-arenaceous Cretaceous

Laboratory

11


Major experiments carried out at the

laboratory

From the surface:• analysis of regional seismic profiles,• drilling of deep boreholes and measure-

ment of mechanical properties, permea-bility and diffusion,

• drilling of deviated boreholes to survey,at large scale, the geological layers,

• 2D and 3D geophysical survey cam-paigns (underground auscultation withseismic waves),

• hydro-geological monitoring,• seismic (earthquake) monitoring network.

Within the rock:

• While sinking the shafts: layer survey, wa-ter collection and flow-rate measure-ments in limestone layers overlying theCallovo-Oxfordian formation, wall defor-mation measurements, real-time monito-ring of rock mechanical behaviour (viasensors), and assessment of rock da-mage by excavation.

• Inside the drifts: wall deformationmeasurements, thermal conductivity

measurements, monitoring of chemicalperturbations, permeability and diffusionmeasurements for water and radioactivesubstances, and performance tests ongrooves filled with swelling clay1.

Conclusions from10 years ofresearch atthe Meuse/Haute-Marne site After 10 years of research, Andra has ac-quired data confirming that the Callovo-

Oxfordian layer of the Meuse/Haute-

Marne site has favourable properties

for an HLLL waste repository:

• the geological environment is stable:very low seismic risk,

• the clay layer is regular and homoge-neous over a large surface area.It does not present any fault,

• the Callovo-Oxfordian has a low per-meability: very low water flow and argil-lites with favourable properties to trapand retain radioactive substances overlong periods of time,

• the rock can withstand mining excava-tion work,

• its characteristics are compatible withthe reversibility requirements,

• the impact of the engineered materials(cement, concrete, metal, etc.) is verysmall and limited to the immediate sur-rounding of the engineered structures,

• water flow in the Callovo-Oxfordian sur-rounding layers is very slow,

• the results obtained in the undergroundlaboratory can be transposed to a 200 km2

area.

Could the repositoryundergo an earthquake?

The deformations associated with tec-tonic plate movements have remainedsmall for the past 150 million years, as inthe rest of the Paris basin. They are es-sentially limited to the Gondrecourt andMarne grabens, on the boundary of thesector studied. Between these faults, theCallovo-Oxfordian layer is regular andpractically flat.

Available data confirms that the region hasvery low seismicity. However, in a cau-tious approach, the engineered structuresproposed for the repository have beendesigned to withstand a hypotheticalearthquake of magnitude 6.1 ± 0.4 at 6 kmfrom the site (most pessimistic hypothe-sis). Finally, the effect of an earthquakewould be, in any case, very low under-ground.

MAIN SHAFT

EXPERIMENTAL DRIFT

SURFACE INSTALLATIONS

AUXILIARY SHAFT

MAIN LEVEL

TECHNICAL DRIFTS

EXPERIMENTAL DRIFTS

- 445 m

- 490 m

NORTH

MEUSE/HAUTE-MARNEUNDERGROUND RESEARCH LABORATORY

FORECAST STATUS IN 2006

1 - Very low permeability material that swells as ithydrates.

Meuse/Haute-Marne site clay

Seismic vibrator trucks (Vibroseis) used

during the geophysical survey campaign

Experimental drift (– 445 metres) of the

Meuse/Haute-Marne laboratory

Scientific team of the Meuse/Haute-Marne

laboratory

2

CL

AY

Are there natural resour-ces nearby that might be

useful in the future?

The site presents no rare natural resour-ces to be preserved. In particular, theaquifer layer located beneath the labora-tory does not have the necessary proper-ties to eventually become an exploitableresource for consumption or geothermalapplications.

Is the absence of faultsconfirmed?

This aspect has been particularly studiedduring the geophysical campaigns.

The 200 km2 explored north and nor-theast of the laboratory revealed no faultsin the sector studied. The only knownfaults are located outside this sector:Marne fault (oriented north-northwest)and Gondrecourt graben (oriented nor-theast), forming the western, southernand eastern boundaries of the sector.

In the Callovo-Oxfordian layer, none of theboreholes, meaning a drilled length of2300 meters in all, intercepted a secon-dary fault. Only a few microstructureswith a maximum size of a few centime-tres have been surveyed. They are all

clogged and do not modify the confine-ment properties of the layer.

Could radioactivesubstances migrate out

from the repository?

The Callovo-Oxfordian layer has a verylow permeability. Water flow is thereforevery limited, preventing the possibletransport of radionuclides: a drop of water

would move away a few centimetres in100,000 years. In addition, the layer has alarge smectite content, a mineral thattends to immobilise elements dissolvedin water. Finally, the chemical composi-tion of the interstitial water in the rockcauses various radioactive substances toprecipitate in solid form, thus preventingtheir dissolution in water. All of these fac-tors result in most of the radioactivitybeing trapped within the repository.

A few radioactive substances could ne-vertheless migrate in the very long term(not before a hundred thousand years),but with no impact on man and the environment.

Could the excavationof the repository drifts

damage the host forma-tion surrounding the repo-sitory?

The laboratory excavation monitoring hasshown that these argillites are very stiff,deform little and slowly.

Drifts excavation creates a damaged zonearound the excavated structure, suscepti-ble of constituting a water pathway. At a depth of 450 metres, there is practi-cally no fracturing, but microfissuring mayoccur around the engineered structures.For example, in the case of a 10 metres

12



Seismicity

around the

Paris basin

Assessment of the damaged zone based on microseismic measurements

SEISMICITY

The Meuse/Haute-Marne siteis marked by a triangle

MOHO DEPTH (according to Dézes and Zieger, 2004)

13



diameter structure, the microfissuredzone may reach a few metres. The rockproperties are little affected. In particular,its very low permeability is preserved.Moreover, preliminary results suggestthat these fissures and fractures tend toheal with time.

Although the introduction of air for ventila-tion contributes to drying the rock, possiblyweakening it, analyses of models and an-cient structures show that this phenome-non is slow in argillites and does not ex-ceed the thickness of the damaged zone.

Can the rock close upthe repository preventing

access to the wastepackages?

Callovo-Oxfordian argillite can deformwith time, but this process is very slow.The displacement of the walls of a repo-sitory structure would amount to lessthan a few centimetres after 1000 years.This provides stable and robust enginee-red structures over long periods of time,thereby guaranteeing reversibility.

How would the heatreleased by the waste

packages affect the rock?

Certain types of waste give off heat. It istherefore important to examine rockbehaviour with respect to heat. Up to ap-proximately 70°C, the argillites remainpractically unaffected. They can withstandsuch temperatures without significant al-teration for approximately 10,000 years,and higher temperatures for shorter pe-riods. These estimates can be used to de-fine the acceptable temperatures for a re-pository. The maximum value adopted forthe temperature in the rock is 90°C, and70°C beyond 1000 years.

What would bethe behaviour of the

engineered materials inthe geological medium?

