203
2011

EDFMESS FR - inis.iaea.org

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Page 1: EDFMESS FR - inis.iaea.org

2011

Page 2: EDFMESS FR - inis.iaea.org
Page 3: EDFMESS FR - inis.iaea.org

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

AVANT-PROPOS

Ce rapport est destiné au président d’EDF et a pour objectif de lui présenter mon jugement sur l’état de la sûreté et de la radioprotection dans le groupe EDF. Il s’adresse aussi à tous ceux qui, de près ou de loin, participent à l’amélioration de la sûreté nucléaire. Il s’adresse également à toutes les personnes qui, en dehors de l’entreprise, dans le nucléaire, ou dans d’autres industries à risques, s’intéressent ou participent au débat. Il est une contribution au partage d’expérience avec les autres opérateurs nucléaires de WANO.1

L’amélioration continue de la performance de sûreté est un engagement fondamental du groupe EDF. Il se traduit en particulier par une démarche de transparence aussi bien en France qu’au Royaume-Uni, aux Etats-Unis et en Chine. Ce rapport y contribue, il constitue un regard indépendant des lignes opérationnelles. Il est fondé sur les informations et les observations recueillies auprès des équipes de terrain ainsi que lors d’entretiens avec des décideurs, des managers, des membres du corps médical, des acteurs du dialogue social et des parties prenantes extérieures, entreprises prestataires notamment. Cette vision de terrain en fait sa première richesse.

Ce rapport met donc l’accent sur les difficultés et les fragilités, plutôt que sur les forces et les progrès, sauf à l’étranger où il s’attarde davantage sur les bonnes pratiques que sur les difficultés. Il pourra donc paraître injuste à tous ceux qui n’ont pas mesuré leurs efforts pour assurer chaque jour, sans discontinuer, l’exploitation d’un outil particulièrement exigeant. Je pense notamment aux personnels de nos centres d’ingénierie et de production nucléaires, qui ont été les premiers affectés par l’accident de Fukushima et mis à forte contribution dans les évaluations qui ont suivi.

Je tiens à remercier tous ceux que j’ai rencontrés, à l’intérieur ou à l’extérieur du Groupe, en France ou à l’étranger, pour la qualité de leur accueil, la franchise de leurs propos et la richesse de nos échanges. Leur ouverture, qui conditionne la pertinence de ce rapport, continue de s’inscrire dans l’esprit de la culture de sûreté.

Merci aussi à mes chargés de mission Christian Thézée, Bruno Coraça, Peter Wakefield qui, cette année encore, n’ont pas ménagé leur aide, sans oublier Jean-Paul Combémorel qui nous a rejoints en début d’année. Une mention toute particulière pour Jacques Dusserre qui vient de quitter l’équipe, après avoir rendu les meilleurs services pendant plus de six ans. Ce rapport garde encore sa marque.

Pour optimiser ce rapport, la forme a été revue par rapport aux années précédentes. Tous les sujets ayant un lien avec la sûreté ont été traités, même si le nombre de chapitres a été réduit. Enfin, bien que ce document n’ait pas vocation d’outil de communication externe, il est comme les années précédentes mis à la disposition du public, en français et en anglais, sur le site Internet d’EDF (www.edf.fr). Cette année, peut-être plus qu’une autre année, je sais qu’il pourra répondre aux interrogations des lecteurs concernant la sûreté nucléaire.

L’Inspecteur général pour la sûreté nucléaire et la radioprotection du groupe EDF

Jean Tandonnet

Paris, le 20 janvier 20121 WANO : World Association of Nuclear Operators.

Page 4: EDFMESS FR - inis.iaea.org
Page 5: EDFMESS FR - inis.iaea.org

3

1

2

3

4

5

6

7

8

9

10

11

12

13

14

Sommaire

1. Mon regard sur la sûreté dans le Groupe 5

2. Le tour d’horizon de l’année 2011 7

2.1. Le management 92.2. L’état des installations 102.3. La maintenance du parc français 102.4. Les relations avec l’Autorité de sûreté nucléaire 122.5. Les projets EPR 142.6. Les acteurs de la santé et de la sécurité 152.7. La radioprotection 172.8. La prévention et la lutte contre l’incendie 182.9. Le renforcement du dispositif de protection des CNPE 192.10. La déconstruction 20

3. Le management de la sûreté 21

4. L’accident de Fukushima Daiichi 29

5. La formation au service des compétences dans les métiers 41

6. Le métier de conduite 47

7. Les ingénieries et les appuis à l’exploitation 53

8. La conduite des projets 57

9. Avec les entreprises prestataires vers une meilleure performance d’ensemble 61

10. Le nucléaire d’EDF hors de France 65

11. Les missions à l’étranger 73

12. Des événements dans le groupe EDF 77

12.1. EDF SA : une usure prématurée de coussinets sur les diesels de secours 7712.2. EDF SA : un incident de pilotage et un arrêt automatique de réacteur 7912.3. EDF Energy : la défaillance de tuyauteries en fonte 8112.4. SOCODEI : un accident industriel à Centraco (Marcoule) 82

13. Un événement hors du groupe EDF : 83

l’accident de la plateforme de forage Deepwater Horizon

14. Annexes 87

14.1. Indicateurs de résultats du parc nucléaire d’EDF SA 8814.2. Indicateurs de résultats du parc nucléaire d’EDF Energy 8914.3. Carte des centrales nucléaires d’EDF SA 9014.4. Carte des centrales nucléaires d’EDF Energy 9114.5. Etapes industrielles des unités nucléaires de production d’EDF SA 9214.6. Etapes industrielles des unités nucléaires de production EDF Energy 9314.7. Table des abréviations 95

Page 6: EDFMESS FR - inis.iaea.org
Page 7: EDFMESS FR - inis.iaea.org

5

1

MON REGARD SUR LA SURETE

DANS LE GROUPE

Je veux, en premier lieu, relever l’absence d’événements majeurs de sûreté cette année dans les centrales nucléaires du groupe EDF, avec toutefois en France un événement de niveau 2 sur l’échelle INES2, portant sur une défaillance matérielle et sans conséquence directe sur la sûreté (cf. chapitre 12) ainsi que des événements marquants en France et au Royaume-Uni, sans conséquences directes non plus, mais qui doivent être considérés comme autant d’alertes et traités avec rigueur par les managers du nucléaire.Je tiens aussi à relever, en France et au Royaume-Uni, les performances globalement bonnes sur les principaux indicateurs de sûreté, dont l’incendie, avec à titre d’illustration, le très faible niveau d’arrêts automatiques réacteurs (AAR) en France, meilleure performance réalisée depuis l’origine par le parc et meilleur niveau de l’inter-comparaison internationale.Ces résultats, conjugués à l’amélioration des résultats de production, sont de nature à donner sérénité et confiance à tous les acteurs des sites et vont dans le bon sens pour la sûreté.

Je considère en outre qu’EDF SA dispose d’une filière indépendante de sûreté à la fois solide, professionnelle et qui donne confiance dans le niveau de transparence dont l’exploitant doit faire preuve. La volonté de la direction d’EDF Energy de se doter du même type d’organisation et de compétences est également un point très positif.

Enfin, la réactivité dont ont su faire preuve les entités et les directions concernées dans le Groupe après la catastrophe de Fukushima, tant aux niveaux nationaux respectifs que dans les organisations internationales, mérite aussi d’être saluée, signant là une attitude d’exploitant responsable.

Pour la radioprotection, les résultats 2011 sont globalement satisfaisants dans le Groupe. J’attire l’attention sur les effets induits en France par les importants chantiers de maintenance et de modernisation à venir, qui nécessiteront une approche encore plus volontariste.

Pour la sécurité au travail, je dois souligner que nos résultats en France sont encore loin des meilleurs résultats identifiés au niveau mondial par WANO et appellent une action managériale vigoureuse. Je considère en effet que l’action des managers sur ce terrain, qui est celui de la rigueur en exploitation, est gage de succès durables pour la sûreté nucléaire.

En même temps que mon appréciation des performances, je formule une alerte et un point de vigilance, concernant le parc français.

L’alerte porte toujours sur le renouvellement des compétences des salariés de l’entreprise et de ceux des entreprises prestataires travaillant à leurs côtés : il faut à la fois réussir la relève de génération pendant la marche des 58 réacteurs, répondre à des besoins massifs d’ingénieurs pour l’extension de la durée de fonctionnement des réacteurs à 60 ans et engager des actions après Fukushima. Ces défis conduisent l’entreprise à prolonger son effort considérable et continu. J’attire l’attention, dans un tel contexte, sur le besoin impératif de priorisation de l’activité, pour ne pas fragiliser la sûreté en exploitation.

2 INES : International Nuclear Events Scale.

Page 8: EDFMESS FR - inis.iaea.org

6

1

Tout en me félicitant des premiers actes concrets pour responsabiliser davantage les unités opérationnelles et leurs équipes de direction, en leur assurant en particulier un meilleur appui du niveau national, je considère qu’il reste encore beaucoup à faire dans cette voie prometteuse et porteuse de sérénité sur le terrain et qu’il convient d’être vigilant.

Pour l’avenir du nucléaire, je veux aussi insister, au-delà des difficultés du chantier ou des projets en cours, sur l’avancée sans égale pour la sûreté que représente le réacteur EPR.

Au-delà de ces points essentiels auxquels je veux donner un relief particulier, je dresse dans la suite de mon rapport, dans mon tour d’horizon de l’année, un bilan plus large de l’ensemble des domaines de l’exploitation et de l’ingénierie. J’aborde dans les chapitres suivants, plus en profondeur, des thématiques à fort enjeu et des événements techniques ayant retenu mon attention.

Page 9: EDFMESS FR - inis.iaea.org

7

2

LE TOUR D’HORIZON

DE L’ANNEE 2011

L’année 2011 a été marquée par les suites du séisme suivi de tsunamis au Japon le 11 mars, plus particulièrement par les conséquences sur la centrale de Fukushima Daiichi. Les destructions ont conduit à la perte de sources froides et d’alimentations électriques sur les six tranches du site. Cet accident souligne que l’exploitant est le premier responsable de la sûreté d’une centrale nucléaire. Il pointe, une fois encore, que le nucléaire est une école d’humilité et surtout d’exigence. Il rappelle également l’interdépendance de l’industrie nucléaire qui subit, dans le monde entier, les conséquences d’un accident local. Compte tenu de l’importance de cet accident, le rapport traite de ses conséquences au fil des différents paragraphes et dans un chapitre particulier. Je tiens à exprimer toute ma compassion aux familles des victimes du tsunami et aux personnes déplacées après l’accident de Fukushima, dont les conditions de vie sont encore profondément bouleversées.

Suite à cet accident, des stress-tests ou tests de résistance ont été demandés dès le 23 mars par le Premier ministre français puis par la Commission européenne. Je note que le groupe EDF a présenté des plans d’action et a rendu, dans les mois qui ont suivi, des évaluations conformément aux exigences françaises et britanniques.Cet accident a marqué les agents des sites de production nucléaire. Aux jeunes générations qui n’ont pas vécu les accidents de TMI3 et de Tchernobyl, il rappelle la spécificité de l’industrie nucléaire et l’importance d’une bonne culture de sûreté, aussi bien à la conception qu’en exploitation.

En France

Je note que, pour EDF, le premier défi a été de tirer le retour d’expérience de Fukushima, mais ce n’était pas le seul. L'ambition d'EDF est de continuer à améliorer la sûreté et la disponibilité de ses centrales nucléaires et de réunir les conditions nécessaires pour assurer le fonctionnement de ses réacteurs pour une durée de 60 ans. Ainsi, comme prévu, l’année a été marquée par le doublement du nombre de visites décennales par rapport à 2010, auxquelles s’ajoutent des opérations lourdes de maintenance, comme le remplacement de générateurs de vapeur (RGV). Je note avec satisfaction que ces investissements importants seront poursuivis dans les prochaines années pour tenir compte de la mi-vie du parc français et des nouvelles exigences de sûreté et de sécurité.En 2011, l’Autorité de sûreté a donné un avis positif pour la poursuite de l’exploitation de la tranche 1 de la centrale de Fessenheim, âgée de 30 ans, pour une période de dix ans, conformément à la loi TSN4 et sous réserve de la réalisation de travaux complémentaires. Après la tranche 1 de Tricastin en 2010, elle est la deuxième tranche à franchir cette étape avec succès.

3 TMI : Three Mile Island (Etats-Unis).4 Loi TSN : Loi sur la Transparence et la Sécurité Nucléaire de 2006.

Page 10: EDFMESS FR - inis.iaea.org

8

2

Le programme de rénovation du parc nucléaire va s'amplifier à partir de 2015 avec les importants travaux des troisièmes visites décennales des tranches du palier 1300 MWe. Je suivrai avec attention en 2012 la préparation de ce Grand carénage5, en particulier sur le volet des compétences. Lors de mes visites dans les CNPE, j'ai mesuré la mobilisation pour se préparer à ce challenge, mais je constate dans le même temps des facteurs d’inquiétudes fortes sur le terrain, suite aux débats récents sur la fermeture anticipée d’un certain nombre de tranches. La sérénité s’en ressent, ce qui me préoccupe pour son impact éventuel sur la sûreté.

Au Royaume-Uni

Je note que le Royaume-Uni poursuit la relance du nucléaire. En juillet, le Parlement a voté la Déclaration nationale de politique nucléaire (National Policy Statement for Nuclear). Ce vote ratifie le programme de construction de centrales nucléaires, dont Hinkley Point C et Sizewell C, ce qui devrait faire d’EDF Energy le premier opérateur à participer au renouveau du nucléaire au Royaume-Uni. Le dossier du Nuclear Site Licence a été déposé auprès de l’Autorité de sûreté en vue d’une approbation mi-2012.Concernant l’accident de Fukushima, j’observe dans la société britannique une sérénité et un flegme qui contrastent avec les réactions d’autres pays d’Europe.

Aux Etats-Unis

J’ai conduit une inspection à la centrale de Ginna, en exploitation depuis 43 ans. Elle est opérée par CENG6, compagnie dans laquelle EDF possède des intérêts. Cela rappelle que le parc français d’EDF, avec une moyenne d'âge proche de 30 ans, est relativement jeune. Je signale que des réacteurs de même conception REP ont déjà reçu, aux Etats-Unis, une autorisation d’exploitation pour 60 ans.J’observe que CENG a confié à EDF la présidence du Comité Sûreté Exploitation de son conseil d’administration. Ce comité a notamment croisé et enrichi les enseignements tirés en France et aux Etats-Unis après Fukushima.J’ai noté cette année que la compagnie EXELON s’est portée acquéreur de Constellation, elle-même actionnaire majoritaire (50,1 %) aux côtés d’EDF (49,9 %), de CENG. Je suivrai avec intérêt l’évolution de cette situation nouvelle.

En Chine

J’ai une nouvelle fois conduit une visite du chantier des deux réacteurs de Taishan et rencontré des acteurs de ce projet. La pose du dôme métallique sur le premier réacteur a été une étape symbolique et je relève l’efficacité des équipes du bureau d’études commun AREVA-CGNPC7 pour rattraper le retard dans la production des plans de montage. Après l’accident de Fukushima, j’approuve que les évolutions des réacteurs reprennent les propositions françaises élaborées pour l’EPR de Flamanville 3, en ajoutant quelques spécificités liées au site.

L’importance des agressions naturelles

A la suite de l’accident de Fukushima, où un phénomène naturel a eu des conséquences désastreuses, je reste particulièrement attentif aux conséquences d’événements naturels sur la sûreté des parcs nucléaires.Dans le groupe EDF, la centrale de Torness au Royaume-Uni a subi en 2011 un début de colmatage de sa prise d’eau, liée à la présence de méduses en quantité exceptionnelle. Dans le Maryland aux

5 Grand carénage : prévu à partir de 2015, il consiste d’une part, lors des visites décennales, à remplacer les composants clés dans la perspective du fonctionnement des centrales jusqu’à 60 ans, d’autre part, dans le cadre des réévaluations périodiques de sûreté, à apporter des améliorations pour renforcer la prévention et la robustesse face aux accidents graves et aux agressions externes (séisme, inondation, etc.).

6 CENG : Constellation Energy Nuclear Group.7 CGNPC : China Guangdong Nuclear Power Company.

Page 11: EDFMESS FR - inis.iaea.org

9

2

Etat-Unis, la centrale de Calvert Cliffs a ressenti un séisme significatif, suivi 4 jours plus tard d’un ouragan qui a momentanément arrêté la production d’une tranche. Sur ces deux sites, je note qu’il n’y a pas eu de conséquences sur la sûreté, compte tenu de la conception robuste des installations. En France, face à la sécheresse sévère et précoce du premier semestre, EDF a dû prendre des dispositions pour éviter des conséquences sur la production du parc. Le retour d’expérience de la sécheresse et de la canicule de 2003 et des conditions climatiques moins sévères que prévu ont permis un été sans difficulté particulière.

En dehors du groupe EDF, la centrale américaine de North Anna a subi, le 23 août 2011 un séisme au-delà du dimensionnement de conception. Après inspection et requalification des deux réacteurs, la NRC8 a autorisé leur redémarrage en novembre. De même, des inondations ont affecté en juin les centrales de Fort Calhoun et de Cooper : bien qu’ils aient été isolés pendant plusieurs semaines, les deux sites n’ont subi aucun dommage grâce à des mesures bien anticipées.

Un motif de satisfaction

Début 2011, les 58 réacteurs français et les 15 réacteurs britanniques étaient tous en production. Dans l’histoire du parc français, c’est la deuxième fois que cette situation se produit. Au-delà du côté anecdotique, je note la baisse du nombre d’indisponibilités fortuites en 2011 ce qui, d’une certaine manière, est une bonne chose pour la sûreté et reflète la pertinence des investissements importants de rénovation depuis 2007.

2.1 Le management

Je complète ce tour d’horizon par un regard sur quatre orientations ayant retenu mon attention.

Le projet Génération 2020

Il exprime la vision de la direction de la DPN9 pour l’avenir et constitue un élément essentiel de la démarche visant à motiver la nouvelle génération de personnels. Les premiers pas du projet montrent déjà des effets bénéfiques. L’attention portée aux problèmes du terrain a amené une plus grande sérénité des équipes dirigeantes des sites. J’attends également des résultats du projet DIN10-DPN, qui doit améliorer les synergies entre ces deux divisions et permettre notamment de disposer d’une ingénierie du parc en exploitation (IPE), répondant mieux aux besoins de l’exploitant. Je reste cependant très attentif au risque de surcharge de l'ingénierie où les projets techniques sont de plus en plus nombreux.

Les directeurs délégués opérations (DDO)

Prioriser les activités en fonction des ressources disponibles demeure, plus que jamais, incontournable pour les directions, tant de l'exploitation que de l'ingénierie. Le risque actuel est probablement de voir des renoncements faits au niveau du terrain sans dépendre de choix managériaux. C’est dans le besoin que la nécessité de gestion nationale se fait sentir, sans tomber dans une centralisation excessive qui faisait l’objet d’une alerte de ma part les années précédentes. Pour la DPN, les DDO ont un rôle d’arbitrage essentiel pour permettre cette adaptation. J'invite également les équipes de direction à réaffirmer deux principes de sûreté qui prennent de la valeur dans ces périodes d’activités intenses et de changements : l'attitude interrogative et le devoir d'alerte.

8 NRC : Nuclear Regulatory Commission (Etats-Unis).9 DPN : Division Production Nucléaire d’EDF SA.10 DIN : Division Ingénierie Nucléaire d’EDF SA.

Page 12: EDFMESS FR - inis.iaea.org

10

2

Une recherche de priorisation et d’efficacité

J'observe que la mise en place du système de management intégré se généralise à l’initiative des sites, sans prescription particulière du niveau national. Comme l'an passé, je constate que ce système de management favorise la synergie entre les projets, clarifie le rôle des acteurs et permet aussi d’identifier et de corriger les lacunes des organisations. Cependant, j’attire de nouveau l’attention, dans l’environnement complexe du nucléaire, sur la nécessité de conserver cet outil au service de la performance.Le système de priorisation apparaît toujours insuffisamment visible pour les managers de première ligne et les préparateurs, qui éprouvent encore des difficultés pour donner à leurs équipes le sens des choix. Dans les CNPE, les managers et les préparateurs me semblent les populations les plus contraintes.

Une politique sûreté pour le Groupe

Je me félicite que la rédaction de la politique sûreté du Groupe ait enfin été engagée. J’attire l’attention sur l’importance à accorder à son déploiement auprès de l’ensemble du personnel concerné.

2.2 L’état des installations

En France, l’état des installations s’améliore après des investissements importants. Cet effort doit être maintenu pour tenir compte d’un parc à mi-vie et d’un retard initial par rapport aux standards internationaux.

L’état des installations de production nucléaire, est en voie d’atteindre le niveau attendu dans la plupart des sites. Plus que le changement visuel, c’est le changement culturel que je retiens. Le retour d'expérience du pilotage du projet O2EI11 est

à tirer pour pérenniser ces acquis sur tous les sites. Je souhaite que les équipes de direction conservent le niveau d'engagement qui a fait le succès de ce projet. Une attention spécifique est à porter aux sites de bord de mer où la corrosion saline rend le maintien au niveau plus délicat. J’insiste pour que la tenue du chantier EPR de Flamanville fasse l’objet d’une attention constante de la part de la DIN et que l’exploitation démarre à un niveau de house-keeping exemplaire. Cela ne me semble pas acquis aujourd’hui si cette priorité n’est pas réaffirmée et partagée.

En complément du projet O2EI dédié aux installations industrielles, je salue l'effort d’adaptation et de remise à niveau des bâtiments tertiaires avec le projet PARTNER12, engagé sur la décennie à venir. Ce projet répond à une attente forte des salariés, d'EDF et des prestataires. L'amélioration de la qualité de vie au travail que l’on peut en attendre est une avancée supplémentaire pour la sûreté.

2.3 La maintenance du parc français

En France, les dix prochaines années sont déterminantes pour le devenir du parc. De nombreux défis sont à relever pour préserver le patrimoine, préparer la prolongation de sa durée de vie, augmenter sa disponibilité, tout en continuant à faire progresser son niveau de sûreté. Pour réussir ce challenge, avec l’ensemble des prestataires, la DPN a lancé son projet Génération 2020 déjà évoqué. Je partage en particulier les orientations du projet visant à responsabiliser le personnel sur l’état et l’entretien de sa machine ainsi que sur ses méthodes de travail.

11 O2EI : Obtenir un Etat Exemplaire des Installations (projet de la DPN).12 PARTNER : PARc Tertiaire Nucléaire Eco-Rénové (projet Direction Immobilier Groupe).

Page 13: EDFMESS FR - inis.iaea.org

11

2

Je voudrais insister sur la place de la maintenance, aux côtés de la conduite, dans ce projet. Avec un parc à mi-vie, la maintenance est devenue un domaine stratégique pour préserver le niveau de sûreté des tranches et le patrimoine de l’entreprise. Dans un passé encore récent, des défaillances de matériels ont affecté la performance des tranches et conduit à s’interroger sur les stratégies de maintenance et l’organisation. Je constate que les choses ont évolué et que la maintenance est partie prenante des changements en cours.

Lors de mes visites, j’ai noté une réelle volonté des équipes locales, en particulier à la maintenance, de reprise en main du fonctionnement et de l’entretien de leurs matériels. Mais dans le contexte actuel de renouvellement des compétences et de déploiement de nombreux projets impactant fortement les équipes de maintenance, en particulier les préparateurs, ce revirement est difficile à conduire. Partout je rencontre des équipes de préparateurs sur-sollicités. Les organisations sont en tension face au volume de travail à réaliser, au manque de priorisation et à la faiblesse des ressources. Le programme Grand carénage, de rénovation lourde, qui se met en place devra en

tenir compte. Je ne manquerai pas de suivre son déroulement dans mes prochaines visites.

Je voudrais, cette année encore, dire combien les organisations du travail sont pénalisantes et peu adaptées au surcroît d’activité qui attend l’entreprise. Je note toutefois que certains CNPE ont commencé à évoluer mais les contraintes restent nombreuses.Noyau dur des futurs grands carénages, la préparation des arrêts de tranche n’est pas encore ni partout à la hauteur des niveaux visés. Elle se met en place encore trop tardivement et ne dispose pas d’assez de moyens, suffisamment tôt en amont du début de l’arrêt. J’observe que les activités répétitives d’un arrêt sur l’autre sont encore trop peu industrialisées.

Certaines difficultés, de mon point de vue surmontables et que je pensais résolues, sont devenues récurrentes et compliquent la tâche des équipes de maintenance. Où que j’aille, on me fait part de problèmes d’obsolescence de matériels et de difficultés pour obtenir des pièces de rechange.Pour les intervenants rencontrés, pouvoir disposer à temps des pièces de rechange nécessaires est encore trop souvent un problème. Même si le projet AMELIE13 commence à porter ses fruits par la professionnalisation de toute la chaîne d’approvisionnement, de nombreux points restent à traiter. A cette difficulté de constitution de stocks et de logistique s’ajoutent des problèmes d’obsolescence de composants ou de matériels. Je suis frappé de voir encore dans nos salles de commande des enregistreurs papier à l’heure du numérique. Ce sujet ne me semble pas assez anticipé ni dimensionné, il mériterait un « changement de braquet ». Dans un environnement industriel en évolution permanente, avec un parc standardisé, l’arrêt de fabrication d’un composant peut devenir critique et conduire à des événements comme celui de cette année sur les diesels (cf. chapitre 12).

La forte proportion de jeunes embauchés dans les équipes (1 pour 1 dans les services maintenance en 2012) doit inciter les managers à rester vigilants sur l’accompagnement des nouveaux arrivants et sur les fondamentaux du métier et de la sûreté. Dans de telles situations, les pratiques de performance humaine (PPH) sont des lignes de défense solides. Il convient de les faire appliquer systématiquement.

Cette année encore, je me suis rendu dans une Agence de maintenance thermique (AMT). Les AMT assurent, sur les tranches nucléaires et conventionnelles, des activités de maintenance spécialisées et à forte valeur ajoutée. Elles ont toujours démontré leurs compétences face aux constructeurs. Elles ne peuvent, par leur mission, envisager la sous-traitance, ce qui impose aux managers d’être très impliqués sur le maintien des compétences dans chaque métier. C’est un thème que je développe au chapitre 5 consacré à la formation. Capables dans leurs domaines de remplacer un fournisseur défaillant, ces entités sont des assurances pour la DPN. Aujourd’hui les AMT sont des entités à taille humaine, tournées à 100 % sur la technique, constituées d’équipes très motivées. 13 AMELIE : Projet visant à transformer la logistique des pièces de rechange.

Page 14: EDFMESS FR - inis.iaea.org

12

2

Je pense que la motivation vient en grande partie de l’attrait des postes. La technique, la qualité, l’autonomie et la variété en sont les principales composantes.

Enfin sous l’angle des résultats, je note cette année un nombre élevé de non-qualités de maintenance. Devant l’augmentation prochaine du volume des travaux, les services maintenance doivent retrouver le niveau de qualité que l’on est en droit d’attendre sur les sites nucléaires et faire face à la charge de travail qui les attend. En 2012, j’y attacherai une importance particulière lors de mes visites.

2.4 Les relations avec l’Autorité de sûreté nucléaire

Pour la pérennité du nucléaire et son acceptabilité, il est essentiel d’avoir une autorité de sûreté forte. L’équilibre des rôles reste essentiel entre l’exploitant et l’autorité de sûreté, et en France son appui technique l’IRSN14. Je mesure tout l’intérêt de discussions techniques de haut niveau, l’IRSN apportant sa très bonne connaissance des centrales d’EDF, mais la formulation des exigences doit être portée par l’Autorité de sûreté

nucléaire (ASN). Face aux nombreuses demandes de l’ASN, j’attire l’attention sur l’importance de définir des priorités pour ne pas disperser les forces d’ingénierie.

L'accident de Fukushima

Cette année, les relations avec les autorités de sûreté ont été marquées par l’accident de Fukushima.

Partout dans le monde, l’exploitant porte la responsabilité de la sûreté nucléaire, l’accident de Fukushima l’a rappelé. J’ai apprécié la réactivité du groupe EDF dans son ensemble, pour réaliser un premier état des lieux de ses propres installations, et répondre ensuite aux diverses sollicitations. En France, les évaluations complémentaires de sûreté (ECS) demandées par l’ASN ont mobilisé une énergie considérable de toutes parts. Elles ont été remises dans les temps. La confiance est restée une valeur partagée et, à mon avis, EDF fait preuve d’une transparence exemplaire. Cette transparence, attendue par nos concitoyens, a été relevée par le Haut Comité pour la transparence et l’information sur la sécurité nucléaire (HCTISN). Je note en particulier l’ouverture de CNPE aux membres du HCTISN et des Commissions locales d’information (CLI), lors d’inspections conduites dans le cadre des ECS.Au Royaume-Uni, je note le travail similaire réalisé par EDF Energy en concertation avec l’ONR15. J’apprécie la cohérence des démarches conduites par les autorités de sûreté française et britannique pour la définition des stress-tests en Europe. La mission d’expertise de l’AIEA16 au Japon, conduite par Mike Weightman, inspecteur en chef de l’ONR, a abouti à un compte-rendu détaillé des événements, utile aux exploitants, après le premier rapport du gouvernement japonais présenté à l’AIEA.Aux Etats-Unis, j’observe le rôle moteur des industriels, au sein du NEI17 et de l’INPO18, en particulier l’engagement des dirigeants de CENG.

Les relations sur le terrain

Au cours de mes inspections sur les sites de production nucléaire et dans les centres d’ingénierie, j’ai ressenti, plus que les années précédentes, des difficultés dans les relations et le traitement des affaires entre les représentants de l’ASN et les équipes d’EDF. Les contraintes industrielles qui pèsent sur l’exploitant et le volume de travail des deux parties en cette période chargée compliquent singulièrement ces rapports. Je suis convaincu que le formalisme ne doit pas prendre le pas sur le fond.

14 IRSN : Institut de Radioprotection et de Sûreté Nucléaire.15 ONR : Office of Nuclear Regulation.16 AIEA : Agence Internationale de l’Energie Atomique.17 NEI : Nuclear Energy Institute.18 INPO : Institute of Nuclear Power Operations.

Page 15: EDFMESS FR - inis.iaea.org

13

2

Les autorisations internes

Je note que le traitement des modifications génériques - article 26 de la loi TSN - n’a pas réellement avancé, ce serait pourtant une source de simplification et d’efficacité. Je regrette également que le champ des autorisations internes, au titre de l’article 27 de la loi TSN n’ait toujours pas été élargi à EDF. Il faut, à mon sens, faire progresser cette possibilité qui est certes exigeante, mais qui présente aussi des aspects vertueux, en clarifiant et en simplifiant les relations entre l’exploitant et l’ASN. Je note la volonté d’EDF d’aller dans ce sens.Dans le domaine des combustibles, des dossiers sont en cours d’instruction sur le périmètre et les modalités d’exécution, en particulier pour dégager plus de souplesse en exploitation. Ils n’ont pas encore abouti et je crains qu’ils ne soient plus une priorité après Fukushima.Dans le domaine de la déconstruction, des autorisations internes efficaces sont mises en place depuis plusieurs années, de manière responsable.

Des points d’attention complémentaires

Je rejoins l’Autorité de sûreté qui fixe deux domaines sensibles pour le parc nucléaire : la sécurité au travail et l’environnement. Sur le premier sujet (traité au chapitre 2.6), j’ai été alerté de difficultés par des CNPE dans leurs relations avec des inspecteurs du travail. Malgré des progrès significatifs, l’environnement reste un sujet sensible. Je partage la perception de l’ASN qui pense que des progrès sont encore possibles pour les rejets d’effluents chimiques, à l’échelle des réalisations de ces dernières années pour les rejets d’effluents radioactifs liquides. L’interface chimie-conduite-maintenance doit pouvoir être renforcée.Je rappelle toutefois la différence potentielle d’échelle des risques entre, d’une part, la sûreté nucléaire, et d’autre part, la sécurité au travail ou la préservation de l’environnement dans l’exploitation quotidienne. Les conséquences pour nos concitoyens sont d’un autre ordre de grandeur. Il reste néanmoins qu’une sûreté nucléaire exemplaire se traduit dans les métiers par une sécurité et une protection de l’environnement à la même hauteur.

Les services d’inspection reconnus (SIR)

Cette année pour la première fois, j’ai rencontré des inspecteurs et des responsables EDF de SIR dans les CNPE. Portant des enjeux de sécurité des personnes, ils contrôlent les appareils à pression, par délégation de l’autorité administrative, pour le compte du directeur de site. Ils sont directement rattachés au directeur d’unité pour garantir leur indépendance, leurs préconisations pouvant nécessiter des arbitrages de haut niveau sur le site.Je recommande d’être particulièrement vigilant sur la gestion des compétences et l’indépendance de ce service. En effet, le gréement en compétences des SIR reste difficile, avec un temps de formation long, et nécessite de disposer d’une pépinière pour faciliter la relève. L’animation nationale de ce métier devrait s’attacher davantage à recueillir des informations de terrain sur la gestion et la résolution de difficultés techniques.

Les relations entre les autorités de sûreté

Pour un groupe international comme EDF, les relations entre les différentes autorités de sûreté sont essentielles et le seront encore plus demain avec l’approbation des dossiers de sûreté des EPR en Chine et au Royaume-Uni. Je suis surpris que les autorités nationales, responsables de la définition des objectifs de sûreté et du contrôle du respect de ces objectifs, n’aient pas une coopération plus active. Cette difficulté doit pouvoir trouver une solution au moment où l’accident de Fukushima a conduit les autorités de sûreté à coopérer plus étroitement. Je note que l’ASN suit cet objectif en souhaitant une harmonisation qui pourrait s’appuyer sur ce qui existe déjà dans le cadre de WENRA19 et de l’INRA.20

19 WENRA : Western European Nuclear Regulator’s Association.20 INRA : International Nuclear Regulator’s Association.

Page 16: EDFMESS FR - inis.iaea.org

14

2

2.5 Les projets EPR

Réacteur de troisième génération, l’EPR constitue par son design une avancée significative pour la sûreté. Quatre EPR sont en cours de construction dans le monde et deux en phase d’études au Royaume-Uni : tous ces projets sont suivis avec attention en France et à l’étranger. Ils sont complexes à conduire et exigeants en matière de compétences. Durant le chantier comme cette année, ou plus tard durant l’exploitation, le bon déroulement et la réussite de Flamanville 3 sont essentiels pour les autres projets, tant leur interdépendance est grande.

L’ampleur du projet de Flamanville 3 et les contraintes de son pilotage ont conduit à plusieurs reprises à des difficultés de réalisation sur le chantier et d’approvisionnement de matériels. Je relève l’annonce faite par EDF d’un décalage à 2016 du couplage de Flamanville 3. Un important travail de refondation du pilotage du projet, initié dès 2010, a redéfini clairement, comme je le notais l’an passé, les responsabilités entre les acteurs, en particulier au niveau des études. Des moyens nouveaux pour disposer d’une planification détaillée des activités ont été mis en place. A ce propos, je me félicite que les dix principaux fournisseurs du projet, réunis dans le cadre d’une instance de coordination nommée F10, soient associés à l’élaboration du planning du projet. Je les encourage à poursuivre et à accentuer cette collaboration. Cet important travail de planification et de coordination éclaire aujourd’hui le management du projet sur les risques, le chemin critique du planning et les enjeux.

Devant l’importance pour les réacteurs EPR en construction de la réussite du démarrage de Flamanville 3 conformément au nouveau planning, je m’interroge sur la nécessité de maintenir certains objectifs d’exploitation spécifiques au parc français, tels que le suivi de charge ou la manœuvrabilité de la tranche, car ceux-ci demandent de nombreuses études de sûreté spécifiques au réacteur de Flamanville 3.

A contrario, je signale deux démarches importantes du projet EPR au regard de la sûreté. D’abord l’objectif retenu par EDF et son partenaire chinois de qualifier les matériels EPR aux conséquences des accidents graves avant le démarrage de la tranche, malgré un surcroît important de travail et de contraintes pour la conduite du projet. Ensuite, dans le domaine des facteurs humains, la qualification en réel sur simulateurs avant le démarrage de l’interface homme-machine de la salle de commande informatisée. Du point de vue de la sûreté, ces deux démarches me semblent exemplaires.

Cette année, je me suis rendu sur le chantier EPR de Taishan en Chine. La dépendance de ce chantier et de celui de Flamanville 3 est frappante même si certaines séquences de construction, en particulier sur le génie civil, sont différentes. Des présentations qui m’ont été faites, je retiens que le retour d’expérience très récent de notre partenaire chinois, sur la mise en service des tranches de Ling Ao, l’a conduit à gréer très tôt les équipes d’essais pour la préparation des procédures et des essais de mise en service de la tranche 1 de Taishan. Un rapprochement des équipes d’essais de Flamanville 3 avec leurs homologues chinois sur ce sujet me semblerait une démarche intéressante et très certainement profitable.

Le chantier de Flamanville 3 a vécu cette année des évènements graves avec deux accidents mortels. La sécurité au travail, tout comme la qualité des travaux, fait l’objet d’une attention particulière du management et sa volonté de redresser la barre est d’autant plus forte que commence la période des montages mécaniques et électriques, alors que la phase de génie civil n’est pas totalement terminée. Néanmoins, lors de ma visite sur le chantier, l’état de l’installation n’était pas encore aux standards affichés. Je m’interroge sur les leviers dont dispose la direction locale vis-à-vis des services études, du projet et des entreprises sous-traitantes pour faire évoluer les choses et garantir la qualité et la sécurité durant les montages. J’attire aussi l’attention, pour ce site de bord de mer, sur les risques de corrosion, la conservation des équipements et la protection des locaux, qui ne me semblent pas encore bénéficier de toute la priorité souhaitée.

Page 17: EDFMESS FR - inis.iaea.org

15

2

L’exploitant déjà sur place, et quasi au complet depuis presque deux ans, continue sa préparation mais le nouveau planning suscite des interrogations dans les équipes, voire des inquiétudes. Certaines formations sur simulateur seront à reprendre et les premiers départs de personnels interviendront avant le démarrage de la tranche, ce dont je m’étonne. De plus, ce planning ne donne pas encore aujourd’hui assez de visibilité sur le contenu des phases préparatoires à l’exploitation.Je reste également surpris en France par la conduite d’un tel projet dans deux divisions séparées. Tout au long du projet, les équipes construction de la DIN et les équipes d’exploitation de la DPN réfèrent à des lignes managériales différentes. Cela conduit, dans une période où les ressources sont rares, à une utilisation non optimisée de celles-ci. Je note que, sur les projets EPR en Chine et au Royaume-Uni, ces équipes sont regroupées depuis le début du projet : leur pluridisciplinarité est un facteur supplémentaire de qualité, d’efficacité et de sûreté. Dans la structure actuelle du chantier, j’encourage les équipes de Flamanville 3 à coopérer davantage durant les phases de montage et de pré-exploitation.

L’EPR de Flamanville 3 constitue une vitrine pour le Groupe et je m’étonne d’apprendre qu’il ne disposera peut-être pas pour son démarrage des derniers standards du parc sur l’incendie et sur l’état des installations (O2EI).

La réussite de Flamanville 3 est une conquête de chaque jour et conditionne la réussite des autres EPR. L’expérience développée en France, en Chine et en Finlande est un atout que le groupe EDF doit savoir utiliser.

2.6 Les acteurs de la santé et de la sécurité

Cette année, j’ai continué à rencontrer, en France et au Royaume-Uni, l’ensemble des structures médicales, notamment les médecins et les services de prévention des risques. Ils sont des acteurs incontournables de la santé, mais aussi de la maîtrise des risques psychosociaux, de la qualité de vie et de la sécurité au travail.La santé des salariés et des prestataires reste une des principales préoccupations du groupe EDF. J’insiste pour que la sécurité, qui ne peut être dissociée de la sûreté, fasse l’objet de la même priorité. L’approche internationale et celle d’EDF Energy conduisent EDF SA à considérer, aujourd’hui plus qu’hier, la sécurité au travail comme un indicateur pertinent du comportement rigoureux en exploitation, donc de la culture de sûreté.

Les équipes médicales et leur environnement

Les relations entre les services médicaux et les équipes de direction des CNPE se sont généralement bien améliorées ces dernières années, ce qui est essentiel. J’ai rencontré, à peu près sur tous les sites, des médecins ouverts au travail avec les équipes de direction. Ce partage de cultures différentes, à l’intersection de la santé, de la sécurité et de la radioprotection, enrichit les analyses de risques et les décisions qui en découlent. Il constitue un levier de progrès pour le management et les acteurs de la santé au travail. J’invite les équipes de direction à renforcer encore les liens avec les médecins.

Dans le domaine de la radioprotection, ces relations permettent :aux équipes médicales de contribuer activement à la prévention et d’exercer une surveillance médicale pertinente, adaptée aux niveaux de risques réellement encourus,aux équipes dédiées à la radioprotection d’avoir une vision enrichie dans les domaines humains, organisationnels et techniques.

Je suis attentif à la charge de travail des équipes médicales qui assurent le suivi du personnel des CNPE, mais aussi le suivi DATR21 de prestataires de plus en plus nombreux. Ce suivi, que les services médicaux interentreprises ne sont pas tous habilités à assurer, concerne environ 3 500 travailleurs prestataires concentrés sur certains sites. Je rappelle que les équipes médicales devront pouvoir assurer cette charge

21 DATR : Directement Apte à Travailler sous Rayonnement.

Page 18: EDFMESS FR - inis.iaea.org

16

2

qui augmentera dans les prochaines années avec l’accroissement des activités de maintenance et les difficultés de recrutement des services médicaux interentreprises. Je note cependant que l’arrivée de nouveaux portiques de détection C222 devrait alléger les contrôles anthropogamma-métriques.

La qualité de vie au travail et les risques psychosociaux

Suite à l’accord « Prévenir les risques psychosociaux et améliorer la qualité de vie au travail des salariés d’EDF SA », signé le 10 novembre 2010 par trois organisations syndicales, j’observe que les démarches de prévention et d’amélioration de la qualité de vie et de la santé au travail, pour certaines déjà prévues par un accord préexistant à la DPN, se poursuivent. Le bien-être au travail est un facteur favorable à la sûreté et il est indispensable de développer encore ces actions, en y impliquant davantage les managers.

L’alcool et les produits stupéfiants

J’ai déjà souligné l’importance des mesures pour interdire le travail sous l’emprise de l’alcool et des stupéfiants. En France, les sites sont sensibilisés mais la prescription de la direction de la DPN, avec les contrôles inopinés d’alcoolémie, n’est pas appliquée partout avec la même vigueur. Dans le domaine des stupéfiants, la détection est laissée à la seule appréciation du médecin, ce qui ne me paraît pas suffisant.Ainsi au Royaume-Uni, EDF Energy s'est fixé l'objectif de contrôler chaque année (alcool et stupéfiants) 20 % du personnel, y compris les prestataires. Les directeurs de site sont informés des résultats, ce qui peut conduire à un suivi médical, voire à la suppression d’habilitations en cas d’infractions. Les résultats montrent une augmentation, d’une année sur l’autre, des infractions. Cette tendance, observée dans un pays comparable, doit nous faire réfléchir sur la politique suivie en France. Les Etats-Unis appliquent les mêmes principes que le Royaume-Uni.

La sécurité au travail en France

Des efforts constants de prévention des risques ont réduit régulièrement le taux de fréquence des accidents du travail. En France, bien que divisé par un facteur supérieur à 2 pendant ces dix dernières années, ce taux (3,8 par million d’heures travaillées, EDF et prestataires) est bien supérieur au taux du reste de l’industrie nucléaire. En 2011, ces résultats stagnent et sont, de plus, marqués par trois accidents mortels : deux sur le chantier de Flamanville et un, sur le site de Gravelines. Je note que la DPI23 a conduit une mission d’évaluation de la sécurité.

Toutefois, les résultats interrogent sur le niveau de maîtrise et de prévention des risques. L’implication des équipes de direction au plus haut niveau est encore insuffisante. Je les engage à afficher très clairement cette priorité sur le terrain. Les entreprises prestataires doivent être impliquées dans cette démarche.Je suis certain que l’extension des champs des évaluations globales d’excellence (EGE) par l’inspection nucléaire sur la sécurité contribuera à de meilleurs résultats. Sur les sites, je considère que le métier de préventeur au sein des services de prévention des risques doit être renforcé.

22 C2 : Portique de détection en sortie de zone contrôlée.23 DPI : Direction Production Ingénierie.

Page 19: EDFMESS FR - inis.iaea.org

17

2

British Energy24 depuis 2005, un exemple à suivre

Bien que partant d’un taux de fréquence de 1,825, le doublement des accidents du travail début 2005 a conduit l’entreprise à réagir fortement. Des règles de comportement ont été édictées au plus haut niveau de British Energy comme de ses prestataires principaux. Elles rappellent les exigences vis-à-vis de l’alcool et de la drogue ainsi que la nécessité d'avoir une attitude interrogative, d'utiliser son expérience professionnelle pour planifier les activités, de porter les équipements de protection, d’organiser un retour d’expérience, de ne pas accepter un chantier non sécurisé, de comparer et d'accepter le challenge des autres. En 2011, le taux de fréquence est de 0,6 par million d’heures.

2.7 La radioprotection

Pour le parc nucléaire d’EDF SA, l’année a été plus chargée en opérations de maintenance avec 9 arrêts décennaux, près du double de l’année précédente.

La dose collective

Le parc est entré dans un cycle d’oscillation continue de la dose collective lié à la typologie des arrêts pour maintenance. Cette oscillation s’inscrit dans le faisceau des résultats internationaux pour des réacteurs de technologie équivalente. L’augmentation du temps d’exposition des intervenants, toutes choses égales par ailleurs, accroît donc mécaniquement la dose collective par réacteur qui passe de 0,62 hSv en 2010 à 0,71 hSv en 2011.Sur les sites, je constate avec satisfaction des opérations visant à l’abaissement du niveau de débit d’équivalent de dose moyen (terme source) dans le cadre de la démarche ALARA. Ces opérations sont, pour certaines, ponctuelles liées à la présence de particules radioactives dosantes ou systématiques avant certaines opérations de maintenance.

La dose individuelle

Les doses individuelles continuent à baisser pour les métiers les plus exposés (échafaudeurs, calorifugeurs, soudeurs…). Ce constat est à mettre à l’actif de la relance de la démarche ALARA qui accompagne l’optimisation des doses sur les chantiers. Le nombre d’intervenants dont la dose sur 12 mois dépasse 16 mSv continue à décroître avec 2 cas signalés en 2011.

Les événements

Deux événements en radioprotection ont été classés en 2011 au niveau INES 1, dont l'un lors d’une opération de maintenance au cours de laquelle un intervenant a reçu une irradiation ponctuelle sur la main correspondant à une dose peau de 430 mSv, proche de la limite réglementaire (500 mSv). Je note une stabilité des événements significatifs en radioprotection (ESR). Je recommande de nouveau une plus grande rigueur sur certains points, en particulier : les accès en zones radiologiques rouge et orange, le port des équipements individuels en zone, les tirs radio (contrôles gammagraphiques), la préparation des interventions en zone contrôlée, les travaux en piscine des bâtiments réacteur et combustible. Je note que les processus d’intervention se sont renforcés sur les sites mais que les comportements attestent encore de difficultés d’adaptation.Les arrêts décennaux, par le volume de maintenance et des contrôles programmés, s’accompagnent d’une forte augmentation des tirs radio (contrôles gammagraphiques). De même pour le chantier EPR de Flamanville 3, le début du montage mécanique conduit à une activité intense de contrôles par tirs radio, opérations rendues plus complexes du fait de la grande variété des métiers présents sur le chantier et de personnes peu informées de ces activités à haut risque. Si le niveau général de maîtrise de ces activités a bien progressé en 2011 sur l’ensemble des sites, les situations de presque incidents

24 British Energy : devenu depuis EDF Energy.25 Taux de fréquence des accidents du travail par millions d’heures travaillées.

Page 20: EDFMESS FR - inis.iaea.org

18

2

survenues appellent à la plus forte vigilance. Je regrette à cet égard que le PSRP (poste de supervision radioprotection) qui doit, à terme, équiper chaque site ait fait l’objet d’une reprogrammation alors que de nombreux sites à l’étranger en sont équipés de longue date.Je note aussi des anomalies récurrentes concernant les restrictions d’accès en zones nucléaires pour les prestataires sous contrats de travail temporaires (CDD et Intérim). La situation ne s’améliore pas en 2011. La difficulté administrative pour EDF devrait pouvoir se résoudre dans le cadre de l’accord sur la « sous-traitance socialement responsable à EDF », proposé aux entreprises prestataires en France. J’estime que la situation doit impérativement être mise sous contrôle dans les meilleurs délais.La maîtrise du terme source a conduit quelques CNPE à s’engager progressivement sur l’ascension de l'« EVEREST radiologique », du nom du projet engagé il y a longtemps et atteint par quelques sites pilotes (Golfech, Civaux et Cattenom). Ce projet, que je soutiens fortement, consiste à rendre l’essentiel des zones nucléaires dépourvues de contamination et de radiation significative, donc accessibles en tenue de travail classique. Sur les trois sites pilotes, cette démarche a favorisé les temps d’intervention en zone nucléaire et donc le « temps métal » (cf. chapitre 9) et amélioré la présence de l’encadrement sur les chantiers. Elle est plébiscitée par l’ensemble des intervenants sur ces sites, malgré un très fort investissement nécessaire à la mise en place d’une nouvelle organisation de la logistique et des interventions.

Enfin, je suggère qu’une attention particulière soit apportée à la valorisation des métiers de prévention en général et de radioprotection en particulier ainsi qu’au maintien, voire au renforcement, des compétences pérennes pour ne pas risquer des à-coups comme ce fut le cas dans le passé.

Le parc d’EDF Energy est très majoritairement constitué de réacteurs AGR26 qui contribuent plus faiblement à la dosimétrie collective et individuelle que les réacteurs REP. Ses résultats 2011 sont aussi satisfaisants. La dose collective se situe à 0.11 hSv/réacteur. Les challenges les plus significatifs de 2011 étaient l’arrêt pour rechargement et travaux de Sizewell B et des travaux dans les réacteurs de Hunterston B, Hinkley Point B et Heysham B.

Les doses individuelles font l’objet, comme en France, d’une forte attention. La dose individuelle annuelle la plus forte se situe à 7,62 mSv.Les sites ont été équipés de nouveaux portiques de détection en sortie de zone contrôlée, permettant de déceler des traces de contamination à des niveaux plus bas, conformément aux standards internationaux.Le nombre d’événements significatifs en radioprotection est le même qu’en 2010. Aucun n’a concerné un niveau de dose individuelle interne ou externe significatif ni atteint un niveau de déclaration à l’autorité.

2.8 La prévention et la lutte contre l’incendie

Le domaine de l’incendie participe aussi à l’amélioration des résultats de sûreté pour l’année 2011. Je constate avec satisfaction une baisse significative du nombre des départs de feux.Le parc d’EDF SA a subi un incendie d’ampleur sur un pôle de transformateur, qui a confirmé la pertinence de l’organisation commune des centres de production nucléaires avec les SDIS27.

Sur les sites français, la mission des OSPP28 détachés sur chaque site se pérennise et le renouvellement de la première génération se passe bien. Après le travail de fond sur l’organisation et la lutte, cette nouvelle génération devra se focaliser sur la prévention et la qualité des exercices et des entraînements, pour lesquels j’ai noté certaines faiblesses dans la motivation des intervenants. Des exercices plus variés et des exercices en zone nucléaire sont nécessaires pour améliorer les performances.

26 AGR : Advanced Gas-cooled Reactor.27 SDIS : Services Départementaux d’Incendie et de Secours.28 OSPP : Officiers Sapeurs Pompiers Professionnels.

Page 21: EDFMESS FR - inis.iaea.org

19

2

Si l’organisation de la lutte contre les incendies se maintient à un niveau très satisfaisant, j’observe sur les sites une progression plus lente de l’appropriation de la culture de prévention par les intervenants, y compris les prestataires. Il en résulte un nombre encore trop élevé de départs de feux d’origine humaine. Dans une grande majorité des sites, les projets O2EI et MRI 29 ont amélioré la prévention du risque incendie, en particulier par des règles strictes de colisage lors des arrêts de tranche. Ces dispositions facilitent la progression des secours et améliorent l’entreposage des matériels et des produits dans les locaux. J’observe aussi que le nombre

d’alarmes incendie intempestives décroît, favorisant la crédibilité de l’ensemble du dispositif de détection. Je constate, hélas aussi, que le gréement de l’équipe de deuxième intervention n’est pas toujours systématique lors de l’apparition d’alarmes.Le contrôle de la sectorisation incendie, pilier de la doctrine de conception française, se heurte encore à des difficultés peu compréhensibles de bases de données sur certains sites. Cette situation, qui dure depuis trop longtemps, doit être corrigée au plus vite.

A l’occasion d’une rencontre avec l’équipe de pilotage du projet AMELIE, je note avec satisfaction que l’organisation de la lutte contre les incendies pour le magasin général de Velaines se met en place avec le SDIS sur le schéma des CNPE et qu’EDF contribue à la mise à niveau de la dotation en matériel du SDIS voisin.

2.9 Le renforcement du dispositif de protection des CNPE

Décidé en 2007, le renforcement du dispositif de protection des sites nucléaires par la Gendarmerie nationale se poursuit. Il sera finalisé pour l’ensemble du parc courant 2012, 2 sites sur les 20 restant à gréer. Pour 18 CNPE, les organisations locales des pelotons spécialisés de protection de la gendarmerie (PSPG) sont maintenant opérationnelles, leurs implantations immobilières ne sont pas encore toutes réalisées mais l’ensemble des projets avance.Des deux côtés, gendarmerie et EDF, les satisfactions s’expriment. Lors de mes visites, j’observe la présence de gendarmes en mission à l’intérieur des sites. Ce constat appelle deux remarques : d’une part, le personnel d’EDF et des entreprises prestataires accepte cette situation nouvelle longtemps crainte, d’autre part, les sites EDF s’apparentent désormais à ce que j’observe de raisonnable à l’étranger. Cette nouvelle organisation facilite aussi les démarches administratives préalables à la délivrance des autorisations d’accès aux sites. Les procédures d’EDF, des gendarmeries et des préfectures se sont optimisées à la satisfaction générale.La protection matérielle se renforce aussi et de nombreux matériels anciens sont remplacés. Les conditions de travail dans les postes de surveillance des blocs de sécurité s’améliorent nettement.Les événements d’intrusion par une association internationale, le 5 décembre 2011, ont eu un retentissement médiatique important, but recherché par les intrus. J’estime que ces actes irresponsables et intolérables ne traduisent en rien une vulnérabilité du dispositif mis en place par EDF et la gendarmerie. Les intrus n’ont pas pu atteindre de zones sensibles et les forces de l’ordre ont procédé avec discernement à leur interpellation.De manière plus générale, je considère cependant, cette année encore, que la coordination entre la direction de la sécurité du Groupe, la mission sécurité de la DPN, ses appuis d’ingénierie à la DIN (au CNEPE) et les services locaux de protection des sites mériterait une organisation plus proche de celle de la sûreté nucléaire et de ses contrôles.

L’année a été marquée par plusieurs inspections de l’autorité en matière de sécurité des sites : le Haut Fonctionnaire de Défense et de Sécurité (HFDS) rattaché au ministère de l’Ecologie, du Développement durable, du Transport et du Logement (MEDDTL).

29 MRI : Maîtriser le Risque Incendie (projet de la DPN).

Page 22: EDFMESS FR - inis.iaea.org

20

2

Je note aussi qu’en fin d’année 2011, à la demande du gouvernement français et en accord avec EDF SA, l’AIEA a mené en France une visite de type IPPAS30 au CNPE de Gravelines et a formulé à cette occasion des appréciations et recommandations fondées sur la cohérence avec le référentiel international (guidelines INFCIRC/225/Rev.5) et les meilleures pratiques internationales. Comme toutes les visites de ce type, elle sera suivie de visites de contrôle pour le traitement des recommandations et d’assistance éventuelle.Je me félicite que, comme pour la sûreté, ce domaine soit ouvert au regard externe.

2.10 La déconstruction

Les seules unités de production nucléaire en déconstruction dans le groupe EDF se trouvent en France.

