17
CLEFS CEA - No. 53 - WINTER 2005-2006 26 The groundwork of current solutions From the outset, nuclear waste has been managed in terms of the most advanced technologies available at the time it is produced. Advances in characterization and decontamination have in the meantime allowed improvements in sorting, and, in some cases, downgrading of some categories of waste. Such downgrading allows simpler long-term management modes to be used. These advances further allow retrieval and conditioning to be considered for older, legacy waste. Throughout the various stages in a waste item’s life, the “package” stands as the “basic unit” for management purposes, ensuring durable confinement of radioactivity. P. Stroppa/CEA Storage hall for drums of waste at CEA Saclay, awaiting consignment to ANDRA (in yellow) or Centraco (red-brown), after characterization on the automated CAMDICES line. Nuclear waste management and processing: between legacy and anticipation T he aim of radioactive waste management is to pro- tect humankind and its environment from the effects of that waste’s constituent materials, particu- larly from radiological hazards. This management is governed by a few major principles: curbing quanti- ties, by producing as little waste as practicable; sorting; safe conditioning, taking on board long-term aspects; and isolation from man and the environment. This takes the form of a deployment of pathways which, as a rule, involve, apart from the primary-waste genera- ting stage (R&D, operation, maintenance, decommis- sioning), stages of processing–conditioning, possibly of interim storage, followed by final disposal. What are the steps in the formation of a package? A conditioned waste item is known as a package, and the latter must possess mechanical properties meeting the requirements of handling, transport, storage and disposal (see Box C, What stands between waste and the environment? p. 28; Box D, From storage to dispo- sal, p. 50). Its formation thus involves an ensemble of steps, from raw waste characterization to characteri- zation of the package itself, vouchsafing the function of durable radioactivity confinement, over the times- cale considered. Raw waste characterization consists in determining the quantity, nature and location of the radioactivity the waste bears, and its physical–chemical form, for the purposes of optimizing conditioning (matrix for- mulation…). The waste treatment step, after this, addresses two major objectives: on the one hand, volume reduction through specifically-designed methods (cut- ting, crushing, decontamination…) and processes (organic waste incineration, melting metallic waste, compaction…), and, on the other hand, adjustment of waste shape and form to suit it to process and condi- tioning matrix. Waste conditioning operations allow putting waste into a form suited for future manage-

Nuclear waste management and processing: between legacyand

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

CLEFS CEA - No. 53 - WINTER 2005-200626

The groundwork of current solutions

From the outset, nuclear waste has been managed in terms of the most advancedtechnologies available at the time it is produced. Advances in characterization anddecontamination have in the meantime allowed improvements in sorting, and, in some cases,downgrading of some categories of waste. Such downgrading allows simpler long-termmanagement modes to be used. These advances further allow retrieval and conditioning to beconsidered for older, legacy waste. Throughout the various stages in a waste item’s life, the“package” stands as the “basic unit” for management purposes, ensuring durable confinementof radioactivity.

P. S

trop

pa/C

EA

Storage hall for drums ofwaste at CEA Saclay,

awaiting consignment toANDRA (in yellow) or

Centraco (red-brown),after characterization on

the automatedCAMDICES line.

Nuclear waste managementand processing: betweenlegacy and anticipation

The aim of radioactive waste management is to pro-tect humankind and its environment from the

effects of that waste’s constituent materials, particu-larly from radiological hazards. This management isgoverned by a few major principles: curbing quanti-ties,by producing as little waste as practicable; sorting;safe conditioning, taking on board long-term aspects;and isolation from man and the environment. Thistakes the form of a deployment of pathways which, asa rule, involve, apart from the primary-waste genera-ting stage (R&D, operation, maintenance, decommis-sioning), stages of processing–conditioning, possiblyof interim storage, followed by final disposal.

What are the steps in the formation of apackage?

A conditioned waste item is known as a package, andthe latter must possess mechanical properties meetingthe requirements of handling, transport, storage and

disposal (see Box C, What stands between waste andthe environment? p.28; Box D,From storage to dispo-sal, p. 50). Its formation thus involves an ensemble ofsteps, from raw waste characterization to characteri-zation of the package itself, vouchsafing the functionof durable radioactivity confinement,over the times-cale considered.Raw waste characterization consists in determiningthe quantity, nature and location of the radioactivitythe waste bears, and its physical–chemical form, forthe purposes of optimizing conditioning (matrix for-mulation…). The waste treatment step, after this,addresses two major objectives:on the one hand,volumereduction through specifically-designed methods (cut-ting, crushing, decontamination…) and processes(organic waste incineration, melting metallic waste,compaction…), and, on the other hand, adjustmentof waste shape and form to suit it to process and condi-tioning matrix. Waste conditioning operations allowputting waste into a form suited for future manage-

CLEFS CEA - No. 53 - WINTER 2005-2006 27

ment purposes, through compaction, or incorpora-tion, whether homogeneous (e.g. effluent cementa-tion,vitrification of fission product and actinide solu-tions, encapsulation in bitumen) or heterogeneous(encasing in cement), into an appropriate confinementmatrix.Some radioactive waste is directly conditionedin suitable containers. The following pages describethe conditioning processes in use, and advances pre-sently in hand, these being the outcome of researchefforts carried out, in the main, under the aegis of theFrench Act of 30 December 1991.Package characterization allows the quantity, nature,and location of the radioactivity contained to be acces-sed, for the purposes of directing the package to themost appropriate outlet (and for the purposes of veri-fication that associated regulatory acceptance criteriaare met). These obligatory steps are complemented bythose relating to demonstration of safety for a packagetype, in its storage or disposal site. The aim, for exam-ple, may be to demonstrate the quality of confinementover a timescale relevant to a pathway type (see Box D).This is the object of investigations on the long-termbehavior of packages, covered in the second chapterof this issue of Clefs CEA.At the outcome of the process,primary waste has under-gone a number of treatment and conditioning opera-tions, bringing it to the final form of a waste package.It may then be taken in by the final disposal site, forshort-lived intermediate- and low-level waste, andvery-low-level waste (provided,in France,that ANDRAdisposal specifications have been met, and that rele-vant agreements have been issued), or be kept in sto-rage (for long-lived low- and intermediate-level waste),pending deployment of long-term management path-ways.

Management pathways and wastedowngrading

With respect to day-to-day management,a distinctionmust be made between operational pathways, i.e. forwhich a final outlet is available (Aube Disposal Center,Morvilliers Disposal Center; see Industrial solutions forall low-level waste, p. 32), and interim storage, for suchwaste as is covered,as regards its long-term fate,beyondthat storage stage, by the Act of 1991. Turning to path-ways now in prospect, relying on the deployment ofnew installations (now planned, or in construction),these will benefit from advances in treatment tech-niques and processes, conducive to lower waste pro-duction. Finally, as regards legacy waste (mainly wasteproduced earlier than the 1990s, and stored on site atthe installations), and waste from decommissioningoperations at installations to be shut down, retrievaland conditioning operations will be addressed by aspecific development effort.As final installations are in place only for VLLW andLILW-SL waste, considerable resources are beingdeployed by producers to minimize the quantities ofwaste, in an effort directed at the ILW-LL and HLW-LL waste categories. This has resulted, in particular, inimproved processes, to achieve lower waste produc-tion, and in the deployment of treatment and charac-terization methods,with the aim of downgrading waste,initially classified as of the ILW-LL type, to LILW-SLwaste.Thus, for the older, legacy waste mentioned ear-lier, which is often taken to be of the ILW-LL type, dueto inadequate knowledge, characterization by way ofcurrent methods, more accurate than the older onesas they are, allows certain items to be directed straightto operational industrial disposal facilities.At the same

Waste from basic nuclearinstallations at CEA’s SaclayCenter, placed in 60-literstainless-steel drums, aftercompaction. These drumswill subsequently beencased in mortar insidepackages, which may be putinto concrete overpacks, foracceptance by ANDRA.

P. S

trop

pa/C

EAJo

ly/C

EA

Alceste shielded cell in the Chicade installation, dedicated to characterization, through destructive or non-destructive methods, of low- or intermediate-level waste packages, and to the investigation of waste decontamination and confinement processes.