In the chemical environment of the repo-sitory, concrete degradation takes severaltens of thousands of years. Once therepository is closed, metallic materialscorrode very slowly (absence of oxygen),producing minerals that stabilise theirdegradation. The studies show also thatthe chemical impact of corrosion on theargillites is very limited.

Do we have informa-tion concerning water

flow in the layerssurrounding the argillites?

In the layers surrounding the Callovo-Oxfordian, the overall waterflow is ho-rizontal and directed from the plateauxlocated south and east of the site towardsthe centre of the basin. Based on the re-sults of the topographic analysis for thepast two million years and the predictionsregarding climatic changes, it is possible toappraise possible variations in water flowover the next 500,000 to 1 million years. Itappears that flow directions will undergorelatively few changes.

To which zone can theresults obtained in the

underground laboratorybe transposed?

Andra has defined a geographic zonewhere the properties of the argillites aresimilar to those found at the laboratorysite. After exploring a 700 km2 areaaround the laboratory, using precise map-ping methods, it has been determinedthat the transposition zone extends overapproximately 200 km2 to the north andwest of the laboratory.

Why does Andra studyclay formations in

Switzerland? Between 1996 and 2005, Andra has car-ried out experiments in the Mont Terri la-boratory in Switzerland. The Mont Terri ar-gillites are somehow similar to theCallovo-Oxfordian ones. Andra has testedvarious tools and methods, acquiredscientific data on clay behaviour, valida-ted models and performed full-scale engi-neering tests. A major result has beenachieved: the observations obtained atsample scale remain valid at larger scales.

Shaft sinking

Reading of experimental data

2

CL

AY

C waste repository zone

C0 waste repository zone

B waste repository zone

Operating unit

Disposal cell

Module

Shafts

Connecting drifts

C wasterepositorysub-zone

C.IM.0SES.04.0596.C_E

The repository is located on a single le-

vel, in the middle of the geological

layer, so as to benefit as much as possiblefrom the thickness of the argillite barrier.

It is organised into distinct zones

according to package type (B, C, CU),

separated from one another and subdivi-ded into modules. The modules areconstructed and put into service as theneed arises.

The engineered structures are designed

to minimise mechanical disturbances.

Their architecture is simple, with a gene-rally half-circular cross-section and theirdimension is limited. The cells are spacedapart and oriented parallel to the directionof maximum stress in the rock so as to

not interact mechanically. A lining sup-ports the engineered structures for seve-ral centuries, and void spaces within thecells are limited.

The engineered structures receiving

the C waste and spent fuel packages

are designed to limit the disturbances

associated with the large amount of

heat given off, by means of sufficientspacing between cells and a suitable ar-rangement of the packages. The tempera-ture must remain below 100°C in contactwith the packages and 90°C within therock while the architecture is designed tokeep the temperature below 70°C, oreven much lower, beyond 1000 years.

The engineered structures are arranged

in a dead-end fashion so as to limit wa-

ter flow. In the event that it is decided toclose them, they will be sealed with low-permeability swelling clay plugs.

The cell and package materials

(concrete, steel, etc.) are chosen to last aslong as possible and preserve a physico-chemical environment that retains the ra-dioactive substances.

The reversibility requirement is integra-

ted as of the repository design phase. Itentails privileging the use of durable ma-terials, maintaining the technical possibi-lity to retrieve the packages, and organi-sing the repository operation or closure invarious stages and in a modular manner.

14


Repository installations

Andra has designed a simple and robust repository:a modular architecture grouping together packages of the samecategory and allowing flexible operation.

Safe and reversible architecture

General organisation of the repository

C.IM.OSES.04.0596.B

15


Disposal of BwasteB waste gives off very little heat. Thepackages delivered by waste producersare placed in high-performance reinforcedconcrete containers to constitute dispo-sal packages. These parallelepiped contai-ners are approximately 1.5 to 2 metreshigh and weigh 6 to 25 metric tonnes.They are designed to last several centu-ries and ensure a good resistance to falls.Prototypes have been used to demons-trate their feasibility and to perform full-scale tests.

The containers are placed in concrete dis-posal cells 250 metres long and 12 me-tres in diameter. The cells are equippedwith a high-performance concrete circularlining with guaranteed stability for severalcenturies. A disposal chamber adapted tothe geometry of the stacked packages isarranged inside the lining.

Disposal of CwasteC waste is confined in a glass matrix whe-rein the radioactive substances are en-trapped. This type of waste gives off largeamounts of heat. High temperatures canaccelerate the dissolution of the glass incontact with water. In order to preventwater ingress on the glass during its hightemperature phase, each C waste packagedelivered by the waste producer is placedin a leak-tight cylindrical steel container1.3 to 1.6 metres long and 60 cm in dia-meter (weight: 1.7 to 2 metric tonnes).The thickness of the container (approxi-mately 5 cm) is calculated to withstandcorrosion and ensure leak-tightness for atleast 4000 years.

The C waste disposal cell is a micro-tun-nel 40 metres long and 70 cm in diame-ter. It is designed to preserve the contai-ner from corrosion and is not ventilatedin order to limit the ingress of oxygen fa-vouring corrosion. In order to meet thetemperature limits, the cells are spacedapproximately 10 metres apart and eachone receives a small number of packages(6 to 8).

Possible disposalof spent fuel (CU)Although the geological disposal of spentfuel is not currently foreseen, Andra hasconsidered this possibility so as to coverall possible configurations.

The concept is similar to that used for Cwaste, but it has been decided to emplaceswelling clay between the steel containerand the argillites so as to take into accountthe large amounts of heat given off byspent fuel. Their thick steel container (thick-ness slightly over 10 cm) can withstand cor-rosion for 10,000 years. The container itselfhas been designed to prevent the occur-rence of an uncontrolled nuclear chainreaction.Containers receiving several spent fuel as-semblies weigh 43 metric tonnes, whilethose only receiving one assembly weigh 8to 10 metric tonnes.The spent fuel disposal cell is approxima-tely 45 metres long and 2.5 to 3 metres indiameter.

In order to avoid heat accumulation, the cellsare excavated approximately 20 metresapart and number of packages per cell isreduced (3 or 4).


C waste disposal cell

Prototypes of standard B waste containers

0.57mto 0.64m

1.30m to 1.60m

Primary package

Lid

Steel over-pack

Disposal package

SleeveSeparator

Ø excavated: 0.7m approx.

Length: approx. 40m

2

CL

AY

C.IM.OSES.04.0269.C

4

31

How would the occupa-tional safety be ensured

inside the repository?

The underground architecture is designedso that disposal operations can be perfor-med simultaneously with the construc-tion of new cells.

Nevertheless, the design provides for aseparation of activities: to avoid all risksof interference, the traffic associated withconstruction and closure operations is se-parated from that associated with dispo-sal activities (which present a radiologicalhazard).

If necessary, personnel can be evacuatedvia the access shafts and rescue teamscan quickly access underground. Asmoke evacuation system is also plannedin case of fire. The drift network meets allregulatory safety requirements.

