study on status of safety margins assessment practices -- vtt-r-06265-15

8/19/2019 Study on Status of Safety Margins Assessment Practices -- VTT-R-06265-15

1/41

RESEARCH REPORT VTT-R-06265-15

FOUND WP1.1 - Status of safety margins assessment practices

Study on Status of Safety Margins

Assessment Practices

Authors: Otso Cronvall

Confidentiality: Public


2/41

1 (40)

Report’s title

Study on Status of Safety Margins Assessment PracticesCustomer, contact person, address Order reference

State Nuclear Waste Management Fund (VYR), VTT Dnro: SAFIR 20/2015Project name Project number/Short nameFOUND 2015 101984/ FOUND 2015

Author(s) Pages

Otso Cronvall 40/-Keywords Report identification code

Safety factor, safety margin, YVL Guide, ASME, NPP piping VTT-R-06265-15Summary

This study concerns collection and review of Finnish safety factor definitions applied to structural integrityanalyses of nuclear power plant (NPP) piping components as well as representative analysis examples showingthe effect and underlying uncertainties concerning the safety factors.

On global level, International Atomic Energy Agency (IAEA) provides information and definitions on safetymargins and safety factors of NPP components. The detailed definitions of the safety factors are a countryspecific issue, mainly as presented by the local regulators. The top level IAEA documents are the safetystandards, which are or will be adopted by all European regulators as the basis for their assessments of NPPs.There exist no specified values for safety margins in safety guides or regulatory demands.

In Finland, the significant documents concerning the safety factors of NPP components are YVL Guides.However, they do not give the actual safety factor data, but instead refer to certain applicable standards. Theseare mainly ASME code Sections III and XI.

Both deterministic and probabilistic crack growth analyses were performed for two representative BWR pipingcomponents. For comparison purposes, the analyses were carried out both with and without safety factors.

Based on the results of this study, in the structural integrity analyses, safety margins are maintained byintroducing various kinds of safety factors and conservative assumptions. This makes it difficult, or evenimpossible, to define what is the total safety margin of a structure or component, e.g. how much more loading itcould bear beyond that caused by plant operation. Structural integrity analyses of structures and componentswould benefit from such more specific definitions of safety factors that would give a more accurate assessment of the total safety. For the time being, establishment of a value that could represent the total safety status (safetymargin) of a structure or component has to be done with probabilistic terms by risk analysis approach.

Confidentiality PublicEspoo 4.1.2016Written by

Otso CronvallSenior Scientist

Reviewed by

Kim CaloniusResearch Scientist

Accepted by

Petri KinnunenDeputy Head of Research Area

VTT’s contact addressP.O. Box 1000, 02044 VTT

Distribution (customer and VTT)SAFIR2018 Reference Group 5,

VTT Archive (2)

The use of the name of the VTT Technical Research Centre of Finland (VTT) in advertising or publication in part of this report is only permissible with written authorisation from the VTT Technical Research Centre of Finland.


3/41


2 (40)

Contents

Foreword ..................................................................................................................... 3

List of abbreviations ..................................................................................................... 4

1 Introduction ............................................................................................................. 5

1.1 General ........................................................................................................... 51.2 On technical background ................................................................................ 5

2 Scope and objectives of the study ........................................................................ 11

3 Review of safety factors concerning degradation analyses of NPP components . 12

3.1 YVL guides on safety factors and margins .................................................... 123.2 YVL guides on analyses requiring application of safety factors .................... 143.3 Relevant standards on analyses applying safety factors .............................. 15

4 NPP piping integrity analyses ............................................................................... 20

4.1 Analysis cases and input data ....................................................................... 204.1.1 Geometry data ................................................................................... 204.1.2 Material property data ........................................................................ 204.1.3 Load and stress data ......................................................................... 204.1.4 Considered degradation mechanism and associated model .............. 23

4.1.5 SCC induced initial crack postulates according to NURBIT code ...... 244.1.6 SCC induced initial crack postulates developed by VTT .................... 244.1.7 Model for the effect of inspections ..................................................... 254.1.8 Summary on input data and list of analysis cases ............................. 26

4.2 Applied analysis procedures ......................................................................... 284.2.1 Deterministic VTTBESIT .................................................................... 294.2.2 Probabilistic VTTBESIT...................................................................... 29

4.3 Deterministic analysis results ........................................................................ 304.4 Probabilistic analysis results ......................................................................... 31

5 Summary and conclusions.................................................................................... 34

References ................................................................................................................ 37


4/41


3 (40)

Foreword

This report has been prepared under the research project FOUND, and therein for WP 1.1Status of safety margins assessment practices. The project is a part of SAFIR2018, which is anational nuclear energy research program. FOUND WP 1.1 work in 2015 was funded by theState Nuclear Waste Management Fund (VYR) and the Technical Research Centre of Finland (VTT), which is gratefully acknowledged. The work was carried out at VTT.

Espoo 4.1.2016

Author


5/41


4 (40)

List of abbreviations

ASME American Society of Mechanical EngineersASTM American Society for Testing and MaterialsBWR Boiling water reactor FE Finite elementFFS Fitness-for-serviceHAZ Heat affected zoneIAEA International Atomic Energy AgencyIWM Fraunhofer-Inst itut für Werkstoffmechanik LWR Light water reactor

NDT Non-destructive testing

NPP Nuclear power plantPOD Probability of detectionSCC Stress corrosion crackingSSM Swedish Radiation Safety Authority (Strålsäkerhetsmyndigheten in Swedish)STUK Radiation and Nuclear Safety Authority of Finland (Säteilyturvakeskus in

Finnish)VTT Technical Research Centre of Finland (Teknologian tutkimuskeskus VTT)VYR State Nuclear Waste Management Fund WRS Weld residual stressYVL Regulatory guides on nuclear safety


6/41


5 (40)

1 Introduction

This study is part of the SAFIR2018 research program project FOUND. The overall objectiveof this project is cross-disciplinary assessment of ageing mechanisms for safe managementand extension of operational lifetime of Finnish nuclear power plants (NPPs). This involvesdeveloping deterministic, probabilistic and risk informed approaches in computational and experimental analyses with education of new experts. FOUND is an acronym of Analysis of Fatigue and Other cUmulative ageing to exteND lifetime.

1.1 General

This study concerns collection and review of Finnish safety factor definitions applied to

structural integrity analyses of NPP piping components as well as representative analysisexamples showing the effect and underlying uncertainties concerning the safety factors.

After this Chapter the structure of this report is as follows. The scope and objectives of thestudy are described in Chapter 2. The review of safety factors concerning degradationanalyses of NPP components is presented in Chapter 3. This is followed with bothdeterministic and probabilistic computational NPP piping analyses in Chapter 4. Thesummary and conclusions are presented in Chapter 5, including also some suggestions for future research.

1.2 On technical backgroundTo provide technical background for safety factors, a short summary of safety margins of NPPcomponents and their relation to safety factors is described in the following. This is carried out in international level, as based on documents published by the International AtomicEnergy Agency (IAEA). The detailed definitions of the safety factors are a country specificissue, mainly as presented by the local regulators.

The IAEA publishes internationally important documents concerning safety of application of nuclear energy. The top level IAEA documents are the safety standards, which are or will beadopted by all European regulators as the basis for their assessments of NPPs. The INSAG

reports are complementary reports but are not adopted by the national regulators and,therefore, are not fully followed in the national safety assessments. IAEA publishes also other important documents, such as TECDOC series reports. These documents are presently notadopted by national regulators.

The following IAEA safety standards have connections with safety assessment and safetymargin assessments:

NS-R-1 Safety of Nuclear Power Plants: Design [1],SSR-2/2 Safety of Nuclear Power Plants: Commissioning and Operation [2],SSG-2 Deterministic Safety Analysis for Nuclear Power Plants - Specific Safety Guide[3],

NS-G-2.14 - Conduct of operations at nuclear power plants [4], NS-G-2.2 - Operational limits and conditions and operating procedures for nuclear power plants [5], NS-G-2-12 - Ageing management for nuclear power plants [6].


7/41


6 (40)

Instead of specifying what safety margins actually mean, the covered IAEA documents useterms limits and levels. In addition, there exist several kinds of margins. Some of theserepresent demanded margins to be used in safety assessment, others represent uncertainties

concerning knowledge of methods/data that are included in the assessment. There exist almostalways margins that are remaining margins for which there exist no demands in the safetyguides or by the regulators. Only in IAEA report NS-R-2 - Safety of Nuclear Power Plants:Operation [7] are safety factors mentioned. Therein it is mentioned in connection to safetyassessment that the strategy for the review and the safety factors to be evaluated shall beapproved or agreed by the regulatory body. However, presently the report SSR-2/2 [2]supersedes report NS-R-2 [7], and in the former report nothing is mentioned concerningsafety factors.