Cette année, j’ai visité le chantier de déconstruction du premier REP construit en France, à Chooz dans les Ardennes. La déconstruction de cette tranche atypique, dont la partie nucléaire est située dans une caverne, se poursuit depuis quelques années. Les travaux, engagés en 2010, concernent maintenant le circuit primaire principal. La décontamination des générateurs de vapeur était en cours afin de les déclasser dans l’échelle des déchets nucléaires. Sous la maîtrise d’ouvrage de la DIN (le CIDEN31), ces travaux sont réalisés par des entreprises qui doivent acquérir les compétences et l’expérience nécessaires à

ces chantiers, qui seront tous très singuliers pour les réacteurs des quatre filières en cours de déconstruction : eau-lourde (Brennilis), REP (Chooz), Uranium Naturel Graphite Gaz (UNGG32 à Bugey, Chinon et Saint-Laurent) et neutrons rapides (RNR33 à Creys-Malville). Ce travail constitue un challenge intéressant, il appelle des solutions industrielles novatrices pour répondre à des difficultés techniques spécifiques. Les solutions pour stocker définitivement les différents types de déchets se heurtent au contexte politique et social national, pouvant conduire à des retards dans les travaux de déconstruction ou la réalisation d’installations d’entreposage provisoire de déchets très onéreuses. Ce fut le cas récemment à Brennilis où les travaux peuvent reprendre depuis novembre 2011.Je constate que la stratégie d’EDF de déconstruire au plus tôt est aujourd’hui confrontée à des obstacles administratifs, juridiques et techniques, qui sont alimentés par des mouvements associatifs ou qui sont du ressort des autorités de l’Etat, pour les filières de déchets radioactifs.

30 IPPAS : International Physical Protection Advisory Service. L’IPPAS a été créé par l’AIEA pour aider les Etats à renforcer leur référentiel de sécurité nucléaire au sens de la protection des sites en particulier.

31 CIDEN : Centre d’Ingénierie Déconstruction et Environnement Nucléaire.32 UNGG : Uranium Naturel Graphite Gaz.33 RNR : Réacteur à Neutrons Rapides.

Page 23: EDFMESS FR - inis.iaea.org

21

3

LE MANAGEMENT DE LA SURETE

L’engagement des managers, à chaque niveau de responsabilité

dans l’entreprise, est aujourd’hui une force bien visible sur

tous les sites du Groupe. Cependant, la priorisation des projets

n’est pas encore suffisamment perçue par les agents sur le

terrain. La responsabilisation accrue des directions locales

améliorerait l’efficacité des actions et la sérénité des équipes.

Le management de la sûreté doit à la fois obtenir de bons résultats aujourd’hui et préparer les bons résultats de demain.Si les résultats de cette année sont de nouveau satisfaisants dans l’ensemble, avec de vrais succès dans certains domaines, je me suis attaché à identifier les progrès à consolider et les domaines à améliorer pour bien préparer les résultats de demain.

Une politique sûreté pour le groupe EDF

Je me félicite de voir engagée la rédaction de la première politique sûreté Groupe, applicable à toutes les entités exerçant une responsabilité nucléaire à EDF SA en France, EDF Energy au Royaume-Uni, CENG aux Etats-Unis et TNPJVC34 en Chine.Sa validation, au début de l’année 2012, permettra de réaffirmer, au plus haut niveau du Groupe, la priorité absolue à accorder à la sûreté nucléaire, en toutes circonstances et en tout lieu, ainsi que l’engagement d’EDF dans le renforcement du rôle de WANO après Fukushima. Elle permettra également

34 TNPJVC : Joint venture entre la compagnie chinoise CGNPC (70 %) et EDF (30 %).

Page 24: EDFMESS FR - inis.iaea.org

22

3

de réafficher, voire de préciser, les différents niveaux de responsabilité de sûreté nucléaire dans l’organisation du Groupe. J’encourage tous les managers d’unités opérationnelles, sites de production et centres d’ingénierie, à saisir l’opportunité de sa diffusion, pour réaffirmer la prééminence de

la sûreté dans le comportement de tous les salariés du groupe EDF et des entreprises prestataires.

Des directeurs de CNPE responsabilisés

Je constate que les directeurs d’unités, en particulier sur les sites de production, ont le sentiment

d’être écoutés, compris et mieux aidés dans l’exercice de leur mission délicate.Le principe de subsidiarité, affiché avec force au niveau national, devient une réalité, avec un meilleur équilibre entre ce qui doit relever du niveau national, et ce qui revient au niveau local où s’exercent les responsabilités opérationnelles les plus directes.

La création des directeurs délégués opérations (DDO) à la direction de la DPN en est une bonne illustration. Les prérogatives qui leur sont attribuées en font des interlocuteurs reconnus des

directions de CNPE, avec un réel pouvoir d’arbitrage pour la priorisation des grandes actions et des ressources. Tournés vers les sites, ils en ont une connaissance fine, ce qui est très précieux dans la gouvernance d’un parc de 19 unités. Leur mission, qui recouvre l’ensemble des champs d’intégration, est bien là pour faciliter le travail des directeurs d’unité sans les déresponsabiliser.

Renforcer encore la subsidiarité.

Je considère ces évolutions très positives. Il faudra savoir en assurer la pérennité et, plus largement, veiller à ce que cette subsidiarité nouvelle reste bien dans le cadre défini.

Une filière indépendante forte et écoutée sur les sites

Une des caractéristiques de l’exploitation nucléaire en France repose sur l’existence, aux côtés de la ligne managériale, d’une ligne de contrôle indépendante, dotée de compétences significatives : la filière indépendante de sûreté (FIS). Cette ligne est en place à trois niveaux : site de production ou centre d’ingénierie, division opérationnelle, tête du Groupe.Je constate que cette filière indépendante de sûreté, sur la quasi-totalité des sites visités, est solide et dotée de personnels qualifiés. Elle est écoutée par les directeurs d’unités qui organisent des rencontres régulières avec les chefs de mission sûreté et les ingénieurs sûreté (IS).J’observe également avec intérêt les initiatives prises d’EDF Energy pour faire évoluer les moyens de contrôle sûreté en « temps réel » sur les sites, qui méritent en effet d’être renforcés.En 2011, un benchmark avec deux CNPE du parc français puis une expérimentation sur trois sites britanniques témoignent de la vitalité croissante des échanges entre les exploitants des deux parcs.

Un besoin de compétences

process.

Je relève toutefois un point de vigilance sur le renouvellement des compétences

des ingénieurs sûreté et sur la fragilité des équipes d’audit locales. Faute de pouvoir s’appuyer sur des compétences suffisantes en conduite, le gréement actuel des emplois d’IS privilégie de manière excessive les profils de jeunes ingénieurs.

Les programmes de vérification sont bien structurés et planifiés à l’année, sur la base d’un cadre défini nationalement. Je note que ce cadre laisse toutefois peu de place à l’initiative de contrôle locale, ce qui mérite attention, d’autant que je tiens de nouveau à souligner la qualité croissante des diagnostics

annuels réalisés par les unités. Les CNPE, en particulier, élaborent leur analyse annuelle de sûreté (AAS) en s’appuyant bien sur les regards croisés du management et de la filière indépendante de sûreté.

Je considère que les comités locaux de sûreté (GTS/CTS) constituent un levier de progrès puissant, s’ils sont à la fois de véritables lieux d’expression des managers de l’unité et de support au processus de décision.

Page 25: EDFMESS FR - inis.iaea.org

23

3

Je rencontre également à l’étranger des exploitants qui cherchent à renforcer ces moyens de contrôle indépendants, sur les sites ou au niveau corporate. Que ce soit avec les Chinois de CGNPC, les Sud-Africains d’ESKOM35 ou les Russes de Rosenergoatom, les actions de coopération se développent avec EDF, principalement avec l’inspection nucléaire de la DPN, et témoignent d’une prise de conscience

croissante dans ce domaine.

Une autorité des chefs d’exploitation à renforcer

Sur les sites de production nucléaire, les chefs d’exploitation assurent une mission de contrôle sûreté essentielle, 24 heures sur 24, par délégation directe du directeur d’unité. J’observe que cette mission délicate est bien reconnue dans les unités et qu’elle est exercée avec compétence et un grand sens des responsabilités. Ces acteurs clés ont également pour mission de s’assurer de la bonne marche de

l’exploitation, en supervisant les activités engagées sur les tranches et en procédant aux arbitrages et aux priorisations nécessaires.

Je m’interroge, en France, sur le niveau de recul de ces chefs d’exploitation qui peuvent parfois être trop « embarqués » par le temps réel ou par des activités dont ils devraient être déchargés. Leurs homologues dans les centrales à l’étranger, particulièrement aux Etats-Unis et au Royaume-Uni, me paraissent plus sereins dans la gestion de leurs activités quotidiennes, dotés d’appuis nombreux et efficaces, et disposant du recul indispensable au bon exercice de cette fonction « ligne de défense ».

Je mesure aussi, lors d’une de mes visites en CNPE, combien le rétablissement d’une relation

privilégiée et directe entre le directeur du site et ses chefs d’exploitation permet de mobiliser efficacement les équipes de conduite sur les objectifs de sûreté et, plus généralement, d’exploitation. L’existence de ce type de liens, entre ces deux niveaux de management essentiels pour une bonne maîtrise de la sûreté en exploitation, mérite une attention particulière.

Décider en toute sûreté.

Je considère aussi qu’une attention particulière est à accorder sur chaque site au bon déroulement des processus de décisions opérationnels. Ces processus doivent toujours placer le chef d’exploitation dans une position de décideur incontestable, bénéficiant des appuis techniques, voire de la direction,

indispensables. Le juste positionnement des décideurs de niveau direction de site (appelé « PCD1 » en astreinte) mérite la même attention.

Je note la volonté en France de la DPN de redonner sens et efficacité à une démarche OSRDE36 essoufflée, sous une forme plus simple pour être davantage utilisée. Initiée en son temps par l’un des mes prédécesseurs, elle recommande à l’exploitant nucléaire d’évaluer régulièrement, à partir de cas concrets, la pertinence des arbitrages passés ayant pu engager la sûreté. J’estime que le renouvellement important des managers sur les sites, renforce ce besoin de relance. Cette question concerne aussi EDF Energy et CENG, où des initiatives sur l’"operationnal decision making" sont prises en ce sens (cf. chapitres 10 et 12).

Des managers mobilisés sur les sites

Je constate, cette année encore, le haut niveau d’engagement et de mobilisation des managers sur l’ensemble des sites visités. Sur tous les fronts, exerçant une mission parfois ingrate, ils méritent une attention particulière de la part des directions locales.

Je me réjouis de la montée en puissance de véritables formations au métier de manager, appréciées de tous. Sur ce plan, le programme d’EDF Energy pour le développement du leadership est également remarquable. La volonté affichée de recentrer les managers sur le métier et sur le « focus opérationnel »

35 ESKOM : Compagnie d’électricité sud africaine.36 OSRDE : Observatoire Sûreté Radioprotection Disponibilité Environnement.

Page 26: EDFMESS FR - inis.iaea.org

24

3

répond très largement à leurs attentes et à celles de leurs équipes de travail. Pour autant, ces managers de proximité sont toujours aussi expressifs quand je les rencontre sur leurs difficultés de logistique quotidienne et sur le poids croissant de ce que chacun nomme le tertiaire diffus. J’attire donc à nouveau l’attention sur cette réalité, à laquelle il convient de s’attaquer résolument. Les bonnes pratiques que j’observe sur certains sites, consistant, à l’instar des pratiques internationales, à mettre en place des appuis dédiés dans les services, méritent d’être développées.

Les indicateurs de résultat

Je veux d’abord souligner le travail commun engagé cette année par les parcs français et britanniques pour définir davantage d’indicateurs de résultats homogènes et rendre plus efficace l’inter-comparaison (cf. annexe 14).Dans le parc français, le nombre d’ESS déclarés en 2011 à l’ASN demeure élevé (10,6 par réacteur), avec toutefois une réduction sensible des événements classés au niveau 1 de l’échelle internationale INES (0,91 par réacteur). Ce point positif est à retenir, malgré la déclaration d’un ESS de niveau 2 sur l’échelle INES (début 2011), suite à des défauts matériels sur des groupes électrogènes de secours du palier 900 MWe (cf. chapitre 12).

A EDF Energy, le nombre d’ESS déclarés en 2011 baisse (4,7 par réacteur), avec des pratiques de déclaration différentes au Royaume-Uni et en France, compte tenu des exigences des autorités de sûreté respectives. Plus comparables, le nombre d’événements classés sur l’échelle INES, tous de niveau 1 en 2011, est supérieur à celui de 2010 et du parc France (1,3 par réacteur).

Je considère que le niveau de transparence d’EDF SA et d’EDF Energy est bon, reposant sur une détection de plus en plus fine des écarts sur le terrain, par les techniciens eux-mêmes, et sur des pratiques managériales sans ambiguïté, témoignant ainsi d’une solide culture de sûreté.

Trois résultats méritent d’être relevés car ils illustrent le retour sur investissement d’actions engagées il y a plusieurs années au niveau national comme au niveau local, et portent sur des domaines clés : les arrêts automatiques de réacteur, le respect des STE37 et l’incendie.

La réduction des arrêts automatiques de réacteur

Record à battre.

Les AAR38 atteignent cette année un niveau très bas en France (0,50 AAR/réacteur/an), nouveau record dans l’histoire du parc. Ce résultat place les réacteurs du parc EDF France parmi les meilleurs au monde, avec 38 réacteurs qui n’ont pas

connu d’AAR sur l’ensemble de l’année. Il est lié aux actions de long terme engagées il y a plusieurs années pour améliorer à la fois la fiabilité des matériels et celle des interventions humaines. Je note toutefois la proportion, plus élevée cette année, d’AAR d’origine humaine qui fait écho au risque d’essoufflement du projet performance humaine, au regard des pratiques de fiabilisation.A EDF Energy, comme aux Etats-Unis, il y a lieu de considérer le cumul AAR et arrêts manuels, compte tenu des pratiques d’exploitation différentes. Les procédures britanniques demandent plus fréquemment aux opérateurs de déclencher manuellement l’arrêt du réacteur, avant l’action des automatismes. Les résultats 2011 sont stables par rapport à l’année précédente (1,96 par réacteur) et toujours en retrait par rapport aux meilleures performances mondiales.

37 STE : Spécifications Techniques d’Exploitation.38 AAR : Arrêt Automatique Réacteur.

Page 27: EDFMESS FR - inis.iaea.org

25

3

La maîtrise du référentiel d’exploitation

Sur le parc France, le respect des spécifications techniques d’exploitation (code de la route de l’exploitation nucléaire) atteint son meilleur niveau depuis 2004. Très adaptées aux besoins, les actions d’accompagnement et de formation engagées par les CNPE sont à l’origine de ces résultats. Je recommande de veiller à la pérennisation de ces bonnes pratiques. La réduction du nombre de sorties du domaine de fonctionnement autorisé, point sur lequel j’avais insisté dans mon précédent rapport, contribue également à ce meilleur respect des STE et traduit des progrès de pilotage depuis la salle de commande.

A EDF Energy, les résultats sont stables dans ce domaine, sans retrouver les bons niveaux des années 2007-2009. C’est un des axes de progrès prioritaires sur lequel est engagé le parc britannique.

La maîtrise du risque incendie

Je veux souligner de nouveau (cf. chapitre 2) les bons résultats du parc EDF en matière d’incendie tant ce risque est majeur et mérite une approche déterminée.La réduction du nombre de départs de feu en 2011 est spectaculaire (proche de 20 %) et met en lumière la pertinence des actions engagées ces dernières années. Les sites sont tous dotés d’organisation incendie robustes, s’appuyant sur le réseau des officiers sapeurs-pompiers professionnels (un par site) et sur des relations de coopération remarquables avec les SDIS.La prévention de ce risque, sur les chantiers principalement, n’a pas encore atteint partout le

niveau visé mais partout les plans d’action existent et sont sous contrôle.Je relève que le dispositif de pérennisation du projet MRI est solide et me donne confiance dans l’aboutissement complet de ce programme ambitieux. Ce domaine illustre bien ce qu’un exploitant responsable, avec l’aiguillon de l’autorité de sûreté, peut faire de meilleur en matière de réexamen périodique de sûreté.

En Grande-Bretagne, les résultats 2011 sont également bons avec une diminution sensible des départs de feu. Toutefois, les "peer reviews" de WANO conduites ces deux dernières années sur les sites mettent souvent en lumière des fragilités dans l’entretien des systèmes d’incendie et la maîtrise des charges calorifiques.

Mes sujets de préoccupation

Les non-qualités de maintenance sur les chantiers et les non-qualités de conduite, relevées principalement sur les lignages et la configuration des circuits, et trop souvent encore sur les opérations de pilotage depuis la salle de commande, montrent une maîtrise encore insuffisante, voire en recul, des activités « cœur de métier ». Dans un contexte de très fort renouvellement des hommes et des femmes, la maîtrise de certaines « règles de l’art » et connaissances du métier doit faire l’objet d’un traitement prioritaire.

Du sens et du métier.

Ce qui me préoccupe le plus, ce sont les causes assez communes de ces écarts qui mettent souvent en lumière les difficultés de surveillance des chantiers et des prestataires, la perte du sens de certaines analyses de risque et un contrôle

technique, à charge des services EDF, trop fragile. Ce point méritera dans les années à venir la mobilisation de beaucoup d’acteurs pour inverser la tendance.

Page 28: EDFMESS FR - inis.iaea.org

26

3

Aller au bout des actions engagées.

Les pratiques de fiabilisation, après un effort de formation sans précédent ces dernières années, ne sont pas encore assez généralisées. Certains métiers sont en pointe, comme les automaticiens et les équipes de manutention combustible, mais je constate encore la distance qui sépare les pratiques observées en salle de commande en France et à EDF Energy.

Je mesure l’ampleur des actions engagées, mais j’attire aussi l’attention sur le risque d’essoufflement

du projet performances humaines (PPH) si les managers n’affichent pas davantage leurs exigences sur ces nouvelles pratiques. Je considère que la priorité doit être accordée aux évolutions dans les métiers de conduite.Enfin, tout en me félicitant de l’inversion de tendance concernant les sorties du domaine de fonctionnement autorisé, j’alerte sur la récurrence d’événements significatifs touchant la maîtrise de la réactivité, en particulier des opérations de dilution du circuit primaire principal, déjà mentionnée dans mon précédent rapport. Ce domaine mérite des actions déterminées pour la formation à la conduite de ces opérations de pilotage, qui ne doivent en aucun cas être banalisées. L’événement traité au chapitre 12 illustre cette préoccupation.

Un dispositif de contrôle national en évolution

La DPN s’appuie sur un processus sûreté solide. Les revues annuelles permettent de bien dégager les grandes questions issues des analyses d’événements du parc en exploitation et de bien définir les orientations d’amélioration inscrites dans les CAP39 des sites. Les forces et faiblesses du parc identifiées au terme de ce travail sont les bonnes. Le calendrier de ce processus a été revu et se situe désormais en amont des revues de sûreté des CNPE, de plus en plus cadencées par le calendrier type des revues des SMI. Cette évolution répond à une attente forte des sites, de plus en plus nombreux à avoir mis en œuvre le SMI. Elle permet une définition anticipée des priorités, donc une bonne dynamique de réalisation des actions au plus tôt dans l’année.

Je relève avec satisfaction que l’inspection nucléaire a pu cette année réaliser l’ensemble de son programme de contrôles, contrairement aux dernières années, comme j’avais déjà eu l’occasion de le regretter dans mes précédents rapports. Le renforcement significatif du nombre d’inspecteurs en est la raison principale. Il convient maintenant de veiller à la pérennité de cette évolution positive.Je note la décision de simplifier le dispositif de contrôle des CNPE, en faisant évoluer les évaluations globales de sûreté (EGS) vers des évaluations globales d’excellence (EGE). Ces EGE, couvrant plus largement les attendus d’exploitation, par exemple la sécurité au travail et la capacité à produire, seront réalisées sur chaque site tous les quatre ans, en alternance avec les peer reviews de WANO dont la fréquence est portée à quatre ans également (six ans auparavant). Chaque site sera ainsi évalué par l’inspection nucléaire ou par les équipes de WANO tous les deux ans. Je me félicite de ce renforcement, comparable à ce qui est mis en œuvre par l’INPO aux Etats-Unis par exemple. J’attire toutefois l’attention sur la priorité qui devra toujours être accordée à la sûreté nucléaire dans ces évaluations et sur la nécessité de dégager des ressources supplémentaires, de bon niveau, que réclament ces évolutions.

Décidé en 2011, le renforcement positif des moyens accordés à WANO pour assurer un meilleur exercice de ses activités d’évaluation des sites nucléaires dans le monde nécessitera également de faire appel à des compétences expérimentées, venant des CNPE. Je considère qu’il y a là un point de vigilance quant à la capacité de l’entreprise à faire face à ces besoins.

Le dispositif de contrôle de sûreté de la DIN, assuré par la mission d’appui et d’évaluation (MAE), s’inspire de l’expérience de l’inspection nucléaire, et me semble toujours pertinent. Il nécessite, lui aussi, d’être doté des ressources en inspecteurs, lui permettant d’assurer pleinement ses missions.

A EDF Energy, il n’existe pas de service d’inspection interne pouvant procéder à des évaluations ou audits de grande ampleur, comparables aux EGS. Le choix a été fait de confier le contrôle des sites aux peer reviews de WANO, réalisées tous les trois ans, et privilégiant le regard externe. Au niveau central, 39 CAP : Contrat Annuel de Performances.

Page 29: EDFMESS FR - inis.iaea.org

27

3

le département SRD40 rassemble des ressources d’ingénierie importantes, délivrant une bonne vision et analyse des résultats et tendances de sûreté du parc et des sites (comparable avec la mission assurée à la DPN par le groupe performances sûreté nucléaire, GPSN). Son maillage avec les inspecteurs résidents des sites est perfectible et je note avec satisfaction, comme je l’ai déjà évoqué, qu’il est en cours de ré-interrogation, avec le renforcement des équipes d’inspection de site.

Des grands projets mieux priorisés, une réalité de terrain encore difficile

Accompagner les changements.

Pour avoir mis en garde par le passé devant le risque de surcharge lié à la conduite d’un trop grand nombre de ces projets en parallèle dans les CNPE (cf. chapitre 8), je me félicite des décisions d’étalement dans le temps prises cette année. J’observe lors de mes visites que ces décisions ont permis de construire des

programmations pluriannuelles plus robustes, prenant mieux en compte la capacité d’absorption de

chaque CNPE. Cette évolution importante n’est toutefois pas encore perçue sur le terrain.Le nombre, encore considérable, d’évolutions des pratiques professionnelles et des organisations méritera, dans la durée, un accompagnement renforcé en matière de conduite du changement et d’écoute des difficultés rencontrées par les équipes.

En conclusion, le management de la sûreté est porté par des acteurs de terrain mobilisés, mais dont l’action est trop souvent dispersée par des priorités encore trop nombreuses et par des appuis logistiques insuffisants. Les performances de sûreté sont satisfaisantes, avec des succès indiscutables et des fragilités bien identifiées par les managers. L’amélioration de ces résultats passe par une priorité à accorder aux apprentissages-métiers pour les nombreux jeunes rejoignant l’exploitation nucléaire, et par encore plus d’autonomie à accorder au niveau local.

40 SRD : Safety and Regulation Department.

Page 30: EDFMESS FR - inis.iaea.org
Page 31: EDFMESS FR - inis.iaea.org

29

4

L’ACCIDENT DE FUKUSHIMA DAIICHI

Une catastrophe naturelle déclenche un événement qui façonne

le futur de l’industrie nucléaire, bouscule sa gouvernance

mondiale et l’oblige à explorer sa conception au-delà du

référentiel. Le groupe EDF a fait preuve d’une grande réactivité

pour répondre aux exigences de réévaluation de sûreté et de

projection sur l’avenir. Le défi pour les ingénieries ne devra

pas réduire l’appui au parc en exploitation dans le domaine

de la sûreté.

Avis au lecteur : ce chapitre est fondé sur les analyses internationales connues au 31 décembre 2011 et ne couvre pas toute les dimensions de cette catastrophe.

Le 11 mars 2011, un tremblement de terre d’une rare violence s’est produit en mer, à 180 km du site nucléaire de Fukushima Daiichi de TEPCO41, comportant 6 réacteurs à eau bouillante (REB). Il a provoqué l’arrêt automatique des réacteurs en fonctionnement, par activation des protections contre les séismes. Des disjoncteurs électriques et le réseau d’alimentation électrique ont été endommagés. Les diesels de secours ont automatiquement démarré pour alimenter les systèmes de sécurité; 41 minutes après le tremblement de terre, une première série de 7 tsunamis est arrivée sur le site. La hauteur maximale des vagues a été de 14 à 15 mètres alors que la protection du site était dimensionnée pour 5,7 mètres, ce qui a conduit au noyage de la plateforme usine, en particulier des diesels de secours et des locaux de distribution électrique. Seules quelques batteries sont restées disponibles sur le

41 TEPCO : Compagnie d’électricité japonaise.

Page 32: EDFMESS FR - inis.iaea.org

30

4

réacteur n° 3 ainsi qu’un diesel pour les réacteurs 5 et 6. Les actions des intervenants ont été rendues très difficiles par la perte de l’éclairage sur les 4 installations impactées.Sans refroidissement, un début de fusion du cœur a commencé le jour de l’accident sur le réacteur n°1. Sur les réacteurs 2 et 3, un refroidissement par des turbopompes restées disponibles a retardé la fusion.L’hydrogène produit par la détérioration du combustible a généré des explosions sur les réacteurs 1 et 3, ce qui a fortement compliqué les interventions. La montée en pression des enceintes de confinement a nécessité leur décompression. Pendant les périodes d’éventage, des débits d’équivalent de doses de 12 mSv/h ont été mesurés en limite de site. Cent personnes ont reçu une dose supérieure à 100 mSv mais, selon l’AIEA, la contamination relâchée dans l’atmosphère a été 3 000 fois plus faible qu'à

Tchernobyl.

Une mobilisation forte face à l’évènement

L’ensemble du monde nucléaire a été très réactif pour comprendre cet accident et en tirer les premiers enseignements :

11 mars 2011 : accident à la centrale nucléaire de Fukushima,18 mars 2011 : WANO, par la SOER 2011-0242, prescrit à tous ses membres d’évaluer, sous un mois, le niveau de préparation aux accidents de leurs installations et d’en rendre compte dans des rapports,mars 2011 : les gouvernements commencent à demander à leurs autorités réglementaires de diligenter des enquêtes,mai/juin 2011: la mission d'experts internationaux de l’AIEA, qui réunit des représentants de douze nations, mène une enquête au Japon,juin 2011 : conférence ministérielle de l’AIEA sur la sûreté nucléaire,octobre 2011 : les membres de WANO décident à l'unanimité de modifier le mode de fonctionnement de l'association et de renforcer ses moyens et son autorité,1er semestre 2012 : revues de pairs des autorités de sûreté des pays de l’Union européenne,juin 2012 : remise des conclusions à la Commission européenne.

Les premiers enseignements communiqués par les autorités japonaises

Un rapport immédiat sans concessions.

Lors de la conférence ministérielle de l’AIEA à Vienne en juin, le gouvernement japonais a présenté un rapport très complet sur l'accident. Le chapitre consacré aux « leçons tirées de l'accident à ce jour » intéresse directement EDF et tous les exploitants nucléaires. Il ne faut pas oublier que les interventions d'urgence ont été menées alors que le tremblement de terre et le tsunami avaient

détruit les équipements collectifs, notamment les réseaux électriques, de communication et de transport, dans une grande partie de la zone environnant la centrale, et causé plus de 20 000 morts ou disparus. En outre, de nombreuses répliques sismiques ont entravé ces interventions. Le rapport relate les événements de façon analytique, avec une grande franchise, puis classe les enseignements qui s'en dégagent en cinq catégories : la suffisance des mesures de prévention, l’adéquation de la réponse technique à cet accident grave, l’adéquation des réponses des secours, la robustesse des dispositifs de sûreté sur les sites, les réflexions sur la culture de sûreté. Tous les enseignements de ce rapport du gouvernement japonais mériteront d’être examinés de plus près. A titre d’exemple, j’en ai sélectionné quelques-uns dans chaque catégorie.

42 SOER 2011-02 : Significant Operating Experience Report - “Fukushima Daiichi Nuclear Station Fuel Damage Caused by Earthquake and Tsunami”.

Page 33: EDFMESS FR - inis.iaea.org

31

4

La suffisance des mesures de prévention

Les protections du site contre les effets des tsunamis n’étaient pas suffisantes.Les mesures de gestion des accidents graves n’ont pas été prédéfinies avec la rigueur nécessaire, car généralement considérées comme relevant d’une démarche volontaire des exploitants et non d’une exigence légale. De plus, les directives de gestion des accidents graves n'ont pas été révisées depuis leur élaboration en 1992, ni renforcées ou améliorées.Les mesures prises pour garantir l’autonomie de chaque tranche durant un accident n’étaient pas suffisantes.

L’adéquation de la réponse technique à cet accident grave

Le contexte de réponse aux accidents doit être complété par des mesures telles que le renforcement des protections radiologiques dans les salles de commande et les centres

de secours, les systèmes de ventilation et de conditionnement d'air doivent être améliorés, les équipements auxiliaires doivent être renforcés, notamment en se dotant de systèmes de communication et d'éclairage autonomes.Une quantité adaptée de dosimètres individuels, de combinaisons de protection et d’équipements d’urgence doit être disponible sur le site. A Fukushima, la plupart des dosimètres personnels et des dispositifs de mesure de dose étaient inutilisables car ils avaient été immergés dans l'eau de mer.

L’adéquation des moyens de secours

La répartition des rôles et responsabilités entre les organisations de TEPCO et du gouvernement, au niveau central et local est à clarifier.La réponse aux propositions d'assistance des autres pays et la communication auprès de la communauté internationale doivent être améliorées. Par exemple, il faut dresser une liste des fournitures et matériels nécessaires pour réagir efficacement à tout type d'accident, définir à l'avance des points de contact dans chaque pays en cas d'accident et améliorer les dispositifs de partage de l'information.

La robustesse des dispositifs de sûreté sur les sites

La prise en compte des nouvelles connaissances et expertises, notamment pour la conception des systèmes, doit être réévaluée. Un nouveau cadre légal clarifiera le statut des rénovations.L’amélioration de la compétence des acteurs, exploitants et régulateurs, doit conduire au niveau des formations à mettre l'accent sur la sûreté nucléaire, la préparation et la réponse aux urgences nucléaires, la gestion de crise et la médecine nucléaire.

Les réflexions sur la culture de sûreté

Les organismes ou acteurs de la réglementation nucléaire nationale, auxquels il incombe d'assurer la sécurité nucléaire du public, doivent mener une réflexion pour déterminer s'ils ont véritablement pris en compte les progrès dans les connaissances avec la réactivité et la rapidité nécessaires.

••

Page 34: EDFMESS FR - inis.iaea.org

32

4

Mes premiers enseignements

Un retour d’expérience

inédit.

Pour ma part, j’ai retenu dans l’immédiat et avant toute analyse plus approfondie :• l’importance du dimensionnement initial et l’intérêt d’une réévaluation

systématique des protections lors des revues décennales de sûreté, conformément à la loi française,

l’intérêt d’avoir des moyens rapidement disponibles, donc disposés dans un environnement sur site et faciles à déployer, compte tenu de la rapidité avec laquelle un début de fusion du cœur intervient sur les réacteurs à eau pressurisée ou bouillante, en l’absence totale de refroidissement et d’alimentation électrique,

l’intérêt des recombineurs d’hydrogène ou des systèmes de ventilation plus robustes,

compte tenu des effets dévastateurs des explosions d’hydrogène sur la mise en œuvre des moyens de secours,

l’intérêt de prévoir des branchements pour les moyens de secours, pour pouvoir aider un site pratiquement isolé et partiellement détruit,

la nécessité d’intégrer à la gestion de crise le scénario d’un accident affectant plusieurs tranches

d‘un même site,

la nécessité de disposer de compétences de haut niveau, rapidement disponibles, dans la chaîne de décision des pouvoirs publics et politiques,

les conditions très sévères dans lesquelles sont intervenus avec courage les personnels de

TEPCO et leurs prestataires, pour rétablir le minimum de fonctions, sans beaucoup d’aide au début en raison de l’isolement du site.

Les actions internes au groupe EDF en France

EDF a mis une grande énergie, dès les premiers jours, pour proposer et étudier des mesures immédiates de soutien. C’est ainsi qu’avec AREVA et le CEA, EDF a sollicité les moyens

d’intervention en milieu hostile du groupe INTRA43. Le groupe EDF a aussi cherché à évaluer l’impact à long terme pour son parc nucléaire. Cela a été facilité par le modèle intégré

d’EDF avec, aux côtés de l’exploitant, une ingénierie forte de conception et d’exploitation.

Les « revues de paquetage »

La conformité est une priorité.

Je note la réactivité d’EDF pour tirer des conclusions rapides qui ont servi avec celles de l’INPO à la rédaction du cahier des charges de la SOER 2011-02 de WANO. Dès le mois de mars, l’objectif a été d’évaluer sur chaque site, la conformité des installations et organisations au référentiel de dimensionnement,

notamment vis-à-vis des risques d’inondation, de séisme, de perte de la source froide ou de perte de l’alimentation électrique. Ce travail a été apprécié par l’ASN.J’insiste sur la priorité à accorder à la correction des écarts de conformité encore présents, en particulier pour les situations accidentelles de perte de sources d’énergie ou de refroidissement.Je relève que ces analyses n’ont pas mis en évidence de risque significatif pour la sûreté des installations.

43 INTRA : groupe d’INTervention Robotique sur Accident.

Page 35: EDFMESS FR - inis.iaea.org

33

4

La démarche des évaluations complémentaires de sûreté

Les évaluations complémentaires de sûreté (ECS), à la demande du gouvernement et de l’ASN ont permis une évaluation au-delà du domaine de dimensionnement de chacun des sites. Ces évaluations ont été remises à l’ASN mi-septembre 2011 et ont été rendues publiques.La démarche consiste à étudier des situations extrêmes sur une base purement déterministe, indépendamment de leur caractère plausible ou de toute probabilité d’occurrence. Elle vise en particulier à déterminer à partir de quand interviendraient des rejets significatifs dans l’environnement et à mettre en évidence les effets falaise (ou à seuil), face à la perte des moyens de protection des installations ou de maîtrise des situations accidentelles.Au-delà du dimensionnement actuel, ont été étudiés un tremblement de terre 1,5 fois supérieur au dimensionnement, des crues multipliées par un facteur 1,3 par rapport aux crues millénaires majorées, des pluies 2 fois supérieures au dimensionnement. Enfin, a été prise en compte la perte totale des alimentations en eau et en électricité pendant huit jours, le dimensionnement actuel étant de 24 heures sans source électrique externe ou de 100 heures sans source froide pour tout site. La priorité pour EDF est de finaliser les études dans les rapports d’évaluation, de conforter les parades proposées et de les mettre en œuvre, d’accélérer pour le parc en exploitation certaines modifications ou remises en conformité.

Ce travail de grande qualité effectué lors des ECS a démontré la capacité d’EDF à maîtriser les bases de conception de ses installations. J’observe également que ce travail a permis aux sites de se réapproprier

leurs référentiels de conception. C’est une très bonne chose au moment où le renouvellement de génération risque de faire disparaître une partie de la mémoire.

Enfin, à la demande du HCTISN, l’ECS comprend une troisième partie, spécifiquement française, sur les conditions de recours aux entreprises prestataires dans les centrales nucléaires en exploitation. J'aborde ce volet au chapitre 9.

Une amélioration continue

Depuis leur mise en service, les centrales du parc nucléaire d’EDF font l’objet de nombreuses évolutions et modifications destinées à améliorer leur niveau de sûreté, notamment lors des visites décennales. Ces évolutions sont réalisées selon un principe d’amélioration continue de la sûreté, en vigueur en France depuis l’origine du parc nucléaire et codifié par la loi TSN depuis 2006.

ECS : une sûreté confirmée.

Les ECS ont permis à EDF de porter un avis positif sur le dimensionnement actuel des installations, avis largement lié aux réévaluations successives déjà réalisées lors des réexamens décennaux de sûreté du parc, et aux dispositions prévues dès la phase de conception pour l’EPR. Je confirme, pour ma part, cet avis.

Je note également que, pour les accidents graves extrêmes, ces analyses confirment l’apport que constituent, pour la protection des populations vis-à-vis d’éventuels rejets radioactifs, les moyens mis en place suite aux accidents de TMI et de Tchernobyl.

Les parades proposées

Robustesse et résilience.

Jusqu’à mars 2011, des situations aussi extrêmes que celles étudiées n’avaient pas fait l’objet de recherche de parades spécifiques compte tenu de leur caractère peu plausible. L’étude de parades supplémentaires est donc proposée pour renforcer

la robustesse des installations et leur résilience.

Pour la robustesse, des dispositifs supplémentaires, comme le renforcement de sources électriques et d’appoint en eau, sont retenus. Ces dispositifs font partie du « noyau dur » capable de résister à des agressions hors dimensionnement. Ce noyau dur est une avancée majeure mais j’attire l’attention sur la nécessité de limiter son périmètre pour ne pas trop compliquer les installations.

Page 36: EDFMESS FR - inis.iaea.org

34

4

Pour la résilience, l’objectif est de disposer de moyens de crise locaux et nationaux, matériels et humains, capables de répondre avec flexibilité à des situations très diverses et imprévues. La création de la FARN, Force d’action rapide nucléaire, opérationnelle sur site en moins de 24 heures, répond à cette préoccupation. Il est envisagé de la gréer à terme avec environ 300 personnes réparties sur plusieurs sites et qui conserveront leurs métiers propres à temps partiel.

J’adhère pleinement aux avancées proposées dans ces deux domaines et j'approuve la priorité d’EDF de mettre en œuvre ces parades en cohérence avec les réexamens décennaux. Mais leur application doit tenir compte des réalités sur le terrain. La problématique des compétences, qui se pose aujourd’hui de manière aiguë, notamment pour le personnel de conduite, et la nécessité de faire face à l’augmentation de charge de l’ingénierie me conduisent à conseiller une grande vigilance dans les conditions de

mise en œuvre de ces mesures. Mon appréhension est de voir, par une surcharge difficilement supportable par les sites et les centres d’ingénierie, une gestion des priorités qui conduise à faire des impasses dans des domaines concernant la sûreté au quotidien. Il va sans dire que les ressources financières devront tenir compte de ces nouvelles mesures. La disponibilité d’un tissu de fournisseurs industriels de qualité sera aussi une clé de la réalisation de ce plan d’actions.

La gestion de crise

La gestion d’un accident grave repose sur trois pôles de responsabilité clairement définis : celui de l’exploitant, celui de l’autorité de sûreté et celui des pouvoirs publics, répartis tous trois sur deux niveaux, nationaux et locaux.

A EDF, les moyens de gestion de crise au plan national ont toujours fait l’objet d’une attention particulière et l’organisation de crise a confirmé, lors d’incidents ou d’exercices, sa capacité d’appui aux sites en situation incidentelle. Il faudra cependant examiner, aux niveaux local et national et pour l’ensemble des parties prenantes, les conséquences d’un événement affectant plusieurs réacteurs sur un même site. Les conditions d’intervention des personnels en situation de crise ont également fait l’objet d’une réflexion et il faudra prendre en compte le retour d’expérience de Fukushima sur le contrôle et la protection radiologique des intervenants et leurs conditions de vie dans les locaux de crise.

Pour les pouvoirs publics, j’attire l’attention, à la lumière de l’accident japonais et de l’importance des populations déplacées, sur la formation des décideurs publics à la gestion de crise nucléaire et leur entraînement qui pourraient, par exemple, être réalisés sur un simulateur de gestion de crise.

Le niveau de sûreté, l’avis de l’ASN

Dans son avis du 3 janvier, à l’issue des ECS, l’ASN réaffirme le rôle de l’exploitant, premier responsable de la sûreté de ces installations. L’ASN assurant le contrôle au nom de l’Etat.

Elle « considère que les installations présentent un niveau de sûreté suffisant pour qu’elle ne demande l’arrêt immédiat d’aucune d’entre elles. Dans le même temps, elle considère que la poursuite de leur exploitation nécessite d’augmenter dans les meilleurs délais, au-delà des marges de sûreté dont elles disposent déjà, leur robustesse face à des situations externes ».

L’ASN va donc imposer aux exploitants un ensemble de dispositions et renforcer les exigences de sûreté relatives à la prévention des risques naturels (séisme et inondation). Elle souligne en particulier les propositions d’EDF pour la définition du « noyau dur » de dispositions matérielles et organisationnelles pour faire face aux situations extrêmes et pour la mise en place progressive, à partir de 2012, de la Force d’action rapide nucléaire.

Page 37: EDFMESS FR - inis.iaea.org

35

4

Elle note que les facteurs sociaux, organisationnels et humains ont fait l’objet d’une attention particulière et retient les priorités sur le renouvellement des effectifs et l’organisation du recours à la sous-traitance.

Elle demande de renforcer le processus de traitement des non-conformités et les référentiels sur les aspects des risques liés aux autres activités industrielles.

Au Royaume-Uni

Le board d’EDF Energy Nuclear Generation a agi rapidement pour entreprendre des évaluations dans les centrales, à la suite d'une réunion tenue dès le 14 mars. Cette démarche a été renforcée par la publication de la SOER 2011-02 de WANO, le 18 mars. En parallèle, le gouvernement mandate l'inspecteur en chef de l'ONR pour mener une évaluation de sûreté, avec un premier rapport en mai et un rapport final en septembre 2011.

La méthode de travail

Après l'accident de Fukushima, EDF Energy au Royaume-Uni et EDF SA en France ont maintenu

une communication étroite et continue sur les méthodes d'évaluation et les hypothèses retenues. Cette démarche garantit une approche cohérente dans le Groupe.

En parallèle aux efforts nationaux, la Commission européenne, via l’European Nuclear Safety Regulators Group (ENSREG), a élaboré un document indiquant comment les stress tests devaient être menés dans l'Union. Il y a de nombreux points communs entre les rapports de l’ONR et la spécification de l'ENSREG, l'ONR a cependant étendu son champ d'application à la gestion post-accidentelle, ainsi qu’à l'ouverture et à la transparence de l'industrie nucléaire.

L’originalité de la démarche de l'ONR est d’assurer le retour d’expérience de l’accident en utilisant les pratiques courantes et les processus réglementaires existants.

Des réacteurs de types différents

Des réacteurs avec une grande

inertie.

Il convient de noter que, par conception, les réacteurs AGR réagissent

différemment des REP dans les scenarii d'accidents graves. Les réacteurs AGR ont une densité de puissance environ 40 fois inférieure à celle des REP et une inertie thermique importante. En cas de perte de refroidissement, leur montée en température est beaucoup plus lente. De plus, les gaines du combustible sont

en acier inoxydable et ne produisent pas d‘hydrogène à très haute température en présence de vapeur d’eau comme c’est le cas pour les REP dont les gaines contiennent du zirconium.

Sizewell B, réacteur REP de conception récente, a été évalué de manière cohérente avec le parc d'EDF en France.

Le niveau de sûreté, l’avis de l’ONR

Dans le cadre de cette évaluation globale de sûreté, EDF Energy, dans ses rapports de sûreté, n’a pas identifié de manque pour chacune de ses centrales nucléaires.

Les rapports (intermédiaire et final) de l’ONR indiquent : « en considérant les causes directes de l'accident de Fukushima, nous (ONR) ne voyons aucune raison pour restreindre l'exploitation des centrales nucléaires au Royaume-Uni » et « l'industrie nucléaire britannique a réagi de façon responsable et appropriée, affichant un leadership en matière de sécurité et une solide culture de sûreté nucléaire dans sa réponse à ce jour ». Les conclusions de ces rapports confirment

Page 38: EDFMESS FR - inis.iaea.org

36

4

que l'approche britannique pour identifier les référentiels de conception des installations nucléaires est solide pour les agressions externes. L’exigence d'effectuer des réexamens périodiques des dossiers de sûreté et de les soumettre à l'ONR offre un moyen robuste de s'assurer que les installations nucléaires sont suffisamment améliorées en fonction des progrès technologiques. Dans le cas contraire, elles seraient mises à l’arrêt.

Les changements en cours d’étude

Des modifications techniques des installations sont envisagées pour améliorer la résistance aux agressions naturelles externes, dont la possibilité de connexion rapide de matériels de sauvegarde. Le

stockage hors site d’équipements de sauvegarde sera renforcé pour permettre aux équipes de secours d’intervenir plus rapidement.

Une gestion de crise renforcée.

Les moyens des centres régionaux de gestion de crise seront renforcés et pourront être utilisés pour tout type de crise, nucléaire ou non.

Dans le domaine de l’organisation de crise, la formation, les procédures d’accidents graves, les exercices de crise et les moyens de communication seront améliorés.

Pour l'ouverture et la transparence, la création de centres d’accueil du public dans chaque site est par exemple envisagée.

Aux Etats-Unis

Aux Etats-Unis, la réaction de l'industrie nucléaire a été bien organisée par l’INPO, le NEI et l’EPRI44, avec la création d'un comité de pilotage de la réponse à Fukushima, chargé de coordonner les efforts de sept groupes de travail.

Comme ailleurs dans le monde, les exploitants ont tous répondu aux SOERs de WANO et à leur équivalent de l’INPO, et ont tenu les revues de site exigées.Chacun s'accorde à reconnaître que le NEI, avec l'appui des acteurs industriels, a recueilli et mis à disposition les informations relatives à l'accident de façon très efficace pour répondre aux besoins du

public, des médias et des politiques.

La NRC a créé une "near term task force" composée d’experts missionnés pour réexaminer ses processus et réglementations aux États-Unis puis formuler des recommandations pour l’adaptation éventuelle de sa politique.Dans mes échanges avec l’INPO, un sujet est revenu comme un leitmotiv : après un accident nucléaire important dans n’importe quel pays, tous les grands organismes nationaux et internationaux

sont appelés à travailler ensemble et sous la pression des événements. Ils doivent s’appuyer sur des relations et des réseaux déjà bien rodés. En amont de tout incident, les responsables des différents organismes doivent se connaître et savoir quel rôle ils seraient appelés à jouer.En matière de gestion de crise, il me semble que les sites américains sont bien organisés avec leur administration locale mais, au niveau fédéral, la FEMA45 ne me paraît pas dotée d’une organisation de crise nucléaire robuste.

44 EPRI : Electric Power Research Institute.45 FEMA : Federal Emergency Management Authority.

Page 39: EDFMESS FR - inis.iaea.org

37

4

Constellation Energy Nuclear Group (CENG)

Un exploitant très actif.

CENG est très engagé dans les actions post-Fukushima de l’industrie américaine. Son CNO pilote un des sept groupes de travail mentionnés précédemment et représente CENG au comité de pilotage. De plus, CENG coordonne la réponse à l’IER 11-446 relative à la perte totale des alimentations électriques. En interne,

CENG a lancé des actions relatives à :

l’évaluation de la tenue au séisme et l’étude des marges à l’inondation,

l’autonomie des batteries, l’amélioration de l’instrumentation piscine, le renforcement des lignes d’éventage des REB, la mise à niveau des guides d’accidents graves,

l’achat de matériels mobiles et les études de connexions rapides,

les évaluations sur sites de novembre 2011 à avril 2012 et la réception des matériels mobiles au premier semestre 2012.

Des réacteurs de types différents

CENG exploite deux réacteurs à eau bouillante (REB) à Nine Mile Point, sur les rives du lac Ontario. La conception de base du réacteur 1 de Nine Mile Point (NMP) est très similaire à celle des premières unités de Fukushima (réacteur de type Mark-One).Sur leurs deux autres sites, Ginna et Calvert Cliffs, CENG exploite trois réacteurs à eau pressurisée (REP).

Le niveau de sûreté, l’avis de la NRC

En juillet 2011, les experts de la NRC47 réunis dans le Near Term Task Force ont rendu un rapport avec douze thèmes de travail et une trentaine de recommandations et de prescriptions. Ils ont déclaré les centrales américaines « sûres et ne nécessitant aucune mise à l’arrêt ». Je constate que ces experts expriment le besoin d’une clarification des exigences relatives aux « événements de dimensionnement » et aux « événements hors dimensionnement ». De plus, ils signalent une difficulté de cohérence des initiatives volontaires de l'industrie qui échappent à la rigueur du contrôle continu de la NRC.

Ces recommandations rendues publiques en juillet 2011 ont été classées par la NRC en octobre 2011, par la lettre SECY11-137, en trois lots :

court terme : évaluations sismiques et inondation, perte des alimentations électriques, fiabilisation des lignes d’éventage des réacteurs à eau bouillante, renforcement des procédures de gestion des accidents graves,

moyen terme : capacité d’appoint aux piscines de stockage du combustible, organisation de crise,

long terme : capacité d’éventage des réacteurs à eau pressurisée, maîtrise du risque hydrogène.

Les analyses et les initiatives des autres acteurs internationaux

La communauté internationale de sûreté nucléaire (Global Nuclear Safety Regime)

Ce terme a été inventé par l’INSAG48, groupe consultatif international pour la sûreté nucléaire créé par l’AIEA. Il décrit l'initiative collective internationale visant à définir le niveau de performance attendu de tous les exploitants et autorités réglementaires, à examiner leur performance ainsi qu’à renforcer leurs 46 IER : INPO Event Report.47 NRC : Nuclear Regulatory Commission48 INSAG : International Nuclear Safety Advisory Group.

Page 40: EDFMESS FR - inis.iaea.org

38

4

compétences et leurs capacités. Cette communauté, qui rassemble l’ensemble des initiatives

des réseaux nationaux et internationaux, est décrite en détail par le rapport INSAG-21. Alors que tous les acteurs concernés s’attachaient à comprendre les événements au fil de leur déroulement à Fukushima Daiichi, plusieurs mesures importantes ont été prises au niveau mondial pour renforcer cette communauté internationale. Je pense qu'il est tout à fait essentiel que les acteurs clés de l'industrie nucléaire, notamment EDF, s’impliquent plus activement, par une contribution effective et continue, dans la vie de cette communauté.

WANO

WANO engage sa crédibilité.

Je note que l’Association mondiale des opérateurs nucléaires (WANO) a réagi promptement à l'accident, en étroite liaison avec l’INPO, et veillé à ce que tous ses membres contrôlent rapidement l'état de leurs centrales ainsi que leur niveau de préparation aux accidents et aux crises. Cet exercice s'est révélé utile, mettant en

lumière plusieurs domaines à améliorer tels que la formation, la documentation, la maintenance et le contrôle des équipements de secours.En outre, une commission de 14 membres a été créée afin de recommander des modifications des programmes et du fonctionnement de WANO. Lors de l'Assemblée générale biennale de WANO à Shenzhen en octobre 2011, les membres de WANO ont voté à l'unanimité les recommandations. Elles portent sur :

l’élargissement du champ d’activité de WANO,

l’élaboration d'une stratégie mondiale intégrée de réponse aux événements,

l’amélioration de la crédibilité de WANO, notamment en modifiant en profondeur le processus des revues de pairs,

le renforcement de la visibilité de l’association,

l’amélioration de la qualité de l'ensemble des produits et services proposés.

Une nette augmentation des ressources sera nécessaire. En particulier, les membres devront mettre à la disposition de l'association et de ses programmes des moyens et des personnes compétentes. Cette contribution restera toujours faible face aux coûts directs induits par l'accident de Fukushima et à son impact sur la confiance du public. Le groupe EDF a soutenu activement cette commission et ses recommandations, qui amélioreront la crédibilité et le pouvoir de WANO.Je pense que la responsabilité de tous les opérateurs nucléaires sera de veiller à la bonne exécution de ce programme, dans les délais et en qualité.

L’AIEA

Fin mai 2011, une mission d'experts internationaux de l’AIEA regroupant des représentants de 12 pays s'est rendue au Japon. Ils ont visité les centrales nucléaires de Fukushima Daiichi, Fukushima Daini et Tokai Daini. Leurs constatations viennent corroborer les enseignements cités précédemment.

Page 41: EDFMESS FR - inis.iaea.org

39

4

Certains points supplémentaires ont été dégagés par cette mission :

la réponse apportée sur place, dans des conditions extrêmement difficiles, par des personnels spécialisés à la fois dévoués et déterminés, s'est révélée exemplaire. Elle a permis d'adopter l'approche la plus adaptée pour garantir la sûreté, eu égard au caractère exceptionnel de la situation,

un soutien logistique très professionnel a été créé, en particulier les dispositions prises au « village J » (centre d’accès au site, situé à 20 km) pour assurer la protection des intervenants appelés à se rendre sur les sites,

le plan d’actions, prévu pour la reprise de contrôle des réacteurs accidentés, est un élément important et reconnu. Il devrait s'inscrire dans un plan plus vaste pouvant aboutir à la décontamination des zones hors sites touchées par les rejets radioactifs, afin de permettre aux populations évacuées de reprendre une vie normale. Le monde entier serait ainsi témoin des mesures qui peuvent être prises pour répondre à ce type d'événements nucléaires extrêmes,

la réglementation nucléaire doit être conçue pour prendre en considération des agressions externes extrêmes et faire l’objet de réexamens périodiques. Elle devra aussi veiller à préserver

l'indépendance de l’autorité de sûreté et la clarté des rôles de chacun en toutes

circonstances, conformément aux normes de l’AIEA.

Une somme considérable d’informations a été recueillie lors de cette mission. J’ai été frappé d’apprendre que sur le site voisin de Fukushima Daini, où l'inondation a causé des dégâts moins importants qu’à Fukushima Daiichi, le personnel de TEPCO et celui des entreprises prestataires ont installé plus de 9 km de câbles électriques provisoires en 16 heures, ce qui a permis de retrouver des fonctions de sûreté.

En conclusion, comme TMI et Tchernobyl, Fukushima est un événement qui façonne l’avenir de l’industrie nucléaire.

Je considère que la réaction des exploitants nucléaires du groupe EDF a été rapide et proportionnée, tant par des initiatives promptes que par la qualité des réponses aux demandes de leurs autorités de sûreté respectives.

J’estime que ce travail sans précédent de remise en question a confirmé que la sûreté des réacteurs du groupe EDF est au niveau attendu du référentiel actuel. La prise en compte d’agressions naturelles encore plus violentes augmentera leur niveau de sûreté.

J’attire toutefois l’attention sur les capacités d’ingénierie qui doivent être mobilisées sur ces études et travaux, à un moment où, en France en particulier, les équipes d’ingénierie sont déjà fortement mobilisées par d’autres exigences de sûreté. Elles doivent, de plus, faire face à un important renouvellement de leur personnel.

L’organisation de crise nucléaire dans les pays d’implantation du Groupe, associant les pouvoirs publics, les opérateurs et les autorités de sûreté, mérite d’être réinterrogée pour évaluer sa capacité à faire face à un accident du type Fukushima, mettant plusieurs réacteurs d’un même site en péril et entraînant un important déplacement de populations. Les Etats et leurs services devront prendre toute leur place dans ce travail que j’appelle de mes vœux.

Je considère que les décisions prises en octobre 2011 par les opérateurs nucléaires du monde entier, regroupés au sein de WANO, sont nécessaires car elles renforcent très nettement le contrôle

international des activités des exploitants nucléaires. Je me félicite qu’EDF y ait pris toute sa part.

Page 42: EDFMESS FR - inis.iaea.org
Page 43: EDFMESS FR - inis.iaea.org

41

5

LA FORMATION AU SERVICE DES

COMPETENCES DANS LES METIERS

L’entreprise fait face en France à un renouvellement sans

précédent de générations. La préparation du Grand carénage

du parc d’EDF SA va mobiliser rapidement de nombreuses

compétences. La maîtrise en toute sûreté de ces défis passe par

l’engagement des managers sur la formation et les métiers.

La problématique est étudiée dans ce chapitre sur son volet français. Cette année, des recrutements

importants se sont poursuivis avec l’embauche de près de 2 000 techniciens et ingénieurs dans le secteur nucléaire. Ils vont compenser en premier lieu les départs en retraite qui se profilent, 40 % du personnel DPN-DIN devant partir entre 2010 et 2015, le pic maximum des départs étant en 2014. Ces départs sont la traduction du gréement massif pour la construction du parc nucléaire à la fin des années 70 et au début des années 80 : ils ont été insuffisamment anticipés dans la décennie précédente.

Ces recrutements auront aussi à faire face à la charge de travail due au Grand carénage et aux modifications issues du retour d'expérience de l’accident de Fukushima. Les temps de formation et de professionnalisation dans le nucléaire étant longs, il est nécessaire de constituer des pépinières pour anticiper les départs.

Page 44: EDFMESS FR - inis.iaea.org

42

5

Pour faire face à ces défis, la DPN et la DIN ont lancé un programme « compétences » très ambitieux. J’observe que ce programme part d’une analyse fine de l’état des lieux et s’inspire des bonnes pratiques internationales.

Des compétences

rares.

J’insiste sur le renouvellement des experts qui, même s’il est faible en nombre, est vital pour l’avenir de l’entreprise. Il doit être anticipé domaine par domaine, y compris à la division du combustible et à la R&D. Sur les sites, il en est de même pour les compétences rares qui demandent de l’anticipation compte tenu de l’expérience professionnelle nécessaire pour ces emplois.