How many shafts wouldconnect the repository to

the surface?

Four shafts connect the surface to the un-derground installations. These shafts areequipped with reliable systems used inthe mining industry and are specialisedaccording to function: waste packagetransfer, personnel transport, service(transport of muck and large equipment),and exhaust (air return ventilation).

16


Unit inoperation

Unit underconstruction

Sub-zone 1Under construction-operation

}Units waitingto be closed

Sub-zone 1Under construction-operation

Construction of zone 2Access infrastructure

}Sub-zone 1Under construction-operation

Construction ofsub-zone 2Access infrastructure

re

Sub-zone 1End of operation

Sub-zone 2Beginning ofconstruction


Demonstrator of spent fuel disposal package

Surface facilities (project)

Phases in the construction and operation of a C waste repository zone

Which installationswould be built at

surface ?

The surface installations cover approxima-tely 100 hectares comprising various areas:

• nuclear area, where waste packagesdelivered by waste producers are recei-ved and subsequently conditioned indisposal containers,

• industrial area, with the technical facili-ties and materials required for under-ground works,

• administrative area.

In addition, a specific area could be set upto receive the excavated muck that mightbe used as drift backfill.

C.IM.OSES.04.0528.B

2

17


In the surface installations, waste packa-ges delivered to the site are removedfrom their transport casks and placed incontainers: then they are inspected, andstored temporarily. Most of these opera-tions are performed by remote-controlleddevices in shielded compartments, withoutthe presence of humans.

Each container is then placed in a shiel-ding cask ensuring the radiological pro-tection of personnel and then transferredto the underground installations.

For B waste packages, a remote-control-led carriage extracts the container fromthe cask and emplaces it in the cell. For Cwaste packages, a mobile robot integra-ted in the cask pushes the container intothe cell. This process is being testedwithin the scope of a European project(ESDRED).

Spent fuel packages of small diametercan be emplaced using the same me-thods, whereas those with a larger dia-meter (i.e., containing 4 fuel assemblies)are lifted by air cushion support pushedby a self-propelled carriage. After a suc-cessful preliminary test carried out inSweden, this process is also currentlybeing tested within the scope of theabove-mentioned European project.

The disposal facilityin operationThe construction of engineered structures, the industrialoperation and closure of repository installations areperformed progressively, and can be silmutaneous. This flexibility allows for changes in repository design ormanagement methods according to the lessons learnt feedback. In addition to these main activities, specific maintenance and monitoring activities are required toensure the reversible management of the disposal system.

From waste packages reception to their disposal in cells

Scheme of a disposal package transfer

2

CL

AY

Stages of the progressiveclosure of engineered structures

In keeping with the reversibility require-ment, the repository would be closed instages, i.e., disposal cell sealing, backfil-ling and sealing of drifts and then shafts.Sealing prevents water circulation insidethe repository. Drift backfilling limitsdeformations in the geological medium.

B waste disposal cells are first closedwith a radiological protection concretewall ensuring occupationnal safety, andthen by approximately 30 metres ofswelling clay in order to prevent waterflow. C waste and spent fuel disposalcells are closed with a metal plug ensu-

Inserting the packageinto the transfer cask

thoughput

The transfer caskcarried by the surface transfer vehicle

Insertion into the cage ofthe packagetransfer shaft

Transfer caskexiting the shaft and transferred through drifts

Lifting down the transfer cask in the shaft

Dockingthe transfer cask

Packageemplacement

Depositing the cask

Transfer cask moved by shuttle

C.IM.OSES.04.0430.B

ring radiological protection, to which is added a 3 metre swelling clay plug.The drifts are then backfilled and sealedsimilarly to B waste disposal cells. Theshafts are filled with concrete at the baseand sealed with swelling clay to a heightof 30 metres. They are then backfilled withargillite from the site, with a swelling clayinsulating plug (10 to 15 metres) at eachporous level.

What are the risks asso-ciated with construction

and operation? How to prevent them?

The main risks are those encountered inindustrial, mining and tunnelling activities:fire, handling and traffic accidents, fallingblocks, electrical hazards, etc., calling forconventional preventive measures.

All aspects of repository operation aredesigned to prevent the risk of radioac-tive exposure. These measures which arethe current standard in the nuclear industry,consist in protection screens, remote-controlled systems and robots, confine-ment of radioactive materials, limiting therelease of radioactive gases and monito-ring the absence of contamination.

The risk of criticality (uncontrolled nuclearchain reaction) has been taken into ac-count for spent fuel packages, whereasB and C waste packages do not containthe quantities of materials required forsuch a reaction. To prevent this risk andsimilarly to existing storage facilities,packages and cells are respectively suffi-ciently apart from each other and drytreatment processes are used, since wa-ter increases the reactivity of the radioac-tive elements.

In order to prevent the packages frombeing damaged by falling, handlingheights are limited and the resistance ofthe containers and casks is over-designedwith respect to the possible fall height.The integrity of the primary package isthus preserved in the event of a fall. Therisk of falls in shafts has been conside-

red, despite the fact that current miningshaft installations make it highly unlikely.In the event of a fall, a shock absorber li-mits damage to the cask. C waste andspent fuel containers placed inside thecask withstand the shock, whereas Bwaste containers may be slightly fissu-red, without affecting the primary packa-ges. Nevertheless, to avoid this risk, fil-ters are planned so as to trap radioactivesubstances possibly released into the airof the shaft.

What about the risk of fire?

The risk of fire has been assessed in a de-tailed study. The simulation results showthat the repository design (with driftsconnected together at regular intervals)allows personnel to escape away from the

fire, quickly access a parallel drift suppliedwith fresh air and reach safely the surface.

In addition, the disposal packages wouldnot suffer damage possibly leading to thedissemination of radioactive substances.

What radioactive dosewould a person working

inside the repository beexposed to?

The estimated dose would be far belowthe current regulatory limits, whichare 20 mSv/year2 for personnel and1 mSv/year for the public.

18


2 - 1 mSv = one milli-Sievert. The Sievert is the unitto measure radioactivity effect on human beings

The disposal facility in operation

C.IM.0SES.04.0460.B

Transfert du colis en position basse

Disposal of packages in a B waste disposal cell

C.IM.OSES.04.0460.B

19


Freedom ofchoice for futuregenerationsThe approach to reversibility proposed byAndra can be defined as the possibility tomanage the repository in a flexible mannerand in stages. The objective is to leave fu-ture generations the freedom to make deci-sions concerning repository management.

The repository design (modular architec-ture, simplified operation, dimensioningand choice of durable materials, etc.) isintended to offer the widest possiblerange of choices. Reversibility means thepossibility to retrieve emplaced packages,to intervene in the disposal process andto modify the design of the engineeredstructures.

Various closurestagesAfter waste package emplacement:

cells are not sealed but closed with devi-ces protecting the personnel. All under-ground infrastructures remain accessible.

