An Overview of Differences in Nuclear SafetyRegulatory Approaches and RequirementsBetween United States and Other Countries
Prepared byH. P. Nourbakhsh
October 2004
Advisory Committee on Reactor Safeguards (ACRS)U.S. Nuclear Regulatory CommissionWashington, DC 20555-0001
ii
ABSTRACT
This report has been prepared for use by the NRC Advisory Committee on Reactor Safeguards(ACRS) in support of its ongoing effort to inform the Commission on significant differences inregulatory approaches and requirements between the United States and other countries. Thisreport, which is based on review of a number of documents issued by various internationalorganizations, provides an overview of regulatory approaches and discusses differences in specificregulatory requirements of current interest in the United States.
The views expressed in this report are solely those of the author and do not necessarily representthe views of the ACRS.
iii
CONTENTSPage
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivACRONYMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. GENERAL OVERVIEW OF REGULATORY APPROACHES IN THE WORLD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3. DIFFERENCES IN REGULATORY REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . 73.1 Design-Basis Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.1 Acceptance Criteria for Emergency Core Cooling System . . . . . . . . 73.1.2 Extent of Fuel Failure that is Assumed in Radiological Assessment . . 83.1.3 Strainer Blockage Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Periodic Safety Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.3 Protection Against Severe Accidents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143.4 Risk-Informed Regulations and Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . 163.5 Materials Degradation Issues and Aging Management . . . . . . . . . . . . . . . . . 17
4. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
iv
TABLES
Page
1. Operating reactors in various countries . . . . . . . . . . . . . . . . . . . . . . . . . . . 32. Operating and under construction reactors by type . . . . . . . . . . . . . . . . . . 43. The extent of fuel failure that is assumed in radiological assessments . . . 94. Summary of the BWR strainer modifications in different countries
after the Barsebäck-2 event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105. Summary of the PWR stainer modifications in different countries
after the Barsebäck-2 event . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116. Periodic safety review requirements in various countries . . . . . . . . . . . . 137. Operating license periods in various countries . . . . . . . . . . . . . . . . . . . . 18
v
ACRONYMS
Acronym Definition
ABWR Advanced Boiling Water ReactorACRS Advisory Committee on Reactor SafeguardsAGR Advanced Gas-cooled ReactorALARA As Low As reasonably AchievableALARP As Low As reasonably PracticableBWR Boiling Water ReactorCFR Code of Federal RegulationsCSNI Committee on the Safety of Nuclear InstallationsCNRA Committee on Nuclear Regulatory ActivitiesCSS Containment Spray SystemDCH Direct Containment HeatingECCS Emergency Core Cooling SystemEPR European Pressurized Water ReactorGCR Gas-Cooled ReactorGE General ElectricIAEA International Atomic Energy AgencyLOCA Loss-of-Coolant AccidentLWR Light Water ReactorNEA Nuclear Energy AgencyNEI Nuclear Energy AgencyNII Nuclear Installations InspectorateNRC Nuclear Regulatory CommissionOECD Organization for Economic Cooperation and DevelopmentPHWR Pressurized Heavy Water ReactorPWR Pressurized Water ReactorPRA Probabilistic Risk AssessmentRCS Reactor Coolant SystemSAM Severe Accident Management SRM Staff Requirements MemorandumSSCs Systems, Structures, and ComponentsTMI-2 Three Mile Island Unit 2U.K. United KingdomU.S. United States
1
1 INTRODUCTION
The purpose of this report is to provide anoverview of differences in nuclear safetyregulatory approaches and requirementsbetween United States (U.S.) and othercountries.
In an April 28, 2003 Staff RequirementsMemorandum (SRM) [1], resulting from theApril 11, 2003 meeting with AdvisoryCommittee on Reactor Safeguards (ACRS),the Commission stated that "In the course ofits routine activities of reviewing and advisingthe Commission on reactor issues, theCommittee should explore and consider otherinternational regulatory approaches. Wherethere are significant differences in regulatoryapproaches and requirements, theCommission should be informed." This reporthas been prepared for use by the ACRS inresponding to the Commission request.
This report focuses on regulatoryrequirements pertinent to western-designedlight water reactors (LWRs). It does notaddress requirements relating to nuclearmaterials and waste safety, or safeguard andsecurity issues.
A number of documents issued by variousinternational organizations, in particular theEuropean Commission and the Organizationfor Economic Cooperation and Development/Nuclear Energy Agency (OECD/NEA), werereviewed for the preparation of this report.
The European Commission has sponsoredmany studies to support its activities towardharmonization of safety requirements andpractices in an enlarged European Union.The results of these studies [2-7] on review ofsafety philosophies and practices inEuropean Union member states were themajor source of information for developingthis report.
The OECD/NEA reports [8-11] on thescientific and technological background ofnuclear safety criteria, rules and guidelines,and applied assessment methods werereviewed to identify the safety issues forwhich there may not yet be a commontechnical position among internationalcommunities.
The adoption of the Convention on NuclearSafety in 1994 legally binds the participatingcountries to maintain a high level of safety.The Convention obliges parties to submitreports on the implementation of theirobligations for “peer review” at regularmeetings of the parties held by theInternational Atomic Energy Agency (IAEA).The National Reports on the Convention ofNuclear Safety [12] were also utilized for thepreparation of this report.
The report begins with a general overview ofregulatory approaches in various countries.It then discusses differences in the specificregulatory requirements in the areas ofcurrent interest in the U.S.. They are:
• Design-basis assessment• Periodic safety reviews• Protection against severe accidents• Risk-informed regulations and practices• Materials degradation issues and aging
management
2
2 GENERAL OVERVIEW OF REGULATORY APPROACHES IN THE WORLD
Regulatory policies differ from country tocountry. These differences reflect thedifferences in culture, social, economic, andgovernmental systems between countries [2].Regulatory regimes fall broadly within twocategories, prescriptive or otherwise. In aprescriptive regime, the requirements onmethodologies, standards, and qualityassurance are prescribed by the regulatoryauthority. The licensee must demonstratethat the plant complies with these regulatoryrequirements. In addition regulatoryguidelines, describing methods acceptable tothe regulatory authority, may be providedwhich the licensee can follow forimplementing specific portions of regulations.The U.S. Nuclear Regulatory Commission(NRC) regulations fall in this category.