The demands on never passing safety limits have to be fulfilled by the NPPs independent of the age of the plant. Guides and standards specify methods to be used in performing safety

assessments and in ways to determine safety limits. These guides and standards specify thatconservative approach is to be used in such assessments. The conservative approachrepresents different kinds of margins. Used data and calculation methods include certaindegree of uncertainty. The uncertainty part of the margins will be decreased when knowledgeconcerning the issue in question is increased.

There exist no specified values for safety margins in safety guides or regulatory demands.Safety margins consist of margins from several different structures, systems and components.It is not possible to calculate an overall value for safety margins with the commonly used deterministic methodology [8].

In every safety assessment, there exists a remaining margin between the output from thesafety assessment and the safety limit. This remaining margin, which can only be quantified with best estimate analyses, is not part of the safety margin and there are no quantitativerequirements for it [8].

Establishment of a value that could represent the total safety status (safety margin) of a NPPhas to be done by probabilistic terms with risk analysis approach. Distances to different risk limits can give insight in the total safety margins of the plant. This also allows rankingdifferent safety issues against each other.

In performing deterministic safety assessment, several kinds of margins can be used with highconfidence to support that specified limits are not violated. In each assessment, the followingmargins exists:

Margins between safety limit and incorrect function or failed barrier, M1 ; Thismargin is specified in international standards or regulatory demands, by specifying thesafety limits. Deterministic safety assessments have to prove that this margin exists for allinitiating events and barriers by never exceeding the specified safety limits.Margins between conservative and non-conservative safety assessment, M2 ;Performing safety assessment includes that certain methodology is used, as steered byguidance or regulatory demands. The methodology specifies certain conservatism to beused. This includes the application of safety factors. The desired degree of conservatism

affects the choice of analysis codes, procedures and data to be used in the assessment.This margin is represented by the difference in outputs between realistic assessment and conservative assessment. The degree of conservatism depends on the knowledge of the

phenomena in question and on the ability of the analysis procedure to represent a certain


8/41


7 (40)

phenomenon. By research and other developments, the demands on used conservatism can be changed. When more knowledge and analysis procedures are available, it is in mostcases accepted to reduce the conservatism. Some types of conservatism are specified indifferent guidance or regulations and cannot be reduced without change in regulations.

Margins on operating position or effects of ageing on material, M3 ; In performingsafety assessment, the current condition of systems, structures and components needs to be specified. For this purpose, ageing effects to material properties shall be specified in aconservative way, so as to represent a broad group of initiating positions of the events and material conditions. With operating experiences and with modern knowledge on materialageing, the conservatism used for the initiating parameters represent the M3 margin in thesafety assessment. Deterministic safety assessments have no specific demands on the M3margin itself for any initiating events and barriers. It is important that a positive marginexists for all operational modes.Remaining margin, R ; The output value of safety assessments shall be shown not tooverride the specified safety limits for the specific event/case. The differences between the

output value of the assessment and the value of the safety limit represent the remainingmargin. There exist no specific demands on the value of this margin. It can be zero or verylarge. In some cases, the methodology defines how to assure that this margin is largeenough to cover the uncertainty in the assessment. In most cases, the uncertainty isincluded in the output value from the assessment. The change in remaining margin can beevaluated when modifications are performed at the plant. In most cases, a change inremaining margin does not affect the other margins.Uncertainties, U ; The margins M1 to M3 include effects of chosen conservatism and uncertainties. The basic demand in the safety assessment is to assure that the specified limit is not exceeded. Uncertainties exist in all parts of a deterministic safety assessment.Especially the analytical margins of type M2 include a certain degree of uncertainty, e.g.

based on how well the used thermo-hydraulic code can represent a certain scenario. Thisuncertainty is part of the margins but has to be included only as long as the uncertaintyexists based on lack of knowledge, code validations or data. It is important that the degreeof uncertainties is quantified or at least positioned towards used conservatism in theassessment. The uncertainties shall be part of the value for the margin and may never exceed the value of a specific margin.

Safety margins in component level or structural integrity level

Each NPP component has to fulfil demands specified in the final safety analysis report(FSAR) and in some cases specified in national or international norms/standards. On thecomponent level, these demands can vary in different event classes. Safety cases are evaluated with different load combinations. Concerning piping components, different load combinationswill, as based on the frequencies of occurrence, be evaluated against different acceptancecriteria, i.e. safety limits. To have different acceptance criteria in different event classes isalso valid for other components. For instance, US NPP code ASME prescribes different stresslimits for different event classes.

The safety limits are developed in such a way that demands on system level and plant levelwill be fulfilled. Concerning NPP piping, the criteria are set to certify that the loss-of-coolant-

accident (LOCA) frequencies support the event frequencies on these different events specified in the FSAR. For certain assessments of components, specific demands are specified for dataconservatism (M3 safety margins) and on methodologies (M2 safety margins) to be used for


9/41


8 (40)

the evaluations. For instance, the ASME code identifies specific data and methodologies to beused in the assessment of stresses on piping or maximum pressure in structural systems.

In these assessments for each component, there are safety margins in:

The conservatism in specifying the acceptance criteria for the different components ineach event class (M1 safety margins).The conservatism in data and methodologies used in the assessment (M2 and M3 safetymargins).The conservatism of the selection of material (including specified material properties),which is part of the M3 safety margin value.

The existing margins in categories M1, M2 and M3 can be changed by changing the method for specifying the acceptance criteria or by altering the definition of the conservatism in used assessment methodologies.

As the definitions of conservatism and acceptance criteria differ between different eventclasses, modifications that change the event classes will also affect the margins associated with the acceptance criteria. With ageing of components, e.g. by corrosion and fatigue, themargins associated with the acceptance criteria will decrease, depending on to what extent theassessment has quantified the ageing effects in the initial assessment in the M3 margins.These R margins between the actual capacity of the system and the acceptance criteria in theFSAR are those that the utilities are allowed to change, reduce and/or increase. These marginsare not specified in any FSAR and, therefore, are of type R, besides the margins described above. These remaining margins have been described in the IAEA report IAEA-TECDOC-1332 [9] as the safety margin. However, the remaining margins are described as “licensingmargin” in the IAEA report IAEA-TECDOC-1418 [10]. Thus, the definitions by the IAEAconcerning the R margins are at least to some extent ambiguous.

Uncertainties U exist in all margins. The M2 margins include uncertainties in the used methodologies and their validation levels. The safety factors belong mainly to the M2 safetymargins. As for the M3 margins, they include uncertainties in the initial material properties.

The remaining margins R can be changed by ageing, plant modifications and changes inoperating procedures.

The overall integrity margin or component margin is therefore the following sum: M1+ M2+M3.

An overall presentation of the various types of NPP safety margins is shown in Figure 1.2-1.An illustration of the NPP design margins is shown in Figure 1.2-2. An illustration of the NPPsafety margins associated with the acceptance criteria is shown in Figure 1.2-3.


10/41


9 (40)

Figure 1.2-1. The various types of NPP safety margins [10].

Figure 1.2-2. Illustration of NPP design margins [10].


11/41


10 (40)

Figure 1.2-3. Illustration of the NPP safety margins associated with the acceptance criteria[10] .


12/41


11 (40)

2 Scope and ob jec tives o f the study

The scope and objectives of the study are described briefly in the following.

The scope of this study covers:review of safety factors concerning analyses of NPP degradation mechanisms,deterministic NPP piping integrity analyses,

probabilistic NPP piping integrity analyses,summary and conclusions on the studied issues.

The covered documents concerning the safety factors of NPP components are:Finnish YVL guides by the authority STUK (Radiation and Nuclear Safety Authority of Finland, Säteilyturvakeskus in Finnish),

relevant international standards.

The performed analyses concern:a small representative set of Finnish boiling water reactor (BWR) piping components,

both deterministic and probabilistic approaches, considering the uncertainties caused bythe used input data to analysis results.

The objectives of this study are to:review domestic safety factors concerning degradation analyses of NPP pipingcomponents,through representative computational examples for NPP piping components show and assess the effect of uncertainties in input data to degradation analysis results.


13/41


12 (40)

3 Review of safety factors concerning degradation analysesof NPP components

A review of safety factors concerning degradation analyses of NPP piping components is presented in the following. The scope is limited to domestic safety factor definitions, mainlygiven by STUK in the YVL Guides. The acronym YVL corresponds to Finnish words that astranslated to English stand for: regulatory guides on nuclear safety. However, in a number of cases the YVL Guides refer to international standards, so those are mentioned too with

presenting the relevant definitions therein. The issues considered in this chapter are:YVL guides on safety factors and safety margins,YVL guides on analyses requiring application of safety factors,relevant international standards on analyses applying safety factors.