Les compétences

La vision de terrain

L’effort de rationalisation et de productivité, conduit dans les années 2000, se rattrape aujourd'hui difficilement. Les temps de formation et surtout de professionnalisation étant longs, le renouvellement

des compétences, malgré un effort sans précédent, reste la préoccupation majeure des managers des CNPE.

A la conduite, où les temps de professionnalisation sont les plus importants, l’expérience collective diminue. Alors que l’on comptait 7 expérimentés pour 1 jeune en 2007, en 2012 il n'y aura plus que 2 expérimentés pour 1 jeune, la situation s'améliorant les années suivantes. Ce constat appelle des évolutions dans l’organisation des activités et du contrôle au sein des équipes.

Dans les métiers de la maintenance, la problématique est différente mais reste réelle avec un expérimenté pour un jeune en 2012. Cet afflux de jeunes salariés pose la question de la cohabitation de plusieurs générations. La relation au travail des plus jeunes est différente, il faut que l’entreprise en tienne compte.

L’ingénierie est sans doute le domaine où le manque de ressources expérimentées se fait

davantage sentir, alors que le projet EPR est dans sa phase la plus active, que l’IPE49 doit assurer la performance des 58 tranches et que l'instruction des suites de l’accident de Fukushima mobilise près de 300 ingénieurs à la DIN et à la R&D. S’y ajoutent les ressources nécessaires à la déconstruction. Face à cette charge de travail importante et croissante, il faut sans doute réfléchir à une manière de travailler différente. Pourquoi ne pas rechercher et favoriser le renfort d’autres ingénieries ?

A la division combustible, une filière métier est développée pour compenser d’ici à la fin de la décennie de nombreux départs en retraite, fragilisant certaines compétences d’expertise. La création de la communauté combustible en 2010 favorise la gestion transversale des compétences de toute la filière, réparties dans les différentes directions de l’entreprise. Dans ce domaine, les recrutements annuels, d’une dizaine de personnes, semblent suffisants pour les responsabilités de la DCN.

L’ingénierie, une priorité

Les besoins en compétences de l'ingénierie, à la DIN et à la DPN, pour les dix prochaines années s’établissent dans une fourchette de 6000 à 6500 ingénieurs.

Renforcer les compétences de

l’ingénierie.

En 2011, plus de 400 nouveaux arrivants ont rejoint la DIN, tendance qui sera maintenue au cours des trois prochaines années. Si la majorité des besoins est couverte par l’embauche de jeunes diplômés, les recrutements avec expérience professionnelle sont particulièrement recherchés. J’attire l’attention sur des spécialités très concurrentielles (génie civil, chimie, chef de projet) ou spécifiques

49 IPE : Ingénierie du Parc en Exploitation.

Page 45: EDFMESS FR - inis.iaea.org

43

5

(contrôle-commande, radiochimie). Je ne peux qu’encourager ce type de recrutement. Et je souligne l’effort particulier du CEIDRE50 pour recruter des chimistes de grande qualité.

J’ai souligné, en 2010, la création d’une pépinière de jeunes, formés par la R&D pour une durée de deux à trois ans et destinés aux métiers de la production et de l’ingénierie. Ces embauches communes DPI-R&D, même limitées, sont une excellente initiative.

La mobilité

Je constate que le marché interne de l’emploi n’a pas retrouvé sa fluidité, même si je salue le travail engagé sur les parcours professionnels inter-divisions de la DPI. Devant le manque de compétences expérimentées, les directeurs d’unité continuent, dans leur majorité, à préférer conserver leurs agents. Je ne peux qu’encourager la politique des ressources humaines à prendre en compte les spécificités de la jeune génération, qui ne répond pas aux mêmes attentes que les anciens. Je me félicite de la mise en place des CMPP51 dans chaque unité de la DPN et de la DIN. Je m’attacherai à les rencontrer à l’occasion de mes prochaines visites.

Je note aussi les hésitations, pour ne pas dire les réticences, de la maîtrise à franchir le pas pour devenir cadre. La perception des missions et des contraintes des emplois de cadre et la perte temporaire d’avantages spécifiques en sont les principales explications. La mobilité fonctionnelle progresse insuffisamment et l’ascenseur social se grippe quant aux emplois opérationnels et aux emplois cadres.

Les nombreux départs en retraite sont une source de perte d’expérience. Je m'interroge sur la manière de résoudre les contraintes liées à ces flux. Ne pourrait-on pas trouver un moyen plus attractif pour garder le personnel qui le souhaite, ne serait-ce qu’à temps partiel. Toute l’imagination a-t-elle été mise au service de cette cause ?

Les nouveaux défis après Fukushima

Gérer un accident

multi tranche.

L’accident de Fukushima aura sur notre parc nucléaire des incidences dans le domaine des ressources humaines. Les ECS révèlent le besoin de compétences complémentaires pour permettre de réaliser des interventions minimales dans des situations complexes exceptionnelles. Jusqu’à présent, la prise en compte d’un accident sur une seule tranche permettait un renfort par les autres tranches du

site. L’hypothèse déterministe d’un accident sur toutes les tranches d’un même site remet en cause ce renfort mutuel. Ce renforcement demandera des ressources de conduite expérimentées pour les prochaines années.

Il faudra aussi compter sur la création de la FARN, Force d’action rapide nucléaire, qui sera gréée par du personnel de la DPN : environ 360 personnes, détachées pour 35 % de leur temps dans cette nouvelle activité. Je m’attacherai à suivre la mise en place de ce nouveau dispositif.

De plus, les décisions prises par les exploitants réunis au Biennal General Meeting de WANO, en octobre 2011, conduisent à renforcer le personnel détaché à cette organisation d’environ 200 ingénieurs expérimentés. Pour le groupe EDF, ce sont sans doute 46 cadres expérimentés qu’il faudra mettre à disposition du centre de WANO Paris. J’insiste sur la qualité du personnel qui sera détaché et sur la mise en œuvre de ces décisions qui me semblent essentielles pour la sûreté et l’avenir de l’industrie nucléaire.

50 CEIDRE : Centre d’Expertise et d’Inspection dans les Domaines de la Réalisation et de l’Exploitation.51 CMPP : Chef de Mission Parcours Professionnels.

Page 46: EDFMESS FR - inis.iaea.org

44

5

La coopération avec EDF Energy

Une coopération s’est engagée avec EDF Energy pour réaliser des recrutements en commun au Royaume-Uni. Une vingtaine d’ingénieurs ont ainsi rejoint la DIN cet automne pour un détachement de deux ans avant de rejoindre EDF Energy.

Des échanges de personnel se développent également entre EDF Energy et les divisions nucléaires de la DPI. J’encourage ces échanges qui devraient monter en puissance avec le projet EPR UK.

La formation

Les besoins en formation ont plus que doublé entre 2008 et 2011. Cette tendance à la hausse se confirme. Aujourd’hui, les possibilités de recrutement sont limitées par la capacité d’accompagnement, de formation et de professionnalisation des jeunes embauchés. Face à de tels volumes, la qualité de la

formation est une clé de la réussite du renouvellement des compétences.

Les moyens ont été considérablement renforcés avec la création des académies de métier production nucléaire et ingénierie nucléaire. Les académies production nucléaire assurent plus de 700 000 heures-stagiaires, pour moitié par des formateurs et pour moitié par des opérationnels des sites.

Les moyens de formation

Les académies des métiers servent avant tout à la formation initiale des agents qui rejoignent les directions nucléaires. Elles sont toujours autant appréciées par les agents et leur hiérarchie. Il faut poursuivre. Je salue la création d'académies des métiers dites des « savoirs spécifiques », prenant le relais des académies des « savoirs communs », dans des domaines de plus en plus étendus.Je note également avec intérêt la création d’une formation destinée aux directeurs d’unités nouvellement nommés à la DPN. Elle complète l’éventail des formations destinées aux managers, des MPL aux chefs de services et aux chefs de projets arrêts de tranche. Sur le site de Torness, en Ecosse, j’ai observé une formation au leadership qui m’a semblé novatrice et que je recommande.

Je pense qu’une attention doit être portée à la formation dite « juste à temps ». L’entraînement avant une opération délicate est toujours très bénéfique pour une bonne maîtrise de la sûreté. J’encourage ce type de formation proche du processus opérationnel, en particulier sur simulateurs. Le développement de maquettes spécifiques, par type de matériels, répond également à ce type de formation. Lors de mes visites, j'ai vu de très bonnes initiatives en ce sens, qui mériteraient d’être mieux connues.Les chantiers écoles des sites sont d'excellents supports pour réaliser des formations plus générales ou ciblées sur les performances humaines. J’invite les sites à continuer à bien utiliser ces chantiers écoles, pour leur personnel et celui des entreprises prestataires. Pour la conduite, les simulateurs "pleine échelle" sont des outils incontournables d'apprentissage. Il convient de les faire évoluer pour être les plus représentatifs de chacune des salles de commande opérationnelles. Je note que le taux de charge des simulateurs est de plus en plus élevé, ce dont je me félicite. J’attire toutefois l’attention sur la situation du CNPE de Gravelines qui, avec six tranches, devrait disposer d'un deuxième simulateur.J’insiste pour que les services de conduite s’organisent mieux pour permettre aux opérateurs de réaliser des formations à la demande. A l’étranger, le passage sur simulateur est plus systématique et mieux organisé car intégré au planning des équipes.

De nouveau, je rappelle la nécessité de disposer de plus de formateurs expérimentés. A l’étranger, les centres de formation des centrales sont souvent deux à trois fois mieux gréés que ceux des sites français. Au-delà de cet effort important, il convient de valoriser davantage ce métier pour attirer les meilleurs professionnels et leur assurer un bon retour en exploitation. J’ai vu sur un site britannique une pratique de détachements de courte durée, entre six mois et un an, qui paraît donner satisfaction.

Page 47: EDFMESS FR - inis.iaea.org

45

5

Une plus grande autonomie

La formation à la main des

managers.

Au-delà des moyens, c’est sans doute la responsabilisation des managers qui peut faire l’objet de progrès. Dans mes visites sur les sites, j’ai observé trop souvent des managers insuffisamment impliqués dans l’entraînement de leurs équipes. Je suggère que les services s’approprient davantage la formation et que les managers puissent dire ce qui fait défaut dans l’exercice des fonctions et emplois

de leurs services, et pour pouvoir adapter les formations en conséquence. L’engagement sur les compétences passe aussi par la définition des besoins en formation, par l’appropriation des programmes et par la présence des managers sur le terrain de la formation.

La question de l’autonomie locale dans le domaine de la formation a été soulevée avec le besoin d’un rapprochement entre l’UFPI52, maître d’œuvre de la formation, et les sites nucléaires. J’ai noté des expérimentations conduites sur les sites de Golfech, Civaux, Chooz et Tricastin et le projet d’extension à six autres CNPE en 2012. Je me suis fait présenter celles-ci sur les deux premiers sites. Je suis

convaincu qu’une certaine autonomie est bénéfique car elle responsabilise les managers dans le domaine de la formation. L’UFPI doit contribuer à ce changement important. Une telle manière de faire ne conduira au succès que si toutes les parties prenantes sont acquises à cette autonomie des sites.

Le Systematic Approach to Training, SAT

On m’a présenté cette année, lors de mes visites à Koeberg, à Sizewell et à l’INPO, la démarche SAT, reconnue au plan international dans de nombreuses compagnies. Elle définit la formation à partir des besoins réels de compétences des services, sur la base d’une description des emplois portée par le management du site, propriétaire de la formation et de ses priorités. La centrale de Koeberg, en Afrique du Sud, a fait ce choix. J'apprécie également la méthode pragmatique et préservant l’esprit de la démarche, utilisée par le site de Sizewell, pour aller vers l’accréditation totale de son service formation d’ici fin 2012, comme les autres sites d’EDF Energy.

Je note que la DPN ne s’est pas fixé à ce stade d’objectif d’accréditation, privilégiant la réussite de la démarche de fond engagée selon le référentiel de l’INPO.

Une formation au plus près du geste technique.

Pour tous les métiers techniques de la DPN, un référentiel est en cours d’élaboration selon ce processus, en partant de l’analyse des activités à réaliser par les métiers. Les experts des métiers et de l’UFPI envisagent de définir plus de 1 500 stages de maîtrise d’œuvre UFPI pour la DPN. Mon souhait, à travers ce travail qui se poursuivra jusqu’en 2015, est de voir se développer des

formations moins théoriques, plus ciblées sur les métiers.

La formation EPR

La formation des équipes a commencé en 2008, avec une bonne anticipation. Le simulateur de la salle de commande est déjà en service, ce qui est important quand 50 % du personnel de la conduite est constitué de jeunes embauchés et que la salle de commande est totalement informatisée. La préparation de la formation se fait, là aussi, dans le cadre de l’approche SAT, ce qui est un gage de qualité au moment où il faudra assurer un million d’heures-stagiaires sur toute la période précédant la mise en exploitation. En Chine, je note que cette problématique du manque d’expérience à la conduite est intelligemment compensée, avant de passer sur l’EPR, par l’habilitation de tous les opérateurs sur les réacteurs chinois dont la salle de commande est aussi informatisée, les CPR 1000.

52 UFPI : Unité de Formation Production Ingénierie.

Page 48: EDFMESS FR - inis.iaea.org

46

5

L’apprentissage et le tutorat

J’observe cette année une inquiétude, très compréhensible, des plus expérimentés sur la manière de transmettre leur expérience et leur savoir-faire avant de partir en retraite. Les choix de l’alternance et du tutorat par la DPN permettent de répondre en partie à cette question. Le tutorat est, à mon avis, un enjeu majeur pour la réussite de l’alternance. Le nombre, la qualité et la motivation des tuteurs constituent cependant un point de vigilance. Le flux de jeunes embauchés étant tellement important, j’ai rencontré parfois des tuteurs avec seulement trois ans d’ancienneté dans l’entreprise : c’est sans doute un peu juste.

Le lycée André-Malraux de Montereau

Le lycée de Montereau est « partenaire » du nucléaire d’EDF de longue date. Je salue ce partenariat choisi pour sa proximité avec le bassin nucléaire : Nogent, Belleville, Dampierre, St-Laurent, et pour la présence d’une équipe pédagogique dynamique et partie prenante d’un projet ambitieux. Je l'ai constaté lors de ma visite de cet établissement. Le chantier école "robinetterie" est particulièrement bien équipé.

Ce lycée assure une formation initiale bac professionnel « environnement nucléaire » et une formation continue à l’usage d’EDF et des entreprises prestataires du nucléaire. Les élèves diplômés trouvent tous un emploi en fin de formation. A côté des formations classiques de robinetiers et d’électroniciens, j’ai noté une académie des encadrants prestataires. En 2011, pour la première fois, un BTS « environnement nucléaire » a été développé en formation par apprentissage, pour former de futurs responsables de chantiers ou chargés d’affaires.

Je souhaite que l’Education nationale reconnaisse tout l’intérêt des formations relevant de besoins industriels et sanctionnées par des diplômes. Les relations du lycée avec EDF sont excellentes et l’entreprise doit maintenir le soutien apporté à ce lycée, notamment par la poursuite de la coopération avec le CNPE de Nogent et la fourniture de matériel.

En conclusion, pour permettre à l’exploitation et à l’ingénierie de compenser les départs et d’assurer leur charge de travail, il faut continuer à rapprocher la formation du terrain et lui donner les moyens nécessaires. L’intégration des meilleures pratiques internationales est très prometteuse. Le succès reposera sur l’implication forte des équipes de direction dans la formation.

Page 49: EDFMESS FR - inis.iaea.org

47

6

LE METIER DE CONDUITE

Les services Conduite concentrent beaucoup d’enjeux majeurs

de sûreté nucléaire. La maîtrise de ces activités « cœur de

métier » y reste perfectible. Le renforcement de la formation

des équipes et une plus grande affirmation managériale des

chefs d’exploitation seraient de nature à améliorer la qualité

d’exploitation et la sérénité.

Les résultats d’un site nucléaire, en sûreté comme en production, sont très étroitement liés à l’existence d’une force professionnelle de conduite, reconnue par les autres métiers et régulant les activités au quotidien avec autorité.La compétence de ses acteurs, la place qui doit lui être accordée par la direction des sites, la capacité à irriguer les autres métiers en compétences « process » sont autant de conditions de réussite que l’on devrait retrouver sur tous les sites du Groupe.

Un effort de recrutement considérable, des ressources humaines encore

fragiles

Dans le parc France, les métiers de conduite sont confrontés eux aussi à un très important renouvellement humain qui portera, en 2012, le taux d’encadrement dans les équipes à deux agents expérimentés environ pour un jeune débutant.

Page 50: EDFMESS FR - inis.iaea.org

48

6

Investir encore dans les

compétences.

Dans tous les sites visités cette année, je mesure l’ampleur des recrutements

réalisés et l’importance de l’effort des CNPE pour former et intégrer cette nouvelle génération. Je constate toutefois que, dans la plupart des sites, malgré cet effort sans précédent, les GPEC53 sont encore fragiles et tendues. Le gréement des différents emplois aux cibles affichées - techniciens, opérateurs,

têtes d’équipe - s’améliore sensiblement mais demandera encore du temps et de la continuité dans l’effort. D’autant que les compétences issues des métiers de conduite sont très attendues dans nombre d’autres services et manquent encore cruellement pour des emplois de la plus haute importance : instructeurs sur simulateurs, ingénieurs sûreté, chefs d’arrêts, responsables de sous-projets au sein des équipes d’arrêts de tranche et de tranche en marche, ingénierie.

Des progrès dans le respect du référentiel, des fragilités récurrentes dans la

réalisation des activités

Sur un point essentiel à mes yeux, je relève avec satisfaction que la maîtrise des spécifications

techniques d’exploitation (STE), « code de la route » de la conduite des réacteurs, est en progrès

sensible. Ce résultat est le fruit d’un important travail d’accompagnement et de formation réalisé sur de nombreux sites, en associant étroitement chefs d’exploitation et ingénieurs sûreté.Les progrès enregistrés sur la plupart des sites, en matière de condamnations administratives, méritent également d’être soulignés.

Le métier d’abord.

J’observe cependant une permanence des écarts de surveillance en salle de commande et une dégradation sensible des écarts de lignage et de configuration des circuits. Sur ce dernier point, les causes identifiées couvrent à peu près toute la gamme du savoir-faire attendu d’un exploitant de conduite,

illustrant bien une vraie question de compétences clés ou cœur de métier. Ces activités dites de terrain, réalisées en dehors de la salle de commande sur l’ensemble de l’installation, dans des conditions parfois difficiles, mériteraient à mon sens une plus grande attention de la part des managers, de conduite comme de direction. Une meilleure observation des pratiques, parfois mal connues, et des problèmes quotidiens rencontrés par ces techniciens serait de nature à inverser cette tendance.Je mesure également certaines difficultés dans l’application des procédures et des consignes d’exploitation en fonctionnement normal, traduisant parfois un manque de confiance dans des

documents modifiés récemment. Un très important travail de standardisation des procédures d’exploitation, dans le cadre du projet PHPM54, a en effet été mené ces dernières années, améliorant la prise en compte du retour d’expérience et rationalisant cette activité documentaire primordiale. Mais à chaque modification de consigne ou de gamme est très souvent associée une perte de repères. Ce point vaut aussi pour les activités de maintenance. Il est donc essentiel de veiller, pendant cette phase de transition, à assurer aux utilisateurs un accompagnement de qualité. Point positif, les chefs d’exploitation et les opérateurs ont toujours une grande confiance dans les procédures de conduite en situation incidentelle et accidentelle, procédures qu’ils testent régulièrement sur simulateur. Ce qui est un facteur de sérénité important.Enfin, je relève dans la plupart des sites visités les difficultés de réalisation des pratiques de

fiabilisation à la conduite, en salle de commande ou lors des activités sur le terrain. Un investissement dans la formation sans précédent a pourtant été réalisé dans les CNPE, tant sur les chantiers écoles que sur les simulateurs. Pour autant, à ce stade, le « déclic » ne s’est pas produit, et nous sommes encore loin de ce que j’ai observé ailleurs, en particulier dans les salles de commande d’EDF Energy.Devant le risque perceptible d’essoufflement des actions engagées dans le cadre du projet PPH55, projet novateur et désormais peu coûteux en ressources, j’incite le management local à prendre des

initiatives pour provoquer ce « déclic », en s’appuyant en particulier sur les jeunes générations.

53 GPEC : Gestion Prévisionnelle des Emplois et Compétences.54 PHPM : Projet Homogénéisation des Pratiques et des Méthodes.55 PPH : Projet Performances Humaines.

Page 51: EDFMESS FR - inis.iaea.org

49

6

Une autorité des chefs d’exploitation qui doit davantage s’affirmer

Les chefs d’exploitation assurent d’abord une mission essentielle de contrôle sûreté, 24 heures sur 24, par délégation directe du directeur d’unité. Ces acteurs clés doivent également s’assurer de la bonne marche de l’exploitation, en supervisant l’ensemble des activités engagées sur les tranches, en procédant aux arbitrages quotidiens et en intégrant les contraintes des services de maintenance. Ils managent tous une équipe de conduite et ont donc pour mission de porter les exigences de rigueur

d’exploitation auprès de chacun de leurs agents et de développer des collectifs de travail apprenants et ouverts sur l’externe.

A la lumière de mes visites, je m’interroge sur leur niveau de recul, tant je les vois trop souvent embarqués par le temps réel ou par des activités dont ils pourraient être déchargés. Les chefs d’exploitation que j’observe dans les centrales d’autres pays ou compagnies, particulièrement aux Etats-Unis et au Royaume-Uni, me paraissent beaucoup plus sereins et disponibles dans la gestion de leurs activités quotidiennes, dotés d’appuis nombreux et efficaces, et disposant du recul indispensable au bon exercice de cette fonction de ligne de défense.

Les actions engagées dans le parc en France pour mieux organiser le travail en production ou en arrêt de tranche, et inspirées des meilleures pratiques internationales comme l’AP 92856 ou le COPAT, vont dans le bon sens, en renforçant les appuis autour du chef d’exploitation. Je recommande de bien veiller à toujours lui faire toute sa place dans ces nouvelles organisations.

Des chefs d’exploitation

managers.

J’invite également les directions locales à accorder une grande attention à la dimension managériale de leurs chefs d’exploitation qui, pour réussir dans la conduite d’évolutions importantes de certaines pratiques professionnelles, ont besoin de voir leur autorité soutenue et leurs initiatives encouragées.

La formation et l’entraînement des équipes, une priorité absolue

La formation initiale des agents de conduite est de qualité, avec un cursus remarquable pour les conducteurs de réacteurs (opérateurs) et, c’est à relever, la généralisation sur les sites que j’ai visités d’académies « des savoirs spécifiques » pour les techniciens et les agents du terrain. Ces académies, co-animées par des techniciens très expérimentés, répondent au besoin et permettent une bonne

transmission des savoirs entre les générations.

J’observe avec satisfaction la mise en place généralisée d’une évaluation plus formelle des compétences des opérateurs sur simulateur, par une méthode pragmatique et réaliste. Cette évolution, inspirée des pratiques internationales, est de nature à donner confiance à l’externe, en particulier à l’autorité de sûreté.

La formation, une activité

à part entière.

Néanmoins, la formation et l’entraînement des équipes de conduite sont encore loin de la cible DPN des 15 jours d’entraînement sur simulateur par an et par opérateur, atteinte sur trop peu de sites. Elle se heurte à des problèmes

d’organisation du travail, voire de disponibilité des simulateurs sur les sites 4 et 6 tranches.

Je relève d’excellentes initiatives de terrain, planifiant par exemple avant chaque arrêt de tranche un entraînement dédié aux équipes concernées, à l’instar des pratiques des sites de CENG et d’EDF Energy. Cette pratique est encore trop peu répandue.Au vu de mes observations sur la plupart des sites visités, je regrette le nombre restreint de journées de formation réalisées en équipes constituées, en dehors du travail posté. Ce sont en effet de très bonnes occasions pour développer les collectifs de travail et rapprocher le manager de son équipe, collectifs

56 AP 928 : Pratique performante issue de l’INPO, définissant l’organisation du travail, en arrêt de tranche et tranche en fonctionnement.

Page 52: EDFMESS FR - inis.iaea.org

50

6

précieux pour faire face à toutes les conditions d’exploitation, normales ou incidentelles. Les réelles difficultés d’organisation du travail, qui rendent ces pratiques trop rares, méritent selon moi d’être réexaminées.

Plus généralement, je considère encore insuffisantes la présence et l’implication des managers dans la « vie du simulateur », chefs de service conduite et direction. Cette proximité, que l’on perçoit sur les sites d’EDF Energy par exemple, est un réel facteur de progrès et d’amélioration de l’efficacité de la formation.Je me réjouis que le programme compétences, construit par la DPN après un vrai travail d’ouverture sur l’international, soit désormais sur les rails. Il faudra lui accorder du temps et des moyens mais il affiche clairement l’ambition de responsabiliser davantage les managers locaux sur la formation.

Des évolutions des métiers à mener à terme

La définition d’objectifs ambitieux pour l’ensemble des métiers techniques a fait l’objet de projets métiers, engagés il y a plus de cinq ans par le parc nucléaire. Dans ce cadre, le noyau de cohérence des métiers conduite (NCC) a défini de nouveaux contours pour ces métiers, allant de la technicité accrue des agents de terrain de haute maîtrise, au renforcement des compétences en salle de commande, avec l’instauration du troisième opérateur et de l’opérateur pilote de tranche.

Le renforcement des têtes d’équipe autour des chefs d’exploitation, avec les chefs d’exploitation-délégués (CED) et des délégués sécurité en exploitation (DSE), est également un élément important de ces évolutions des métiers.

Force est de constater que ces évolutions ne sont que partiellement effectives dans les CNPE que j’ai visités, et se heurtent parfois encore à des réticences notables des acteurs concernés. Elles sont également freinées dans de nombreux services conduite par le rythme du gréement en compétences, et ce malgré des pépinières aujourd’hui bien remplies. Cela demandera tout simplement du temps et de la continuité dans l’engagement de ressources.

La salle de commande : un environnement de travail à sacraliser

S’il est un lieu sur un site nucléaire qui exige calme et sérénité, c’est bien la salle de commande. Je note avec satisfaction que des progrès importants ont été accomplis ces dernières années et j’observe, lors de mes visites, des initiatives concrètes prises sur le terrain et portées par une directive nationale qui fait sens. Toutefois, nous sommes encore loin des meilleures pratiques observables à l’étranger.Les accès ne sont pas toujours limités au strict nécessaire, en particulier lors des opérations d’arrêt de tranche, et l’autorité des acteurs de conduite ne semble pas toujours suffisamment affirmée, ou soutenue, pour faire respecter ces exigences. La réalisation des pratiques de fiabilisation dans les salles de commande qui, de facto, sacralisent les opérations qui s’y déroulent, est encore loin d’être effective.

La modernisation de l’instrumentation et des dispositifs utilisés en salle de commande, qui peut également constituer un facteur important de sérénité, ne fait pas l’objet à mon sens d’une priorité suffisante. A titre d’illustration, je regrette toujours l’absence de dispositif de consultation des données d’exploitation déporté en dehors des salles de commande et consultable par les services de maintenance, ce point étant particulièrement criant sur les tranches du palier 900 MWe.

A Flamanville 3, je constate avec satisfaction que ces questions font l’objet d’une bonne intégration dans l’ensemble du projet EPR, allant jusqu’à remettre en question l’organisation des activités en salle de commande. Sans préjuger de la validité de l’évolution proposée, je relève que celle-ci, ainsi que l’ergonomie et l’interface homme machine, fait l’objet d’un important programme de qualification

sous l’angle facteur humain.

Page 53: EDFMESS FR - inis.iaea.org

51

6

En conclusion, les métiers de conduite sont engagés, comme d’autres, dans un renouvellement de génération sans précédent. Le Groupe devra accompagner le haut niveau d’investissement humain actuellement engagé en France. Le plein succès de cet investissement, déterminant pour la sûreté en exploitation, passe par la réussite des changements importants engagés sur le contenu des

métiers, par une amélioration sensible du dispositif de formation et par une ouverture encore

accrue des agents de conduite vers l’externe.

Les directions locales doivent veiller à « penser conduite » dans leurs réflexions et leurs décisions pour viser à un meilleur niveau de sûreté et des performances d’exploitation élevées.

Page 54: EDFMESS FR - inis.iaea.org
Page 55: EDFMESS FR - inis.iaea.org

53

7LES INGENIERIES ET LES APPUIS

A L’EXPLOITATION

Le parc nucléaire français est au tout début d’un défi sans

précédent. Il ne pourra le relever qu’avec l’appui effectif et

réactif de ses propres ingénieries. Celles-ci doivent mieux

s’organiser et rechercher des solutions pour absorber, dans

des délais courts, une charge de travail considérable.

Sur le parc français, EDF ambitionne de reconquérir sa place de leader du nucléaire par son niveau de sûreté et ses performances. Porter la durée de vie des tranches nucléaires à 60 ans, intégrer le retour d’expérience de l’exploitation et les suites de l’accident de Fukushima sont ses enjeux pour les prochaines années. Face à ces challenges de natures différentes, EDF dispose d’une ingénierie intégrée qui tient une place essentielle, particulièrement pour ses soutiens à l’exploitation.L’ingénierie d’EDF est vaste et diverse et je ne peux chaque année balayer la totalité de ses missions et composantes. Cette année, je me suis particulièrement intéressé aux aspects « service après-vente du concepteur », appelé aussi l’IPE (ingénierie du parc en exploitation), et aux moyens centraux, appuis de l’exploitant.

Réussir le Grand carénage.

La DPN ne peut aborder son programme Grand carénage sans un appui solide de l’ingénierie de conception de la DIN. De même, chaque CNPE ne saurait réussir seul les challenges qui l’attendent. Les appuis à l’exploitant, qu’ils viennent

des centres d’ingénierie de la DIN ou des unités centrales de la DPN, sont

indispensables. Si la performance au quotidien est bien dans les mains des CNPE, la qualité du soutien

Page 56: EDFMESS FR - inis.iaea.org

54

7

qu’ils reçoivent pour l’atteindre est déterminante. A cet égard, je tiens à souligner que malgré les nombreuses démarches lancées ces dernières années, ce soutien reste toujours difficile à mettre en œuvre. Les initiatives pour clarifier les responsabilités et simplifier les interfaces n’ont pas encore apporté les résultats attendus.

Le processus IPE

Je constate, lors de mes visites dans les CNPE et les centres d’ingénierie, que le processus IPE reste

complexe et son fonctionnement lourd. L’organisation et les décisions de l’IPE sont encore très centralisées avec de nombreux niveaux de reporting. Je perçois le décalage entre les besoins de l’exploitant, confronté à des difficultés pour maintenir et améliorer ses performances, et les possibilités du processus IPE, qui n’est pas aujourd’hui en mesure de répondre aux demandes de modifications en moins de quatre ou cinq ans. Ce constat est assez unanimement partagé par les centres d’ingénierie de la DIN.

Réactivité dans la réponse

aux CNPE.

Trop souvent, l’ingénierie nationale ne peut pas prendre en charge les demandes des sites dans des délais satisfaisants, soit parce que son organisation et ses ressources ne le lui permettent pas, soit parce que ses priorités sont autres. En effet, je constate que les centres d’ingénierie en charge de l’IPE font constamment face à des arbitrages complexes pour optimiser leurs ressources. Les suites de

l’accident de Fukushima cette année en sont une illustration supplémentaire et aggravante.Dans cet environnement changeant, le rôle des équipes communes, têtes de pont avancées de

l’IPE sur les sites, est vital. Je comprends à quel point elles facilitent l’appropriation des modifications par les services des sites et, en retour, alertent les centres d’ingénieries sur les contraintes liées à la mise en œuvre de certains dossiers. J’ai apprécié de rencontrer des équipes motivées et fières de leur métier, même devant l’ampleur de la charge de travail et des programmations tardives et changeantes. La distance entre la production et l’ingénierie sur les sites en France est moins marquée à l’étranger, car les équipes d’ingénierie locale y sont plus fortes. Face à ce constat, les équipes communes sont, de mon point de vue, un bon facilitateur du dialogue entre les CNPE et leur ingénierie.

Les projets de l’ingénierie pour le parc

Le projet DIN/DPN

Les constats ci-dessus fragilisent le parc nucléaire face aux défis qui l’attendent. L’exploitant a

besoin pour ses CNPE d’une ingénierie plus réactive et plus proche de ses enjeux et de

ses préoccupations, a contrario l’ingénierie de la DIN a besoin de recul et de vision à plus long terme. Lancé en 2010, le projet « DIN/DPN pour le parc en exploitation » doit y répondre. J’approuve ses orientations qui consistent, premièrement, à mettre en place des compétences de conception sur chaque site et, deuxièmement, à différencier clairement à la DIN ce qui relève de l’ingénierie patrimoniale dédiée aux grands projets de long terme, de l’appui de conception réactif aux besoins immédiats de l’exploitant. Je regrette toutefois le retard pris par ce projet cette année, même si celui-ci est en grande partie dû à la mobilisation des équipes pour les ECS, suite à l’accident de Fukushima.

Renforcer les ingénieries

locales.

J’avais déjà alerté dans mes précédents rapports sur la faiblesse des ingénieries locales et la surcharge de demandes que cela entraînait pour les centres d’ingénierie. Je salue le choix de mettre en place des compétences de conception sur chaque site et j’encourage les directions de la DIN et de la DPN à gréer au plus vite ces forces d’ingénierie indispensables à l’exploitant. Il me semble bon de

redonner aux CNPE les leviers d’action qu’ils ont perdus et je me félicite de voir qu’un des objectifs de ce projet est de diminuer les temps de réponse à leurs demandes.

Page 57: EDFMESS FR - inis.iaea.org

55

7

Faire face au doublement de la charge d’activité

Rechercher des ingénieries

partenaires

Les centres d’ingénierie en charge de l’IPE font constamment face à des arbitrages complexes pour optimiser leurs ressources. Je ne pense pas que ceux-ci se fassent toujours au bon niveau. Un tel mode de régulation de la charge IPE ne permet plus de faire face au surcroît d’activité futur. Les volumes d’activités pour le

parc vont rapidement plus que doubler. Même si de gros efforts de recrutement, que je salue, sont faits, l’ingénierie ne pourra pas doubler ses ressources dans un si court laps de temps (cf. chapitre 5). L’entreprise ne peut plus, comme par le passé, étaler les activités pour lisser la charge. Ne faut-il pas préparer une autre approche du faire et du faire-faire, davantage tournée vers des partenariats et des coopérations avec d’autres ingénieries ? Dans ce contexte, l’ingénierie intégrée d’EDF gagnerait à s’ouvrir plus sur l’externe pour rechercher des collaborations ou des solutions déjà éprouvées.

La gouvernance de l’IPE

Aujourd’hui l’exploitant doit prendre une part plus active dans la fixation des priorités de travail de l’ingénierie et dans les renoncements à faire sur son parc. Dans le cadre du projet DIN/DPN, un groupe de travail a proposé des évolutions de la gouvernance de l’IPE, en particulier sur le fonctionnement, au niveau des directions, des directoires et des comités, encore très nombreux. L’Instance de commandite technique, créée par les directions de la DIN et de la DPN, fixera les grandes orientations du parc. Je suivrai avec intérêt son travail et ses résultats qui, je le pense, amélioreront le fonctionnement de l’IPE.

Le projet Méthodes d’ingénierie 2012

Je suis frappé de voir qu’il n’existe pas encore, dans les centres d’ingénierie visités ces dernières années, d’outils de gestion de projet, tels qu’on en trouve maintenant sur le marché et dans de nombreux bureaux d’études. Les ingénieurs travaillent encore trop souvent avec leurs propres outils

développés localement. Cela ne facilite ni les interfaces ni la performance globale de l’ingénierie. Devant la très forte augmentation de l’activité pour le parc en exploitation et pour ses nouvelles affaires, la DIN a lancé une analyse en profondeur de ses modes de pilotage et de suivi de projets. Le projet Méthodes d’ingénierie 2012 issu de cette analyse a vu ses premières actions cette année. Les grands projets de la DIN du nouveau nucléaire comme de l’IPE doivent rapidement en tirer bénéfice. Le travail sur les outils et les méthodes est une source de productivité indispensable pour l’ingénierie, qui ne doit pas oublier que la charge de travail future est considérable et bien au-dessus de ses moyens actuels.

Les moyens centraux de la DPN

A la DPN, deux unités centrales, l’UTO et l’UNIE57, regroupent les appuis et les expertises nécessaires à l’exploitation et à la maintenance du parc. A mi-chemin entre l’ingénierie de conception et l’ingénierie locale, ces deux unités portent pour le parc des projets à forts enjeux de sûreté et de performance.

Cette année, je me suis rendu à l’UNIE où j’ai constaté la qualité des travaux des équipes et leur contribution aux performances techniques du parc. L’UNIE pilote pour le compte de la direction de la DPN beaucoup de projets et de processus pour lesquels elle a acquis un savoir-faire élevé et reconnu. Sa réactivité pour faire face aux nombreuses affaires qui lui sont confiées est notable.

57 UNIE : UNité d’Ingénierie d’Exploitation.

Page 58: EDFMESS FR - inis.iaea.org

56

7

Développer l’assistance aux

CNPE.

Dans le contexte actuel d’amélioration de la sûreté et de recherche de la performance, je m’interroge sur le besoin de piloter ou de traiter depuis Paris des projets majoritairement spécifiques aux sites. Je vois là un positionnement assez particulier des moyens centraux du parc en France, dont les missions sont trop souvent tournées vers le pilotage (grands projets) ou l’organisation (processus) au

détriment de l’assistance et de l’appui aux sites qui reste encore rare. A mon sens, ce point mérite une analyse et un benchmark avec les pratiques d’autres grands exploitants américains ou britanniques.

Cette année dans le cadre de mes visites, je me suis rendu à l’INPO à Atlanta (cf. chapitre 11). L’INPO assure pour l’ensemble du parc américain quatre missions fondamentales : l’évaluation, le retour d’expérience, l’assistance et la formation. Orientées sur la performance et faites par des auditeurs expérimentés, les évaluations intègrent systématiquement des revues des fonctions de sûreté, sur site et dans la compagnie. Le retour d’expérience vient du parc américain et de l’international. Les événements sont analysés par les équipes d’ingénieurs de l’INPO et partagés entre toutes les compagnies. Une

assistance est proposée systématiquement aux sites dont les résultats déclinent et elle

s’impose aux sites en difficulté. Elle comporte des visites de soutien, des analyses des difficultés et la création d’indicateurs de suivi. La formation, mise à disposition des compagnies et de leurs sous-traitants, regroupe de nombreuses sessions accessibles en permanence. Celles-ci permettent de s’assurer de la qualification et du savoir-faire des intervenants en vue de leur habilitation.

Un exemple à suivre.

Avec des tranches de technologies et d’âges très divers et des opérateurs différents, le parc nucléaire américain peut pourtant apparaître, dans son mode de management de la sûreté et de la performance, comme plus homogène que le parc français, pourtant standardisé avec un seul opérateur. Je m’interroge sur les

causes d’une telle différence. Il me semble que l’INPO, en laissant l’initiative des solutions aux sites mais en apportant son soutien lorsque les performances se dégradent, et en axant ses missions sur l’assistance, contribue à cette homogénéité. Face aux enjeux futurs du parc français, je note qu’une réflexion est en cours sur les missions et le positionnement de l’UNIE et de l’Inspection nucléaire. Je ne manquerai pas d’en suivre les recommandations.

En conclusion, le parc nucléaire français doit s’appuyer sur une ingénierie solide et performante pour relever le défi du Grand carénage. La DIN et la DPN recherchent ensemble les solutions les plus performantes pour répondre aux besoins des sites. La charge de travail de l’ingénierie à court et à moyen terme ne pourra se solder sans une ouverture plus forte vers l’externe et la recherche d’autres méthodes de travail. A l’image d’autres exploitants, les moyens centraux du parc doivent renforcer l’appui et l’assistance aux sites pour en resserrer les résultats.

Page 59: EDFMESS FR - inis.iaea.org

57

8

LA CONDUITE DES PROJETS

Le management en mode projet est un facteur de sûreté et

de performance économique. Le succès des projets passe

par un pilotage professionnel et des objectifs adaptés à la

capacité d’assimilation de l’organisation concernée. Ils doivent

répondre davantage à des exigences en matière de conduite

du changement et d’accompagnement, surtout s’ils sont à fort

enjeu. C’est le cas des projets EPR, défis considérables pour le

Groupe.

De nombreuses initiatives importantes pour la sûreté nucléaire sont gérées sous forme de « projets » dans l'ensemble du Groupe. Ces initiatives sont extrêmement variables, par leur taille et leur degré de réussite (sûreté, durée, qualité et coût). S’ils ne se déroulent pas de façon suffisamment efficace et rationnelle, les projets peuvent, par leur nombre même, avoir des effets négatifs sur d'autres modes de travail. En revanche, les projets bien menés constituent un outil précieux pour le management.

Relever des challenges techniques complexes

Dans certains cas, je constate que des problèmes techniques spécifiques affectant la sûreté et la production sont gérés sous forme de projets. Ils couvrent par exemple les tests et études post-Fukushima ou certains problèmes ne concernant qu'une ou deux tranches, comme les dômes des enceintes chaudes des réacteurs AGR. En règle générale, ces activités sont bien organisées et ciblées, le travail en mode projet constitue un atout indéniable.

Page 60: EDFMESS FR - inis.iaea.org

58

8

Les arrêts de tranche programmés

Encore une marche

à franchir.

Tous les arrêts de tranche programmés, comme les rechargements de combustible, les révisions décennales, le remplacement des générateurs de vapeur et les arrêts de tranche réglementaires, sont autant de « projets ». Ce que j’observe varie, en

qualité et en résultats. Les unités les plus performantes du Groupe disposent des compétences nécessaires, préparent les lots de travaux et les pièces de

rechange en temps utile, fixent le contenu de l’arrêt, impliquent les principales entreprises prestataires dans la préparation et la planification, puis exécutent les activités en toute sûreté, de façon rationnelle et efficace. Le faible taux d’indisponibilité fortuite de ces unités confirme la qualité du travail réalisé. Alors que des travaux de plus grande ampleur sont prévus, notamment en France avec le programme Grand carénage, il faut prendre en compte dès à présent la marche à franchir par certaines centrales pour atteindre l'excellence.

La conduite du changement au niveau d’un parc

Le trop est l’ennemi

du bien.

Les initiatives menées à l'échelle d’un parc sont nombreuses, notamment en France. Mon propos n'est pas de les décrire individuellement mais il est utile d'en citer certaines : O2EI (Obtenir un État Exemplaire des Installations), INTEP (Introduire de Nouvelles Technologies en Exploitation du Parc), MOPIA (Mettre en Œuvre une Politique Industrielle Attractive), AMELIE (pièces de rechange), PPH

(Projet Performance Humaine), SAT (Systematic Approach to Training, démarche systématique de formation), SDIN (Système D'Information du Nucléaire), BMA (Bibliothèque des Modèles d’Activité), PHPM (Projet d'Harmonisation des Pratiques et des Méthodes), REX (Retour d'EXpérience), COPAT (Centre Opérationnel de Pilotage des Arrêts de Tranche), AP 928 (Work Management), AP 913 (fiabilité des matériels). Dans un parc de 58 tranches sur 19 sites, de nombreux projets fonctionnent selon le principe des sites pilotes, qui anticipent le déploiement des nouvelles méthodes. C’est une bonne pratique que j’approuve. Elle implique de faire des choix pour déterminer dans quelle mesure

standardiser ou conserver une marge d’adaptation sur chaque site. Je pense que chaque projet doit s’efforcer de trouver l'équilibre qui lui semble le meilleur, en tenant toujours compte du principe de « subsidiarité » qui consiste, autant que possible, à responsabiliser et à encourager des personnes formées à travailler en faisant preuve de réflexion et d’engagement.

Les projets réussis, et ils sont nombreux, poursuivent des objectifs très clairs et leur avancement est mesuré et suivi en permanence. Lorsqu’ils aboutissent, ou atteignent un degré de maturité suffisant, leur intégration à l'activité courante doit être étudiée avec soin et gérée efficacement pour pérenniser leurs avantages. Je constate que ce n'est pas toujours le cas. En outre, le nombre d'initiatives gérées

sous la forme de « projets parc » me semble supérieur à ce que la plupart des sites peuvent

prendre en charge, avec leurs campagnes de production et leurs ressources humaines actuelles.Je note toutefois, en 2011, l'effort de cadencement dans le temps de ces projets. Je considère que la programmation pluriannuelle dans chaque CNPE, intégrant les possibilités et les capacités de chaque site, est une bonne pratique. Les systèmes de régulation au niveau du parc, comme le programme Méthodes de maintenance et d'exploitation 2015 pour les projets AP913-AMELIE-SPP-SDIN-BMA, répond également à ce besoin et doit faciliter la réussite, sur les sites, de chacun de ces projets importants. Ces évolutions positives méritent une grande continuité dans la durée, tant le déploiement de l'ensemble de ces projets est étalé sur une longue période (jusqu’en 2017).La création des directeurs délégués opérationnels (DDO) est un point positif : ils facilitent la mise en place des programmes pluriannuels, l’analyse des risques des projets, la priorisation et l’arbitrage.

Je remarque aussi que, dès que l'on affecte une structure de projet (appelée directoire en France) à une problématique technique importante (par exemple les générateurs de vapeur ou les alternateurs), celle-ci court le risque de ne plus être coordonnée correctement aux autres projets ni aux activités courantes.

Page 61: EDFMESS FR - inis.iaea.org

59

8

Au Royaume-Uni, j'observe un mode de travail dans lequel, pour certains domaines techniques, les gestionnaires du parc collaborent avec les sites afin d'élaborer des transition maps (plannings de déploiement) sur cinq ans. La chronologie et l'ordre de déploiement des étapes inscrites dans ces plannings sont laissés à la discrétion de chaque site. Cette approche me semble pragmatique. Néanmoins je relève qu’avec 8 sites la gestion du parc au Royaume-Uni est plus simple qu’en France avec 19 sites.

Faire face au pic d’activité arrivant

Penser différemment.

L’objectif du projet DIN et DPN pour le parc en exploitation est de garantir que la DPN et la DIN planifient et trouvent les moyens de gérer efficacement le futur pic d'activité considérable lié aux troisièmes et aux quatrièmes arrêts décennaux, à l’allongement de la durée vie des centrales et aux modifications post-Fukushima,

sans oublier les mesures requises pour les EPR. Je comprends des échanges avec la direction de la DPN « qu’aplanir le pic d’activité en retardant les travaux n'est pas une bonne solution mais qu’il faut, au contraire, y faire face ». Je partage totalement cette position.

Des décisions importantes doivent être prises pour définir la part des travaux à externaliser et ceux à réaliser en interne, les méthodes de travail spécifiques à adopter ainsi que l’organisation du travail, les matériels de remplacement et les pièces de rechange, ainsi que les compétences et les effectifs précis à fournir, par EDF et par les principales entreprises prestataires.

Des projets neufs

Une synergie à construire.

Les projets EPR du Groupe déjà engagés, Flamanville, Taishan et Hinkley Point, sont gigantesques. Ils n’ont pas tous atteint le même stade et chacun se caractérise par un contexte particulier et une structure de projet spécifique. Du point de vue de la conduite des projets et de la sûreté, je relève plusieurs

points indispensables.

Un leadership clair et une structure organisationnelle aussi simple que possible pour atteindre l'excellence. On obtient les meilleurs résultats lorsque les directions générales des principaux acteurs du projet communiquent régulièrement et entretiennent une dynamique d’ensemble où les parties concernées conjuguent leurs efforts sur un même objectif. Cela favorise la tenue des plannings, mais aussi la diffusion d’une véritable culture de qualité lors de la conception, de la fabrication, de la construction et des essais, sur laquelle se fonde à son tour la culture de sûreté nucléaire, impérative pour le démarrage de la centrale.

Un processus de planification fiable et global couvrant tous les aspects de la contribution des différents acteurs et assurant les interconnexions nécessaires. Par exemple, les liens entre les études et les approvisionnements pour la démonstration de sûreté (licensing safety case au Royaume-Uni) ainsi que leur calendrier d’exécution doivent être traçables et clairs pour tous : architecte ensemblier, concepteur qui commande le matériel, constructeur, contrôleur qualité, rédacteur des procédures (conduite, maintenance et formation), équipes chargées des tests de construction et de la mise en service, etc. Une telle planification permet de suivre la progression du projet et de gérer les perturbations et modifications toujours possibles, en toute sûreté et de façon contrôlée. Il est de plus possible de planifier les activités de surveillance afin que des ressources adaptées soient disponibles au moment opportun pour garantir une qualité conforme aux intentions des concepteurs.

Des équipes de projet disposant de ressources adaptées et des acteurs clés, géographiquement

regroupés autant que possible. Le bureau d'études commun créé en 2010 par le chaudiériste et son client chinois, pour le projet de Taishan, en est une bonne illustration.

Les standards et méthodes le plus éprouvés pour gérer l’information, la documentation et les travaux (par exemple, conception des tuyauteries et cheminement des câbles).

Page 62: EDFMESS FR - inis.iaea.org

60

8

Un niveau de sécurité du travail et de house-keeping exemplaire, à tous les stades de la construction. Chacun doit être conscient de la nécessité d'entretenir et de protéger les matériels livrés sur le chantier, avant et après le montage. Tout cela est particulièrement important sur les chantiers situés en bord de mer. Dans certains cas, j'ai malheureusement constaté que cela n'était pas fait correctement.

Des compétences et de l'expérience dans tous les domaines, en particulier pour la passation des marchés, la planification, le management de la sûreté, la surveillance et la coordination des travaux.

En conclusion, les travaux traités en mode projet sont une composante essentielle pour l’efficacité de l'industrie nucléaire. Comme je l'ai indiqué dans mon dernier rapport, sûreté nucléaire et performance économique sont étroitement liées. Une gestion efficace des projets est un des moyens pour les optimiser conjointement.

Page 63: EDFMESS FR - inis.iaea.org

61

9

AVEC LES ENTREPRISES PRESTATAIRES

VERS UNE MEILLEURE PERFORMANCE

D‘ENSEMBLE

Les relations contractuelles mettent en évidence d’année en

année des insatisfactions de part et d’autre, car les exigences

augmentent aussi vite que les corrections apportées au fil

du temps, au terme d’échanges fructueux. Les solutions et

les engagements complexifient souvent la forme sans aller

suffisamment sur le fond de ce que le nucléaire exige. La

Charte de progrès et de développement durable et l’accord

récent sur la Sous-traitance socialement responsable à EDF

doivent impérativement faire évoluer le modèle des relations

industrielles d’EDF SA.

Pour le parc nucléaire, l’enjeu porté par les entreprises prestataires est très important à la fois pour la performance technico-économique et pour la sûreté nucléaire. La maintenance est globalement sous-traitée à hauteur de 80 %, voire davantage avec l’accroissement des visites décennales, le Grand carénage et les travaux engagés dans le cadre du retour d’expérience de l’accident de Fukushima.De très nombreux dispositifs d’amélioration des conditions d’intervention des entreprises prestataires sont déployés sur le parc, à l’initiative d’EDF. Je souhaite lister, en les actualisant, les voies de progrès qui s’imposent.

Page 64: EDFMESS FR - inis.iaea.org

62

9

Mieux partager les enjeux stratégiques avec les entreprises prestataires

Le projet MOPIA58 est une boîte à outils bien garnie mais qui ne produit pas tous les effets escomptés. De nombreuses dispositions ont créé de meilleures conditions de travail pour les entreprises, en particulier par une plus grande visibilité contractuelle (durée des contrats) et par le développement d’une

motivation des entreprises (mieux-disance, bonus…). Je crains que ces dispositions, globalement bien reçues par les entreprises, ne constituent pas pour autant une évolution suffisante dans le partage d’objectifs stratégiques. La qualité des travaux, la préparation et la durée des chantiers pour une meilleure maîtrise de la durée des arrêts ne progressent pas de manière significative.Le rebond attendu n’est pas du seul ressort des entreprises, EDF doit prendre sa part dans cette transformation, j’y reviens par la suite. Je pense que c’est là le problème majeur du dossier et je m’interroge sur les raisons qui peuvent conduire à cette situation. Est-il si difficile pour un client de reprendre la main sur les exigences qu’il impose ? Les pratiques passées ont-elles marqué le pli de manière aussi permanente ?

La formation, levier de progrès.

Je pense qu’une première réponse concerne la formation des personnels des

entreprises prestataires. Cette formation doit toucher à mon sens tous les personnels, pas seulement les intervenants. Intervenir sur des marchés nucléaires impose à tous les acteurs d’une entreprise de prendre conscience des exigences

(sûreté, sécurité, radioprotection, environnement, assurance qualité…) associées à ces activités. EDF déploie certes des accompagnements pour les entreprises primo-intervenantes mais l’appropriation de la culture de sûreté doit être sans cesse promue et enrichie par l’entreprise elle-même, avec le soutien d’EDF si besoin. La vision stratégique commune ne peut exister qu’à cette condition. EDF doit se donner les moyens d’obtenir une telle situation pour l’ensemble de ses entreprises prestataires. Un des obstacles réside dans la part de chiffre d’affaires EDF, parfois faible, pour certaines entreprises.Aux Etats-Unis, l’INPO, association de 26 opérateurs nucléaires exploitant 104 tranches, assure pour leur compte la formation et l’évaluation des intervenants de tous niveaux des entreprises prestataires en matière de sûreté nucléaire. Le contrôle est sélectif, le management des opérateurs et des prestataires s’implique. Je suis convaincu que ce dispositif est plus performant que le processus actuel d’EDF.

Mieux anticiper les interventions sur les sites

Anticipation et préparation,

gages de succès.

Là encore, je renvoie dos à dos EDF et ses entreprises prestataires. Je constate, malgré des progrès perceptibles, qu’EDF ne fixe pas le contenu de ses arrêts suffisamment tôt et ne passe pas tous les contrats dans des délais raisonnables. Les entreprises ne bénéficient pas alors de disponibilités suffisantes pour

préparer en parallèle leurs interventions sur les différents sites où elles

ont été retenues.

Une partie des éléments perturbateurs pour la planification vient des très nombreuses modifications réalisées sur le parc EDF au fil de l’eau ou par lots. Ces modifications concernent des aspects techniques, dont la sûreté, bien entendu. Aucun arrêt pour maintenance ne ressemble au précédent. Cette situation existe aussi à l’étranger, avec des référentiels techniques qui semblent évoluer moins vite. Le bénéfice de cette situation pour EDF est d’être en mouvement permanent pour améliorer la sûreté mais l’effort nécessaire est-il compatible avec les ressources et compétences de la DIN, de la DPN et des entreprises prestataires ?A Sizewell B, site d’EDF Energy au Royaume-Uni, j’ai rencontré lors d’une visite d’inspection en février une équipe de préparateurs qui travaillaient avec des représentants d’entreprises prestataires, 18 mois avant l’arrêt. Certains sites d’EDF engagent des politiques partenariales volontaristes impliquant des entreprises prestataires très tôt avant l’arrêt de tranche, lorsque les ressources le permettent.

58 MOPIA : Mettre en Œuvre une Politique Industrielle Attractive.

Page 65: EDFMESS FR - inis.iaea.org

63

9

Améliorer les conditions d’intervention

Repenser et simplifier

le référentiel.

Pour les conditions matérielles de vie sur les sites, je note d’année en année une amélioration continue, reconnue par les représentants d’entreprises rencontrés. Pour autant, les très nombreuses démarches technico-administratives préalables à l’engagement d’un chantier et les aléas de planification conduisent

à des temps d’attentes inacceptables. Appelé « temps métal », le temps consacré au geste professionnel, objet du contrat, est anormalement faible. J’identifie de manière récurrente les facteurs majeurs de dégradation suivants :

l’organisation et l’environnement du travail pèsent trop sur l’intervenant et son encadrement. La sécurité, la radioprotection, les permis de feux, les dossiers techniques, les consignations pour travaux sont de la responsabilité du chef de travaux des entreprises et donc autant de préalables à l’arrivée sur le chantier,

la fragilité de la programmation des interventions évoquée précédemment engendre des temps d’attente car des scenarii alternatifs ne sont pas suffisamment envisagés,

les pièces de rechange manquantes interrompent des chantiers et reportent les intervenants sur d’autres chantiers planifiés à d’autres moments de l’arrêt et finalement désorganisent le planning général,

des outillages sont parfois indisponibles à la pointe de l’activité en arrêt de tranche et les temps d’attente sont parfois très longs aux magasins,

la logistique est à améliorer impérativement pour créer des conditions d’intervention en ligne (par exemple : montage et équipement de sas, échafaudages, démontages de calorifuges).