CLEFS CEA - No. 53 - WINTER 2005-200628

The groundwork of current solutions

C

Raw, solid or liquid radioactive wasteundergoes, after characterization

(determination of its chemical and radio-logical makeup, and of its physical–che-mical properties), conditioning, a termcovering all the operations consisting inbringing this waste (or spent fuel assem-blies) to a form suitable for its transport,storage, and disposal (see Box D). Theaim is to put radioactive waste into asolid, physically and chemically stableform, and ensure effective, lasting confi-nement of the radionuclides it contains.For that purpose, two complementaryoperations are carried out. As a rule,waste is immobilized by a material –whether by encapsulation or homoge-neous incorporation (liquid or powderedwaste, sludges), or encasing (solid waste)– within a matrix, the nature of, and per-formance specification for which dependon waste type (cement for sludges, eva-poration concentrates and incinerationashes; bitumen for encapsulation ofsludges or evaporation concentratesfrom liquid effluent treatment; or avitreous matrix, intimately binding thenuclides to the glass network, for fis-sion product or minor actinide solutions).This matrix contributes to the confine-ment function. The waste thus conditio-ned is placed in an impervious contai-

ner (cylindrical or rectangular), consis-ting in one or more canisters. The whole– container and content – is termed apackage. Equally, waste may be com-pacted and mechanically immobilizedwithin a canister, the whole forming apackage.When in the state they come in as sup-plied by industrial production, they areknown as primary packages, the pri-

mary container being the cement ormetal container into which the conditio-ned waste is ultimately placed, to allowhandling. The container may act as initialconfinement barrier, allotment of func-tions between matrix and container beingdetermined according to the nature ofthe waste involved. Thus, the whole obtai-ned by the grouping together, within onecontainer, of a number of primary

Cross-section of an experimental storage borehole for a spent fuel container (the lower part of theassembly may be seen, top right), in the Galatée gallery of CECER (Centre d’expertise sur leconditionnement et l’entreposage des matières radioactives: Radioactive Materials Conditioningand Storage Expertise Center), at CEA’s Marcoule Center, showing the nested canisters.

A. G

onin

/CEA

What stands between waste and the environment?

from ever more effective analytic methods, requiringhowever as they do samples to be taken, non-destruc-tive methods have been developed,allowing the nature,quantity and location of that radioactivity to be ascer-tained precisely.These methods are based on the coupling of a numberof imaging or spectroscopy techniques (see Figure 1).Some allow radiography of the package, by means ofan external radiation source, while others take advan-tage of the radiation emitted by the radioactive atomsheld in the package, the intensity of which indicatesthe quantity of radioactivity,while its energy spectrumcharacterizes the radionuclides involved.Methods exhibiting detection thresholds, with regardto sorting waste,adequate to direct it to the Aube Center,are available, allowing a reduction in the volume ofILW-LL ultimate waste. By way of example, actinidedetermination for contents lower than the acceptancethreshold for that facility is now feasible in 1-m3 conc-rete packages.

Treatment processes

One major method, to achieve volume reduction forlow- and intermediate-level waste,particularly in clea-nup and decommissioning work, is raw waste decon-tamination. The point is to concentrate radioactivityinto a restricted volume,subsequently segregated from

Figure 1. Inspection, by means of radiography

and coupled active and passivetomography methods, of a waste

package. From left to right:radiography (black and white) of a

package of loose waste, carried outusing an external cobalt-60 source;

transmission tomography of thatsame package; and coupled

emission tomography of that slice,pinpointing a hot spot at center. C

EA

time, for current waste production (operational ordecommissioning waste),downgrading may be achie-ved through sorting operations,complemented by cha-racterizations.Waste sorting is thus an operation of paramount impor-tance, which may be carried out at a number of stagesin package fabrication, notably at the pre- and post-treatment stages.

Package characterization

Investigations on package characterization have madeit possible to achieve improved knowledge, and cha-racterization,of radioactive content.In particular,aside

CLEFS CEA - No. 53 - WINTER 2005-2006 29

the initial waste, for which the residual activity levelbecomes sufficiently low to allow downgrading, or, atany rate, whose radioactivity will have quite signifi-cantly come down. Investigations carried out in thisarea have resulted in efficient processes,making it pos-sible to avoid use of aqueous solutions (caustic soda…),yielding as they do considerable volumes of salineeffluents. These processes include gels, foams, surfac-tant solutions, as well as decontamination by meansof lasers or supercritical fluids. Gels, amenable todrying and vacuuming, ultimately only yield a low-volume,solid residue (flakes).Surfactant-based foamsact on a volume to be decontaminated that is typicallyten times larger than that of the solution initially used.Use of surfactant solutions, allowing for instance treat-ment of surfaces covered with contaminated greasycoatings, was successfully employed in the cleanup ofthe APM (Atelier pilote de Marcoule: Marcoule PilotWorkshop) facility. Coupled use of surfactants in anacid medium and mechanical agitation enabled strip-ping of the superficial layer of TBP (the extractant inthe Purex process) coating the tanks in that facility,allowing their effective decontamination.Laser decon-tamination processes,or processes of supercritical fluidextraction, in turn, yield no effluent. The former pro-cess allows ablation, by means of a laser pulse, of thecontaminated surface,while the latter makes use of thepeculiar solvent properties of CO2 in the supercritical

phase (at moderate pressure and temperature) to dis-solve the surface contaminant, then recover it throughdecompression of the fluid in gaseous phase. Whilethese two processes show promise, they do howeverstill require development work before they reach qua-lification scale.The first three of these methods, on the other hand,have reached a stage of development such as to makeindustrial applications possible.Formulations of whollymineral gels,which may be vacuumed,have been deve-loped, allowing subsequent safe disposal of the solidresidue (flakes), after encapsulation if required, withno risk of physical–chemical alteration of this residuefrom the effects of radiolysis or temperature. Recent

ILW-LL packages may ensure confinementof the radioactivity of this type of waste.If a long-term storage stage is found tobe necessary, beyond the stage of indus-trial storage on the premises of the pro-ducers, primary waste packages mustbe amenable to retrieval, as and whenrequired: durable primary containersmust then be available, in such condi-tions, for all types of waste.In such a case, for spent fuel assemblieswhich might at some time be earmar-ked for such long-term storage, or evenfor disposal, it is not feasible to demons-trate, on a timescale of centuries, theintegrity of the cladding holding the fuel,forming the initial confinement barrierduring the in-reactor use stage. Securingthese assemblies in individual, imper-vious cartridges is thus being conside-red, this stainless-steel cartridge beingcompatible with the various possiblefuture management stages: treatment,return to storage, or disposal. Placingthese cartridges inside imperviouscontainers ensures a second confine-ment barrier, as is the case for high-level waste packages.In storage or disposal conditions, thewaste packages will be subjected to avariety of aggressive agents, both inter-nal and external. First, radionuclide

radioactive decay persists inside the pac-kage (self-irradiation process). Emissionof radiation is concomitant with heatgeneration. For example, in confinementglasses holding high-activity (high-level)waste, the main sources of irradiationoriginate in the alpha decay processesfrom minor actinides, beta decay fromfission products, and gamma transitions.Alpha decay, characterized by produc-tion of a recoil nucleus, and emission ofa particle, which, at the end of its path,yields a helium atom, causes the majorpart of atom displacements. In particu-lar, recoil nuclei, shedding considerableenergy as they do over a short distance,result in atom displacement cascades,thus breaking large numbers of chemi-cal bonds. This is thus the main causeof potential long-term damage. In suchconditions, matrices must exhibit ther-mal stability, and irradiation-damageresistance.Stored waste packages will also be sub-jected to the effects of water (leaching).Container canisters may exhibit a deg-ree of resistance to corrosion processes(the overpacks contemplated for glas-ses may thus delay by some 4,000 yearsthe arrival of water), and the confine-ment matrices must be proven to exhi-bit high chemical stability.

Between the containers and the ultimatebarrier provided, in a radioactive wastedeep disposal facility, by the geologicalenvironment itself, there may further beinterposed, apart, possibly, from an over-pack, other barriers, so-called engi-neered barriers, for backfill and sealingpurposes. While these would be point-less as backfill in clay formations, theywould have the capability, in other envi-ronments (granite), of further retardingany flow of radionuclides to the geo-sphere, notwithstanding degradation ofthe previously mentioned barriers.

CEA

Decontamination of tanksthrough coupled use ofsurfactant solutions in an acidmedium and mechanicalagitation during cleanup ofAPM (Atelier pilote deMarcoule: Marcoule PilotWorkshop).

CEA

Technological demonstratorsof ILW-LL packages for

bituminized sludges.