After cell sealing: cells are sealed with aswelling clay plug. Cell heads remain ac-cessible. Due to the slowness of defor-mation processes and the absence of wa-ter for several centuries, cell linings aresubject to little degradation.

After closure of a module (several cells):access drifts to C waste and spent fuelmodules are backfilled with argillite, butconnecting drifts associated with themodule remain accessible. B wastemodules - since there is only one cell permodule - are not concerned. Stability ofengineered structures is ensured in thevery long term.

Reversible management

After closure of a repository zone: driftswithin the zone are sealed and backfilled.Main drifts remain accessible.

Post-closure: this stage begins afterbackfilling and sealing of main drifts andthen shafts. It corresponds to the end ofthe disposal process. The installation be-comes passive, which means that it

Metallic plug Concrete plug

Permanentsleeve

Clayplug

continues to ensure waste confinementwithout human intervention.

This progressive process is not the onlyone. It is a possible option, itself providingflexibility.

1

2 3

4

5

- Cells are filled with packages

- Cell heads are accessible

- Disposal cells are sealed- Secondary connecting drifts are accessible

- Secondary connecting drifts are backfilled - Main connecting drifts are accessible

- Main connecting drifts are backfilled - Shafts are sealed

Possible stages in the operation and progressive closure of a repository

Sealed C waste disposal cell

2

CL

AY

C.IM.OSES.05.0074.A

C.IM.OSES.04.0624.A

20


Disposalcell

Ventilationreturn drift

Small drift

Borehole

Small drift

Radiologicalprotection wall

Road header

Air returnduct

> Monitoring the repository as input for the decision making process

Choices concerning repository management (maintaining the same level of reversibility, returning tothe previous level, switching to a lower reversibility level, etc.) are based on a scientific understanding of therepository evolution over a period of several centuries.

An observation programme will be implemented to obtain feedback intended to improve repository design andmanagement.

Measurement devices (deformation, temperature, pressure, etc.) and data transmission networks will be placedin instrumented observation cells, shafts, drifts, seals and backfills as of their construction. These devices willneed to operate for long periods in a difficult environment (radiation and temperature).

The experience acquired in the field of civil engineering provides the best practice guidelines: choice of high qualitytools, optimal distribution of redundant observation systems, integration of these means as of the repository designphase.

Reversible management

How long would it bepossible to access the

waste?

Andra does not set a predetermined dura-tion for reversibility.

One key issue of reversibility is the me-chanical stability of the disposal cells,which will last for at least 200 to 300years without specific maintenance.Given the safety margins adopted, thecells should remain stable even longer.The observation programme will allow forregular reassessment of their lifetime.

The final stage is the mechanical ruptureof the cell lining. Beyond this stage, miningoperations and specific radiological protec-tion measures need to be implemented inorder to retrieve the waste packagesblocked by the geological formation.

To extend this duration, specific technicalmeasures need to be adopted (thoroughmaintenance, reinforcement of engineeredstructures, reconstruction, etc.).

How might wastepackages be effectively

retrieved?

The equipment and methods used toretrieve the waste packages are similar tothose used to emplace them. However,conditions vary depending on the stage atwhich the decision is made. For example,if a seal has already been implemented,it must be deconstructed in order to accessthe disposal cell again and cell re-equipmentis necessary to achieve package retrieval.

B waste disposal cell deconstruction process

9.9.2

Backfilledventilation ducts

Concrete blocks

Swelling clay plug

Connectingdrift

Seal abutment (concrete)

C.IM.OSES.04.0165.B

C.IM.OSES.04.0621.B

21


Apprehending the repository complexity A repository constitutes a complex sys-tem comprising numerous components(waste packages, disposal structures,geological medium) and whose evolutiondepends on various phenomena (ther-mal,hydraulic,chemical and mechanical).Studies have provided an assessment ofthe conditions in which radioactive subs-tances might be released by the wastepackages and possibly migrate into theenvironment.

Main evolutionsexpectedHeat will have little impact

The repository is designed to limit thetemperature to 90°C at all points.Maximum temperatures are reachedafter one or more tens of years, andapproximately 1000 years (for C wastedisposal cells) and 6000 years (for spentfuel disposal cells) are required to returnto approximately 40°C. These durationsare much shorter than the time requiredfor deterioration of the waste packages:therefore heat has little effect on therelease and transfer of radioactive subs-tances. Moreover, the increase in tempe-rature does not modify the mineralcomposition of the argillites or their confi-nement capabilities.

Hydraulic evolution with controlled

consequences

The repository disturbs the initial hydraulicequilibrium of the geological formations.These disturbances remain limited to therepository and the Callovo-Oxfordian layer.Between 100,000 and 200,000 years, anew hydraulic equilibrium is establishedand chemical and mechanical processesdevelop more significantly.

After a few hundred thousand years and upto a million years, climatic changes and ero-sion will progressively modify the directionof water flows in the layers surrounding theCallovo-Oxfordian, orienting them towardsthe natural outlets north and west of theMeuse/Haute-Marne site, without chan-

ging flow velocity, which will remain verylow. The Callovo-Oxfordian layer is too deepto be affected by erosion.

A progressive mechanical evolution

The mechanical effects associated withthe repository are limited to the Callovo-Oxfordian layer, within a few metres zone.The excavation damaged zone (EDZ) ischaracterised by the appearance ofmicro-fissures whose density decreasesas the distance from the wall increases.Slight fracturing may also occur, depen-ding on the depth and orientation withrespect to pressure stresses within therock. The calculation results indicate thatfracturing does not occur at a depth of

Long-term evolutionof the repositoryA reversible disposal system can be closed in the event that such adecision is taken. It must then be possible for the repository to evolvesafely in the long term, without human intervention. Andra has thereforestudied the evolution of the repository according to various scenariosto make sure that its impact on the environment is very low in all cases.These studies are based on current scientific and technological know-ledge and take into account all kinds of uncertainty, from repository closure up to one million years in the future.

Evolution of the excavation damaged zone (EDZ)

2

CL

AY

22


Long-term evolution of the repository

Three barriers prevent the release ofradioactive substances: first barrierconstituted by the waste packages,second barrier constituted by the dispo-sal cells, and third barrier constituted bythe geological medium.

Waste package degradation will takeplace very progressively over time:package corrosion, concrete degradation,glass matrix dissolution depending onthermal and hydraulic conditions.

It will therefore be possible for radioac-tive substances to migrate out from thepackages. These substances will then be

either retained or dissolved, depending onthe chemical environment encountered.

The slowness of the diffusion process de-lays migration, and most of these subs-tances eventually disappear through na-tural radioactive decay. Only a few mobileelements such as iodine 129, chlorine 36,caesium 135, selenium 79 and carbon 14migrate significantly outside the cells.

These elements then diffuse very slowlyinto the argillaceous rock.