Ces faiblesses doivent être corrigées dans les meilleurs délais, en vue des grands travaux de demain.A l’étranger, sur les sites les plus performants, je constate que la force des services de planification est en général plus importante. Au-delà de la maîtrise du planning nominal, le chemin critique (fil rouge de l’arrêt) est suivi sur des intervalles très fins et des scenarii de substitution sont préparés et dotés en moyens pour faire face à l’aléa avec un minimum d’effet. J’observe aussi que les intervenants

bénéficient généralement d’une assistance plus grande et que parfois les spécialistes arrivent sur leur chantier juste à temps, quand tout est prêt, un peu comme des chirurgiens.

Des non-qualités de maintenance

Pour une meilleure visibilité.

L'année 2011 aura été marquée, plus que les années précédentes, par des non qualités de maintenance avec des responsabilités partagées entre EDF SA et ses entreprises prestataires. Les marchés nucléaires d’EDF n’ont pas la même envergure chaque année, compte tenu de la nature des arrêts (nombreuses visites décennales en cours, par exemple), du volume des modifications et des

programmes spécifiques de travaux (par exemple O2EI, Plan d’action incendie, Grand carénage). Sans contrat suffisamment pérenne, les besoins en compétences ne peuvent pas être raisonnablement anticipés par les entreprises prestataires. De plus, des pressions sur les montants des contrats se

traduisent souvent par des réductions d’encadrement.

Motiver davantage le personnel intervenant

Une reconnaissance plus réactive.

Primes, bonus et malus entrent souvent dans des pots annuels, ce qui peut être compréhensible pour la gestion d’une entreprise. Mais au final, le bilan peut être nul, les malus détruisant les effets des bonus et des primes, ce qui est très démotivant pour ceux qui ont mérité le bonus.

J’observe qu’aux Etats-Unis, où la motivation des acteurs est une préoccupation majeure du management des entreprises, les bonus sont payés aux personnels concernés, au plus près de la fin des interventions.

Page 66: EDFMESS FR - inis.iaea.org

64

9

Parvenir à une surveillance de qualité

La surveillance d’EDF se développe pour assurer la qualité des chantiers majeurs. Les agents chargés de surveillance se professionnalisent sur le terrain et suivent de nouveaux stages spécifiques dans les activités complexes. Malgré cela, ce métier manque d’homogénéité entre les sites et entre les divisions, il est peu reconnu et sans parcours professionnel associé. C’est pourtant un métier à part entière et le volume des travaux des prochaines années le justifiera pleinement.

Travailler mieux pour surveiller

moins.

Un dispositif idéal conduirait les entreprises à « bien faire du premier coup », avec un système de contrôle externe d’EDF plus justement proportionné. La prévention

serait alors créatrice de richesse.

A la DIN, dans les équipes communes sur les CNPE, le processus de surveillance est soumis à des fluctuations d’activités saisonnières et renforcé lors des pics par la sous-traitance. Cette situation mérite d’être évaluée de manière continue.

Je vois pourtant, lors de mes visites sur le terrain et dans les unités centrales, beaucoup d’acteurs qui déploient une énergie très importante, et des interfaces organisationnelles supplémentaires pour limiter les conséquences de toutes ces situations. Je souhaite leur rendre hommage et reconnaître leur travail mais, en même temps, je déplore cette dépense d’énergie due à des faiblesses de l’organisation

qui peuvent nuire à la sûreté.

De nombreuses pistes d’amélioration, menées au niveau national sont prometteuses comme le COPAT et la préparation des arrêts, ou encore AMELIE59, sous réserve de laisser suffisamment de place à l’initiative locale et au pragmatisme. A titre d’exemple, aller vers une plus grande représentativité des fiches d’évaluation prestataires à l’échelle des bassins industriels est une bonne décision. En effet, l’évaluation au niveau national d’un prestataire composé de plusieurs agences régionales aux performances contrastées n’avait pas grand sens.

Une autre difficulté m’est souvent rapportée sur les sites visités : certains chantiers ou marchés

génériques sont réalisés par des entreprises choisies par le niveau national sans que les sites

n’aient leur mot à dire. Les sites ayant établi des relations performantes avec d’autres prestataires voient alors leur travail anéanti.

Dans la liste des ECS demandées par l’ASN après l’accident de Fukushima, le thème du « recours à la sous-traitance » a été retenu. Il s’agit d’une exception française et je ne vois pas de lien direct entre ce thème et l’accident. Au Japon, les niveaux de sous-traitance sur un même contrat peuvent être nombreux aux yeux des Européens. Je n’ai jamais eu connaissance de telles pratiques en France.

En conclusion, les entreprises et groupements d’entreprises, les organisations professionnelles, et EDF SA doivent, à mon sens, faire encore évoluer leurs modèles d’activités vers plus de partenariat. La Charte de progrès et de développement durable puis l’accord sur la Sous-traitance socialement responsable à EDF vont dans le bon sens, mais constituent-ils le levier suffisant pour atteindre rapidement les ambitions retenues pour le parc nucléaire ?

59 AMELIE : projet visant à transformer la logistique des pièces de rechange.

Page 67: EDFMESS FR - inis.iaea.org

65

10

LE NUCLEAIRE D’EDF

HORS DE FRANCE

Le contexte d'exploitation diffère selon les pays mais on

retrouve partout une même priorité accordée à la sûreté

nucléaire et des performances challengées en permanence.

Aucun événement grave n’a affecté la sûreté des 20 réacteurs

du périmètre EDF hors de France. Les nombreuses pratiques

performantes identifiées dans le Groupe méritent d’être mieux

partagées au sein d’EDF, pour l’exploitation, la construction et

la mise en service des nouveaux réacteurs.

Cette année, j’ai visité une partie des installations nucléaires du groupe EDF hors de France. EDF est l'actionnaire principal d'EDF Energy qui exploite quinze réacteurs au Royaume-Uni et prévoit d'y construire quatre EPR. En outre, le Groupe détient une participation de 49,9 % dans CENG qui exploite cinq réacteurs aux Etats-Unis et de 30 % dans les deux EPR en construction en Chine.

Page 68: EDFMESS FR - inis.iaea.org

66

10

Des points communs

Je relève des aspects qui contribuent indéniablement, et de façon mesurable, à la sûreté nucléaire.

Les formations sont souvent bien organisées et associées à une méthode d'accréditation par une commission comprenant des acteurs externes à l'entreprise. Pour les opérateurs de conduite, ces formations sont intégrées dans le roulement des équipes, ce qui garantit un entraînement et une évaluation réguliers. Un haut degré d'implication caractérise la ligne managériale qui veille à la satisfaction des besoins courants des centrales.

Les comités de supervision de la sûreté nucléaire tirent souvent bénéfice de la présence de personnalités externes qualifiées.

Des actions sont engagées pour éviter que les personnels qui suivent des procédures ne se concentrent trop sur les gestes qu’ils font et ne perdent de vue l'objectif premier de leur action.

La communication et la synergie entre les entreprises du groupe EDF se renforcent sur des aspects techniques et de sûreté, comme les domaines de la chimie, de l'ingénierie, des alternateurs, et via des participations à des revues de pairs.

Les salles de commande se caractérisent par une ambiance attentive et calme, et j’y ai observé une utilisation systématique des outils de performance humaine.

Les arrêts de tranche font l’objet d’une préparation très anticipée, donc de qualité. Le planning est protégé des travaux fortuits par le recours à des équipes dédiées.

Chaque réunion formelle débute par un message relatif à la sûreté ou à la sécurité, permettant au management d’exprimer clairement ses attentes, tout en incitant le personnel à réfléchir à la sûreté nucléaire.

Les évaluations de sûreté des sites s’appuient sur des organisations externes reconnues (INPO, WANO).

EDF Energy au Royaume-Uni

Les pratiques managériales

Les leaders forgent

la culture de sûreté.

Le management opérationnel du parc britannique s’appuie sur un dispositif composé de trois Chief Nuclear Officers, chacun chargé de 2 à 3 sites. Les centrales bénéficient ainsi d’un soutien rapproché et efficace. J’attire l’attention sur la nécessité de maintenir une autonomie suffisante des directeurs de sites.

Je perçois une forte volonté d'améliorer à la fois la sûreté nucléaire et la performance économique, par exemple en réduisant fortement les « pots de demandes d’intervention » en attente, ou en investissant régulièrement dans le maintien en état des matériels d’exploitation, sur la base d’un système de gestion des risques facilitant les prises de décision et les arbitrages de ressources. Je constate toutefois que ce dispositif peut encore être mis en défaut (cf. chapitre 12).Des enquêtes de perception de la culture de sûreté sont réalisées régulièrement auprès des salariés d’EDF Energy et des prestataires principaux. Avec un taux de réponse très élevé (90 %), elles sont comparées aux enquêtes similaires réalisées par des compagnies d’électricité des Etats-Unis. Leurs résultats sont utilisés par les managers pour des actions ciblées et justifiées, sur les sites et dans les services centraux.Je suis également impressionné par les efforts réalisés pour proposer des formations innovantes

au leadership, à tous les niveaux. Ces formations portent sur le « professionnalisme nucléaire », les outils de performance humaine, le tutorat ou le développement du leadership via un travail en groupes réunissant des salariés de différents métiers et niveaux hiérarchiques (diagonal slices). Des images fortes et parlantes y sont utilisées pour montrer comment l'exécution concrète des travaux peut lentement

Page 69: EDFMESS FR - inis.iaea.org

67

10

dériver en-deçà des attentes de la direction et comment, en même temps, les dangers, risques et défauts matériels se cumulent sans être décelés jusqu'à disparition des marges de sûreté.Je relève les filières indépendantes de contrôle de la sûreté nucléaire sur les sites ne me semblent pas avoir encore assez de poids et d’autorité. Je note avec satisfaction qu’une nouvelle approche, après un benchmark avec des missions sûreté-qualité de CNPE français, fait l’objet d’une expérimentation prometteuse dans trois centrales britanniques.Pour le contrôle de sûreté des sites et du corporate, EDF Energy a fait le choix d’utiliser les "peer reviews" de WANO réalisées tous les trois ans. Il n’existe pas de service d’inspection interne pouvant procéder à des évaluations ou des audits de grande ampleur comparables aux EGS60 de l’inspection nucléaire en France.

Je note en 2011 une réorientation stratégique de niveau parc fixant les priorités des comités

de sûreté nucléaire. Outre l'examen des modifications techniques et organisationnelles importantes impactant la sûreté, l'accent est plus clairement mis sur les dossiers opérationnels soulevés par les sites. Cette évolution, appréciée des directeurs de centrales, me paraît aller dans la bonne direction pour la sûreté.

Une dynamique sur le terrain

Savoir créer les conditions.

Sur les sites, je perçois une dynamique constructive avec la plupart des

grandes entreprises prestataires. Les objectifs et les formations en rapport avec la sûreté et les comportements sur le lieu de travail sont définis et renforcés conjointement. Des safety representatives (délégués sûreté-sécurité volontaires)

sont en place dans les services d’EDF Energy et j’ai apprécié leur enthousiasme et leur engagement. Les contrats avec les entreprises couvrent généralement de longues périodes et les incitations financières en faveur de la sécurité au travail et de la qualité sont claires. D’importants efforts sont faits pour atteindre les objectifs du programme Zero Harm61. Je relève une forte volonté de simplification et d’élimination des tâches administratives inutiles associées aux processus clés. Ces actions sont issues de l'analyse réalisée après l'accident mortel de Heysham 1, en 2010.La recherche du progrès continu pour améliorer la qualité a été renforcée après certains événements de non-qualités de maintenance. Les chargés de surveillance ( field supervisors ), EDF Energy et prestataires, sont reconnus comme des facteurs de progrès déterminants. Nombreux, ils sont formés et habilités pour remplir cette fonction. Je me félicite de cette ambition, de l’autorité qui leur est accordée et des moyens qui sont consacrés.Je relève néanmoins des prises de décisions techniques parfois trop largement déléguées aux entreprises prestataires, tant pour des opérations de maintenance que pour des affaires d'ingénierie locale. Ce point mérite une attention particulière, tant il peut potentiellement fragiliser la responsabilité sûreté de l’exploitant nucléaire, maître d’ouvrage.

Des points notables en exploitation, des initiatives intéressantes

Dans toutes les centrales, les revues de pairs de WANO (peer reviews) mettent en avant des voies d’amélioration pour l’entretien des équipements incendie et la maîtrise des charges calorifiques.Pour la radioprotection, les personnels britanniques bénéficient toujours d’un bas niveau

d'exposition résultant de la conception des réacteurs AGR. Je constate également avec satisfaction que le site de Sizewell B se maintient dans le meilleur décile des statistiques mondiales de dose pour les réacteurs REP. Je note que la maîtrise des contaminations est affichée comme un des objectifs prioritaires pour l’ensemble du parc.J’ai observé dans une centrale un processus très réactif, one stop process, qui permet de corriger rapidement les écarts identifiés lors de l’application des procédures d’exploitation. Cette pratique répond à la nécessité de maintenir à jour les procédures avec un très haut niveau de qualité, qui justifie l’exigence d’un strict respect de celles-ci. Elle me paraît mériter un benchmark avec le parc français.

60 EGS : Evaluation Globale de Sûreté.61 Zero Harm : programme de performances d’EDF Energy.

Page 70: EDFMESS FR - inis.iaea.org

68

10

Trois événements marquants

Sur l’ensemble des événements techniques de l'année 2011, je retiens d’abord celui, classé au niveau 0 de l’échelle INES, ayant affecté une centrale de bord de mer, lors d’une arrivée massive et subite de méduses (800 tonnes), ce qui illustre bien la nécessité de se préparer à toutes sortes d’agressions externes. En l’occurrence, les opérateurs ont toujours maîtrisé la situation et mis à l’arrêt les réacteurs en toute sûreté. L’utilisation réactive et innovante de chalutiers a également permis de faire face à cet incident.

Dans une autre centrale, un événement classé au niveau 1 de l’échelle INES a affecté les circuits de refroidissement auxiliaires du réacteur suite à une rupture de tuyauteries, pourtant identifiées sensibles à la corrosion. Cet événement est très significatif de la maîtrise des processus de décision

en matière de management de projet. Il est détaillé au chapitre 12.Un dernier événement, portant sur la maîtrise des opérations d’exploitation « cœur de métier de conduite », classé au niveau 1 de l’échelle INES a également retenu mon attention. A la suite d'un arrêt de tranche court, lors du redémarrage, les vannes de l’alimentation de secours en eau des générateurs de vapeur n’ont pas été réouvertes, comme attendu, pendant près de 9 heures. L’analyse a posteriori de ce non-respect des spécifications techniques et de cet écart de lignage de circuit a mis en évidence un contrôle incomplet avant démarrage, un manque de rigueur dans la documentation et la tenue des journaux de bord, et une mauvaise communication dans les équipes et entre elles.A eux trois, ces événements mettent en lumière trois points de vigilance cruciaux pour tout exploitant nucléaire. Ils ont tous fait l’objet d’analyses approfondies par les sites concernés, de partage d’expérience avec tous les autres sites et avec le niveau central d’EDF Energy.

Les défis techniques du parc AGR

Une technologie unique et exigeante.

Le parc de réacteurs AGR fait aujourd’hui face à des défis techniques complexes, nécessitant la mobilisation d'une ingénierie très pointue. Cette ingénierie est de plus confrontée à des réacteurs uniques au monde, offrant peu de possibilités de synergies et de partage de retour d'expérience avec d'autres exploitants. Certaines problématiques techniques doivent être gérées sans les

équipementiers et les concepteurs qui ont cessé leur activité.Dans le domaine de l’étude et de la compréhension des phénomènes à enjeu de sûreté, je note que les travaux sur le graphite des cœurs des réacteurs AGR, la fissuration des tubes de chaudière et les dépôts de carbone sur le combustible de certains réacteurs se poursuivent. Ces efforts soulignent la nécessité pour l'industrie nucléaire britannique d'adopter une bonne planification stratégique de long terme en R&D. Le premier chargement dans ces réacteurs AGR du combustible de nouvelle conception robust fuel, aboutissement d'un long processus d'essais et d'instruction technique avant autorisation, est pour moi un point positif.Je considère par conséquent qu'il faudra veiller à maintenir une ingénierie de haut niveau, jusqu'au terme du fonctionnement de ces réacteurs, voire au-delà pour la déconstruction. Le maintien de ces compétences spécifiques passe par un dispositif de formation accrédité. L’accréditation récente des premiers départements (design authority, engineering, safety and regulation departments), première mondiale pour ces disciplines, répond à ce besoin.

Page 71: EDFMESS FR - inis.iaea.org

69

10

Le nouveau nucléaire, le projet EPR

Dans le cadre du projet de construction des nouvelles tranches nucléaires au Royaume-Uni, plusieurs dossiers importants ont été soumis à l’ONR et aux Autorités environnementales et d'aménagement du territoire cette année. Dans ce contexte, le vote quasi unanime du Parlement britannique en juillet 2011, confirmant l’option nucléaire dans le futur bouquet énergétique, a eu un écho positif considérable en interne.La structure du projet EPR me semble judicieuse : une entité unique est chargée du suivi des études, de la supervision des travaux, de la mise en service, et elle se prépare à devenir la future structure d'exploitation.De nombreuses caractéristiques de conception du réacteur de Sizewell B, seul réacteur REP construit au Royaume-Uni, sont considérées comme des points de référence technique (par exemple : source froide d’ultime secours, délai d’autonomie sur perte des sources électriques). Ces points importants font l’objet d’une étude approfondie en vue d’autoriser l’EPR au Royaume-Uni et de déterminer les actions à réaliser après Fukushima. D’éventuelles différences entre les EPR français et britanniques nécessiteront une communication efficace, mais je recommande avant tout de fixer le design de référence de l’EPR britannique au plus tôt.

Je regrette que le bon fonctionnement du projet soit encore gêné par des problèmes liés aux systèmes informatiques et à leur sécurité, rendant difficile l'accès aux plans et à la documentation, entre le Royaume-Uni, l’architecte ensemblier et les bureaux d'études en France.

CENG aux Etats-Unis

Les pratiques managériales

Je me rends chez CENG depuis plusieurs années. Le dispositif de pilotage de sa flotte de réacteurs (3 REP et 2 REB) est maintenant mature. Je constate que les indicateurs de performances du Tableau de bord d'excellence nucléaire sont utilisés efficacement. Le taux d'indisponibilité fortuite, inférieur à 2 %, confirme que CENG maîtrise bien la fiabilité de ses matériels et de ses opérations de maintenance, ce qui constitue un facteur déterminant et positif pour la sûreté. A un niveau plus détaillé, je note que l'indice de fiabilité des équipements (Equipment Reliability Index) créé par l’INPO, l’EPRI, et les fournisseurs pour permettre aux centrales de comparer leurs performances, est utilisé à bon escient. Son utilisation sur d’autres sites dans le Groupe a débuté et j’invite à en étendre l’usage.

Déjà mentionnée dans mes précédents rapports, la collaboration étroite et bénéfique de CENG avec

l’INPO, l’EPRI et le NEI est manifeste. A l'évidence, ces organismes assurent auprès des nombreuses compagnies d’électricité américaines, souvent de petite taille, un appui déterminant, comparable à celui dont bénéficient les centrales françaises d’EDF via les unités de soutien internes à l’entreprise.

CENG exploite deux réacteurs à eau bouillante (REB) sur le site de Nine Mile Point, sur les rives du lac Ontario. La conception de base du réacteur 1 de Nine Mile Point (NMP) est très similaire à celle des premières unités de Fukushima (réacteur de type Mark-One). Il m’était donc important de recevoir des informations sur cette unité et sur les modifications et améliorations majeures qui lui ont été apportées depuis 1979. L’analyse détaillée des enseignements directs pour le site de Nine Mile Point n’entre pas dans le cadre de ce rapport, mais j'ai constaté avec satisfaction que CENG était très proactif au sein de l’industrie nucléaire américaine dans les démarches et actions post Fukushima.Plus généralement, pour les actions visant à intégrer le retour d’expérience du secteur (notamment TMI, les attentats du 11 septembre 2001 et les nouvelles connaissances), je relève que l'amélioration continue des dispositions de sûreté (design, organisation, formations) ne repose pas uniquement sur l’évolution des règles formulées par la NRC. De nombreuses actions ou des modifications significatives pour la sûreté ont été réalisées à l’initiative des exploitants américains, souvent soumises à la NRC par le NEI. Sans procéder d’une démarche aussi méthodique, elles sont à rapprocher des démarches française et britannique de réexamens décennaux de sûreté.

Page 72: EDFMESS FR - inis.iaea.org

70

10

Mes observations sur site

La centrale de Ginna mise en service en 1968 est le réacteur REP qui cumule la plus longue durée de fonctionnement aux Etats-Unis. Son autorisation de fonctionnement a récemment été étendue jusqu’en 2029 (60 ans). C’est un exemple pour le parc de réacteurs REP du groupe EDF, engagé dans l’instruction de la durée de fonctionnement de ses réacteurs.J’ai assisté, au Comité trimestriel d'examen de la sûreté nucléaire (NSRB62), présidé par une personnalité qualifiée externe à la compagnie. Ce comité est précédé, la veille, par une visite du site et par une série d’entretiens avec le personnel de la centrale, dans chacun des grands domaines de l’exploitation. Des personnes interviewées ont même été amenées à s’exprimer sur leur propre niveau de performance. Ces entretiens ont permis d'obtenir des indications sur la manière dont le personnel détermine,

décrit, voire mesure, sa performance. Le besoin de fixer des objectifs à la fois adaptés à tous les niveaux et cohérents a été rappelé, avec pragmatisme.Je note aussi que les trois sites de CENG s’entraident efficacement pendant les périodes d’arrêts de tranche, par des renforts en personnel très significatifs (environ 50 % du personnel du service de maintenance).

Savoir où l'on veut aller

J’ai aussi observé à plusieurs reprises la direction de CENG et le management local exprimer leurs attentes et leurs exigences, demander au personnel rencontré d'expliquer ses actions et de situer les performances de sa centrale. Cette façon de faire est selon moi très pertinente.

Le responsable de la sûreté de CENG dispose d’un diagnostic solide et les priorités d’amélioration sont claires : les consignations, le lignage des circuits et la sécurité au travail. A mes yeux, tout cela indique que la sûreté nucléaire est constamment renforcée et remise en question.Le management est mobilisé et explicite en matière d’exigences et de sûreté. Un cadre dirigeant a ainsi évoqué en public, au vu de certains dysfonctionnements, le besoin qu’il ressentait de passer de

« dissatisfaction°» à « intolerance », donc que les choses devaient changer. Je pense que ce type d’attitude gagnerait à être développé ailleurs dans le Groupe.Je note la mise en place, dans chaque réunion opérationnelle importante de CENG, d’une fonction de Nuclear Safety Advocate : la personne identifiée pour tenir ce rôle est le promoteur de la sûreté nucléaire dans la réunion. Cette pratique responsabilisante me paraît digne d’intérêt et présente des complémentarités avec le rôle assuré par l’ingénieur sûreté en France sur les sites.

La prise de décision

Dans l’une des centrales de CENG, les équipes d’exploitation ont rencontré des difficultés pour caractériser l’origine de performances irrégulières en essais, d’un matériel important pour la sûreté (une turbopompe de secours). Le soutien technique apporté aux équipes de conduite par l’ingénierie locale s’est polarisé sur une seule hypothèse technique, qui s’est révélée erronée. Le dialogue entre l’ingénierie et le service conduite n’a pas été suffisant, au point que l’autorité de sûreté a remis en question la capacité du site à faire face à ce type de problème.La direction de CENG a alors engagé une analyse approfondie, et sans concession, de l’organisation et de la chaîne de prise de décision en cas d’aléa technique. Cette analyse a conduit à instaurer un nouveau processus, integrated problem identification and resolution process, précisant les responsabilités des différents acteurs et services autour de la conduite. Ce travail semble être efficace et apprécié. Il est maintenant utilisé sur tous les sites de CENG.

62 NSRB : Nuclear Safety Review Board.

Page 73: EDFMESS FR - inis.iaea.org

71

10

Cet événement illustre bien la difficulté de relier efficacement la prise de décision opérationnelle, normalement contrôlée par la conduite, aux nombreux autres processus, traitant notamment de la surveillance des non-conformités d’exploitation, des critères d'opérabilité, des actions correctives, de l'identification des mécanismes de défaillance, de la gestion du risque, de la performance humaine, de la fiabilité des matériels (AP 913) et des analyses à la suite d'incidents. La méthode, autant que la solution retenue, mérite d’être méditée par les exploitants nucléaires tant cette question traverse toutes les organisations existantes. Chaque organisation doit à un moment donné ou à un autre, être réinterrogée par ceux qui la mettent en œuvre. C’est un exercice parfois délicat, mais qui témoigne toujours d’un bon niveau de culture de sûreté.

Des événements marquants

Durant l'été 2011, un séisme a touché la côte est des Etats-Unis. Ressenti sur l’un des sites de CENG, mais d’intensité très inférieure aux études de conception, il n’a pas affecté le fonctionnement des réacteurs et n’a eu aucune conséquence sur la sûreté. Classé au niveau 0 de l’échelle INES, il a toutefois mis en évidence les difficultés rencontrées par les opérateurs pour déterminer, dans les délais requis, l’intensité du séisme depuis la salle de commande, comme le demandent les procédures qui déclenchent l’organisation de crise. Les mesures disponibles nécessitent en effet, pour être correctement interprétées, des connaissances spécifiques, rarement utilisées par les opérateurs. Cette expérience mérite d’être connue sur tous les sites du Groupe pour s’assurer de la bonne maîtrise de ce type de procédures, rarement mises en œuvre mais à fort enjeu.

Autre agression d’origine naturelle, l’ouragan Irène a affecté un site nucléaire de CENG, en septembre 2011. Le détachement d’une plaque métallique provenant du bardage de la salle des machines a provoqué un défaut électrique sur un transformateur, conduisant à l’arrêt automatique de l’un des deux réacteurs. Cet événement, classé au niveau 0 de l’échelle INES, a été bien géré par les équipes. Les investigations ont révélé que le bardage n’était pas installé conformément à la conception et que les schémas de conception n’étaient pas suffisamment explicites. Le retour d’expérience des précédents dommages dus au vent n’a pas été tiré pour les bardages similaires.

TNPJVC en Chine

J’ai visité cette année le chantier des deux réacteurs EPR de Taishan, où j’ai mesuré l’avancement de leur construction, le réacteur 1 ayant désormais un planning très proche de celui de l’EPR de Flamanville 3. La pose du dôme métallique du bâtiment réacteur, pendant ma visite, représente une étape symbolique de cet immense chantier mobilisant plus de 10 000 personnes. J’ai également rencontré les équipes conduisant ce projet, qui m’ont notamment présenté l’évolution du bureau d'études commun AREVA-CGNPC (le chaudiériste et l’ingénierie du client). Ce nouveau bureau s’est révélé efficace car il a permis

de produire les plans de montage dans les délais, point qui avait attiré mon attention lors de ma dernière visite.Les équipes chinoises de TNPJVC et de CGNPC s’appuient sur la forte expérience accumulée

par leurs équipes d’essais et d’exploitation lors des mises en service des réacteurs de type

CPR 1000. Je me félicite du nombre important de futurs exploitants chinois de Taishan accueillis et formés dans les centrales en France.Par ailleurs, je note, comme pour Flamanville 3, la décision de qualifier le matériel aux conséquences des accidents graves avant le démarrage, en s’appuyant au maximum sur la démarche d’EDF SA.Au siège de CGNPC, à Shenzhen, je me suis fait présenter le projet de création d’une filière indépendante de sûreté, au niveau des sites et du parc, dans l’esprit de ce qui existe à EDF SA. Cette démarche renforce la sûreté dans un contexte de croissance rapide du nombre de réacteurs en exploitation. Je me félicite des échanges et de la coopération active engagée avec l’Inspection nucléaire de la DPN, allant jusqu’à la formation des inspecteurs chinois.

Page 74: EDFMESS FR - inis.iaea.org

72

10

Plus généralement, les compagnies chinoises deviennent rapidement des acteurs majeurs du nucléaire dans le monde, avec 14 réacteurs en exploitation et 28 en construction fin 2011. A Shenzhen, en octobre 2011, l’Assemblée générale biennale de WANO a symbolisé ce nouveau statut. Je considère qu’il est essentiel pour le groupe EDF de continuer à rechercher le maximum de partage d’expérience de construction, de mise en service et d’exploitation avec les compagnies chinoises, pour promouvoir un haut niveau de sûreté nucléaire. Je note l’ouverture en 2011 à Shenzhen d’un bureau EDF, au plus près des industriels et des exploitants, pour favoriser le retour d’expérience et les échanges.

En conclusion, je me félicite du développement de synergies de plus en plus actives entres les différentes entités nucléaires du Groupe. Cette diversité est une richesse qui peut être une source de progrès significatifs pour la sûreté nucléaire. J’encourage le développement de ces échanges, sur le nucléaire existant et sur les projets de nouveaux réacteurs. Chacun a en effet beaucoup à gagner à s’ouvrir aux pratiques des autres, quelle que soit son histoire.

Page 75: EDFMESS FR - inis.iaea.org

73

11

LES MISSIONS A L’ETRANGER

Le partage des expériences et des méthodes avec des opérateurs

nucléaires à l’étranger permet de réinterroger ses propres

pratiques ou son propre regard sur son domaine d’activité.

Les échanges avec l’INPO, acteur de référence mondial pour

le contrôle et l’évaluation des performances, sont également

indispensables pour challenger les méthodes du Groupe et

apprécier le contrôle exercé sur les sites de CENG.

L’INPO à Atlanta

J'ai eu le privilège de passer, avec plusieurs responsables de l’état-major, des moments à la fois de qualité et utiles car l’INPO joue un rôle majeur dans l'industrie de l'énergie nucléaire aux Etats-Unis. Son autorité s'assoit sur une implication pleine et entière des présidents des compagnies membres. Il bénéficie d'un financement adapté et son personnel est majoritairement composé (80 %) de salariés permanents. Il joue un rôle essentiel auprès des nombreux opérateurs nucléaires américains qui n’ont pas de fonctions corporate substantielles en interne. Avec plus de 100 réacteurs aux Etats-Unis, auxquels viennent s'ajouter ceux des membres hors Etats-Unis, le personnel de l’INPO évalue et

accompagne un groupe de centrales et d'organisations vaste et diversifié. Il me semble que sa valeur réside en grande partie dans sa base de connaissances professionnelles, son engagement et dans le fait qu'il n'est pas issu des compagnies qu’il évalue. L’INPO porte un regard externe et ne subit ni

contraintes hiérarchiques ni pressions internes de la part des opérateurs évalués. Il lui est donc plus facile de remettre en question la pertinence ou la crédibilité des directives d’une compagnie.

Page 76: EDFMESS FR - inis.iaea.org

74

11

Ses principales fonctions consistent à réaliser des évaluations (34 par an), accréditer les programmes de formation, assurer le retour d’expérience et proposer un accompagnement aux centrales américaines.

Les évaluations

Les évaluations portent sur les compagnies et sur les sites. Après l'évaluation d'un site, une note de performance globale lui est attribuée en toute confidentialité. Cette note est discutée entre les CEO63 de l’INPO et de l'entreprise concernée.

L’INPO organise une conférence annuelle des dirigeants durant laquelle les signaux précurseurs sont examinés en séance restreinte. L’an dernier, il s'agissait de l'augmentation du nombre d'événements significatifs d'exploitation (SOER 10-2) et de l’augmentation du nombre d’arrêts automatiques ou manuels du réacteur, des faiblesses affectant les performances des équipes de conduite et des défaillances des groupes Diesel de secours. Les meilleurs sites sont reconnus et les dirigeants des entreprises aux moindres performances sont invités à expliquer à leurs pairs les mesures prises pour réduire les risques que leur entreprise représente pour l'industrie nucléaire.

L’INPO m'a présenté une excellente synthèse des types de réaction des centrales en cas de baisse des performances, donc du classement INPO :

celles qui renouent immédiatement avec l'excellence, font preuve, dès les premiers signes de dégradation, d’une volonté de reconquête par la mise en œuvre de mesures énergiques et acceptent les remontées d’information critiques, quelle qu'en soit la source. Elles expriment leur insatisfaction vis-à-vis de la situation, sans se chercher d’excuses et s’impliquent à tous les niveaux.celles dont la situation continue de se dégrader sont gérées par des dirigeants qui tardent à reconnaître et à accepter l'existence d'un problème, qui tendent à rationaliser plutôt qu’à résoudre les problèmes, hésitent à opérer des changements décisifs. Ils sont sur la défensive et se réfèrent à leurs performances passées et non aux meilleurs élèves du secteur.

Je pense que nous pouvons en tirer des enseignements utiles.De plus l’INPO propose une démarche d'assistance progressive, avec au terme de chaque évaluation, des missions d’appui ciblées pour chaque axe de progrès identifié. Pour une centrale en difficulté, il crée une commission de supervision spéciale, qui rencontre son CEO tous les six mois.Je note une observation intéressante issue du retour d’expérience américain : « Les centrales les moins performantes ont généralement un problème de pépinière au sein de la conduite, si bien que les autres services manquent de compétences process. Elles ont aussi souvent subi des coupes budgétaires.» J’ai également abordé les actions post-Fukushima avec mes interlocuteurs de l’équipe d’évaluation (cf. chapitre 4). Comme WANO, l’INPO doit désormais inclure les aspects hors dimensionnement dans les objectifs et les critères de performance des évaluations. A ce jour, les directives de gestion des accidents graves (SAMG64) ne couvrent pas les piscines d’entreposage de combustibles et devront donc être complétées.

Les indicateurs de performances

Pour la plupart des indicateurs, les performances de l'industrie américaine plafonnent ou stagnent et « la forte propension aux événements significatifs relevée début 2010 se maintient ».Les niveaux de disponibilité des groupes Diesel de secours n'ont jamais été aussi bas aux États-Unis. En s’appuyant sur leurs études probabilistes, je constate que les exploitants américains multiplient les interventions volontaires de maintenance sur les groupes Diesel pendant le fonctionnement des unités. Dans la plupart des autres pays, cette pratique n'est pas autorisée ou elle est très réglementée par les

63 CEO : Chief Executive Officer.64 SAMG : Severe Accident Management Guidelines.

Page 77: EDFMESS FR - inis.iaea.org

75

11

autorités de sûreté. Je pense que ce sujet mériterait un débat international plus poussé, particulièrement au lendemain de Fukushima.Le rendement du combustible est un domaine où l’INPO travaille en étroite liaison avec les fournisseurs, au travers d’un comité consultatif, le SPAC65. En 2005-2006, seuls 35 % des réacteurs n'avaient enregistré aucun défaut d’étanchéité du gainage sur une période de trois ans. Cette proportion atteint aujourd'hui 60 % (56,5 % pour les REP). Fin 2010, 92 % des centrales fonctionnaient sans aucun défaut. Il s'agit là d'un bon exemple de collaboration entre les exploitants et les fournisseurs en vue d'améliorer la sûreté nucléaire.

Les axes de progrès de l’INPO

L’INPO cherchera, avant validation de tous les constats, à anticiper au maximum la prévention des incidents, par des discussions plus précoces avec les sites. Les observations des équipes de conduite seront aussi renforcées.L’INPO va désormais se concentrer sur les fondamentaux de l'opérateur. Son récent rapport d'événement 11-3 vise à faire évoluer la situation vers une « prévalence du fond sur la forme », de ce que l'on fait sur le processus à proprement parler.L’INPO émet des recommandations utiles sur des thèmes comme la prise de décision opérationnelle, la culture de sûreté nucléaire et le leadership dynamique.Avec le MIT66, il propose aux CEO et à leurs cadres dirigeants nouveaux venus dans le secteur nucléaire, un cursus de formation largement reconnu.On m'a présenté un élément très intéressant d’un autre cursus : l'outil d'apprentissage en ligne NANTel. Egalement accessible aux prestataires, il couvre des aspects comme les outils de performance humaine et les FME (foreign material exclusion, prévention contre les corps migrants). L’INPO commence à travailler directement avec les prestataires et non plus uniquement avec les opérateurs. Plus de 95 000 connexions au site NANTeL ont été enregistrées.Une action est également en cours pour encourager les prestataires à participer davantage à l’enrichissement des bases de retour d’expérience, y compris pendant la construction.L’INPO réalise des vidéos des grands événements marquants de l’industrie nucléaire. Il s'agit d'une excellente initiative, en particulier pour la nouvelle génération de personnels. Je suggère qu’ils soient utilisés partout dans le groupe EDF.

Le nucléaire en Afrique du Sud

En Afrique du Sud, la compagnie ESKOM67, responsable de la production d’électricité pour l’ensemble du pays, essentiellement d’origine charbonnière, exploite le seul site de production nucléaire du continent africain et fait appel, depuis l’origine, aux revues de pairs et à l'assistance de WANO de l’INPO et de l’AIEA. Elle a également conclu un contrat d'assistance technique avec EDF SA, car les centrales de référence des deux unités REP 900 MWe de Koeberg sont françaises. Ces unités sont en service depuis 1984, Lors de ma visite, j’ai constaté que l'inter-comparaison et la recherche des regards internationaux

s’y poursuivent bien. Les ingénieurs d’EDF présents sur le site sont bien intégrés aux équipes et à la structure d’ESKOM. Les pratiques d'EDF constituent une référence pour la conception des modifications et d'exploitation. Ces dernières années, plusieurs modifications importantes ont rendu les installations sud-africaines conformes aux standards68 de l’AIEA (INSAG-12) pour les centrales neuves. ESKOM intègre les études probabilistes de sûreté en exploitation dans les processus de définition des priorités et de prise de décision.

Le service de conduite est bien gréé en personnel et possède une solide expérience. Le programme de

formation est directement accrédité par l’INPO. Les services opérationnels se comportent comme des

65 SPAC : Suppliers Participants Advisory Committee.66 MIT : Massachusetts Institute of Technology.67 ESKOM : Société d’Electricité Sud-Africaine.68 INSAG-12 : Basic Safety Principles for Nuclear Power Plants.

Page 78: EDFMESS FR - inis.iaea.org

76

11

propriétaires de la formation. Les mises à jour du simulateur et des différentes procédures associées sont bien coordonnées avec les modifications implantées sur les réacteurs de la centrale. Les personnels de

conduite s’entraînent sur le simulateur une semaine sur six, selon le roulement de travail des équipes de conduite, conçu pour faire face à cette exigence.

Je suis impressionné par les moyens et les ressources dont dispose le département formation, avec près de 70 instructeurs travaillant dans des installations modernes, ouvert sur l’extérieur (lycées, en particulier) et intégrant complètement les entreprises prestataires.De nombreux personnels supplémentaires ont été recrutés et sont en formation, ou occupent des postes supplémentaires, prévus dans l'optique de la construction d’une nouvelle centrale.D'importants investissements ont été réalisés pour le maintien en état de la centrale. Le remplacement des générateurs de vapeur et une augmentation de la puissance thermique de la chaudière sont en cours de préparation. Le bon état des générateurs de vapeur de Koeberg, derniers de leur génération encore en service, mérite d'être signalé. Il atteste d’un suivi des paramètres chimiques en exploitation de haut niveau, réalisé tout au long de leurs 27 années d’exploitation.Je relève que l’un des réacteurs présente d’excellents résultats de fiabilité pour la période récente, avec un coefficient d’indisponibilité fortuite inférieur à 2 % ; l'autre a été touché par un problème de conception de sa turbine, récemment modifiée, dégradant sensiblement ses résultats de fiabilité. De plus, une contamination de l'air dans le bâtiment du réacteur, pendant un arrêt de tranche, a abouti à une contamination interne mineure de 247 personnes. Les similitudes avec un événement survenu

précédemment à Tricastin, non pris en compte à Koeberg, montrent l’importance du retour

d’expérience.

Je constate que l’un des défis d’ESKOM est de parvenir à préparer et à planifier ses arrêts de

tranche suffisamment à l'avance pour disposer des lots de travaux et des pièces de rechange en temps voulu.Je relève que les responsables des activités nucléaires d’ESKOM ont fait appel à WANO pour organiser des revues de pairs sur leur centrale et bénéficier d’un regard externe sur les différents types

d’organisation envisagés pour piloter la branche nucléaire au sein d’ESKOM, qui exploite très majoritairement des centrales thermiques à flamme. Cet aspect organisationnel a été, à juste titre, considéré comme un enjeu essentiel pour garantir la sûreté nucléaire dans ce type d’environnement.Je perçois également chez mes interlocuteurs une forte préoccupation touchant au contrôle et à la surveillance des constructeurs et des prestataires ainsi qu’aux méthodes à adopter pour induire des comportements adaptés et garantir une qualité régulière des travaux.

Sur le plan du management de la sûreté, je constate qu’ESKOM cherche à perfectionner sa filière

indépendante de sûreté en s’inspirant des principes et des pratiques françaises. Un ingénieur sûreté français est détaché à Koeberg dans cet objectif. Les comités de sûreté nucléaire me semblent bien structurés et l’inter-comparaison des performances et des pratiques y est assurée régulièrement (France et Etats-Unis principalement).

Enfin, j’ai rencontré l’Autorité de sûreté sud-africaine, au niveau local et à Pretoria. Comme elle est confrontée à la difficulté propre aux autorités ayant peu d’installations de production nucléaire à contrôler, je note qu'elle entretient des relations bilatérales avec d’autres autorités de sûreté, dont l’ASN française et la NRC américaine.

Page 79: EDFMESS FR - inis.iaea.org

77

12

DES EVENEMENTS

DANS LE GROUPE EDF

Les événements retenus portent sur les grands métiers de

l’exploitation : maintenance, conduite, ingénierie et projets.

Ils mettent tous en lumière la nécessité de consolider sans

relâche les lignes de défense organisationnelles et humaines.

12.1 EDF SA : une usure prématurée de coussinets sur les diesels de secours

Chaque réacteur nucléaire est équipé de deux groupes électrogènes diesels redondants, assurant l’alimentation électrique autonome des matériels les plus importants pour la sûreté, en cas de perte des sources électriques externes (lignes très haute tension). De plus, en cas de défaillance de ces deux diesels de secours, chaque site dispose d’un groupe électrogène d’ultime secours rapidement raccordable à la tranche en avarie (diesel sur la plupart des sites du palier 900 MWe, turbines à combustion sur les autres sites).Après l’arrêt de la fabrication des coussinets de tête de bielle des groupes diesels électrogènes de secours du palier 900 MWe par le fabricant historique, une nouvelle société, leader dans ce domaine, reprend cette fabrication en 2002.La société diéséliste chargée de la fourniture et de l’entretien des diesels, demande alors à ce fournisseur de fabriquer des coussinets identiques aux précédents. La société diéséliste n’est pas non plus celle qui a assuré la conception et la fourniture initiale de ces matériels.

Page 80: EDFMESS FR - inis.iaea.org

78

12

Après des avaries sur 4 de ces groupes entre 2008 et 2009 et informé d’un retour d’expérience international, ce diéséliste doute du comportement dans le temps de ces pièces et propose une évolution de leur conception.Sans attendre la fin des essais de qualification en usine (18 mois, de septembre 2009 à mars 2010), les coussinets de nouvelle génération, les seuls disponibles, sont montés fin 2009 sur les 16 groupes jusque-là équipés de coussinets de première génération. L’ASN en est informée. A la fin de ces essais, au vu de leur très bonne tenue dans le temps, la qualification de ces nouveaux coussinets est prononcée par le diéséliste et par EDF.Le 22 octobre 2010, le diesel du groupe électrogène d’ultime secours d’un CNPE, équipé de ces coussinets de nouvelle génération, subit une avarie lors d’essais de requalification après maintenance, affectant trois coussinets. L’expertise approfondie révèlera l’usure anormale de 8 autres coussinets du même type.Le 28 novembre 2010, lors d’un essai périodique, le diesel d’un groupe électrogène de secours d’un autre CNPE subit une importante dégradation interne au niveau du moteur.Une cellule technique est mise en place début 2011, à la DPN, et des examens sont lancés sur d’autres sites potentiellement touchés. Ces examens confirment l’existence d’usures prématurées sur d’autres coussinets.L’ingénierie nationale conclut à la présence d’un écart potentiellement générique sur l’ensemble des machines du palier 900 MWe équipées de ces coussinets. L’ASN est immédiatement informée de ces résultats d’expertise, et un événement significatif pour la sûreté, générique, est déclaré au niveau 1 de l’échelle INES.Toutefois, dans l’un des 8 CNPE affectés, deux tranches voient leurs deux diesels de secours affectés, ainsi que le diesel d’ultime secours du site. Ce qui conduit à déclarer pour ce site un événement

significatif de niveau 2 sur l’échelle INES.

La caractérisation précise du défaut matériel à l’origine de ces avaries a conduit à la préconisation rapide de mesures correctives, portant sur un rodage amélioré de ces coussinets en usine, sur l’augmentation de la pression de service du circuit d’huile et sur le changement du type d’huile. En parallèle, le remplacement de ces matériels défectueux est réalisé en trois mois sur l’ensemble des machines concernées (près d’une trentaine).

A court terme, des dispositions de suivi renforcé du comportement de ces matériels en exploitation ont également été mises en œuvre (analyses d’huile plus fréquentes) ainsi qu’un programme de maintenance complémentaire. Une machine témoin, équipée des nouveaux coussinets à rodage amélioré, a été soumise à des essais spécifiques d’endurance.

Enfin, la définition de la solution pérenne fait l’objet de discussions techniques soutenues et régulières avec l’autorité de sûreté.

En conclusion, je retiens d’abord de cet événement sérieux, qui est resté sans conséquence directe sur la sûreté des réacteurs concernés, la réactivité des dispositions prises dès que le défaut a pu être suffisamment cerné. Le rôle des équipes nationales (UNIE, UTO) en pareil cas, face à un défaut potentiellement générique pouvant mettre en difficulté un grand nombre de tranches en même temps, me semble bien assuré, qui plus est face à un diagnostic technique difficile à établir. La qualité

de l’instruction de tels dossiers avec l’ASN, et son appui technique l’IRSN, nécessite ce niveau d’engagement, que les sites seuls ne peuvent pas porter. Le maintien, voire le renforcement, de compétences clés dans ces équipes est donc de la plus haute importance.Je retiens ensuite, de mes échanges avec les spécialistes chargés du traitement de cet aléa, les difficultés rencontrées par le diéséliste pour se mobiliser avec la réactivité voulue et dégager les compétences attendues. Celles-ci avaient été sensiblement affaiblies par le temps et par les évolutions

industrielles du secteur.Concernant des matériels aussi importants pour la sûreté, ces points méritent à mon sens la plus grande attention et la plus grande vigilance, tant ils peuvent fragiliser à terme les solutions techniques les mieux conçues à l’origine.

Page 81: EDFMESS FR - inis.iaea.org

79

12

12.2 EDF SA : un incident de pilotage et un arrêt automatique de réacteur

Un événement survenu en 2011 sur une tranche 900 MWe a retenu mon attention, même s’il est resté sans conséquence directe sur la sûreté de l’installation et a été classé au niveau 1 de l’échelle INES. Il met en effet en lumière une préparation insuffisante d’un transitoire normal d’exploitation (baisse de puissance), un manque d’attitude interrogative face à l’imprévu et une action de pilotage inadéquate en temps réel.

Les faits

La puissance de la tranche doit être abaissée à 65 % de sa puissance nominale (Pn) pour permettre la réparation de l’une des deux turbo-pompes alimentaires (TPA), sur la partie secondaire de l’installation. Cette intervention est programmée de nuit, les travaux de remise en état devant durer environ 14 heures.Une consigne de pilotage temporaire, écrite par l’équipe d’appui du service conduite pour la préparation de cet abaissement de puissance (transitoire), est à la disposition des opérateurs en salle de commande. Cette consigne n’a pas fait l’objet d’une préparation conjointe avec l’ingénierie cœur-combustible, locale ou nationale. Elle spécifie le maintien de l’insertion des grappes de contrôle à une cote déterminée (180 pas). Pendant la durée du fonctionnement à la puissance 65 % Pn, les opérateurs suivent scrupuleusement cette consigne de pilotage à un paramètre appelé Delta I69.

Une fois réparée la turbo-pompe, l’équipe de conduite du matin engage la reprise de puissance, selon la procédure normale. A 78 % Pn, la remontée de charge est interrompue par l’apparition d’un défaut sur le contrôle commande. L’astreinte du service automatismes intervient et traite cette panne.

A la relève entre l’équipe du matin et celle de l'après-midi, l’opérateur chargé de la partie primaire de l’installation estime que le Delta I est à une valeur trop élevée et demande des explications. On lui répond qu'il faut respecter la consigne temporaire de pilotage. Pendant cette relève de l'après-midi, l’astreinte du service informatique est sollicitée pour dépanner le système d’aide au pilotage rendu inopérant, suite à un défaut informatique fortuit. Ce système délivre à l’opérateur des informations sur le niveau et la tendance de l’effet Xénon dans le réacteur, paramètre de pilotage influent. Outil d’aide très visuel, il n’est pas formellement requis pour reprendre la montée de puissance, l’opérateur disposant d’autres moyens d’information. L'indisponibilité de ce système d’aide au pilotage est évoquée lors du briefing de début de quart par le chef d’exploitation délégué (CED). Elle n'est pas estimée "bloquante" par le CED, ni par le chef d’exploitation (CE) qui autorise la reprise de puissance dans ces conditions de surveillance abaissées, considérant l’expérience des opérateurs, comprise entre 5 et 20 ans. Cette indisponibilité ne fait pas l’objet d’une analyse de risque suffisante entre les opérateurs, conscients de cette difficulté, le CED et le CE.

La reprise de puissance est donc entamée dans ces conditions à 13h30. L’opérateur réacteur calcule alors, selon les procédures normales, le volume d’eau à injecter dans le circuit primaire pour porter la puissance de 78 % Pn à 100 % Pn. Il procède à des injections d’eau successives, et progressives, mais sans disposer d’informations sur l’évolution de l’effet Xénon dans le réacteur (effet alors décroissant). Il se concentre sur la surveillance des paramètres clés : puissance, température moyenne du circuit primaire, Delta I.A 14h30, la puissance est amenée à 99 % Pn, les opérateurs gardant une marge de 1 % Pn pour prendre en compte l'indisponibilité du système d’aide au pilotage. Fidèle à la stratégie arrêtée, l’opérateur réacteur poursuit l’injection d’eau pour contrôler la température moyenne du réacteur et ramener les grappes de contrôle à une position moins extraite.

69 Delta I : Valeur de déséquilibre axial de puissance très positive (+ 10 %).

Page 82: EDFMESS FR - inis.iaea.org

80

12

La décroissance, alors rapide, du Xénon continue d’apporter de la réactivité ce qui, combiné à l’effet de l’injection d’eau, entraîne automatiquement une insertion importante des grappes de contrôle, très efficaces, qui font ainsi décroître très rapidement le Delta I. L’apparition de l’alarme « sortie du domaine autorisé » conduit à appeler le CE en salle de commande.A 14h51, les opérateurs appliquent la procédure de conduite couvrant cette situation et baissent la puissance du réacteur de 1,5 % Pn, tout en injectant du bore dans le circuit primaire et en extrayant les grappes de contrôle.A 14h55, les protections automatiques DeltaT-SP prévues à cet effet réduisent la puissance puis arrêtent le réacteur, par chute des grappes de contrôle (arrêt quasi instantané). Les consignes de conduite prévues en pareil cas sont ensuite appliquées sans difficulté. Après analyse des causes de cet arrêt automatique et remise en service du système d’aide au pilotage, le réacteur est remis en service le lendemain matin à 6h18 (divergence) et atteint sa pleine puissance dans la nuit suivante.

Une préparation insuffisante d’un transitoire inhabituel

Ce réacteur est exploité en régime permanent à 100 % Pn. Les transitoires de baisse de puissance prolongée de ce type sont très rares sur cette tranche, ce qui aurait dû conduire à une préparation renforcée, associant toutes les compétences concernées, dont l’ingénieur expert cœur-combustible (IECC) du site. La réalisation d’une simulation de ce transitoire, facile à mettre en œuvre avec les outils dont disposent aujourd’hui les ingénieurs, aurait permis de définir la bonne stratégie de pilotage.

Une attitude trop peu interrogative face à l’imprévu

Les interrogations justifiées de l’opérateur prenant son service, sur la valeur inhabituelle du Delta I, ne sont pas menées assez loin et l’équipe en reste à la conformité à la consigne temporaire. L’analyse du risque induit par l’indisponibilité du système d’aide au pilotage ne va pas assez loin, le partage d’informations est insuffisant, le CE et le CED ne prenant par exemple pas toute la mesure de cette indisponibilité. Un pré-job briefing bien conduit aurait probablement permis d’identifier tous ces points.

Une action de pilotage inadéquate

L’injection d’eau dans le circuit primaire, pour porter la puissance de 78 à 100 % Pn, apporte de la réactivité dans le cœur – c’est bien l’objectif - à un moment où la décroissance de l’effet Xénon apporte également de la réactivité. Cette action inappropriée illustre l’absence de compréhension par les opérateurs, le CED et le CE du sens de l’évolution de l’effet Xénon. L’utilisation du logiciel informatique de secours, « Xénon 0D », consultable via intranet en salle de commande, aurait pu partiellement compenser l’indisponibilité de ce système et apporter les premiers éléments de compréhension.

Je relève que sur ces trois temps différents, la connaissance insuffisante des phénomènes physiques en jeu apparaît à chaque fois.

En conclusion, cet événement souligne le besoin de renforcer le niveau de connaissance chez les exploitants des phénomènes physiques complexes en jeu dans le pilotage des réacteurs, par des formations mieux adaptées. La présence sur chaque site d’ingénieurs experts cœur-combustible constitue désormais une force pour augmenter les compétences dans ce domaine et permettre une meilleure utilisation des outils de simulation.Les pratiques de fiabilisation, ici le pré-job briefing, l’attitude interrogative et, quand il le faut, la capacité à dire « stop », sont autant de principes actifs de culture de sûreté et de lignes de défense fortes face à l’imprévu.

Page 83: EDFMESS FR - inis.iaea.org

81

12

12.3 EDF Energy : la défaillance de tuyauteries en fonte

Les faits

Les deux réacteurs d’une centrale AGR sont en production lorsqu'une perte de pression est détectée dans les circuits d'eau de mer qui assurent le refroidissement des auxiliaires de l’un d’entre eux. Un examen visuel permet d’identifier rapidement une fuite importante au niveau d'une tuyauterie enterrée. Conformément aux procédures, les opérateurs réduisent la puissance et effectuent un basculement du circuit sur l'unité voisine pour assurer le refroidissement. L’unité en question est ensuite mise à l’arrêt pour réparation. Les opérateurs n’ont eu aucune difficulté particulière à maintenir la conformité aux spécifications techniques. La fuite était directement imputable à une corrosion interne, qui avait traversé les parois de la tuyauterie en fonte.

Le contexte

Les tuyauteries en fonte étaient généralement utilisées dans les circuits d'eau de mer des réacteurs AGR britanniques. Le mécanisme de corrosion associé à l'eau de mer est parfaitement connu et documenté. Les entreprises propriétaires des installations avant EDF Energy n'ont pas suffisamment anticipé le remplacement et la remise en état de ces circuits. Plusieurs défaillances en exploitation avaient déjà affecté des unités AGR notamment en 2004 et 2006. Cet incident récent est donc un événement récurrent. Depuis 2003 un vaste programme d'inspection et de remplacement est engagé dans l'ensemble de la flotte AGR, constituée de 14 réacteurs. Les tronçons de tuyauterie directement enterrés dans le sol sont les derniers à être traités.

Les causes

Les analyses menées après cet événement ont fait ressortir les points suivants :

les nombreux groupes de suivi pour superviser le remplacement des tuyauteries défectueuses ont fait preuve d'une certaine tolérance vis-à-vis de la lenteur des travaux,le projet de remplacement de ces tuyauteries n'a pas atteint les résultats escomptés ; il présentait des insuffisances en termes de définition de la conception et de l'organisation requises. Le projet a pris beaucoup de retard, avec de fréquents changements de chef de projet (sept en cinq ans, dont quatre au cours des 12 derniers mois),le processus d’analyse et de gestion des risques existants n’a pas permis de neutraliser les risques en temps opportun,dans le cas présent, le programme d'inspection des tuyauteries n'avait donné aucune alerte quant à la défaillance en question et donc ne donnait pas de possibilité de prévoir un plan d'intervention.