The YVL Guides and most relevant international standards covered in this study are:GUIDE YVL A.8, Ageing management of a nuclear facility [11],GUIDE YVL B.3, Deterministic safety analyses for a nuclear power plant [12],GUIDE YVL E.3, Pressure vessels and piping of a nuclear facility [13],GUIDE YVL E.4, Strength analyses of nuclear power plant pressure equipment [14],GUIDE YVL E.5, In-service inspection of nuclear facility pressure equipment with non-destructive testing methods [15],ASME Section III, Rules for Construction of Nuclear Facility Components [16],ASME Section XI, Rules for In-service Inspection of Nuclear Power Plant Components[17],

IAEA Safety Standard NS-G-2.6, Maintenance, Surveillance and In-Service Inspection in Nuclear Power Plants [18],WENRA Reactor Safety Reference Levels, Issue K [19],Standard ASTM E 1921-11, [20],Standard ASTM E 1820-11 [21],Standard Review Plan 3.6.3, Rev.1, Leak-Before-Break Evaluation Procedures [22].

YVL Guides that contain no information on the safety factors and safety margins, but docontain classifications and other important data are:

GUIDE YVL A.6, Conduct of operations at a nuclear power plant [23],GUIDE YVL A.7, Probabilistic risk assessment and risk management of a nuclear power

plant [24],GUIDE YVL B.2, Classification of systems, structures and components of a nuclear facility [25],GUIDE YVL B.5, Reactor coolant circuit of a nuclear power plant [26].

3.1 YVL guides on safety factors and margins

A review of safety factors and margins for NPP piping components in YVL Guides is presented in the following. The considered YVL Guides are already identified in the

introductory beginning of this chapter.


14/41


13 (40)

YVL guides on safety factors

For NPP piping components, the main YVL Guide on design and other structural analysesinvolving safety factors is YVL E.4 [14]. However, also certain other YVL Guides deal with

the safety factors.

According to chapter “Definitions” of GUIDE YVL E.4 [14]: “Load grouping shall refer toapplying the acceptance limits presented in the standard applied to stress analysis, graded bythe severity of the load and the safety factors required, to the specified service loads so thatthe frequency of occurrence of the load, the post-load opportunities for inspection and repair,and the integrity and operability requirements set for the pressure equipment in the scenario inquestion are taken into consideration.”

According to chapter “6.3 Qualification input information” of GUIDE YVL E.5 [15]: “Thetarget defect sizes that shall be detected and defined correctly during in-service inspections

shall be defined for each inspection area. If the items qualified have been grouped intoqualification groups, the defects shall be defined for each wall thickness according to the moststressed component. Of these, the defects most difficult to detect because of their size shall bechosen as target defects. Target defect size shall be primarily based on defects allowed duringthe nuclear facility’s operation by the standard applied in the design of the component or structure in question. A calculation method that complies with ASME Code, Section XI [17]Subarticle IWB-3600 shall be used to calculate pressure equipment crack growth during theinspection interval or during its remaining service life. The standards list the service loadingsafety factors in question.”

YVL guides on safety margins

For NPP piping components, the main YVL Guide on design and other structural analysesinvolving safety margins is YVL E.4 [14]. However, also certain other YVL Guides deal withthe safety margins.

According to chapter “1 Introduction” of GUIDE YVL A.8 [11], chapter “1 Introduction” of GUIDE YVL E.3 [13] and chapter “1 Introduction” of GUIDE YVL E.4 [14]: “The design,construction, operation, condition monitoring and maintenance of a nuclear power plant shall

provide for the ageing of systems, structures and components (SSCs) important to safety inorder to ensure that they meet the design-basis requirements with the necessary safety marginsthroughout the service life of the facility. Systematic procedures shall be in place for

preventing the ageing of systems, structures and components which may deteriorate their availability, and for the early detection of the need for their repair, modification and replacement. Safety requirements and applicability of new technology shall be periodicallyassessed, in order to ensure that the technology applied is up to date, and the availability of the spare parts and the system support shall be monitored.”

According to chapter “9.3 Ageing management follow-up report” of GUIDE YVL A.8 [11]:“The follow-up report shall provide on the SSCs covered by the ageing management regimean assessment of operability and ageing-induced changes to the safety margins.”

According to chapter “1 Introduction” of GUIDE YVL B.3 [12] and chapter “1 Introduction”of GUIDE YVL E.4 [14]: “The analytical methods employed to demonstrate compliance withsafety requirements shall be reliable and well qualified for the purpose. The analyses shall


15/41


14 (40)

demonstrate the conformity with the safety requirements with high certainty. Any uncertaintyin the results shall be assessed and considered in determining safety margins.”

According to chapter “1 Introduction” of GUIDE YVL E.3 [13]: “The systems, structures and

components that implement or are related with safety functions shall be designed,manufactured, installed and used so that their quality level, and the assessments, inspectionsand tests, including environmental qualification, required to verify their quality level, aresufficient considering the safety significance of the item in question.”

According to chapter “3 Strength analysis report” of GUIDE YVL E.4 [14]: “A construction plan submitted on pressure equipment modifications shall include strength analyses revised asnecessary if the item of pressure equipment’s design pressure or temperature, configuration,wall thicknesses, material values, operations, supporting or other factors relating to the itemof pressure equipment change in a way that may compromise the safety margins to beattained.”

3.2 YVL guides on analyses requiring application of safety factors

For NPP piping components, the main YVL Guide on design and other analyses requiringapplication of safety factors is YVL E.4 [14]. Other YVL Guides do not contain anythingsignificant on that issue.

In general, it is stated in chapter “5 Stress analysis” of GUIDE YVL E.4 [14] that: “A stressanalysis in accordance with the present Guide shall be conducted on pressure equipment or their parts which are to be constructed in accordance with the highest safety and quality

requirements applicable to a nuclear power plant. Items of special importance include SafetyClass 1 primary circuit components including main circulating pumps”.

According to chapter “5.4 Acceptance limits” of GUIDE YVL E.4 [14]: “Stress analysis shalldemonstrate the fulfilment of the service limits set by the applicable standard for the item of

pressure equipment in question or its part in accordance with its classification, service load grouping, number of pressure tests and the analysis methods used. The states of stresscalculated by stress analysis shall, therefore, be classified into stress types which aresignificant for the maintenance of integrity and operability and relative to which each servicelimit is set. This procedure shall be repeated for all analysed structural parts and loads of theitem of pressure equipment.”

According to chapter “6.3 Qualification input information” of GUIDE YVL E.5 [15]: “Acalculation method that complies with ASME Code, Section XI [17] Subarticle IWB-3600shall be used to calculate pressure equipment crack growth during the inspection interval or during its remaining service life. The standards list the service loading safety factors inquestion.”

According to chapter “2 Scope of application” of GUIDE YVL E.4 [14]: “Loadings comprisethe mechanical and thermal stress factors arising in the nuclear power plant’s normaloperation, and design basis and design extension operational conditions, against which the

pressure equipment design shall present adequate safety margins, demonstrated by strengthanalyses. The strength analyses include stress analysis, brittle fracture analysis and analysesdemonstrating the implementation of the leak before break principle (LBB). To be analysed are also loadings when their values are to be derived by calculations from the operation of systems, behaviour of buildings or loading conditions exerted on pressure equipment by local


16/41


15 (40)

phenomena. The LBB principle is presented as a procedure for excluding pipe break induced impact loads from the design bases.”

According to chapter “6.6 Operational occurrences and accidents” of GUIDE YVL E.4 [14]:

“The acceptance criterion shall be in accordance with the safety margin required in theapplicable standard as regards the growth of a postulated crack. An analysis of the safetymargin provided by crack arrest may be considered as an additional justification on a case-by-case basis.”

According to chapter “7.7 Analysis methods” of GUIDE YVL E.4 [14]: “LBB analysis shalldemonstrate by analysis of fracture and fluid mechanics the safety margins required inreference [22] as regards leak detection and critical defect size.”

According to chapter “7.7 Analysis methods” of GUIDE YVL E.4 [14]: “Determination of critical defect size shall be based on crack growth resistance values representing the materials

and welding procedures used in manufacturing, which are determined by testing inaccordance with an applicable standard [21] at a temperature corresponding to normaloperation. The crack growth resistance values must be valid for the stable crack extension,

based on which critical defect size and the safety margin for it are determined. In usingextrapolation, the method applied and its qualification data shall be presented.”