In a less prescriptive regime, the emphasis ison principles which are largely qualitative(except perhaps for certain parameters, e.g.,dose limits). The licensee must comply withthese principles, but may choose its ownmethodology of meeting them. Europeanregulations are generally less prescriptivethan those in the U.S. There is, however,some degree of variation among theEuropean countries [2].
The interactions between regulator and utilityvary from country to country. Generally suchinteractions are formal with respect tolicensing, but less formal from the point ofview of safety research. However, there aredifferences within the licensing relationships.Some regulators encourage a collaborativeapproach and a continuing dialog through thestages of a licensing application, others adopta very formal approach [2]. There is a non-adversarial relationship between the plantoperator and the regulatory authority in manycountries. This is, in part , due to the fact thatthe plants are owned by the governmentinstitutions in these countries.
The safety approaches and practices in thewestern countries have been largely open topublic knowledge and scrutiny. This hasencouraged a collaborative safetyconsciousness over many years. TheEastern European and the former Sovietcountries are moving toward more opensafety practices. This has been facilitatedthrough the influence of various IAEA, OECD,U.S., and European initiatives.
There is a strong influence of the U.S.regulatory system in setting the basis forlicensing requirements in many countries.This is because a large number of plants inoperation in other countries are of U.S.design or derived from U.S. designs, whichmust be licensable in their country of origin.Some countries (e.g., Spain, Holland, andBelgium) completely follow the regulations ofthe country from which their nuclear powerplants were purchased. They follow the U.S.NRC regulations for their Westinghousepressurized water reactors ( PWRs) and theGeneral Electric (GE) boiling water reactors(BWRs), and the German regulations for theirSiemens (KWU) plants [2].
Operating reactors by country and by typeare presented in Table 1, and Table 2respectively. The LWR technology wasinitially developed in the U.S., with GEpioneering the BWRs and Westinghousedeveloping the PWRs. The main nuclearelectricity production in Europe and the FarEast, in common with the rest of the world,now comes from LWRs. The exception is theUnited Kingdom (U.K.) where advanced(oxide fueled) gas-cooled reactors (AGRs)provide a large fraction of the nuclear-generated electricity. The U.K. alsosubsequently elected to follow the LWRroute. The LWR also provides the basicconcepts for the WWER reactors developed
1 Based on information in IAEA (PRIS) database [13], last updated on May 19, 2004
3
Table 1 Operating Reactors in Various Countries1
CountryNo. of Operational Units
Total MW(e)
PWR BWR Other Total
United States Of America 69 35 0 104 98298France 58 0 FBR 1 59 63363Japan 23 27 ABWR 3
FBR 154 45464
Russian Federation 0 0 LWGR 15 WWER 14FBR 1
30 20793
United Kingdom 1 0 AGR 14GCR 12
27 12052
Republic of Korea 15 0 PHWR 4 19 15850Germany 12 6 0 18 20643Canada 0 0 PHWR 17 17 12113India 0 2 PHWR 12 14 2550Ukraine 0 0 WWER 13 13 11207Sweden 3 8 0 11 9451Spain 7 2 0 9 7584China 7 0 PHWR 2 9 6587Belgium 7 0 0 7 5760Taiwan 2 4 0 6 4884Czech Republic 0 0 WWER 6 6 3548Slovak Republic 0 0 WWER 6 6 2442Switzerland 3 2 0 5 3220Bulgaria 0 0 WWER 4 4 2722Finland 0 2 WWER 2 4 2656Hungary 0 0 WWER 4 4 1755Republic of Lithuania 0 0 LWGR 2 2 2370Brazil 2 0 0 2 1901South Africa 2 0 0 2 1800Mexico 0 2 0 2 1310Argentina 0 0 PHWR 2 2 935Pakistan 1 0 PHWR 1 2 425Slovenia 1 0 0 1 656Romania 0 0 PHWR 1 1 655Netherlands 1 0 0 1 449Armenia 0 0 WWER 1 1 376
2 Based on information in IAEA (PRIS) database [13], last updated on May 19, 2004
4
Table 2 Operating and Under Construction Reactors by Type2
TypeOperational Under Construction
No. of Units Total MW(e) No. of Units Total MW(e)
PWR 214 204335 6 6111
BWR 90 78025 1 1067
WWER 50 33040 8 7534
PHWR 39 19972 8 3135
LWGR 17 12589 1 925
AGR 14 8380 0 0
GCR 12 2484 0 0
ABWR 3 3955 3 3904
FBR 3 1039 0 0
in Russia and used in other EasternEuropean countries and Finland.
The LWRs in other countries are quite similarto the designs developed in the U.S.. TheFrench PWRs are very similar to theWestinghouse PWRs since France hadbought the license for their design fromWestinghouse. The PWRs and BWRsdesigned by the KWU (Siemens) are similarto the U.S. designs but are different in theconfigurations of their containments. TheBWRs designed by ABB Atom are similar tothe Mark-II BWR plants of GE, except, withsome modifications (e.g., internal pumps).
The LWRs in other countries, having beencommissioned after the U.S. LWRs, weredesigned and constructed in accordance withthe design criteria and safety philosophy
developed in the U.S.. The U.S. safetyphilosophy of defense in depth was adoptedby the regulatory authorities in westernEurope, Japan, and Korea, not only for thebarriers to the release of radioactivesubstances, but also in the design,construction, quality assurance, inspection,and operational practices. However, theremay be differences in the implementation ofthe defense-in-depth principle, e.g., in levelsof diversity and redundancy required from thesafety systems. Requirements for threetrains of safeguard in France and four trainsof safeguard in Germany ( because of on-linemaintenance) and the requirement fordiversity of instrumentation for all safety-related measurements in Germany areexamples of such dif ferences inimplementation of the defense-in-depthprinciple. There are also some country-
5
specific regulatory requirements regardingthe effectiveness of the various barriers. Onesuch example is the requirement in Germanyto design the containment to withstand thecrash of a light fighter plane.
The 1979 accident at Three Mile Island Unit2 (TMI-2), led to the reexamination of thedesign basis and the consideration ofregulations for protection against severeaccidents. The reexamination of the designbasis was prompted by the fact that the TMI-2accident initiated with a small-break loss -of-coolant accident (LOCA), whoseconsequences should have been bounded bythose of a large-break LOCA, but becamemuch more severe due to misunderstandingof the event by the operators. The event-based procedures have been modified tosymptom-based procedures in most Westernplants [2].