CLEFS CEA - No. 53 - WINTER 2005-200630

The groundwork of current solutions

development work has made it possible to go from thelaboratory scale to that of industrial production: indeed,an operating license has recently been granted.Moreover,spray application processes may be used with this typeof gel. The latter are also covered by patents, jointlytaken out by Areva–Cogema, and a number of licen-ses have been agreed to. Use of these processes resultsin a quite significant gain in terms of absorbed dosefor operators, compared with traditional swabbingprocesses, using rags soaked with a decontaminantsolution,e.g.cerium-based (see Figure 2). Indeed,theyinvolve spraying the gel onto the walls requiring decon-tamination, letting it act and leaving it to dry, thenvacuuming it while brushing the surface, if an opti-mized decontamination factor is sought, as evidencedby operational feedback.This vacuumable gel technique has just been success-fully used for decontamination of a cell of the ISAIinstallation at Marcoule. The gel was applied over allstainless steels walls, its formulation not allowing, asyet,effective use on painted surfaces. Initial assessmentis positive,since the decontamination factor was foundto be greater than 10,allowing the cell’s ambient radio-activity to be brought down sufficiently for operatorsto begin decommissioning work. This technique ishighly attractive on the export side, for major cleanupand decommissioning sites,where these gels have alreadyundergone preliminary testing.Foam decontamination, in turn, offers the advantagethat it may be used for complicated geometries, suchas the inside of a fission product storage tank, whileensuring a gain by a factor 10 as between total volumesubjected to decontamination, and final volume offoam settling back, after a few hours, into the form ofa liquid solution. Such tanks are, as a matter of fact,fitted with internal structures, mainly comprisingcoolant water circuits (warranted by the heat releasedby the fission products stored).This technique is beingtargeted by full-scale formulation and qualification

investigations. Pilots do in fact allow testing (in inac-tive configuration) foam filling and draining for a 20-m3 tank, or filling a compressed-air line networkby means of a foam train, displaced along the entirelength of the network. This qualification work mayrequire feedback with regard to formulation, relating,for example, to the mechanical properties required forthe foam.A second form of waste treatment,concentrating radio-activity into a restricted volume, consists in incinera-ting, where applicable, its organic fraction. Three pro-cess families have been developed. The first two(conventional incineration in the Iris pilot,at Marcoule,and supercritical-water incineration, by the so-calledhydrothermal oxidation [HTO] process) have reachedthe industrial development stage. The third process, acoupled incineration–vitrification process (in parti-cular through use of a plasma torch), is described inSafe conditioning obtained through constantly impro-ved processes, p. 44.The Iris pilot has allowed full-scale development oforganic waste incineration (overwhelmingly,solid wasteto date).The initial pyrolysis step is carried out at mode-rate temperature (around 500 °C),allowing extractionof combustion gases, chlorinated gases in particular,in conditions of controlled materials corrosion, follo-wed by treatment of these gases by afterburning, dustfiltration and scrubbing. The second step consists inincineration of the pitch residue thus obtained. Thisprocess, adapted for nuclear materials, is operationalat CEA’s Valduc Center,for contaminated organic waste.Support for this application, of an industrial charac-ter,has allowed broadening the range of waste treated:thus, pure PVC waste may now be treated, this beingparticularly chlorine-rich.As for HTO incineration, this is used for partly orga-nic, partly aqueous, contaminated organic effluents.This technique allows complete oxidation of organicmatter into CO2 and H2O, mineral matter thus beingpresent, at the outcome of the process, in the form ofan aqueous solution. For this purpose, the processemploys water in the supercritical phase, at about 250bars and 500 °C. A double-walled tubular reactor wasdeveloped, to minimize corrosion effects when themineral fraction is chlorine-rich,for instance.This pro-cess was developed successfully to a throughput of100 g/h, and is to be used in the Delos installation ofthe Atalante facility,at Marcoule,to incinerate effluentsfrom CEA research activities, with a throughput of1kg/h.

Retrieval of legacy waste…

One major development, in terms of waste manage-ment, concerns operations of legacy waste retrievaland conditioning,scheduled by Areva and CEA throughto 2035 at La Hague, Marcoule, and other CEA sites.

… at Areva–La Hague…

At the La Hague site,plans are in hand to proceed withretrieval and vitrification conditioning of so-calledUMo fission product solutions (uranium–molybde-num solutions, yielded by treatment of UNGG fuel),currently in storage.Owing to their high molybdenumcontent, these cannot be vitrified using current facili-C

EAC

EA

Decontamination bymeans of a vacuumable

gel of a cell in the ISAIinstallation, at CEA’s

Marcoule Center.

Figure 2. Decontamination bymeans of a vacuumablecerium gel.

CLEFS CEA - No. 53 - WINTER 2005-2006 31

ties,and R&D investigations have allowed deploymentof a suitable process, consisting in fabricating the glassinvolved,or rather a vitroceramic matrix, in a cold cru-cible. Projected production is 800 standard vitrifiedwaste containers.Vitrification of UMo solutions shouldbe initiated after 2010.Further retrieval and conditioning operations concer-ning stored legacy waste, predominantly produced inthe 1970s and 1980s, are also planned for the comingyears. Areva is considering making use of its recentinstallations (put on stream between 1990 and 2002)to carry out its treatment and conditioning. The mainoperations concern sludges stored at the STE2 facility(retrieved from 2005), and waste stored in the HAO(Haute Activité oxyde: High-Level Oxide) silo,silos 115and 130, and at the SOC (Stockage organisé des coques:Organized Hull Disposal) site. Retrieval at the HAOsilo and silo 130 is scheduled from 2010.The companyis planning to carry out compaction conditioning, atthe Hull Compaction Workshop (ACC: Atelier de com-pactage des coques; see Industrial solutions for long-lived,high- and intermediate-level waste, p. 36), of a majorpart of the waste being stored in silos, mostly of theILW-LL type. Graphite sleeves, due to be taken out ofstorage at that point,are of the LLW-LL type,and shouldultimately be conditioned in concrete packages.Thereis also a certain amount of operational waste, at firstblush suited for surface disposal.Areva is further plan-ning to effect retrieval and conditioning of variouswaste categories (resins, powdered graphite, etc.), sto-red in the settling tanks of the UP2-400 plant, alongwith treatment and conditioning of used solvents fromthat same plant.Most of this waste will be conditioned by 2020, andthe final operations should come to an end around2030.Some of these operations will require prior agree-ment from the French Nuclear Safety Authority(Autorité de sûreté nucléaire). Cleanup operations atshutdown workshops of UP2-400 should largely be inhand by 2020.

… at Marcoule…

Cleanup operations at the Marcoule site will carrythrough. Part of the waste may be directed to surfacedisposal sites, provided agreements are forthcomingfrom ANDRA. For other waste (HLW and ILW-LL),the aim is to bring it all into two storage facilities, thatat AVM (Atelier de vitrification de Marcoule: MarcouleVitrification Workshop), for storage in wells, for vitri-fied HLW waste packages, and at the modular storagehall specifically built for these operations, EIP(Entreposage intermédiare polyvalent: MultipurposeInterim Storage), put in service in 2000, for ILW-LLwaste.The strategy for magnesium structural waste (clad-ding from spent fuel of the UNGG type) is to condi-tion it in stainless steel drums.Retrieval operations forsuch waste, stored in 17 trenches at the site, have notbeen initiated as yet. Drums of bitumen produced bythe Liquid Effluent Treatment Station (STEL: Stationde traitement des effluents liquides) have been stored,since the unit started up in 1996, for some 90%, inconcrete bunkers adjoining it and in semi-subterra-nean trenches in the site’s Northern area, from whichthey have been gradually retrieved from 2000 on. As

of 31 December 2004,4,184 packages containing drumstaken from the trenches in the Northern area were instorage at EIP.By 2020, nearly half of the retrieval and conditioningoperations for legacy waste stored at the Marcoule sitewill have been completed,along with most of the instal-lation shutdown and decommissioning operations(with the major exception of support workshops usedfor the treatment and conditioning of legacy waste notyet conditioned by that time).

… and at other CEA sites

CEA has launched a major program for the retrievalof legacy waste disposed before the 1990s.Some of thatwaste has been kept at R&D installations, other itemsare stored in dedicated installations in CEA centers.The packages resulting from this program will be direc-ted for disposal at the Aube Center, or stored at CEAinstallations.This mainly involves solid waste,as a rule kept in drums,stored in wells, vaults or trenches at the various sites,and solid waste stored in the 5 trenches of basic nuclearinstallation (INB) 56 at Cadarache,in a variety of forms(loose in vinyl bags,metal drums,concrete overpacks);current schedule is for retrieval of this waste to be com-pleted by 2008, with site rehabilitation achieved by2009. Other solid waste, stored in hangars at INB 56,already conditioned and/or requiring reconditioning,is to be subjected to radiological measurements for thepurposes of suitability sorting for the LILW-SL andILW-LL pathways. Such operations are dependent on

the coming on stream of the Radioactive WasteConditioning and Storage (CEDRA: Conditionnementet entreposage des déchets radioactifs) installation, cur-rently being built,clearance of all packages being spreadover 6 years, from the unit’s startup date. It should benoted that, while retrieval of legacy solid waste doesrequire specific tools or installations, the waste resul-ting from these specific treatments, for its part,may betaken in by conventional pathways.