3.3 Relevant standards on analyses applying safety factors

The relevant standards containing safety factors applicable to NPP piping components areconsidered in the following. The YVL Guides do not give the actual safety factor data, but

instead refer to these standards. These are ASME code Sections III [16] and XI [17].

In chapter “5.2 Applicable standards” of GUIDE YVL E.4 [14] it is stated that: “Inconducting a stress analysis of pressure equipment and pumps in accordance with the presentGuide, the standard ASME Boiler and Pressure Vessel Code Section III, Division 1 [16] shallapply as a general rule. The mandatory regulations stated in its Articles NB 3200 and NB3650 as well as in its Subsections NF and NG apply in these cases in so far as STUK has not

presented more detailed requirements.”

After that the YVL E.4 [14] moves on to state that another design and strength analysisstandard for Safety Class 1 pressure equipment can be applied too, on the condition that it has

been applied previously in the construction of NPPs of the same type.

To provide technical background associated with the application of the safety factors, thedefinitions of the safety classes of the NPP components, service level loads and necessarystress components are summarised first. This description is limited to metallic components.

According to YVL B.2 [25] the NPP components divide into safety classes, corresponding tocriteria ensuring structural resistance, integrity and leak-tightness, as follows:

Safety Class 1 shall include nuclear fuel as well as structures and components whosefailure could result in an accident compromising reactor integrity and requiring immediate

actuation of safety functions. Safety Class 1 includes the reactor pressure vessel (RPV)and those components of the primary circuit whose failure results in a primary circuit leak that cannot be compensated for by systems relating to normal plant operation.


17/41


16 (40)

Safety Class 2 components include main components and piping of the emergency corecooling system, structures of the core support and reactor shutdown system, primarycircuit piping supports and brackets, fuel storage racks, certain less risk significant small-diameter piping and components, as well as components which can be isolated from the

reactor coolant system.Safety Class 3 includes mainly structures and components relating to barriers to thedispersion of radioactive substances or structures relating to the handling of radioactivematerials not assigned to higher safety classes.

The four service level loads A, B, C and D are defined in Appendix H of ASME Section XI[17] as follows:

Level A service loadings are any loadings arising from system start-up, operation in thedesign power range, hot stand-by, and system shutdown, and excepting only thoseloadings covered by other service level loads.Level B service loadings correspond to incidents of moderate frequency. These loadingsare deviations from Level A service loadings that are anticipated to occur often enoughthat design capability should withstand them without operational impairment. The eventscausing Level B service loadings include transients resulting from any single operator error or control malfunction, transients caused by a fault in a system component requiringits isolation from the system, and transients due to loss of load or power. These eventsinclude any abnormal incidents not resulting in a forced outage and also forced outagesfor which the corrective action does not include any repair of mechanical damage.Level C service loadings correspond to infrequent incidents. These are deviations fromLevel A service loadings and require shutdown for correction of the loadings or repair of damage in the system. The conditions have a low probability of occurrence, but areincluded to provide assurance that no gross loss of structural integrity will result as aconcomitant effect of any damage developed in the system. The total number of

postulated occurrences for such events may not exceed 25.Level D service loadings correspond to limiting faults. These are combinations of loadings associated with extremely low probability, postulated events whoseconsequences are such that the integrity and operability of the nuclear energy system may

be impaired to the extent that only considerations of public health and safety are involved.

For NPP piping components, the necessary stress components are defined in Article NB-3000of ASME Section III [16] as follows:

General primary membrane stress, P m, is any normal stress or a shear stress developed

by an imposed loading which is necessary to satisfy the laws of equilibrium of externaland internal forces and moments. The basic characteristic of a primary stress is that it isnot self-limiting. A thermal stress is not classified as a primary stress. A general primarymembrane stress is distributed so in the structure that no redistribution of load occurs as aresult of yielding. An example of primary stress is general membrane stress due to internal

pressure or to distributed live loads.Local primary membrane stress, P L, corresponds to membrane stress produced by

pressure or other mechanical loading and associated with a discontinuity that would causeexcessive distortion in the transfer of load to other portions of the structure. When the sizeof the more severely stressed region remains within certain limits, it can be considered local, with the limiting stress value being membrane stress intensity of 1.1 times thedesign stress, S m.Primary bending stress, P b, is the variable component of normal stress, which in turn isthe stress component normal to the plane of reference. The variation may or may not be


18/41


17 (40)

linear across the thickness. The loads causing the primary bending stress are the same asthose causing the general primary membrane stress.Expansion stress, P e, results from restraint of free end displacement of the piping system.Secondary stress, Q, is a normal stress or a shear stress developed by the constraint of

adjacent material or by self-constraint of the structure. The basic characteristic of asecondary stress is that it is self-limiting. Local yielding and minor distortions can satisfythe conditions which cause the stress to occur and failure from one application of thestress is not to be expected. Examples of secondary stresses are general thermal stress, and

bending stress at a gross structural discontinuity.Peak stress, F , is that increment of stress which is additive to the primary plus secondarystresses by reason of local discontinuities or local thermal stress, including the effects of stress concentrations. The basic characteristic of a peak stress is that it does not cause anynoticeable distortion and is objectionable only as a possible source of a fatigue crack or a

brittle fracture. Examples of peak stresses are thermal stress in the austenitic steelcladding of a carbon steel component, the stress at a local structural discontinuity, and surface stresses produced by a thermal shock.

The use of safety factors applicable to NPP piping components are often associated with localcrack like flaws and their allowable sizes. The maximum sizes of such NPP piping flaws thatneed not be analysed further are defined in Subarticle IWB-3514 of ASME Section XI [17], inSubarticle IWB-3514.2 for ferritic piping and in Subarticle IWB-3514.3 for austenitic piping.For surface flaws detected by in-service inspection (ISI), the maximum depth of these flawsvaries between 5.5 and 14.8 % of wall thickness. Thus, NPP piping flaws deeper than thatneed to be analysed.

According to Subarticle IWB-3640 of ASME Section XI [17], NPP piping containing flawsexceeding the acceptance standards of IWB-3514.1 may be evaluated by analytical

procedures to determine acceptability for continued service to the next inspection or to theend of the evaluation period. A pipe containing flaws is acceptable for continued service for aspecified evaluation time period if the criteria of Subarticles IWB-3642, IWB-3643, or IWA-3644 are satisfied.

This corresponds to the statement given in GUIDE YVL E.5 [15], that “A calculation method that complies with ASME Code Section XI [17] Subarticle IWB-3600 shall be used tocalculate pressure equipment crack growth during the inspection interval or during itsremaining service life.”

The above mentioned Subarticles IWB-3642, IWB-3643 and IWA-3644, in turn, refer toAppendix C or H of ASME Section XI [17] on determining the maximum allowable flawsizes.

According to Appendix C of ASME Section XI [17], the flaws in a NPP piping componentare evaluated by comparing the maximum flaw dimensions at the end of the evaluation period with the maximum allowable flaw size, or by comparing the applied pipe stress with themaximum allowable stress for the flaw size at the end of the evaluation period. The followingmaximum allowable flaw size and stress component definitions are used:

a allow is maximum allowable flaw depth,

lallow is maximum allowable flaw length,S c is maximum allowable bending stress for circumferentially flawed pipe,S a is maximum allowable hoop membrane stress for an axially flawed pipe,S t is maximum allowable membrane stress for a circumferentially flawed pipe.


19/41


18 (40)

The evaluation for maximum allowable flaw size or maximum allowable stress requires theapplication of the safety factors. According to Appendix C of ASME Section XI [17], thesafety factors are applied individually to membrane and bending stresses as SF m and SF b,

respectively. The safety factors depend on the service level and flaw orientation. Theconsidered loading conditions are those associated with Service Levels A, B, C, and D, for the piping system design. Test conditions are evaluated as Service Level B.

According to Appendix C of ASME Section XI [17], the maximum allowable stress for theflawed pipe is a function of pipe stresses, the required safety factors, pipe material properties,end-of-evaluation-period flaw length and depth, flaw orientation, and pipe failure mode.

The safety factor values for primary membrane and primary bending stresses whencalculating the maximum allowable depths for circumferential and axial flaws for pipingcomponents are as presented in Tables 3.3-1 and 3.3-2, as according to Appendix C of ASME

Section XI [17].

Table 3.3-1. For circumferential flaws in Class 1, 2, and 3 piping, the safety factors to beapplied on primary membrane and primary bending stresses in calculating maximumallowable flaw depth [17].

Service Level Membrane Stress, SF m Bending Stress, SF bA 2.7 2.3B 2.4 2.0C 1.8 1.6D 1.3 1.4

Table 3.3-2. For axial flaws in Class 1, 2, and 3 piping, the safety factors to be applied on primary membrane stress in calculating maximum allowable flaw depth [17].