The Chernobyl accident in 1986 affectedopinion in Western Europe about the safetyof nuclear power plants in general,contributing to the decisions of somecountries (e.g., Germany) to tighten safetyrequirements for new plants, implying designmodifications, or to phase out nuclear powerstations, either immediately (e.g., Italy), orover a period of time (e.g., Sweden,Germany).
In most countries, the principles of traditionaldeterministic approach have been acceptedover many years to demonstrate the reliabilityand safety of design. ( deterministic approachrefers to an approach that specifies certaindesign and operational conditions and appliesbounding criteria to demonstrate acceptableplant performance.) Systems, structures,and components (SSCs) are designed andmanufactured to accepted standards,regulations, codes of practice etc. to ensurethat the SSCs can perform their intendedfunctions. The single-failure criterion hasbeen commonly adopted, as has the 30minutes rule, i.e., that the safety objectives
can be met without operator interventionwithin the first 30 minutes into an accident.
The majority of licensing submittals havebeen based on the evaluation model (EM)methodology. This was established on thepremise that deliberate modelingconservatisms are included to compensatefor lack of knowledge of the governingphenomena. This methodology was basedon the Appendix K of the U.S. Code ofFederal Regulations (10CFR Part 50).However, with improved understanding of thephenomena, there have been moves tochange the conservative biases andassumptions of the evaluation modelmethodology, allowing the licensee to movefurther toward best-estimate methodologies.Within the U.S. this led to a revision of theemergency core cooling system (ECCS) rule(10CFR 50.46) in 1988 enabling licensees toapply best-estimate methodologies, with theprovision that due allowance is given to anyremaining uncertainties in code, data, ormodeling. The move toward best-estimatemethodologies is also a common trend inmost countries [2].
In light of increased realization of the impactof human factors on plant safety, regulatoryauthorities now require the utilities to considerhuman factor engineering concepts in thedesign and operational aspects of plants.There is an international recognition of theimportance of safety culture andmanagement. There is an evolvingconsensus on what constitutes goodperformance on the part of an organizationbut less on how it can be measured.
The ALARA (or ALARP) principles aregenerally adopted to ensure that risks arereduced to a level acceptable to theregulatory body to be “as low as reasonablyachievable (or practicable).” Most countriesfollow this approach in qualitative terms. Inprinciple risk may be quantified via a cost-benefit analysis, whereby the costs to
6
industry are compared with the benefits tosociety. The extent to which cost-benefitanalysis is encouraged or allowed varies fromcountry to country. In most Europeancountries, safety improvements are generallyintroduced without the requirement for formalcost-benefit analysis [2]. Nevertheless,cost/benefit is informally considered byregulatory authorities in all of these countries.The issue of cost may become moreimportant as competition grows in Europe.
Basic deterministic safety assessments arenow generally complemented by probabilistic
risk assessments (PRAs) to verify the overalldesign and system of operation. PRAs areconducted by many countries to demonstratethat there are no sudden increase in risk foraccidents that are outside of the design basis.Most countries with nuclear power plantshave performed PRAs and have found thatsuch assessments often lead to theidentification of plant vulnerabilities. However,there is not much support, so far, in manyother countries for formally considering riskinformation in regulatory decisionmaking as itis in the U.S..
7
3 DIFFERENCES IN REGULATORY REQUIREMENTS
Despite considerable similarities in theobjectives and actual implementation ofnuclear safety regulatory approaches, thereare differences in regulatory requirementsacross the world. Indeed, efforts toharmonize safety requirements andregulatory practices within the EuropeanUnion have been unsuccessful so far.
Reasons for the differences in regulatoryrequirements relate to national energy policy(mainly in support of public acceptance);national industrial tradition (e.g., giving morecredit to redundancy or diversity, or creditinga software-based system as opposed to hard-wired controls); consistency with nationalregulatory or legislative system (e.g.,compliance with probabilistic safety criteria onindividual and societal risk as applicable tothe environmental policy); country-specificconditions (e.g., differences in geographysuch as flooding for Netherlands andseismic for Japan); and uncertaintiesassociated with the severe-accidentphenomena.
Some of the areas where differences insafety requirements exist are discussedbelow.
3.1 Design-Basis Assessment
There is an internationally accepted rule thatthe licensee should provide a comprehensivesafety assessment to confirm that the designof an installation fulfils the safety objectivesand requirements. This assessment issubmitted in a safety analysis report. Specificapproval by the regulatory body is requiredbefore the start of operation. The U.S. NRCregulation (10CFR50.71) requires thelicensee to update periodically ( the intervalbetween updates should not exceed 24months) the final safety analysis report(FSAR) originally submitted as part of theapplication for the operating license. The
update should include the effects of: allchanges made in the facility or procedures asdescribed in the FSAR; all safety analysesand evaluations performed by the licensee insupport of approved license amendments,and all analyses of new safety issuesperformed by or on behalf of licensee atCommission request. In many othercountries, a safety analysis report is updatedevery 10 years as a part of periodic safetyreviews (see section 3.2). These reviewsmust take account of existing operationalexperience and any other information relevantto safety that is currently available.
The accident sequence groups and theaccidents to be analyzed in the safetyanalysis report may be prescribed by theregulator (e.g., U.S. NRC), but if not, aredefined by the licensee as part of its safetycase submission (e.g., United Kingdom). Theimplementation of either approach is similar.There are, however, some differences incertain acceptance criteria and the licensingcalculations due to various degree ofconservatism made at each step of thecalculation. Some of these differences aresummarized below.
3.1.1 Acceptance Criteria for EmergencyCore Cooling System
Most countries use acceptance criteria forECCS that are based on those specified inAppendix K to 10CFR Part 50. Germany hasalso established an additional acceptancecriterion to limit the fraction of failed fuel cladunder LOCA conditions. The 10% fuel failurecriterion in Germany was originallyestablished to limit the radiologicalconsequences in case of a LOCA (seesection 3.1.2). The original intention of thiscriterion has since been broadened. Besidethe radiological aspects, this criterion hasbeen used for the evaluation of core loading.If the core is loaded with new fuel rods or new
8
loading strategies are applied, thecompliance with the 10% fuel failure criterionhas to be demonstrated again by theapplicant [6].