> Marc ButezNuclear Energy Division

Real Estate and Cleanup DirectorateCEA Saclay Center

> Gilles Bordier* and Xavier Vitart**Nuclear Energy Division

CEA Valrhô–Marcoule* and Saclay** Centers> Isabelle Hablot

Treatment Business UnitAreva

Waste storage hall atCEA/Saclay, comprising 100wells, each with a 10-drumcapacity.

P. S

trop

pa/C

EA

Mockup of the CEDRA project, a solid radioactive waste treatment and storage installation at CEA’s Cadarache Center.

CEA

A

According to the International AtomicEnergy Agency (IAEA), radioactive

waste may be defined as “any material forwhich no use is foreseen and that containsradionuclides at concentrations greaterthan the values deemed admissible by thecompetent authority in materials suitablefor use not subject to control.” French lawin turn introduces a further distinction,valid for nuclear waste as for any otherwaste, between waste and final, or “ulti-mate,” waste (déchet ultime). Article L.541-1 of the French Environmental Codethus specifies that “may be deemed aswaste any residue from a process of pro-duction, transformation or use, any sub-stance, material, product, or, more gene-rally, any movable property left derelict orthat its owner intends to leave derelict,”further defining as ultimate “waste, be itthe outcome of waste treatment or not,that is not amenable to further treatmentunder prevailing technological and eco-nomic conditions, in particular by extrac-tion of the recoverable, usable part, ormitigation of its polluting or hazardouscharacter.”Internationally, experts from IAEA and theNuclear Energy Agency (NEA) – an OECDorganization – as those in the EuropeanCommission find that long-lived waste pro-duced in countries operating a nuclearpower program is stored securely nowa-days, whilst acknowledging a final solu-tion is required, for the long-term mana-gement of such waste. They consider burialin deep geological structures appears, pre-sently, to be the safest way to achieve finaldisposal of this type of waste.

What constitutes radioactivewaste? What are the volumescurrently involved?Radioactive waste is classified into a num-ber of categories, according to its level ofradioactivity, and the radioactive period,or half-life, of the radionuclides itcontains. It is termed long-lived wastewhen that period is greater than 30 years,short-lived waste otherwise. The Frenchclassification system involves the follo-wing categories:– very-low-level waste (VLLW); thiscontains very small amounts of radionu-clides, of the order of 10–100 Bq/g (bec-querels per gram), which precludes consi-dering it as conventional waste;– short-lived low and intermediate levelwaste (LILW-SL); radioactivity levels forsuch waste lie as a rule in a range from

a few hundred to one million Bq/g, ofwhich less than 10,000 Bq/g is from long-lived radionuclides. Its radioactivity beco-mes comparable to natural radioactivityin less than three hundred years.Production of such waste stands at some15,000 m3 per year in France;– long-lived low-level waste (LLW-LL);this category includes radium-bearingwaste from the extraction of rare earthsfrom radioactive ore, and graphite wastefrom first-generation reactors;– long-lived intermediate-level waste(ILW-LL), this being highly disparate, whe-ther in terms of origin or nature, with anoverall stock standing, in France, at45,000 m3 at the end of 2004. This mainlycomes from spent fuel assemblies (clad-ding hulls and end-caps), or from opera-tion and maintenance of installations; thisincludes, in particular, waste conditionedduring spent fuel reprocessing operations(as from 2002, this type of waste is com-pacted, amounting to some 200 m3

annually), technological waste from theoperation or routine maintenance of pro-duction or fuel-processing plants, fromnuclear reactors or from research cen-ters (some 230 m3 annually), along withsludges from effluent treatment (less than100 m3 annually). Most such waste gene-rates little heat, however some waste ofthis type is liable to release gases;– high-level waste (HLW), containing fis-sion products and minor actinides parti-tioned during spent fuel reprocessing (seeBox B), and incorporated at high tempe-rature into a glass matrix. Some 120 m3

of “nuclear glass” is thus cast every year.This type of waste bears the major partof radioactivity (over 95%), consequentlyit is the seat of considerable heat release,this remaining significant on a scale ofseveral centuries.Overall, radioactive waste conditioned inFrance amounts to less than 1 kg per year,per capita. That kilogram consists, forover 90%, of LILW-SL type waste, bearingbut 5% of total radioactivity; 9% of ILW-LL waste, less than 1% HLW, and virtuallyno LLW-LL waste.

What of the waste of tomorrow?From 1991, ANDRA compiled, on a yearlybasis, a geographical inventory of wastepresent on French territory. In 2001,ANDRA was asked by government to aug-ment this “National Inventory,” with thethreefold aim of characterizing extantstocks (state of conditioning, processing

traceability), predicting future waste pro-duction trends to 2020, and informing thepublic (see An inventory projecting into thefuture). ANDRA published this referenceNational Inventory at the end of 2004. Tomeet requirements for research in com-pliance with the directions set out in theFrench Act of 30 December 1991 (seeRadioactive waste management research:an ongoing process of advances), ANDRA,in collaboration with waste producers,has drawn up a Dimensioning InventoryModel (MID: Modèle d’inventaire dedimensionnement), for the purposes ofarriving at estimates of the volume ofwaste packages to be taken on board inresearch along direction 2 (disposal). Thismodel, including as it does predictions asto overall radioactive waste arisings fromthe current reactor fleet, over their entirelifespan, seeks to group waste types intofamilies, homogeneous in terms of cha-racteristics, and to formulate the mostplausible hypotheses, with respect toconditioning modes, to derive the volu-mes to be taken on board for the purpo-ses of the investigation. Finally, MID setsout to provide detailed stocktaking, inten-ded to cover waste in the broadest pos-sible fashion. MID (not to be confused withthe National Inventory, which has theremit to provide a detailed account ofactual waste currently present on Frenchterritory) thus makes it possible to bringdown the variety of package families to alimited number of representative objects,and to specify the requisite margins oferror, to ensure the design and assess-ment of disposal safety will be as robustas feasible, with respect to possible futurevariations in data.To ensure consistency between investi-gations carried out in accordance withdirection 2 and those along direction 3(conditioning and long-term storage), CEAadopted MID as input data. MID subsu-mes waste packages into standard pac-kage types, then computes the numberand volume of HLW and ILW-LL packa-ges, according to a number of scenarios,all based on the assumption that currentnuclear power plants will be operated for40 years, their output plateauing at400 TWhe per year.Table 1 shows the numbers and volumesfor each standard package type, for thescenario assuming a continuation of cur-rent strategy, with respect to spent fuelreprocessing: reprocessing of 79,200 UOXfuel assemblies and storage of 5,400 MOX

What is radioactive waste?

A

assemblies discharged from the currentPWR fleet, when operated over 40 years.

What forms does it come in?Five types of generic packages (also foundin MID) may be considered:• cementitious waste packages: ILW-LLwaste packages employing hydraulic-bin-der based materials as a conditioningmatrix, or as an immobilizing grout, or yetas a container constituent;• bituminized sludge packages: LLW andILW-LL waste packages, in which bitumenis used as confinement matrix for low- andintermediate-level residues from treat-ment of a variety of liquid effluents (fuelprocessing, research centers, etc.);• standard compacted waste packages(CSD-C: colis standard de déchets compac-tés): ILW-LL packages obtained throughcompaction conditioning of structural wastefrom fuel assemblies, and technologicalwaste from the La Hague workshops;• standard vitrified waste packages(CSD-V: colis standard de déchets vitrifiés):

HLW packages, obtained mainly throughvitrification of highly active solutions fromspent fuel reprocessing;• spent fuel packages: packages consis-ting in nuclear fuel assemblies dischargedfrom reactors; these are not considered tobe waste in France.The only long-lived waste packages to begenerated in any significant amounts bycurrent electricity production (see Box B)are vitrified waste packages and standardcompacted waste packages, the other typesof packages having, for the most part,already been produced, and bearing but asmall part of total radioactivity.