Mes commentaires

Après les incidents dans les centrales, ce sont souvent les opérateurs de salle de commande et les exploitants de l’installation qui se retrouvent sur la sellette. Toutefois, dans le cas présent, je considère que nous devons nous intéresser, en amont, à la façon dont la direction et les ingénieries d’appui perçoivent les risques, établissent les priorités et, surtout, prennent des mesures pour minimiser les risques. Ici, un phénomène de vieillissement avéré touche des circuits de refroidissement importants pour la sûreté du réacteur. Le retour d’expérience en exploitation est bien connu, tant en interne qu’à l'international. La politique de remplacement de ces tuyauteries est définie et les ressources financières nécessaires disponibles. Le projet de remplacement des tuyaux en fonte de la centrale a débuté en 2003. On dénombre au moins six instances de projet, d'investissement, de liaison et autres chargées d’évaluer la progression du projet et de définir les priorités. Toutes ces instances se sont inquiétées de la lenteur des opérations de remplacement mais aucune n'était suffisamment influente pour accélérer les travaux. Le projet de remplacement de ces tuyauteries s’inscrivait dans une série de projets majeurs qui

Page 84: EDFMESS FR - inis.iaea.org

82

12

ont tous pris du retard sur leur calendrier de réalisation. Je relève aussi qu’un autre projet, important pour la sûreté, est également confronté à des retards et à des difficultés portant sur la remise en état du circuit de refroidissement de secours des réacteurs.Cet événement montre qu’en matière de sûreté nucléaire, de nombreux maillons de la chaîne doivent être suffisamment robustes si l’on veut garantir la défense en profondeur. Il est nécessaire, mais non suffisant, de percevoir les risques et de les classer. Il faut définir des mesures d'atténuation des risques et leur allouer un financement adapté. Les projets doivent ensuite être gérés de façon professionnelle jusqu'à leur aboutissement (qualité, délais et coûts).Le rôle moteur de la direction dans l’évolution des situations, la solidité et la réactivité des services d'ingénierie ainsi qu’une gestion professionnelle des projets sont des composantes vitales de la sûreté nucléaire.

12.4 SOCODEI : un accident industriel à Centraco (Marcoule)

Le 12 septembre 2011 vers 12h15, un accident est survenu sur le site de Centraco, au plancher de travail d’un four de fusion de déchets métalliques par induction. Cette unité de fusion de Centraco traite des déchets de faible et de très faible activité et à vie courte, issus de l'exploitation courante des installations nucléaires, de la maintenance et des modifications des procédés, du démantèlement des sites nucléaires.L’usine est une installation nucléaire de base régulièrement inspectée par l’autorité de sûreté nucléaire.Un salarié de fonderie qui se trouvait à proximité du four est décédé. Quatre autres personnes ont été blessées, dont une dans un état grave.L'incendie déclenché par l'explosion dans le local a été maîtrisé à 13h06 par les équipes d’intervention de l’usine.L’installation Centraco appartient à 100 % au groupe EDF et, dans le cadre de la mission de l’IGSN, je formule quelques observations sur le déroulement de l’accident :

il n’a engendré aucun rejet radioactif ou chimique à l’extérieur,Centraco étant installé sur la commune de Codolet et jouxtant le centre du CEA70 de Marcoule, avec lequel Centraco entretient des liens techniques et opérationnels, la crise a été gérée par l’organisation du centre. L’organisation nationale de crise EDF a été mise en alerte dès la connaissance de l’accident,une erreur de calcul a conduit à sous-estimer l’activité contenue à l’intérieur du four. Les causes de cette erreur ont été identifiées par la suite.

L’analyse de l’accident n’étant pas définitivement établie, j’y reviendrai éventuellement dans mon prochain rapport.

Si les enjeux strictement radiologiques de cet événement sont limités, il s’agit d’un accident industriel grave en raison de ses conséquences humaines. L'application des critères relatifs à l'échelle INES conduit à classer l’événement au niveau 1 sur cette échelle, en raison de la faible activité radiologique contenue dans le four de fusion.

70 CEA : Commissariat à l’Energie Atomique et aux Energies Alternatives.

••

Page 85: EDFMESS FR - inis.iaea.org

83

13

UN EVENEMENT TECHNIQUE

HORS DU GROUPE EDF

L’accident de la plateforme de forage Deepwater Horizon

Le contexte

Le 19 mars 2008, British Petroleum (BP) obtient du Mineral Management Service (MMS, Etats-Unis) le permis d’exploration de la zone Mississippi Canyon Block 252, située dans la partie centrale du golfe du Mexique. Le permis accordé sera exploité pendant 10 ans par BP, qui en détient 65 % aux côtés d’Anadarko Petroleum 25 % et de MOEX Offshore 10 %.Le 22 mai 2009, MMS accorde un permis de forage du puits Macondo, situé dans la zone sous 1 500 mètres d’eau.Après des débuts de forage rendus difficiles par le premier procédé retenu et par l’ouragan Ida, BP fait appel à la plateforme Deepwater Horizon détenue et exploitée par TRANSOCEAN, sous contrat avec BP. Dans le cadre de ce contrat, cette plateforme a déjà effectué une trentaine de forages dont les 2/3 sont des forages d’exploration. La plateforme Deepwater Horizon est arrivée sur le site le 31 janvier 2010, les activités de forage ont commencé le 6 février 2010.L’accident s’est produit le 20 avril 2010 alors que s’achevait le forage offshore profond du puits d’exploration. Onze personnes ont perdu la vie.Sur Macondo, des problèmes techniques, des prises de décisions non conservatives et une réactivité insuffisante ont entraîné une détection de l’éruption et une décision de fermeture du bloc obturateur de puits trop tardives. Les gaz accompagnant la remontée accidentelle du puits se sont enflammés sur la plateforme, qui a sombré.A la tragédie humaine s’ajoutait une catastrophe écologique se soldant, pendant trois mois, par le plus important déversement d’hydrocarbures à la mer jamais rapporté sur la côte des Etats-Unis.Le 27 mai 2010, face à ce que l’opinion publique a appelé l’american Tchernobyl ou le 11-septembre écologique, le Président Obama impose un moratoire sur les forages offshore.

Le rapport de l’événement, rédigé par une commission d’experts mandatés par le gouvernement, analyse par la méthode de « l’arbre des causes » 8 facteurs clés de l’accident, conduisant chronologiquement à la perte de la plateforme. Ces lignes de défense sont elles-mêmes regroupées en quatre facteurs critiques :

le confinement du puits est défaillant, et n’est pas détecté en tant que tel (problème de ciment et de barrières mécaniques),la surveillance du puits ne détecte pas la remontée d’hydrocarbures et le pilotage du puits est perdu (interprétation des tests de pression, surveillance en salle de commande et temps de réaction),les hydrocarbures remontant sur la plateforme ne sont pas confinés (les gaz ne sont donc pas non plus collectés et s’enflamment),le système d’urgence (Blow Out Preventer) ne remplit pas sa fonction et ne condamne pas le puits (le système d’isolement du puits ne fonctionne pas et l’incendie est alimenté par les gaz).

Si cette marée noire n’est pas la plus importante que le monde ait connue, c’est pourtant celle qui aura marqué le plus le secteur pétrolier à long terme en forçant les acteurs à redéfinir ensemble la conception, la pratique et le contrôle de la sûreté des procédures et des installations.

Page 86: EDFMESS FR - inis.iaea.org

84

13

Les enseignements

Je ne vais pas commenter les aspects techniques qui n’ont rien de commun avec l’industrie nucléaire, mais examiner, avec le regard de la culture de sûreté, les facteurs qui ont participé, favorisé et conduit à l’explosion de la plateforme. Je note à cet égard que cet accident constitue avec Fukushima les deux derniers accidents industriels majeurs, ayant un impact environnemental et liés à la production d’énergie. Il présente un intérêt pour l’industrie nucléaire dans la mesure où l’Etat américain a demandé assez tôt à l’INPO de faire part de son expérience dans l’industrie nucléaire nationale. Et puis, le monde de l’énergie est un monde d’interdépendances, où tout événement local a des résonances globales.

Je me suis rapproché du Groupe TOTAL et je constate que les voies de travail sur le retour d’expérience de cet accident sont très similaires à celles identifiées par les opérateurs nucléaires : un travail avec les entreprises prestataires, le renforcement de la filière indépendante de sûreté avec le superviseur de forage, le renforcement des compétences du personnel (en particulier par des mises en situations exceptionnelles sur simulateurs), le développement d’outillages spécifiques pour faire face à l’improbable, des inspections des opérations de forage profonds en cours et des matériels de sécurité associés, la consolidation des capacités à lutter contre une pollution majeure, le contrôle de l’efficacité des plans d’urgence dans les différents contextes…

Des enseignements pour la culture de sûreté

Dans le groupe EDF, j’ai observé des actions de retour d’expérience assez contrastées en matière, à la suite de cet accident. Je tiens d’abord à saluer la démarche exemplaire d’EDF Energy, assez tôt après l’accident. Cette démarche était du type bottom-up (montante) : sur chaque site de production nucléaire, des groupes de travail ont été créés pour consacrer un moment à la réflexion, à l’analyse et aux propositions visant à renforcer encore la sûreté des sites nucléaires. L’ensemble des sites a ensuite établi une synthèse commune présentée en comité de direction d’EDF Energy. Cette démarche s’est construite avec tous les strates de l’organisation des sites et du niveau corporate, marquant ainsi la volonté de faire progresser à tous les niveaux la culture de sûreté. EDF Energy a finalement dégagé 3 thèmes à approfondir afin de « challenger » sa propre organisation en matière de sûreté nucléaire :

le management des prestataires (stratégie industrielle, passation de marchés, achats, encadrement, surveillance, contrôle…),les boucles de contrôle de la sûreté, en interne et avec les autorités,la « gestion » des barrières de sûreté et la maîtrise de leur dégradation éventuelle.

Les agents impliqués sur les sites ont été très intéressés par la démarche. C’est la première fois qu’un tel travail était effectué sur un événement industriel, hors du nucléaire. C’est aussi un travail qui a fortement décloisonné localement les spécialités professionnelles et qui va déboucher sur une production co-construite par tous les sites nucléaires. Enfin, EDF Energy a sollicité son autorité de sûreté (ONR) pour un travail en commun sur la base du partage des analyses effectuées.

Pour sa part, EDF SA a organisé une réflexion de niveau corporate, en s’appuyant sur ses compétences fortes d’analyse : le département MRI de la R&D, la direction du contrôle des risques Groupe et des représentants de la DPI. Une rencontre de partage d’analyse a été organisée avec EDF Energy. Fukushima est survenu en mars 2011, déviant brutalement toutes les ressources sur cet événement majeur aux conséquences multiples pour le Groupe. L’organisation top-down (descendante) choisie par EDF SA n’aura donc pas percolé. C’est une occasion perdue pour développer une démarche de renforcement de la culture de sûreté à tous les niveaux de responsabilité, jusque sur le terrain et dans des délais raisonnables.

••

Page 87: EDFMESS FR - inis.iaea.org

85

13

En conclusion, l’accident de BP a cela de commun avec les autres grands accidents industriels :

qu’il somme les acteurs mondiaux du secteur de revoir profondément les conditions de la sûreté de leurs opérations,qu’il crée plus de solidarité entre les acteurs de l’industrie. La sûreté est devenue un enjeu et un bien communs qu’il faut apprendre à partager, une seule défaillance chez l’un pouvant provoquer de très graves conséquences pour tous.

Dans le nucléaire, le principe d’évaluation des systèmes et des organisations par un tiers extérieur, de préférence un pair, permet une évaluation objective par des acteurs compétents, tout en limitant les risques de conflits d’intérêts puisque les pairs n’ont a priori aucune utilité à se montrer conciliants. Cette pratique est mondialement adoptée depuis longtemps.Enfin, l’harmonisation des standards internationaux de sûreté est une voie sur laquelle chaque secteur d’activité industrielle exposé à des risques doit s’engager avec détermination pour que « la rigidité des maillons la chaîne soit la même ». L’industrie aéronautique est un exemple à suivre en la matière. Cette mesure doit être encore approfondie dans l’industrie nucléaire mondiale.

Page 88: EDFMESS FR - inis.iaea.org
Page 89: EDFMESS FR - inis.iaea.org

87

14

ANNEXES

Page 90: EDFMESS FR - inis.iaea.org

88

14

14.1 - Les indicateurs de résultat du parc nucléaire d’EDF SA (sur dix années)

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Nombre d’événements classés dans l’échelle INES (1 et plus), par réacteur * 1,00 1,20 0,88 0,76 1,22 0,80 1,15 1,17 1,17 0,91

Nombre d’événements significatifs pour la sûreté, par réacteur * 7,70 8,14 7,62 9,54 10,21 10,80 10,34 10,93 10,45 10,57

Nombre de cas de non-conformité aux STE, par réacteur 1,30 1,57 1,16 1,48 1,55 1,70 1,70 1,39 1,55 1,36

Nombre d’erreurs de lignage**, par réacteur 0,74 0,93 0,50 0,66 0,69 0,57 0,62 0,53 0,77 0,71

Nombre d’arrêts du réacteur, par réacteur (et pour 7000 heures de criticité )***• Automatiques 1,19 1,13 1,01 0,93 0,89 0,87 0,51 0,71 0,69 0,50

• Manuels 0 0 0 0 0 0 0 0 0,01 0,05

Dose opérationnelle collective moyenne, par tranche en service (en hSv)

0,97 0,89 0,79 0,78 0,69 0,63 0,66 0,69 0,62 0,71

Dosimétrie individuelle :

• Nombre de personnes dont la dose est supérieure à 20 mSv 0 1 0 0 0 0 0 2 0 0

• Nombre de personnes entre 16 et 20 mSv 154 74 73 28 17 20 14 10 3 2

Nombre d’événements significatifs de radioprotection - 160 177 173 112 99 107 102 91 92

Disponibilité (%) 82,2 82,7 82,8 83,4 83,6 80,2 79,2 78,0 78,5 80,7

Indisponibilité fortuite (%) 2,5 3,0 3,5 3,2 3,3 3,7 4,4 4,6 5,2 2,2

Taux de fréquence des accidents du travail (Tf) **** 8,5 7,5 5,5 5,5 5,6 4,6 4,4 4,3 4,1 3,9

Page 91: EDFMESS FR - inis.iaea.org

89

14

14.2 - LES INDICATEURS DE RESULTAT DU PARC NUCLEAIRE D’EDF ENERGY

Précaution à prendre en compte en cas de comparaison du tableau de résultats d’EDF SA avec celui du Royaume-Uni :

Lignes 3, 4 et 5 : les pratiques de déclaration des événements sont différentes au Royaume-Uni et en France, compte tenu des exigences des Autorités de sûreté respectives. Ligne 7 : les réacteurs des deux parcs nucléaires ne sont pas de même technologie (essentiellement AGR pour le Royaume-Uni et REP en France). Les AGR sont, par conception, de l’ordre 10 fois moins-dosants (référence WANO). Ligne 9 : les valeurs des années antérieures ont été modifiées pour prendre en compte des critères de déclaration différents entre EDF Energy et EDF SA, et permettre une meilleure intercomparaison.

N° Indicateurs 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

1Nombre d’événements classés dans l’échelle INES (1 et plus), par réacteur

5,67 5,27 5,60 5,67 3,13 1,20 1,13 0,80 0,93 1,33

2Nombre d’événements classés dans l’échelle INES (0 et plus), par réacteur

10,53 10,80 9,60 9,13 7,53 4,93 4,53 5,47 5,60 4,7

3Nombre d’événements au niveau le plus haut déclarés à la NII, par réacteur

- 5,47 3,60 2,67 1,53 0,40 0,67 0,33 0,67 0,46

4Nombre de cas de non-conformité aux STE, par réacteur - - - 1,00 0,73 0,13 0,27 0,13 0,60 0,33

5Nombre d’erreurs de lignage, par réacteur - - 0,40 1,09 0,69 0,13 0,27 0,13 0,60 0,33

6

Nombre d’arrêts du réacteur, non programmés, par réacteur (et pour 7000 heures de criticité)• Automatiques 1,70 1,51 1,30 0,74 0,73 0,44 1,13 0,82 0,58 0,74• Manuels 1,96 1,39 2,18 1,28 2,54 1,48 1,04 1,44 1,68 1,22

7

Dose opérationnelle collective moyenne, par tranche en service (en hSv)• REP 0,296 0,351 0,032 0,352 0,524 0,045 0,264 0,337 0,271 0,537• AGR 0,103 0,073 0,026 0,055 0,152 0,071 0,167 0,100 0,018 0,084

8

Dosimétrie individuelle :• Nombre de personnes dont la

dose est supérieure à 20 mSv 0 0 0 0 0 0 0 0 0 0• Nombre de personnes dont la

dose est supérieure à 15 mSv 0 0 0 0 0 0 0 0 0 0

9Nombre d’événements significatifs de radioprotection 98 192 184 368 249 58 38 31 43 43

10

Disponibilité (%) :• Parc EDF Energy 74,6 78,2 69,9 71,9 66,1 62,8 51,2 71,0 65,7 72,0• PWR 88,9 86,1 89,4 83,7 85,3 98,4 89,2 87,4 45,6 82,5• AGR 73,6 77,6 68,5 71 64,7 60,2 48,5 69,8 67,1 71,3

11

Indisponibilité fortuite (%)• Parc EDF Energy 15,0 11,7 15,5 12,3 17,0 20,3 20,4 13,2 19,6 13,0• PWR 2,4 2,4 9,7 0,0 0,3 0,4 2,1 0,9 54,3 3,4• AGR 15,9 12,4 15,9 13,1 18,2 21,7 21,8 14,0 17,1 13,7

12Taux de fréquence des accidents du travail (pour 200000 h travaillées)

NC NC 0,51 0,37 0,22 0,27 0,35 0,11 0,007 0,12

13Taux de fréquence des accidents du travail (Tf)* NC NC 2,6 1,9 1,1 1,4 1,8 0,6 0,4 0,6

* Tf : taux de fréquence des accidents du travail avec arrêt pour 1 million d’heures travaillées

Page 92: EDFMESS FR - inis.iaea.org

90

14

14.3 - La carte des centrales nucléaires d’EDF SA

FlamanvillePaluel

Penly

Gravelines

Chooz

Cattenom

Nogents/SeineDampierre

Belleville

Fessenheim

Chinon

Civaux

Blayais

Golfech

Bugey

St-Alban

Creys-Malville

Cruas

Tricastin

St-Laurent

Brennilis

Clermont-Ferrand

Nîmes

Grenoble

Lyon

Bordeaux

Bourges

Paris

AmiensCherbourg

Tours

Strasbourg

Marseille

Page 93: EDFMESS FR - inis.iaea.org

91

14

14.4 - La carte des centrales nucléaires d’EDF Energy

Construction ou Projet

Nombre par type AGR REP EPR Ingénierie

Exploitation

Unités d’appui aux centrales

Hunterston B

Hinkley Point B

Hinkley Point C

Barnwood

East Kilbride

Dungeness B

Sizewell B

Sizewell C

Hartlepool

Torness

Heysham 1

Heysham 2

4

14 1

2

Londres

Page 94: EDFMESS FR - inis.iaea.org

92

14

14.5 - Les étapes industrielles de chacune des unités de production d’EDF SA

AnnéeMise enService

Unité de Production

Puissanceen MWe*

VD1 VD2 VD3Année

Mise enService

Unité de Production

Puissanceen MWe*

VD1 VD2 VD3

1977 Fessenheim 1 880 1989 1999 2009 1984 Cruas 4 915 1996 2006

1977 Fessenheim 2 880 1990 2000 2011 1984 Gravelines 5 910 1996 2006

1978 Bugey 2 910 1989 2000 2010 1984 Paluel 1 1330 1996 2006

1978 Bugey 3 910 1991 2002 1984 Paluel 2 1330 1995 2005

1979 Bugey 4 880 1990 2001 2011 1985 Flamanville 1 1330 1997 2008

1979 Bugey 5 880 1991 2001 2011 1985 Gravelines 6 910 1997 2007

1980 Dampierre 1 890 1990 2000 2011 1985 Paluel 3 1330 1997 2007

1980 Dampierre 2 890 1991 2002 1985 St-Alban 1 1335 1997 2007

1980 Gravelines 1 910 1990 2001 2011 1986 Cattenom 1 1300 1997 2006

1980 Gravelines 2 910 1991 2002 1986 Chinon B3 905 1999 2009

1980 Gravelines 3 910 1992 2001 1986 Flamanville 2 1330 1998 2008

1980 Tricastin 1 915 1990 1998 2009 1986 Paluel 4 1330 1998 2008

1980 Tricastin 2 915 1991 2000 2010 1986 St-Alban 2 135 1998 2008

1980 Tricastin 3 915 1992 2001 1987 Belleville 1 1310 1999 2010

1981 Blayais 1 910 1992 2002 1987 Cattenom 2 1300 1998 2008

1981 Dampierre 3 890 1992 2003 1987 Chinon B4 905 2000 2010

1981 Dampierre 4 890 1993 2004 1987 Nogent 1 1310 1998 2009

1981 Gravelines 4 910 1992 2003 1988 Belleville 2 1310 1999 2009-

1981 St-Laurent B1 915 1995 2005 1988 Nogent 2 1310 1999 2010

1981 St-Laurent B2 915 1993 2003 1990 Cattenom 3 1300 2001 2011

1981 Tricastin 4 915 1992 2004 1990 Golfech 1 1310 2001 -

1982 Blayais 2 910 1993 2003 1990 Penly 1 1330 2002 2011

1982 Chinon B1 905 1994 2003 1991 Cattenom 4 1300 2003 -

1983 Blayais 3 910 1994 2004 1992 Penly 2 1330 2004 -

1983 Blayais 4 910 1995 2005 1993 Golfech 2 1310 2004 -

1983 Chinon B2 905 1996 2006 1996 Chooz B1 1500 2010 -

1983 Cruas 1 915 1995 2005 1997 Chooz B2 1500 2009 -

1984 Cruas 2 915 1997 2007 1997 Civaux 1 1495 - -

1984 Cruas 3 915 1994 2004 1999 Civaux 2 1495 - -

VD1 : 1ère visite décennale VD1VD2 : 2ème visite décennale VD2VD3 : 3ème visite décennale VD3

(*) Puissance continue Nette (PCN)

Page 95: EDFMESS FR - inis.iaea.org

93

14

14.6 - Les étapes industrielles de chacune des unités de production d’EDF

Energy

Année demise enservice

Unité de production

Réacteur numéro

PuissanceMW RUP

(1)

Date prévue de mise à l’arrêt

défi nitif(2)

1976 Hinkley Point B R3 435 2016

1976 Hinkley Point B R4 435 2016

1976 Hunterston B R3 460 2016

1976 Hunterston B R4 430 2016

1983 Dungeness B R21 520 2018

1983 Dungeness B R22 520 2018

1983 Heysham 1 R1 585 2019

1983 Heysham 1 R2 575 2019

1983 Hartlepool R1 595 2019

1983 Hartlepool R2 595 2019

1988 Heysham 2 R7 605 2023

1988 Heysham 2 R8 605 2023

1988 Torness R1 600 2023

1988 Torness R2 605 2023

1995 Sizewell 1191 2035

(1) RUP (Reference Unit Power) : Puissance électrique de référence de l’unité de production déclarée par EDF Energy dans les transactions journalières au 17 janvier 2012.

(2) Dates de mise à l’arrêt défi nitif, incluant toutes les décisions d’extension de la durée de fonctionnement formellement prises à la date du 17 janvier 2012.

Page 96: EDFMESS FR - inis.iaea.org
Page 97: EDFMESS FR - inis.iaea.org

95

14

14.7 - TABLE DES ABREVIATIONS

AAAR Arrêt Automatique Réacteur

AEN Agence pour l’Energie Nucléaire de l’OCDE

AGR Advanced Gas-cooled Reactor

AIEA Agence Internationale de l’Energie Atomique

ALARA As Low As Reasonnably Acheavable

AMELIE Projet visant à transformer la logistique des pièces de rechange

AMT Agence de Maintenance Thermique

ANDRA Agence Nationale pour la gestion des Déchets RAdioactifs

ASN Autorité de Sûreté Nucléaire

BBR Bâtiment Réacteur

BK Bâtiment Combustible

CCAP Contrat Annuel de Performances

CSB Chemical Safety and hazard investigation Board

CEFRI Comité français de certification des Entreprises pour la formation et le suivi du personnel travaillant sous Rayonnements Ionisants

CEIDRE Centre d’Expertise et d’Inspection dans les Domaines de la Réalisation et de l’Exploitation

CENG Constellation Energy Nuclear Group

CGNPC China Guangdong Nuclear Power Company

CIDEN Centre d’Ingénierie Déconstruction et Environnement Nucléaire

CIEST Comité Inter-Entreprises sur la Sécurité et les conditions de Travail

CIPN Centre d’Ingénierie pour le Parc Nucléaire

CIPR Commission International de Protection Radiologique

CLI Commission Locale d’Information

CMPP Chargé(e) de Mission Parcours Professionnels (DPN)

CNEN Centre National d’Equipement Nucléaire

CNPE Centre Nucléaire de Production d’Electricité

COPAT COmité de Pilotage des Arrêts de Tranche

CSN Conseil de Sûreté Nucléaire

CSNE Conseil de Sûreté Nucléaire de Production d’Electricité

DDCN Division Combustible Nucléaire

DIN Division Ingénierie Nucléaire

DNMC Daya-Bay Nuclear operation and Management Company

DPI Direction Production Ingénierie

DPN Division Production Nucléaire

EEDEC Engagement de développement de l’Emploi

et des Compétences

EGE Evaluation Globale d’Excellence

EGS Evaluation globale de Sûreté

ENISS European Nuclear Installations Safety Standard

EPR European Pressurised Reactor

EPRI Electric Power Research Institute

ESKOM compagnie d'électricité sud-africaine

ESR Evénement Significatif en Radioprotection

ESS Evénement Significatif de Sûreté

EVEREST Evoluer VERs une Entrée Sans Tenue universelle (Projet de reconquête de la propreté radiologique)

ENSREG European Nuclear Safety Regulators Group

FFAI Fiche d’Action Incendie

FAVL Déchets de Faible Activités à Vie Longue

FEMA Federal Emergency Management Authority (USA)

GGPEC Gestion Prévisionnelle de Emplois et des

Compétences

GV Générateur de vapeur

HHCTISN Haut Comité à la Transparence et à

l’Information pour la Sécurité Nucléaire

HSE Health Safety Environment

Page 98: EDFMESS FR - inis.iaea.org

96

14

IIFOPSE Institut de Formation à la Prévention et à la

Sécurité

IN Inspection Nucléaire de la DPN

INB Installation Nucléaire de Base

INES International Nuclear Events Scale

INPO Institute of Nuclear Power Operators (USA)

INRA International Nuclear Regulator Association

INSAG International Nuclear Safety Advisory Group

INTEP Introduire de Nouvelles Technologies en Exploitation du Parc

INTRA Groupe d’INTervention Robotique sur Accident.

IOP Ingénierie OPérationnelle

IRPA International Radiation Protection Association

ISOE Information System on Occupational Exposure de l’AEN et de l’OCDE

MMAE Mission d’Appui et d’Expertise de la DIN

MOPIA Mettre en Oeuvre une Politique Industrielle Attractive

MPL Manager de Première Ligne

MRI Maîtriser le Risque Incendie (projet de la DPN)

NNDA Nuclear Decommissioning Authority

(Royaume-Uni)

NEI Nuclear Energy Institute

NNB Nuclear New Build

NNSA National Nuclear Safety Administration (Chine)

NRC Nuclear Regulatory Commission (USA)

OO2EI Obtenir un Etat Exemplaire des Installations

(projet de la DPN)

ONR Office of Nuclear Regulation

OSART Operational Safety Analysis Review Team (AIEA)

OSRDE Observatoire Sûreté Radioprotection Disponibilité Environnement

PPAI Plan d’Action Incendie

PARTNER PARc Tertiaire Nucléaire Eco-Rénové

PBMP Programme de Base de Maintenance Préventive

PGAC Prestations Générales d’Assistance aux Chantiers

PHPM Projet d’Harmonisation des Pratiques et des Méthodes

PPH Projet Performances Humaines (DPN)

PSPG Peloton Spécialisé de Protection de la Gendarmerie

PUI Plan d’Urgence Interne

RRDA Projet : Réduction de la Durée des Arrêts

R&D Direction Recherche et Développement

REP Réacteur à Eau Pressurisée

RET Régime Exceptionnel de Travaux

REX Retour d’EXpérience

RGE Règles Générales d'Exploitation

RNR Réacteur à Neutrons Rapides

RTE Réseau de Transport d’Electricité

RTGE Réglementation Technique Générale destinée à prévenir et limiter les nuisances et les risques externes résultant de l’Exploitation des installations nucléaires de base

SSDIN Système D’Information du Nucléaire

SDIS Services Départementaux d’Incendie et de Secours

SEPTEN Services d’Etudes et de Projets Thermiques et Nucléaire (DIN)

SIR Services d’Inspection Reconnus

SOH Socio-Organisationnel et Humain

SPR Service Prévention des Risques

SRD Safety and Regulation Department (EDF Energy)

SRO Senior Reactor Operator

STE Spécifications Techniques d’Exploitation

Page 99: EDFMESS FR - inis.iaea.org

97

14

TTEM Tranche En Marche

TEPCO Tokyo Electric Power Compagny

TNPJVC Joint venture entre la compagnie chinoise CGNPC (70%) et EDF (30%)

TSM Technical Support Mission réalisé par des pairs sous l’égide de WANO

TSN Loi sur la Transparence et la Sécurité en matière Nucléaire

UUFPI Unité de Formation Production Ingénierie

UNE Unistar Nuclear Energy

UNIE UNité d’Ingénierie d’Exploitation

UNGG réacteurs Uranium Naturel Graphite Gaz

UTO Unité Technique Opérationnelle

VVD Visite Décennale

VP Visite Partielle

WWENRA West European Nuclear Regulators

Association

WANO World Association for Nuclear Operators

Page 100: EDFMESS FR - inis.iaea.org
Page 101: EDFMESS FR - inis.iaea.org

99

14

Jacques DUSSERRE, Jean-Paul COMBEMOREL, JeanTANDONNET, Peter WAKEFIELD, Christian THEZEE, Bruno CORACA

Crédit photographique

Couverture : © EDF Médiathèque - H. M. DUCLOS Chapitre 2-5 : © EDF Médiathèque - Alexis MORIN Chapitre 2-6 : © EDF Médiathèque - H. M. DUCLOS Chapitre 2-10 : © Sylvain CONSEIL Chapitre 3 : © EDF Médiathèque Chapitre 4 : © TEPCO Chapitre 5 : © EDF Médiathèque - H. M. DUCLOS Chapitre 6 : © EDF Médiathèque Chapitre 7 : © EDF Médiathèque - H. M. DUCLOS Chapitre 8 : © EDF Médiathèque - Alexis MORIN Chapitre 9 : © EDF Médiathèque Chapitre 10 : © EDF Médiathèque Chapitre 11 : © EDF Médiathèque Chapitre 12 : © EDF Médiathèque Chapitre 13 : © EDF Médiathèque

Page 102: EDFMESS FR - inis.iaea.org
Page 103: EDFMESS FR - inis.iaea.org
Page 104: EDFMESS FR - inis.iaea.org

E.D.F.Présidence IGSN21, avenue de Messine75008 Paris

www.edf.fr

Page 105: EDFMESS FR - inis.iaea.org

2011

E.D.F.Présidence IGSN21, avenue de Messine75008 Paris

www.edf.fr

The Inspector General’s report

on Nuclear Safety and Radiation Protection

ED

F I

GS

N R

EP

OR

T’S

20

11

Page 106: EDFMESS FR - inis.iaea.org
Page 107: EDFMESS FR - inis.iaea.org

1

2

3

4

5

6

7

8

9

10

11

12

13

14

FOREWORD

This report is destined for the President of EDF with the purpose of informing him of my judgement of nuclear safety and radiation protection within the EDF Group. It is also intended for all who, in any way, play a role in improving nuclear safety. It may also prove useful to those outside the company who work in the nuclear field or indeed any branch of industry where there are risks to be managed and who wish to participate in the debate. It also contributes to the pool of information shared with the other nuclear operators under the auspices of WANO.

The continuous improvement of nuclear safety results is a fundamental commitment of the EDF Group. This is notably reflected in a policy of transparency, whether in France, Britain, America or China. This document constitutes an example of this, offering an independent view of the operational situation. It is based on facts and findings derived in our contacts with field staff and meetings with the decision makers, managers, medical personnel and players on both sides of this industry as well as outside stakeholders, especially contractor companies. The resulting snapshot of the situation in the field is particularly instructive.

This report therefore concentrates more on problems and weaknesses rather than strengths and progress, except as concerns happenings outside France, where the focus is more on good practices than difficulties This may seem unfair to all those who have totally committed themselves to the delicate task of operating the facilities on a daily basis. Here, in particular, I have in mind the staff from our engineering centres and power plants who were the first to be affected by the Fukushima accident and have significantly contributed to the subsequent assessments.

I would like to thank all those I met, inside and outside the EDF Group, in France and elsewhere, for welcoming us so kindly, for discussing matters so frankly and creating the conditions for fruitful discussions. Their openness, vital to the relevance of this report, continues to reflect a spirit of nuclear safety culture.

I would also like to thank my advisers Christian Thézée, Bruno Coraça and Peter Wakefield who, once again this year, were unstinting in their efforts, without forgetting Jean-Paul Combémorel who joined us at the start of the year. I address my special thanks to Jacques Dusserre, who has left the team after six years of remarkable service. This year’s report still shows his style.

To optimise this report, the layout has been changed somewhat relative to previous years. All the issues linked to nuclear safety have been covered although the number of chapters has been reduced. Finally, although this document has not been written for the purpose of public relations, as in previous years, it will be available to the general public on the EDF website in both French and English (www.edf.fr). This year, more perhaps than previous years, I hope it will provide answers to the questions that many are posing concerning nuclear safety.

The EDF Group Inspector General for Nuclear Safety and Radiation Protection

Jean Tandonnet

Paris, 20 January 20�2

Page 108: EDFMESS FR - inis.iaea.org
Page 109: EDFMESS FR - inis.iaea.org

1

2

3

4

5

6

7

8

9

10

11

12

13

14

Contents

1 My view of nuclear safety in the EDF group 5

2 A panorama of 2011 92.1 Themanagement 102.2 Thestateoftheinstallations 102.3 MaintenanceoftheFrenchnuclearpowerplants 102.4 RelationswiththeFrenchnuclearsafetyauthority(ASN) 122.5 TheEPRprojects 142.6 Theplayersinthefieldsofhealthandsafety 152.7 Radiationprotection 172.8 Firepreventionandfirefighting 182.9 Increasingnuclearpowerplantsecurity 192.10 Decommissioning 20

3 Nuclear safety management 21

4 The Fukushima Daiichi accident 27

5 Training to develop skills in the professions 39

6 The operations profession 45

7 The engineering services and operations support 49

8 Managing projects 53

9 Together with contractors towards better overall performance 57

10 EDF nuclear activities outside France 61

11 Visits outside France 69

12 Events in the EDF group 7312.1 EDFSA:prematurewearofdieselgeneratorbearings 7312.2 EDFSA:acontrolincidentandanautomaticreactortrip 7412.3 EDFEnergyfailureofcast-ironpipes 7612.4 SOCODEI:anindustrialaccidentatCentraco(Marcoule) 77

13 A technical event outside the EDF group: The Deepwater Horizon drilling platform accident 79

14 Appendices 8314.1 ResultindicatorsforEDFSA’splants 8414.2 ResultindicatorsforEDFEnergy’splants 8514.3 MapoftheEDFSAnuclearpowerplants 8614.4 MapoftheEDFEnergynuclearpowerplants 8714.5 TechnicalkeydatesforeachoftheEDFSAnuclearunits 8814.6 TechnicalkeydatesfortheEDFEnergynuclearunits 8914.7 Tableofabbreviations 91

Page 110: EDFMESS FR - inis.iaea.org
Page 111: EDFMESS FR - inis.iaea.org

1

MY VIEW OF NUCLEAR SAFETY IN THE EDF GROUP

First, I would like to state that no major nuclear safety events occurred in the EDF Group’s nuclear plants this year. There was, however, an event ranked Level 2 on the International Nuclear Event Scale (INES) in France related to equipment failure which had no direct effect on nuclear safety (see Chapter �2), as well as some noteworthy events in France and Britain which also had no direct implications but need to be heeded as warnings and dealt with in a formal way by those in charge.I would also like to highlight, in France and Britain, the overall good performance as shown by the main nuclear safety indicators, including fire, for instance the small number of scrams in France, better performance levels in the plants and more effective international benchmarking.These results, together with our improved generating results, will make all involved feel more calm and confident, with a positive effect on nuclear safety.

I also consider that EDF SA (France) has an independent nuclear safety organisation that is both strong and professional, fostering confidence in the level of openness expected of it. The intention of the EDF Energy (UK) management to set in place the same type of organisational structure and capability is also extremely positive.

Finally, the responsiveness exhibited by the Group’s different bodies and management teams affected by the Fukushima catastrophe, both in France and beyond, also needs to be commended, which is a good sign of an operator aware of its responsibilities.

As regards radiation protection, the 20�� results in the Group are satisfactory overall. I would like to draw attention to the consequences of the forthcoming major maintenance and modernisation jobs in France, which will necessitate an even more proactive approach.

As concerns safety in the workplace, it must be emphasised that our French results are still lagging considerably behind the best results internationally as determined by WANO and vigorous action is required. Here, I consider that manager action in the field, focusing on rigour in operations, is the key to durably improving the nuclear safety situation.

Although I appreciate the levels of performance, I have to sound the warning bugle as I have an issue to raise concerning the French plants.The warning is once more about the renewal of skills in the company's staff and contractor staff that work alongside them: we need to ensure the successful handover to the incoming generation while the �8 reactors continue to operate, we need to meet the massive demand for engineers required to extend the service lives of the reactors to sixty years, and we need to take appropriate action in the Fukushima aftermath. These challenges are obliging the company to prolong its current sustained and strenuous efforts. In such a context, it is important to draw attention to the overriding need to give priority to this action without compromising nuclear operating safety.

Page 112: EDFMESS FR - inis.iaea.org

1

While I am glad that concrete action is now being taken to give more responsibility to the plant staff and managers, while providing them with stronger backup at corporate level, I believe that much still remains to be done in this promising area which is so important to ensuring calm in the field: vigilance is required.

For the sake of the nuclear industry’s future, I would like to emphasise the importance of the EPR technology, which represents an unparalleled advance in nuclear safety, irrespective of any worksite and project-related difficulties.

Beyond these essential issues which I wished to give particular prominence, I treat in the next section “a panorama of 20��”, a wider range of issues concerning operations and engineering. In the chapters that follow, in greater depth, I tackle important technical themes which engage my attention.

Page 113: EDFMESS FR - inis.iaea.org

2A PANORAMA OF 2011

The year 20�� was marked by the aftermath of the earthquake and tsunamis that hit Japan on �� March, more particularly their effect on Fukushima Daiichi Nuclear Power Plant. The resulting damage caused loss of the heat sink and station blackout in all six nuclear units on the site. The accident showed that it is the operator which holds prime responsibility for the safety of a nuclear power plant. It also showed, once again, that the nuclear industry demands humility and above all rigour. I would like to mention the interdependence of the nuclear industry, where a local accident affects the entire world. In view of the scale of the accident, the consequences are indicated below and covered in depth is a later chapter devoted to the case. Here, I would like to express my compassion for the families of the tsunami victims and those displaced after the Fukushima accident, whose living conditions are still strongly affected.

In the wake of the accident, stress tests were demanded as early as the 2� March by the French Prime Minister and later by the European Commission. I note that the EDF Group submitted action plans and, in the following months, made a number of assessments as requested by the French and British authorities.The accident has left its mark on nuclear power plant staff everywhere. I would like to remind the upcoming generations who did not live through the Three Mile Island and Chernobyl accidents that the nuclear industry is special and that proper safety culture is vital in both design and operations.

In France

I note that, for EDF, the first challenge was to learn everything possible from the Fukushima case, but this was not the only one. EDF has set itself the goal of continuously increasing the levels of nuclear safety in its plants and to take the necessary action to extend their service lives to �0 years. Accordingly, the year featured twice as many ten-yearly inspections as in 20�0 as well as major maintenance operations such as steam generator replacement. I observe with satisfaction that these massive investments will be maintained over the next few years to support the French plants reaching mid-lifespan and to accommodate new nuclear and industrial safety requirements.In 20��, the French Nuclear Safety Authority ruled in favour of continued operation of Unit � of Fessenheim Nuclear Power Plant, which is �0 years old, for a further �0 years in accordance with the Nuclear Safety and Transparency Act on the condition that additional works were carried out. After Unit � of Tricastin Nuclear Power Plant in 20�0, it is the second nuclear unit to pass this milestone.

The nuclear power plant refurbishment will intensify from 20�� with the major works associated with the third ten-yearly inspections of the nuclear units belonging to the ��00 MWe series. In 20�2, I will be carefully monitoring this work, referred to as the General Refurbishment, with particular attention to the skills-related aspects. The work, which is to begin in 20��, will consist of two parts: replacement of certain key components during the ten-yearly inspections to prepare for plant life extension to sixty years and, as part of the periodic nuclear safety re-assessments, integration of improvements to strengthen prevention and the ability to survive severe accidents and external events, such as earthquakes and flooding. During my visits to the plants, I observed mobilisation to rise to this challenge, yet noticed some very worrying signs in the field reflecting recent discussion of the early withdrawal from service of some of the nuclear units. This is a source of anxiety, and I am worried about a possible impact on nuclear safety.

Page 114: EDFMESS FR - inis.iaea.org

8

2In the United Kingdom

I note a revival of the nuclear industry in the United Kingdom. In July, the British Parliament approved the National Policy Statement for Nuclear Power Generation. This confirmed the planned construction of nuclear power plants, including Hinkley Point C and Sizewell C, which would make EDF Energy the first nuclear operator to participate in the revival of nuclear power in Britain. The Nuclear Site Licence application was lodged with the British nuclear safety authority and is expected to be approved by the middle of 20�2.With regard to the Fukushima accident, I note a calm and phlegmatic response at odds with the reaction in some other European countries.

In the United States of America

I headed an inspection of Ginna Nuclear Power Plant, which has been in service for 4� years. It is operated by the Constellation Nuclear Energy Group (CENG), a company in which EDF has a stake. It is to be remembered that the EDF plants in France have an average age of around �0 years and are relatively young. I would like to point out that reactors of the same PWR design have already received licenses to operate for up to �0 years.I observe that CENG has appointed EDF to preside over the Operations Nuclear Safety Committee of its board of directors. The committee has notably cross-referenced and enriched the French and American experience feedback on Fukushima.I note that, this year, the company EXELON as made an offer to purchase Constellation, itself the majority shareholder of CENG (of which it has �0.�% and EDF has 49.9%). I will be watching with interest how this new situation develops.

In China

I re-visited the worksite for two reactors at Taishan, and met people involved in the project. Setting the metal dome in place on the first reactor containment was a symbolic step, and I observed the efficiency of the joint AREVA and CGNPC (China Guangdong Nuclear Power Company) design office in catching up the delay in producing the erection drawings. In the wake of the Fukushima accident, I approve of the changes in the reactor which reflect the French proposals determined for the Flamanville � EPR, with a few site-related adaptations.

The importance of natural events

After what happened at Fukushima, where a natural phenomenon has had such disastrous consequences, I remain particularly vigilant about the consequences of natural events on the nuclear safety of nuclear power plant fleets.Within the EDF Group, in 20��, the cooling water intake of Torness Nuclear Power Plant in the UK was almost blocked due to an excessive number of jelly fish. In Maryland in the USA, Calvert Cliffs Nuclear Power Plant experienced a substantial earthquake, followed by a four-day hurricane that momentarily halted one of the nuclear units. There were no consequences in terms of nuclear safety at either of the plants, as the installations were robustly designed. In France, after a long and severe drought in the first half of the year, EDF had to make special arrangements to avoid any impact on the generation of its plants. Experience feedback from the 200� drought and heat wave, and the fact that the weather conditions were not as bad as forecast, ensured that no particular difficulties were encountered in the course of the summer.

Outside the EDF Group, North Anna Nuclear Power Plant in the USA experienced an earthquake of an intensity that exceeded its design basis. After inspection and requalification of its two reactors, the Nuclear Regulatory Commission (NRC) licensed their restarting in November. Similarly, the floods that affected Fort Calhoun Nuclear Power Station and Cooper Nuclear Power Station in June for a number of months did not cause any damage as precautionary measures were taken well in advance.

Page 115: EDFMESS FR - inis.iaea.org

9

2Grounds for satisfaction

At the start of 20��, the �8 French reactors and the �� British reactors were all generating power. This is only the second time this has happened in the history of the French plants. Meanwhile, I note a drop in the number of unplanned inoperability cases in 20�� which, to a certain extent, is good in terms of nuclear safety and demonstrated the pertinence of the major investment in refurbishment since 200�.

2.1 The management

I continue this review with four initiatives that have caught my attention.

The 2020 Generation Campaign

This embodies the thinking of the management of EDF SA’s Nuclear Operations Division (DPN) and is an intrinsic part of the action to motivate the upcoming generation of staff. This new campaign is already having beneficial effects. The attention paid to the problems faced in the field has increased the ability of the plant management teams to function in a climate of calm. I also await the results of a campaign, which is being conducted by the same plant management teams in cooperation with EDF SA's Nuclear Engineering Division (DIN) and is designed to increase the synergy between the two divisions, particularly the availability of an engineering force for the plants in service that corresponds more precisely to the needs of the operator. However, I remain attentive to the risk of overloading the engineering force, which has an increasing number of technical projects to handle.

The Associate Directors for Operations

More than ever, prioritising activities as a function of the resources available is a fact of life for the management, in both operations and engineering. The risk now being faced is the likelihood of actions being left undone in the field without reflecting managerial choices. It is in times of need that corporate-level management is seen to be a necessity, although the excessive centralisation as I have mentioned in my previous reports needs to be avoided. For the DPN, the Associate Directors for Operations play an essential role of arbitration so the necessary adaptations can take place. I also call on the management teams to re-affirm the two basic principles of nuclear safety that become all the more important in periods of intense activity and change: a questioning attitude and the duty to remain on the alert.

Establishing priorities and increasing efficiency

I note that integrated management is becoming universal in the plants, without it having been specifically asked for at corporate level. This year again, I note that this management system fosters synergy between separate campaigns and clarifies the roles of the players, which makes it possible to identify and correct any shortcomings in the organisational structure. However, I would again draw attention to the need to preserve this performance-enhancing tool in the complex environment of the nuclear industry.The prioritisation system still seems to be insufficiently clear to the front-line managers. In the plants, the managers and planners appear to be those confronted with the most constraints.

A nuclear safety policy for the EDF Group

I am glad that a start has finally been made with writing the nuclear safety policy of the EDF Group. I would like to draw attention to the importance of deploying it with all the personnel concerned.

Page 116: EDFMESS FR - inis.iaea.org

�0

22.2 The state of the installations

In France, the state of the installations is improving as a result of substantial investment. It is vital to sustain this effort as the plants are now in the middle of their service lives and we are lagging behind international standards.

The state of the nuclear generating facilities should soon reach the anticipated level at most of the sites. Rather than the change in appearance, it is the change in culture that I find most noteworthy. The experience feedback from the Better

Housekeeping Project managers needs to be assessed if the achievements are to be maintained in all the plants. I hope that the management teams maintain the levels of commitment that have made this project a success. Special attention is required in the case of costal sites where salt corrosion makes it more difficult to keep up to standard. I strongly recommend that the orderliness of the EPR construction site at Flamanville be the subject of constant attention by the DIN and that housekeeping be of the highest standard when the plant enters service. At the present time, it cannot be taken for granted that this priority is clearly imparted and shared.

Alongside the Better Housekeeping Project which relates to the generating facilities, I commend the efforts to adapt and refurbish the administration buildings as part of the nuclear site administration building green renovation project – known as the PARTNER Project - which is to run over the next ten years. This project has been eagerly awaited by both EDF and contractor staff. The improved living conditions at work that are expected to result from the project represent another step forward for nuclear safety.

2.3 Maintenance of the French nuclear power plants

In France, the next ten years will be decisive for the future of the French plants. A number of challenges need to be overcome to protect investments, to plan for plant life extension and to increase availability levels while continuing to make progress with nuclear safety. To succeed in this, the DPN has launched the previously mentioned 2020 Generation Campaign. I personally share the ambitions of this campaign which are to foster responsible attitudes in the staff regarding the state and maintenance of the facilities, as well as their work methods.

I emphasise the importance of the maintenance force alongside the nuclear unit operating staff in this campaign. With the plants half way through their service lives, maintenance has acquired strategic importance in keeping the safety levels of the nuclear units up to standard and protecting the company’s assets. In the recent past, equipment failures were affecting the performance of the nuclear units, which led to questions being posed about the maintenance strategies and the organisational structure. I observe that things have evolved, and that maintenance is now being given due consideration in the current changes.

During my visits, I noted a real willingness of the local staff, particularly the maintenance personnel, to get personally involved in the operation and maintenance of their equipment. However, in the current context of skill renewal and the launching of numerous campaigns that strongly affect the maintenance staff, particularly the planners, this change is difficult to implement. Everywhere I found the planning staff were heavily overloaded. The organisational structures are under pressure due to the amount of work to be carried out, the lack of prioritisation and the lack of resources. Allowance will need to be made for this in the General Refurbishment

planning. I will make a point of monitoring its progress in my forthcoming visits.

Page 117: EDFMESS FR - inis.iaea.org

��

2This year again, it needs to be said that the way work is organised is proving a hindrance and is ill-adapted to the expected increase in the company’s workload. However, I note that certain plants have begun to change things but this is far from being the general case.The planning of the unit outages, which will represent the main workload of future major refurbishments, is still not up to the desired standard. It is not yet being organised sufficiently in advance and manpower is not being assigned to it long enough before the outage is due to begin. I note that repetitive outage activities are still not sufficiently standardised.

Certain issues, which I believed were surmountable and already resolved, appear to be chronic and are complicating the work of the maintenance staff. Wherever I go, I hear about obsolete equipment and difficulties with obtaining spares.For the field workers I met, having spares available in due time is still a major obstacle. Even if the AMELIE spare parts management project may be beginning to bear fruit by streamlining the entire supply chain, much still needs to be sorted out. In addition to the difficulty in building up stocks and arranging distribution, there are problems associated with obsolescence of components and equipment. I am still struck by the number of recorders in the control rooms that still use rolls of paper in this digital age. It seems that there has not been proper forward planning in quality or quantity, therefore a paradigm shift is required. As the industrial environment constantly evolves, and as the plants are standardised, halting production of a component can cause critical problems and lead to events of the type that occurred this year with the diesel generators (see Chapter �2).

Many newcomers have joined the plant teams on a one-for-one basis in the maintenance departments since 20�2. This being the case, the managers must remain on the alert to ensure that these newcomers get a proper grounding in the basics of their profession and nuclear safety. In such situations, human performance practices constitute strong lines of defence. These need to be applied systematically.

Once again this year, I visited an EDF Joint Maintenance Agency. In our nuclear and conventional power stations, the Joint Maintenance Agencies carry out specialised maintenance operations where the financial stakes are high. They have always demonstrated their competence to the construction companies. Given their role, subcontracting is closed to them, which means that their managers must take an active part in maintaining skill levels in all areas. I cover this issue in detail in Chapter � which is devoted to training. As these bodies can step in to replace a failing supplier, they provide the EDF Nuclear Operation Divisions with a backup. At the present time, the Joint Maintenance Agencies are small teams totally focused on technical matters, consisting of highly motivated staff. It appears to me that their motivation mainly stems from the attractiveness of the posts. These are characterised by technology, quality, autonomy and variety.

Finally, as regards results, I note this year that a number of shortcomings in maintenance have come to light. Faced with the forthcoming increase in workload, the maintenance departments need to re-achieve the levels of quality that are to be expected of them in the nuclear plants and to be ready to meet the load. In 20�2, I will be particularly attentive to this during my visits to the plants.

Page 118: EDFMESS FR - inis.iaea.org

�2

22.4 Relations with the French nuclear safety authority (ASN)

There can be no long-term acceptability of nuclear power without a strong nuclear safety authority. An appropriate balance of roles is essential between the operator and the safety authority, and in France, with their technical support, the Institute for Nuclear Safety and Radiation Protection (IRSN). I am aware of the importance of the high-level technical discussions with IRSN providing its in-depth knowledge of the EDF plants, yet the actual requirements need to be stated by the ASN. Confronted by

the numerous requirements of the ASN, I would like to draw attention to the importance of prioritisation, so as to avoid overloading the engineering forces.

The Fukushima accident

This year, relations with the nuclear safety authorities were marked by the Fukushima accident.

In all countries, it is the operator that bears the responsibility for nuclear safety, as the Fukushima accident reminds us. I appreciated the responsiveness of the EDF Group as a whole in first making a rapid review of its own facilities and then answering the questions put to it. In France, the supplementary nuclear safety assessments (stress tests) requested by the ASN required a lot of time and effort. They were submitted in due time. Confidence has prevailed and, in my opinion, EDF has demonstrated exemplary openness. The openness, expected by the French public, was cited by the High Committee for Nuclear Safety and Transparency (HCTISN). In particular, I note the opening of the nuclear plants to the members of the HCTISN and the Local Information Commissions during inspections conducted in the context of supplementary nuclear safety assessments. I note the similar work carried out in the United Kingdom by EDF Energy in consultation with the Office of Nuclear Regulation. I appreciate the consistency of the action taken by the French and British safety authorities in devising stress tests in Europe. The International Atomic Energy Agency's expert appraisal mission in Japan headed by Mike Weightman, the chief inspector of the Office of Nuclear Regulation, resulted in the publication of a detailed report on the events for the nuclear operators after an initial report from the Japanese government presented to the International Atomic Energy Agency.In the USA, I observed the power of the utility company lobby, expressed through the Nuclear Energy Institute (NEI) and the Institute of Nuclear Power Operations (INPO), and especially the strong commitment of the managers at the Constellation Energy Nuclear Group.

Relations in the field

During my inspections of the nuclear generating sites and the engineering centres, I noticed, more than previous years, difficulty in relations and the settling of cases by the ASN representatives working with the EDF teams. The business constraints faced by the operator and heavy workload of both parties during this busy period have greatly strained the relations between them. I remain convinced that substance must prevail over form.

Internal licenses

I note that no real progress has been made with the processing of generic modifications, as provided for in Article 2� of the Nuclear Safety and Transparency Act, although this would offer simplicity and effectiveness. I also regret that the scope of internal licensing, as provided for in Article 2� of the Act, has still not yet been broadened within EDF SA. It seems to me that progress is needed in this area, for although the requirements are stringent, it has the virtue of clarifying and simplifying relations between the operator and the ASN. I note that it is EDF’s desire to act accordingly.

Page 119: EDFMESS FR - inis.iaea.org

��

2 In the field of fuel, the plans are being assessed in terms of the scope and the manner of execution, especially with regard to allowing greater flexibility in operation. No conclusions have yet been reached, and I fear that this is no longer a priority after the Fukushima accident. In the field of decommissioning, effective internal authorisations have long been implemented in a responsible manner.

Other issues requiring attention

I concur with the French nuclear safety authority, which has identified two sensitive issues for the nuclear plants: industrial safety and environmental protection. Concerning the first, which is covered in Section 2.�, I have been alerted to the difficulties of the plants in their relations with health and safety inspectors. Despite considerable progress, environmental protection remains a delicate issue. I share the ASN’s view that progress remains possible with the discharge of chemical effluents, as has been possible with the releases of liquid radioactive effluents over the last few years. Interfacing between chemistry, operations and maintenance needs to be improved.It is important to also be aware of the difference between the scale of the risks associated with nuclear safety on one hand, and with industrial safety and protection of the environment on the other. The consequences for the general public are of another order of magnitude. It nevertheless remains true that excellence in nuclear safety means excellence in industrial safety and protection of the environment, for all the professions involved.

Official Inspection Departments (SIR)

This year, for the first time, I met the EDF managers and inspectors of the SIRs in the plants. Charged with issues relating to staff safety, they check, on behalf of the plant director, all pressurised equipment using the powers delegated to them by the government. They report directly to the plant directors to assure their independence, and their recommendations may necessitate arbitration at the highest level in the plant.I recommend the greatest vigilance concerning the management of skills in the SIRs and maintaining their independence. Indeed, ensuring the SIRs have sufficient manpower remains difficult, lengthy training is required and potential replacements need to be identified at an early stage to ensure that the posts are filled. Corporate-level promotion of this profession needs to focus more on collecting data in the field and the management and resolution of technical difficulties.

Relations with the nuclear safety authorities

For an international group like EDF, relations between the different nuclear safety authorities are a vital issue, and will become all the more so in the future with the approval of the nuclear safety submissions for EPR-type reactors in China and the United Kingdom. I am surprised that the national authorities responsible for setting the goals for nuclear safety and checking that they are met do not cooperate more actively. It should be possible to overcome this difficulty now that the Fukushima accident has led the safety authorities to cooperate more closely. I note that the ASN is working towards this by seeking harmonisation that could build on the foundations laid in the context of the Western European Nuclear Regulator's Association (WENRA) and the International Nuclear Regulator’s Association (INRA).