There is a common understanding among theGerman licensing authorities that if thecompliance with all these acceptance criteriacan be proven, there is no need to limit thefuel burn-up or to restrict the core loading[6].
3.1.2 Extent of Fuel Failures that isAssumed in Radiological Assessment
The extent of fuel failure that is assumed inradiological assessments varies from countryto country. In performing the design-basisaccident analyses, the commonly appliedpractice includes the use of conservativeassumptions regarding system performanceand components failure. Following a largeLOCA, it is assumed that a fraction of the fuelis failed allowing release of the radionuclidesfrom the fuel into the containmentatmosphere. This release of fission productsinto the containment (“in-containment sourceterm”) has a wide range of regulatoryapplications, including the basis for (1) theadequacy of the leaktightness of thecontainment, (2) the performancerequirement of fission-product cleanupsystems such as sprays and filters, (3) post-accident habitability requirements for thecontrol room, and (4) the radiationenvironment for qualification of safety-relatedequipments.
The determination of source term inside thecontainment involves assumptionscorresponding to to various physical stagesin the release of fission products, includingfraction of core failure, release from damagedfuel, airborne part of release and release intoreactor coolant system and sumps, chemicalbehavior of iodine in the aqueous and gasphases, and natural and spray removal in thecontainment atmosphere.
Some countries (Belgium and Spain) followthe U.S. and assume a source termcorresponding more to a core-melt accidentdecoupled from the LOCA thermal-hydrauliccalculations, while other countries take intoaccount the physical phenomena during aLOCA still with conservative assumptions.
Table 3 shows the extent of fuel failure that isassumed in radiological assessments indifferent countries. Many countries (e.g.,Belgium, United Kingdom, Spain) follow theU.S. and assume 100% fuel failures during alarge LOCA. Some European countries(e.g., Germany, Switzerland, Netherlands)assume 10% of fuel failure during a LOCA.In France, a 100% fuel failure assumption isused for the radiological consequencesevaluation of the 900 and 1300 MWe plants.However, for N4 plants, a 33% fuel failureassumption has been proposed by the utilityand is under assessment by the regulatorybody (IPSN). The utility position is that thisvalue is sufficiently conservative to constitutea decoupling assumption avoiding a specificsafety demonstration for each core refueling,taking into account a previous Framatomestudy for the 1300 MWe French nuclearpower plant design for which 7% of cladfailure was predicted [6].
9
Table 3 The Extent of Fuel Failure That Is Assumed In Radiological Assessments
Country Extent of Fuel Failures inRadiological Assessment
Belgium 100%
France 100% (33% proposed for N4 plants)
Germany 10%
Netherlands 10%
Spain 100%
Switzerland 10%
United Kingdom 100%
United States 100%
3.1.3 Strainer Blockage Issue
The 1992 clogging of intake strainers forcontainment spray water in Barsebäck-2, aBWR in Sweden, renewed the focus ofregulators around the world on safetyquestions associated with strainer cloggingwhich, until then, had been considered asresolved.
Although the Barsebäck incident in itself wasnot very serious, it revealed a weakness inthe implementation of defense-in-depthconcept in the design, which under othercircumstances could have led to the failure ofthe ECCS and containment spray system(CSS). The Barsebäck-2 event alsodemonstrated that larger quantities of fibrousdebris could reach the strainers than hadbeen predicted by models and analysismethods developed for the resolution of thestrainer blockage issue[8,14].
The Barsebäck-2 incident prompted action onthe part of regulators and utilities in othercountries. Research and development effortsof varying intensity were launched in many
countries. Extensive studies have beenperformed to assess the amount of insulationmaterials that could be dislodged during pipebreak events inside the containment. Inmany countries, the analyses were based onthe double cone model developed by theNRC [14]. The analyses have also includedspecific studies of the transport of insulationmaterials and other debris in the containment,and of strainer pressure drops. Such effortsresulted in a number of corrective actionsbeing taken in BWRs and some PWRsaround the world. For a number of plants,actions were taken as direct responses torequirements issued by regulatory authorities,while for other plants back-fitting measureswere introduced voluntarily or because ofanticipated requirements [9].
The modifications of the ECCS and/or CSSsuction strainers carried out in differentcountries are summarized in Tables 4 and 5for BWRs and PWRs respectively. Themodifications have resulted in new strainerdesigns with significantly enlarged filteringarea. Most of the new strainers have goodself-cleaning properties. In some BWRs, the
10
Table 4 Summary of the BWR Strainer Modifications in Different Countries After the Barsebäck-2 Event
CountryBWR
Strainer Modifications Comment
United States ofAmerica
34 out of 34 unitsmodified
New strainers with significantlyincreased area
Established a schedule to removeparticulate and other debris fromthe suppression pool
Japan None of 28 unitsmodified
More than 95% of the insulationsare replaced by non-fiber typeones (e.g., Reflective MetallicInsulation).
The absence of foreign materialsin the suppression pool isensured through inspection andmaintenance practices
Sweden 9 out of 9 unitsmodified
New strainers with 15 to 40-foldarea increase
Germany 4 out of 6 unitsmodified
Stainers were enlarged
Spain 2 out of 2 unitsmodified
New strainers with significantarea increase
Switzerland 2 out of 2 unitsmodified
New strainers with 7 to 30-foldarea increase
Finland 2 out of 2 unitsmodified
About 10-fold area increase
11
Table 5 Summary of the PWR Stainer Modifications in Different Countries After the Barsebäck-2 Event
CountryPWR
Strainer Modifications Comment
United States of America One out of 69 unitsmodified
Davis-Besse is the only U.S. PWRplant that its licensee voluntarilyenlarged its sump screen area - by30 fold
France EDF plans Backfits tosump strainers at all 58units
EDF is finishing design studies forthe backfit, which is expected toconsist of replacing the current filterswith a system of pierced piping thatprovides more strainer surface.
Japan None of 23 units modified More than 95% of the insulations arereplaced by non-fiber type ones(e.g., Reflective Metallic Insulation).
The absence of foreign materials inthe recirculation sumps is ensuredthrough inspection and maintenancepractices.