Service Level Membrane Stress, SF mA 2.7B 2.4C 1.8D 1.3

To start a structural integrity analysis of NPP piping components with safety factors onmaximum allowable flaw size or/and maximum allowable stress, the sequence used todetermine the failure mode and analysis method for NPP piping components are needed first.These are determined in Article C-4000 of ref. [17]. Therein it is stated that for flaws inwrought base metal, non-flux welds, weld metal, or cast product in which ferrite content isless than 20 %, plastic collapse is the controlling failure mode. For flaws in flux welds of wrought pipe, elastic-plastic analysis methods shall be applied.

The next step is to use the procedures described in Articles C-5000, C-6000 or C-7000 of ref.[17], as applicable to the failure mode, to determine the maximum allowable flaw depth,a allow , or the maximum allowable applied stress, S c or S a , and the maximum allowable flaw

length limit, lallow :Article C-5000 gives a limit load criteria based procedure,Article C-6000 gives an elastic-plastic fracture mechanics (EPFM) criteria based

procedure, and


20/41


19 (40)

Article C-7000 gives a linear-elastic fracture mechanics (LEFM) criteria based procedure.

Article C-5000 of ref. [17] contains:for circumferential flaws tabulated solutions with safety factors for maximum allowable

flaw depth, and analytical solutions for maximum allowable applied stress,for axial flaws both tabulated and analytical solutions with safety factors for maximumallowable flaw depth.

Article C-6000 of ref. [17] contains:for circumferential flaws tabulated solutions with safety factors for maximum allowableflaw depth, and analytical solutions for maximum allowable applied stress,for axial flaws both tabulated and analytical solutions with safety factors for maximumallowable flaw depth.

Article C-7000 of ref. [17] contains for both circumferential and axial flaws analyticalsolutions with safety factors for maximum allowable mode I stress intensity factor, K I , whichis to be compared to corresponding fracture toughness, K Ic . Actually, the fracture toughness of the material is determined by two properties K Ia and K Ic , which represent critical values of thestress intensity factor K I . Whereas K Ia is based on the lower bound of crack arrest critical K I values measured as a function of temperature. K Ic is based on the lower bound of staticinitiation critical K I values measured as a function of temperature.

Concerning conducting a stress analysis of pressure equipment and pumps, GUIDE YVL E.4[14] refers to ASME Section III, Division 1 [16], and therein in particular to Subarticles NB-3200 and NB-3650. These both are parts of Article NB-3000, which concerns Class 1components. Actually, Article NB-3000 refers to safety margins or safety factors only in onesubarticle. This is in NB-3222.4 Analysis for Cyclic Operation, where it is stated that: “Thefatigue curves are obtained from uniaxial strain cycling data in which the imposed strainshave been multiplied by the elastic modulus and a design margin has been provided so as tomake the calculated stress intensity amplitude and the allowable stress intensity amplitudedirectly comparable. Where necessary, the curves have been adjusted to include the maximumeffects of mean stress, which is the condition where the stress fluctuates about a mean valuethat is different from zero.” Thus, ASME Section III [16] does not refer to safety margins butdesign margins. This corresponds to the purpose of ASME Section III [16], which concernsthe design of NPP components. Whereas it is ASME Section XI [17] that concerns theoperation of the NPP components. According to NUREG/CR-6909 report [27], the safety

factor values of 2 on stress and 20 on load cycles have been used in defining the fatiguecurves. These safety factors intend to cover the effects of variables that can influence fatiguelife in air and light water reactor (LWR) environments, and they are broadly classified into thefollowing three groups:

material variability and data scatter,loading sequence,environment.


21/41


20 (40)

4 NPP pi pi ng integrit y anal yses

The scope, conduct and results of the performed deterministic and probabilistic NPP pipingcomponent degradation potential analyses are described in the following. These analyses werefracture mechanics based crack growth computations, and they were performed for two NPP

piping components. The analyses consider the impact on the results caused by loads with and without safety factors, component dimensions, initial flaws as well as inspections.

4.1 Analysis cases and input data

In this study, two representative BWR piping cross-sections are considered. These are takenfrom the analyses presented in an earlier VTT study, see ref. [28].

4.1.1 Geometry data

The main dimensions of the considered BWR piping cross-sections are [28]:BWR-1; outer diameter = 168 mm, wall thickness = 14 mm,BWR-2; outer diameter = 273 mm, wall thickness = 22 mm.

4.1.2 Material property data

The considered material is austenitic stainless steel SA-376 TP304. Relevant material property data of this steel is presented in Table 4.1-1.

Table 4.1-1. Relevant material property data of austenitic stainless steel SA-376 TP304 [29]. Linear interpolation of the values between the four temperatures.

Temperature[ C]

Young’s modulus[GPa]

Yield strength[MPa]

Tensile strength[MPa]

20 195 207 517100 189 170 485200 183 144 442286 177 131 437

Temperature[ C]

Coefficient of thermalexpansion [10 -6 /K]

Thermal conductivity[W/mK]

Specific heat[J/kgK]

20 15.3 14.8 483100 16.1 16.2 511200 17.0 17.9 537286 17.5 19.1 550

4.1.3 Load and stress data

The pressure and temperature in the computations are 7.0 MPa and 286 °C, corresponding tonormal operational conditions in Finnish BWR units [30]. The commonly used assumptionthat yearly time in operation is 8000 hours is also applied here, this corresponds to 10.9months. The necessary stress parameters for the BWR piping components as caused by theoperational BWR loading while taking into account the associated boundary conditions are


22/41


21 (40)

presented in Table 4.1-2, where Pm is applied membrane stress, P b is applied primary global bending load/stress, and P e is applied secondary global bending load/stress. The operationalBWR loading corresponds to ASME Section XI [17] Level A service loadings. Table 4.1-3

presents the membrane and bending stress data of Table 4.1-2 multiplied with Level A safety

factor values, for which data see Table 3.3-1.

Table 4.1-2. The necessary stress components for BWR piping components as caused by theoperational BWR loads.

Pipe component P m[MPa]

P b + Pe[MPa]

BWR-1 17.5 56.3BWR-2 18.2 29.3

Table 4.1-3. The necessary stress components for BWR piping components as multiplied with ASME Section XI [17] Level A safety factor values.

Pipe component P m[MPa]

P b + Pe[MPa]

BWR-1 47.3 129.4BWR-2 49.2 67.4

As for weld residual stresses (WRSs), two commonly used recommendations were selected for the present analyses, which reflect the variation these distributions have as compared toeach other. For further comparison, also the case without WRSs is included in the analyses.The WRS recommendations used in the analyses are:

as-welded state WRSs from the R6 Method, Revision 4 [31],as-welded state WRSs from the SSM handbook [32].

In the analyses, only axial (perpendicular to weld) WRSs are needed. For the two considered piping cross-sections, Figures 4.1-1 to 4.1-4 show in axial direction both the as-welded stateWRSs and total stresses through wall, as corresponding to operational BWR conditions.Figures 4.1-1 and 4.1-2 show these stress components without safety factors, i.e. as such, and Figures 4.1-3 and 4.1-4 show them as multiplied with ASME Section XI [17] Service level Asafety factor values. The WRSs are depicted with dashed curves and the total stresses withcontinuous curves, respectively. The coordinate system in these figures is such that the origin

is in the inner surface of the pipe wall, whereas in the legends p+T+acronym corresponds tototal stresses and acronyms only to the considered WRSs, respectively, and No WRSs tostresses caused only by process loads.


23/41


22 (40)

Figure 4.1-1. Axial as-welded state WRSs and total stresses through wall as such for pipingcomponent BWR-1.

Figure 4.1-2. Axial as-welded state WRSs and total stresses through wall as such for pipingcomponent BWR-2.


24/41


23 (40)

Figure 4.1-3. Axial as-welded state WRSs and total stresses through wall taking into account Service level A safety factor values for piping component BWR-1.

Figure 4.1-4. Axial as-welded state WRSs and total stresses through wall taking into account Service level A safety factor values for piping component BWR-2.

4.1.4 Considered degradation mechanism and associated modelThe considered degradation mechanism is stress corrosion cracking (SCC). The loadingconcerning SCC is stationary operational conditions, including WRSs when the analysed


25/41


24 (40)

crack postulate is located in weld or heat affected zone (HAZ) adjacent to it. SCC wasselected as the considered degradation mechanism because austenitic stainless steels, such asSA-376 TP304 here, are susceptible to SCC. Piping welds and HAZs are locations with hightensional stresses due to WRSs, and BWR environment provides corroding conditions. The

fracture mechanics based crack growth equation used in the analyses depicting intermediate(stage 2) SCC is [33, 34]:

n I CK dt

da(4.1-1)

where a [mm] is crack depth, t [year] is time and K I [MPa m] is the mode I stress intensityfactor (SIF-I), whereas C and n are experimentally determined material, environment and temperature specific constants. Applicable best estimate values for these constants are

presented in Table 4.1-3.