Slow and limited release of radioactivesubstances

> Mobility of radioactive substances

Radioactive substances are grouped into three families according totheir solubility and retention in a clay medium:

> Mobile elements (ex: iodine, chlorine)

> Medium mobility elements (ex: caesium)

> Low mobility elements (ex: uranium, plutonium)

The geological repository would mainly comprise medium and lowmobility elements.

Schematization of the Callovo-Oxfordian argillite texture and porosity

500 metres and is moderately initiated atapproximately 600 metres. At the scaleof several thousands to tens of thousandsof years, the evolution of the argillitestends to heal these fractures.

Chemical evolution will be slow and

limited

Concrete degradation and metal corro-sion are very slow processes. Chemicalperturbations are limited to a few metresat the most, which is little in comparisonwith the thickness of the geological layer.The retention properties of the argilliteswill therefore be preserved.

Waste package deterioration (and thesubsequent release of radioactive subs-tances) also takes place very slowly, overperiods of up to several hundreds of thou-sands of years for C waste packages.

1 2 3

water (various types)

10 nm

ClayClay

Bioclastic matter

Quartz grain

100 nm 1 µm



How long wouldradioactive substances

be retained in the wastepackages?

Non-gaseous radioactive substances re-main confined for periods of up to severalhundred thousand years, depending onthe waste packages. The very small quan-tities of gaseous substances released byB waste packages do not exit the reposi-tory, with the exception of carbon 14,which dissolves in water or disperses intothe geological medium (due to the parti-cularly small quantities released).

The ingress of water on B waste packa-ges leads to the progressive release of

1 & 2 : clay-water system ; 3 & 4 : mineralogical assembly

radioactive substances over a period ofseveral tens to several thousands of years(neglecting the possible role of theconcrete containers). According to cau-tious estimates, the C waste and spentfuel containers remain leak-tight for 4000and 10,000 years, respectively. After theirdegradation, water comes into contactwith the glass and spent fuel assemblies,and these substances dissolve for severalhundreds of thousands of years.

How long would ittake for them to reach

the environment?

At least one hundred thousand years arerequired for the most mobile radioactivesubstances to be transferred to the boun-daries of the Callovo-Oxfordian layer.

Half of the elements moves downwards(Dogger layer), while the other half movesupwards (Oxfordian layer). Only the mostsoluble elements with the longest radioac-tive half-lives have enough time to reachthe layers above and below the Callovo-Oxfordian during the next million years.

2

CL

AY

4 5

Corrosion process observed with an

electronic beam microscope

23

Bioclastic matter

QuartzCalcite

1 mm 50 m

24


Research laboratoryResearch laboratoryResearch laboratory

Gondrecourt graben

Marne graben

Poissons fault

Brouthières-Soulaincourt fault

Direction of flow

Outcropping Oxfordianlimestone

Joinville graben

Marne/Poissonsfracturing zone

Callovo-Oxfordian

Marne river

Ornain river

Saulx river

Joinville

Ligny-en-Barrois

Gondrecourt-le-Château

LITHOLOGICAL CROSSSECTIONS

LimestoneMarls and clays

Direction of flow

Outcropping Oxfordianlimestone

Gondrecourt graben

Marne graben

Poissons fault

0 km 2.5 km 5 km Scale:

Brouthières-Soulaincourt fault

Joinville graben

Marne/Poissonsfracturing zone

Callovo-Oxfordian

Marne river

Ornain river

Saulx river

Joinville

Ligny-en-Barrois

Gondrecourt-le-Château

In one million years

Present-day situation

D.PL.FSTE.04.0058.B

Research laboratoryResearch laboratoryResearch laboratory


Schematic representation of the evolution of water flows in the calcareous Oxfordian layer, at the current state and in a million years

D.PL.FSTE.04.0058.B

25


Several evolutionscenariosAndra has performed calculations in twoconfigurations:

• Normal evolution scenario, based oncautious hypotheses. This scenario is notintended to represent the future reality,but rather to encompass the most proba-ble situations through a cautious approach.

• Altered evolution scenarios, integra-ting low-probability events and possibleincidents.

In keeping with international practice, thecalculations cover a period of millionyears, integrating climatic evolutions.

The impact of the repository is calculatedusing modelling: in case of low uncer-tainty, the most scientifically supportedmodel is used; in case of high uncertainty,a penalising model is adopted. Risks anduncertainties are therefore integrated asof the repository design phase.

Normal evolutionThe repository meets the safety criteria.

Water flow in the repository is very low.Most of the radioactive substancesmigrate very slowly through the Callovo-Oxfordian layer, and the argillites presentgood retention properties.

The argillaceous layer delays the release ofradioactive substances over hundreds ofthousands of years. After one million years,nearly all the radioactive elements are com-pletely attenuated. Only iodine 129 andchlorine 36 show significant flows at theCallovo-Oxfordian layer boundaries.

Various safety margins

In order to ensure maximum repositorysafety, a number of cautious choices havebeen made: many of the parameters cho-sen are among the most pessimistic, nofavourable properties have been attributedto the layers surrounding the Callovo-Oxfordian (even though they possesssome), and impacts on populations havebeen calculated on the basis of particularlypenalising water producing wells.

Altered evolutionDespite their low probability, Andra hasexplored possible disposal system mal-functions and external failures of humanor natural origin. It has been verified thattheir consequences would remain accep-table. This analysis has been comparedwith the results of studies conducted atthe international level:

• seal failure (drifts and B waste disposalcells) and defective plug (C waste andspent fuel disposal cells): very low

impact, due to the layer low permeabi-lity and the proposed architecture,

• defective C waste and spent fuel contai-ners: earlier manifestation of impacts,which are substantially the same as inthe normal evolution scenario due to theretention properties of the geologicalmedium,

• borehole penetrating the repository atvarious locations: effect limited to the im-pact zone, due to repository compartmen-talization, sealing and rock properties.

Repository safetyand impact on manTo assess the long-term safety of a repository, Andra has conductedstudies to assess its impact on man and the environment.

> Maximum acceptable impact

The repository impact on man and the environment must be comparedwith the threshold established by the Basic Safety Rule, i.e.,0.25 mSv/year. This radioactive dose corresponds to one fourth of theregulatory limit for public exposure of non-natural origin and approxi-mately one tenth of the annual dose due to natural radioactivity.

The calculated impact is the individual dose received by a groupconsisting of most exposed persons. For this purpose, "hypotheticalcritical groups representative of individuals susceptible of receivingthe highest doses, including individuals living in at least partialautarky" have been adopted (Basic Safety Rule RFS III.2.f).

2

CL

AY

What radioactive dosemight local populations

be exposed to?

Andra has assessed radioactive doses ba-sed on penalising criteria. Iodine 129,chlorine 36 and selenium 79 are the main

radioactive substances susceptible ofreaching the environment in the very longterm. Under current conditions as underthe future predictable conditions in a mil-lion years, the dose is at least 10 timessmaller than the limit of 0.25 mSv/year inthe case of spent fuel, and 100 times

smaller for B and C waste. Therefore, un-der severe hypotheses, the dose recei-ved would be 1000 smaller than naturalradioactivity in the case of B and C waste.