Page 120: EDFMESS FR - inis.iaea.org

�4

22.5 The EPR projects

The third-generation EPR technology is specially designed to offer a significant advance in nuclear safety. Four such reactors are under construction in the world, and two have reached design stage in the United Kingdom: all these projects are being attentively watched in France and beyond. These projects are complex and require special skills. During the worksite period, like this year, or later during operation, regular progress and success with Flamanville � are vital for favourable outcomes with the other projects, for they all greatly depend on each other.

The scale of the Flamanville � project and the project management constraints have resulted in many difficulties in carrying out the site work and in the procurement of equipment. EDF recently announced that the commercial operation of Flamanville � has been put back to 20��. Major changes in the way the project was organised were made as early as 20�0, which clearly re-established, as I have previously pointed out, the division of responsibilities between the different players, particularly regarding the design issues. New resources for establishing a detailed schedule of the activities have been set in place. Concerning this point, I am glad to see that the ten main suppliers of the project, associated in a coordination group designated F�0, were involved in drawing up the project schedule. I would like to encourage them to continue and intensify their collaboration. This major task of scheduling and coordination is throwing light on the way the project is managed in terms of the risks, the schedule’s critical path and what is at stake.

In light of these EPRs under construction and how important it is to successfully complete Flamanville � in line with its new schedule, I question the need to maintain certain operating objectives that are specific to the French plants, such as load following and nuclear unit controllability, as these demand numerous ad-hoc nuclear safety studies being carried out specifically for Flamanville �.

Conversely, I hail two important innovations of the EPR project with regard to nuclear safety. Firstly, the objective set by EDF and its Chinese partner to qualify the EPR-type reactor's equipment for the consequences of severe accidents before unit start-up, despite that fact that this considerably augments the workload and the constraints on project management. Secondly, in the field of human factors, performing simulator-based qualification under real conditions before commissioning the man-machine interface of the computerised control room. From the point of view of nuclear safety, these two initiatives appear to be exemplary.

This year, I visited the EPR-type reactor worksite in Taishan in China. The dependence of this worksite on the Flamanville � one is striking, even though some of the building work, particularly the civil engineering, is different. From the presentations I was given, I gathered that the very recent feedback from our Chinese partner on the commissioning of the Ling Ao Nuclear Power Plant units led to the formation of testing teams at a very early stage so as to plan the procedures and tests for commissioning Taishan �. Contacts between the Flamanville � test teams and their Chinese counterparts concerning this issue could be worthwhile and in all probability profitable.

This year the Flamanville � worksite was the scene of some serious events, including two fatal accidents. Work safety, like the quality of the work, is the subject of special attention by the management and its desire to get things back on an even keel is all the greater as the erection of the mechanical and electrical equipment is about to begin, although the civil engineering phase is not entirely complete. Be this as it may, during my visit to the worksite, the state of the installations was not yet up to the announced standards. I wonder what means are available to the local managements with regard to the design services of the project team and subcontractors to get things changed and to guarantee quality and industrial safety during the erection work. I would also like to draw attention to the risk of

Page 121: EDFMESS FR - inis.iaea.org

��

2corrosion at this coastal site, as the preservation of equipment and the protection of the plant areas do not yet appear to be given sufficient priority.

The operator is already on site, with almost full staff numbers for nearly two years now. They are getting everything prepared, but the new schedule has caused them disquiet and even anxiety. Some of the simulator training programmes need to be changed, and staff turnover will begin before the plant has even started up, which I find worrying. Moreover, the schedule still does not yet make it possible to see far enough ahead to establish the content of the planning for commercial operation.I am also surprised that, in France, the management of such a project is entrusted to two different divisions. Throughout the project, the construction teams of the Nuclear Engineering Division and the operations teams of the Nuclear Operations Division report to different managements. This results in the sub-optimal use of manpower at a time when it is scarce. I note that, in the projects involving EPR-type reactors in England and China, the teams are combined from the start: multidisciplinary teams are considered positive in terms of quality, effectiveness and nuclear safety. Given the current arrangements at the worksite, I would encourage the Flamanville � teams to cooperate more during the erection and pre-operations phases.

The Flamanville � EPR showcases the EDF Group and I have been concerned to learn that it may not reach the latest standards existing in the other French plants as regards fire safety and the state of the installations (Better Housekeeping Project, O2EI).

Success with Flamanville � is a daily challenge with a direct bearing on success with the other EPR projects. The experience feedback from France, China and Finland is a key asset for the EDF Group that needs to be fully exploited.

2.6 The players in the fields of health and safety

This year, I continued to meet medical staff, essentially doctors and health and safety inspectors, in both France and the United Kingdom. They are the key players in the field of health, but also as concerns management of psychosocial risks, quality of life and industrial safety.The health of company and contractor staff is one of the main concerns of the EDF Group. I emphasise that industrial safety cannot be separated from nuclear safety and has to be given the same priority. The international approach, and that of EDF Energy, is leading EDF SA to increasingly consider industrial safety as a meaningful indicator of rigour in everyday operation, and hence of nuclear safety culture.

The medical staff and their working conditions

Relations between the medical services and the plant management staff have generally improved in recent years, which is highly important. In almost all the plants I visited, I encountered medical practitioners who were willing to work openly with the management staff. These crossroads of cultures, where health, industrial safety and radiation protection meet give greater depth to risk analysis and the resulting decisions. It offers leverage in achieving progress for the managers and the occupational health staff. I encourage the management teams to develop closer links with the medical practitioners.

In the field of radiation protection, these relations help:

the medical staff to actively contribute to prevention and exercise their medical surveillance in a pertinent manner, adapted to the actual risk levels,the radiation protection staff to gain a clearer understanding of human, organisational and technical matters.

I am keeping a watch on the workloads of the medical staff charged with monitoring the plant staff, as well as Radiation Worker Accreditation (DATR) for the increasingly numerous contactor staff. This monitoring, which the contractor joint medical services are not all empowered to provide, involves

Page 122: EDFMESS FR - inis.iaea.org

��

2some �,�00 contractor workers who are concentrated at certain sites. It is to be remembered that the medical staff will need to be able to perform this task in the next few years when there will be an increase in the amount of maintenance work and the contractor joint medical services will be faced with recruitment problems. I note however that the arrival of the new Type C2 controlled area exit detection portals should reduce the demand for whole-body counting.

Quality of life and psychosocial risks

Subsequent to the agreement which involves preventing psychosocial risks and improving the working lives of EDF SA staff signed on �0 November 20�0 by three trades unions, I observe that the initiatives to protect and improve the life and health in the workplace, some of which were already in hand under a prior agreement within the Nuclear Operations Division, are continuing. Wellbeing in the workplace fosters nuclear safety and it is indispensable to go further with this action, and for the managers to be more directly involved in it.

Alcohol and drugs

I have previously stressed the importance of the measures to bar anyone under the influence of alcohol or drugs from the workplace. In France, the awareness of the problem has increased, but the requirements of the Nuclear Operations Division management, including random alcohol testing, are not being applied with equal rigour everywhere. As concerns drugs, detection is at the discretion of the occupational physician alone, which is insufficient in my opinion.Meanwhile, in the United Kingdom, EDF Energy has set itself the goal of checking 20% of the personnel for alcohol and drugs every year, including contractor staff. The plant managers are informed of the results, which can entail medical surveillance and even loss of accreditation for violators. The statistics show a year-on-year increase in the number of violations. This trend, which is also observed in comparable countries, should lead us to call the current French policy into question. In the USA, the policy is similar to that in the UK.

Industrial safety in France

The constant efforts to mitigate risks have helped to progressively reduce the frequency of occupational accidents. Although the incidence of such accidents has been reduced by a factor or two over the last ten years, the figure in France stands at �.8 per million hours worked for EDF and contractor staff, which is well above that for the nuclear industry as a whole. In 20��, the figures did not improve, and there were three fatal accidents: two on the Flamanville worksite and on Gravelines Nuclear Power Plant. It is however to be noted that the Production and Engineering Directorate (DPI) conducted an industrial safety

assessment.The results call into question the adequacy of the risk management and mitigation. The involvement of the management teams at the highest level remains insufficient. I call on them to clearly announce this as a priority in the field. The contractors need to be involved in this initiative.I feel sure that extending the scope of the overall assessments of excellence by the Nuclear Inspectorate to the field of industrial safety will help to ensure that better results are obtained. On a plant level, I consider that the profession of health and safety adviser in the risk mitigation services needs to be given more prominence.

British Energy (now EDF Energy) setting an example since 2005

Although starting from a frequency factor of �.8 occupational accidents per million hours worked, when this figure doubled at the beginning of 200�, the company took some decisive action. Rules

Page 123: EDFMESS FR - inis.iaea.org

��

2of behaviour were laid down at the highest levels of British Energy and the main contractors. These reiterated the rules on alcohol and drugs, as well as the need to maintain a questioning attitude, to use one's own professional experience to plan activities, to use personal protective equipment, to organise experience feedback, to refuse a non-secured work area, and to compare one's self with and accept the challenges of others. In 20��, the frequency factor dropped to 0.�.

2.7 Radiation protection

For the EDF SA's nuclear plants, there was a heavy maintenance operation workload with 9 ten-yearly outages, more than twice as many as the previous year.

Collective dose

The plants have entered a period of cyclic oscillation of the collective dose synchronised with the maintenance outage profile. This oscillation lies within the band corresponding to the international results for reactors of equivalent technology. The increase in worker exposure time when all other factors are unchanged therefore automatically increases the collective dose per reactor, which increased from 0.�2 man-sieverts in 20�0 to 0.�� man-sieverts in 20��.In some of the plants, I note with satisfaction that the operations designed to reduce the mean values of the dose equivalent (source term) as part of the ALARA approach. These operations are, in some instances, occasionally affected by the presence of particles emitting radioactivity or systematically affected by certain maintenance operations.

Individual doses

Individual doses continued to drop among the professions with the highest exposure (scaffolding erectors, heat lagging workers, welders etc.). This improvement can be ascribed to the revival of the ALARA approach, which accompanies work area dose optimisation. The number of radiation workers whose dose over twelve months exceeded �� mSv continued to decrease, with two cases reported in 20��.

Events in 2011

Two radiation protection events were classified as Level � on the International Nuclear Event Scale, one of which occurred during a maintenance operation when a worker’s hand was irradiated, corresponding to a skin dose of 4�0 mSv, close to the regulatory limit of �00 mSv. I note stability in the number of significant radiation protection events. I recommend, once again, greater rigour with certain issues, such as: access to red and orange radiation hazard areas, the wearing of personal protective equipment in radiation hazard areas, radiography work, the planning of jobs in controlled areas and work in the reactor and fuel building pools. I note that the field work processes have been reinforced in the plants but behavioural patterns still indicate difficulty in adaptation.As the ten-yearly outages involve a considerable amount of maintenance work and scheduled testing, they are accompanied by a great increase in the amount of radiography. Similarly, at the Flamanville � EPR worksite, the start of mechanical equipment erection has resulted in a strong increase in the amount of radiography, with the operations complicated by the great number of different professions being present and some individuals having little awareness of the high risks associated with the work. Although the overall level of management of these activities improved substantially in 20�� in all the plants, a number of narrowly averted incidents indicate the need for greater vigilance. I thus regret that the Radiation Protection Surveillance Station that is planned to be installed in every plant has been re-scheduled whereas many plants outside France have long had them.I also note recurrent violations of the restrictions on access to radiation hazard areas for service personnel with temporary employment contracts (temporary workers and fixed-term contract workers). The situation did not improve in 20��. It should be possible to solve EDF’s administrative difficulties under a framework agreement on “socially responsible subcontracting in EDF” proposed to the

Page 124: EDFMESS FR - inis.iaea.org

�8

2contractor companies in France. I consider that the situation absolutely must be brought under control as soon as possible.The need to control the source term has led some plants to envisage joining the campaign to enable entry into controlled areas without a special protective suit (EVEREST Project), long since adopted by a few pilot plants (Golfech, Civaux and Cattenom Nuclear Power Plants). This project, which I strongly support, consists in ridding the radiation hazard areas of contamination and any significant sources of radiation, hence accessible in normal working clothes. In the three pilot plants, this initiative has made it possible to stay longer in the radiation hazard areas and thus increase the time spent on the job (see Chapter 9) and supervisor presence in the work areas. It is supported by all those working in the plants, despite the difficulty in achieving the situation, which necessitates setting in place a new organisational structure for logistics and field work.

Finally, I suggest that particular care be paid to upgrading the health and safety professions in general and that of radiation protection in particular, as well as to preserving, or even strengthening, the necessary skills so as to avoid the stop-and-go in the recruitment process seen in the past.

EDF Energy’s plants are mainly advanced gas reactors (AGRs) which generate lower collective and individual doses than pressurised water reactors (PWRs). The 20�� results are satisfactory. The collective dose stands at 0.�� man-sieverts per reactor. The greatest challenges in 20�� were the refuelling and maintenance outage of Sizewell B and work on the reactors of Hunterston B, Hinkley Point B and Heysham B.As in France, the individual doses are the subject of considerable attention. The highest annual individual dose was �.�2 mSv.The plants have been equipped with new detection portals at the exits of controlled areas that can detect very small traces of contamination, in accordance with international standards.The number of significant radiation protection events was the same as in 20�0. None involved a significant internal or external individual dose level or reached a level which required special reporting to the authorities.

2.8 Fire prevention and fire fighting

I am glad to be able to report that the 20�� nuclear safety results were positively affected by a significant drop in the number of fires.In the EDF-SA plants there was a fire in a transformer that confirmed the relevance of the organisational system combining the nuclear plants and the county accident response services (French acronym SDIS).

In the French plants, the fire brigade officers seconded to each plant are becoming a regular feature and the renewal of the first generation is proceeding smoothly. After background work on organisation and response, this new generation will need to focus on prevention and the quality of the exercises and drills, in which I have observed a certain lack of motivation among the participants. More varied exercises and drills carried out in radiation hazard areas are needed to improve the performance levels.

Although the fire fighting organisation has remained at the level that is extremely satisfactory, in the plants I observe slower progress in the uptake of prevention culture among the field workers, particularly contractor staff. The result has been that the number of fires of human origin was still too high. In the great majority of the plants the Better Housekeeping Project and the campaign to control fire hazards launched by the Nuclear Operations Division improved fire hazard mitigation, especially as a result of the strict rules about forming packages during unit outages. These arrangements facilitate the progress of the response forces and improve the way products and equipment are placed in interim storage in the facilities. I observed that the number of false fire alarms had decreased, increasing the

Page 125: EDFMESS FR - inis.iaea.org

�9

2credibility of the detection system as a whole. I regret to say that I also noticed that the second-line response team is not always at full strength when alarms have gone off.Checking the fire compartmentation system - the cornerstone of French practice - is again encountering difficulties due to obscure problems with databases in certain plants. This situation has been going on too long and needs to be rectified as soon as possible.

When I met the management team of the AMELIE spare parts management campaign, I was pleased to see that the organization of the fire-fighting arrangements for the general warehouse at Velaines had been set in place conjointly with the county accident response services as per nuclear plant practice and that EDF had contributed additional equipment to the nearby fire station.

2.9 Increasing nuclear power plant security

As was decided in 200�, the security of the French nuclear plants is being stepped up by the military police. The process is to be complete during 20�2, and 2 of the 20 plants are still awaiting secondment of personnel. For the remaining �8 plants, locally organised squads of specialised military police guards are now operational, their quarters have not yet all been built but work is advancing.Both the military police and EDF have expressed their satisfaction. During my visits, I observed the presence of military police on duty inside the plants. This calls for two observations: the EDF and contractor staff accept this new situation that had long been apprehended, and the EDF plants are now much more like what I see to be reasonable in other countries. The new arrangements also facilitate the administrative procedures involved when issuing authorisations to enter the plants. The procedures of EDF, the military police and the local authorities are optimised in a satisfactory way.The security installations have been reinforced and many of the old ones have been replaced. The working conditions in the surveillance stations of the security facilities have been significantly improved.Intrusions by an international association on � December 20�� were widely reported in the media, which was the intruder’s goal. I consider that such irresponsible and unacceptable acts in no way reflect any vulnerability of the system set in place by EDF and the military police. The intruders were unable to gain access to any sensitive areas and the guards handled the matter with discretion.However, more generally, it is once again my impression this year that the coordination between the Group’s security management, the Nuclear Operations Division security committee, the Nuclear Engineering Division’s engineering support (in the Electromechanical Department) and the local site protection services deserve an organisational system more close to that adopted for nuclear safety and the associated oversight.

The year was marked by a number of security-related inspections of the plants by the High Commissioner for Defence and Security (HFDS) attached to the Ministry for Ecology, Sustainable Development, Transport and Housing (MEDDTL)

I also note that, at the end of 20��, at the request of the French government and in agreement with EDF SA, the IAEA conducted an International Physical Protection Advisory Service (IPPAS) inspection, created by the IAEA to assist states in strengthening their nuclear safety rule set with particular reference to site security at Gravelines Nuclear Power Plant. On this occasion, it stated its findings and recommendations in terms of coherence with the international rule sets (INFCIRC/22�/Rev.� Guidelines) and the best international practices. As is the case with visits of this type, it will be followed up with visits to check that the recommendations are being integrated and to provide assistance as appropriate.I am glad to see that, as in the case of nuclear safety, this field is being opened up to outside appraisal.

Page 126: EDFMESS FR - inis.iaea.org

20

22.10 Decommissioning

The only nuclear generating facilities under decommissioning in the EDF Group are located in France.

This year, I visited the decommissioning site of the first French pressurized water reactor, located in Chooz in the Ardennes. The dismantling of this somewhat special facility, the nuclear part of which is located in a cavern, has been going on for a number of years. The work, begun in 20�0, now mainly relates to the reactor coolant system. The steam generators were being decontaminated to reduce their status on the nuclear waste scale. With the Nuclear Engineering Division acting as client (the Nuclear Environment and Decommissioning Engineering Centre, CIDEN), the work is being carried out by contractors, who have

to acquire the skills and experience necessary for such work that is highly specific to the four types of reactors under decommissioning: heavy water reactor (Brennilis), pressurized water reactor (Chooz), gas-cooled natural uranium fuelled reactor (Bugey, Chinon and Saint-Laurent) and sodium-cooled fast reactor (Creys-Malville). The work represents an interesting challenge, demanding new industrial solutions to specific technical problems. Finding solutions for permanent disposal of the different types of waste is complicated by the social and political climate in France, and can result in delays in the decommissioning work or the creation of costly interim storage facilities. This was recently the case at Brennilis where work was resumed in November 20��.I observe that EDF’s policy of decommissioning as soon as possible is now confronted with administrative, legal and technical obstacles that are being exacerbated by non-governmental associations and by the government authorities in the case of the treatment of radioactive waste.

Page 127: EDFMESS FR - inis.iaea.org

2�

3

NUCLEAR SAFETY MANAGEMENT

The commitment of the managers, at every level of responsibility in the company, is now clearly perceptible in the EDF Group's plants. The need, however, for proper prioritisation of projects is not yet sufficiently visible by staff in the field. Giving more responsibility to local managements would increase the effectiveness of action and reassure the staff.

Nuclear safety management must be designed to ensure good results both in the present and the future.Although the results again this year are basically satisfactory with real success in some areas, I have made a point of identifying areas where progress needs to be consolidated and where improvement is needed to plan ahead for the results of the future.

A nuclear safety policy for the EDF Group

I am glad to see that a start has been made on drafting the first EDF Group nuclear safety policy, covering all entities responsible for nuclear safety in EDF SA in France, EDF Energy in the United Kingdom, the Constellation Energy Nuclear Group in the United States and the Taishan Nuclear Power Company Joint Venture Company in China (in which EDF has a �0% stake).Its official release in early 20�2 will make it possible to re-assert, at the highest level, the absolute priority being given to nuclear safety under all circumstances and in all places, as well as EDF's commitment to strengthening the role of WANO after the Fukushima accident. It will also make it possible to re-state, if not clarify, the different levels of responsibility for nuclear safety in the organisational structure of the EDF Group. I encourage all managers of operations departments, power plants and engineering centres to take the opportunity of its publication to re-affirm that nuclear safety is paramount in every action of all the employees of the EDF Group and of its contractors.

Responsibilities of nuclear power plant directors

I note that the head of departments, especially in the power plants, need to feel that they are being listened to, understood and properly supported in their delicate task.The principle of subsidiarity, clearly announced at corporate level, is now adopted, with a better balance between the respective duties at corporate level and at plant level, where the operations responsibilities are the most direct.The introduction of the Associate Directors for Operations attached to the directors of the Nuclear Operations Division is a good example of this. Their prerogatives make them the official representatives of the nuclear power plant managements, with real power to arbitrate in the prioritisation of activities and the disposal of resources. Their main focus is the plants, of which they have an intimate understanding, which is of vital importance when managing a set of nineteen. Their job, which extends to all areas of integration, has the basic purpose of facilitating the work of the plant directors without absolving them of their responsibilities.

Page 128: EDFMESS FR - inis.iaea.org

22

3

Further increasing subsidiarity

This innovation is comparable to the arrangements in place for a number of years in EDF Energy, with three Chief Nuclear Officers for 8 plants. The latters’ role is however more focused on everyday performance management and, as there are few plants, they take a very active role in the field.

I consider these changes to be very positive. It needs to be made sure that they become permanent and, more generally, that this new subsidiarity remains within a well-defined context.

An independent organisation well-respected in the plants

One of the characteristics of the nuclear industry in France is the existence of an independent checking network with special skills alongside the managerial hierarchy: the Independent Nuclear Safety Network. This network is active at three levels: generating plant or engineering centre, operations division and the top of the EDF Group.I find that this Independent Nuclear Safety Network is robust and staffed with qualified personnel, in almost all the plants that I visited. It is heeded by the plant directors who regularly organise meetings with the heads of the Nuclear Safety Advisory Units and the Safety Engineers.I also observe with interest the EDF Energy initiatives to upgrade the “real time” nuclear safety checking systems in the plants, which need to be reinforced.In 20��, benchmarking with two French nuclear plants, followed with an experiment involving three British plants bore witness to the way fruitful exchanges are developing between the operators of the two different sets of plants.

A need for process skills

I would, however, like to point out the need for vigilance about the renewal of safety engineering skills and the frailty of the local audit teams. With the necessary operations skills in short supply, young engineers are filling Safety Engineer positions to an excessive degree.

The inspection programmes are carefully designed and scheduled for the year in accordance with a framework laid down at corporate level. I note that the framework leaves little margin for initiative in local checking, which should be examined, and I especially wish to again stress the improvement in the quality of the annual diagnostics carried out by the plants. For example, the plants establish their Annual Nuclear Safety Analyses by combining the different perspectives of the management and those of the Independent Nuclear Safety Networks.

I consider that the Local Safety Committees offer powerful leverage for progress, offering forums where views can be openly expressed by plant managers and assist in the decision-making process.

In other countries, I also meet operators that seek to reinforce their independent checking capability, at both plant and corporate levels. The staff working for the Chinese operator CGNPC, the South African operator ESKOM and the Russian operator Rosenergoatom, all have cooperation agreements that link them to EDF, principally with the Nuclear Inspectorate of the Nuclear Operations Division, and show a growing awareness of its importance.

The authority of the operations shift managers needs to be reinforced

In the nuclear power plants, the operations shift managers are responsible for conducting nuclear safety checks 24/�, under power delegated directly to them by the plant director. I note that this delicate task is respected in the plants, and that it is performed with competence and proper awareness of the responsibility. These key players also have the job of ensuring that operations are proceeding correctly, by supervising the activities performed in the nuclear units and handling the associated arbitration and prioritisation.

In France, I wonder about the ability of the operations shift managers to take stock of the situation as they are sometimes completely taken up with the current issues or activities that should be entrusted to

Page 129: EDFMESS FR - inis.iaea.org

2�

3

others. Their counterparts in nuclear power plants in other countries, particularly the USA and the UK, appear to be calmer in performing their everyday tasks, having broad and effective support; they are able to step back from the situation to the benefit of this “line of defence”.

During my visits to the nuclear power plants, I also noticed how a special direct relationship has been reinstated between the plant director and the operations shift managers. This makes it possible to effectively mobilise the shift teams about the nuclear safety goals, and operations in general. This type of link between these two management levels is vital for proper control over nuclear safety in operations and thus merits special attention.

Deciding in perfect safety

I also consider that special attention needs to be paid to the proper functioning of the decision-making processes related to operations in each plant. In these processes, the operations shift manager must always be the final decision-maker, and benefit from the indispensable support of expert advisers, if not the plant

management. The proper positioning of plant management level decision-makers (designated PCD� in the on-call rota system) merits the same attention.

I note the desire of the Nuclear Operations Division in France to give new life and effectiveness to the flagging Nuclear Safety, Radiation Protection, Availability and Environment Watch (OSRDE) initiative by recreating it in a simpler form so it can be used more often. Originally introduced by one of my predecessors, this initiative draws on concrete examples and encourages the nuclear operator to regularly assess the relevance of trade-offs made in the past that have implications for nuclear safety. I consider that, with the wave of new plant managers arriving, the need to revive this initiative is pressing. The same applies in EDF Energy and the Constellation Energy Nuclear Group, where operational decision-making reflects this (see Chapters �0 and �2).

Motivated managers in the plants

Once again this year, I observe the high level of commitment and motivation of the managers in all the plants visited. On all fronts, performing what can be a thankless task, they deserve special attention on the part of the plant managements.

I am glad to observe that proper training programmes for managerial positions are now becoming available and are universally appreciated. In this context, the EDF Energy programme for the development of leadership is truly remarkable. The stated intention of focusing the manager on their profession and on operations accurately reflects their aspirations and the needs of their teams. Yet, when I meet the front-line managers, they readily mention the difficulties they have with everyday logistics and the increasing deadweight that they call bureaucratisation. I again wish to draw attention to this real problem, which needs to be solved by resolute action. In accordance with international practice, the good practices that I observe in some plants consist in setting up dedicated advisory support in the departments; they merit further development.

Performance indicators

I would like to first emphasise the joint work carried out this year in the French and British plants to better establish coherent performance indicators and thus make benchmarking more meaningful (see Appendix �4).In the French plants, the number of nuclear safety significant events declared in 20�� remained high (at �0.� per reactor), although there was a perceptible drop in the number of events ranked Level � on the International Nuclear Event Scale (0.9� per reactor). This positive point is to be borne in mind, despite the declaration of one nuclear safety significant event at INES level 2 (at the beginning of 20��) after equipment failures in some of the 900 MWe series reactor standby generators (see Chapter �2).

Page 130: EDFMESS FR - inis.iaea.org

24

3

In EDF Energy, the number of nuclear safety significant events declared in 20�� dropped (4.� per reactor), with different declaration rules in France and the United Kingdom reflecting the requirements of their respective nuclear safety authorities. The number of events ranked on the International Nuclear Event Scale (all Level �), which is more directly comparable, was higher than in 20�0 and higher than in the French plants (�.� per reactor).

I consider that the degree of transparency in EDF SA and EDF Energy is satisfactory. It is based on more rigorous detection of minor discrepancies in the field, by the technicians themselves, and on unambiguous managerial practices reflecting solid nuclear safety culture.

Three results are particularly noteworthy, as they illustrate the return on investment in the action taken some years ago at both corporate and plant levels, relating to key issues: reactor scrams, compliance with the technical specifications for operation, and fire.

Reduced number of scrams

A record to beat

This year, the number of scrams in France was particularly low (0.�0 per reactor per year), the lowest level ever recorded in the French plants. This result places the EDF’s French reactors on an equal footing with the best in the world, with 38 reactors where not a single scram occurred during the entire year. It

is the result of long-term action taken over a number of years to improve both the reliability of equipment and human action. However, this year I note a higher proportion of scrams of human origin, suggesting that the Human Performance Campaign might be running out of steam, in terms of the practices designed to increase reliability.

At EDF Energy, as in the United States, there is a risk of lumping automatic scrams and manual shutdowns together, as the operating practices are different. British procedures more frequently require the operators to trip the reactor manually, before automatic action is triggered. The 20�� results are stable compared with the previous year (�.9� per reactor) and still not as good as the best results worldwide.

Managing the operations rule set

In the French plants, compliance with the technical specifications for operation (the nuclear operator's ‘highway code’) achieved its best levels since 2004. The support and training work undertaken in the plants meets the requirements very well, which explains these results. I recommend that these good practices be made permanent. The reduced number of excursions outside the authorised operation envelope, a point that I emphasised in my report last year, has also contributed to greater compliance with the technical specifications for operation and reflects the progress made with directing operations from the control room.

At EDF Energy, the results are stable in this area, without re-achieving the high performance levels of 200� to 2009. This represents one of the main avenues of progress that the British plants are taking.

Managing the fire risk

I would like to again emphasise the good results of the EDF plants in the area of fire (see Chapter 2), as this hazard represents a major issue and calls for determined action.The reduced number of fires in 20�� is spectacular (nearly 20%) and shows the relevant action taken in recent years. The plants all have excellent organisational systems in place to address the fire hazard, relying on a network of professional fire brigade staff (one per plant), and close cooperation with the county accident response services.

Page 131: EDFMESS FR - inis.iaea.org

2�

3

Fire prevention, especially in work areas, has not yet attained the level intended but fire action plans are in place everywhere and are being monitored.I note that arrangements have been made to ensure that the campaign to control fire hazards is made permanent which gives me confidence that this ambitious programme will be taken to completion. This is a clear illustration of the level that any responsible operator can achieve when pressed by the nuclear safety authority.

In the United Kingdom, the results were also good, with a significant reduction in the number of fires. The peer reviews, however, carried out by WANO in the last two years in the plants have frequently revealed weaknesses in the maintenance of the fire systems and management of the fire loads.

My main concerns

Sub-standard maintenance in work areas and sub-standard control of operations - most related to alignment and system configuration - are still too frequently coming to light in the directing of operations from the control room, demonstrating that the core skills are still insufficient, if not deteriorating. In a context of particularly high staff turnover, the acquisition of proper professional skills must be a priority.

Relevance and professionalism

I am most concerned about the common causes of these discrepancies which, all too often, reveal problems with the supervision of contractor work, failure to properly understand certain risk analyses, and technical checking by EDF staff that is not up to standards. In coming years, major efforts will be required to

reverse this trend.

Going the whole way with the actions in

progress

The use of error prevention tools, despite an unprecedented training drive in recent years, is still not sufficiently widespread. Certain professions are doing well, such as automation and fuel handling, but there is still a long way to go to close the gap between the practices observed in the control rooms in France compared with EDF Energy.

I am aware of the scale of the action taken, but consider that attention must also be drawn to the risk of burnout with the Human Performance Campaign if the managers do not more clearly express their requirements concerning these new practices. I consider that priority must be given to the changes in the operating department.Finally, while I am glad to report the reversal in the trend as concerns excursions outside the authorised operating envelope, I also have to report the recurrence of significant events involving reactivity control, particularly during reactor coolant dilution, as mentioned in my previous report. Determined action is required here to ensure proper training in carrying out such operations, which must in no case be standardised. The event described in Chapter �2 illustrates this type of issue.

An evolving corporate-level checking system

The Nuclear Operations Division relies on a solid nuclear safety process. Annual reviews make it possible to clearly establish the major issues by analysing the events in the plants in service and accurately determining the policy needed to improve matters, which are stated in the Annual Performance Contracts of the plants. The strengths and weaknesses of the plants identified in this work are good to know. The timing of the process has been revised and is now situated before the plant nuclear safety reviews, which are increasingly being linked to the standard timetables for the reviews of the Integrated Management System. This change has been made to accommodate a pressing need of the plants, who are adopting the Integrated Management System in increasing numbers. This enables priorities to be established further in advance, hence scheduling the action plans earlier in the year.

Page 132: EDFMESS FR - inis.iaea.org

2�

3

I note with satisfaction that, this year, the nuclear inspectorate was able to complete its checking programme, unlike in previous years, as I have previously had cause to regret in my reports. The significant increase in the number of inspectors is the main reason. It now remains to ensure that this positive development remains a permanent one.I take note of the decision to simplify the arrangements for checking the plants, by converting the overall safety reviews into overall excellence reviews. These overall excellence reviews - more broadly covering operations, for instance safety in the workplace and generating capacity - will be carried out every four years in each plant, alternating with the WANO peer reviews, the frequency of which is now changed to every four years (previously every six). Each plant will thus be evaluated by the nuclear inspectorate or by WANO teams every two years. This is very gratifying, and comparable to what is carried out by the Institute of Nuclear Power Operations (INPO) in the United States, for instance. I would also like to draw attention to the priority that must always be given to nuclear safety in these assessments and the need to provide the additional resources of adequate quality.

The reinforcement of resources made available to WANO, which was decided in 20��, to enable it to better conduct its nuclear plant assessment activities throughout the world will also require the secondment of experienced staff from the plants. I believe this issue will need to be monitored closely in order to satisfy these demands.

The Nuclear Engineering Division’s nuclear safety checking system, provided by its Assessment and Support Unit, is modelled on the nuclear inspectorate and still appears to be pertinent. It is also necessary to have sufficient inspectors to properly perform its duties.

At EDF Energy there is no internal inspection service that can make large-scale audits or assessments comparable to the overall safety reviews. The choice was made to rely on the WANO peer reviews for checking the plants, these being carried out every three years and featuring an external viewpoint. At corporate level, the Safety and Regulation Department possesses substantial engineering resources providing a clear picture and analysis of the results and the nuclear safety trends in the plants collectively and individually (comparable to the task performed by the Nuclear Safety Performance Group in the Nuclear Operations Department). Its linkage, with resident inspectors in the plants, remains perfectible and I note with satisfaction that, as mentioned, its strength is being reassessed, including reinforcement of the plant inspection teams.

The major projects are better prioritised yet the conditions in the field remain difficult

Supporting change

Having raised the alarm in the past about the risk of an excessive workload associated with a large number of campaigns being conducted in parallel in the nuclear plants (see Chapter 8), I welcome the decisions that were taken this year to stagger them over time. During my visits, I have observed that decisions

have enabled stronger multi-year campaigns to be organised that better reflect the ability of each plant to handle things. The results of this major change are, however, not yet being felt in the field.The still very great number of changes in professional practices and in organisation will require, on a long-term basis, greater backing in the form of support and help in alleviating the difficulties encountered by staff.

In conclusion, nuclear safety management is maintained by motivated players in the field but their action is all too frequently impeded by there being too many priorities and by insufficient logistical support. Nuclear safety levels are satisfactory, with undeniable successes and also weaknesses clearly identified by the managers. Improvement of the results requires priority being given to professional apprenticeships for the numerous incomers to the world of nuclear power generation, with more independence being granted at plant level.

Page 133: EDFMESS FR - inis.iaea.org

2�

4

THE FUKUSHIMA DAIICHI ACCIDENT

This natural catastrophe caused an event that shapes the future of the nuclear industry, transforms its worldwide governance and forces it to give thought to plant design beyond the original design basis. The EDF Group has shown a high degree of responsiveness in meeting the demand to re-assess nuclear safety and look at the future. This challenge facing the engineering forces must not have an impact on the support they provide to the operating plants in the field of nuclear safety.

Note: this chapter is based on international analysis available on 31 December 2011 and does not cover all aspects of the catastrophe.

On 11 March 2011, an extremely powerful earthquake occurred out at sea some �80 kilometres from Fukushima Daiichi Nuclear Plant operated by the Japanese corporation TEPCO, consisting of six boiling water reactors (BWRs). This caused reactors that were operating to automatically trip by activating the earthquake protection systems. Some circuit breakers and the electrical power system were damaged. The standby diesel generators automatically started to power the safety systems. Forty one minutes after the earthquake, the first of a series of seven tsunamis hit the site. The maximum wave height was �4 to �� metres whereas the site protection was designed to withstand �.� metres, which flooded the plant platform, including the standby diesel generators and the electrical power distribution equipment. Only a few batteries remained operable in Reactor �, and one diesel generator

Page 134: EDFMESS FR - inis.iaea.org

28

4

for Reactors � and �. Response was greatly complicated by the loss of lighting in the four facilities affected.Without cooling, core melt began that same day in Reactor �. In Reactors 2 and �, core melt was delayed as a result of cooling by the action of turbine-driven pumps that remained operable.The hydrogen produced by the deterioration of the fuel resulted in explosions in Reactors � and �, which greatly complicated response. The pressure rose in the containment vessels, making it necessary to vent them. During the venting, dose equivalent rates of �2 mSv/h were measured at the site boundary. A total of �00 people received a dose above �00 mSv but the contamination released into the atmosphere was 3,000 times less than at Chernobyl according to the IAEA.

An immediate reaction to the event

Throughout the world, the nuclear industry quickly responded to the event to understand and learn from it without delay:

11 March 2011: the accident occurs in Fukushima Nuclear Power Plant,18 March 2011: WANO issues a Significant Operating Experience Report entitled “Fukushima Daiichi Nuclear Station Fuel Damage Caused by Earthquake and Tsunami” (SOER 20��-02), giving all members � month to assess the degree of accident preparedness in their facilities and to submit reports,March 2011: governments begin requesting their nuclear safety regulators to conduct investigations,May-June 2011: A team of international experts from the IAEA representing �2 different countries conducts an investigation in Japan.June 2011: a ministerial conference on nuclear safety is held by the IAEA,October 2011: the WANO members unanimously decide to modify the way in which the association operates so as to reinforce its resources and authority,First half of 2012: peer reviews will be carried out by the nuclear safety authorities of the European Union countries,June 2012: the conclusions will be submitted to the European Commission.

First inferences reported by the Japanese authorities

A prompt and uncompromising

report

During the ministerial conference with the IAEA in Vienna in June, the Japanese government presented a comprehensive report on the accident. The section devoted to “the lessons learnt from the accident to date” was very relevant to EDF and all the other nuclear operators. It must be borne in mind that the emergency response was conducted in a situation where the earthquake and

tsunami had destroyed the public services, particularly the electrical power, communication and transport networks, as well as much of the area surrounding the plant, leaving more than 20,000 dead or missing. In addition, numerous aftershocks hindered the work. The lessons of the accident can be divided into five categories: the adequacy of the preventive measures, the adequacy of the technical response to this severe accident, the adequacy of the relief effort, the strength of the safety infrastructure, and the nuclear safety culture. The Japanese government's report merits a detailed study to find out all that can be learnt from it. To serve as examples, I have selected some aspects that can be learnt from each category.

Adequacy of the preventive measures

The arrangements for protecting the site against the effects of tsunamis were insufficient.The arrangements for managing severe accident situations were not defined in sufficient detail, being generally considered to be voluntary action by the operator rather than a legal requirement. Furthermore, the severe accident management guidelines had not been reviewed since their promulgation in �992, having been neither made more stringent nor improved.

••

••

••

Page 135: EDFMESS FR - inis.iaea.org

29

4

The measures taken to guarantee stand-alone operability of the individual nuclear units in an accident were insufficient.

Adequacy of the technical response to this severe accident

The accident response arrangements need to be supplemented with measures such as reinforcing radiological protection in the control rooms and emergency centres, upgrading the ventilation and air conditioning systems and upgrading the auxiliary systems, in particular by providing self-contained lighting and communication equipment.A suitable quantity of personal dosimeters, protective suits and emergency equipment must be available on-site. Most of the personal dosimeters and dose-measuring devices at Fukushima were un-usable as they had been damaged by the seawater.

Adequacy of the nuclear emergency response

The division of roles and responsibilities between TEPCO and government organisations at corporate and plant level need to be clarified.The response to offers of assistance from other countries and communication with the international community need to be improved. For instance, there should be a list of the supplies and equipment necessary to react effectively to any type of accident, the points of contact in the event of an accident should be established well in advance in each country and the arrangements for sharing information should be improved.

Strength of the safety infrastructure

Measures to take into account new knowledge and expertise, particularly as regards system design, need to be reassessed. A new legal framework will clarify the status of renovation work.Increasing the levels of skill in the human resources, operators and regulators should involve, in terms of training, focus on nuclear safety, nuclear emergency planning and response, emergency management and nuclear medicine.

Nuclear safety culture

The organisations and individuals involved in national nuclear regulations who are responsible to the public for nuclear safety, need to give consideration to whether they have made proper allowance for new knowledge with the responsiveness and the promptness needed.

The first lessons I have learnt

Unparalleled experience feedback

For my part, my first impressions before further study are:

• The importance of the initial design basis and the systematic reassessment of protection systems during ten-yearly nuclear safety reviews, as is required by French law,

The importance of having resources that can be rapidly deployed and can be quickly located on site so they can be easily mobilised. This is paramount when we consider the speed at which core meltdown can begin in pressurized and boiling water reactors in the total absence of coolant and electrical power.

The need for hydrogen recombiners or more powerful ventilation systems, in view of the devastating effects of the hydrogen explosions on relief operations.

The importance of providing connection points for emergency equipment, to provide assistance in plants that are isolated and partially destroyed,

Page 136: EDFMESS FR - inis.iaea.org

�0

4

The need to improve emergency management by including scenarios where an accident affects more than one nuclear unit in the same plant,

The need for skills of high levels that can be rapidly integrated into the decision-making systems of the government and political authorities,

The extremely harsh conditions in which the TEPCO and contractor staff bravely took action to re-establish a minimum of functionality, with little assistance due to the site being isolated.

Internal actions within the EDF Group in France

Within a few days, EDF mobilised their resources to offer and arrange immediate support measures. The French vendor AREVA, the French atomic energy commission CEA, and EDF investigated what means were available for response in hostile environments from the INTRA robotics group. The EDF Group also sought to assess the long-term impact on its nuclear plants. This was facilitated by EDF's integrated structure, with a strong engineering force to assist with design and operations issues.

“Kit inspections”

Compliance is a priority

I note EDF’s responsiveness in rapidly drawing conclusions which, together with those of INPO, were used to draft the specification in WANO's SOER 20��-02. By March, the goal was set of assessing each plant’s proper compliance of the installations and organisational structure with the design basis rule set, with

particular regard to the risks of flooding, earthquake, loss of the heat sink and station blackout. This work was positively received by the French nuclear safety authority (ASN).

I stress the importance of priority being given to rectifying the persisting discrepancies, particularly for accidents consisting of loss of cooling and of power supplies.I note that these analyses did not reveal any significant risks for the nuclear safety of the installations.

Supplementary nuclear safety assessments process

Supplementary nuclear safety assessments (or stress tests), carried out at the request of the French government or the ASN have made it possible to make assessments extending beyond the design basis of each of the plants. The results were submitted to the ASN in mid-September 20��, and were made public.The process consists in studying extreme situations on a purely deterministic basis, independently of the plausibility or probability of occurrence. It focuses on determining at what point substantial discharges into the atmosphere would occur and demonstrating the presence of any cliff edge effects associated with the loss of means used to protect the installations or control accident situations.For conditions beyond the current design basis, studies investigated earthquakes �.� times higher than the design basis ones, floods �.� times greater than the once in a thousand �,000-year flood levels increased by a margin and rainfalls two times greater that the design basis ones. Finally, allowance was made for total loss of the water and power supplies for eight days, while the current design basis corresponds to 24 hours with no off-site power supply or �00 hours with no heat sink for the entire site. EDF’s priority has been to finalise the studies in the assessment reports, to substantiate the proposed countermeasures and to implement them, as well as to hasten the implementation of certain modifications and actions to restore compliance with the requirements.

The high quality work carried out during the supplementary nuclear safety assessments demonstrated EDF’s ability to manage the design basis of its facilities. I also note that this work enabled the plants to re-appropriate their design rule sets. This is encouraging at a time when generation turnover risks losing some of the organisational memory.

Page 137: EDFMESS FR - inis.iaea.org

��

4

Finally, at the request of the High Committee for Transparency and Information Concerning Nuclear Matters, the supplementary nuclear safety assessment includes a third part (specific to France) relating to the conditions for using contractors in the nuclear plants in service. I mention this issue again in Chapter 9.

Continuous improvement

From their commissioning, the EDF nuclear power plants have been the subject of numerous upgrades and modifications intended to increase their safety levels, particularly during the ten-yearly inspections. These changes are made in accordance with the policy on continuous improvement of nuclear safety that has prevailed in France since the construction of the very first nuclear plants, and codified in the Nuclear Safety and Transparency Act since 200�.

Stress tests : confirmation of nuclear safety

The stress tests have enabled EDF to substantiate the existing design basis of the facilities, a position largely borne out by the successive reviews already carried out during the ten-yearly re-assessments of plant nuclear safety, and by rules set out from the design stage for the EPR-type reactor. For my part, I agree with this.

I also note that, for extremely severe accidents, these analyses confirm the accuracy, for the protection of populations against radioactive discharges, of the findings concerning the resources implemented in the wake of the Three Mile Island and Chernobyl power plant accidents.

Countermeasures proposed

Robustness and resilience

Up to March 20��, situations as extreme as those studied had not been followed by the definition of specific countermeasures in view of their low probability. The study of additional countermeasures is therefore being proposed to increase the robustness and the resilience of the installations.

As regards robustness, additional arrangements have been selected such as reinforcing the electrical power supplies and the water make-up systems. These arrangements constitute part of a ‘hard core’ of measures capable of withstanding beyond-design-basis events. Planning these measures is a major step forward but I would like to draw attention to the need for restricting their perimeter so as not to excessively complicate the installations.As concerns resilience, the goal is to have both human and material means available at corporate and plant levels that are able to respond flexibly to a wide range of unplanned situations. The creation of the Nuclear Rapid Response Force (FARN), operational on-site in less than 24 hours, answers this concern. The FARN should eventually have a staff of around �00 divided between the plants and who will continue to perform their original professions on a part time basis.

I fully agree with the steps planned in these two areas and I approve of the priority that EDF is giving to implement these countermeasures in concordance with the ten-yearly re-appraisals. Their application, however, must allow for the actual situations in the field. The problem of skills that is currently so acute, especially for operations staff, and the need to prepare for an increased engineering workload leads me to advise everyone to keep a careful watch over how these measures are implemented. My worry is that, as a result of an increasingly heavy workload, the plants and the engineering centres will find it difficult to handle their tasks, priorities will be managed in a way that cuts out issues relating to everyday nuclear safety. Obviously, allowance must be made in budgeting for these new measures. The availability of a pool of quality industrial suppliers will be vital in implementing these plans.

Emergency management

Severe accident management features three clearly-defined centres of responsibility: the operator, the nuclear safety authority and the government, all acting at two levels: local and national.

Page 138: EDFMESS FR - inis.iaea.org

�2

4

At EDF, the national-level emergency management resources have always been the subject of special attention and the associated organisational structure has proved its ability to support the plants in actual and simulated incidents. There is, however, a need to examine the consequences of an event affecting multiple reactors in the same plant, at both local and central level, for all the parties involved. Thought has also been given to the conditions under which response personnel would be working in an emergency, and it will be necessary to take into account experience feedback from Fukushima on their radiological monitoring and protection, as well as on their living conditions in the affected facilities.

I would like to draw the attention of the government authorities to what the accident in Japan has revealed, the number of evacuees involved, and the implications for training their nuclear emergency decision-makers, which could be carried out on an emergency simulator for instance.

Level of nuclear safety, the ASN’s ruling

In the ruling it issued on � January 20�2 after the supplementary nuclear safety assessments, the ASN re-affirmed the fact that the operator bears prime responsibility for the nuclear safety of the facilities. The ASN is to perform checking in the name of the government.

It “considers that the safety level of the facilities is sufficient for it not to request immediate action in any of them. Notwithstanding, it considers that their continued operation requires increasing without delay their ability to resist external events beyond the existing nuclear safety margins.”

The ASN will therefore impose a set of requirements on the operators and make the existing ones on nuclear safety more stringent with regard to protection against natural hazards (earthquake and flooding). It notably stresses EDF’s proposals to define a ‘hard core’ of material and organisational provisions to deal with extreme situations and the progressive setting in place of the Nuclear Rapid Reaction Force from 20�2.

It notes that social, organisational and human factors have been the subject of special attention and describes the priorities regarding the renewal of personnel and the way recourse to sub-contracting is organised.

It requests improving the process for handling cases of discrepancies and violation of the rule sets as regards hazards associated with other industrial activities.

In the United Kingdom

The board of EDF Energy Nuclear Generation rapidly took action to assess its plants, after a meeting held on �4 March. This initiative was backed up by the publication of SOER 20��-02 by WANO on �8 March. Meanwhile, the government appointed the chief inspector of the Office of Nuclear Regulation to conduct a nuclear safety assessment, and he issued a preliminary report in May and a final one in September 20��.

Working method

After the Fukushima accident, EDF Energy in the United Kingdom and EDF SA in France maintained close communications on the methods of assessment and the hypotheses adopted. This initiative guaranteed that the approach would be consistent within the group.At the same time this was taking place at national level, the European Commission, via the requirements of European Nuclear Safety Regulators Group, drew up a document indicating how stress tests were to be conducted with the European Union. There are many points in common between the Office for Nuclear Regulation (ONR) and the European Nuclear Safety Regulators Group requirements,

Page 139: EDFMESS FR - inis.iaea.org

��

4

however the former extended its scope to include post-accident management as well as openness and transparency in the nuclear industry.The originality of the Office’s approach is to ensure that experience feedback from the accident is obtained and handled using current practices and existing regulatory processes.

Reactors of different types

High-inertia reactors

It is to be noted that, due to the way they are designed, advanced gas-cooled reactors respond differently from pressurized water reactors to severe accidents. Advanced gas-cooled reactors have a power density that is some 40 times less than pressurized-water reactors and substantial thermal

inertia. In the event of loss of cooling, the rate at which their temperature rises is far slower. In addition, the fuel cladding is made of stainless steel that does not produce hydrogen at very high temperatures in the presence of water vapour, unlike in pressurized water reactors in which the cladding is made of zirconium.Sizewell B is a pressurized water reactor of recent design and was assessed in a manner consistent with that of EDF's rectors in France.

The level of nuclear safety, the Office for Nuclear Regulation’s ruling

In the context of this overall nuclear safety assessment, EDF Energy did not identify any shortcomings in any of its plants.In the Office’s preliminary and final reports, it is stated that “in view of the direct causes of the Fukushima accident, we (the ONR) see no reason to place restrictions on the operation of nuclear power stations in the United Kingdom." and "the British nuclear industry has responded in a responsible and appropriate manner, showing leadership in the field of safety and solid nuclear safety culture in its reaction to date.". The reports conclude that the British approach to establishing design rules for nuclear facilities provides proper allowance for external events. The requirement to make periodic reassessments of the nuclear safety case and to submit them to the Office constitutes a reliable way of making sure that the nuclear facilities have been upgraded to reflect progress in technology. Otherwise, they would be closed down.

Changes being studied

Technical modifications in the facilities are being planned to increase their ability to resist natural events, including the possibility of rapidly connecting emergency equipment. The storage of emergency equipment off-site will be increased to enable response teams to take action more quickly.

Improved emergency

preparedness

The resources of the regional emergency response centres are to be reinforced and could be used in any type of emergency, nuclear or otherwise.

Training, severe accident procedures, emergency exercises and communication systems are all being upgraded in the field of emergency preparedness.

As regards openness and transparency, the creation of centres to receive the public in every plant is being envisaged.

Page 140: EDFMESS FR - inis.iaea.org

�4

4

In the United States

In the United States, the response to the Fukushima accident by the nuclear industry was carefully organised by INPO, the Nuclear Energy Institute (NEI) and the Electric Power Research Institute (EPRI), involving the creation of a steering committee charged with coordinating the efforts of seven working groups.

Like elsewhere in the world, the operators all responded to WANO’s SOERs and the equivalent reports by INPO, by conducting the required plant reviews.All agree that the NEI, with the assistance of industry partners, collected and made available information on the accident in a highly effective manner, satisfying the demands of the general public, the media and politicians.The Nuclear Regulatory Commission (NRC) created a near-term taskforce consisting of experts with the task of re-examining the American processes and regulations, before issuing recommendations for contingent changes in policy.In my exchanges with INPO, one issue came up often: after a serious nuclear accident in any country, all the main national and international organisations would have to work together as the events unfolded. They would need to rely on well-established relationships and networks. Before any such occurrence, the heads of the different organisations need to know each other and understand what roles they would be called to play.As regards emergency response, it appears that the American plants are well organised with the local administration, but at federal level the Federal Emergency Management Authority (FEMA) does not appear to have solid arrangements in place for a nuclear crisis.

Constellation Energy Nuclear Group (CENG)

An extremely active operator

CENG is deeply committed to post-Fukushima actions in the American industry. Its Chief Nuclear Officer is supervising one of the seven above-mentioned working groups, and represents CENG in the steering committee. Furthermore, CENG is coordinating the response to INPO Event Report ��-4 on station blackout.

Internally, CENG is intiating action relating to:

assessment of earthquake resistance and study of flooding margins,

capacity of the batteries, improvement of the in-pool instrumentation, reinforcing of the boiling-water reactor vent lines, and bringing the severe accident guidelines up to standard,

purchasing of mobile equipment and studies on rapid hook-ups,

on-site assessments from November 20�� to April 20�2, as well as delivery of mobile equipment in the first half of 20�2.

Different types of reactors

CENG operates two boiling-water reactors at Nine Mile Point, on the shores of Lake Ontario. The basic design of Reactor � at Nine Mile Point is very similar to that of the first nuclear units at Fukushima (Mark I type).In their other two plants (Ginna and Calvert Cliffs), CENG operates three pressurized water reactors.

Page 141: EDFMESS FR - inis.iaea.org

��

4

Level of nuclear safety, the NRC’s ruling

In July 20��, after the Near Term Task Force of experts from the NRC issued a report featuring twelve themes and some thirty recommendations and requirements. They declared the American plants to be “safe and that none need be shut down”. I note that the experts expressed the need to clarify the requirements relating to "design basis events" and "beyond design basis events". Furthermore, they report lack of coordination of the voluntary initiatives of the industry, where rigorous checking by the NRC does not occur.

These recommendations, which were made public in July 20��, were divided by the NRC into three categories in October 20��, under Letter SECY��-���:

short term: earthquake and flood assessments, loss of electrical power supplies, increased dependability of boiling-water reactor vent lines, reinforcement of severe accident management procedures,medium term: fuel storage pool water make-up ability, emergency preparedness,long term: ability to vent pressurized water reactors, mastery of the hydrogen risk.

Analyses and initiatives by other international players

The international nuclear safety community (Global Nuclear Safety Regime)

The term is used by the International Nuclear Safety Advisory Group of the IAEA. It expresses the collective international enterprise that seeks to define a level of performance expected of all operators and regulators, to monitor their performance, and to build their competence and capability. It includes all the national and international efforts and networks. More detail is given in the report INSAG-2�. Post-Fukushima, a number of important steps have been taken in the international arena and in strengthening the global nuclear safety regime. I consider that it is vital for the key players of the nuclear industry, particularly EDF, to contribute more actively by their effective and sustained participation in the life of this community.

WANO

WANO commits its credibility

I note that WANO reacted promptly to the Fukushima accident in close liaison with INPO and made sure that all its members rapidly checked their plants and their accident preparedness. The process was found to be useful, revealing a number of areas where improvements could be made, such as training,

documentation, maintenance and checking of emergency equipment.Amongst other things, a fourteen-member commission was set up to recommend modifications to the programmes and the functioning of WANO. During the biannual general meeting of WANO held in Shenzhen in October 20��, the WANO members unanimously approved the recommendations. They related to:

broadening the scope of WANO’s activities,

establishing an integrated worldwide strategy for response to events,

increasing the credibility of WANO, notably by modifying the peer review procedures in depth,

increasing the visibility of the association,

improving the quality of all the products and services on offer.

••

Page 142: EDFMESS FR - inis.iaea.org

��

4

A substantial increase in its resources will be necessary. In particular, the members will need to make resources and competent staff available to the association and its programmes. This contribution will remain modest compared with the direct costs resulting from the Fukushima accident and its impact on public confidence. The EDF Group has actively supported this commission and its recommendations which should increase the credibility and power of WANO.I think that all the nuclear operators should be responsible for ensuring that this programme is properly executed as concerns timing and content.

The IAEA

At the end of May 20��, an international team of IAEA experts with representatives from 12 countries went to Japan. They visited Fukushima Daiichi, Fukushima Daini and Tokai Daini Nuclear Power Plants. Their findings bore out the inferences mentioned above. The visit brought other issues to light:

The response undertaken on the spot, under extremely difficult conditions, by specialised personnel who proved themselves both committed and determined was exemplary. It made it possible to adopt the most appropriate approach for ensuring nuclear safety in view of the exceptional nature of the situation.