Table 4.1-3. Best estimate values for constants C and n used in BWR SCC analyses for austenitic stainless steel in normal water chemistry [35]. The dimensions are: [C] =(mm/year)/(MPa m) and n is dimensionless.

C n2.710E-07 4.96

4.1.5 SCC induced initial crack postulates according to NURBIT code

The depth and length of SCC induced initial circumferential cracks according to analysis code NURBIT [36, 37] are presented in the following. The depth of the initiated cracks is taken to be 1.0 mm. The probability density function for initial crack length, f al(l0), was estimated froma total of 98 intergranular SCC cases in Swedish stainless steel girth welds in straight pipes,as collected from nine BWR units. The probability density function for the initial crack lengthis:

010 212exp2 l R H e Rl

R f i

iial

(4.1-2)

where l0 [mm] is initial crack length, Ri [mm] is inner radius of pipe cross-section and H is theHeaviside step function. The last two factors in equation (4.1-2) are due to the truncation. The

parameter was chosen with 0 equal to 9.380 so that the mean values of the observed and fitted distributions coincided. This corresponds to a mean value for 1/ 0 of 10.66 % of theinner pipe circumference. The nucleation frequency of these initial cracks is assessed as4.08E-04 per year per weld.

4.1.6 SCC induced initial crack postulates developed by VTT

The assessment of depth and length of SCC induced initial circumferential crack postulatesdeveloped by VTT [38] is presented in the following. This treatment is based on the sameflaw data as was used for the assessment of the corresponding initial cracks included in the

NURBIT code.


26/41


25 (40)

A recursive method based on fracture mechanics and statistical curve fitting was used toassess the probabilistic distributions for depth and length of cracks initiating due to SCCduring plant operation. The first step in the applied approach is to convert the size data

concerning detected grown SCC induced cracks to dimensionless form in relation to pipe wallthickness and inner circumference. Then, with recursive fracture mechanics based analyses,the thus obtained data is matched with the assumed initial size criteria, corresponding here torespective mode I stress intensity factor threshold values, K I,threshold , for SCC initiated cracks.Finally, the obtained data is converted to probabilistic form and suitable reliabilitydistribution functions are fitted to them.

The fitted linear probabilistic density function for estimated initial depths, f (a 0), of SCCinduced circumferential cracks is:

0125.000004.0 %00 aa f (4.1-3)

where the unit of the initial crack depth a 0% is %, and in terms of actual dimensions a 0 islimited to region from 0.2 mm to 0.76 mm. For instance, the actual depth of an initial crack with depth of 50 % is: a 0 = (50/100)×(0.76-0.2)+0.2 = 0.48 mm. This being the median depthfor SCC induced initial circumferential crack postulates developed by VTT.

The fitted linear probabilistic density function for estimated initial lengths, f (l0), of SCCinduced circumferential cracks is:

00 103.0exp103.0 ll f (4.1-4)

where the unit of the initial crack length l0 is %, in relation to the inner circumference of pipecomponent.

4.1.7 Model for the effect of inspections

A quantitative measure of inspection effectiveness is needed in order to calculate thereduction in failure probability associated with inspection. The probability for finding a crack like flaw in the inspections is typically presented in the form of probability of detection(POD) functions, which describe this probability as a function of the flaw size, i.e. flaw depthor length. Worldwide several material and degradation mechanism specific POD functionsapplicable to NPP components have been published. POD functions according to report

NUREG/CR-3869 [39] have been selected for the present analyses. These POD functionshave the following form:

wallwall t a B Bt a lnPOD 21 (4.1-5)

where a [mm] is crack depth, t wall [mm] is wall thickness, B1 and B2 are given parameter values, see ref. [39], whereas (*) is normalised Gaussian distribution function. The

NUREG/CR-3869 [39] defines the POD values as a function in relation to crack depth

through wall separately for ferritic steels and stainless steels and for inspection quality of three levels, those being Poor, Good and Advanced. Concerning POD for stainless steels, it islimited to consider IGSCC. For the POD value distributions in stainless steel piping, see


27/41


26 (40)

Figure 4.1-3 below. Note that the presented POD model and data correspond to ultrasonictesting procedure with near side access to weld.

Figure 4.1-6. Distributions of POD function values concerning intergranular SCC for stainless steel piping, according to NUREG/CR-3869 [39].

4.1.8 Summary on input data and list of analysis cases

A summary the input data to be used in the computational analyses is presented in thefollowing, see also Table 4.1-4. The list of the considered analysis cases is presented in Table4.1-5.

The scope and used input data of the computational analyses cover:two BWR pipe components, for their geometry data see Table 4.1-4 below,

base material is austenitic stainless steel SA-376 TP304, for material property data seeTable 4.1-1,SCC as the considered degradation mechanism, with the considered flaw postulate being acircumferentially oriented semi-elliptic crack opening to inner surface and located in HAZadjacent to weld,for parameter values used in the SCC growth rate equation, see Table 4.1-3,for deterministic analysis cases, the initial crack depth, a 0, is 1.0 mm, and initial crack length, l0, is 6.0 mm, respect ively,for probabilistic analysis cases, two sets of probability density distributions for sizes of SCC induced initial cracks, see Chapters 4.1.5 and 4.1.6, and Table 4.1-4 below,

operational BWR conditions as the considered process loads, see Chapter 4.1-3, theassumed yearly time in operation is 8000 hours,two distributions for the WRSs, see Chapter 3 and Table 4.1-1, additionally included isthe case of no WRSs, i.e. only operational process loading induced stresses,


28/41


27 (40)

effect of inspections to pipe component failure probabilities is taken into account withPOD functions, see Chapter 4.1.7,inspection intervals of 3 and 10 years, for quality of inspections see Table 4.1-4 below,additionally included is the case of no inspections.

As the assumed yearly time in operation is 8000 hours, corresponding to approximately 11months, it realistically leaves for the yearly maintenance outage, for other possible timesunder shut-down, and for anticipated/typical yearly load transients approximately one month.These load events are not taken into account in the computational analyses, as SCC concernsonly stationary operational conditions. No piping component replacements or changes in the

process water chemistry were considered, thus it was assumed that SCC is in effect throughthe whole of the assumed time in operation.

Table 4.1-4. Input data used in computational degradation potential analyses.Pipe size Outer diameter [mm] Wall thickness [mm] ReferenceBWR-1 168 14 [28]BWR-2 273 22 [28]Initial crack sizes by Cause for crack initiation Median crack depth [mm] Reference

NURBIT distribution SCC 1.0 [36, 37]VTT distribution SCC 0.5 [38]WRSsrecommendations

Maximum value [MPa] Minimum value [MPa] Reference

R6 Method, Rev. 4 310 120 [31]SSM handbook 270 -140 [32]POD from Scope NDT quality options Reference

NUREG/CR-3869 intergranular SCC, austeniticstainless steels and ferriticsteels

advanced, good, poor, of which good is used here

[39]

CCDP value No. of degradation states Time in operation [year] Reference0.00001 10 or 8 (*) 60 [38]

(*) depends on wall thickness.

Table 4.1-5a. List of deterministic analysis cases.Case no. Cross-section WRSs according to Stress components

D1 BWR-1 R6 Method, Rev. 4 as suchD2 BWR-1 R6 Method, Rev. 4 with safety factors

D3 BWR-1 SSM handbook as suchD4 BWR-1 SSM handbook with safety factors

D5 BWR-1 no WRSs as such

D6 BWR-1 no WRSs with safety factorsD7 BWR-2 R6 Method, Rev. 4 as such

D8 BWR-2 R6 Method, Rev. 4 with safety factors

D9 BWR-2 SSM handbook as such

D10 BWR-2 SSM handbook with safety factors

D11 BWR-2 no WRSs as suchD12 BWR-2 no WRSs with safety factors


29/41


28 (40)

Table 4.1-5b. List of probabilistic analysis cases, stress loads as such.Case no. Cross-section Initial crack sizes WRSs Inspection intervalP1 BWR-1 NURBIT R6 Method, Rev. 4 3 yearsP2 BWR-1 NURBIT R6 Method, Rev. 4 10 years