26


Repository safety and impact on man

> Very cautious parameters

To predict repository evolution, including in the normal evolution scenario, Andra has adopted cautious,even pessimistic parameters. For example:

> Repository located in the thinnest part of the Callovo-Oxfordian layer and at the greatest depthin the transposition zone.

> For B waste, concrete overpacks assumed to be non-watertight.

> For C waste and spent fuel, overpacks with pessimistic lifetimes (4000 and 10,000 years, respectively)as compared to available data.

> Absence of retention properties in layers surrounding the Callovo-Oxfordian.

> Impact of chemical elements

Boron, selenium, nickel and antimony are elements susceptible of being present in high-level long-lived waste.The results of calculations up to a million years show that their impact is negligible.

Progression of selenium after 100,000, 500,000 and one million years

27


Repository safety and impact on man

What would the dosebe if all repository safety

functions were degradedsimultaneously?

Andra has also studied an extreme situa-tion where all safety functions would nolonger be normally ensured: permeabilityof rock and materials higher than expec-ted, very low performance of seals,release of radioactive substances by allpackages, and pessimistic values fortransport and retention in argillite. The

impact would still be smaller than thedose limit of 0.25 mSv/year.

In general, the incidental or altered evolu-tion scenarios only produce a modestincrease of the dose, which remainssignificantly below the regulatory limits.

Even in highly unlikely situations, geolo-gical disposal constitutes an efficientand robust concept to protect man andthe environment from the waste empla-ced therein.

Zone within whichthe water tablefluctuates

Spraying

Roottransfer

Irrigation

Leaftransfer

Desaturatedzone

Irrigation/spraying

Transfer ofelements

Outlet

VALLEY ZONE PLATEAU ZONE

Crops and pastureson alluvium

Low transfer towardsthe ground (non-satured)

River

Alluvial aquifer

Calcaire du Barrois(Tithonian)

~20 m

D.PL.FSTE.04.0152.B

Model of radionuclide transfers from the geosphere to the biosphere at the Meuse/Haute-Marne site

2

CL

AY

D.PL.FSTE.04.0152.B

A globalapproach

Simultaneously with the research on clay,Andra has studied the possibility of a HLLLwaste repository in a granite formation.From 1994 to 1996, it performed geologicalsurveys to site an underground laboratoryin the granite of the Vienne department.Following the recommendation by theFrench National Review Board (CNE), theGovernment authorities discarded this site.In 1999, the French Government organiseda consultation mission to collect theopinions of populations in 15 sites deemed

geologically favourable by a committee ofnational and international experts. This mis-sion could not be completed, so the follo-wing year Andra set up a research pro-gramme to value the data acquired inforeign underground laboratories and in va-ried geological contexts.

In the absence of a specific site, this re-search work was not able to assess thefeasibility of a repository in a particular lo-cation. Its purpose was to assess the ge-neral interest of the granite medium andto propose generic concepts capable ofmeeting the long-term safety objectiveswithin the French geological context.

The research covered four aspects:

• granite medium

• generic design of a reversible disposalsystem in a granite formation, i.e., archi-tecture, waste conditioning, operationand closure, while sharing certain dataused for the research on clay (particu-larly as regards packages and materials)

• repository evolution (thermal, mechani-cal, chemical and hydraulic)

• long-term safety analyses

28


Research on arepository in a graniteformation

Experimentation in the Äspö laboratory (Sweden)

29


ScientificCo-operationsAndra has engaged in various scientificpartnerships, namely with the FrenchGeological Survey (BRGM), the FrenchAtomic Energy Commission (CEA), theFrench National Centre for ScientificResearch (CNRS) and the Paris School ofMines (Ecole des Mines de Paris), notablyin order to transpose to the Frenchcontext, the laboratory results obtainedabroad . It has also relied strongly on re-search conducted in foreign laboratories,namely in Sweden and Switzerland, andhas actively participated in studiesconducted from the surface in Finland.

Dossier 2005Granite Andra has organised its knowledge ongranite into five reference knowledgedocuments. Four of these documents areshared with the Dossier 2005 Argile(those concerning materials, radioactivesubstances and their migration). The fifthone contains the data available on Frenchgranite formations.

Three volumes summarise the know-ledge acquired:

• proposal of generic options for safe andreversible disposal system architectures

• analysis of repository evolution

• safety analysis. In the absence of anyspecific site, the purpose of this analy-sis is to verify that none of the conside-red criteria rules out feasibility, and toidentify the essential parameters todesign a repository and to carry out theworks at a possible site.

What knowledge hasAndra acquired from

foreign granite formations?Large amounts of information concerninggranite have been acquired at the interna-tional level. Andra has participated inexperiments conducted in the under-ground laboratories of Sweden (Äspö),Switzerland (Grimsel), Canada (Lac duBonnet), on the Olkiluoto site (Finland)and in Japan. The main cooperations haveinvolved research on granite formations(geological structure, survey methods,underground water flows, rock capacityto retain radioactive substances), reposi-tory structures and safety analyses.

Research on a repository in a granite formation

Swedish repository concept for the disposal of spent fuels in a granite medium 3

GR

AN

ITE

What propertiesare required for arepository?In the absence of a specific site, the repo-sitory design is based on the shared pro-perties of French granite formations.

Granite has interesting properties for anHLLL waste repository: hardness, resis-tance, low porosity, very low permeabi-lity and good thermal conductivity.

Most granite formations extend in depth, of-fering great flexibility for repository design.

The possible variations in rock composi-tion from one point to another in a forma-tion do not significantly modify its proper-ties. On the other hand, the fracturespresent in the granite massif must be ta-ken into account. Small fractures (up to afew tens of metres long) can affect therock local permeability. Faults (up to seve-ral kilometres long) are far less numerousand constitute privileged water transportpathways, but this transport occurs deepunderground and slowly. Finally, the deepunderground chemical environment in gra-nite formations provides the preservationof repository materials and the immobili-sation of most radioactive substances.

Different types ofgranite formations

The granitic zones studied were initiallydescribed based on surface mapping andon their deep underground characteristicsextrapolated from geological knowledgeand calculations. Andra then classified theFrench granite formations into differentcategories and appraised their propertiesfor repository design purposes.

The differences between the French gra-nite formations in terms of mechanicalresistance and water composition do notput at stake the proposed design options.On the other hand, their thermal proper-ties (for example, with temperaturesreaching 17 to 30°C at a depth of 500metres) can modify the repository dimen-sioning.

Large fractures are arranged differentlydepending on the formations.

30


Characteristics of Frenchgranite formationsAs in the case of clay, the repository study consists of identifyingthe characteristics of the geological medium and designing architecturesbased on these characteristics so as to confine the waste over very longperiods, while meeting the reversibility requirement.