Germany 2 out of 12 units modified Stainers were enlarged.
Spain None of 7 units modified
Belgium 2 out of 7 units modified 6-fold area increase
Switzerland None of 3 units modified
Sweden 1 out of 3 units modified New strainers with > 7-fold areaincrease
Netherlands 1 out of 1 unit modified New strainer installed (50% areaincrease)
There is no mineral or fiberglassinsulation of noteworthy importancearound the primary components aswell as in the sump area.
Finland 2 out of 2 WWER-440/213 units modified
New strainer design with significantarea increase
12
design includes the capability to back-flushthe strainers [9].
Large fractions of the thermal insulationmaterials utilized on piping and othercomponents inside the containment have alsobeen replaced. The newly installed insulationmaterials vary both within and amongcountries. They are primarily reflectivemetallic insulation, nuclear grade fiberglass,mineral wool, and calcium silicate. The sameinsulation material (e.g., mineral wool) areinstalled differently in different countries (i.e.,Jacketed or encapsulated in cassettes). Theadministrative measures taken in othercountries include a periodic cleanup of thesuppression pool and containment sumps,with the aim to minimize the presence offoreign materials, and the control andeventual improvement of the containmentcoating.
In U.S., all BWR licensees were required toimplement appropriate measures to ensurethe capability of the ECCS to perform itssafety function following a LOCA. The U.S.nuclear industry addressed the NRCrequirements by installing large capacitypassive strainers in each BWR plant andestablishing a schedule to remove particulateand other debris from the suppression pools.Most U.S. BWR licensees followed theguidance prepared by the U.S. BWR OwnersGroup during the development of theircorrective actions.
As a result of research findings related toresolving the BWR ECCS strainer blockagesafety issue, the NRC conducted furtherresearch to determine if the transport andaccumulation of debris in a containmentfollowing a LOCA would impede the operationof the ECCS in operating PWRs. Theresearch program included debris transporttests, debris settling tests, debris generationtests, computational simulations, and variousengineering analyses. The results of thesestudies indicated the need for accurate plant-
specific assessment of adequacy of therecirculation function of the ECCS and CSSfor each operating PWR. The NuclearEnergy Institute (NEI) also recognized thisneed and has developed guidance for suchplant-specific assessment, which is underreview by the NRC staff.
The issue of strainer blockage in PWRs havebeen particularly troublesome. Continuingresearch revealing new modes of blockagehas shown that the prompt actions taken bysome European plants may not havecompletely alleviated the problem of strainerblockage. Indeed, redesign may be requiredof these plants. There is a strong evidencethat plant owners throughout the world do nothave a definitive solution to the issue.
3.2 The Periodic Safety Reviews
In contrast to U.S. NRC, most regulatoryauthorities in the world have a requirementthat the nuclear power plants be subject to anoverall assessment on a periodic basis, inaddition to the permanent supervision theregulatory body exerts on these plants. Table6 presents a comparison of internationalpractices with respect to periodic safetyreview activities.
The periodic safety review is a safety conceptmainly developed in the European countriesand was introduced later in the IAEAdocuments [16]. The periodic safety reviewsare complementary to the routine reviews ofnuclear power plant operation (includingmodifications to hardware and procedures,significant events, and operating experience)and special safety reviews following majorevents of risk significance. The frequency ofreview varies from country to country;typically every ten years ( see Table 6). Theperiodic safety review necessitates licenseesto take into account advances in technologyunconstrained by licensing basis as in U.S..
13
Table 6 Periodic Safety Review Requirements in Various Countries
Country Periodic SafetyReview
Frequency
Comment
United States OfAmerica
None Required Requiring the licensee to maintain thelicensing basis for the facility or activity
France Every 10 years(normally)
Linked to the statutory 10 year outageprogram interval, but the in-depthsafety assessment performed onrequest , at the regulator’s discretion
Japan Every 10 years Limited scope, concentrated mainly onaging behavior, without the evaluationof the overall plant design
United Kingdom Every 10 years Comprehensive safety reviewsRequirements stipulated in conditionsattached to licenses
Republic of Korea Every 10 years Comprehensive safety reviews
Germany Every 10 years Comprehensive safety reviews
Canada None required License renewed every 2-5 yearssubject to satisfactory safetyperformance
Sweden Every 10 years Comprehensive safety reviews
Spain Every 10 years Comprehensive safety reviews
Belgium Every 10 years Comprehensive safety reviews
Czech Republic Every 10 years Comprehensive safety reviews
Switzerland Every 10 years Comprehensive safety reviews(Regulatory requirement for facilities tocomply with the state-of-the-art inscience and technology)
Finland Every 10 years Comprehensive safety reviews
Hungary Every 10 years Comprehensive safety reviews
Mexico Every 10 years Comprehensive safety reviews
Netherlands Every 10 years Comprehensive safety reviews
14
The objective of these periodic safety reviewsare to assess the cumulative effects of plantaging and plant modifications, operatingexperience, technical developments, andsiting aspects. The reviews include anassessment of plant design and operationagainst current safety standards andpractices in order to propose any eventualimprovement. The reviews also examine anextension of the original design basis of theplant, in particular postulated initiating events(internal and external) not considered earlier.The reassessment of the original designbasis in Europe is strongly linked to theresearch on severe accidents and theirmanagement strategies [3]. Deterministic safety analyses are used insafety reassessments made in the periodicsafety reviews. However, it is now commonto complement the deterministic analyseswith a PRA (level 1 or 2), in particular todetermine the modifications that significantlyimprove the safety [3].
3.3 Protection Against SevereAccidents
The desire for protection against severeaccidents is shared by all of the regulatoryauthorities in the Western World. It has alsobeen argued that the severe accident is avery low probability event; it deserves aresponse, but the cost/benefit should be afactor. This argument has been accepted bythe U.S. NRC and it is a part of the regulatorypractice ( backfit rule, 10CFR50.109).However, most regulatory authorities of theEuropean Union Member States do notformally accept this argument. Nevertheless,cost/benefit is silently considered by all ofthese authorities[2].