What is happening to this waste atpresent? What is to be done in thelong term?The goal of long-term radioactive wastemanagement is to protect humankind andits environment from the effects of thematerials comprised in this waste, mostimportantly from radiological hazards. Anyrelease or dissemination of radioactive

materials must thus be precluded, throughthe lasting isolation of such waste from theenvironment. This management is guidedby the following principles: to produce aslittle waste as practicable; limit its hazar-dous character as far as feasible; take intoaccount the specific characters of eachcategory of waste; and opt for measuresthat will minimize the burden (monitoring,maintenance) for future generations.As for all nuclear activities subject to controlby the French Nuclear Safety Authority(Autorité de sûreté nucléaire), fundamentalsafety regulations (RFSs: règles fonda-mentales de sûreté) have been drawn upwith respect to radioactive waste mana-gement: sorting, volume reduction, pac-kage confinement potential, manufactu-ring method, radionuclide concentration.RFS III-2.f, in particular, specifies the condi-tions to be met for the design of, anddemonstration of safety for an undergroundrepository, and thus provides a basic guidefor disposal investigations. Industrial solu-tions (see Industrial solutions for all low-level waste) are currently available for nighon 85% (by volume) of waste, i.e. VLLW andLILW-SL waste. A solution for LLW-LLwaste is the subject of ongoing investiga-tion by ANDRA, at the behest of waste pro-ducers. ILW-LL and HLW waste, contai-ning radionuclides having very longhalf-lives (in some cases, greater thanseveral hundred thousand years) are cur-rently held in storage installations comingunder the control of the Nuclear SafetyAuthority. What is to become of this wastein the long term, beyond this storage phase,is what the Act of 30 December 1991addresses (see Table 2).For all of these waste types, the FrenchNuclear Safety Authority is drawing up aNational Radioactive Waste ManagementPlan, specifying, for each type, a manage-ment pathway.

MID standard package types Symbols Producers Categories Number Volume (m3)

Vitrified waste packages CO — C2 Cogema* HLW 42,470 7,410

Activated metal waste packages B1 EDF ILW-LL 2,560 470

Bituminized sludge packages B2 CEA, Cogema* ILW-LL 105,010 36,060

Cemented technological waste packages B3 CEA, Cogema* ILW-LL 32,940 27,260

Cemented hull and end-cap packages B4 Cogema* ILW-LL 1,520 2,730

Compacted structural and technological waste packages B5 Cogema* ILW-LL 39,900 7,300

Containerized loose structural and technological B6 Cogema* ILW-LL 10,810 4,580waste packages

Total B 192,740 78,400Total overall 235,210 85,810

Short-lived Long-livedHalf-life < 30 years Half-life > 30 years

for the main elements

Very-low-level Morvilliers dedicated disposal facility (open since 2003)waste (VLLW) Capacity: 650,000 m3

Low-level waste Dedicated disposal facility under(LLW) investigation for radium-bearing

waste (volume: 100,000 m3)Aube Center and graphite waste

(open since 1992) (volume: 14,000 m3)Intermediate-level Capacity: 1 million m3

waste (ILW) MID volume estimate: 78,000 m3

High-level waste MID volume estimate: 7,400 m3

(HLW)

Table 1. Amounts (number, and volume) of waste packages, as predicted in France for 40 years’ operation of the current fleet of reactors, according to ANDRA’sDimensioning Inventory Model (MID).

Table 2. Long-term management modes, as currently operated, or planned, in France, byradioactive waste category. The orange area highlights those categories targeted byinvestigations covered by the Act of 30 December 1991.

(1) According to the Dimensioning Inventory Model (MID)

* renamed Areva NC in 2006

(next)

B

Most high-level (high-activity) radio-active waste (HLW) originates, in

France, in the irradiation, inside nuclearpower reactors, of fuel made up fromenriched uranium oxide (UOX) pellets, oralso, in part, from mixed uranium andplutonium oxide (MOX). Some 1,200 ton-nes of spent fuel is discharged annuallyfrom the fleet of 58 pressurized-waterreactors (PWRs) operated by EDF, sup-plying over 400 TWh per year, i.e. morethan three quarters of French nationalpower consumption.The fuel’s composition alters, during itsirradiation inside the reactor. Shortly afterdischarge, fuel elements contain, on ave-rage,(1) some 95% residual uranium, 1%plutonium and other transuranic ele-ments – up to 0.1% – and 4% of productsyielded by fission. The latter exhibit verysignificant radioactivity levels – to theextent this necessitates managementsafety measures requiring major indus-trial resources – of some 1017 Bq per tonneof initial uranium (tiU) (see Figure 1).The uranium found in spent fuel exhibitsa makeup that is obviously different fromthat of the initial fuel. The greater theirradiation, the higher the consumptionof fissile nuclei, and consequently thegreater the extent by which the uraniumwill have been depleted of the fissile iso-tope 235 (235U). Irradiation conditionsusually prevailing in reactors in the Frenchfleet, with an average fuel residence timeinside the reactor of some 4 years, for a

burnup rate close to 50 GWd/t, result inbringing down final 235U content to a valuequite close to that of natural uranium(less than 1%), entailing an energy poten-tial very close to the latter’s. Indeed, eventhough this uranium remains slightlyricher in the fissile isotope than naturaluranium, for which 235U content standsat 0.7%, the presence should also benoted, in smaller, though significant,amounts, of other isotopes having adverseeffects in neutronic or radiological terms(232U, 236U), that had not figured in theinitial fuel (see Table 1).

The plutonium present in spent fuel isyielded by successive neutron captureand decay processes. Part of the Pu isdissipated through fission: thus aboutone third of the energy generated is yiel-ded by “in situ recycling” of this element.These processes further bring about theformation of heavy nuclei, involving, whe-ther directly themselves, or through theirdaughter products, long radioactive half-lives. These are the elements of the acti-nide family, this including, essentially,plutonium (from 238Pu to 242Pu, the odd-numbered isotopes generated in partundergoing fission themselves duringirradiation), but equally neptunium (Np),americium (Am), and curium (Cm), knownas minor actinides (MAs), owing to the

Waste from the nuclear power cycle

(1) These figures should be taken as indicative values. They allow orders of magnitude to bepinpointed for enriched-uranium oxide fuel, taken from the main current French nuclear powerpathway; they do depend, however, on a number of parameters, such as initial fuel composition andirradiation conditions, particularly irradiation time.

Figure 1.The main elements found in spent nuclear fuel.

element isotope half-life(years) isotope quantity isotope quantity isotope quantity isotope quantity

content (g/tiU) content (g/tiU) content (g/tiU) content (g/tihm)(%) (%) (%) (%)

234 246,000 0.02 222 0.02 206 0.02 229 0.02 112

235 7.04·108 1.05 10,300 0.74 6,870 0.62 5,870 0.13 1,070U

236 2.34·107 0.43 4,224 0.54 4,950 0.66 6,240 0.05 255

238 4.47·109 98.4 941,000 98.7 929,000 98.7 911,000 99.8 886,000

238 87.7 1.8 166 2.9 334 4.5 590 3.9 2,390

239 24,100 58.3 5,680 52.1 5,900 48.9 6,360 37.7 23,100

Pu 240 6,560 22.7 2,214 24,3 2,760 24.5 3,180 32 19,600

241 14.4 12.2 1,187 12.9 1,460 12.6 1,640 14.5 8,920

242 3.75·105 5.0 490 7.8 884 9.5 1,230 11.9 7,300

UOX 33 GWd/tiU UOX 45 GWd/tiU UOX 60 GWd/tiU MOX 45 GWd/tihm(E 235U: 3.5%) (E 235U: 3.7%) (E 235U: 4.5%) (Ei Pu: 8.65%)

Table 1.Major actinide inventory for spent UOX and MOX fuel after 3 years’ cooling, for a variety of enrichment and burnup rates. Burnup rate and quantity areexpressed per tonne of initial uranium (tiU) for UOX, per tonne of initial heavy metal (tihm) for MOX.

H1

Li3

Na11

K19

Rb37

Cs55

Fr87

Be4

Mg12

Ca20

Sr38

Ba56

Ra88

Sc21

Y39

Ln

An

Ti22

Zr40

Hf72

Rf104

V23

Nb41

Ta73

Db105

Cr24

Mo42

W74

Sg106

Mn25

Tc43

Re75

Bh107

Fe26

Ru44

Os76

Hs108

Co27

Rh45

Ir77

Mt109

Ni28

Pd46

Pt78

Uun110

Ac89

Th90

Pa91

U92

Np93

Pu94

Am95

Cm96

Bk97

Cf98

Es99

Fm100

Md101

No102

Lr103

Cu29

Ag47

Au79

Zn30

Cd48

Hg80

Ga31

In49

TI81

Ge32

Sn50

Pb82

As33

Sb51

Bi83

Se34

Te52

Po84

Br35

I53

At85

Kr36

Al13

Si14

P15

S16

Cl17

Ar18

B5

C6

N7

O8

F9

Ne10

He2

Xe54

Rn86

La57

Ce58

Pr59

Nd60

Pm61

Sm62

Eu63

Gd64

Tb65

Dy66

Ho67

Er68

Tm69

Yb70

Lu71

■ heavy nuclei■ fission products

long-lived radionuclides

■ activation products■ fission and activation products

actinides

lanthanides

B (next)lesser abundance of these elements, com-pared with that of U and Pu, the latter beingtermed major actinides.Activation processes affecting nuclei of non-radioactive elements mainly involve struc-tural materials, i.e. the materials of thetubes, grids, plates and end-fittings thatensure the mechanical strength of nuclearfuel. These materials lead, in particular,to formation of carbon 14 (14C), with a half-life of 5,730 years, in amounts that arehowever very low, much less than one gramper tonne of initial uranium (g/tiU) in usualconditions.It is the products yielded by fission of theinitial uranium 235, but equally of the Pugenerated (isotopes 239 and 241), knownas fission products (FPs), that are theessential source of the radioactivity ofspent fuel, shortly after discharge. Over300 radionuclides – two thirds of whichhowever will be dissipated through radio-active decay in a few years, after irradia-tion – have been identified. These radio-nuclides are distributed over some40 elements in the periodic table, fromgermanium (32Ge) to dysprosium (66Dy),with a presence of tritium from fission, i.e.from the fission into three fragments (ter-nary fission) of 235U. They are thus cha-racterized by great diversity: diverse radio-active properties, involving as they do somehighly radioactive nuclides having very