Location of granite formations in France

G.PL.FSTE.05.0064.A

31


Nevertheless, the zones where a reposi-tory could be sited meet characteristicsshared by all the French granite forma-tions considered.

Three configurations of massifs were de-fined, this classification allowing architec-tures to be adapted and generic safetyanalyses to be conducted.

Where are the graniteformations considered

for these generic studieslocated?

In the absence of a specifically designa-ted site, Andra has studied the variousgranite formations so as to determinetheir typology and assess their proper-ties. These studies concerned 78 zonesof over 20 km2 each, located in the MassifCentral and Massif Armoricain forma-tions, quite apart from large faults.

Can the presenceof fractures rule out a

repository?

Although fractures in a granite can allowwater to circulate and therefore be detri-mental, they also give rise to phenomenasusceptible of immobilising or delayingthe migration of radioactive substances.Indeed, radioactive substances can beentrapped therein, and clogging may alsooccur. In situ experiments, notably in theÄspö underground laboratory in Sweden,have allowed these phenomena to beunderstood. Nevertheless, the identifica-tion of “sound” (without faults) graniteblocks remains of major importance,since such blocks may host a repository.

Characteristics of French granite formations

Details of a main fault structure

Unfinished obelisk in the Assouan granite quarry (Egypt)

3

GR

AN

ITE

G.PL.FSTE.05.0119.A

Located apart from large faults, the repo-sitory is divided, according to waste ca-tegories (B, C) and spent fuel, in zonessufficiently separated from one anotherto prevent thermal or chemical interac-tions between waste packages. Eachzone is divided into modules comprising aset of disposal cells. In order to limit wa-ter flow, these cells are implemented ingranite “blocks” with very low permeabi-lity and no faults. The means of access(shaft and/or ramp) connecting the sur-face to the underground installations areimplemented so as to avoid drainage ofsurface waters.

The thickness available at a depth of bet-ween 300 and 1000 metres provides thenecessary flexibility to adapt the architec-ture to granite fracturing. The repositorycan thus be designed on two levelsapproximately one hundred metres apart.

Clay seals toprevent water flowThe drifts and means of access of therepository may encounter water conduc-ting fractures. In order to limit water

flows in the repository and delay themigration of radioactive substancestowards the environment, the drifts andmeans of access are backfilled duringtheir progressive closure. In keeping withthe reversibility rationale, it will still bepossible to reopen them and access thewaste packages. Swelling clay seals arealso implemented in the drifts at locationsintersecting the faults. The disposal cellsare sealed with low-permeability swellingclay plugs.

Waste disposalpackagesensuring long-termleak-tightness

The packages delivered by waste produ-cers are placed in containers to constitutedisposal packages. The B and C wastecontainers are similar to those proposedfor the clay medium repository (concretecontainer for B waste, steel one for C

32



Repository design adapted to granite fractures

Diagram of a possible repository layout in a granite medium

Copper container demonstrator tested by the Swedish organisation SKB

C.IM.OSES.05.0169.C

Surfaceinstallations

Ramp Disposal tunnels(B waste) Connecting

drifts Exploratorydrifts

Repositorymodule(C waste)

Shaft

33


waste). For spent fuels, Andra has selec-ted the copper container adopted inSweden and Finland. This spent fuel cop-per container concept is currently beingtested at full scale in the Äspö laboratory(Sweden).

Physicaland chemicalenvironmentfavourable forwaste packagesThe C waste and spent fuel disposal cellsare equipped with a swelling clay barrierthat limits chemical contacts betweenwaste packages and granite groundwaterand provides shelter from small fractures.The materials constituting the repository,i.e., concrete, steel or copper (for thepackages), swelling clay, backfill and seals,provide a favourable physical and chemi-cal environment for waste packages andthe retention of radioactive substances.

Architecturelimitingthe effectsof heatGranite is a resistant rock. The drifts andcells are designed to ensure long-termmechanical stability.

In order to control the consequences ofthe heat given off by C waste and spentfuel packages, the maximum temperaturein the engineered structures is limited to90°C by restricting the number of packa-ges per cell and spacing the cells apart.

What would a disposalcell in the granite rock

look like?

The proposed solution for the B wastedisposal cell is a horizontal tunnel wherepackages are stacked in several rows. Itslength (70 to 200 metres) is adapted togranite fracturing. It is approximately 10metres high and 10 to 20 metres wide.Packages are handled by remote controland placed in the disposal cell through asafety dual-gate airlock. Once the cell isclosed, the dual-gate airlock is backfilledand the access drift is sealed with a swel-ling clay plug (preventing water flow).A similar principle is being studied inJapan.

For C waste packages, the principle of avertical borehole (12 metres long and2 metres in diameter) has been adopted.This borehole receives between two andfive packages, depending on the heatgiven off, and opens into a drift above.A swelling clay barrier is emplaced bet-ween the packages and the rock.

How would the driftsbe organised?

The tunnels are implemented at a dis-tance of several tens of metres from wa-ter conducting faults while the two-levelrepository concept makes use of the lesspermeable rock between the faults.

Is a disposal system ina granite formation safe

and reversible?

The principles of safety and reversibilityfor a disposal system in a granite forma-tion are similar to those for a disposal sys-tem in a clay formation.

At this stage, in the absence of a specificsite, a detailed safety assessment cannotbe carried out. However, the studiesconducted so far have not identified anyaspect ruling out the feasibility of a gra-nite repository with regard to the long-term safety. For the reversibility require-ment, it provides guarantees similar tothose of a repository in a clay formation.


C waste repository concept in a granite medium

3

GR

AN

ITE

C.IM.OSES.05.0262.A

Deep geological disposal has been inves-tigated since the 1960's in Western na-tions. In France, research has progressedsignificantly since the 1991 Act, with allresources mobilised to produce solidscientific results. Quite significant resultshave been achieved in all fields of research,yielding a precise view concerning theproperties of all repository components.

Assets of the Meuse/Haute-Marne site

In the case of research on the clay me-dium, highly detailed investigations havebeen conducted for over 10 years in theMeuse/Haute-Marne site. The under-ground laboratory has produced impor-tant data and constitutes an extremely va-luable asset to supplement previouslyacquired results whenever decided.

Simultaneously with the programmeconducted in France, the research perfor-med in foreign underground laboratorieshas enabled the validation of Andra ap-proach.

Mobilisation of leading scientists

Andra has mobilised the best laboratoriesavailable both in France and abroad foreach field of research. The production ofresults has been discussed in accordancewith the requirements of the scientificcommunity and the aim for excellence.This guarantees the quality of the workproduced.

Regular external assessment

Andra has called on external experts tobenchmark its research with the best in-ternational practices. In 2002 and 2003,international experts reviewed the resultsof a scientific report concerning clay(Dossier 2001 Argile) and formulated veryencouraging conclusions. Their recom-mendations were integrated in the

Dossiers 2005 Argile and Granite.