The first significant regulatory action forsevere accident mitigation was the hydrogenrule (10CFR50.44) issued by U.S. NRC soonafter the TMI-2 accident. The rule required
control of the hydrogen that is produced in asevere accident. Decisions were made toinert the BWR Mark-I and Mark-IIcontainments and install igniters for hydrogencontrol in BWR Mark-III and the icecondenser containments. The PWR plantswith large dry containments (including thoseoperating with a sub-atmospheric internalpressure) were exempted from hydrogencontrol, because of the large volume of theircontainments. Regarding hydrogen control,the BWR Mark-I and Mark-II plants inEuropean countries followed suit in inertingcontainment atmosphere. The PWRs inEurope have gone through a long evaluationprocess and most of them (except theWestinghouse-designed plants) have decidedto install catalytic hydrogen recombiners ofsufficient capacity to address severe accidenthydrogen production [2].
The phenomenology of the severe accident isextremely complicated. The severe accidentevaluation methodologies are associated withlarge uncertainties. In fact, suchuncertainties have led different parties toreach to different conclusions from researchresults obtained for several severe accidentphenomena. For example, there is a largeuncertainty associated with the coolability ofa melt/debris attacking the concrete basemat,by flooding with water. This has introduceddifferent approaches for severe accidentmanagement strategies. For example, U.K.,Spain, Belgium, Sweden, and Finland will addwater to their PWR cavities and their BWRlower dry-wells in order to fragment the melt,to facilitate its cooling, and possibly delay thebasemat melt-through. On the other hand,the Germans do not have either the facility, orthe desire to add water to their PWR cavitiesin order to avoid the possibility of steamexplosions.
The European plant owners, with theencouragement of the regulatory authorities,have developed severe accident
15
management measures. An excellentexample is the containment filtered vent,which has been installed in the SwedishBWRs. Containment vents are beingconsidered for installation in severalEuropean BWRs and PWRs [2]. Sand filtershave been installed in French PWRs, as abackfit. The U.S. plants on the other hand,have not been partial to containment venting.Hard vents are being installed in U.S. BWRswith Mark-I containments, but no U.S. PWRis installing a filtered vent system.Correspondingly, some of the Westinghous-designed plants in Europe (e.g., in U.K.,Spain, Belgium) are not consideringinstallation of vents on their containments [2].
A severe accident management measure ofvery wide acceptance by the PWR plants inEurope and in the U.S. is that of reactorcoolant system (RCS) depressurization in theevent of a severe accident, in order to avoidthe potential for early failure of thecontainment by direct containment heating(DCH). RCS depressurization was includedin the design of the U.K. PWR, primary foraccident prevention, and has been introducedin French PWRs as a backfit following highpressure melt ejection and DCH studies.This severe accident management measure,however, cannot be accomplished on someplants whose safety valves do not havesufficient relief capacity.
Another example of severe accidentmanagement measures is that of cooling thevessel from outside in order to retain the coredebris inside the vessel. With the reactorintact and debris retained in the lower head,phenomena such as ex-vessel steamexplosion and core-concrete interaction,which occur as a result of core debrisrelocation to the reactor cavity, could beprevented. This is the so-called severeaccident management strategy of In-VesselMelt Retention which has been approved forthe Loviisa plant in Finland and has beenincorporated in the design of the AP600 and
AP1000 passive plants. Reactor vesselintegrity is assumed if RCS is depressurizedand the cavity adequately flooded.Cooperative, international researchprograms, RASPLAV and MASCA areproducing results that suggest this approachmay not work for plants with power densitieshigher than that in the Loviisa plant.
Future reactors are expected to have greaterprovision against severe accidents. Theextension of the design to cover severeaccidents, as proposed in Germany andbeing adopted by the French, wouldrepresent a significant departure fromcurrently accepted safety practices in manycountries. Whether such an objectivebecomes a regulatory requirement or not in aparticular country will clearly have a majorimpact on different national approaches tosafety [2].
In Europe, there is now a desire to extend thedesign basis to deal specifically with severeaccidents, but the ways to achieve this havenot been defined. Much of the currentcapability for severe accident mitigationarises from the strength of the containment.However, if (some) severe accidents are tobe included in the design basis, there is acase for the containment to be designed forhigher loads, possibly with a smaller safetymargin. This is an area in which standardshave yet to emerge, although currentdocuments imply that “best estimate” shouldbe sufficient for severe accident assessments[2].
Inclusion of severe accidents in the designbasis poses technical challenges in otherareas, such as steam explosionassessments. There are questions on howconservative should the loading be, orwhether it is possible to show that the “designloading” is always conservative. Currently thisis an area where probabilistic arguments,supported by deterministic analyses, havebeen accepted [2]. Current proposal for new
16
reactor systems have focused on“evolutionary designs”. These designs areessentially modifications to existing LWRs,usually with some advanced safety features.Examples of such safety features are morepassive systems, the use of ex-vesselflooding in AP600 and AP1000, and theprovision of a debris retention device in theEuropean Pressurized Water Reactor (EPR)design and the WWER-1000 reactorscurrently under construction in China.However, there is no general agreement onwhat additional features should be included infuture designs, what the design basis shouldbe, and how the improvements to safetyshould be quantified [2].
3.4 Risk-informed Regulations and Practices
The U.S. NRC has led the development ofthe quantitative risk analysis for nuclearpower plants. Though PRAs have been usedextensively in the past, they were usuallylimited to a variety of applications on a caseby case basis as deemed necessary oruseful. The NRC is now moving toward amuch expanded use of PRAs in what istermed risk-informed regulatory approach. In1995, the NRC adopted a policy thatpromotes increasing the use of probabilisticrisk analysis in all regulatory matters to theextent supported by the state-of-the-art tocomplement the deterministic approach. Thecurrent regulatory framework is based largely,but not entirely on a deterministic approachthat employs safety margins, operatingexperience, accident analyses, andprobabilistic assessment of the risk, andrelies on a defense-in-depth philosophy.