short lifespans, and conversely othershaving radioactive half-lives counted inmillions of years; and diverse chemicalproperties, as is apparent from the ana-lysis, for the “reference” fuels used inPWRs in the French fleet, of the break-down of FPs generated, by families in theperiodic table (see Table 2). These FPs,along with the actinides generated, are,for the most part, present in the form ofoxides included in the initial uranium oxide,which remains by far the majority consti-tuent. Among some notable exceptionsmay be noted iodine (I), present in the formof cesium iodide, rare gases, such as kryp-ton (Kr) and xenon (Xe), or certain noblemetals, including ruthenium (Ru), rho-dium (Rh), and palladium (Pd), which mayform metallic inclusions within the oxidematrix.Pu is recycled nowadays in the form ofMOX fuel, used in part of the fleet (some20 reactors currently). Residual U may inturn be re-enriched (and recycled as a sub-stitute for mined uranium). Recyclingintensity depends on market prices fornatural uranium, the recent upturn inwhich should result in raising the currentrecycling rate (about one third being recy-cled at present).Such U and Pu recycling is the foundationfor the reprocessing strategy currentlyimplemented in France, for the major partof spent fuel (some two thirds currently).

For the 500 kg or so of U initially contai-ned in every fuel element, and after par-titioning of 475 kg of residual U and about5 kg Pu, this “ultimate” waste amountsto less than 20 kg of FPs, and less than500 grams MAs. This waste managementpathway (otherwise know as the closedcycle), consisting as it does in reproces-sing spent fuel now, to partition recove-rable materials and ultimate waste, dif-fers from strategies whereby spent fuelis conserved as-is, whether this be due toa wait-and-see policy (pending a decisionon a long-term management mode), or toa so-called open cycle policy, wherebyspent fuel is considered to be waste, anddesignated for conditioning into contai-ners, and disposal as-is.In the nuclear power cycle, as it is imple-mented in France, waste is subdivided intotwo categories, according to its origin.Waste directly obtained from spent fuel isfurther subdivided into minor actinidesand fission products, on the one hand, andstructural waste, comprising hulls (seg-ments of the cladding tubes that had heldthe fuel for PWRs) and end-caps (fittingsforming the end-pieces of the fuel assem-blies for these same PWRs), on the otherhand. The process used for spent fuelreprocessing, to extract U and Pu, alsogenerates technological waste (operatio-nal waste, such as spare parts, protec-tion gloves…) and liquid effluents.

UOX 33 GWd/tiU UOX 45 GWd/tiU UOX 60 GWd/tiU MOX 45 GWd/tihmfamily (E 235U: 3.5%) (E 235U: 3.7%) (E 235U: 4.5%) (Ei Pu: 8.65%)

quantity (kg/tiU) quantity (kg/tiU) quantity (kg/tiU) quantity (kg/tihm)

rare gases (Kr, Xe) 5.6 7.7 10.3 7

alkali metals(Cs, Rb) 3 4 5.2 4.5

alkaline-earthmetals (Sr, Ba) 2.4 3.3 4.5 2.6

Y and lanthanides 10.2 13.8 18.3 12.4

zirconium 3.6 4.8 6.3 3.3

chalcogens(Se, Te) 0.5 0.7 1 0.8

molybdenum 3.3 4.5 6 4.1

halogens (I, Br) 0.2 0.3 0.4 0.4

technetium 0.8 1.1 1.4 1.1

Ru, Rh, Pd 3.9 5.7 7.7 8.3

miscellaneous:Ag, Cd, Sn, Sb… 0.1 0.2 0.3 0.6

Table 2.Breakdown by chemical family of fission products in spent UOX and MOX fuel, after 3 years’ cooling,for a variety of enrichment and burnup rates.

After discharge, spent fuel is stored in cooling pools, to allow its radioactivity to come down significantly. Shown here is a storage pool at Areva’s spent fuel reprocessing plant at La Hague.

Mag

num

/Har

ry G

ruya

ert

C

Raw, solid or liquid radioactive wasteundergoes, after characterization

(determination of its chemical and radio-logical makeup, and of its physical–che-mical properties), conditioning, a termcovering all the operations consisting inbringing this waste (or spent fuel assem-blies) to a form suitable for its transport,storage, and disposal (see Box D). Theaim is to put radioactive waste into asolid, physically and chemically stableform, and ensure effective, lasting confi-nement of the radionuclides it contains.For that purpose, two complementaryoperations are carried out. As a rule,waste is immobilized by a material –whether by encapsulation or homoge-neous incorporation (liquid or powderedwaste, sludges), or encasing (solid waste)– within a matrix, the nature of, and per-formance specification for which dependon waste type (cement for sludges, eva-poration concentrates and incinerationashes; bitumen for encapsulation ofsludges or evaporation concentratesfrom liquid effluent treatment; or avitreous matrix, intimately binding thenuclides to the glass network, for fis-sion product or minor actinide solutions).This matrix contributes to the confine-ment function. The waste thus conditio-ned is placed in an impervious contai-

ner (cylindrical or rectangular), consis-ting in one or more canisters. The whole– container and content – is termed apackage. Equally, waste may be com-pacted and mechanically immobilizedwithin a canister, the whole forming apackage.When in the state they come in as sup-plied by industrial production, they areknown as primary packages, the pri-

mary container being the cement ormetal container into which the conditio-ned waste is ultimately placed, to allowhandling. The container may act as initialconfinement barrier, allotment of func-tions between matrix and container beingdetermined according to the nature ofthe waste involved. Thus, the whole obtai-ned by the grouping together, within onecontainer, of a number of primary

Cross-section of an experimental storage borehole for a spent fuel container (the lower part of theassembly may be seen, top right), in the Galatée gallery of CECER (Centre d’expertise sur leconditionnement et l’entreposage des matières radioactives: Radioactive Materials Conditioningand Storage Expertise Center), at CEA’s Marcoule Center, showing the nested canisters.

A. G

onin

/CEA

What stands between waste and the environment?

C (next)ILW-LL packages may ensure confinementof the radioactivity of this type of waste.If a long-term storage stage is found tobe necessary, beyond the stage of indus-trial storage on the premises of the pro-ducers, primary waste packages mustbe amenable to retrieval, as and whenrequired: durable primary containersmust then be available, in such condi-tions, for all types of waste.In such a case, for spent fuel assemblieswhich might at some time be earmar-ked for such long-term storage, or evenfor disposal, it is not feasible to demons-trate, on a timescale of centuries, theintegrity of the cladding holding the fuel,forming the initial confinement barrierduring the in-reactor use stage. Securingthese assemblies in individual, imper-vious cartridges is thus being conside-red, this stainless-steel cartridge beingcompatible with the various possiblefuture management stages: treatment,return to storage, or disposal. Placingthese cartridges inside imperviouscontainers ensures a second confine-ment barrier, as is the case for high-level waste packages.In storage or disposal conditions, thewaste packages will be subjected to avariety of aggressive agents, both inter-nal and external. First, radionuclide

radioactive decay persists inside the pac-kage (self-irradiation process). Emissionof radiation is concomitant with heatgeneration. For example, in confinementglasses holding high-activity (high-level)waste, the main sources of irradiationoriginate in the alpha decay processesfrom minor actinides, beta decay fromfission products, and gamma transitions.Alpha decay, characterized by produc-tion of a recoil nucleus, and emission ofa particle, which, at the end of its path,yields a helium atom, causes the majorpart of atom displacements. In particu-lar, recoil nuclei, shedding considerableenergy as they do over a short distance,result in atom displacement cascades,thus breaking large numbers of chemi-cal bonds. This is thus the main causeof potential long-term damage. In suchconditions, matrices must exhibit ther-mal stability, and irradiation-damageresistance.Stored waste packages will also be sub-jected to the effects of water (leaching).Container canisters may exhibit a deg-ree of resistance to corrosion processes(the overpacks contemplated for glas-ses may thus delay by some 4,000 yearsthe arrival of water), and the confine-ment matrices must be proven to exhi-bit high chemical stability.