Highly-professional logistical support was set up, especially by the arrangements taken in "Village J” (centre providing access to the site, located 20 km away) to provide protection for the response workers that needed to go to the sites.

The action plan set up to regain control of the accident-stricken reactors is important and well-recognised. It needs to form part of a broader plan possibly leading to decontamination of areas outside the sites affected by radioactive discharges in order to enable the evacuees to return to normal life. The entire world would thus bear witness to the measures that might be necessary to cope with this type of extreme nuclear event.

The nuclear regulations need to be designed to make allowance for extreme off-site events and be the subject of periodic re-appraisals. They also need to preserve the independence of the nuclear safety authority and clearly indicate the role of all the players under such circumstances, as required under the IAEA's standards.

A large amount of information was obtained by the team. I was impressed to learn that at the neighbouring site of Fukushima Daini, where flooding caused less damage than at Fukushima Daiichi, TEPCO and contractor staff installed nine kilometres of temporary electrical cables in sixteen hours, making it possible to re-establish nuclear safety functions.

Page 143: EDFMESS FR - inis.iaea.org

��

4

In conclusion, as in the case of the Three Mile Island and Chernobyl accidents, the Fukushima accident has affected the future of the nuclear industry.

I consider that the reaction of the nuclear operators of the EDF Group was prompt and proportionate, both in terms of the actions speedily put in hand and the quality of the response to the requests of their respective nuclear safety authorities.

I note that during this unprecedented assessment procedure, the nuclear safety of the EDF Group's reactors proved to be adequate in terms of the current rule set. The allowance for even more severe natural events will increase the level of their nuclear safety.

I would once again like to draw attention to the engineering resources that will need to be mobilised for these studies and works at a time when, in France in particular, the engineering staff are already busy with other nuclear safety requirements. They will also need to cope with large-scale turnover among their personnel.

Nuclear accident preparedness in the countries where the Group has interests, involving government authorities, operators and safety authorities, will need to be re-assessed to determine its ability to cope with an accident such as Fukushima, where a number of reactors on the same site are threatened and large-scale evacuation is required. The states and their services need to take decisive action in this field, and I encourage their progress.

I consider the decisions taken in October 20�� by nuclear operators from around the world under the aegis of WANO are necessary as they greatly strengthen the international oversight of the nuclear operators’ activities. I am glad to be able to report that EDF has played an active role in this process.

Page 144: EDFMESS FR - inis.iaea.org
Page 145: EDFMESS FR - inis.iaea.org

�9

5

TRAINING TO DEVELOP SKILLS IN THE PROFESSIONS

In France, the company has been confronted with an unprecedented generation turnover. Planning for the general refurbishment of the EDF SA plants will soon put great pressure on many skills. Managing these challenges without compromising nuclear safety will require managerial commitment to training and the professions.

The French aspect of this state of affairs is considered in this chapter. This year, massive recruitment is continuing with the hiring of nearly 2,000 engineers and technicians in the nuclear sector. First, they will be compensating for 40% of the personnel in the engineering and operations divisions that will be retiring between 20�0 and 20��, with departures peaking in 20�4. This is the back of the wave of staff that were hired to build the nuclear plants in the late seventies and early eighties, which was not properly planned for in the previous decade.

Recruited staff will also have to cope with the General Refurbishment workload and the modifications due to the experience feedback from the Fukushima accident. As the time required to train and become a true professional is long in the nuclear industry, it is necessary to make special arrangements to cope with the demand.The Nuclear Operations Division and the Nuclear Engineering Division have accordingly set in place a very ambitious programme. I observe that this programme is based on an in-depth analysis of the situation and was inspired by the best practices worldwide.

Page 146: EDFMESS FR - inis.iaea.org

40

5

Rare skills

I would like to stress that the replacement of experts is vital to the future of the company, although they may be few in number. It will need to be planned ahead field by field, including in the nuclear fuel and the R&D divisions. In the plants, the same applies to the rare skills that require foresight to be acquired in view of

the corresponding professional experience needed.

Skills

The landscape

The effects of the rationalisation and productivity drive in the first half of the last decade are proving difficult to overcome. As the time taken to train and become a true professional is long, the renewal of skills remains a major concern for the plant managers despite an unprecedented effort.

In the operations branch where the time taken to become a true professional is even longer, collective experience is diminishing. In 200�, there were � experienced staff for each new arrival, in 20�2 there will only be � for each new arrival, with the situation improving in the following years. This situation calls for changes in the way activities are organised and checking is conducted in the teams.

In the maintenance profession, the situation is different but remains problematic with � experienced staff member for each new arrival in 20�2. The influx of new employees raises the question of how different generations will get along together. The incomers’ approach to work is different and the company will need to take this into account.

It is in engineering where the dearth of experienced staff will be most keenly felt, whereas the EPR-type reactor work is the most intense, the engineering force for the plants in service needs to maintain the performance levels of the �8 reactors and the issues raised by the Fukushima accident are keeping nearly �00 engineers busy in the Nuclear Engineering Division and the Research and Development Division. To this must be added the resources needed for decommissioning work. In view of this large and increasing workload, it will no doubt be necessary to work differently. Why not seek and encourage reinforcement by other engineering companies?

In the Nuclear Fuel Division, a professional trajectory has been developed to compensate for the number of staff taking retirement before the end of the decade, compromising some of the expert appraisal skills. The creation of the fuel community in 20�0 is fostering cross-functional management of skills throughout the branch, divided between the different divisions of the company. Here, the annual recruitment of some ten individuals appears to be sufficient to meet the needs of the Nuclear Fuel Division.

Engineering, a priority

The engineering personnel needs of the Nuclear Engineering Division and the Nuclear Operations Division for the next ten years amount to �,000 to �,�00 engineers.

Reinforcing engineering skills

In 20��, more than 400 incomers joined the Nuclear Engineering Division, a trend that will continue for the next three years. Although most of the needs are covered by hiring young graduates, candidates with professional experience are especially coveted. I would like to draw attention to the highly-competitive fields

(civil engineering, chemistry and project managers) and the highly-specific ones (instrumentation and control, and radiochemistry). I can only encourage the recruitment of such staff. In particular, I would like to stress the efforts of the Construction and Operation Expert Appraisal and Inspection Centre to recruit chemists of the highest calibre.

Page 147: EDFMESS FR - inis.iaea.org

4�

5

I would like to emphasise the creation, in 20�0, of special training arrangements for incomers in the Research and Development Division lasting two to three years to ready them for the operations and engineering professions. These joint Nuclear Engineering Division and Research and Development Division hiring actions may be limited in number but constitute an excellent initiative.

Mobility

I observe that the internal job market has not regained its fluidity, but I commend the work carried out on cross-division career trajectories by the Production and Engineering Directorate. Faced with the lack of experienced personnel with the skills needed, most of the plant managers prefer to keep their staff. I can only encourage that the human resources policy take into account the particularities of the incoming generation, which does not have the same expectations as the outgoing generation. I am glad to see the setting up of Professional Career Supervisor positions in every department of the Nuclear Operations Division and the Nuclear Engineering Division. I will make a point of meeting them in my forthcoming visits.

I also note the hesitation, if not the reluctance, of the supervisors to make the decisive step of accepting an executive position. The perception of the duties and constraints of an executive position and the temporary loss of certain advantages are the main reasons. Progress with functional mobility has been insufficient and social promotion stalls at the transition from technician to executive.

The many people retiring are draining the pool of experience. I am left wondering about how to solve the problems associated with these trends. Might it be possible to render jobs more attractive to keep the personnel needed, even if this means consenting to work part time? Has full use been made of imagination in finding the best solutions?

New challenges after Fukushima

Managing a multi-unit

accident

The Fukushima accident will have an impact on the plants’ human resources. The stress tests show the need for supplementary skills in carrying out the minimum amount of response work in exceptional complex situations. Up to now, the postulation of an accident in one nuclear unit of a plant offered the possibility of reinforcements arriving from the other units. The deterministic hypothesis of an

accident in all the nuclear units in the same plant would prevent mutual backup. To meet the need for reinforcement, additional experienced operations personnel will be needed in the coming years.

Consideration also needs to be given to the creation of the Nuclear Rapid Reaction Force, which will consist of Nuclear Operations Division staff: some ��0 staff, seconded to this new activity for ��% of their time. I will be watching how this new system is set in place.

Furthermore, the decisions taken by the operators that attended the WANO Biannual General Meeting in October 20�� involved providing the organisation with more manpower and some 200 experienced engineers have been seconded to it. For the EDF Group, this means that 4� experienced executives will need to be made available to the WANO centre in Paris. I stress that high-calibre personnel must be sent and the undertaking be fulfilled as they appear to be essential for nuclear safety and the future of the nuclear industry.

Cooperation with EDF Energy

Cooperation has been arranged with EDF Energy to conduct recruitment jointly in the United Kingdom. Some twenty engineers thus joined the Nuclear Engineering Division this autumn on secondment for two years before returning to EDF Energy.

Page 148: EDFMESS FR - inis.iaea.org

42

5

Exchanges of personnel are also developing between EDF Energy and the nuclear departments of the Nuclear Technology Division. I encourage these exchanges, which should intensify with the construction of EPR-type reactors in the UK.

Training

Training requirements more than doubled between 2008 and 20��. This increase is expected to continue. At the present time, scope for recruitment is limited by the capacity to receive new arrivals and train them into professionals. In view of the numbers, the quality of training is vital to successfully renewing the skills.

The training potential has been considerably reinforced with the creation of the ‘nuclear generating and engineering’ vocational academies. The ‘nuclear generating’ academies provide more than �00,000 trainee-hours, half of which are given by instructors and half by plant operations staff.

Training resources

The vocational academies also provide basic training for staff joining the nuclear divisions. All are equally appreciated by both the operatives and their managers. This must be continued. I hail the creation of ‘specialist knowledge’ vocational academies, following on from the ‘common core’ academies, in an increasing number of areas.I also note with interest the creation of a training scheme for the plant directors newly appointed by the Nuclear Operations Division. This rounds off the range of training intended for managers, from front-line supervisors to heads of departments and unit outage project leaders. At the Torness site in Scotland, I saw a leadership training scheme that seemed highly innovative and which I recommend.

I think that more attention needs to be paid to “just in time” training. Training before a delicate operation is always highly beneficial in terms of mastering nuclear safety. I am all for this type of training that is closely integrated with the operations process, particularly when simulators are used. The development of ad hoc mock-ups for specific types of equipment also forms part of this type of training. During my visits, I noticed some excellent initiatives which deserve to be more widely known.The practical training centres in the plants are excellent for carrying out more general training or training focused on human performance. I encourage the plants to use these practical training centres to the best advantage of their staff and that of their contractors. The full-scale simulators are vital learning tools for operations. They need to be upgraded to be more representative of the different control rooms in service. I am glad to see that the simulators are being used more and more frequently. Here, I would like to draw attention to the fact that Gravelines Nuclear Power Plant has six nuclear units and needs a second simulator.I insist that the operations departments organise themselves better and enable the operators to have training on demand. In other countries, simulator sessions are more systematic and better organised, being integrated into the teams’ timetables.

Once again, I would like to point out that there is a need for more experienced instructors. The training centres of the plants in other countries commonly have twice or three times more staff. Beyond this major effort, there is a need to make better use of the potential of this profession to attract the top professionals and make sure that their feedback to operations will be good. On one British site, I observed a short-duration secondment system, lasting between six months and a year, which appeared to be giving satisfaction.

Page 149: EDFMESS FR - inis.iaea.org

4�

5

Greater independence

Hands-on training of managers

Apart from the resources, there is room for progress when it comes to transferring more responsibility to the managers. In my visits to the plants, I have all too often observed managers with insufficient concern for training their staff. I suggest that the departments take greater charge of training and that the managers be free to say what is lacking in the staffing and functioning of their departments so

they can have the training modified accordingly. The commitment to skills also means identifying the training requirements by appropriating the programmes and by getting managers more involved in the training.

The question of local independence in training has been raised, with the need for liaison between the nuclear plants and the Operations Engineering Training Unit, which acts a prime contractor for training. I have taken note of the experiments conducted at Golfech, Civaux, Chooz and Tricastin Nuclear Power Plants and the planned extension to include six other plants in 20�2. I had these shown to me at two of the first plants. I am convinced that a degree of independence is beneficial as it places more responsibility on the managers when it comes to training. The Operations Engineering Training Unit needs to support this important transition. This alternative approach will only be successful if all involved actually agree on granting more independence to the plants.

Systematic Approach to Training

This year, during my visits to Koeberg and Sizewell Nuclear Power Plants and to INPO, I was given presentations of the internationally-recognised Systematic Approach to Training that is used in many companies. This consists in determining the scope of training on the basis of the actual needs for skills in departments using job descriptions supplied by the plant management for which the training is performed and which has defined the management’s priorities. Koeberg Nuclear Power Plant, which is located in South Africa, has opted for it. I also appreciate the pragmatic approach preserving the spirit of the method that was used at Sizewell Nuclear Power Plant to achieve full accreditation of its training department by the end of 20�2, as is the case at the other EDF Energy plants.

I note that the Nuclear Operations Division has not yet set itself the goal of accreditation, preferring to concentrate on the approach in depth according to the INPO's rule set.

Training that is closer to the

technician’s task

For all the technical professions in the Nuclear Operations Division, a rule set is being accordingly prepared on the basis of the activities that the professions are to perform. The experts of the professions and the Operations Engineering Training Unit are considering designing more than �,�00 courses for the Nuclear Operations Division with the Operations Engineering Training Unit as prime

contractor. By means of this work which is due to continue until 20��, my wish is for the training to be less theoretical and more closely focused on the professions.

EPR training

Staff training began in 2008 well in advance of schedule. The control room simulator is already in service, which is important as �0% of the operating personnel are actually new arrivals in the company and that the control room is totally computerised. The training is, here again, planned in accordance with the Systematic Approach to Training methodology, which is a guarantee of quality at a time when it will be necessary to provide one million trainee-hours throughout the period leading up to commercial operation. In China, I note that problems associated with the lack of experience is cleverly countered by accrediting all operators on Chinese reactors with computerised control rooms, such as the CPR �000 series, before moving on to the EPR-type reactor.

Page 150: EDFMESS FR - inis.iaea.org

44

5

Apprenticeship and mentoring

This year, I observe a readily-understandable disquiet among the most experienced staff as to the manner of transmitting their experience and know-how before retiring. The adoption of sandwich courses and mentoring by the Nuclear Operations Division provides at least a partial response to this question. Mentoring is, in my opinion, vital to the success of sandwich courses. The number, quality and motivation of the tutors will all call for vigilance. The inflow of new arrivals is so great that I have, on occasions, encountered mentors who have only been in the company for three years, which can hardly be sufficient.

Lycée André Malraux in Montereau

The Lycée de Montereau has been a partner of EDF in the nuclear industry for many years. I commend this partnership, selected for its proximity to the nuclear plants at Nogent, Belleville, Dampierre and St. Laurent, and for the presence of the dynamic teaching team that is a stakeholder in this ambitious project. I discovered all this when I visited the establishment. The ‘valve system’ practical training centre is particularly well equipped.

The Lycée provides initial vocational training in nuclear technology up to school-leaving level, and in-service training for EDF and nuclear contractor staff. All students awarded diplomas are able to find jobs. Alongside the conventional training for valve and electronic technicians, I noted a course for contractor staff supervisors. In 20��, for the first time, a “nuclear environment” vocational diploma was developed for apprenticeship training for future work area managers and supervisors.

It is my wish that the French education system recognise the importance of the industry’s requirements and the awarding of official diplomas. Relations between EDF and the Lycée are excellent and the company needs to maintain its support, notably by continuing the cooperation with Nogent Nuclear Power Plant and through the supply of equipment.

In conclusion, to make it possible to compensate for departures and cope with the workload in operations and engineering, we need to make training more closely reflect the needs in the field and to assign the necessary resources to it. The integration of the best international practices is a very promising development. Success will depend on the full involvement of the management teams in training.

Page 151: EDFMESS FR - inis.iaea.org

4�

6

THE OPERATIONS PROFESSION

The operations departments are key players in the field of nuclear safety. Scope remains for improvement in the activities that are central to the profession. Upgraded staff training and stronger leadership on the part of the operations shift managers would help to improve the quality of operations and foster calmness.

For a nuclear plant, the results in terms of nuclear safety as in production are closely linked to the presence of a truly professional operations team that is recognised as such by the other professions and manages everyday activities with authority.The competence of the players, its standing in the eyes of the plant management and its ability to encourage process skills in other professions, are all factors that are essential to success and should be found in all the EDF Group's plants.

A major recruitment drive, weakness in human resources remains

In the French plants, the operations professions are also confronted with high levels of staff turnover which, in 20�2, will result in the supervision factor reaching two experienced staff for each new arrival.

Page 152: EDFMESS FR - inis.iaea.org

4�

6

Invest even more in skills

In all the plants visited this year, I saw the scale of recruitment having taken place and the intensity of efforts made in the plants to train and integrate the incoming generation. It seems that despite the unprecedented efforts, the forward planning of jobs and skills is still not up to standard in most of the

plants. Filling the different positions in accordance with the scheduling — as regards technicians, operators and team leaders — has substantially improved but will require more time and sustained efforts. This is especially the case as many skills derived from the operations professions are in short supply in many other departments and still severely lacking for positions of the highest importance: simulator instructors, safety engineers, outage supervisors, sub-project leaders in both the unit outage and unit in-service teams and engineering.

Progress in compliance with the rule set and recurrent shortcomings in the execution of activities

Concerning one vital issue, I note with satisfaction that compliance with the technical specifications for operation, the ‘highway code’ for reactor operations, is an area where there has been substantial progress. This is the result of large amounts of support and training work carried out in many of the plants, including the active involvement of the operations shift managers and safety engineers.The progress achieved in most of the plants in the field of administrative lockouts also needs to be emphasised.

The profession comes first

However, I note regular control room surveillance discrepancies and perceptible deterioration as regards errors in alignment and system configuration. Concerning this last point, the causes identified cover virtually the entire range of what operations staff ought to know, clearly illustrating that there is a problem

with the key skills central to the profession. It seems that the so-called field activities, performed outside the control room throughout the plant under possibly difficult conditions, merit greater attention on the part of the plant and operations managers. Closer observation of the practices, which are often little known, and the everyday problems encountered by the technicians could reverse this trend.I can also see how certain difficulties in the application of operating rules and procedures during normal operation sometimes reflect a lack of confidence in documentation that has recently been updated. Much work on standardising the operating procedures as part of the Methods and Practices Harmonisation Campaign has indeed been carried out in recent years, providing greater allowance for experience feedback and rationalising this vital documentary activity. But every modification of the rules or instructions seems to result in people losing their bearings. This is also true of maintenance work. It is therefore essential to take care to provide the users with proper backup during this transitional phase. It is positive that the operations shift managers and the operating staff still have plenty of confidence in the incident and accident situation operating procedures, which they regularly test on simulators. This is a factor that greatly inspires calm.

I also observed, in most of the plants that I visited, that the error prevention tools are not being used systematically both in the control room and in field activities. An unprecedented investment in training has taken place in the plants, both in the practical training centres and on simulators. Still, at the present juncture, the penny still has not dropped and we are still far from what I can observe elsewhere, particularly in the control rooms of EDF Energy.There is a real risk that the human performance practice campaign is running out steam. Since this is an innovative project that is no longer consuming many resources, I call on the plant managements to take action to ensure that the penny drops, particularly with the new intake.

The operations shift managers need to assert their authority

The operations shift manager’s task is to provide nuclear safety checks 24/�, under powers delegated directly to him by the plant director. These key players also need to ensure operations are proceeding smoothly, by supervising all the activities begun by the nuclear unit staff, determining the best trade-offs and allowing for the constraints of the maintenance services. They all manage an operations team

Page 153: EDFMESS FR - inis.iaea.org

4�

6

and thus have the task of setting the standard for rigour in operations for each of their staff and developing a work atmosphere that encourages willingness to learn and openness to the outside.

In light of my visits, I wonder if they are able to see things in perspective, as all too often they are caught up in the work in hand or activities of which they could well be relieved. The operations shift managers that I have observed in nuclear power plants in other countries and belonging to other companies, particularly the USA and the UK, appear to be calmer and freer in the execution of their everyday tasks. They have broad and effective support and are able to step back from the situation to the benefit of this line of defence.

The initiative taken in the French plants to better organise work during operations and outages, which are inspired by the best international practices, such as AP 928 and outage control centre (OCC), are a step in the right direction and increase the support from which the operations shift manager benefits. I recommend that care should be taken to ensure this position is given its proper standing in any new organisational setups.

Operations shift supervisors are real managers

I would also invite the plant managements to carefully consider the managerial dimension of their operations shift managers who, to succeed in inducing major changes in certain professional practices, need to see their authority backed up and their initiative encouraged.

Training and drilling the staff: an absolute priority

The basic training of the staff is of high quality with a substantial syllabus for the reactor operating personnel. It should also be noted that specialist knowledge academies for technicians and field staff are becoming universal. These academies, co-organised by highly-experienced technicians, fulfil the requirements and ensure that knowledge is properly passed on to the next generation.

I observe with satisfaction the widespread introduction of a more formal assessment of operator skills on simulators, using a realistic and pragmatic method. This change, inspired by international practices, is of a nature to inspire confidence in outsiders, especially the nuclear safety authority.

Training: an activity in its own right

Nevertheless, the training and drilling of operations staff still falls far short of the Nuclear Operations Division's target of �� days of drill on a simulator per year for each operator, which is only being met in a few plants. Problems are arising with work organisation and even simulator availability in the plants with four or six nuclear units.

I note excellent initiatives in fieldwork, for instance scheduling ad hoc training for the staff concerned before each unit outage, as is the case in the plants of the Constellation Energy Nuclear Group and EDF Energy. The practice is still insufficiently widespread.From what I have seen in most plants that I visited, there is a regrettably small number of training days that are carried out with entire teams outside shift work. These actually constitute excellent opportunities for developing team spirit and bringing the managers closer to their teams, which is very important for properly handling operations under normal, incident and accident conditions. The real problems encountered with work organisation, which make such practices rare, appear to need re-examination.

More generally, I consider that the presence and the involvement of managers in simulator usage remain insufficient, for both heads of operations departments and plant management staff. Such involvement, so perceptible in the EDF Energy plants for instance, provides leverage for progress and increases the effectiveness of training.I am glad to see that the skills programme, drawn up by the Nuclear Operations Division after hard work on opening up internationally, is now going ahead. It needs to be supported with time and resources, but the goal is clearly announced; fostering more responsible attitudes to training in the plant managers.

Page 154: EDFMESS FR - inis.iaea.org

48

6

Changes to be progressively made in the professions

The announcement of ambitious objectives for all the technical professions has been the subject of plans for the professions that were set in hand more than five years ago in the nuclear plants. This involved an operations professions steering group laying down new job descriptions for these professions, extending from increased technical knowledge for expert field staff to further skills for control room personnel, with the introduction of a third operator and the unit control operator.Reinforcement of the team leaders under the operations shift managers, with associate operations shift managers and operational safety associate managers is also a key feature of the changes in the professions.

It is plain that these changes are only partially complete in the plants that I visited, and there are still strong reservations among those concerned. Their introduction is also being slowed in many operations departments by the rate at which skilled staff can be supplied, despite the fact that the special training schemes are now heavily subscribed. This is simply going to take time and continued investment.

The control room: a working environment that needs to be sacrosanct

If there is a place in a nuclear plant where calm needs to prevail, it is indeed the control room. I note with satisfaction that considerable progress has been made in recent years and, during my visits, I see concrete initiatives being taken in the field, boosted by a corporate-level directive that makes a great deal of sense. However, we are still far from the best practices that can be observed outside France.Access is still not limited to the strict minimum, particularly during unit outages, and the operations staff do not always seem to be sufficiently firm or imposing to ensure that the rules are obeyed. Practices to increase reliability in the control rooms which, by their very nature, render the operations sacrosanct, are still very far from being fully adopted.

The modernisation of the instrumentation and devices used in the control rooms, which can also greatly contribute to calmness there, is not, in my opinion, being given high enough priority. For instance, I am still sorry to see that there is no means of viewing operational data remotely outside the control rooms where it can be accessed by the maintenance services; this is particularly a problem in the 900 MWe nuclear units.

For Flamanville �, I note with satisfaction that these issues have been taken into account in the EPR-type reactor design, going as far as reviewing the entire organisation of the control room activities. Without questioning the validity of the proposed change, I note that the ergonomics and the man-machine interface are the subject of a major human factors qualification programme.

In conclusion, the operations professions, like others, are faced with a situation of unprecedented generation turnover. The EDF Group will have to invest massively in human resources in France. Total success with this investment is vital for operating safety. This, however, depends on how successful the major changes in the content of the professions prove to be, by not only substantially improving the training system but also encouraging greater openness in the operating staff to the outside world.The plant managements will need to focus on operations in their deliberations and decisions to strive for a higher level of nuclear safety and better operating results.

Page 155: EDFMESS FR - inis.iaea.org

49

7THE ENGINEERING SERVICES AND OPERATIONS SUPPORT

The French nuclear power plants now face an unprecedented challenge. They will only be able to overcome it with the effective and responsive support of their own engineering services. These services need to be better organised and rapidly find ways of coping with a heavy workload.

For the French plants, EDF has the ambition to regain its place of leader in the nuclear industry by its levels of nuclear safety and performance. Extending plant life to �0 years and integrating the experience feedback from operations and the aftermath of the Fukushima accident will be part and parcel of its challenges for the next few years. To deal with these issues, EDF will be making good use of its integrated engineering capability, which plays a vital role, especially in its support for operations.EDF’s engineering force is vast and diverse, and it is not possible to discuss all its tasks and component parts every year. This year, I have chosen to focus on the “designer after-sales service” it provides (referred to as engineering support for the plants in service) and on the corporate level bodies that provide the operator with support.

Succeed with the Major

Refurbishment

The Nuclear Operations Division cannot implement its General Refurbishment programme without the proper backing of the Nuclear Engineering Division’s design engineering force. Also, unassisted, no nuclear plant can prevail against the challenges that await it. Operator support organisations are indispensible, whether they are the engineering centres of the Nuclear Engineering

Page 156: EDFMESS FR - inis.iaea.org

�0

7

Division or the corporate departments of the Nuclear Operations Division. Although everyday operation is a matter for the plants, the quality of the support they receive to maintain the proper standards, is the determining factor. I wish to stress that, despite the numerous initiatives launched in recent years, it still remains difficult to obtain this support. The actions taken to clarify the responsibilities and simplify the interfaces have not yet achieved the desired results.

Engineering support for the plants in service – the process

I have observed during my visits to the plants and engineering centres that the process involved in providing engineering support for the plants in service remains complex and ponderous. The organisation and decisions of engineering support for the plants in service are still extremely centralised, with reporting at many levels. I can see the gap between the operators’ needs, faced with difficulty in improving or even maintaining its levels of performance, and what the engineering support for the plants in service can actually provide. For instance, it cannot currently respond to modification requests in less than four or five years. This seems to be the case in most engineering centres of the Nuclear Engineering Division.

Responsiveness to plant request

All too often, the corporate-level engineering force is unable to satisfy the plant requests in due time, either because its organisation and resources prevent it from doing so, or because it has other priorities. Indeed, I find that the engineering centres in charge of engineering support for the plants in service are constantly

having to make complex trade-offs to optimise their resources. The aftermath of the Fukushima accident this year is another example of this, and one that has made the situation worse.In this changing situation, the role of the joint teams, the bridgeheads of the engineering support for the plants in service, is vital. I understand how much they facilitate the uptake of modifications by the plant departments and, in return, warn the engineering centres of the constraints associated with handling certain issues. I appreciated meeting teams that were both motivated and proud of their profession, despite the size of the workload and the tardy and fluctuating scheduling. The gap between production and engineering in France is less perceptible elsewhere, as the plant engineering teams are more substantial. This being the case, the joint teams are, as I see it, in a good position to facilitate dialogue between the plants and their engineering support.

Engineering projects for the plants

The joint Nuclear Engineering Division and Nuclear Operations Division project

The above situation reduces the capacity of the nuclear plants to cope with the challenges they face. The operator needs a more responsive engineering force for its plants that is more attuned to its needs and preoccupations. Meanwhile the Nuclear Engineering Division engineering force needs to step back and see things in proper perspective. The “joint Nuclear Engineering Division and Nuclear Operations Division project for the plants in service” launched in 20�0 is intended to remedy this. I approve its goals which consist in, first, providing design skills in each plant and, second, making it clear to the Nuclear Engineering Division what corresponds to asset-related engineering for the major long-term projects and what corresponds to responsive design-related support to satisfy the operator’s immediate needs. However, I regret the delay in the project this year, even if this may have a lot to do with the mobilisation of the staff for the stress tests after the Fukushima accident.

Reinforcing the in-plant engineering capability

In my previous report, I have already warned about the weakness of the plant engineering capability and the flood of requests that this generates for the engineering centres. I commend the decision to provide design capability in each plant and encourage the managers of the Nuclear Engineering Division and the Nuclear Operations Division to provide staff as soon as possible for these engineering forces that are indispensable to the operator. It seems appropriate to

return the powers to the plants and I am glad to see that one of the project goals is to reduce the time taken to respond to their requests.

Page 157: EDFMESS FR - inis.iaea.org

��

7

Coping with ever-increasing workloads

Finding engineering

partners

The engineering centres in charge of engineering support for the plants in service are constantly having to make complex trade-offs to optimise their resources. I am not convinced that this is always taking place at the correct level. This mode of regulation incumbent on the engineering support for the plants in service will not be appropriate for the increased workload in the future. The amount of plant-

related activity is set to more than double. Even if the massive recruitment, which I commend, takes place, the engineering force will not be able to double its manpower in such a short lapse of time (see Chapter �). The company will no longer be able to stagger the jobs to smooth the load, as it has done in the past. Perhaps preparations need to be made for another approach to doing things and having them done that is more orientated towards partnerships and cooperation with other engineering forces? Under the circumstances, EDF’s integrated engineering capability would gain from opening up more to the outside and seeking collaboration or other tried and tested solutions.

Governance of the engineering support for the plants in service

Today, the operator needs to participate more actively in establishing the priorities for engineering work and what needs to be abandoned for the plants. As part of the joint Nuclear Engineering Division and Nuclear Operations Division project, a working group has proposed changes in the governance of the engineering support for the plants in service, particularly in the functioning of the still very numerous managements, directorates and committees. The technical steering committee jointly set up by the managements of the Nuclear Engineering Division and the Nuclear Operations Division will establish the basic policy for the plants. I will be watching with interest its work and the results which, I believe, will improve the functioning of the engineering support for the plants in service.

The 2012 engineering methods project

At the engineering centres that I visited this year, I was struck to see that there were still no ‘off-the-shelf’ type project management tools that are found in many engineering offices. The engineers are all too often still using their own locally-developed applications. This does not facilitate either interfacing or overall engineering performance. Faced with the great increase in work for the plants in service and for its new jobs, the Nuclear Engineering Division has begun an in-depth analysis of its methods of directing and monitoring projects. The result of this analysis was the 20�2 engineering methods project, which kicked off this year. The Nuclear Engineering Division’s big projects concerning new nuclear facilities, as well as the engineering support for the plants in service, should rapidly benefit from this project. The work on the tools and methods can greatly boost the productivity of the engineering force, which must take care not to forget that the future workload will be considerable and well above their current capacity.

The corporate departments of the Nuclear Operations Division

In the Nuclear Operations Division, two corporate departments - the Central Technical Support Department and the Operations Engineering Unit - centralise the support and expert appraisal capacity necessary for plant operation and maintenance. Halfway between design engineering and in-plant engineering, these two departments carry out projects for the plants that are particularly important in terms of nuclear safety and performance.This year, I visited the Operations Engineering Unit where I note the high quality of staff work and their contribution to the technical performance of the plants. The Operations Engineering Unit directs, on behalf of the Nuclear Operations Division management, a number of projects and processes for which it has acquired, and is recognised to have, special know-how. Its responsiveness in dealing with the many cases entrusted to it is renowned.

Page 158: EDFMESS FR - inis.iaea.org

�2

7

Developing assistance to the

plants

Under the current circumstances, when both progress in nuclear safety and higher performance levels are being sought, I question the need to direct or manage from Paris projects that essentially relate to the plants. I see this as a rather unusual positioning of the corporate operating departments in France, of which the tasks are too frequently orientated towards directing (major projects) or

organising (processes) at the expense of providing assistance and support for the plants, which remains a side issue. As I see it, this merits analysis and comparison with the practices of other big operators in America and Britain.

This year, my visits included a trip to the Institute for Nuclear Power Operations in Atlanta (INPO, see Chapter ��). INPO handles four vital functions for all the American plants: assessment, experience feedback, assistance and training. The assessments focus on performance and are conducted by experienced auditors; they systematically integrate nuclear safety function reviews at plant and corporate levels. The experience feedback originates for the American plants and from all over the world. The events are analysed by the INPO engineering staff and pooled between all the companies. Assistance is systematically proposed to the plants whose results are deteriorating and imposed on any plants in difficulty. The action includes support visits, analysing difficulties and defining monitoring indicators. Training is available to all the companies and their contractors, featuring numerous sessions that are permanently available. This ensures that all staff members are properly qualified and have the necessary know-how before authorisation.

An example to follow

With nuclear units of many technologies and ages belonging to different operators, the American plants can nevertheless be seen as more homogeneous in their way of managing nuclear safety and performance than the French plants, although the latter are standardised and have a single operator. I am left

wondering about the causes of such a difference. It appears that INPO is contributing to this state of affairs by leaving it to the plants to take the initiative but providing its support when performance levels deteriorate and focusing its action on assistance. In view of the challenges that the French plants will soon be facing, I note that thought is now being given to the tasks and stance of the Operations Engineering Unit and the Nuclear Inspectorate. I will make a point of keeping abreast of the conclusions reached.

In conclusion, the French nuclear plants need to be able to count on a reliable and effective engineering force to rise to the challenge of the General Refurbishment. The Nuclear Engineering Division and the Nuclear Operations Division are jointly seeking optimal solutions for satisfying the plants’ need for assistance. The short- and medium-term engineering workload will not be cleared without opening up more to the outside and finding new work methods. As with other operators, the corporate operations departments need to provide more effective support and assistance to improve the results.

Page 159: EDFMESS FR - inis.iaea.org

��

8

MANAGING PROJECTS

Project-mode management fosters nuclear safety and cost-effectiveness. The success of projects requires directing them professionally and setting goals that are consistent with what the relevant organisational structure can absorb. They need to be more focused on how to instigate and support change, particularly when the stakes are high. This is the case for the EPR projects which represent major challenges for the Group.

Many campaigns important for nuclear safety are managed in the form of projects throughout the EDF Group. These campaigns are very diverse in terms of scale and success (nuclear safety, duration, quality and cost). If their development is not sufficiently effective and rational, projects can have negative effects on other modes of work simply because of their sheer number. Conversely, well-organised projects constitute a precious management tool.

Taking up complex technical challenges

In some cases, I have seen that specific technical problems with implications for nuclear safety and production are managed in the form of projects. They include, for instance, the post-Fukushima tests and studies as well as certain problems affecting only one or two nuclear units, such as the hot box domes of the advanced gas-cooled reactors. As a general rule, these activities are properly organised and targeted, and project-mode work is a useful asset.

Page 160: EDFMESS FR - inis.iaea.org

�4

8

Scheduled unit outages

Another step to take

All the scheduled unit outages, such as those for refuelling, ten-yearly inspections, steam generator replacement and regulatory outages all constitute “projects”. I have noticed that they vary in their quality and results. The nuclear units with the best performance levels in the Group have the skills needed, divide the work

into appropriate packages, arrange to have the spares in due time, fix the content of the outage, get the main contractors involved in the planning and scheduling, and then carry out the activities in a safe, rational and effective manner. The levels of unplanned capacity loss factor (UCLF) in such nuclear units show the quality of the work. At a time when larger-scale work is being contemplated, particularly in France with the General Refurbishment, we should think about the efforts that need to be made in some plants to attain excellence.

Instigation of change at national level

Perfect is the enemy of good

Many campaigns are carried out at national level, particularly in France. This is not the place to describe them individually but some of them need to be mentioned: the O2EI Project (better housekeeping campaign), the INTEP Project (campaign to introduce new technology into the plants in service), the MOPIA Project (campaign

to set in place an attractive business policy) the AMELIE Project (spare parts management campaign), the PHH Project (human performance practices campaign), the SAT Project (systematic approach to training campaign), the SDIN Project (nuclear technical information system campaign), the BMA Project (campaign for the harmonization of practices), the REX Project (experience feedback campaign), the OCC Project (campaign for continuous monitoring of installations during unit outages) the AP 928 Project (campaign for work management), and the AP 9�� project (campaign to increase equipment reliability). With �8 reactors in �9 plants, many projects are pilot-plant based to prepare for the deployment of new methods. This is good practice, of which I approve. It involves making choices to determine how much we should standardise and how much initiative we should leave to each plant. I think that in every project, the proper balance needs to be found, with due allowance for the principle of subsidiarity under which, whenever possible, responsibilities are delegated and individuals are encouraged to show thoughtfulness and commitment in their work.

The successful projects –and there are many– maintain precise objectives and their progress is continually assessed and monitored. When they are completed or have made sufficient progress, the integration of their content into standard procedures needs to be studied with care and managed effectively to preserve what has been achieved. I have found that this is not always the case. Furthermore, the number of campaigns managed as fleet projects seems to be greater than most of the plants can handle with their current operating cycles and human resources.I note, however, in 20�� that efforts were being made to stagger the projects in time. I consider that the multi-year scheduling in each plant, reflecting the possibilities and capabilities of each plant, is a good practice. The nuclear plant governance systems (e.g. the 20�� operations and maintenance methods programme for the AP9��, AMELIE, SPP, SDIN and BMA Projects) also meet this need and should facilitate success in the plants with all these projects. These positive trends show that perseverance is called for as they are now staggered over a considerable length of time (until 20��).The creation of the position of Associate Director for Operations is a positive point: these directors help set up multi-year programmes and analyse project risks, while prioritizing and determining trade-offs.

I also note that as soon as a project management team is assigned to a major technical problem (such as one associated with steam generators or alternators) there is a risk that it will no longer be properly coordinated with other parallel projects and activities.

In the United Kingdom, I have observed work organisations for certain technical domains in which supervisors responsible for all the plants collaborate with individual plants to draw up transition maps (deployment schedules) over five years. The timing and order of deployment of the steps mentioned in

Page 161: EDFMESS FR - inis.iaea.org

��

8

the schedules are left to the discretion of each plant. This approach comes across as being pragmatic. Nevertheless, it is clear that with 8 plants, the British situation is simpler than in France where there are �9 plants.

The approaching peak in activity

Thinking differently

The goal of the joint Nuclear Engineering Division and Nuclear Operations Division Project is to guarantee that the two divisions schedule and field the resources necessary to effectively manage the approaching peak in activity associated with the third and fourth ten-yearly outages, plant life extension and post-Fukushima

modifications, as well as the arrangements needed for the EPR-type reactors. From the exchanges with the Nuclear Operations Division management, I gather that "to reduce the activity peak by delaying work is not the correct solution and that, on the contrary, it is necessary to confront it”. I totally agree with this.

Key decisions will need to be made about what parts of the work are to be outsourced and what are not, the specific work methods to be adopted and the way the work is organised, what replacement and spare parts will be used and the skills and manpower to be engaged by EDF and its main contractors.

New projects

Synergy to be developed

The Group’s current EPR-type reactor projects (at Flamanville, Taishan and Hinkley Point) are gigantic. They are not all equally advanced, and each has its own specific context and work organisation. As concerns project management and nuclear safety, I note the following vital points:

Leadership must be unambiguous and the organisational structure must be as simple as possible if excellence is to be achieved. The best results are obtained when the general managements of the main players in the project communicate regularly and maintain synergy where the parties involved combine their efforts to achieve the same goal. This facilitates respecting schedules, as well at the dissemination of a veritable culture of quality at the stages of design, fabrication, erection and testing, on which a proper nuclear safety culture can in its turn be based, which is indispensable for plant start-up.

A scheduling process that it reliable and comprehensive, covering all aspects of the contributions from the different players and all necessary linkages. For example, the links between design and procurement for the licensing safety case and the timetable for their execution must be indicated and clear to all: architect engineer, the designer that orders the equipment, the vendor, the quality control staff, those writing the procedures (operations, maintenance and training), the teams in charge of erection and commissioning, etc. Such scheduling makes it possible to follow the progress with the project and to manage any contingent issues and modifications in total safety and under proper control. It is also possible to schedule the surveillance activities in advance to ensure that suitable resources are available at the due time to guarantee a level of quality that corresponds to the designers' intentions.

Project teams having the right resources and the key players grouped in the same geographical location insofar as possible. The joint engineering office set up in 20�0 by the boilermaking contractor and its Chinese client for the Taishan Project is a good example of this.

Using tried and tested standards and methods for managing data, documentation and work organisation (for instance the design of piping and the routing of cables).

Exemplary safety in the workplace and housekeeping at all stages of construction. Everyone must be aware of the need to preserve and protect the equipment delivered to the worksite, before and after erection. This is particularly important in the case of coastal sites. I have observed that sometimes this is not done properly.

Page 162: EDFMESS FR - inis.iaea.org

��

8

Skills and experience in all domains, in particular the placing of contracts, scheduling, nuclear safety management and the surveillance and coordination of work.

In conclusion, work being carried out in project mode is essential to ensuring effectiveness in the nuclear industry. As I indicated in my report last year, nuclear safety and cost-effectiveness are closely linked. Effective project management is one way of optimising both at the same time.

Page 163: EDFMESS FR - inis.iaea.org

��

9

TOGETHER WITH CONTRACTORS TOWARDS BETTER OVERALL

PERFORMANCE

Year on year, dissatisfaction with contractual relations can be found on both sides since the requirements grow as fast as the corrections that are made after each fruitful exchange. The solutions and the commitments frequently result in superficial complexity without treating matters as comprehensively as the nuclear industry demands. It must be made sure that the charter for progress and sustainable development, as well as the recent agreements on socially responsible subcontracting at EDF, usher in change in the status quo as regards industrial relations in EDF SA.

The role of contractors in the nuclear plants is a vital one in terms of both cost-effectiveness and nuclear safety. Some 80% of all maintenance work is contracted out and this proportion may rise with the increase in the number of ten-yearly inspections, the General Refurbishment and the work required as a result of the experience feedback from the Fukushima accident.A wide range of means of improving the conditions under which contractors operate are being deployed in the plants at the initiative of EDF. Here I wish to give an up-to-date list of the current avenues of progress.

Page 164: EDFMESS FR - inis.iaea.org

�8

9

Better sharing of strategic issues with contractor companies

The MOPIA Project (campaign to set in place an attractive business policy) offers a full set of tools but has not been as effective as expected. Many arrangements have been made to improve the business conditions for contractors, particularly clearer contract terms and the development of incentive schemes (best bidder, bonuses etc.). I am afraid that although these arrangements have, in general, been well received by the contractor companies, they do not offer sufficient progress as regards sharing the strategic issues. There does not appear to be any substantial progress in the quality of work or in the planning and duration of the field work to better control the outage lengths.The desired change will not come from the contractors alone; EDF will need to take action, a point that I will return to later. It seems that this is the main problem here and I wonder about the reasons for the current situation. Is it really so difficult for a client to regain control of the requirements that it imposes? Have past practices taken such a hold as to be unchangeable?

Training fosters progress

I think that the first step should relate to the contractor training set-up. Training should, as I see it, extend to all staff, not only field workers. All working in the nuclear industry must be conscious of the associated requirements relating to nuclear safety, industrial safety, radiation protection, the environment, quality

assurance, etc. It is true that EDF provides support for all companies when they begin working in nuclear field, but the acquisition of a nuclear safety culture must be constantly fostered and developed by the contractor company itself, with the assistance of EDF if necessary. A joint strategic vision cannot otherwise exist. EDF needs to provide itself with the means to create such a situation for all its contractors. One of the obstacles resides in the fact that the EDF budget is relatively small for certain companies.In the USA, INPO – an association of 2� nuclear operators with a total of �04 nuclear units – provides the training and assessments of workers at all levels in the contractor companies. Checking is selective and the operator and contractor management teams are directly involved. I am convinced that this arrangement is more effective than that currently used at EDF.

Planning work at the plants further ahead

Foresight and planning mean

success

Here, again, I call on EDF and its contractors to work this out between them. Despite perceptible progress, I find that EDF does not clearly define its outage programmes sufficiently early and does not award all the contracts in due time. Under the circumstances, the contractors do not have the resources available to concurrently plan their work in the different plants where

they have been chosen.Scheduling can be disturbed by the very numerous modifications incorporated into the EDF plans, which are issued either as soon as they are specified or in batches. These modifications cover technical issues, of which nuclear safety is one. No outage for maintenance resembles the previous one. This type of situation also exists in other countries, though their technical rule sets seem to evolve more slowly. The positive side of the situation for EDF is that there is on-going action to increase nuclear safety. Is the required effort, however, compatible with the resources and skills of the Nuclear Engineering Division, the Nuclear Operations Division and the contractor companies?At Sizewell B, an EDF Energy plant in the United Kingdom, I encountered a team of planners during a visit who were working with contractor representatives �8 months before an outage. Some EDF plants have pro-active partnership policies that call for contractor involvement long before unit outages when the resources make this possible.

Page 165: EDFMESS FR - inis.iaea.org

�9

9

Improving working conditions

Rethink and simplify the rule

set

As concerns the physical conditions of work at the plants, I observe that the situation is improving year by year and this is confirmed by the contractor representatives that I meet. Yet, the many technical and administrative formalities that have to be endured before a job can actually begin and the unpredictability in the scheduling generate unacceptable waits. The

‘hands-on’ time when the professional action covered by the contract is really taking place, is only a small proportion of the total time spent. I have identified the main factors that regularly cause trouble:

Organisation of the working environment which places a heavy burden on the field worker and his supervisor. The contractor job supervisor is responsible for industrial safety, radiation protection, fire prevention, the technical case files and lockout for field work, which all amount to a sum of preparations before arriving in the work area.

Instability of the scheduling of jobs, as previously mentioned, which results in delays as adequate consideration is not given to alternative scenarios.

Missing spare parts which can interrupt jobs and result in workers being switched to other jobs scheduled at other times during the outage, with the final result of the overall schedule being upset.

Special tools which are sometimes unavailable when unit outage activity is peaking and often result in long waits before the stores can issue them.

Logistics which need to be improved to create a proper workflow (for instance, airlock equipment and assembly, scaffolding and removal of heat lagging).

These weakness need to be rectified as soon as possible in light of the big jobs ahead.In the best-organised plants outside France, I find that the number of scheduling staff is generally considerably greater. Apart from controlling the official schedule, the critical path (the red line for the outage) is monitored at very close intervals and alternative scenarios are prepared and assigned resources to be able to cope with contingencies with minimum disturbance. I also note that field workers are generally given far more support and that sometimes specialists arrive in the work area just in time, a bit like a surgeon in the operating theatre.

Sub-standard maintenance

For greater visibility

The year 20�� has been marked by more cases of sub-standard maintenance than previous years, with the responsibility shared between EDF SA and its contractors. The amount of EDF contracting work varies from year to year, depending on the reasons for the outages (e.g. a large number of ten-yearly inspections in progress),

the number of modifications and specific work programmes (e.g. the O2EI Project, the fire action plan and the General Refurbishment). Without sufficiently long contracts, contractors can hardly be expected to retain staff with the required skills. Furthermore, pressure to reduce contract prices often results in cutbacks in supervision.

More staff motivation

More responsive recognition

Bonuses and penalties are frequently totted up annually, which is readily understandable in terms of company management. Yet the final result is often zero since the penalties cancel out the bonuses, which is very discouraging for those that deserve bonuses.

I have observed in the US that motivating the players is a major concern in company management and bonuses are paid out very soon after jobs are completed.

Page 166: EDFMESS FR - inis.iaea.org

�0

9

Bringing surveillance up to standard

In EDF, surveillance is being developed to provide quality assurance for major jobs. The staff members in charge of surveillance hone their professional skills in the field and attend special new courses on complex activities. Nevertheless, the profession differs between plants and divisions, is given little credit and has not established a career trajectory. Yet it is a full-fledged profession and the volume of work in the next few years will make it a very important one.

Better work means less supervision

The ideal system would enable contractors to get everything right the first time, with external checking by EDF sized accordingly. Prevention would then pay dividends.

In the joint teams in plants at the Nuclear Engineering Division, surveillance is affected by seasonal ebb and flow in the activities, and reinforced by sub-contracting to cope with waves. This situation merits being continuously assessed.

What I have seen, however, during my visits at plant and corporate level is many players taking energetic action, and additional organisation interfaces appearing to limit the effects of such situations. I hail them and respect their work but I still deplore the waste of time and trouble caused by weaknesses in organisation which are prejudicial to nuclear safety.

Numerous avenues of progress identified at corporate level show promise, such as the OCC Project (continuous monitoring of installations during unit outages), the arrangements for outage planning, as well as the AMELIE Project (spare parts management), provided sufficient scope is left for local initiative and pragmatism. For instance, giving more prominence to contractor assessment records throughout industrial catchment areas is a good decision. Indeed, corporate-level assessment of a contractor organisation consisting of a number of regional offices with different levels of performance is hardly meaningful.Another difficulty which I often hear mentioned in the plants is that some jobs and generic contracts are carried out by companies selected at corporate level without the plants having a voice in the matter. The plants having built up fruitful relationships with other contractors thus see their work wasted.

As part of the stress tests requested by the ASN after the Fukushima accident, the issue of "reliance on sub-contracting" was selected. This corresponds to a particularity of the French system and I see no direct link between it and the accident. In Japan, the number of sub-contracting levels for the same job can appear numerous in the eyes of Europeans. I have never heard of such practices in France.

In conclusion, the contractor companies, the joint ventures and the organisations representing the professions and EDF SA must, I believe, do more to shape their mode of operation as one of partnership. The charter for progress and sustainable development and the agreement on socially-responsible sub-contracting in EDF represent steps in the right direction, but do they provide sufficient leverage to rapidly fulfil the ambitions announced for the nuclear plants?

Page 167: EDFMESS FR - inis.iaea.org

��

10

EDF NUCLEAR ACTIVITIES OUTSIDE FRANCE

The operations context differs with the country but everywhere the same priority is given to nuclear safety and the constant challenge of achieving high performance. There were no serious events that affected nuclear safety in any of the 20 nuclear reactors in EDF’s scope outside France. The numerous good practices identified within the Group need to be more effectively shared in the operation, construction and commissioning of new reactors.

This year, I visited some of the nuclear facilities of the EDF Group outside France. EDF has a majority stake in EDF Energy, which operates fifteen reactors in the United Kingdom and plans to build four EPR-type reactors there. In addition, the EDF Group has a 49.9% shareholding in Constellation Energy Nuclear Group which operates five reactors in the United States and a �0% shareholding in two EPR-type reactors under construction in China.

Page 168: EDFMESS FR - inis.iaea.org

�2

10

Points in common

I have noted some issues that undeniably contribute to nuclear safety in a measurable manner.

Training is often well organised and associated with a method of accreditation by a commission including players from outside the company. For operations staff, training is integrated into the shift cycle, guaranteeing regular training and assessment. The line management is characterised by its high degree of involvement, striving to satisfy current plant needs.

The nuclear safety supervision committees frequently benefit from the presence of highly-qualified external members.

Care is being taken to try to stop staff following the procedures unthinkingly and losing sight of the ultimate goal of their action.

Communication and synergy between the companies of the EDF Group is improving as regards technical matters and nuclear safety, for instance in the fields of chemistry, engineering, alternator technology and by participation in peer reviews.

The control rooms are characterised by an atmosphere of calm and attentiveness, and I have observed the systematic use of human performance tools there.

The unit outages are the subject of planning well in advance and hence of excellent quality. The schedule is protected against the effects of contingencies by calling in dedicated teams.

Each formal meeting begins with a message about nuclear safety or industrial safety, enabling the management to clearly state its expectations, while encouraging the staff to reflect on nuclear safety.

The plant nuclear safety assessments rely on recognised outside organisations (INPO and WANO).

EDF Energy in the United Kingdom

Managerial practices

The leaders are forging nuclear

safety

For British plant operations, three Chief Nuclear Officers (CNOs) are each in charge of two or three plants. The plants thus benefit from close and effective support. I would like to draw attention to the need to ensure that the plant directors have sufficient independence.

I am aware of a strong will to increase both nuclear safety and cost-effectiveness, for instance by substantially reducing the number of outstanding work requests or by regularly investing to keep operating equipment up to standard. This is done on the basis of a risk management system that facilitates decision making and the judicious utilisation of resources. I have, however, observed that this system can fail (see Chapter �2).Surveys of safety culture perception are regularly carried out on the employees of EDF Energy and its main contractors. The results of the surveys (with very high response rates of 90%) are compared with those of similar surveys carried out by electricity companies in the United States. The results are used by the managers to determine justified, targeted action, at plant and corporate levels.I was also impressed by the efforts made to offer innovative leadership training at all levels. The training courses relate to ‘nuclear professionalism’, human performance tools, mentoring or the development of leadership by working with groups comprising employees from different professions and hierarchical levels, which are referred to as diagonal slices. Striking images with clear messages are used to show how the way in which work is actually done can deteriorate to a point where it fails to match up to what is expected by the management and how, at the same time, dangers, risks and equipment faults can accumulate without being detected until the safety margins are lost.I note that the independent organisations for checking nuclear safety in the plants do not yet appear to have sufficient status or authority. I observe with satisfaction that a new approach, after benchmarking

Page 169: EDFMESS FR - inis.iaea.org

��

10

with nuclear safety and quality teams from the French plants, has been experimented with and found promising in three British nuclear power plants.For nuclear safety checking at plant and corporate level, EDF Energy has opted to rely on peer reviews conducted by WANO every three years. At EDF Energy there is no internal inspection organisation that can make large-scale audits or assessments comparable to the evaluations conducted by the Nuclear Inspectorate in France.I note that, in 20��, there was a strategic re-orientation at fleet level which set the priorities for the nuclear safety committees. Apart from examining the major organisational and technical changes affecting nuclear safety, emphasis is more clearly placed on the operations-related issues raised by the plants. This development is appreciated by the plant directors and appears to me to be positive in terms of nuclear safety.

Dynamic relationships in the field

Knowing how to create

favourable situations

In the plants, I have noticed dynamic relationships with most of the big contractor companies. The nuclear safety objectives and training courses determining workplace behaviour patterns are specified and imposed jointly. Voluntary safety representatives are in place in all EDF Energy departments, and I appreciated their enthusiasm and commitment. The contracts are generally long-term with straightforward financial motivation for workplace safety and quality.

Major efforts are made to achieve the goals of the EDF Energy’s ‘Zero Harm’ Campaign. I note a strong desire to simplify or eliminate any pointless administrative procedures associated with the key processes. These actions stem from the analysis made after a fatal accident in Heysham � in 20�0.The drive for continuous improvement of quality has been re-invigorated after the discovery of instances of sub-standard maintenance. The action of the EDF Energy and contractor field supervisors is recognised as vital for achieving progress. These supervisors, who are numerous, receive special training and accreditation for their role. This is highly commendable, in view of the authority given them and the resources assigned.I note, however, that making technical decisions is sometimes too readily delegated to contractor companies as regards both maintenance operations and local engineering matters. This point merits careful attention as it has the potential to fragment the nuclear operator’s responsibility as owner for nuclear safety.

Noteworthy points concerning operation, and interesting initiatives

In all the plants, WANO peer reviews indicate scope for progress in the maintenance of fire equipment and the management of fire loads.As regards radiation protection, British staff benefit from the low levels of exposure inherent in the design of their advanced gas-cooled reactors. I also noted with satisfaction that the Sizewell B plant is maintaining itself in the top decile results worldwide for the doses received in pressurized-water reactors. I note that contamination control has been announced as a top priority for all the plants.I observed a highly-effective practice in one plant: the one-stop process that quickly corrects any discrepancies when operating procedures are applied. This practice answers the need to keep the procedures up to date, justifying strict compliance with them. This seems to call for benchmarking by the staff of the French plants.

Page 170: EDFMESS FR - inis.iaea.org

�4

10

Three significant events

I have chosen some of the technical events that occurred in 20�� to consider in detail. The first was classified Level 0 on the International Nuclear Event Scale and occurred in a coastal plant as a result of an invasion of jellyfish (800 tonnes), illustrating the need to be prepared for any kind of external event. In point of fact, the operators were at all times in control of the situation and the reactors where stopped in total safety. The prompt and resourceful use of trawlers also helped to cope with the event.