The NRC has applied information gainedfrom PRAs extensively to complement otherengineering analyses in improving issue-specific safety regulation, and in changing thecurrent licensing bases for individual plants.Using risk insights, the NRC has modified itsoversight process and its requirements for
maintenance (10CFR 50.65). The NRC isconsidering further revisions to its reactorregulations (10CFR Part 50) to focusrequirements on programs and activities thatare most risk significant. However, theserevisions would provide alternatives, that arestrictly voluntary, to current requirements.The agency is also considering changes to 10CFR Part 50 that could lead to incorporatinga new set of design-basis accidents, revisingspecific requirements to reflect risk-informedconsiderations, or deleting certainregulations. The main driving force behindthe move toward risk-informing the currentregulations and processes is the expectationthat the use of risk insights can result in bothimproved safety and a reduction inunnecessary regulatory requirements, henceallowing both the NRC and licensees to focusresources on equipment and activities thathave the greatest risk significance.
Within Europe, deterministic safetyassessments are now often complementedby PRAs to verify the overall design andsystem of operation. An example is that ofspecifying as a safety target (goal) the coremelt frequency of 10-5 or 10-6 and theconditional probability of containment failureof 10-1 or 10-2. The use of safety evaluationbased on probabilistic arguments is, so far,confined to resolution of severe accidentsafety issues. This trend is not uniformacross the European Union. For Example, theGerman regulatory and technical supportorganization views of PRA are not asfavorable as those of the comparableSpanish organizations[2].
There is not much support, so far, in Europefor formally considering risk-informedregulations and practices, as it is in the U.S..The exception is the U.K. where the currentNuclear Installations Inspectorate (NII)licensing guidelines adopt a risk-basedapproach. However, most regulatoryauthorities in Europe declare that theyconsider risk information informally [2].
17
The aim of the most European regulatoryauthorities is to improve safety, not just tomaintain it. Therefore, they encourage thedevelopment and the use of PRAs forimproving safety, and not for reducingregulatory requirements. There is aconsiderable reluctance to use the results ofquantitative risk assessment to reduceregulatory requirements regardless of thecalculated risk significance of theserequirements. Uncertainty in the quantitativeresults, concern over the completeness of theanalyses, and lack of properly dealing withorganizational or safety culture issues areusually cited as the bases for this reluctance[17].
3.5 Materials Degradation Issuesand Aging Management
Safe control of aging of nuclear power plantsis an important concern for plant owners andsafety regulatory authorities in the world. Theoptimal ageing management of nuclear powerplants require knowledge on materialsdegradation phenomena and evaluationtechniques.
Contrary to U.S., there is no expiration timefor the operating license in many countries(see Table 7). The periodic safety review(typically every ten years) is the principalmethod applied to reactors to ensure that theplant is adequately safe for a further period ofoperation. However, according to differentcountries, the operating authorization givenby the regulatory authority to the plantoperator is not associated with the sameformal process. Formal aging managementevaluation processes exist in some countries,for quite short periods (i.e., one year in Spain,two in the U.K.); in others, it appears througha requirement of ability for safetydemonstration at any moment (in France andBelgium). In practice, safety agingmanagement is implemented through theperiodic safety review approach, widely
accepted in many countries (see section 3.2).
The material degradation issues have beenthe subject of numerous studies in differentcountries and by several internationalorganizations [4]. These studies have led tothe establishment of various programs orprojects specifically dedicated to themanagement of aging of SSCs.
Aging management begins with plant design.Many design criteria explicitly or implicitlyaddress aging. The ”long-lived” SSCs in anuclear plant, for example, were originallydesigned with sufficient margins to meetminimum lifetime requirements. Current agingmanagement programs aim essentially atmanaging the gradual degradation of SSCsas a result of their physical aging in order toensure permanently satisfying the safetycriteria.
The various aging aspects leading to slowdegradation of SSCs are evaluated duringperiodic safety assessment. However,aspects related to more quick changes ( inparticular those affecting active components)are managed on a continuos basis throughan appropriate maintenance and componentqualification. In the U.S., the original plant life isestablished by the regulatory process. TheAtomic Energy Act and NRC regulations limitthe initial operating licenses of nuclear powerplants to 40 years, but also permit suchlicenses to be renewed. The original 40-yearterm was selected on the basis of economicand antitrust considerations, rather than bytechnical limitations. However, the selectionof this term may have resulted in individualplants being designed on the basis of anexpected 40-year service life. 10 CFR Part54, known as the “license renewal rule,”establishes the technical and proceduralrequirements for renewing operating licenses.Under the license renewal rule, the applicantmust perform a screening review of all SSCs
18
within the scope of the rule to identify“passive” and “long-lived” structures andcomponents. The applicant mustdemonstrate that it will manage the effects of
aging such that the SSCs will function asintended throughout the 20-year period ofextended operation.
Table 7 Operating License Periods in Various Countries
Country License Period Approach
United States Of America Fixed term (40 years, with 20-year renewaloption)
France Lifetime
Japan Lifetime
United Kingdom Lifetime
Republic of Korea Lifetime
Germany Lifetime
Canada Fixed term (2-5 years)
Sweden Lifetime
Spain Variable (5-10 years)Case-by-case, no fixed term but moving to 10-year standard for nuclear facilities thatcomplete periodic safety reviews
Belgium Lifetime
Czech Republic Lifetime
Switzerland Lifetime ( except for 2 plants with term licensesbased on historical technical concerns)
Finland Fixed term (10-20 years)
Hungary Lifetime
Mexico Fixed term (30 years)
Netherlands Lifetime
19
4 SUMMARY AND CONCLUSIONS
Despite considerable similarities in theobjectives and actual implementation ofnuclear safety regulatory approaches, thereare differences in nuclear safety regulatoryrequirements between the United States andother countries.
There is a strong influence of the U.S.regulatory system on setting the basis forlicensing requirements in many countries.This is because a large number of plants inoperation in other countries are of U.S.design or derived from U.S. designs. TheU.S. safety philosophy of defense in depthwas adopted by the regulatory authorities inWestern Europe, Japan, and Korea, not onlyfor the barriers to the release of radioactivesubstances, but also in the design,construction, quality assurance, inspection,and operational practices. However, theremay be differences in the implementation ofthe defense in depth principle, e.g., in levelsof diversity and redundancy required from thesafety systems.
In most countries, the principles of traditionaldeterministic approach have been acceptedover many years to demonstrate the reliabilityand safety of design. Systems, structures,and components are designed andmanufactured to accepted standards,regulation, codes of practice etc. to ensurethat the SSCs can perform their intendedfunctions.