Between the containers and the ultimatebarrier provided, in a radioactive wastedeep disposal facility, by the geologicalenvironment itself, there may further beinterposed, apart, possibly, from an over-pack, other barriers, so-called engi-neered barriers, for backfill and sealingpurposes. While these would be point-less as backfill in clay formations, theywould have the capability, in other envi-ronments (granite), of further retardingany flow of radionuclides to the geo-sphere, notwithstanding degradation ofthe previously mentioned barriers.

CEA

Technological demonstratorsof ILW-LL packages for

bituminized sludges.

D

The object of nuclear waste storageand disposal is to ensure the long-

term confinement of radioactivity, inother words to contain radionuclideswithin a definite space, segre-gated from humankind and theenvironment, as long as requi-red, so that the possible returnto the biosphere of minuteamounts of radionuclides canhave no unacceptable health orenvironmental impact.According to the Joint Con-vention on the Safety of SpentFuel Management and on theSafety of Radioactive WasteManagement, signed on 5 Sep-tember 1997, “storage” means“the holding of spent fuel or ofradioactive waste in a facility thatprovides for its containment,with the intention of retrieval.”This is thus, by definition, aninterim stage, amounting to adelaying, or wait-and-see solu-tion, even though this may be fora very long time (from a fewdecades to several hundredyears), whereas disposal may befinal.Used from the outset of the nuclearpower age, industrial storage keepsspent fuel awaiting reprocessing, andconditioned high-level waste (HLW), orlong-lived intermediate-level waste

(ILW-LL) in conditions of safety, pen-ding a long-term management modefor such waste. Retrieval of stored pac-kages is anticipated, after a period oflimited duration (i.e. after a matter of

years, or tens of years).Long-term storage (LTS) may becontemplated, in particular, in the eventof the deferred deployment of a dispo-sal facility, or of reactors to carry out

recycling–transmutation, or simply toturn to advantage the natural decay ofradioactivity (and hence the falling offof heat release from high-level waste),before putting the waste into geologi-

cal disposal. By “long term” ismeant a timespan of up to 300years. Long-term storage maytake place in a surface or sub-surface facility. In the formercase, the site may be protected,for instance, by a reinforced-concrete structure. In the lattercase, it will be located at a depthof some tens of meters, and pro-tected by a natural environment(for instance, if buried in a hill-side) and its host rock.Whichever management stra-tegy is chosen, it will be impe-rative to protect the biospherefrom the residual ultimate waste.The nature of the radioelementsthe latter contains means a solu-tion is required that has the abi-lity to ensure their confinementover several tens of thousandyears, in the case of long-livedwaste, or even longer. On suchtimescales, social stability is amajor uncertainty that has to be

taken on board. Which is why disposalin deep geological strata (typically, 500m down) is seen as a reference solu-tion, insofar as it inherently makes fordeployment of a more passive techni-cal solution, with the ability to stand,with no increased risk, an absence ofsurveillance, thus mitigating a possibleloss of memory on the part of society.The geological environment of such adisposal facility thus forms a further,essential barrier, which does not existin the storage case.A disposal facility may be designed tobe reversible over a given period. Theconcept of reversibility means the designmust guarantee the ability, for a varietyof reasons, to access the packages, oreven to take them out of the facility, overa certain timespan, or to opt for the finalclosure of the disposal facility. Suchreversibility may be envisaged as a suc-cession of stages, each affording adecreasing “level of reversibility.” Tosimplify, each stage consists in carryingout one further technical operation brin-ging the facility closer to final closure,making retrieval more difficult than atthe previous stage, according to well-specified criteria.

From storage to disposal

ANDRA design for the disposal of standard vitrified waste packages in horizontal galleries,showing in particular the packages’ various canisters, and some characteristics linked to potential reversibility of the disposal facility.

CEA design study for a common container for the long-term storage and disposal of long-lived, intermediate-level waste.

AN

DR

A

A. G

onin

/CEA

lid

1.30m to 1.60 m

0.57 m to 0.64 m

primary package

disposal packageslining

intercalary

excavated diameter: 0.7 m approx.

length: 40 m approx.

steel overpack

E

Transmutation is the transformationof one nucleus into another, through

a reaction induced by particles with whichit is bombarded. As applied to the treat-ment of nuclear waste, this consists inusing that type of reaction to transformlong-lived radioactive isotopes into iso-topes having a markedly shorter life, oreven into stable isotopes, in order toreduce the long-term radiotoxic inven-tory. In theory, the projectiles used maybe photons, protons, or neutrons.In the first case, the aim is to obtain, bybremsstrahlung,(1) through bombardmentof a target by a beam of electrons, pro-vided by an accelerator, photons able tobring about reactions of the (γ, xn) type.Under the effects of the incoming gammaradiation, x neutrons are expelled fromthe nucleus. When applied to substan-ces that are too rich in neutrons, andhence unstable, such as certain fissionproducts (strontium 90, cesium 137…),such reactions yield, as a rule, stable sub-stances. However, owing to the very lowefficiency achieved, and the very highelectron current intensity required, thispath is not deemed to be viable.In the second case, the proton–nucleusinteraction induces a complex reaction,known as spallation, resulting in frag-mentation of the nucleus, and the release

of a number of particles, including high-energy neutrons. Transmutation by wayof direct interaction between protons isuneconomic, since this would involve, inorder to overcome the Coulomb barrier,(2)

very-high-energy protons (1–2 GeV),requiring a generating energy greaterthan had been obtained from the processthat resulted in producing the waste. Onthe other hand, indirect transmutation,using very-high-energy neutrons (ofwhich around 30 may be yielded, depen-ding on target nature and incoming protonenergy), makes it possible to achieve verysignificantly improved performance. Thisis the path forming the basis for thedesign of so-called hybrid reactors, cou-pling a subcritical core and a high-inten-sity proton accelerator (see Box F, Whatis an ADS?).The third particle that may be used is thusthe neutron. Owing to its lack of electriccharge, this is by far the particle best sui-ted to meet the desired criteria. It is “natu-rally” available in large quantities insidenuclear reactors, where it is used to trig-ger fission reactions, thus yielding energy,while constantly inducing, concurrently,transmutations, most of them unsought.The best recycling path for waste wouldthus be to reinject it in the very installa-tion, more or less, that had produced it…

When a neutron collides with a nucleus,it may bounce off the nucleus, or pene-trate it. In the latter case, the nucleus,by absorbing the neutron, gains excessenergy, which it then releases in variousways:• by expelling particles (a neutron, e.g.),while possibly releasing radiation;• by solely emitting radiation; this isknown as a capture reaction, since theneutron remains captive inside thenucleus;• by breaking up into two nuclei, of moreor less equal size, while releasing concur-rently two or three neutrons; this is knownas a fission reaction, in which considera-ble amounts of energy are released.Transmutation of a radionuclide may beachieved either through neutron captureor by fission. Minor actinides, as elementshaving large nuclei (heavy nuclei), mayundergo both fission and capture reac-tions. By fission, they transform intoradionuclides that, in a majority of cases,are short-lived, or even into stable nuclei.The nuclei yielded by fission (known asfission products), being smaller, are onlythe seat of capture reactions, undergoing,on average, 4 radioactive decays, with ahalf-life not longer than a few years, asa rule, before they reach a stable form.Through capture, the same heavy nucleitransform into other radionuclides, oftenlong-lived, which transform in turnthrough natural decay, but equallythrough capture and fission.

What is transmutation?

(1) From the German for “braking radiation.” High-energy photon radiation, yielded by accelerated(or decelerated) particles (electrons) following a circular path, at the same time emitting brakingphotons tangentially, those with the highest energies being emitted preferentially along the electronbeam axis.

(2) A force of repulsion, which resists the drawing together of same-sign electric charges.