P3 BWR-1 NURBIT R6 Method, Rev. 4 No inspectionsP4 BWR-1 NURBIT SSM handbook 3 yearsP5 BWR-1 NURBIT SSM handbook 10 yearsP6 BWR-1 NURBIT SSM handbook No inspectionsP7 BWR-1 NURBIT no WRSs 3 yearsP8 BWR-1 NURBIT no WRSs 10 yearsP9 BWR-1 NURBIT no WRSs No inspectionsP10 BWR-2 VTT R6 Method, Rev. 4 3 yearsP11 BWR-2 VTT R6 Method, Rev. 4 10 yearsP12 BWR-2 VTT R6 Method, Rev. 4 No inspectionsP13 BWR-2 VTT SSM handbook 3 years

P14 BWR-2 VTT SSM handbook 10 yearsP15 BWR-2 VTT SSM handbook No inspectionsP16 BWR-2 VTT no WRSs 3 yearsP17 BWR-2 VTT no WRSs 10 yearsP18 BWR-2 VTT no WRSs No inspections

Table 4.1-5c. List of probabilistic analysis cases, stress loads multiplied with Service level Asafety factor values.Case no. Cross-section Initial crack sizes WRSs Inspection interval

P19 BWR-1 NURBIT R6 Method, Rev. 4 3 yearsP20 BWR-1 NURBIT R6 Method, Rev. 4 10 yearsP21 BWR-1 NURBIT R6 Method, Rev. 4 No inspectionsP22 BWR-1 NURBIT SSM handbook 3 yearsP23 BWR-1 NURBIT SSM handbook 10 yearsP24 BWR-1 NURBIT SSM handbook No inspectionsP25 BWR-2 VTT R6 Method, Rev. 4 3 yearsP26 BWR-2 VTT R6 Method, Rev. 4 10 yearsP27 BWR-2 VTT R6 Method, Rev. 4 No inspectionsP28 BWR-2 VTT SSM handbook 3 yearsP29 BWR-2 VTT SSM handbook 10 years

P30 BWR-2 VTT SSM handbook No inspections

4.2 Applied analysis procedures

The fracture mechanics based analysis code VTTBESIT was used in all crack growthanalyses. This code comprises parts developed by VTT [40, 41] and by IWM [42, 43, 44], thelatter being the Fraunhofer-Institut für Werkstoffmechanik, in Germany.

VTTBESIT computes the SIF-I values over the crack postulate fronts. The analysis codetreats only the mode I loading, in which the direction of the loading is perpendicular to thecrack surface (crack opening mode). These computations are carried out with BESIT60 codemodule by IWM. This module is based on the weight/influence function method. Solutions


30/41


29 (40)

are provided for "infinite" and semi-elliptical surface crack postulates in straight plates and hollow cylinders [42, 43, 44].

VTTBESIT uses the BESIT60 solely to computation of SIF-I values, and applies them as part

of the necessary input data for crack growth computations as well as for the determining theend-of-life crack sizes [40].

VTTBESIT calculates the crack growth as a series of increments, until some pre-defined ending criterion is met, or until the crack tip reaches the opposite surface of the componentwall. Two crack growth models are provided in the analysis code: Paris-Erdogan equation for fatigue induced crack growth, and rate equation for SCC. In addition to deterministic crack growth computations, another version of VTTBESIT exists for probabilistic crack growthanalyses. These two code versions are described briefly in the following.

4.2.1 Deterministic VTTBESITThe analysis flow of the deterministic VTTBESIT is as follows [40, 41]:1. Reading of the deterministic input data.2. Crack growth analysis; the magnitude of crack growth in each time step is calculated from

the respective crack growth equation the ending criterion of the analysis is that crack depth reaches some pre-determined or the opposite pipe surface.

3. Writing the time or load cycle dependent SIF-I and crack growth results to a separate file.

4.2.2 Probabilistic VTTBESIT

The analysis flow of the probabilistic VTTBESIT is as follows [38]:1. Reading of the deterministic input data.2. Random picking of certain input data parameters from the specified distributions; 1) SCC;

probability distributions for initial crack depth and length, 2) fatigue induced crack growth; probability distributions for initial crack depth, length, and frequency of load cycles.

3. Crack growth analysis; the magnitude of crack growth in each time step is calculated fromthe respective crack growth equation the ending criterion of the analysis is that crack depth reaches the opposite pipe surface.

4. For each analysed circumferential piping weld, at least 5000 simulations with Latin

hypercube simulation (LHS) procedure are computed.5. The degradation state to which the crack has grown is computed for each year of the

assumed time in operation and for each simulation these results are used in the ensuing probabilistic Markov process based degradation potential and risk analyses.

6. Writing the time or load cycle dependent crack growth simulation results to a separate file.

The applied discrete time Markov procedure for degradation potential and risk analyses issummarised as follows [45]:1. Construction of degradation matrix transition probabilities from VTTBESIT simulation

results and database analysis of crack initiation frequencies.2. Model for inspection quality, as based on applicable POD functions, which are in turn

used to construct inspection matrix transition probabilities.3. Markov model to calculate pipe leak/break probabilities and risks for chosen inspection

programs, including the case of no inspections.


31/41


30 (40)

4. Writing the time or load cycle dependent leak/break probability and risk results to aseparate file.

4.3 Deterministic analysis resultsThe results from the deterministic VTTBESIT crack growth analyses are described in thefollowing. For analysis cases D1 to D6 the results are presented in Figure 4.2-1, and those for analysis cases D7 to D12 in Figure 4.2-2, respectively. The covered time span is 60 years,corresponding to realistic extended time in operation. In Figures 4.2-1 and 4.2-2, the resultsfor cases with stresses as such are with continuous lines and those for cases with safetyfactored stresses are with dashed lines, respectively.

Figure 4.2-1. Deterministic VTTBESIT crack growth analysis results for cases D1 to D6.

Figure 4.2-2. Deterministic VTTBESIT crack growth analysis results for cases D7 to D12.


32/41


31 (40)

As can be seen from Figures 4.2-1 and 4.2-2, the use of safety factors clearly has an effect tothe crack growth analysis results. Each case with odd number, e.g. Case D1, is to be compared with the following case with even number, e.g. Case D2, the former ones being with stress

loads as such and the latter ones multiplied with safety factor values. For the analysed casesthe increase in tensile stress due to using safety factors varies approximately between 60 to100 MPa. For the analysis cases with WRSs, the cracks grow through wall approximately 20years sooner for cases with safety factors as compared to those without. On the other hand, for the analysis cases without WRSs, the cracks grow very slowly, and during 60 years themaximum growth is less than 3 mm. These latter results reveal the significant role of WRSs inSCC growth analyses. This corresponds to the experienced SCC behaviour, as the tensilestresses must be relatively high before SCC becomes significant. The effect of wall thicknessto the analysis results is such for the cases with WRSs that in the cases D7 to D10 with 8 mmthicker wall than in cases D1 to D4, the time for cracks to grow through wall is approximately10 years longer.

4.4 Probabilistic analysis results

The pipe break probability results computed with the probabilistic VTTBESIT are described in the following. For analysis cases P1 to P18 the results are presented in Figure 4.2-3, and those for analysis cases P19 to P30 in Figure 4.2-4, respectively. The former set of results isfor cases with stress loads as such, whereas the latter set is for cases with safety factored stresses. Again, the covered time span is 60 years. The results in Figures 4.2-3 and 4.2-4 are

presented after 1, 5, 20 and 60 years of operation, the corresponding notations in the legendsare 1y, 5y, 20y and 60y, respectively, while the case names present also the inspection

strategy with inspections every 3 or 10 years as insp int 3y or 10 y, and no inspections as noinsp.

Figure 4.2-3a. VTTBESIT pipe break probability results for cases P1 to P9.


33/41


32 (40)

Figure 4.2-3b. VTTBESIT pipe break probability results for cases P10 to P18.

Figure 4.2-4a. VTTBESIT pipe break probability results for cases P19 to P24.


34/41


33 (40)

Figure 4.2-4b. VTTBESIT pipe break probability results for cases P25 to P30.

As can be seen from Figures 4.2-3 and 4.2-4, the use of safety factors has an effect to the break probability analysis results, though it is much less visible than for the deterministiccrack growth analysis results. For the analysis cases with WRSs, the pipe break probabilities

are less than half a decade higher than for cases with safety factors as compared to thosewithout. As for the effect of WRSs, the pipe break probabilities are approximately 10 decadeshigher with them than for those without. Again these latter results reveal the significant roleof WRSs in SCC growth analyses. The effect of wall thickness to the analysis results is muchless visible than for deterministic crack growth analysis results. As for the initial crack sizes,the break probabilities results corresponding VTT crack distributions are slightly smaller thanthose for cases with NURBIT crack distributions. However, this effect soon evens out,already after 5 years in operation. For the covered analysis cases the effect of inspections ishardly visible. This is outcome is not according to expectations, but is explained by the used applied loading. Namely, the used as-welded state WRSs, as taken accordingrecommendations from the R6 Method, Revision 4 [31] and the SSM handbook [32], are quite

high, often of the scale of material yield stress or higher. Thus, it is recommendable to usemore accurate WRS data in the analyses, obtained with finite element (FE) simulations and/or from experiments.