There is an internationally accepted rule thatthe licensee should provide a comprehensivesafety assessment to confirm that the designof an installation fulfils the safety objectivesand requirements. The accident sequencegroups and the accidents to be analyzed inthe safety analysis report may be prescribedby the regulator (e.g., U.S. NRC), but if not,are defined by the licensee as part of hissafety case submission (e.g., U.K.). Theimplementation of either approach is similar.
There are, however, some differences incertain acceptance criteria and the licensingcalculations due to various degree ofconservatism made at each step of thecalculation. Some of these differences werediscussed in this report.
Basic deterministic safety assessments arenow generally complemented by PRAs toverify the overall design and system ofoperation. However, there is not muchsupport, so far, in many other countries forformally considering risk information inregulatory decisionmaking as it is in the U.S..
The desire for protection against severeaccidents is shared by all of the regulatoryauthorities in the Western World. It has alsobeen argued that the severe accident is avery low-probability event; it deserves aresponse, but the cost/benefit should be afactor. This argument has been accepted bythe U.S. NRC and it is a part of the regulatorypractice ( backfit rule, 10CFR50.109). Mostregulatory authorities of the European UnionMember States do not formally accept thisargument.
The Barsebäck-2 incident prompted anumber of corrective actions being taken inBWRs and some PWRs around the world.Actions were taken as direct responses torequirements issued by regulatory authoritiesfor many plants, while for other plants back-fitting measures were introduced voluntarilyor because of anticipated requirements. Theissue of strainer blockage in PWRs havebeen particularly troublesome. Continuingresearch revealing new modes of blockagehas shown that the prompt actions taken bysome European plants may not havecompletely alleviated the problem of strainerblockage. Indeed, redesign may be requiredof these plants. There is a strong evidencethat plant operators throughout the world donot have a definitive solution to the issue.
20
In Europe, there is now a desire to extend thedesign basis to deal specifically with severeaccidents, but the ways to achieve this havenot been agreed. Future reactors areexpected to have greater provision againstsevere accidents. The extension of thedesign to cover severe accidents, asproposed in Germany and being adopted bythe French, would represent a significantdeparture from currently accepted safetypractices in many countries. Whether suchan objective becomes a regulatoryrequirement or not in a particular country willclearly have a major impact on differentnational approaches to safety.
Contrary to U.S., there is no expiration timefor the operating license in many countries.The periodic safety review (typically every tenyears) is the principal method applied toreactors to ensure that the plant is adequatelysafe for a further period of operation.However, according to different countries, theoperating authorization given by theregulatory authority to the plant operator isnot associated with the same formal process.Formal aging management evaluationprocesses exist in some countries, for quietshort periods; in others, it appears through arequirement of ability for safetydemonstration at any moment.
21
5 REFERENCES
1. U.S. Nuclear Regulatory Commission(USNRC), ”Staff RequirementsMemorandum (SRM), April 11, 2003meeting with Advisory Committee onReactor Safeguards (ACRS),” April28, 2003.
2. European Commission, “NuclearSafety and the Environment, 25 Yearsof Community Activities TowardsHarmonization of Nuclear SafetyCriteria and Requirements -Achievements and Prospects,” EUR20055 EN, October 2001.
3. European Commission, “NuclearSafety and the Environment, 30 Yearsof NRWG Activities TowardsHarmonization of Nuclear SafetyCriteria and Requirements,” EUR20818 EN, November 2002.
4. European Commission, “NuclearSafety and the Environment, SafeManagement of NPP Ageing in theEuropean Union,” EUR 19843 EN,May 2001.
5. European Commission, “NuclearSafety and the Environment,European Safety Practices on theApplication of the Leak Before Break(LBB) Concept,” EUR 18549 EN,January 2000.
6. European Commission, “NuclearSafety and the Environment, FuelCladding Failure Criteria,” EUR 19256EN, September 1999.
7. European Commission, “NuclearSafety and the Environment,Determination of the In-containmentSource Term for a Large-Break Lossof Coolant Accident,” EUR 19841 EN,April 2001.
8. OECD/NEA, “Knowledge Base forEmergency Core Cooling SystemRecirculation Reliability,” prepared byUSNRC for the Principal WorkingGroup1 (PWG-1) International TaskGroup, Committee on the Safety ofNuclear Installations, Organization forEconomic Cooperat ion anddevelopment (OECD) Nuclear EnergyAgency (NEA), NEA/CSNI/R (95) 11,February 1996.
9. OECD/NEA, “Knowledge Base forStrainer Clogging-ModificationsPerformed in Different Countriessince 1992,” Committee on the Safetyof Nuclear Installations, Organizationfor Economic Cooperation anddevelopment (OECD) Nuclear EnergyAgency (NEA), Final Report,NEA/CSNI/R(2002)6, October 2002.
10. OECD/NEA, “Nuclear SafetyResearch in OECD Countries, Areasof Agreement, Areas for FurtherAction, increasing Need forCollaboration,” Nuclear EnergyAgency, OECD, 1996.
22
11. OECD/NEA, “International Practicesw i t h Re s p e c t t o L i c e n c eperiods/Terms for Nuclear Facilities inN E A M e m b e r C o u n t r i e s , ”NEA/CNRA/R(2002) 1, September2002.
12. International Atomic Energy Agency(IAEA), National Reports on TheConvention of Nuclear Safety,http://www.iaea.or.at/ns/nusafe/scv_nrpt.htm
13. International Atomic Energy Agency(IAEA), Power Reactor InformationSystem (PRIS)
http://www.iaea.org/programmes/a2/index.html
14. Rao, D.V., C.J. Shaffer, M.T. Leonardand K.W. Ross, “ Knowledge Base forthe Effect of Debris on PressurizedWater Reactor Emergency CoreCooling Sump Performance,” LosAlamos National Laboratory,NUREG/CR-6808, LA-UR-03-0880,February 2003.
15. U.S. Nuclear Regulatory Commission(USNRC), “Containment EmergencySump Performance,” NUREG-0897,Revision 1, 1985.
16. International Atomic Energy Agency,“Periodic Safety Review of NuclearPower Plants,” IAEA SafetyStandards Series, Safety Guide No.NS-G-2.10, 2003.
17. Gupta, O., and J. M. Lanore, “Viewsof the French Regulatory Body onRisk-Informed Approaches and on theUse of PSA,” (Undated).