E (next)

The probability, for a neutron, of causinga capture or a fission reaction is evalua-ted on the basis, respectively, of its cap-ture cross-section and fission cross-sec-tion. Such cross-sections depend on thenature of the nucleus (they vary consi-derably from one nucleus to the next, and,even more markedly, from one isotopeto the next for the same nucleus) andneutron energy.For a neutron having an energy lowerthan 1 eV (in the range of slow, or ther-mal, neutrons), the capture cross-sec-

tion prevails; capture is about 100 timesmore probable than fission. This remainsthe case for energies in the 1 eV–1 MeVrange (i.e., that of epithermal neutrons,where captures or fissions occur at defi-nite energy levels). Beyond 1 MeV (fastneutron range), fissions become moreprobable than captures.Two reactor pathways may be conside-red, according to the neutron energyrange for which the majority of fissionreactions occur: thermal-neutron reac-tors, and fast-neutron reactors. The ther-

mal neutron pathway is the technologyused by France for its power generationequipment, with close to 60 pressurized-water reactors. In a thermal-neutronreactor, neutrons yielded by fission areslowed down (moderated) through colli-sions against light nuclei, making upmaterials known as moderators. Due tothe moderator (common water, in thecase of pressurized-water reactors), neu-tron velocity falls off, down to a few kilo-meters per second, a value at which neu-trons find themselves in thermalequilibrium with the ambient environ-ment. Since fission cross-sections for235U and 239Pu, for fission induced bythermal neutrons, are very large, aconcentration of a few per cent of thesefissile nuclei is sufficient to sustain thecascade of fissions. The flux, in a ther-mal-neutron reactor, is of the order of1018 neutrons per square meter, persecond.In a fast-neutron reactor, such as Phénix,neutrons yielded by fission immediatelyinduce, without first being slowed down,further fissions. There is no moderator inthis case. Since, for this energy range,cross-sections are small, a fuel rich infissile radionuclides must be used (up to20% uranium 235 or plutonium 239), if theneutron multiplication factor is to be equalto 1. The flux in a fast-neutron reactor isten times larger (of the order of 1019 neu-trons per square meter, per second) thanfor a thermal-neutron reactor.

Figure.Simplified representation of the evolution chain of americium 241 in a thermal-neutron reactor(shown in blue: radionuclides disappearing through fission). Through capture, 241Am transformsinto 242mAm, this disappearing predominantly through fission, and into 242Am, which mainly decays(with a half-life of 16 hours) through beta decay into 242Cm. 242Cm transforms through alpha decayinto 238Pu, and through capture into 243Cm, which itself disappears predominantly through fission.238Pu transforms through capture into 239Pu, which disappears predominantly through fission.

241Am432ans

242Cm163days

242Am16h

242mAm141

years

238Pu87.8

years

240Pu6.5.103

years

239Pu2.4.104

years

241Pu14.4

years

243Cm32

years

86%

98%36%82%

capture

fission

alphadecay

beta-minusdecay

electroncapture

24%

86%

4%

1%95%

11%

15%

12%

2%

5%13% 64% 2% 69%

85%14%

7%

3%86%

F

An ADS (accelerator-driven system) isa hybrid system, comprising a

nuclear reactor operating in subcriticalmode, i.e. a reactor unable by itself to sus-tain a fission chain reaction, “driven” byan external source, having the ability tosupply it with the required comple-ment of neutrons.(1)

Inside the core of a nuclear reactor,indeed, it is the fission energy fromheavy nuclei, such as uranium 235or plutonium 239, that is released.Uranium 235 yields, when under-going fission, on average 2.5 neu-trons, which can in turn induce afurther fission, if they collide with auranium 235 nucleus. It may thusbe seen that, once the initial fissionis initiated, a chain reaction may develop,resulting, through a succession of fis-sions, in a rise in the neutron population.However, of the 2.5 neutrons yielded bythe initial fission, some are captured, thusnot giving rise to further fissions. Thenumber of fissions generated from oneinitial fission is characterized by the effec-tive multiplication factor keff, equal to theratio of the number of fission neutronsgenerated, over the number of neutronsdisappearing. It is on the value of this coef-ficient that the evolution of the neutronpopulation depends: if keff is markedlyhigher than 1, the population increasesrapidly; if it is slightly higher than 1, neu-tron multiplication sets in, but remainsunder control; this is the state desired atreactor startup; if keff is equal to 1, thepopulation remains stable; this is the state

for a reactor in normal operating condi-tions; and, if keff is lower than 1, the neu-tron population dwindles, and becomesextinct, unless – as is the case for a hybridsystem – an external source provides aneutron supply.

From the effective multiplication factor,a reactor’s reactivity is defined by the ratio(keff –1)/keff. The condition for stability isthen expressed by zero reactivity. To sta-bilize a neutron population, it is sufficientto act on the proportion of materials exhi-biting a large neutron capture cross-sec-tion (neutron absorber materials) insidethe reactor.In an ADS, the source of extra neutronsis fed with protons, generated with anenergy of about 100 keV, then injectedinto an accelerator (linear accelerator orcyclotron), which brings them to an energyof around 1 GeV, and directs them to aheavy-metal target (lead, lead–bismuth,tungsten or tantalum). When irradiatedby the proton beam, this target yields,through spallation reactions, an intense,high-energy (1–20 MeV) neutron flux, onesingle incoming neutron having the abi-lity to generate up to 30 neutrons. The lat-

ter then go on to interact with the fuel ofthe subcritical neutron multipliermedium, yielding further neutrons (fis-sion neutrons) (see Figure).Most hybrid system projects use as a core(of annular configuration, as a rule) fast-

neutron environments, since thesemake it possible to achieve neu-tron balances most favorable totransmutation, an operation thatallows waste to be “burned,” butwhich may equally be used to yieldfurther fissile nuclei. Such a sys-tem may also be used for energygeneration, even though part ofthis energy must be set aside topower the proton accelerator, apart that is all the higher, the more

subcritical the system is. Such a systemis safe in principle from most reactivityaccidents, its multiplication factor beinglower than 1, contrary to that of a reac-tor operated in critical mode: the chainreaction would come to a halt, if it wasnot sustained by this supply of externalneutrons.A major component in a hybrid reactor,the window, positioned at the end of thebeam line, isolates the accelerator fromthe target, and makes it possible to keepthe accelerator in a vacuum. Traversedas it is by the proton beam, it is a sensi-tive part of the system: its lifespandepends on thermal and mechanicalstresses, and corrosion. Projects are moo-ted, however, of windowless ADSs. In thelatter case, it is the confinement cons-traints, and those of radioactive spalla-tion product extraction, that must be takenon board.

What is an ADS?

(1) On this topic, see Clefs CEA, No. 37, p. 14

Principle schematic of an ADS.

window

protonsource

accelerator

subcritical reactor

spallationtarget

providingexternalneutrons

100 keV 1 GeV

The characteristics of the major part of the radioactive waste generated in France are determined by those of the French nuclearpower generation fleet, and of the spent fuel reprocessing plants, built in compliance with the principle of reprocessing such fuel, topartition such materials as remain recoverable for energy purposes (uranium and plutonium), and waste (fission products and minoractinides), not amenable to recycling in the current state of the art.58 enriched-uranium pressurized-water reactors (PWRs) have been put on stream by French national utility EDF, from 1977(Fessenheim) to 1999 (Civaux), forming a second generation of reactors, following the first generation, which mainly comprised 8 UNGG(natural uranium, graphite, gas) reactors, now all closed down, and, in the case of the older reactors, in the course of decommis-sioning. Some 20 of these PWRs carry out the industrial recycling of plutonium, included in MOX fuel, supplied since 1995 by theMelox plant, at Marcoule (Gard département, Southern France).EDF is contemplating the gradual replacement of the current PWRs by third-generation reactors, belonging to the selfsame pres-surized-water reactor pathway, of the EPR (European Pressurized-Water Reactor) type, designed by Areva NP (formerly Framatome–ANP),a division of the Areva Group. The very first EPR is being built in Finland, the first to be built in France being sited at Flamanville(Manche département, Western France).The major part of spent fuel from the French fleet currently undergoes reprocessing at the UP2-800(1) plant, which has been opera-ted at La Hague (Manche département), since 1994, by Areva NC (formerly Cogema,) another member of the Areva Group (the UP3plant, put on stream in 1990–92, for its part, carries out reprocessing of fuel from other countries). The waste vitrification workshopsat these plants, the outcome of development work initiated at Marcoule, give their name (R7T7) to the “nuclear” glass used for theconfinement of long-lived, high-level waste.A fourth generation of reactors could emerge from 2040 (along with new reprocessing plants), a prototype being built by 2020. Thesecould be fast-neutron reactors (i.e. fast reactors [FRs]), either sodium-cooled (SFRs) or gas-cooled (GFRs). Following the closingdown of the Superphénix reactor, in 1998, only one FR is operated in France, the Phénix reactor, due to be closed down in 2009.

The industrial context 1

(1) A reengineering of the UP2-400 plant, which, after the UP1 plant, at Marcoule, had been intended to reprocess spent fuel from the UNGG pathway.