The results achieved by Andra and its part-ners are published in international scienti-fic journals and therefore subject to criticalreview by the scientific community.

The feasibility ofa repository in aclay formation hasbeen establishedFavourable conditions in the Meuse/

Haute-Marne site

The Callovo-Oxfordian layer has very inte-resting properties corresponding to thoseexpected for a repository design in a claymedium. These properties are, a priori,met in an area of over 200 km2.

Engineering studies based on cautiouschoices have defined simple and robustrepository concepts adapted to the cha-racteristics of the argillaceous layer.

Reversibility: a demonstrated priority

Andra has developed an approach beyondthe mere possibility of retrieving thewaste packages. It can be defined as thepossibility of progressive and flexible re-pository management in stages, to givefuture generations a freedom of choice.In addition, Andra has decided not to seta predetermined duration for reversibility.

The reversible disposal system can thusserve two purposes. It can be initially ma-naged as a temporary storage, with subse-quent retrieval of the waste if so decided.But it can also be closed progressively soas to evolve in a safe manner withouthuman intervention.

No significant impact on the environment

A major achievement of the research pro-gramme is the construction of the reposi-tory history over the next hundreds ofthousands of years so as to understandthe system evolution together with therisks and their associated uncertainties.

The analysis results show that the safetyobjectives have been met. The cautious,even pessimistic choices made allow forsignificant safety margins. These conclu-sions are valid for normal situations, butalso for altered configurations (failure ofrepository components or intrusion intothe repository).

The consequences for man and theenvironment comply with applicableregulations and recommendations, withsignificant margins.

34


Status of progressand new perspectivesFifteen years of considerable progress in research

35


The studies concerning granite haveshown that a repository in a French gra-nite formation cannot be ruled out. In theabsence of a specifically identified site,the analysis is based on the design of ge-neric architectures. Various possible tech-nical options for a reversible disposal sys-tem have been considered.

A repository in a granite formationis conceivable

The research relies strongly on program-mes developed in Sweden and Finland fora repository in a granite formation. Variousexperiments have been conducted inpartnership with foreign laboratories.

The main uncertainty concerns the exis-tence of granite sites without a too highfracture density, wich would be too de-manding for the architectures.

Status of progress and new perspectives

4

BIL

AN

Meuse/Haute-Marne underground research laboratory

required to determine the possiblelocation of a repository within this zone,

• finally, certain repository componentshave been designed using simplifiedand particularly pessimistic models.Within the scope of a more finalised ap-proach, it would be useful to quantifythese safety margins and reduce theyet remaining uncertainties.

If the assessments confirm the relevanceof Andra results and if the Parliament de-cides to pursue research on a geologicalrepository in a clay layer, Andra could thencarry on its activities within a finalisedperspective.

Initially, the current basic feasibility phasewould be followed by a developmentphase covering a period of approximatelyfive years. This development phase wouldallow to both address potential issues rai-sed by assessment experts in 2006 andfocus on technological implementation.

As well, during this phase, the necessaryinformation to site the possible locationof a repository installation would be col-lected. For example, this could entail a

seismic campaign covering a large zone.As far as the Meuse/Haute-Marne labora-tory is concerned, it would constitute anessential asset to pursue data acquisitionand conduct , directly in situ, technologi-cal tests.

This phase could conclude with an overallassessment. Assuming that scientific andtechnical results receive favourable ap-praisal, it would then be possible to ad-vance to an industrial developmentphase. As an indication, such an approachcould lead to the effective implementa-tion of an industrial disposal facility by theyear 2025.

36


After 2006: what perspectivesfor research on clay formation?

Status of progress and new perspectives

The research programme conducted overthe past 15 years has yielded the neces-sary information to determine the basicfeasibility of a repository in a clay forma-tion. Nevertheless, uncertainties remain.

Without anticipating any decisions thatthe Parliament may deem necessary, va-rious aspects must be considered in or-der to assess the perspectives providedby the research results:

• the experiments have been conductedover short periods of time. Legitimatecaution suggests that experimental sys-tems should be allowed to pursue know-ledge acquisition over the next years,

• repository structures have not been yettested in full scale. It would be useful toproduce , in situ, disposal cell prototy-pes. An engineering consolidation planwould be necessary if it is decided togradually focus on an industrial objec-tive,

• a detailed survey of the zone coveringover 200 km2 around the Meuse/Haute-Marne site has not been conducted.Complementary investigations are

Photo credits:

p. 2Studio Durey

p. 3Top: CogemaBottom: CEA

p. 4Les Films Roger LeenhardtTop and bottom

p. 6Graphix Images

p. 7Graphix ImagesPhilippe DemailAndra

p. 8Graphix Images

p. 9OCDE – AEN

p. 11Laboratory cross-section: AndraAndraGraphix ImagesPhilippe Demail

p. 13Philippe DemailGraphix Images

p. 15T. Duvivier

p. 16T. Duvivier

p. 28SKB

p. 31H. PasteauAndra

p. 32SKB

p. 35Graphix Images

© A

nd

ra

26

5 V

A -

Ap

ril

20

06

- G

ra

ph

ic d

es

ign

: M

ed

iag

ér

an

ce

01

42

12

20

48

- I

SB

N 2

-95

10

10

8-7

-7

Agence nationale pour la gestion des déchets radioactifs

Parc de la Croix Blanche - 1/7, rue Jean Monnet - 92298 Châtenay-Malabry Cedex - Tél. : 01 46 11 80 00www.andra.fr

The Andra Publication Series

Essential SeriesIn a few pages, documents in the Essential Series providesimple and illustrated explanations with a view to furtheringknowledge on radioactive waste and Andra.

Reference SeriesWith standard information concerning Andra methods andprogress reports on its investigations or activities, theReference Series presents various technical and otherinformation, especially on the location of radioactive waste.

Periodical SeriesOn a regular basis, Andra publishes various brochures relatingto the environmental monitoring of its disposal and researchfacilities. The Periodical Series includes those publications aswell as the respective news bulletins of each site.

Discovery SeriesVideos, CD-ROMs, synthesis images and comic strips… areworth more than a thousand words. The Discovery Series usesvivid illustrations to explain to a broad public the underlyingprinciples of radioactive-waste management.

Science and Technology SeriesTaking stock of current knowledge, presenting ongoingresearch as well as the Agency's methods and approachesconstitute the objectives of the Science and TechnologySeries. Intented for a specialised public, it provides varioussyntheses and monographs published under the aegis ofAndra or in partnership with other scientific organisations.

Report SeriesSummaries, reports and seminar proceedings published in

the Report Series highlight the advances of Andra's ongoing

investigations.

Industrial Practices SeriesThe Industrial Practices Series includes documents dealingwith the acceptance criteria and the management ofradioactive waste.

Andra research on the geological disposal of high-level ... · Dossier 2005 Research on deep disposal of radioactive waste A general interest task Andra (French National RadioactiveWaste

Documents