35/41


34 (40)

5 Summary and conclusions

This study concerns collection and review of Finnish safety factor definitions applied tostructural integrity analyses of NPP piping components as well as representative analysisexamples showing the effect and underlying uncertainties concerning the safety factors.

On international level, IAEA provides information and definitions on safety margins and safety factors of NPP components. The detailed definitions of the safety factors are a countryspecific issue, mainly as presented by the local regulators. The top level IAEA documents arethe safety standards, which are or will be adopted by all European regulators as the basis for their assessments of NPPs. The INSAG reports are complementary reports but are not adopted

by the national regulators and, therefore, are not fully followed in the national safetyassessments. IAEA publishes also other important documents, such as TECDOC series

reports. These documents are presently not adopted by national regulators.

Instead of specifying what safety margins actually mean, the IAEA documents use termslimits and levels. In addition, there exist several kinds of margins. Some of these representdemanded margins to be used in safety assessment, others represent uncertainties concerningknowledge of methods/data that are included in the assessment. There exist almost alwaysmargins that are remaining margins for which there exist no demands from the safety guidesor the regulators.

The demands on never passing safety limits have to be fulfilled by the NPPs independent of the age of the plant. Guides and standards specify methods to be used in performing safety

assessments and in ways to determine safety limits. These guides and standards specifyconservatism to be used in such assessments. The conservatism represents different kinds of margins. Used data and calculation methods include certain degree of uncertainty.

There exist no specified values for safety margins in safety guides or regulatory demands.Safety margins consist of margins from several different structures, systems and components.It is not possible to calculate an overall value for safety margins with the commonly used deterministic methodology [8]. Establishment of a value that could represent the total safetystatus (safety margin) of a NPP has to be done with probabilistic terms by risk analysisapproach.

In performing deterministic safety assessment, the following margins exist:Margins between safety limit and incorrect function or failed barrier, M1,Margins between conservative and non-conservative safety assessment, M2,Margins on operating position or effects of ageing on material, M3,Remaining margin, R,Uncertainties, U.

In component level or structural integrity level assessments there are safety margins in:The conservatism in specifying the acceptance criteria for the different components ineach event class (M1 safety margins).

The conservatism in data and methodologies used in the assessment (M2 and M3 safetymargins).The conservatism of the selection of material (incl. specified material properties), which is

part of the M3 safety margin value.


36/41


35 (40)

The overall integrity margin or component margin is therefore the following sum: M1+ M2+M3.

In Finland, the significant documents concerning the safety factors of NPP components are:

Finnish YVL guides by the authority STUK (Radiation and Nuclear Safety Authority of Finland, Säteilyturvakeskus in Finnish),relevant international standards.

The YVL Guides that contain information on safety factors and safety margins are:GUIDE YVL A.8, Ageing management of a nuclear facility [11],GUIDE YVL B.3, Deterministic safety analyses for a nuclear power plant [12],GUIDE YVL E.3, Pressure vessels and piping of a nuclear facility [13],GUIDE YVL E.4, Strength analyses of nuclear power plant pressure equipment [14],GUIDE YVL E.5, In-service inspection of nuclear facility pressure equipment with non-destructive testing methods [15],

However, the YVL Guides do not give the actual safety factor data, but instead refer to certainapplicable standards. These are ASME code Sections III [16] and XI [17]. Some maindefinitions for safety factors given in these refs. are summarised in the following.

According to Appendix C of ref. [17], the evaluation for maximum allowable flaw size or maximum allowable stress requires the application of the safety factors. These are to beapplied individually to membrane and bending stresses as SF m and SF b, respectively. Thesafety factors depend on service level and flaw orientation. The considered loading conditionsare those associated with Service Levels A, B, C, and D, for the piping system design. Test

conditions are evaluated as Service Level B.According to Appendix C of ref. [17], the maximum allowable stress for the flawed pipe is afunction of pipe stresses, the required safety factors, pipe material properties, end-of-evaluation-period flaw length and depth, flaw orientation, and pipe failure mode.

The safety factor values for primary membrane and primary bending stresses whencalculating the maximum allowable depths for circumferential and axial flaws for pipingcomponents are as presented in Tables 3.3-1 and 3.3-2, as taken from Appendix C of ref. [17].

Article C-7000 of ref. [17] contains for both circumferential and axial flaws analytical

solutions with safety factors for maximum allowable mode I stress intensity factor, K I , whichis to be compared to corresponding fracture toughness, K Ic . Actually, the fracture toughness of the material is determined by two properties K Ia and K Ic , which represent critical values of thestress intensity factor K I . Whereas K Ia is based on the lower bound of crack arrest critical K I values measured as a function of temperature. K Ic is based on the lower bound of staticinitiation critical K I values measured as a function of temperature.

Fatigue end-of-life curves are applied for design of NPP components against cyclic loading.According to Article NB-3000 of ref. [16] the fatigue curves are obtained from uniaxial straincycling data in which the imposed strains have been multiplied by the elastic modulus and adesign margin has been provided so as to make the calculated stress intensity amplitude and

the allowable stress intensity amplitude directly comparable. The safety factor values of 2 onstress and 20 on load cycles have been used in defining the fatigue curves. These safetyfactors intend to cover the effects of variables that can influence fatigue life in air and LWR environments.


37/41


36 (40)

As for performed NPP piping integrity analyses, two representative BWR piping cross-sections were considered. Their material is austenitic stainless steel SA-376 TP304. Theapplied pressure and temperature are 7.0 MPa and 286 °C, corresponding to normaloperational conditions in Finnish BWR units [30]. For comparison purposes, the process load

induced stresses were applied in the analyses both with and without safety factors. The WRSsused in the analyses are the recommendations in the R6 Method, Revision 4 [31] and the SSMhandbook [32]. The considered degradation mechanism is SCC, while the considered flaw

postulate is a circumferentially oriented crack opening to inner surface. Both deterministicand probabilistic crack growth analyses were performed, using VTTBESIT code [40, 41, 42,43, 44] developed by VTT and IWM as well as Markov application [45] developed by VTT.

According to the deterministic crack growth analysis results, the use of safety factors clearlyhas an effect to the crack growth analysis results. For the analysed cases the increase in tensilestress due to using safety factors varies approximately between 60 to 100 MPa. For theanalysis cases with WRSs, the cracks grow through wall approximately 20 years sooner for

cases with safety factors as compared to those without. On the other hand, for the analysiscases without WRSs, the cracks grow very slowly, and during 60 years the maximum growthis less than 3 mm. These latter results reveal the significant role of WRSs in SCC growthanalyses. This corresponds to the experienced SCC behaviour, as the tensile stresses must berelatively high before SCC becomes significant. The effect of wall thickness to the analysisresults is such for the cases with WRSs that in the cases D7 to D10 with 8 mm thicker wallthan in cases D1 to D4, the time for cracks to grow through wall is approximately 10 yearslonger.

According to the pipe break probability results, the use of safety factors has an effect to the break probability analysis results, though it is much less visible than for the deterministicanalysis results. For the analysis cases with WRSs, the pipe break probabilities are less thanhalf a decade higher than for cases with safety factors as compared to those without. As for the effect of WRSs, the pipe break probabilities are approximately 10 decades higher withthem than for those without. Again these latter results reveal the significant role of WRSs inSCC growth analyses. The effect of wall thickness to the analysis results is much less visiblethan for deterministic crack growth analysis results. For the covered analysis cases the effectof inspections is hardly visible. This is outcome is not according to expectations, but isexplained by the used applied loading. Namely, the used as-welded state WRSrecommendations are quite high, often of the scale of material yield stress or higher. Thus, itis recommendable to use more accurate WRS data in the analyses, obtained with finiteelement (FE) simulations and/or from experiments.

Based on the results of this study, in the structural integrity or fitness-for-service (FFS)analyses, safety margins are maintained by introducing various kinds of safety factors and conservative assumptions. This makes it difficult, or even impossible, to define what is thetotal safety margin of a structure or component, e.g. how much more loading it could bear

beyond that caused by plant operation, including experienced and anticipated transientloading events. Further, possible new degradation mechanisms and unexpected local loadingconditions challenge the structural safety margins in ways that are difficult to predict.Structural integrity analyses of structures and components would benefit from such morespecific definitions of safety factors that would give a more accurate assessment of the total

safety. This is s

study on status of safety margins assessment practices -- vtt-r-06265-15

Documents