In another plant, an event classified Level � on the same scale affected the reactor auxiliary cooling systems after pipe failure, where sensitivity to corrosion had previously been identified. This event was highly significant in terms of controlling the decision-making processes at project management level. It is detailed in Chapter �2.A last event relating to actions central to the operations profession was classified Level � on the scale also caught my attention. After a short unit outage, the steam generator auxiliary feed valves were not re-opened as expected during re-starting and stayed that way for nearly nine hours. Subsequent analysis of this violation of the technical specifications and system alignment error revealed incomplete checking before starting, lack of rigour in the documentation and log-keeping, as well as unsatisfactory communication within and between the teams involved.Together, these three events throw into stark relief three vital issues for any nuclear operator. They have been the subject of in-depth analysis in the relevant plants and the experience feedback has been shared with the other plants and at corporate level within EDF Energy.

Technical challenges in advanced gas-cooled reactors

A unique and demanding technology

The advanced gas-cooled reactors are now facing complex technical problems that necessitate the mobilisation of highly-specialised engineering staff. Their task is also complicated by the fact that this reactor design is not found anywhere else in the world, offering little scope for joint assistance and the pooling of experience feedback with other operators. Some of the technical

problems now have to be managed without the assistance of the equipment makers and designers, who are no longer in business.As concerns the study and comprehension of phenomena with nuclear safety implications, I note that work is continuing on the graphite cores of the advanced gas-cooled reactors, the cracking of steam generator piping and the deposition of carbon on the fuel in certain reactors. These efforts show that the British nuclear industry needs to adopt a proper policy of strategic long-term research and development work. The first loading into the advanced gas-cooled reactors of the newly-designed "robust fuel”, the culmination of a long process of testing and technical study before licensing, is a positive point.Consequently, I consider that it will be necessary to maintain a high-level engineering capability until these reactors reach the end of their service lives, and possibly during their decommissioning. Maintaining the special skills will require an accredited training system. The recent accreditation of the first departments (design authority, engineering, safety and regulation departments), for the first time anywhere in the world for these disciplines, fulfils this need.

New nuclear plants: the EPR project

As part of the plans to build new nuclear generating facilities in the United Kingdom, a number of submissions were made to the Office for Nuclear Regulation and the authorities for environmental protection and regional development. News of the nearly unanimous vote in Parliament in July 20�� that confirmed that nuclear power would be included in the future energy mix was received with enthusiasm inside the company.

Page 171: EDFMESS FR - inis.iaea.org

��

10

The structure of the EPR Project appears to be judicious; a single body is charged with monitoring the studies, supervising the work and handling commissioning, and is preparing to become the future operating organisation.Numerous features of the Sizewell B reactor design, the only PWR in the United Kingdom, are considered as technical points of reference, such as the ultimate heat sink and the stand-alone operability time on loss of the electrical power supplies. These important points have been the subject of an in-depth study with a view to licensing the EPR type in the United Kingdom and to determining what post-Fukushima actions are required. Any differences between the reactors of the EPR type in France and the United Kingdom will need to be carefully explained in the public arena, and I recommend that the reference design of the British installation be settled as early as possible.

I regret that the proper functioning of the project is still hindered by problems relating to the computer systems and their security, making it difficult to access the drawings and documentation across borders with the architect-engineer in the United Kingdom and the design offices in France.

Constellation Energy Nuclear Group in the United States

Managerial practices

I have visited Constellation Energy Nuclear Group (CENG) for a number of years in succession. The arrangements for managing its reactor line-up of � pressurized-water reactors and 2 boiling-water reactors are now well-established. I note that the nuclear excellence dashboard performance indicators are being used effectively. The unplanned capacity loss factor is below 2% which confirms that CENG is in proper control of the reliability of its equipment and maintenance operations, which is vital for nuclear safety and a positive factor. On a more detailed level, I note that the equipment reliability index – created by the Institute for Nuclear Power Operations, the Electric Power Research Institute (EPRI) and the suppliers – used by the plants to compare their levels of performance is being used productively. Its use in other plants of the Group has begun and I encourage more widespread adoption.

As previously mentioned in my reports, the close and beneficial collaboration between CENG, the Institute for Nuclear Power Operations, EPRI and the Nuclear Energy Institute (NEI) is plain to see. These organisations clearly provide numerous American electricity companies, some of which are small, with vital support which is comparable to that from which EDF’s French plants benefit via the company’s internal support departments.

CENG operates two boiling-water reactors at Nine Mile Point on the shores of Lake Ontario. The basic design of Reactor � at Nine Mile Point is very similar to that of the first nuclear units at Fukushima (Mark � type). I therefore made a point of obtaining information on this plant and the main modifications and improvements made since �9�9. Detailed analysis of the lessons directly applicable to Nine Mile Point Plant is outside the scope of this report, but I note with satisfaction that CENG has been very proactive in the American nuclear industry actions post-Fukushima.More generally, for the use of operational experience feedback (particularly TMI, the 9/�� attack and recent advances in knowledge), I note that the continuous improvements of the nuclear safety arrangements (design, construction and training) are not achieved only by the rules laid down by the Nuclear Regulatory Commission (NRC). Many actions and modifications that are significant in terms of nuclear safety have been made on the initiative of the American operators, frequently submitted to the NRC via the Nuclear Energy Institute. This can be compared to the French and British processes of ten-yearly nuclear safety re-assessments but is not as methodical.

Page 172: EDFMESS FR - inis.iaea.org

��

10

My observations in the field

Ginna Nuclear Power Plant was commissioned in �9�8 and is the pressurized-water reactor that has been in service for longer than any other in the United States. Its operating license was recently extended to 2029, corresponding to sixty years. This is an example for all the pressurized-water reactors operated by EDF, as it prepares to extend the service lives of its reactors.I attended one of the quarterly meetings of the Nuclear Safety Review Board chaired by a qualified personality from outside the company. This meeting is always preceded, the day before, by a walkdown in the plant and a series of meetings with the plant personnel in each of the main fields of operation. Those I interviewed were all asked to express their feelings about their own levels of performance. The interviews helped reach an understanding as to the manner in which the personnel determine, describe and even assess their personal performance. The need to set goals suited to all levels in a consistent way was brought up in a pragmatic manner.

I also note that CENG’s three plants assist each other effectively during unit outages by staff transfers that can be very substantial (around half the maintenance department staff).

Knowing where we are going

On a number of occasions, I also observed the CENG central management and the plant managements requesting personnel to explain their actions and rank the performance levels of their plant. This way of doing things seem to be very appropriate.

The CENG nuclear safety chief has a solid diagnosis available and the priorities for improvement are clear: lock-outs, circuit alignment and safety in the workplace. As I see it, this means that nuclear safety is constantly being critically reviewed and upgraded.The managers are motivated and demanding in terms of requirements and nuclear safety. One of the managers recently mentioned in a meeting that, in view of certain shortcomings, there was a need to move on from dissatisfaction to intolerance and that things accordingly needed to change. I think that this type of attitude needs to be more general within the Group.I also note the appointment, at CENG operations meetings, of a Nuclear Safety Advocate who has the role of promoting nuclear safety during the meeting. This practice, which fosters feelings of responsibility, seems to be worthy of interest and offers parallels with the role of the safety engineer in the French plants.

Decision makingIn one of CENG’s plants, the operations team encountered difficulties in determining the cause of irregular test performance by an item of equipment important for nuclear safety (a standby turbine-driven pump). The technical support provided for the operations teams by the local engineering force focused on a single technical cause assumption which proved to be wrong. The dialogue between the engineering and operations staff was deficient to the point where the nuclear safety authority questioned the ability of the plant to deal with this type of problem.The CENG management then agreed to an in-depth analysis of the organisational structure and decision-making chain in the event of a contingency. This analysis led to setting up a new process: integrated problem identification and resolution process, specifying the responsibilities of the different people and organisations linked to operations. This initiative appears to have been effective and well appreciated. It is now used in all CENG plants.This event clearly illustrates the difficulty in effectively connecting operations-related decision making – normally handled by the operators – to the numerous other processes. These include adverse condition monitoring, operability criteria, the corrective actions programme, failure mechanism identification, risk management, human performance, equipment reliability (AP9��) and post-incident analysis. The method, just as much as the solution adopted, needs to be given careful consideration by nuclear operators as the issues cut clear across all the organisational structures. At some time or another, every organisation needs to be re-appraised by those who put it into place. This can be a complex process, but it is always a sign of a good nuclear safety culture.

Page 173: EDFMESS FR - inis.iaea.org

��

10

Noteworthy events

An earthquake occurred in the east of the United States during the summer of 20��. The effects were perceptible in one of CENG’s nuclear plants, but the intensity was well below the design basis and had no effect on reactor operation or any implications for nuclear safety. Classified Level 0 on the International Nuclear Event Scale, it nevertheless revealed that the operators had difficulty in determining, within the time limits, the intensity of the earthquake from the control room as required in the procedures for emergency plan initiation. It is to be noted that special knowledge, rarely used by operators, is needed to correctly interpret the readings available. This finding needs to be made known to all the EDF Group plants to ensure proper application of this type of procedure which is rarely used but extremely important.Another event of natural origin, Hurricane Irene, affected a CENG plant in September 20��. A sheet of metal that came loose from the turbine hall cladding gave rise to an electrical fault in a transformer, causing one of the two reactors to scram. This event, which was classified Level 0 on the International Nuclear Event Scale, was correctly managed by the staff. Investigations showed that the cladding had not been properly installed and that the design drawings were not sufficiently explicit. Previous wind-damage feedback about similar cladding was not taken into account.

Taishan Nuclear Power Company Joint Venture Company in ChinaThis year, I visited the construction site for two reactors of the EPR type at Taishan. I saw how the work had progressed, with the schedule for Reactor � now very close to that for the Flamanville � EPR. The metal dome on the reactor building, which was installed during my visit, was a symbolic moment for this immense site where more than �0,000 people are now working.I also met the teams managing the project, who explained the development of the joint AREVA and China Guangdong Nuclear Power Company’s design office (the client’s engineering

and boilermaker forces). This new office has been found very effective and has been able to produce erection drawings on time, a situation that I noticed during my previous visit.The Chinese staff of the Taishan Nuclear Power Company Joint Venture Company and China Guangdong Nuclear Power Company rely on the extensive experience accumulated by their testing and operations teams through the commissioning of CPR 1000 reactors.As is the case for Flamanville �, I also noted that it has been decided to qualify the equipment for the consequences of severe accidents before start-up by making maximum use of the EDF SA approach.At the headquarters of China Guangdong Nuclear Power Company, I was shown the plan to create an independent safety organization at plant and corporate level, corresponding to EDF SA practice. This will increase nuclear safety as the number of reactors in service is rapidly increasing. I am glad to see the active cooperation and exchanges with the Nuclear Operations Division’s Nuclear Inspectorate, which extends to the training of Chinese inspectors.More generally, Chinese companies are quickly becoming major players in the nuclear field throughout the world, with �4 reactors in service and 28 under construction at the end of 20��. Holding the two-yearly general assembly of WANO in Shenzhen in October 20�� was a symbol of their new status. I consider that it is essential for the EDF Group to continue to do its utmost to share its construction, commissioning and operating experience with the Chinese companies so at to promote a high level of nuclear safety. I note the opening of an EDF office in Shenzhen in 20��, closer to the industrial concerns and operators, which will facilitate experience feedback and the exchange of information.

In conclusion, I am glad to see increasingly active cooperation between the different nuclear bodies of the Group. This diversity is an asset that can be a driving force for substantial progress in the field of nuclear safety. I encourage the development of such exchanges on the existing nuclear facilities and the projects for new reactors. All have much to gain from opening up to the practices of others, regardless of their background.

Page 174: EDFMESS FR - inis.iaea.org
Page 175: EDFMESS FR - inis.iaea.org

�9

11

VISITS OUTSIDE FRANCE

Sharing experiences and methods with nuclear operators in other countries enables us to question our own practices and put our entire field of activity in proper perspective. Exchanges with INPO, the leading player worldwide in performance assessment, are also indispensable for challenging the Group's methods and understanding the checking carried out in CENG’s plants.

The INPO headquarters in Atlanta, Georgia

I had the privilege of spending quality time with a number of the top managers, which proved extremely useful as INPO plays a major role in the nuclear power industry in the USA. Its authority results from the full and unreserved support of the presidents of its member companies. It is adequately funded and most of the staff (80%) are full-time employees. It plays an essential role in its relations with the many American nuclear operators who do not have any substantial internal corporate-level functions. With more than �00 reactors in the United States, as well as others belonging to members in other countries, INPO makes assessments of and provides support for a wide and diverse number of plants and organisations. It appears that its value essentially resides its considerable professional knowledge, its commitment and its impartiality since it belongs to none of the companies it assesses. INPO has an external viewpoint and is not exposed to management constraints or internal pressure by the operators evaluated. It is thus easier for it to call into question the pertinence and the credibility of company policy.

Page 176: EDFMESS FR - inis.iaea.org

�0

11

Its main functions consist of making assessments (�4 per year), accrediting training programmes, managing experience feedback and offering support for American power plants.

The assessments

The assessments focus on companies and plants. After the assessment of a plant, an overall performance ranking is confidentially communicated to it. This ranking is discussed between the Chief Executive Officers of INPO and the company involved.

INPO organises an annual conference for managers during which warning signs are examined in a closed session. Last year, the pressing issues were the increase in the number of significant operating events (Publication SOER �0-2) and of automatic and manual reactor trips, underperformance by the operations teams and failure of backup diesel generators. The best plants are identified and the managers of the low-performance ones are asked to explain to their peers what measures are being taken to reduce the risk that their company represents for the nuclear industry as a whole.

At INPO, I was shown an excellent summary of the types of plant reactions in cases of deteriorating performance and hence deteriorating INPO rankings:

There are plants that immediately seek to re-achieve excellence, exhibiting a will to turn things around from the first signs of deterioration by taking aggressive action and accepting critical feedback from the hierarchy whatever its source. They make it plain that they are not satisfied with the situation, without seeking excuses and show their commitment at all levels.There are plants in which the situation continues to deteriorate, where the managers are slow to recognise and accept that there is a problem, tending to explain away rather than rectify the issues and hesitate to make decisive changes. They react defensively, quoting their good performance levels in the past rather than the best levels currently achieved by others.

I think that there is much to be learned from this.

Furthermore, INPO offers a system of progressive support with targeted missions after each assessment for each avenue of progress identified. For a plant in difficulty, it creates an ad hoc oversight commission that meets its chief executive officer every six months.I particularly remember an interesting observation in some American feedback: "the worst plants generally have staffing pipeline problems in operations, with the result that the other departments lack process skills. The have also often had budget cuts.”I also discussed the post-Fukushima action with members of the assessment team that I met (see Chapter 4). Like WANO, INPO must also include beyond-design-basis situations in its performance assessment criteria and objectives. At present, the Severe Accident Management Guidelines do not cover fuel interim storage pools and therefore need to be supplemented.

Performance indicators

Most of the indicators show that performance levels in the industry in America have reached maxima or are stagnating and “the strong propensity for significant events detected in early 2010 is continuing”.The operability factors for standby diesel generators have never been as low in the United States. On the basis of their probabilistic studies, I see that the American operators are increasing the amount of elective maintenance work on diesel generators during plant operation. In most other countries, this is not authorised or is tightly controlled by the safety authorities. I consider that this question needs to be the subject of international debate, particularly in light of the Fukushima accident.Fuel reliability is an issue on which INPO is working in close liaison with the suppliers, through the Supplier Participant Advisory Committee. In 200�-200�, only ��% of the plants could boast no cladding leaks over a period of three years. The proportion has now reached �0% (��.�% for pressurized-water

Page 177: EDFMESS FR - inis.iaea.org

��

11

reactors). In 20�0, 92% of the plants had no leaks. This is a good example of collaboration between the operators and suppliers with a view to increasing nuclear safety.

Avenues of progress identified by INPO

INPO’s intention is to discuss with the sites potential areas for improvement before the facts have been fully researched and established so as to forestall any incidents as far as possible. They also intend to increase the observation of the operations teams.INPO also intends to concentrate on the operator’s fundamental skills and knowledge. In the recent Significant Operating Experience Report ��-�, it seeks to foster a situation where substance prevails over form, focusing on what is done in the actual process.INPO issues useful recommendations on topics such as operational decision-making, nuclear safety culture and dynamic leadership.Together with the Massachusetts Institute of Technology, it runs a training syllabus for chief executive officers and top executives who have recently started working in the nuclear sector.I was also shown a particularly interesting feature of another syllabus: the e-learning tool NANTel. This system, which is also available to service providers, covers issues such as human performance tools and foreign material exclusion. INPO is beginning to work directly with contactors and not only with operators. More than 9�,000 connections to the NANTel site have been logged.Action is also in progress to encourage service-providers to participate more in building up experience feedback records, including during the construction phase.

INPO makes videos of the main events in the nuclear industry. This is an excellent initiative, particularly for the new generation of staff. I recommend their use throughout the EDF Group.

The nuclear industry in South Africa

In South Africa, the electricity company ESKOM is in charge of generating electricity throughout the country, mainly from coal. It also operates the only nuclear power plant in Africa. From the outset, it has cooperated in peer reviews and benefited from the assistance of WANO, INPO and the IAEA. It has also signed a technical support contract with EDF SA as the reference plants for its two 900 MWe pressurized-water reactors are French. They have been in service since �984.During my visit, I observed that it is actively engaged in benchmarking and encourages international scrutiny. The EDF engineers at the plants are well integrated with the local staff and the ESKOM hierarchy. EDF practices constitute a reference point for defining modifications and for operations. In recent years, a number of major modifications have brought the South African facilities up to conformity with the IAEA standards (Publication INSAG �2 relating to basic safety principles for nuclear power plants) for new nuclear plants. ESKOM has used probabilistic studies of nuclear safety in operations as part of the process for determining priorities and making decisions.

The operations department is well staffed and has solid experience. The training programme is directly accredited by INPO. The line staff act as if they own the training. The simulator updates and the different procedures associated with them are carefully coordinated with the changes incorporated in the plant’s reactors. The operations personnel carry out simulator training for one week out of six, with the shift rota arranged accordingly.

I was impressed by the staffing and the resources of the training department, with nearly �0 instructors working in modern facilities, open to the outside (especially schools) and fully integrating the contractor companies.Many additional staff have been recruited and are under training or occupy supplementary positions set up as part of plans to build a new plant.Substantial investments have been made to keep the plant up to standard. Replacement of the steam generators and increasing the thermal rating of the steam supply system are being planned. It is worth pointing out the good condition of the Koeberg steam generators, which are the last of

Page 178: EDFMESS FR - inis.iaea.org

�2

11

their generation still in service. This attests to the scrupulous monitoring and control of the chemistry parameters throughout their 2� years of operation.I noted that one of the reactors has excellent reliability results for recent years, with a coefficient of unplanned capacity loss factor of less than 2%, while the other reactor was affected by a turbine design problem, recently rectified, which significantly reduced the reliability results. Furthermore, airborne contamination in the reactor building during a unit outage resulted in the slight internal contamination of 24� people. The similarity with an event that occurred earlier in Tricastin Nuclear Power Plant in France which was not given consideration in Koeberg shows the importance of experience feedback.I observed that one of the challenges with which ESKOM is faced is being able to plan and schedule the unit outages sufficiently far in advance to be able to have the necessary job packages and spare parts available in due time.I also observed that the managers of the ESKOM nuclear activities invited WANO to organise peer reviews in their plant and also give an outside view of the different types of organisational structures envisaged to supervise the nuclear branch of ESKOM which mostly operates fossil-fired plants. Rightly, this organisational aspect is considered vital to guaranteeing nuclear safety in this type of situation.I also noticed that those I spoke to were very concerned about the checking and surveillance of vendors and service providers, as well as the methods to be adopted for fostering suitable behaviour and guaranteeing regular quality of work.

As regards nuclear safety management, I observed that ESKOM was seeking to perfect its independent nuclear safety organisation on the basis of French principles and practices. A French nuclear safety engineer has been seconded to Koeberg for this purpose. The nuclear safety committees seem to be well organised, with performance benchmarking and practices being regularly compared with those in France and the United States.

Finally, I met with some South African nuclear safety authority staff, both in the plant and in Pretoria. Faced with the difficulty common to all authorities with few nuclear facilities to check, I note the bilateral relations with other nuclear safety authorities, including the French ASN and the US NRC.

Page 179: EDFMESS FR - inis.iaea.org

��

12

EVENTS IN THE EDF GROUP

The events covered in this chapter relate to the main operations professions: maintenance, operating, engineering and projects. They show the need to continuously consolidate the human and organisational lines of defence.

12.1 EDF SA: premature wear of diesel generator bearings

Each nuclear reactor is equipped with two redundant diesel-driven generating sets capable of providing backup power for stand-alone operation of the main safety-related equipment in the event of station blackout (high-voltage lines). Furthermore, in the event the two backup diesel generators fail, each plant has a last-resort standby generating set that can rapidly be hooked up to a nuclear unit in time of need (diesel-driven in most of the 900 MWe series plants and gas-turbine-driven in the others).After the original maker ceased production of the connecting rod big-end bearing shells for the standby diesel generating sets of the 900 MWe series units, another company – the current leader the field – began producing them in 2002.The diesel engine company responsible for engine supply and maintenance then asked this supplier to manufacture bearing shells identical to the previous ones. It should be pointed out that the diesel engine company is also not the same company that originally designed and manufactured the equipment.After faults in 4 of the generating sets between 2008 and 2009 and in light of international experience feedback, the diesel engine company began to doubt the durability of the components and proposed a design change.Without waiting for the end of the in-shop qualification tests (�8 months from September 2009 to March 20�0), the new generation of bearing shells (the only ones available) were fitted at the end of 2009 to �� sets previously equipped with the former bearing shells. The ASN was kept informed. After the tests, as they were found to have very long service lives, the diesel engine company and EDF announced that the new-type bearing shells were qualified.On 22 October 20�0, three of the new generation of bearing shells failed during requalification tests after maintenance on a nuclear plant’s last-resort generator diesel engine. In-depth expert appraisal revealed abnormal wear of 8 other bearing shells of the same type.During a surveillance test on 28 November 20�0, serious internal damage was detected in the diesel engine of the standby generator at another nuclear plant.A technical team at the Nuclear Operations Division was assembled in early 20�� and examinations were begun at other plants possibly affected. These examinations confirmed the presence of premature wear in other bearing shells.The corporate engineering force's conclusion was the existence of a potentially generic discrepancy in all the 900 MWe facilities with engines equipped with these bearing shells. The ASN was immediately informed of the expert appraisal results and a generic significant event in terms of nuclear safety was declared as Level � on the International Nuclear Event Scale.However, in one of the 8 plants affected, two of the nuclear units had both standby diesel generators implicated, as well as the plant's last-resort diesel generator. This resulted in the declaration of a Level 2 significant event on the International Nuclear Event Scale.

Precise characterisation of the physical defect causing the failures led to the corrective action rapidly being determined, involving improved honing of the bearing shells in the factory, increasing the pressure in the lubrication system and changing the oil type. Meanwhile, the defective parts were replaced over a three-month period in all the engines affected (around thirty).

Page 180: EDFMESS FR - inis.iaea.org

�4

12

In the short term, arrangements were implemented to intensify the monitoring of the behaviour of this equipment in service (more frequent oil analysis) as was an additional maintenance programme. A test engine equipped with the new better-honed bearing shells was subjected to ad hoc endurance tests.

Finally, finding a permanent solution was the subject of intense, regular technical discussions with the ASN.

In conclusion, the first thing I noticed about this serious event, which did not directly impact the nuclear safety of the reactors affected, is the promptness with which action was taken as soon as the fault had been properly identified. When faced with a potentially generic fault affecting a large number of plants simultaneously, the corporate-level forces (the Operations Engineering Unit and the Central Technical Support Department) appear to have been effective, even though the technical diagnosis was hard to establish. The proper handling of such cases with the ASN and its technical advisers from the institute for nuclear safety and radiation protection (IRSN) requires response at a level that the plants cannot provide. Maintaining, and even developing, the key skills in these organisations is thus of the highest importance.When I talked to the specialists in charge of dealing with this contingency, I also noticed the difficulty in getting a diesel engine company to respond quickly enough and make the skills available. These skills had been dissipating over time and as a result of restructuring in the sector.In the case of equipment so important for nuclear safety, this seems to show the need for greater care and vigilance, as such trends can eventually degrade even the best designs.

12.2 EDF SA: a control incident and an automatic reactor trip

An event occurred in 20�� in one of the 900 MWe units that caught my attention, even though there were no direct consequences for the nuclear safety of the installation. It was ranked Level � on the International Nuclear Event Scale. It revealed inadequate planning for a normal operating transient (a power drop), the absence of a questioning attitude when faced with something unexpected, and poor hands-on reactor control.

The facts

The power level of the nuclear unit was due to be dropped to ��% rated power to repair one of the turbine-driven feedwater pumps on the secondary side of the installation. This operation was scheduled to take place during the night, and repair work was to take approximately �4 hours.A temporary control order written by the operations department support team to prepare for the power drop (transient) was available to the control room operators. The order was not the subject of joint planning with the core and fuel engineering forces at either plant or corporate level. It specified maintaining the control rod clusters inserted by a certain amount (�80 steps). During operation at ��% rated power, the operators scrupulously observed this operating order, following flux tilt parameter Delta I.

Once the turbine-driven pump was repaired, the operating staff commenced power escalation according to the normal procedure. At �8% rated power, the rod lifting operation was interrupted due to a fault signalled by the instrumentation and control equipment. The automation department duty service was called to deal with this.

At the handover between the morning and the afternoon shifts, the operator in charge of the primary side of the installation considered that the Delta I reading was too high and requested explanations. He was told that he had to respect the temporary control order. During the handover to the afternoon shift, the computer department duty service was called in to repair the control assistance system that was inoperable due to a random computer fault. This system supplied the operator with the level and trend of the xenon-effect, an important control parameter. This display-orientated tool is not strictly required for resuming power escalation, as the operator has other sources of information. The inoperability of

Page 181: EDFMESS FR - inis.iaea.org

��

12

the control assistance system was mentioned during the pre-shift briefing by the associate operations shift manager. This was not considered to be a reason to stop by either the associate operations shift manager or the main operations shift manager, who authorised resumption of the power escalation with reduced monitoring capability, as the operators were experienced, having between � and 20 years of service. This instance of inoperability was not the subject of sufficient risk analysis between the operators, the associate operations shift manager and main operations shift manager.

Power escalation was therefore begun under these conditions at ��:�0 hours. The reactor operator then calculated the volume of water to be injected into the reactor coolant system to bring the power from �8% to �00% rated power using the normal procedures. He proceeded to make successive and progressive injections of water without having data on the xenon-effect in the reactor (which was then decreasing). He concentrated on monitoring the key parameters, i.e. the power, average reactor coolant system temperature and Delta I.At �4:�0 hours, the power was brought to 99% rated, the operators leaving a �% margin to allow for the inoperability of the control assistance system. Continuing to apply the chosen approach, the reactor operator continued to inject water to control the average reactor temperature and return the control rods to a less-extracted position.The rapid xenon decay continued to add reactivity which, combined with the effect of the water injection, resulted in the automatic and deep insertion of the control rods. As the action of the rods is highly effective, Delta I then rapidly decreased. This set off the alarm signalling an excursion outside the authorised envelope, resulting in the operations shift manager being called to the control room.At �4:�� hours, the operators applied an operating procedure covering this type of situation and dropped the reactor power by �.�% nominal, while injecting boron into the reactor coolant system and extracting the control rods.At �4:�� hours, the automatic DeltaT-SP protection functionality provided for the purpose reduced the power then shut down the reactor by dropping the control rods, resulting in virtually instantaneous shutdown. The operating instructions planned in this type of situation were then applied without difficulty. After analysis of the causes behind this automatic reactor trip and repairing the control assistance system, the reactor was put back into service the next day at 0�:�8 hours (criticality) and reached full power the following night.

Insufficient preparation for an unusual transient

This reactor is continuously run at �00% rated power. Extended power drop transients of this type are very rare in this type of nuclear unit, which should have meant more careful preparations, combining all the skills involved, including the plant's expert core/fuel engineer. Replicating the transient on a simulator, which is an easy matter with the resources now available to engineers, would have made it possible to devise an appropriate control strategy.

An insufficiently questioning attitude when faced with the unexpected

The perfectly justified questions of the operator beginning his shift about the unusual Delta I reading were disregarded and the team continued applying the temporary instructions. Analysis of the risk represented by the inoperability of the control assistance system was not taken far enough with, for instance, the main and associate operations shift managers failing to grasp all the implications of the situation. A proper pre-job briefing would probably have made it possible to identify all the issues.

Deficient control action

The water injected into the reactor coolant system to bring the power level from �8% to �00% nominal increased the core reactivity – which was indeed its purpose – at a time when the xenon-effect decay was also increasing reactivity. This inappropriate action illustrated that the operating staff and the main and associate operations shift managers had not fully understood the way the xenon-effect was developing. Use of the standby computer application Xenon 0D, which can be accessed on the control

Page 182: EDFMESS FR - inis.iaea.org

��

12

room intranet, would have partially compensated for the inoperability of this system and provided the essentials needed for comprehension.

I note that at all three moments in time, there was insufficient understanding of the physical phenomena involved.

In conclusion, this event clearly shows the need to increase operating staff knowledge of the complex physical phenomena involved in reactor control by means of more suitable training. The presence in each plant of expert core/fuel engineers now offers a way of increasing skill levels in this area and making better use of the simulation facilities.

Practices to increase reliability (here pre-job briefing), an interrogative attitude and, when necessary, the ability to say "stop", are all expressions of nuclear safety culture and represent strong lines of defence against the unexpected.

12.3 EDF Energy failure of cast-iron pipes

The facts

The two reactors of an advanced gas-cooled reactor plant were in commercial operation when pressure loss was detected in the seawater systems used for cooling the auxiliaries of one of them. A visual examination made it possible to rapidly find a large leak in a buried pipe. In accordance with the procedures, the operators reduced power and switched over to the system of the other reactor to provide cooling. The reactor involved was then shut down for repair. The operating staff had no great difficulty in ensuring compliance with the technical specifications. The leak was clearly caused by internal corrosion which had eaten through the wall of a cast-iron pipe.

The context

Cast-iron pipes were commonly used for the seawater systems of the British advance gas-cooled reactors. The way seawater causes corrosion is well known and documented. The owners of the EDF Energy facilities had not carried out replacement and refurbishment of these systems early enough. A number of failures in service have already occurred in advanced gas-cooled reactors, notably in 2004 and 200�. The recent incident was thus a recurrent event. Since 200�, a vast inspection and replacement programme has been pursued in all �4 advanced gas-cooled reactors. The pipe sections buried in the ground were the last to be replaced.

The causes

Analyses carried out after the event revealed the following salient points:

The numerous groups charged with supervising the replacement of the defective pipes tolerated slowness of the work,The pipe replacement project did not have the planned results: there were shortcomings in the design specifications and the work organisation. There was delay in project implementation and frequent changes in project manager (seven in five years, including four in the last �2 months),The risk analysis and management process failed to neutralise the risks in due time.In the present case, the pipe inspection programme gave no pre-alert of the type of failure that occurred and thus precluded an action plan being made.

••

Page 183: EDFMESS FR - inis.iaea.org

��

12

My comments

After incidents in the plants, the control room staff and field workers are often under the spotlight. In the present case, however, I think we need to go back and consider how the management and the engineering support perceive the risks, establish the priorities and, above all, take measures to minimise the risks. Here, a known phenomenon of aging was affecting cooling systems important for reactor safety. Both internal and international operating experience feedback was well known. The pipe replacement policy had been laid down and the necessary budgeting provided. The cast-iron pipe replacement project began in the plant in 200�. There were at least six committees (e.g. project investment, liaison and others) charged with evaluating the progress with the project and setting the priorities. All these bodies were worried by the slowness of the replacement work but none had the clout to speed up the work. The pipe replacement project was but one of a series of major projects with calendar slippage. I observe that another project relating to the refurbishment of the emergency core cooling systems of the reactors and hence important for nuclear safety is also experiencing delays and problems.This event shows that, in the field of nuclear safety, every link in the chain must be strong if defence in depth is to be assured. It is necessary but not sufficient to identify the risks and rank them. Risk mitigation measures need to be determined and budgeted. Projects must then be managed in a professional manner until completion (quality, deadlines and costs).The driving force that management provides to deal with situations, the resourcefulness and responsiveness of the engineering forces, and the professional management of projects, are all vital aspects of nuclear safety.

12.4 SOCODEI: an industrial accident at Centraco (Marcoule)

On �2 September 20�� at around �2:�� hours, an accident occurred in the Centraco site, on the working deck of an induction furnace for melting down metal waste. The Centraco furnace receives short-lived low and very-low level waste produced during the routine operation of the nuclear facilities, maintenance work, process changes and nuclear facility decommissioning.The plant is a licensed nuclear facility that is regularly inspected by the nuclear safety authority.A furnace worker who was near the furnace was killed. Four other people were injured, one seriously.The fire caused by the explosion in the area was brought under control by ��:0� hours by the plant response staff.The Centraco facility is wholly-owned subsidiary of EDF and, as part of the IGSN's mission, I have some remarks to make about the way the accident developed:

It did not cause any radioactive or chemical discharges,The Centraco facility is located in the district of Codolet, alongside the centre belonging to the French Atomic Energy Commission (CEA) in Marcoule. It maintains close operations-related and technical relations with this centre, and the emergency was managed from there. The EDF corporate emergency management system was mobilised as soon as the accident became known,A calculation error resulted in the radioactivity inside the furnace being underestimated. The causes of the error were subsequently determined.

As analysis of the accident is not yet fully complete, I will return to this event is my next report.

Although the radiological potential of the event was limited, it was a serious industrial accident as there was a fatality. Application of the International Nuclear Event Scale criteria resulted in classifying the event as Level � due to the small amount of radioactivity contained in the furnace.

••

Page 184: EDFMESS FR - inis.iaea.org
Page 185: EDFMESS FR - inis.iaea.org

�9

13

A TECHNICAL EVENT OUTSIDE THE EDF GROUP

The Deepwater Horizon drilling platform accident

The context

On �9 March 2008, Mineral Management Service (MMS) in the United States granted British Petroleum (BP) an exploration permit in Mississippi Canyon Block 2�2 in the central part of the Gulf of Mexico. The permit was to last �0 years, with BP holding ��%, Andarko Petroleum holding 2�% and MOEX Offshore holding �0%.On 22 May 2009, MMS granted a drilling permit for the Macondo well located �,�00 metres underwater.After early drilling attempts proved difficult with the first method and further complicated by Hurricane Ida, BP called in the Deepwater Horizon platform which is owned and operated by Transocean, under contract. Within the scope of this contract, the platform made some thirty boreholes, two thirds of which were exploratory. The Deepwater Horizon arrived on location on �� January 20�0 and drilling commenced on � February 20�0.The accident occurred on 20 April 20�0 when a deep production well was being completed. There were eleven fatalities.Technical problems, non-conservative decision-making and insufficient responsiveness led to excessive delay in detecting leakage and deciding to close the valves on the production casing. The gas escaping up the well caught fire at platform level and the platform sunk.The tragic loss of human life was accompanied by an ecological disaster caused by the discharge of oil over a period of three months to produce the greatest ever oil spill near the coast of the United States.

With the press calling the accident the ‘American Chernobyl’ and the ‘Ecological 9/��’, President Obama announced a moratorium on offshore drilling on 2� May 20�0.A report on the event was drawn up by a commission of government-appointed experts, in which fault tree analysis identified eight successive factors that led to loss of the platform. These lines of defence can be classified under four groups of critical factors:

The well containment was faulty, but this was not actually detected (problem relating to the cement and mechanical barriers),Well monitoring failed to detect the leak and control over the well was lost (interpretation of pressure tests, control room surveillance and response time),The oil spill on platform level was not contained (nor was the gas, which caught fire),The blow-out preventer emergency system failed to close off the well (the well isolation system did not work and the fire was fed with gas).

Although the resulting oil spill was not the worst the world has seen, it proved to have the greatest impact on the oil industry in the end by forcing the players to review the design, utilisation, monitoring and safety of their procedures and installations.

The lessons

I have no comments to make about the technical aspects of the case, which have nothing in common with the nuclear industry, but wish to examine the factors that resulted in the platform loss in the light of nuclear safety culture. Here, this accident and the Fukushima accident together constitute the two most recent industrial accidents with effects on the environment associated with the production

••

Page 186: EDFMESS FR - inis.iaea.org

80

13

of energy. It is relevant to the nuclear industry because the American government asked INPO at an early stage to make available its experience in the national nuclear industry. Furthermore, the world of energy is an interdependent one, where every local event has a global resonance.

I contacted the French oil consortium TOTAL and discovered that the factors of experience feedback derived from the accident are much the same as those identified by the nuclear industry: work with contractor companies, reinforcement of an independent safety organisation with the drilling supervisor, increasing the staff skills (particularly for dealing with exceptional situations on simulators), development of special means to cope with improbable events, inspections of current deep-drilling operations and the associated safety equipment, strengthening the ability to combat widespread pollution, checking the effectiveness of the emergency plans in different situations, etc.

Lessons for nuclear safety culture

In the EDF Group, I observed considerable diversity in the reactions to this accident. First, I would like to commend the exemplary approach adopted by EDF Energy soon after the accident. This was of the bottom-up type, with each nuclear generating site forming working groups to discuss, analyse the facts and propose ways of further reinforcing nuclear safety in the nuclear plants. The inputs from the plants were then amalgamated in a single report that was submitted to the management of EDF Energy. This process involved all the hierarchical strata at plant and corporate levels, embodying the intent to make progress in safety culture at all levels. EDF Energy finally identified three topics for further examination to challenge its own organisational structure with regard to nuclear safety:

management of the service providers (business strategy, awarding contracts, purchasing, supervision, surveillance, checking, etc.).nuclear safety checking loops, both internal and with the authorities,management of the nuclear safety barriers and making provision for their contingent deterioration.

The staff involved in the plants were extremely interested in the process. It is the first time such an operation has taken place for an industrial event that occurred outside the nuclear industry. This job also broke down many of the barriers between the different professions since it will lead to a joint submission by all the nuclear plants. Finally, EDF Energy has called on the nuclear safety authority (the ONR) to work jointly on pooling the analyses.

EDF SA, for its part, organised consultation at corporate level using the strong analytical capabilities of the industrial hazard management department at the Research and Development Division, the risk checking directorate of the Group, and representatives of the Production and Engineering Directorate. A meeting to share analysis was organised with EDF Energy. The Fukushima accident occurred in March 20��, with the result that the focus immediately switched to this major event with multiple consequences for the EDF Group. The result was that the top-down approach chosen by EDF SA did not filter through. It was a missed opportunity for creating an impetus to reinforce nuclear safety culture at all levels of responsibility down to the field, within a reasonable timescale.

••

Page 187: EDFMESS FR - inis.iaea.org

8�

13

In conclusion, the BP accident has the following features in common with other major industrial accidents:

it was a call to the international players in the sector to make in-depth reviews of the safety of their operations,it created more solidarity between the industry players; safety has become a pressing issue and due regard needs to paid by all the players as the failure of any one of them can have grave repercussions for all the others.

In the nuclear field, the principle of appraising the technical and organisational systems by an outside third party – preferably a peer – has enabled objective assessment by competent players, while limiting the risk of conflict of interest as peers have no automatic tendency to be conciliating. The practice has long been universal.Finally, international harmonisation of safety standards is a path on which every industrial sector exposed to risks needs to resolutely set foot so that a level playing field is created. The aeronautical industry has set an excellent example of this. This is an area where progress is still needed in the nuclear industry throughout the world.

Page 188: EDFMESS FR - inis.iaea.org
Page 189: EDFMESS FR - inis.iaea.org

8�

14

APPENDICES

Page 190: EDFMESS FR - inis.iaea.org

84

14

14.1 - Result indicators for EDF SA’s plants

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Number of events classified on INES as 1 or greater, per reactor * 1.00 1.20 0.88 0.76 1.22 0.80 1.15 1.17 1.17 0.91

Number of significant nuclear safety events, per reactor * 7.70 8.14 7.62 9.54 10.21 10.80 10.34 10.93 10.45 10.57

Number of cases of non-compliance with the technical specifications for operation, per reactor

1.30 1.57 1.16 1.48 1.55 1.70 1.70 1.39 1.55 1.36

Number of alignment errors, per reactor ** 0.74 0.93 0.50 0.66 0.69 0.57 0.62 0.53 0.77 0.71

Number of scrams, per reactor and per 7,000 hours of criticality (the average value for all reactors) • Automatic 1.19 1.13 1.01 0.93 0.89 0.87 0.51 0.71 0.69 0.50

• Manual 0 0 0 0 0 0 0 0 0.01 0.05

Average collective dose in operations, per nuclear unit in service (in man-sieverts

0.97 0.89 0.79 0.78 0.69 0.63 0.66 0.69 0.62 0.71

Exposure of individuals:

• Number of individuals with doses above 20 mSv 0 1 0 0 0 0 0 2 0 0

• Number of individuals with doses between 16 and 20 mSv 154 74 73 28 17 20 14 10 3 2

Number of significant radiation protection events - 160 177 173 112 99 107 102 91 92

Availability (%) 82.2 82.7 82.8 83.4 83.6 80.2 79.2 78.0 78.5 80.7

Unplanned capacity loss factor (%) 2.5 3.0 3.5 3.2 3.3 3.7 4.4 4.6 5.2 2.2

Occupational accident frequency factor (Tf) *** 8.5 7.5 5.5 5.5 5.6 4.6 4.4 4.3 4.1 3.9

* Excluding generic events, i.e. events due to design defects

** Any configuration of a system or its services that deviates from the expected situation and is the cause of a significant event

*** Tf: frequency of occupational accidents with sick leave for 1 million hours of work

Page 191: EDFMESS FR - inis.iaea.org

8�

14

14.2 - Result indicators for EDF Energy’s plants

Consideration needs to be given to the following factors when comparing the results of EDF SA with the British ones:

Lines 3, 4 and 5: the event declaration procedures are not the same in the United Kingdom and France as a result of the respective nuclear safety authority requirementsLine 7: : the reactors of the two sets of plants do not share the same technology (mostly AGRs in the UK and PWRs in France). The AGR design means that radiation exposure is some ten times lower (WANO).Line 9: To facilitate benchmarking, the values of previous years have been modified to allow for the fact that the declaration criteria of EDF Energy and EDF SA are different.

N° Indicator 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

1Number of events ranked 1 or more on INES, per reactor 5.67 5.27 5.60 5.67 3.13 1.20 1.13 0.80 0.93 1.33

2Number of events ranked 0 or more on INES, per reactor 10.53 10.80 9.60 9.13 7.53 4.93 4.53 5.47 5.60 4.7

3

Number of events of highest level declared to the Nuclear Installations Inspectorate, per reactor

- 5.47 3.60 2.67 1.53 0.40 0.67 0.33 0.67 0.46

4

Number or cases of non-compliance with the technical specifications for operation, per reactor

- - - 1.00 0.73 0.13 0.27 0.13 0.60 0.33

5Number of alignment errors, per reactor - - 0.40 1.09 0.69 0.13 0.27 0.13 0.60 0.33

6

Number of unscheduled reactor shutdowns, per reactor and for 7,000 hours of criticality • Automatic 1.70 1.51 1.30 0.74 0.73 0.44 1.13 0.82 0.58 0.74• Manual 1.96 1.39 2.18 1.28 2.54 1.48 1.04 1.44 1.68 1.22

7

Average collective dose in operations, per nuclear unit in service (in man-sieverts)• PWR 0.296 0.351 0.032 0.352 0.524 0.045 0.264 0.337 0.271 0.537• AGR 0.103 0.073 0.026 0.055 0.152 0.071 0.167 0.100 0.018 0.084

8

Exposure of individuals:• Number of individuals with

doses above 20 mSv 0 0 0 0 0 0 0 0 0 0• Number of individuals with

doses above 15 mSv 0 0 0 0 0 0 0 0 0 0

9Number of significant radiation protection events 98 192 184 368 249 58 38 31 43 43

10

Availability (%):• EDF Energy plants 74.6 78.2 69.9 71.9 66.1 62.8 51.2 71.0 65.7 72.0• PWR 88.9 86.1 89.4 83.7 85.3 98.4 89.2 87.4 45.6 82.5• AGR 73.6 77.6 68.5 71 64.7 60.2 48.5 69.8 67.1 71.3

11

Unplanned capacity loss factor (%)• EDF Energy plants 15.0 11.7 15.5 12.3 17.0 20.3 20.4 13.2 19.6 13.0• PWR 2.4 2.4 9.7 0.0 0.3 0.4 2.1 0.9 54.3 3.4• AGR 15.9 12.4 15.9 13.1 18.2 21.7 21.8 14.0 17.1 13.7

12Occupational accident frequency factor (for 200,000 hours worked) - - 0.51 0.37 0.22 0.27 0.35 0.11 0.007 0.12

13Occupational accident frequency factor (Tf) *

- - 2.6 1.9 1.1 1.4 1.8 0.6 0.4 0.6

* Tf: frequency of occupational accidents with sick leave for 1 million hours of work.

Page 192: EDFMESS FR - inis.iaea.org

8�

14

14.3 - Map of the EDF SA nuclear power plants

Construction/Project

Operation

Decommissioning

300 MWe 900 MWe 1300 MWe 1450 MWe 1600 MWe

FlamanvillePaluel

Penly

Gravelines

Chooz

Cattenom

Nogents/SeineDampierre

Belleville

Fessenheim

Chinon

Civaux

Blayais

Golfech

Bugey

St-Alban

Creys-Malville

Cruas

Tricastin

St-Laurent

Brennilis

Clermont-Ferrand

Nîmes

Grenoble

Lyon

Bordeaux

Bourges

Paris

AmiensCherbourg

Tours

Strasbourg

Marseille

Page 193: EDFMESS FR - inis.iaea.org

8�

14

14.4 - Map of the EDF Energy nuclear power plants

Construction or Project

Nomber of reactor per type AGR REP EPR Engineering

Operational

Engineering existingnuclear sites

4

14 1

2

Hunterston B

Hinkley Point B

Hinkley Point C

Barnwood

East Kilbride

Dungeness B

Sizewell B

Sizewell C

Hartlepool

Torness

Heysham 1

Heysham 2

London

Page 194: EDFMESS FR - inis.iaea.org

88

14

14.5 - Technical key dates for each of the EDF SA nuclear units

Year com-missioned

Nuclear Unit

Powerin

MWe(*)VD1 VD2 VD3

Year commis-sioned

Nuclear Unit

Powerin

MWe(*)VD1 VD2 VD3

1977 Fessenheim 1 880 1989 1999 2009 1984 Cruas 4 915 1996 2006

1977 Fessenheim 2 880 1990 2000 2011 1984 Gravelines 5 910 1996 2006

1978 Bugey 2 910 1989 2000 2010 1984 Paluel 1 1330 1996 2006

1978 Bugey 3 910 1991 2002 1984 Paluel 2 1330 1995 2005

1979 Bugey 4 880 1990 2001 2011 1985 Flamanville 1 1330 1997 2008

1979 Bugey 5 880 1991 2001 2011 1985 Gravelines 6 910 1997 2007

1980 Dampierre 1 890 1990 2000 2011 1985 Paluel 3 1330 1997 2007

1980 Dampierre 2 890 1991 2002 1985 St-Alban 1 1335 1997 2007

1980 Gravelines 1 910 1990 2001 2011 1986 Cattenom 1 1300 1997 2006

1980 Gravelines 2 910 1991 2002 1986 Chinon B3 905 1999 2009

1980 Gravelines 3 910 1992 2001 1986 Flamanville 2 1330 1998 2008

1980 Tricastin 1 915 1990 1998 2009 1986 Paluel 4 1330 1998 2008

1980 Tricastin 2 915 1991 2000 2011 1986 St-Alban 2 135 1998 2008

1980 Tricastin 3 915 1992 2001 1987 Belleville 1 1310 1999 2010

1981 Blayais 1 910 1992 2002 1987 Cattenom 2 1300 1998 2008

1981 Dampierre 3 890 1992 2003 1987 Chinon B4 905 2000 2010

1981 Dampierre 4 890 1993 2004 1987 Nogent 1 1310 1998 2009

1981 Gravelines 4 910 1992 2003 1988 Belleville 2 1310 1999 2009

1981 St-Laurent B1 915 1995 2005 1988 Nogent 2 1310 1999 2010

1981 St-Laurent B2 915 1993 2003 1990 Cattenom 3 1300 2001 2011

1981 Tricastin 4 915 1992 2004 1990 Golfech 1 1310 2001

1982 Blayais 2 910 1993 2003 1990 Penly 1 1330 2002 2011

1982 Chinon B1 905 1994 2003 1991 Cattenom 4 1300 2003

1983 Blayais 3 910 1994 2004 1992 Penly 2 1330 2004

1983 Blayais 4 910 1995 2005 1993 Golfech 2 1310 2004

1983 Chinon B2 905 1996 2006 1996 Chooz B1 1500 2010

1983 Cruas 1 915 1995 2005 1997 Chooz B2 1500 2009

1984 Cruas 2 915 1997 2007 1997 Civaux 1 1495 2011

1984 Cruas 3 915 1994 2004 1999 Civaux 2 1495

VD1 : First ten-yearly inspectionVD2 : Second ten-yearly inspectionVD3 : Third ten-yearly inspection (*) Net continuous power

Page 195: EDFMESS FR - inis.iaea.org

89

14

14.6 - Technical key dates for the EDF Energy nuclear units

YearCommissioned

Nuclear UnitReactor Number

Reference Unit Power (MW)

(1)

Planned date of withdrawal from

service(2)

1976 Hinkley Point B R3 435 2016

1976 Hinkley Point B R4 435 2016

1976 Hunterston B R3 460 2016

1976 Hunterston B R4 430 2016

1983 Dungeness B R21 520 2018

1983 Dungeness B R22 520 2018

1983 Heysham 1 R1 585 2019

1983 Heysham 1 R2 575 2019

1983 Hartlepool R1 595 2019

1983 Hartlepool R2 595 2019

1988 Heysham 2 R7 605 2023

1988 Heysham 2 R8 605 2023

1988 Torness R1 600 2023

1988 Torness R2 605 2023

1995 Sizewell 1191 2035

(1) Reference Unit Power means the rated electrical power of the generating unit as declared by EDF Energy in its daily transactions on 17 January 2012.

(2) Dates of withdrawal from service, including all life extension decisions made on 17 January 2012.

Page 196: EDFMESS FR - inis.iaea.org
Page 197: EDFMESS FR - inis.iaea.org

9�

14

14.7 - Table of abbreviations

AAAR Automatic reactor shutdown

AEN OECD Nuclear Energy Agency

AGR Advanced Gas-cooled Reactor

AIEA International Atomic Energy Agency (IAEA)

ALARA As Low As Reasonably Achievable

AMELIE Project to transform spare part logistics

AMT Joint Maintenance Agency

ANDRA French National Radioactive Waste Management Agency

ASN French Nuclear Safety Authority

BBR Reactor Building

BK Fuel Building

CCAP Annual Performance Contract

CSB Chemical Safety and hazard investigation Board

CEFRI French committee for the certification of companies in training and monitoring radiation workers

CEIDRE Construction and Operation Expert Appraisal and Inspection Centre

CENG Constellation Energy Nuclear Group

CGNPC China Guangdong Nuclear Power Company

CIDEN Nuclear Environment and Decommissioning Engineering Centre

CIEST Inter-Contractor Work Conditions and Safety Committee

CIPN Nuclear Equipment Engineering Department

CIPR International Commission for Radiological Protection

CLI Local Information Commission

CNEN Nuclear Engineering Department for New Build

CNPE Nuclear Power Plant

COPAT Outage Control Centre (OCC)

CSN Nuclear Safety Council

CSNE Plant Safety Review Committee

DDCN Nuclear Fuel Division

DIN Nuclear Engineering Division

DNMC Daya-Bay Nuclear operation and Management Company

DPI Production and Engineering Directorate

DPN Nuclear Operations Division

EEDEC Skills and Employment Development Pledge

EGE Overall Excellence Evaluation

EGS Overall Nuclear Safety Assessment

ENISS European Nuclear Installations Safety Standard

EPR European Pressurised Reactor

EPRI Electric Power Research Institute

ESR Significant radiation protection event

ESS Significant nuclear safety event

EVEREST EDF campaign to allow entry into controlled areas in street clothes

ENSREG European Nuclear Safety Regulators Group

FFAI Fire Action Sheet

FAVL Long-lived low-level radioactive waste

GGPEC Advanced planning of jobs and skills

GV Steam generator

HHCTISN High Committee for Transparency and

Information on Nuclear Safety

HSE Health and Safety Executive

IFOPSE Industrial Safety and Prevention Training Institute

Page 198: EDFMESS FR - inis.iaea.org

92

14

IIN Nuclear Inspectorate (part of EDF Nuclear

Operations Division)

INB Licensed Nuclear Facility

INES International Nuclear Events Scale

INPO Institute of Nuclear Power Operations (USA)

INRA International Nuclear Regulator Association

INSAG International Safety Advisory Group

INTEP EDF Research and Development Division initiative to introduce new technologies in the plants in service

INTRA Post-accident robotic response group

IOP Operations Engineering

IRPA International Radiation Protection Association

ISOE Information System for Occupational Exposure of OECD AEN

MMAE Nuclear Engineering Division Support and

Expertise Unit

MOPIA Project to set in place an attractive business policy

MPL Front Line Manager

NNDA Nuclear Decommissioning Authority (UK)

NEI Nuclear Energy Institute

NNB Nuclear New Build

NNSA National Nuclear Safety Administration (China)

NRC Nuclear Regulatory Commission (USA)

OO2EI Better Housekeeping Campaign (a Nuclear

Operations Division initiative)

ONR Office of Nuclear Regulation (UK)

OSART Operational Safety Analysis Review Team (IAEA)

PPAI Fire Action Plan

PARTNER Ecologically-sensitive nuclear plant administrative facility refurbishment

PBMP Basic Preventive Maintenance Programme

PGAC Worksite General Assistance Services

PHPM Methods and Practices Harmonization Project

PUI On-Site Emergency Plan

RRDA Project to reduce the durations of outages

R&D Research and Development Division

REP Pressurised water reactor

RET Exceptional Work Permit

REX Experience feedback

RTE Power grid

RTGE General Technical Rules intended to avoid or mitigate the off-site nuisances and hazards associated with the operation of licensed nuclear facilities

SSDIN Nuclear Technical Information System

SDIS County Fire and Emergency Services

SEPTEN Basic Design Department (Nuclear Engineering Division)

SIR Official Inspection Departments

SOH Socio-organizational and human approach

SPR Risk Management Department

SRD Safety and Regulation Department (EDF Energy)

SRO Senior Reactor Operator

STE Technical Specifications for Operation

TTEM Unit in service

TNPJVC Joint venture between the Chinese company CGNPC (�0%) and EDF (�0%)

TSM Technical Support Mission by peers organised by WANO

TSN Nuclear Safety & Transparency Act

Page 199: EDFMESS FR - inis.iaea.org

9�

14

UUFPI Operations Engineering Training Unit

UNE Unistar Nuclear Energy

UNIE Operations Engineering Unit

UNGG Gas-cooled graphite-moderated reactor

UTO Central Technical Support Department

VVD Ten-yearly inspection

VP Partial inspection

WWENRA West European Nuclear Regulators

Association

WANO World Association of Nuclear Operators

Page 200: EDFMESS FR - inis.iaea.org
Page 201: EDFMESS FR - inis.iaea.org

9�

14

Jacques DUSSERRE, Jean-Paul COMBEMOREL, JeanTANDONNET, Peter WAKEFIELD, Christian THEZEE, Bruno CORACA

Photo CREDITS

Cover : © EDF Médiathèque - H. M. DUCLOS Chapter 2-� : © EDF Médiathèque - Alexis MORIN Chapter 2-� : © EDF Médiathèque - H. M. DUCLOS Chapter 2-�0 : © Sylvain CONSEIL Chapter � : © EDF Médiathèque Chapter 4 : © TEPCO Chapter � : © EDF Médiathèque - H. M. DUCLOS Chapter � : © EDF Médiathèque Chapter � : © EDF Médiathèque - H. M. DUCLOS Chapter 8 : © EDF Médiathèque - Alexis MORIN Chapter 9 : © EDF Médiathèque Chapter �0 : © EDF Energy Chapter �0 : © TNPJVC Chapter �� : © ESKOM Chapter �2 : © EDF Médiathèque

Page 202: EDFMESS FR - inis.iaea.org
Page 203: EDFMESS FR - inis.iaea.org

2011

E.D.F.

Présidence IGSN

21, avenue de Messine

75008 Paris

www.edf.fr

The Inspector General’s report

on Nuclear Safety and Radiation Protection

ED

F IG

SN

R

EP

OR

T’S

2

01

1

079999_EDFMESS_Couv.indd 1-3 30/03/12 07:25